All ER (EC) 1998

[1998] All ER (EC) 1

Fantask A/S and others v Industriministeriet (Erhvervsministeriet)

(Case C-188/95)
EUROPEAN COMMUNITY; Taxation, Other European Community
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, RAGNEMALM AND WATHELET
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, KAPTEYN, MURRAY, EDWARD,
PUISSOCHET (RAPPORTEUR), HIRSCH, JANN AND SEVN_ADVOCATE GENERAL JACOBS
29 APRIL, 26 JUNE, 2 DECEMBER 1997
European Community Companies Indirect taxes on raising capital Danish rules imposing charge on
registration of new companies and increases in capital Variable element of charge calculated as
proportion of nominal value of capital raised Whether registration charge compatible with Community
law Whether Community provisions having direct effect Council Directive (EEC) 69/335, arts 10, 12.
European Community State liability Infringement of Community law attributable to state Charges on
raising or increasing capital levied unlawfully over long period of time without Danish authorities or
persons charged being aware of unlawfulness National law allowing dismissal of claims for recovery of
charges in such circumstances Whether compatible with Community law.

The plaintiff companies were subject to charges levied by the Danish authorities on the registration of new
public and private limited companies and on increases in their capital. The amounts charged consisted of
a fixed basic charge and a supplementary charge calculated in proportion to the nominal value of the
capital raised. When the Danish Court of Auditors found the charging system to be unlawful, the
supplementary charge was abolished and the plaintiffs requested a refund of the charges they had paid
between 1983 and 1992. The requests were rejected and the plaintiffs commenced proceedings against
the relevant authorities, contending inter alia that the supplementary charge was contrary to arts 10 1 and
122 of Council Directive (EEC) 69/335 concerning indirect taxes on the raising of capital, which
respectively prohibited the charging of any taxes in respect of registration or any other formality required
in order to set up a business and laid down a list of exempt taxes and duties which included duties paid
by way of fees or dues. The Danish court stayed the proceedings and referred to the Court of Justice of
the European Communities questions regarding the compatibility of the charges at issue with the
directive, the application of national rules on state liability and limitation periods to actions for the recovery
of charges imposed in contravention of Community law and whether arts 10 and 12 had direct effect.
1
Articles 10, so far as material, is set out at p 5 j, post
2
Article 12, so far as material, is set out at p 5 j, post

Held (1) For the purposes of art 12 of Directive 69/335, fees or dues charged on the registration of
public and private limited companies and on increases in their capital could not increase indefinitely in
proportion to the nominal value of the transaction in question but had to be calculated solely on the basis
of the costs of the formalities involved. Those costs could also include the proportion of overheads which
could be attributed to the effecting of registration and the costs of minor services performed without
charge. Moreover, member states could impose flat rate charges of a fixed amount for an indefinite period
provided that they checked at regular intervals that the charges did not exceed the average costs incurred
(see p 33 d to j and p 37 c d, post); Ponente Carni SpA v Amministrazione delle Finanze dello Stato
Joined cases C-71/91 and C-178/91 [1993] ECR I-1915 applied.
(2) The application of a general principle of national law under which the courts of a member state
should dismiss claims for the recovery of charges levied over a long period in breach of Community law
without either the authorities of that state or the persons liable to pay the charges having been aware that
they were unlawful would, in the circumstances, have made it excessively difficult to obtain recovery of
charges which were contrary to Community law. It would, moreover, have the effect of encouraging
infringements of Community law which had been committed over a long period. It followed that charges
levied in breach of the directive could not be justified on the ground that they had been imposed by a
member state over a long period without either that state or the persons liable to them having been aware
that they were unlawful (see p 34 h j and p 37 e, post); SCS Peterbroeck Van Campenhout & Cie v
Belgium Case C-312/93 [1996] All ER (EC) 242 applied.
(3) Community law did not prevent a member state which had not properly transposed a directive from
resisting an action for the repayment of charges levied in breach thereof by relying on a limitation period
under national law which ran from the date on which the charges in question became payable, provided
that such period was not less favourable for actions based on Community law than for actions based on
national law and did not render virtually impossible or excessively difficult the exercise of rights conferred
by Community law (see p 36 e and p 37 f, post); Comet BV v Produktschap voor Siergewassen Case
45/76 [1976] ECR 2043 and Johnson v Chief Adjudication Officer (No 2) Case 410/92 [1995] All ER (EC)
258 applied; Emmott v Minister for Social Welfare Case C-208/90 [1991] ECR I-4269 distinguished.
 (4) The prohibition laid down in art 10 of the directive and the derogation from that prohibition in art
12(1)(e) were expressed in sufficiently precise and unconditional terms to be invoked by individuals in
their national courts in order to contest a provision of national law which infringed the directive (see p 36 j
and p 37 g, post); Comitato di Coordinamento per la Difesa della Cava v Regione Lombardia Case C-
236/92 [1994] ECR I-483 applied.

Notes
For an introduction to Community provisions on free movement of capital, see 52 Halsburys Laws (4th
edn) paras 17011702.
For private enforcement of Community law, see 51 Halsburys Laws (4th edn) paras 234235, 370.

Cases cited
Amministrazione delle Finanze dello Stato v Denkavit Italiana Srl Case 61/79 [1980] ECR 1205.
Amministrazione delle Finanze dello Stato v SpA San Giorgio Case 199/82 [1983] ECR 3595.
Argos Distributors Ltd v Customs and Excise Comrs Case C-288/94 [1996] ECR I-5311.
Bautiaa Srl v Directeur des Services Fiscaux, Socit Franaise Maritime SA v Directeur des Services
Fiscaux Joined cases C-197/94 and C-252/94 [1996] ECR I-505.
BP Supergas Anonimos Etaira Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece Case C-62/93
[1995] All ER (EC) 684, [1995] ECR I-1883, ECJ.
Brasserie du Pcheur SA v Germany, R v Secretary of State for Transport, ex p Factortame Ltd Joined
cases C-46/93 and C-48/93 [1996] All ER (EC) 301, [1996] ECR I-1029, ECJ.
Comateb (Socit) v Directeur Gnral des Douanes et Droits Indirects Joined cases C-192218/95
[1997] ECR I-165.
Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043.
Comitato di Coordinamento per la Difesa della Cava v Regione Lombardia Case C-236/92 [1994] ECR I-
483.
Conradsen (P) A/S (Advokatrdet as representative of) v Ministeriet for Skatter og Afgifter Case 161/78
[1979] ECR 2221.
Corbeau (Criminal proceedings against) Case C-320/91 [1993] ECR I-2533.
Denkavit Internationaal BV v Kamer van Koophandel en Fabrieken voor Midden-Gelderland Case C-2/94
[1996] ECR I-2827.
Elida Gibbs Ltd v Customs and Excise Comrs Case C-317/94 [1997] All ER (EC) 53, [1996] ECR I-5339,
ECJ.
Emmott v Minister for Social Welfare Case C-208/90 [1991] ECR I-4269.
Felicitas Rickmers-Linie KG & Co v Finanzamt fr Verkehrsteuern, Hamburg Case 270/81 [1982] ECR
2771.
FMC plc v Intervention Board for Agricultural Produce Case C-212/94 [1996] ECR I-389.
Francovich v Italy Joined cases C-6/90 and C-9/90 [1991] ECR I-5357.
Haahr Petroleum Ltd v benr Havn Case C-90/94 (1997) ECJ Transcript, 17 July.
Johnson v Chief Adjudication Officer (No 2) Case 410/92 [1995] All ER (EC) 258, [1994] ECR I-5483.
Marshall v Southampton and South West Hampshire Area Health Authority (No 2) Case C-271/91 [1993]
4 All ER 586, [1994] QB 126, [1993] 3 WLR 1054, [1993] ECR I-4367, ECJ.
Marshall v Southampton and South West Hampshire Health Authority (Teaching) Case 152/84 [1986] 2 All
ER 584, [1986] QB 401, [1986] 2 WLR 780, [1986] ECR 723, ECJ.
McDermott v Minister for Social Welfare Case 286/85 [1987] ECR 1453.
Ministeriet for Skatter og Afgifter v Investeringforeningen Dansk Sparinvest Case 36/86 [1988] ECR 409.
Palmisani v Istituto Nazionale della Previdenza Sociale (INPS) Case C-261/95 (1997) ECJ Transcript, 10
July 1997.
Peterbroeck Van Campenhout & Cie (SCS) v Belgium Case C-312/93 [1996] All ER (EC) 242, [1995]
ECR I-4599, ECJ.
Ponente Carni SpA v Amministrazione delle Finanze dello Stato Joined cases C-71/91 and C-178/91
[1993] ECR I-1915.
R v IRC, ex p Unilever plc [1996] STC 681, CA.
3
R v Secretary of State for Social Security, ex p Sutton Case C-66/95 [1997] All ER (EC) 497, ECJ.
Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989.
Steenhorst-Neerings v Bestuur van de Bedrijfsvereniging voor Detailhandel, Ambachten en Huisvrouwen
Case C-338/91 [1993] ECR I-5475.
Texaco A/S v Middelfart Havn, Olieselskabet Danmark amba v Trafikministeriet Joined cases C-114
115/95 (1997) ECJ Transcript, 7 July 1997.
Van Schijndel v Stichting Pensioenfonds voor Fysiotherapeuten Joined cases C-430431/93 [1996] All
ER (EC) 259, [1995] ECR I-4705, ECJ.
Waldrich Siegen Werkzeugmaschinen GmbH v Finanzamt Hagen Case C-38/88 [1990] ECR I-1447.

Reference
By order of 8 June 1995, the stre Landsret (the Eastern Regional Court) referred to the Court of Justice
of the European Communities for a preliminary ruling under art 177 of the EC Treaty eight questions (set
out at p 29 b to p 30 d, post) on the interpretation of Council Directive (EEC) 69/335 concerning indirect
taxes on the raising of capital, as amended. Those questions were raised in actions brought by Fantask
A/S and a number of other companies or groups of companies against the Industriministeriet
(Erhvervsministeriet) (the Danish Ministry of Trade and Industry) relating to charges levied on registration
of new public and private limited companies and on the capital of such companies being increased.
Written observations were submitted on behalf of: Fantask A/S, by T Rrdam, of the Copenhagen Bar;
Norsk Hydro Danmark A/S, Tryg Forsikring skadesforsikringsselskab A/S and Tryg Forsikring
livsforsikringsselskab A/S, by K Michelsen, C Heg Madsen and H Aasmul-Olsen, of the Copenhagen
Bar; Aalborg Portland A/S, by K Dyekjr-Hansen, of the Copenhagen Bar; Forsikrings-Aktieselskabet
Alka, Robert Bosch A/S, Uponor A/S, Uponor Holding A/S and Pen-Sam ApS and others, by V Thorup, H
Stenbjerre, J Boe and L Normann Jrgensen, from the firm Kromann and Mnter, of the Copenhagen
Bar; the Danish government, by P Biering, Head of Division in the Ministry of Foreign Affairs, acting as
agent, assisted by K Hagel-Srensen, of the Copenhagen Bar; the French government, by C de Salins,
Assistant Director in the Legal Affairs Directorate of the Ministry of Foreign Affairs, and F Pascal,
Administrative Attach in the same directorate, acting as agents; the Swedish government, by E
Brattgrd, Adviser in the Trade Department of the Ministry of Foreign Affairs, acting as agent; the UK
government, by J E Collins, Assistant Treasury Solicitor, acting as agent, assisted by E Sharpston,
Barrister; and the European Commission, by A C Jessen and E Traversa, of its Legal Service, acting as
agents, assisted by S Helsteen and J Rostock-Jensen, from the firm Reumert & Partnere, of the
Copenhagen Bar. Oral observations were submitted by: Fantask, represented by P Jker Thorsen, of the
Copenhagen Bar; Norsk Hydro Danmark, Tryg Forsikring skadesforsikringsselskab and Tryg Forsikring
livsforsikrings- selskab, represented by H Aasmul-Olsen; Aalborg Portland, represented by L Hennenberg,
of the Copenhagen Bar; Forsikrings-Aktieselskabet Alka, Robert Bosch, Uponor, Uponor Holding and
Pen-Sam and others, represented by H Peytz, of the Copenhagen Bar; the Industriministeriet
(Erhvervsministeriet), represented by K Hagel-Srensen; the Danish government, represented by P
Biering; the French government, represented by G Mignot, Foreign Affairs Secretary in the Legal Affairs
Directorate of the Ministry of Foreign Affairs, acting
4
as agent; the Italian government, represented by D Del Gaizo, Avvocato dello Stato; the UK government,
represented by J E Collins, assisted by E Sharpston; and the Commission, represented by A C Jessen
and E Traversa, assisted by J Rostock-Jensen and H Henrik Skjdt, of the Copenhagen Bar. The
language of the case was Danish. The facts are set out in the opinion of the Advocate General.

26 June 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
The stre Landsret (the Danish Eastern Regional Court) has requested the Court of Justice of the
European Communities to give a preliminary ruling on the interpretation of Council Directive (EEC) 69/335
concerning indirect taxes on the raising of capital, as amended (OJ S Edn 1969 (II) p 412). The directive,
which introduces a harmonised duty on the raising of capital by companies, prohibits the charging of any
other taxes in respect of company registration of capital companies. By virtue of art 12 of the directive
member states nevertheless retain the power to charge duties paid by way of fees or dues. The stre
Landsret seeks guidance on the scope of that expression and also asks the Court of Justice whether, in
the light of the ruling in Emmott v Minister for Social Welfare Case C-208/90 [1991] ECR I-4269, a
member state can rely upon a limitation period in proceedings brought against it even though it has failed
to implement a directive properly.

RELEVANT COMMUNITY RULES

2.
Directive 69/335 has the aim of promoting the free movement of capital by harmonising the taxation
payable on the contribution of capital to companies and firms and by abolishing stamp duty on securities
as well as other indirect taxes with the same characteristics as capital duty or stamp duty. Article 3 of the
directive specifies the companies and firms in respect of which capital duty is payable, which are referred
to in the directive as capital companies, and art 4 specifies the transactions which may attract the duty.
By virtue of art 7 of the directive, which was most recently amended by art 1(2) of Council Directive
85/303 (OJ 1985 L156 p 23), member states must either exempt such transactions from capital duty or
charge duty at a single rate not exceeding one per cent.
3.
As the eighth recital in the preamble to the directive states:
 the retention of other indirect taxes with the same characteristics as capital duty or the
stamp duty on securities might frustrate the purpose of the measures provided for in this Directive
and those should therefore be abolished.
4.
Article 10 of the directive accordingly provides:
Apart from capital duty, Member States shall not charge, with regard to companies, firms,
associations or legal persons operating for profit, any taxes whatsoever (c) in respect of
registration or any other formality required before the commencement of business to which a
company, firm, association or legal person operating for profit may be subject by reason of its legal
form.
5.
However, by virtue of art 12(1): Notwithstanding Articles 10 and 11, Member States may charge (e)
duties paid by way of fees or dues.
6.
The court had occasion to consider the scope of arts 10 and 12(1)(e) in its judgment in Ponente Carni
SpA v Amministrazione delle Finanze dello Stato Joined
5
cases C-71/91 and C-178/91 [1993] ECR I-1915. In issue in that case were Italian charges payable on
the first registration of a company and annually thereafter. The level and structure of the charges were
amended several times during the material period. The charges were invariably higher for public limited
companies than for private limited companies and, during one period, varied according to a public limited
companys capital. The charges were substantial; for 1988, for example, they were fixed at Lit 15m for
public limited companies and Lit 35m for private limited companies.
7.
The court held first that charges such as those in issue fell within the scope of art 10. That was so
even though the revenue from the charge contributed to the financing of the department responsible for
keeping the register. If member states were able to impose, without any limitation under Community law, a
charge other than capital duty on capital companies in respect of one of the essential formalities for their
formation, that would run counter to the objectives of the directive (see [1993] ECR I-1915 (para 30)).
8.
The court then went on to consider the scope of art 12 of the directive. Distinguishing its case law on
the EC Treaty provisions on charges having an equivalent effect to customs duties, the court held that art
12 permitted the charging of fees or dues representing
(37) the consideration for a transaction required by law for an object of public interest.
(38) That may be precisely the case with a charge required as consideration for a transaction
such as the registration of capital companies which is required by national law, in accordance with
Community law, in the interest of both third parties and of the companies themselves. (See [1993]
ECR I-1915 (paras 3738).)
9.
Turning to the calculation of the fees or dues, the court observed ([1993] ECR I-1915 (paras 4143)):
(41) The distinction between taxes prohibited by Article 10 of the Directive and duties paid by
way of fees or dues implies that the latter cover only payments collected on registration or annually,
the amount of which is calculated on the basis of the cost of the service rendered.
(42) A payment the amount of which had no link with the cost of the particular service or was
calculated not on the basis of the cost of the transaction for which it is a consideration but on the
basis of all the running and capital costs of the department responsible for that transaction would
have to be regarded as a tax falling solely under the prohibition of Article 10 of the directive.
(43) For certain transactions such as, for example, the registration of a company, it may be
difficult to determine their cost. In such case the assessment of the cost can only be on a flat-rate
basis and must be fixed in a reasonable manner, taking account, in particular, of the number and
qualification of the officials, the time they take and the various material costs necessary for carrying
out the transaction.
10.
The court added that member states were free to fix different amounts for public and private limited
companies provided that none of the amounts required for any of the companies exceeds the cost of the
transaction of registration (see para 44).
6

RELEVANT NATIONAL RULES

11.
Law No 468 of 29 September 1917 and, more recently, Law No 370 of 13 June 1973 authorised the
competent minister, and later the Danish Trade and Companies Office, to levy certain charges for
registering the formation of public limited companies and increases in their capital. Corresponding
provisions were introduced for private companies by Law No 371 of 13 June 1973.
12.
 Until 1 May 1992 the amounts charged comprised a basic fee plus a supplement calculated
proportionally at the rate of D Kr 1 per D Kr 1000 of the nominal value of the capital raised. The basic
charge was itself variable (on a degressive scale according to capital raised) until 1 January 1974; from
that date it was replaced by fixed basic charges which, during the period from 1 January 1974 to May
1992, ranged from D Kr 500 to D Kr 1,700 for notification of new public and private limited companies and
from D Kr 200 to D Kr 900 for notification of increases in capital of existing public and private limited
companies. From 1 February 1973 until 1 May 1992 the variable supplement was fixed at D Kr 4 per D Kr
1000 of the value of the capital raised.
13.
The Danish Audit Board (the Danmarks Rigsrevision) published a report on 13 May 1992 concluding
that the charges in question were in its experience the most remarkable example of a public authority
more than covering its operating costs. The report questioned the national legal basis for levying the
charges.
14.
The report, an outline of which was released in advance of the official publication date, led to the
abolition of the per mil supplement with effect from 1 May 1992. On the same date the fixed basic charge
for notification of new companies was altered from D Kr 1,700, applicable to both public and private
companies, to D Kr 2,500 for public companies and D Kr 1,800 for private companies; at the same time
the charge for increases in capital was reduced from D Kr 900 to D Kr 600.

THE TRADE AND COMPANIES OFFICE

15.
The Danish Companies Registry was set up in 1918 and on 1 January 1988 changed its name to the
Trade and Companies Office. The office comprises six divisions responsible for various matters relating to
the administration and drafting of legislation in the fields of company and business law.
16.
The order for reference contains two tables showing the operating expenses and income of the office.
The first table was prepared by the Audit Board and lists total operating expenses, total income and
surpluses for the years 1980 to 1990. The office considers that the figures in that table do not properly
reflect its income and expenditure relating to the field of company law and has produced a different set of
figures for the years 1987 to 1991. The table produced by the Audit Board shows surpluses ranging from
D Kr 49m (1980) to D Kr 1394m (1990). The table produced by the office shows smaller surpluses
ranging from D Kr 12m (1987) to D Kr 902m (1991).
17.
The figures produced by the office include its direct and indirect expenditure on administration of the
company laws and also the Ministry of Trades staff costs relating to preparation of company legislation
and administration of the office. The offices expenditure on administration of the company laws includes
the cost of registering company formations, conversions, increases in capital, mergers and other changes
and of ensuring compliance with procedures concerning lists of major shareholdings and prospectuses for
issue of shares. It also includes costs connected with matters such as: preparatory legal work on
regulations in the field of company law and on annual accounts; dealing
7
with complaints before the board of appeal of the Ministry of Trade and complaints to the ombudsman;
administering the rules on loss of capital, loans to shareholders, annual accounts (including examination
of accounts) and bookkeeping; and disseminating information through conferences, articles, leaflets and
meetings with professional organisations and groups. The figures also include a proportion of overheads
relating to matters such as financial management and budgeting, staff management, computer
development, administration of buildings and equipment, library, messenger services and staff training.

THE FACTS AND THE NATIONAL COURTS QUESTIONS

18.
In total eight sets of proceedings have been brought before the stre Landsret by Fantask A/S, Norsk
Hydro Danmark A/S, Robert Bosch A/S, Uponor A/S and Uponor Holding A/S, the Pen-Sam Group, Tryg-
Baltica Forsikring, Skadeforsikringsselskab A/S and Tryg-Baltica Forsikring, Livsforsikringssselskab A/S,
Aalborg Portland A/S and Alka Forsikring A/S. All the plaintiffs are public limited companies registered in
Denmark (although the Pen-Sam group comprises a number of private limited companies).
19.
All the cases concern claims for repayment of charges paid in connection with applications for
company formation or increases in capital. All except Fantask limit their claim to recovery of the per mil
charge. The amounts sought range from D Kr 2,900 (Fantask) to D Kr 48m (the Tryg group).
20.
According to the order for reference the claims of the plaintiffs in the main proceedings are governed
by para 1 of the Danish Law of Limitations of 22 December 1908, which prescribes a limitation period of
five years. Paragraph 2 of the 1908 Law provides that the period runs in principle from the moment at
which payment of the debt could have been demanded by the creditor, normally the moment when the
debt fell due. However, by virtue of para 3, where the creditor has been unaware of his claim through no
fault of his own, the period begins to run only from the moment at which the creditor was, or with normal
diligence would have been, in a position to demand payment of his debt.
21.
 In addition to those statutory rules, the national court refers to a principle, known as the forest fees
principle, developed by the Danish courts, which precludes a taxpayer from obtaining repayment of
overpaid charges if they were levied in accordance with long-standing rules assumed by both the
authorities and the taxpayer to be lawful. The scope and effect of that principle is disputed by the parties
to the main proceedings.
22.
Against that background the national court decided to put the following questions to the Court of
Justice:
(1) Does Community law impose requirements upon the Member States delimitation of the
concept of fees or dues in Article 12(1)(e) of Directive 69/335/EEC or are the individual Member
States free to decide what may be regarded as fees or dues for a specific service?
(2) May the basis for the calculation of duties charged under Article 12(1)(e) of Directive
69/335/EEC by a Member State for registration of formation or increase in capital of a public limited
company or a private limited company include the following types of costs or some of them: the
cost of salaries and pension contributions for officials not involved in effecting the registration, such
as the registration authoritys administrative staff or staff of the registration authority or other
authorities who are engaged on preparatory legal work in the field of company law; the cost of
8
effecting registration of other matters relating to companies, in respect of which the member
state has determined that no specific consideration is to be paid; the cost of performing duties,
other than registration, required of the registration authority in pursuance of company legislation
and legislation related thereto, such as examination of companies accounts and supervision of
companies bookkeeping; payment of interest and depreciation of all capital costs which are
regarded by the registration authority as concerning the field of company law and related fields of
law; the cost of official journeys not connected with the specific work of registration; the cost of the
registration authoritys external dissemination of information and guidance not connected with the
specific work of registration, such as lecturing, preparation of articles and brochures and holding of
meetings with trade organizations and other interested groups.
(3)(a) Is Article 12(1)(e) of Directive 69/335/EEC to be interpreted as meaning that a Member
State is precluded from fixing standardized charges by rules valid without limitation of time? (b) If
that is not possible, is a Member State required to adjust its scale of charges every year or at other
fixed intervals? (c) Is it of any significance for the answer whether charges are fixed in proportion to
the amount of the capital to be raised, as notified for registration?
(4) Is Article 12(1)(e) in conjunction with Article 10(c) of Directive 69/335/EEC to be interpreted
as meaning that the amount charged as consideration for a specific servicesuch as, for example,
registration of the formation or increase in capital of a public limited company or a private limited
companyis to be calculated on the basis of the actual cost of the specific serviceregistration
or can the duty for the individual registration be fixed at, for example, a basic charge together with
D Kr 4 per D Kr 1 000 of the nominal value of the capital subscribed, so that the amount of the duty
is independent of the registration authoritys time used and other costs necessary for effecting the
registration?
(5) Is Article 12(1)(e) in conjunction with Article 10(c) of Directive 69/335/EEC to be interpreted
as meaning that the Member State in calculating any amount to be recovered must work on the
basis that the duty must reflect the cost of the specific service at the time at which the service is
performed, or is the Member State entitled to make a comprehensive assessment over a longer
period, for example an accounting year or within the period in which it will be possible under
national law to assert a claim for recovery?
(6) If national law contains a general principle that, in determining claims for recovery of charges
made without the requisite authority, importance should be attached to the fact that the charge was
made in pursuance of rules which have been in force over a long period without either the
authorities or other parties having been aware that the charge was unauthorized, will Community
law preclude dismissal on those grounds of an action for recovery of charges levied contrary to
Directive 69/335/EEC?
(7) Does Community law make it impossible under national law for the authorities of a Member
State, in cases of claims for recovery concerning charges made contrary to Directive 69/335/EEC,
to contend and establish that national limitation periods start to run from a time at which an unlawful
implementation of Directive 69/335/EEC occurred?
9
(8) Does Article 10(c) in conjunction with Article 12(1)(e) of Directive 69/335/EEC as interpreted
in the foregoing questions result in rights on which citizens in the individual Member States may
rely before the national courts?

QUESTIONS 1 TO 5
23.
By its first five questions the national court requests guidance concerning the scope of the expression
fees or dues in art 12(1)(e) of the directive. It wishes to know whether member states are free to decide
the scope of the term themselves (Question 1), whether certain types of cost may be taken into account in
fixing the level of such fees or dues (Question 2) and whether fixed charges (including charges varying in
proportion to capital raised) not related to the actual cost of specific services may be imposed and the
extent to which they must be periodically reviewed (Questions 3 and 4). The national courts fifth question
concerns the calculation of sums to be reimbursed. It asks whether the calculation must be based on the
cost of the specific service at the time when the service is performed or whether it can be based on an
overall assessment over a longer period such as an accounting year or the period of the claim.
24.
The court has received written or oral argument on those questions from plaintiffs in the main
proceedings, the Danish Ministry of Trade and Industry, Danish and Swedish governments and the
European Commission.
25.
The plaintiffs in the main proceedings consider that the supplementary charge varying according to
nominal capital raised is prohibited by art 10 of the directive and does not fall within the concept of duties
paid by way of fees or dues permitted by art 12(1)(e). The charge bears no relation to the costs incurred
by the administration in effecting the registration. The plaintiffs in the main proceedings are supported by
the Commission. The latter considers that, while the basic charge appears to be a reasonable
remuneration for the registration services concerned, the supplementary ad valorem charge leads to large
surpluses and is unrelated to the specific service received by companies in connection with registration.
26.
The Danish government argues that the directive seeks to harmonise indirect taxes and not fees for
services provided in the general interest. Referring by analogy to the courts ruling on art 90 of the Treaty
in Criminal proceedings against Corbeau Case C-320/91 [1993] ECR I-2533, it argues that it is entitled to
fix fees at a level which is sufficient to cover the costs, both direct and indirect, not only of specific
registration services but of all the administrations activities in the field of company law. It is also entitled to
introduce a degree of solidarity in the charges by imposing a higher burden on larger companies. The
Danish government considers that its view is in accordance with the judgment in Ponente Carni and also
with the principle of subsidiarity, which the court must take into account even in interpreting legislation
existing prior to the entry into force of the Treaty on European Union (the TEU) (Maastricht, 7 February
1992; TS 12 (1994); Cmnd 2485; OJ 1992 C191). The Danish government is supported by the Danish
Ministry of Trade and Industry and by the Swedish government, which considers that the office was
entitled to take all the items of expenditure mentioned in the national courts second question into account
in fixing the amount of its charges.
27.
With reference to the national courts first question I accept the point made by the Danish ministry and
government that the directive does not as such harmonise fees and dues charged for services. It does not
specify what services may be supplied to companies in return for a remuneration or what the level of such
remuneration should be. It is nevertheless clear from the judgment in
10
Ponente Carni [1993] ECR I-1915 (esp para 30) that a charge connected with the registration of a
capital company falls within the prohibition laid down in art 10 of the directive and is lawful only if
permitted by art 12. It is also apparent from that judgment, that the directive imposes limits on what a
member state may lawfully charge by way of fees or dues under art 12(1)(e) (see [1993] ECR I-1915
(paras 4143)). The reason for that is plain. If it did not impose such limits the directive would be
ineffective since member states would be free to circumvent its provisions by imposing taxes other than
capital duty in the guise of fees or dues for supposed services.
28.
In that regard the Danish governments reference to the principle of subsidiarity in art 3b of the EC
Treaty, inserted by art G.5 of the TEU, is inapposite. Notwithstanding that provision, where the
Community has chosen to adopt a directive in an area not falling within its exclusive competence the
court must interpret it in accordance with its wording and aims and in a manner which will ensure that it is
effective.
29.
As regards the national courts second question concerning the costs which be taken into account in
fixing the charges in question, the following principles from the courts ruling in Ponente Carni [1993] ECR
I-1915: a member state may charge fees for certain individualised services which it performs for
companies; such services include a transaction such as the registration of capital companies which is
required by national law, in accordance with Community law, in the interest of both third parties and of the
companies themselves (see [1993] ECR I-1915 (paras 3738)); any fees or duties charged must be
calculated on the basis of the actual be fixed so high as cost of the specific services in question. They
may not to cover all the running and capital costs of the department responsible (see paras 4142);
where it is difficult to determine the cost of certain transactions, such costs may be assessed on a flat-rate
basis. Such assessment must be carried out on a reasonable basis, taking account of the number and
qualifications of the staff used, the time taken and the relevant material costs (see para 43). Different fees
may be fixed for private and public limited companies provided that none of the amounts required for any
of the companies exceeds the cost of the service.
30.
It seems to me that the present case raises two basic questions: first, for what activities may the office
charge fees or dues within the meaning of art 12(1)(e) and secondly, what limits does the directive place
on the manner in which such fees or dues are calculated?

Activities for which charges may lawfully be made

31.
It is clear from the judgment in Ponente Carni that Denmark is entitled to charge fees to cover the
costs of creating and maintaining files for companies in the companies register. That service, which is
provided to the company in the general interest, is expressly required by art 3(1) of Council Directive
(EEC) 68/151 on co-ordination of safeguards which, for the protection of the interests of members and
others, are required by member states of companies within the meaning of the second paragraph of art
58 of the EC Treaty, with a view to making such safeguards equivalent throughout the Community (the
first company law directive; OJ S Edn 1968 (I) p 41), and indeed many of the requirements concerning
deposit of documents and disclosure of information in regard to companies now flow from Community law.
Under art 3(2) of that directive a member state is obliged to ensure that all the documents and particulars
which must be disclosed under art 2 are kept on the file or entered in
11
the register. Article 2(1), as amended by the Act of Accession (1985): EC 27 (1985); Cmnd 9634; OJ
1985 L302 (Spain, Portugal), provides:
Member States shall take the measures required to ensure compulsory disclosure by
companies of at least the following documents and particulars: (a) The instrument of constitution,
and the statutes if they are contained in a separate instrument; (b) Any amendments to the
instruments mentioned in (a), including any extension of the duration of the company; (c) After
every amendment of the instrument of constitution or of the statutes, the complete text of the
instrument or statutes as amended to date; (d) The appointment, termination of office and
particulars of the persons who either as a body constituted pursuant to law or as members of any
such body: (i) are authorized to represent the company in dealings with third parties and in legal
proceedings; (ii) take part in the administration, supervision or control of the company. It must
appear from the disclosure whether the persons authorized to represent the company may do so
alone or must act jointly; (e) At least once a year, the amount of the capital subscribed, where the
instrument of constitution or the statutes mention an authorized capital, unless any increase in the
capital subscribed necessitates an amendment of the statutes; (f) The balance sheet and the profit
and loss account for each financial year. The document containing the balance sheet must give
details of the persons who are required by law to certify it. (g) Any transfer of the seat of the
company; (h) The winding up of the company; (i) Any declaration of nullity of the company by the
courts; (j) The appointment of liquidators, particulars concerning them, and their respective powers,
unless such powers are expressly and exclusively derived from law or from the statutes of the
company; (k) The termination of the liquidation and, in Member States where striking off the
register entails legal consequences, the fact of any such striking off.
32.
Under art 3(4) of the first company law directive those documents and particulars are to be disclosed
by publication in the national gazette appointed for that purpose by the member state; disclosure may be
limited to publication of a reference to the document deposited.
33.
Numerous further requirements flow from later directives. For example, art 3 of Council Directive
(EEC) 77/91 on coordination of safeguards which, for the protection of the interests of members and
others, are required by member states of companies within the meaning of the second paragraph of art
58 of the Treaty, in respect of the formation of public limited liability companies and the maintenance and
alteration of their capital, with a view to making such safeguards equivalent (the second company law
directive; OJ 1977 L26 p 1) contains a detailed list of the information which must be contained in a public
companys statutes or other documents published in accordance with art 3 of the first directive. The
directive also provides for publication of certain transactions such as increases in capital (art 25(1)), offers
of subscription to existing shareholders on a pre-emptive basis and restrictions on pre-emption (art 29(3)
and (4)), reductions in subscribed capital (art 30), redemption of subscribed capital without reduction (art
35(a)), compulsory withdrawal of shares (art 36(1)(e)) and redemption of redeemable shares (art 39(h)).
Council Directive (EEC) 78/855 based on art 54(3)(g) of the Treaty concerning mergers of public limited
liability companies (the third company law directive; OJ 1978 L295 p 36) requires disclosure, in
accordance with art 3 of the first directive, of various
12
matters connected with proposed mergers and acquisitions: see arts 6 and 18. Council Directive
(EEC) 78/660 based on art 54(3)(g) of the Treaty on the annual accounts of certain types of companies
(the fourth company law directive; OJ 1978 L222 p 11) and Council Directive (EEC) 83/349 based on art
54(3)(g) of the Treaty on consolidated accounts (the seventh company law directive; OJ 1983 L193 p 1)
require publication of a companys or groups annual accounts and the annual report, together with the
auditors report: see art 47 of the fourth directive and art 38 of the seventh directive. Article 16 of Council
Directive (EEC) 82/891 based on art 54(3)(g) of the Treaty, concerning the division of public limited
liability companies (the sixth company law directive; OJ1982 L378 p 47) requires publication of divisions
of companies. Council Directive (EEC) 89/666 concerning disclosure requirements in respects of
branches opened in a member state by certain types of company governed by the law of another state
(the eleventh company law directive; OJ 1989 L395 p 36) lays down disclosure requirements concerning
a branch opened in a member state by a company governed by the law of another member state. Article 3
of Council Directive (EEC) 89/667 on single-member private limited-liability companies (the twelfth
company law directive; OJ 1989 L395 p 40) requires disclosure of single member companies.
34.
 It seems to me that matters of that kind, prescribed by Community or by national law, come within the
statement in Ponente Carni [1993] ECR I-1915 (paras 3738) cited at para 8, above. They may be
regarded as entailing services supplied to companies in the general interest by the registration authority.
35.
By its second question the national court inquires about three specific activities: other work of the
office such as preparatory legal work in the field of company law; registration of other matters relating to
companies in respect of which no specific charge is made; examination of company accounts and
supervision of their bookkeeping.
36.
I do not think that the more general activities of the office or the Ministry of Trade and Industry in the
field of company law, such as administration going beyond maintenance of the companies register,
monitoring EC developments and preparatory work on new legislation, can be regarded as a service for
which fees or dues may be charged under art 12(1)(e). Such activities do not entail the provision of
specific services to individual companies, as required by the judgment in Ponente Carni, but rather form
part of the general business of government.
37.
On the other hand it is clear that the services provided to individual companies by the office include a
number of tasks for which no specific charge is made. It appears that the office makes a charge only for
first registration and for registration of increases in capital. As I have stated at paras 31 to 35, above, a
registration authority is responsible for ensuring compliance with numerous other registration and
disclosure requirements. As I shall explain at para 45, below, it is in my view open to a member state to
confine its charges solely to more substantial transactions and to take account of the costs of minor tasks
performed in connection with maintenance of the register in the fees or dues charged for such
transactions.
38.
As regards, more specifically, examination of accounts and accounting records, it must be
remembered that art 51(1) of the fourth company law directive lays down a compulsory audit requirement
for companies (other than very small companies). Council Directive (EEC) 84/253 based on art 54(3)(g) of
the Treaty on the approval of persons responsible for carrying out the statutory
13
audits of accounting documents (the eighth company law directive; OJ 1984 L126 p 20) lays down
rules governing the qualifications of persons or firms who may be approved by member states for the
purpose of carrying out a statutory audit. It may be assumed that such an audit, which is also carried out
partly in the general interest and for which a fee is charged by the auditor, will include inspection of a
companys accounting records and accounts with a view to ensuring that they accurately represent the
affairs of the company and comply with accepted accounting practice and with statutory or stock
exchange requirements.
39.
Against that background it seems to me that it would not be justified for an authority of a member state
to make a further charge to a company for work duplicating the work already carried out by an auditor
approved by that state. That applies particularly to checks on the companys underlying accounting
records, ie its bookkeeping. I am nevertheless willing to accept that the registration authority, which, as
the Danish government points out, is ultimately responsible for ensuring that registration and disclosure
requirements are met, must retain the right to inspect the audited accounts presented to it in order to
ensure that they comply with the various disclosure requirements laid down by national and Community
law (in particular the fourth and seventh company law directives) and to charge an appropriate fee for that
work.

Calculation of the cost of the services

40.
It is clear from the judgment in Ponente Carni [1993] ECR I-1915 (para 41) that any fees or duties
charged must be calculated on the basis of the actual cost of the specific services in question. The court
added that a member state could not charge an amount which
had no link with the cost of the particular service or was calculated not on the basis of the cost
of the transaction for which it is a consideration but on the basis of all the running and capital costs
of the department responsible for that transaction. (See [1993] ECR I-1915 (para 42).)
41.
That passage makes it clear that the fees must reflect the costs of specific services and cannot be
used to finance the general administrative expenditure of the department concerned. In that regard the
Danish governments reference to the judgment in Corbeau [1993] ECR I-2533 is inapposite. There the
court accepted, in the context of art 90 of the Treaty, that it was legitimate to protect an undertaking
entrusted with the performance of certain tasks in the general interest, such as the provision of postal
services, from competition in profitable sectors of its activity in order to allow it to achieve an economic
equilibrium by offsetting less profitable sectors against the profitable sectors (see [1993] ECR I-2533
(para 17)). The rationale for that principle is clear. In the absence of restrictions on competition private
operators not subject to the same obligations as the public-service operator would be able to undercut the
latters prices in the profitable sectors, leaving it solely with the unprofitable sectors (see para 18). That
ruling, which takes account of the varying cost structures of the different sectors of a public-service
undertakings economic activity, has no relevance to the activities of the office. Any profits which the office
makes from company registration will not be used to subsidise other, less profitable sectors of an
economic activity but as a source of finance for government expenditure. In so far as they exceed the
costs of registration its charges are therefore more in the nature of a tax than fees or dues for identifiable
services.
14
42.
The courts statement in the judgment in Ponente Carni [1993] ECR I-1915 (para 42) that the charges
should not be based on all the running and capital costs of the [relevant] department may have been
prompted by the Italian governments suggestion that the fees could be fixed at a level which would
finance the entire machinery for disclosure of documents. As I noted in in my opinion in that case, the only
costs which can be taken into consideration in setting the fees are the administrative costs of effecting the
registration (including, I should add, publication in the appointed national gazette). The other costs of the
system, in particular those involved in providing copies of documents to individuals, cannot properly be
charged to the companies but must be financed by other means, for instance by means of a fee charged
to the recipients of the information as expressly permitted by art 3(3) of the first directive (see the opinion
in Ponente Carni [1993] ECR I-1915 (para 32)).
43.
The present case, in particular the national courts second question, requires the court to go a little
further than in Ponente Carni and offer more detailed guidance on the manner in which the costs of a
registration authority are to be calculated. It would in my view be appropriate to base the calculation of the
relevant costs on the normal principles of cost or management accounting. In other words, the fees may
reflect the direct costs and overheads of the authority attributable to the services in question. Thus, such
costs might include, in addition to direct material costs and the salary and social security costs of the staff
carrying out the services, a proportion of the overheads of the authority such as lighting and heating, staff
management costs, computer operation and development costs, office rents or depreciation, depreciation
of other fixed assets such as furniture and equipment etc. The proportion of such costs referable to
registration services should, where possible, be determined by direct attribution, for example by
identifying the rent payable for the offices used specifically for the services in question. Where costs
relate both to registration services and to other activities such as preparatory work on legislation, it will be
necessary to make an apportionment on the basis of appropriate criteria such as staff employed on the
various types of activity, office space used, computer time used etc.
44.
As regards the more specific items referred to by the national court in its second question, the cost of
official journeys, interest and depreciation costs and the cost of disseminating information in the field of
company law must be left out of account to the extent to which they are not directly connected with the
specific work of registration or filing. On the other hand, the fees charged could, I think, lawfully cover the
cost of publications directly linked to the registration and filing services, eg guides for directors concerning
disclosure requirements.
45.
I accept the Danish governments proposition that in the interests of administrative simplicity the office
should be able to limit its charges to major transactions and pass on the costs of comparatively minor
services (eg recording changes of registered office or directors) in the registration charges which it does
make. The contrary view would require the office to make individual charges for every service which it
performs, however small.
46.
Contrary to the assertion of the plaintiffs in the main proceedings and of the Commission, I do not
think, purely as a matter of principle, that the use of a proportionate charge based on capital raised is
necessarily any more unrelated to the costs of individual services than a flat-rate charge. If it were shown
that on average the costs of registration did to some degree increase in line with capital raised, the
introduction of a proportional element into the total charges could produce a fairer scale of fees.
15
47.
However, while I am willing to accept that registration of the formation of larger public companies and
of large share issues, possibly in conjunction with a merger or restructuring, may involve more than the
average time required for registrations generally, I doubt that the costs of such transactions and the
amount of capital raised increase in direct proportion. The Danish government itself implicitly
acknowledges that when it suggests that its scale of charges introduces a degree of solidarity between
larger and smaller companies. As Norsk Hydro and Tryg-Baltica Forsikring pointed out at the hearing,
such solidarity is difficult to reconcile with the courts statement in Ponente Carni that, although a member
state might charge different fees for public and private companies, none of the amounts required for any
of the companies [should exceed] the costs of the transaction of registration. Contrary to the suggestion
of the Danish ministry, I do not think the courts ruling in Corbeau has any relevance here. Unlike, for
example, a post office which must provide unprofitable services to the population of outlying districts,
there is no particular reason why the office should not be able to structure its charges to cover its costs for
the various services offered to companies of different sizes. That must however be weighed against the
right of a registration office within reason to base its fees on average costs.
48.
In any event it is clear in the present case that the proportional charge goes beyond application of a
solidarity principle. The lack of any direct correlation between the costs of registration and the amount of
capital raised, the high level of the proportional charge (D Kr 4 per D Kr 1000) and the absence of any
ceiling on the charge will inevitably have led, during the period in question, to charges which in total
exceeded the total costs of the services provided to companies. Confirmation of that is provided by the
surpluses shown by the tables set out in the order for reference. While I accept that small or isolated
surpluses may be necessary to cover future pension commitments (depending on how pensions are
funded), to supplement depreciation reserves for replacement of fixed assets or to cover deficits in other
years, the large recurring surpluses shown by the Audit Boards figures or even those of the office are
scarcely consistent with the proposition that the office merely covered its costs. Moreover, it is apparent
from the order for reference that the Audit Board came to the view that the offices charges were
substantially in excess of costs (which in fact include the cost of certain activities which cannot be
regarded as services provided to the companies).
49.
Ultimately, however, it is for the national court, in the light of the criteria set out above, and the figures
which are available to it, to determine the amount of the costs of the office which could lawfully be passed
on to the plaintiffs in the main proceedings and the amount of any repayment due.
50.
With reference to the national courts third question it is clear from the judgment in Ponente Carni that
a member state is entitled to charge flat-rate fees where the costs of specific services would be difficult to
determine. It seems to me that that will normally be the case since individual costing is unlikely to be
practicable in the case of a companies registry responsible for processing large numbers of comparatively
small transactions. A member state is however obliged to review its fees periodically, perhaps every few
years, in order to ensure that they do reflect the costs of the transactions concerned.
51.
By its fifth question the national court asks whether any repayment which falls to be made must be
calculated on the basis that the duty should reflect the cost of the specific service at the moment at which
the service is performed, or whether the calculation can be based on a comprehensive assessment over
a
16
longer period, for example an accounting year or the period within which it is possible under national
law to assert a claim for recovery.
52.
In my view it is appropriate to leave it to the national court to arrive at the best estimate of any
repayment due in the light of the figures available to it. The national court could base its calculation either
on the actual cost to the registration authority of the specific services supplied to each company or, if that
is not possible, on the average cost of the services at or about the relevant time or, if necessary, over a
longer period. If the repayment is based on average costs, it may be appropriate for the national court, if it
is able to do so, to adjust the figures to take account of the different costs of larger and smaller
transactions.

QUESTIONS 6 AND 7
53.
By its sixth question the national court asks whether Community law precludes it from dismissing the
claims for reimbursement on the ground that the charges were levied in pursuance of rules which had
been in force for a long time and had been assumed by all concerned to be lawful. The national courts
seventh question is designed to establish whether the time limit for instituting proceedings could begin to
run before the directive was properly implemented.
54.
The latter question is prompted by the courts ruling in Emmott [1991] ECR I-4269, a case concerning
Council Directive (EEC) 79/7 on the progressive implementation of the principle of equal treatment for
men and women in matters of social security (the equal treatment directive; OJ 1979 L6 p 24). In that
case the Irish authorities failed to grant Mrs Emmott equal benefits pursuant to the directive until 28
January 1988, although the directive should have been implemented by 23 December 1984. Mrs Emmott
instituted judicial review proceedings with a view to obtaining equal benefits from 23 December 1984. The
Irish authorities objected that she had failed to make her application within the three-month period from
the date when the grounds arose as required by Irish law. However, the court held ([1991] ECR I-4269
(paras 2123)):
(21) So long as a directive has not been properly transposed into national law, individuals are
unable to ascertain the full extent of their rights. That state of uncertainty for individuals subsists
even after the Court has delivered a judgment finding that the Member State in question has not
fulfilled its obligations under the directive and even if the Court has held that a particular provision
or provisions of the directive are sufficiently precise and unconditional to be relied upon before a
national court.
(22) Only the proper transposition of the directive will bring that state of uncertainty to an end
and it is only upon that transposition that the legal certainty which must exist if individuals are to be
required to assert their rights is created.
(23) It follows that, until such time as a directive has been properly transposed, a defaulting
Member State may not rely on an individuals delay in initiating proceedings against it in order to
protect rights conferred upon him by the provisions of the directive and that a period laid down by
national law within which proceedings must be initiated cannot begin to run before that time.
55.
 The Danish, French, Italian and UK governments consider that the principle as formulated above is too
broad inasmuch as it has the effect of imposing almost limitless retrospective liability on member states.
Norsk Hydro and Tryg-Baltica, Alka Forsikring and others and the Commission contend that
17
the court should adhere to its ruling in Emmott, so that the five-year limitation period did not begin to
run until 1 May 1992 when Denmark implemented the directive by abolishing the supplementary charge.
They argue that to allow a member state to rely on time limits in a case such as the present would permit
it to escape the consequences of its own unlawful conduct and discourage it from taking steps to remedy
the defects in its rules.
56.
In what follows I shall explain why I consider the governments criticisms of Emmott to be justified and
show that a broad view of Emmott cannot in any event be reconciled with the courts subsequent case
law. I shall however also show that, in the light of other developments in the courts case law on remedies
in the national courts, the concerns which led the Commission to favour a broad view of Emmott can be
accommodated in a coherent system of remedies allowing a proper balance to be struck between the
need for effective protection of Community rights, including those under directives, and the principles of
respect for national procedural autonomy and legal certainty. Against that background I shall, finally, turn
to the Danish rules in issue here.

The criticisms of the Emmott ruling

57.
All the governments have emphasised the far-reaching financial consequences of the Emmott ruling.
For example, at the hearing the Italian government stated that, following the courts judgment in Ponente
Carni, the Italian courts, applying the ruling in Emmott, had set aside the three-year time limit laid down by
national law, exposing the Italian state to substantial repayment claims. Similarly the French government
pointed to a judgment of 9 July 1996 in which the French Cour de Cassation, referring to Emmott, held
that the tax authorities could not rely on the limitation periods laid down by French law in order to resist
claims dating back to the 1970s based on the courts ruling in Bautiaa Srl v Directeur des Services
Fiscaux, Socit Franaise Maritime SA v Directeur des Services Fiscaux Joined cases C-197/94 and C-
252/94 [1996] ECR I-505, from which it followed that a French registration duty of 12% on contributions
to companies of movable property made in connection with a merger was unlawful.
58.
The French government noted further that the ruling in Emmott [1991] ECR I-4269, which is based on
the special characteristics of Community directives, has the paradoxical result that greater protection is
given to rights under directives than to rights conferred by regulations and by the Treaty itself.
59.
Turning to the reasoning underlying the ruling, the French government noted that the courts statement
that, until a directive had been properly transposed into national law, individuals were unable to ascertain
the full extent of their rights was unsupported in the judgment. The French government referred however
to the passage in the opinion of Advocate General Mischo in which he argued that the very nature of the
directive precluded its final date for implementation from being taken as the starting point of the three-
month period for applying for judicial review. He said ([1991] ECR I-4269 (paras 2629)):
(26) The principle that everyone is presumed to know the law cannot be pleaded against
individuals in the case of a directive which has not yet been transposed. A directive binds only the
Member State; it is not addressed to individuals. It is therefore not possible to infer obligations for
individuals from the directive as such [see Marshall v Southampton and South West Hampshire
Health Authority (Teaching) Case 152/84 [1986] 2 All ER 584,
18
[1986] ECR 723]. It follows that the directive also cannot supply a starting point for a time-limit
which could be raised as a bar to claims by individuals.
(27) It may also be borne in mind in this regard that the publication of directives in the Official
Journal of the European Communities, which was brought up by the respondents at the hearing, is
fundamentally different from the publication in the Official Journal of measures binding on
individuals. That is not publication required by law producing legal effects, as in the case of
regulations, but only publication for information.
(28) It is also noteworthy that the text of a directive, once published, does not enable individuals
to know precisely the time-limit for its transposition. It mentions only a period before the expiry of
which the Member States to whom the directive is addressed must have transposed it, as well as
the fact that the period begins to run from the date of notification of the directive to the Member
State. However, that date is not given and there is no reason to suppose that individuals are aware
of it.
(29) Furthermore, although it is certainly true that the interpretation given by the Court in a
preliminary ruling has retroactive effect in so far as it indicates how the rule which is interpreted
should have been understood from the beginning, it is, however, also beyond argument that, before
the Court settled the matter, it is not certain that the directive or a particular article thereof has
direct effect.
60.
 The French government criticises the Advocate Generals reasoning on a number of grounds. First, the
fact that a directive is not addressed to individuals does not, in the French governments view, prevent its
publication from causing a limitation period to commence to run. The French government refers by
analogy to Commission decisions in matters of state aid which, although addressed to member states,
must be challenged before the Court of First Instance of the European Communities by the undertaking
receiving the aid or by competitors within two months of their publication.
61.
Secondly, it argues that it is immaterial whether the publication of a directive satisfies a legal obligation
imposed by the Treaty. What matters is whether it was actually published. Directives have in practice
been published systematically for many years, and art 191 of the Treaty, as amended by the TEU, now
requires publication of most directives.
62.
Thirdly, the French government is unimpressed by the argument that it is not always clear from a
directive when an implementation period running from the date of notification to a member state expires. It
contends that in such circumstances, which are rare, the date of the adoption of the directive will provide
a reasonably accurate guide and a national court, in applying a limitation period, can take account of any
uncertainty existing on that point.
63.
Finally, the French government considers that the fact that it may be uncertain whether a provision has
direct effect does not exempt an individual from his obligation to show a degree of diligence and to take
the necessary steps to safeguard his rights. The French government notes moreover that the same doubt
may exist with respect to a Treaty provision.
64.
Although, as I shall explain below, I think the result in Emmott can be justified, I share the
governments reservations about the broad wording of the ruling.
65.
As the French government states, it is not clear from the judgment precisely why the court took the
view that an individual could not be sure of the
19
full extent of his rights before a directive had been properly implemented. For the reasons given by the
French government I do not first of all think that an individual can be considered not to have notice of a
directive merely because it has not been transposed into national law. The anomalies noted by Advocate
General Mischo, namely the absence of any publication obligation and the fact that the final date for
implementation of a directive was not always completely clear from the directive itself, stemmed from the
fact that the relevant Treaty rules and practice of the Community legislature did not at the time reflect the
enhanced status which the court had accorded to directives in its case law, namely as instruments
conferring rights which individuals can assert in the national courts. It would be somewhat paradoxical to
rely on such anomalieswhich in any event have now been corrected 3to support the conclusionitself
paradoxicalthat rights flowing from directives should be given greater protection than those arising from
the Treaty itself or from measures which have direct application under the Treaty.
3
Directives must now generally be published pursuant to art 191 of the EC Treaty, as amended by art G.63 of the TEU.
The final date for implementation of directives now seems to be either expressly mentioned or fixed by reference to the
publication date.

66.
Moreover, the suggestion that in the absence of correct implementation an individual can never be
sure of the full extent of his rights is in my view difficult to reconcile with the conditions for the direct effect
of directives. The direct effect of a provision of a directive presupposes that the basic content of the right
conferred is sufficiently clear (or, at least, amenable to judicial determination) for it to be protected by the
national legal system. It is inconsistent to recognise that a provision is sufficiently clear to create remedies
for individuals in the national courts while at the same time exempting the individual, on the ground that
the rights conferred upon him are insufficiently clear, from the limits placed by national law on the exercise
of such remedies. The inevitable and wholly anomalous result is that already mentioned, namely to confer
privileged status on directives by comparison with other Community provisions, even those of superior
rank. The recognition of the direct effect of directives was surely intended to ensure that, notwithstanding
defective implementation, the rights which they were intended to confer enjoyed the same degree of
protection in the national courts as those arising from directly applicable Community law.
67.
In any event, I doubt whether an individual is more likely to be uncertain of his rights in the case of an
unimplemented directive than in the case of a Treaty provision. Although not requiring implementation in
national law, Treaty provisions, which by nature are broadly worded, often leave considerable scope for
uncertainty as to the extent of the rights, if any, which they confer on individuals.
68.
The governments arguments concerning the financial consequences of Emmott also raise an
important point of principle. As they correctly observe, the Emmott ruling, if read literally, would expose
member states to the risk of claims dating back to the final date for implementing a directive. For
example, it would permit claims based on Directive 69/335, which was to be implemented by 1 January
1972, to be brought in respect of the last 25 years notwithstanding national limitation periods.
69.
 Moreover, such liability would arise even in the event of a minor or inadvertent breach. Such a result
wholly disregards the balance which must be
20
struck in every legal system between the rights of the individual and the collective interest in providing
a degree of legal certainty for the state. That applies particularly to matters of taxation and social security,
where the public authorities have the special responsibility of routinely applying tax and social security
legislation to vast numbers of cases.
70.
The scope for error in applying such legislation is considerable. Regrettably that is particularly so in
the case of Community legislation, which is often rather loosely drafted. For example, the provision in
issue in Ponente Carni and in the present case is hardly a model of clarity: the directive provides no
indication of the scope of the term fees or dues or how they are to be calculated. The recent cases of
Argos Distributors Ltd v Customs and Excise Comrs Case C-288/94 [1996] ECR I-5311 and Elida Gibbs
Ltd v Customs and Excise Comrs Case C-317/94 [1997] All ER (EC) 53, [1996] ECR I-5339 provide a
further example of how huge repayment claims can arise from a comparatively minor error 4 in
implementing a Community tax directive. In those cases the court found that the fiscal treatment accorded
by the United Kingdom to voucher transactionsused extensively in that member state as a business
promotion techniquewas not in accordance with Council Directive (EEC) 77/388 on the harmonisation
of the laws of the member states relating to turnover taxesCommon system of value added tax: uniform
basis of assessment (OJ 1977 L145 p 1). The resultant repayment claims are reported to be between
200m and 400m (see The Times, 25 October 1996).
4
It may be noted that in his opinions of 27 June 1996 in those two cases Advocate General Fennelly took a different view
from that taken by the court.

71.
It might be objected that it is not unreasonable to require member states to refund overpaid charges
given that they were not entitled to collect them in the first place. However, that view disregards the need
for states and public bodies to plan their income and expenditure and to ensure that their budgets are not
disrupted by huge unforeseen liabilities. That need was particularly clear in Denkavit Internationaal BV v
Kamer van Koophandel en Fabrieken voor Midden-Gelderland Case C-2/94 [1996] ECR I-2827 in which
repayment was sought of the annual levies imposed by the Netherlands Chambers of Trade and Industry
in order to finance their activities. As I noted in my opinion in that case, retrospective claims of up to 20
years would have had catastrophic effects on their finances (see [1996] ECR I-2827 (para 64)).
72.
In short, therefore, my main reservations about a broad view of the Emmott ruling are that it disregards
the need, recognised by all legal systems, for a degree of legal certainty for the state, particularly where
infringements are comparatively minor or inadvertent; it goes further than is necessary to give effective
protection to directives; and it places rights under directives in an unduly privileged position by
comparison with other Community rights. Moreover a broad view cannot be reconciled with the courts
subsequent case law on time limits.

Subsequent case law on time limits

73.
It is clear from the rulings in Steenhorst-Neerings v Bestuur van de Bedrijfsvereniging voor
Detailhandel, Ambachten en Huisvrouwen Case C-338/91 [1993] ECR I-5475 and Johnson v Chief
Adjudication Officer (No 2) Case 410/92 [1995] All ER (EC) 258, [1994] ECR I-5483 that Emmott does not
preclude reliance
21
on rules limiting the period to which a claim may relate back. It might be thought that such time limits
are distinguishable from time limits for bringing proceedings, such as the three-month time limit for
applying for judicial review in Emmott; the latter constitute an absolute bar to a claim whereas the former
merely limit the extent of the claim (see the judgment in Steenhorst-Neerings [1993] ECR I-5475 (para
21)).
74.
As I shall explain at paras 85 and 86, below, the judgments in Steenhorst-Neerings and in particular
Johnson suggest however that in Emmott it was the application of the time limit in the particular
circumstances of the case which denied Mrs Emmott the opportunity to assert her rights under the
directive in the courts. Moreover I doubt whether it is possible to make any general distinction of the kind
mentioned above between the two types of time limit. For example, a five-year time limit for instituting
proceedings, if applied to recurring taxes or benefits, could equally be viewed as a rule limiting to five
years the extent to which a claim may relate back.
Conversely a rule, such as that in Johnson, which limits entitlement to benefits to a period not
exceeding 12 months prior to a claim, could equally be viewed as a time limit barring claims in respect of
a particular period after 12 months; although in most cases the application of such a time limit will merely
reduce the amount of benefit paid, it will lead to complete denial of the claim in all cases in which
entitlement to benefit has ceased 12 months before the claim is made.
75.
Some time limits are of a hybrid nature. For example UK law commonly lays down comparatively short
time limits for appealing against current assessments to tax, while separately allowing a taxpayer to make
a claim in respect of previous years on more limited grounds (see eg s 33 of the Taxes Management Act
1970). The latter type of time limit, although expressed as a time limit on retrospective claims, might
equally be viewed as an extended time limit for bringing proceedings but on more limited grounds.
76.
In any event, as I noted in my opinion in Denkavit, the courts reasoning in Emmott, in particular the
statement that an individual cannot ascertain his rights until a directive has been properly implemented,
would, if it were read without qualification, preclude reliance by member states on either type of time limit.
Both permit a member state, notwithstanding the uncertainty for the individual, to deny claims in respect
of periods in which a directive has not been properly implemented.
77.
Moreover, BP Supergas Anonimos Etaira Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece
Case C-62/93 [1995] All ER (EC) 684, [1995] ECR I-1883 concerned a similar type of time limit to that in
issue in Emmott, namely a three-year time limit for bringing proceedings laid down by Greek law; yet the
court, although referring to Emmott, did not suggest that the ruling precluded reliance on such a time limit
in the absence of proper implementation of Directive 77/388.

Recent case law on state liability in damages

78.
The issues arising from the Emmott ruling cannot however be considered in isolation. It is necessary,
in assessing the adequacy of the protection of Community rights, in particular those under directives, to
have regard also to more recent developments in the case law, in particular the line of cases beginning
with Francovich v Italy Joined cases C-6/90 and C-9/90 [1991] ECR I-5357 in which the court has
recognised the right of individuals to seek
22
compensation from a member state for loss or damage caused by failure to implement a directive. In
the last few months the court has given two further rulings on the subject of damages claims which are of
particular relevance to the present case.
79.
In Socit Comateb v Directeur Gnral des Douanes et Droits Indirects Joined cases C-192218/95
[1997] ECR I-165 the court held that a member state could, on certain conditions, resist repayment of
charges levied in breach of Community law on the ground that repayment would unjustly enrich the trader.
In the framework of repayment proceedings a national court could, where domestic law permitted, take
account of damage incurred by the trader by way of loss of business which wholly or partly negated any
unjust enrichment on his part. The court also recognised the right of the trader to bring a separate
damages claim subject to the conditions laid down in Brasserie du Pcheur SA v Germany, R v Secretary
of State for Transport, ex p Factortame Ltd Joined cases C-46/93 and C-48/93 [1996] All ER (EC) 301,
[1996] ECR I-1029 in the competent courts in accordance with the appropriate procedures of national law
in order to obtain reparation of the loss caused by the overpaid charges (irrespective of whether those
charges had been passed on).
80.
In R v Secretary of State for Social Security, ex p Sutton Case C-66/95 [1997] All ER (EC) 497 the
court held that art 6 of the equal treatment directive did not require a member state to pay an individual
interest on arrears of a social security benefit such as invalid care allowance where the delay in payment
of the benefit was the result of discrimination prohibited by the directive. The court distinguished its
judgment in Marshall v Southampton and South West Hampshire Area Health Authority (No 2) Case C-
271/91 [1993] 4 All ER 586, [1993] ECR I-4367 where it had held that compensation for discriminatory
dismissal could not leave out of account factors, such as effluxion of time, which could reduce its value; in
such a case the interest was an essential component of the compensation. By contrast, amounts payable
by way of arrears of social security benefits did not in any way constitute reparation for loss or damage of
the kind in issue in Marshall (No 2). However, the court pointed out that Mrs Sutton might have a claim in
damages if the conditions laid down in Brasserie du Pcheur were met. It was for the national court to
verify whether that was so and to determine the amount of the damages.
81.
Thus, in those judgments the court recognised that repayment or entitlement claims against state
authorities and damages claims against the state may co-exist as independent remedies in matters of
taxation and social security. Repayment or entitlement claims and damages claims are claims of a
different nature, and what is recoverable under each may differ. For example, interest on arrears of
benefit irrecoverable under an entitlement claim may be recoverable under a damages claim.
82.
It seems to me that, in matters of taxation and social security, an individual should be able, where the
conditions for a Francovich claim are met, to obtain full compensation for loss or damage incurred,
including the amount of the tax overpaid or benefit withheld (subject to the setting off of amounts actually
obtained by means of other remedies). The duty to mitigate loss or damage by using other remedies,
recognised by the court in its judgment in Brasserie du Pcheur [1996] All ER (EC) 301, [1996] ECR I-
1029 (para 84), has no relevance to the restitutionary or entitlement element of the claim, ie the amount of
the overpaid tax or benefit denied. Whereas the duty to mitigate will be relevant in
23
the case of a loss of profits, the loss corresponding to overpaid tax or denial of benefits will not be
aggravated by the delay in bringing proceedings5.
 5
Interest accruing on the principal sum merely reflects the fact that the state rather than the individual has had the use of
the money during the period in question; it cannot be regarded as an aggravation of the loss.

83.
The existence of a wholly independent claim for damages, subject to longer time limits than the
comparatively short ones prescribed for restitutionary and entitlement claims in many member states, is
consistent with the different nature of the claim. Its basis is not merely the unjust enrichment of the state
resulting from simple error in the routine application of technical legislation but a serious violation of
individual rights, calling for a re-appraisal of the balance between such rights and the collective interest in
a measure of legal certainty for the state.
84.
Such an approach has a number of advantages. It provides comprehensive protection of individual
rights within the existing framework of remedies and time limits, making it unnecessary to set aside
national time limits for repayment or entitlement claims. It draws a proper balance between individual
rights and the collective interest in legal certainty; in particular it takes proper account of the degree of
culpability of the state in failing properly to implement a directive and ensures that claims are properly
categorised and brought in the appropriate courts in accordance with appropriate substantive and
procedural conditions, in particular time limits. Moreover, while not giving unduly privileged treatment to
rights under Community directives, such an approach will nevertheless provide a substantial incentive to
member states to implement directives on time and to make every effort to do so properly; it will also
encourage them to repair without delay any inadequacies that become apparent, for example because of
a ruling of the court.

The scope of the Emmott ruling

85.
An important factor in Emmott was that it would have seemed unjust in the particular circumstances of
the case to permit the Irish authorities to rely on the time limit laid down by national law. As I noted in my
opinion in Denkavit, the court in its rulings in Steenhorst-Neerings and Johnson emphasised the following
circumstances: Mrs Emmott had sought payment of the benefits in question on the basis of the courts
judgment in McDermott v Minister for Social Welfare Case 286/85 [1987] ECR 1453 the administrative
authorities had declined to adjudicate on her claim until the litigation concerning the directive pending
before the national courts had been concluded; and the authorities sought to rely on the time limit
notwithstanding the failure correctly to implement the directive.
86.
Consequently, as the court stated in the judgment in Johnson [1995] All ER (EC) 258, [1994] ECR I-
5483 (para 26):
the solution adopted in Emmotts case was justified by the particular circumstances of that
case, in which a time-bar had the result of depriving the applicant of any opportunity whatever to
rely on her right to equal treatment under the directive.
87.
Similar rulings based on principles of equity and good faith are to be found in the case law of national
courts6. I do not in fact think it is necessary to develop any new principle of Community law in order to
explain the result in Emmott. As I suggested in my opinion in Denkavit, the ruling can be seen as an
application,
24
albeit a new application, of the well-established principles laid down in Rewe-Zentralfinanz eG v
Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989 and subsequent cases7, in
particular the principle that the exercise of Community rights must not be rendered excessively or unduly
difficult. The ruling can be read as standing for the proposition that a member state cannot rely on a
limitation period where it is in default both in failing to implement a directive and in obstructing the
exercise of a judicial remedy in reliance upon it, or perhaps where the delay in exercising the remedy is in
some other way due to the conduct of the national authorities. In Emmott the member states default in
obstructing the remedy was compounded by Mrs Emmotts particularly unprotected position as an
individual dependent on social welfare.
6
See, by way of analogy, the judgment of the Baden-Wrttembergischer Verwaltungsgerichtshof of 21 October 1992
(Verwaltungsblatt fr Baden-Wrttemberg 1993, 220). There the court, referring to the principle of good faith applicable
in German administrative law, set aside the time limit laid down by national law for challenging an administrative
measure because the authority concerned led the individual to believe that it was reconsidering the matter. See also,
for English law, the judgment of the Court of Appeal in R v IRC, ex p Unilever plc [1996] STC 681 (where it was held that
it would be an abuse of power for the revenue authorities to rely on a two-year time limit for claiming losses when they
had not done so previously) and RSC Ord 53, which allow the three-month period for applications for judicial review to
be extended if there is good reason for the applicants delay.
7
See the judgments in Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989
(para 5), Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043 (para 13), Amministrazione delle
Finanze dello Stato v SpA San Giorgio Case 199/82 [1983] ECR 3595 (para 12), Emmott [1991] ECR I-4269 (para 16),
Francovich [1991] ECR I-5357 (para 43), Steenhorst-Neerings [1993] ECR I-5475 (para 15), Johnson [1995] All ER (EC)
258, [1994] ECR I-5483 (para 21), SCS Peterbroeck Van Campenhout & Cie v Belgium Case C-312/93 [1996] All ER
(EC) 242, [1995] ECR I-4599 (para 12), Van Schijndel v Stichting Pensioenfonds voor Fysiotherapeuten Joined cases
C-430431/93 [1996] All ER (EC) 259, [1995] ECR I-4705 (para 17) and FMC plc v Intervention Board for Agricultural
Produce Case C-212/94 [1996] ECR I-389 (para 71).

88.
 It seems to me that so understood the Emmott principle, although confined to very exceptional
circumstances, continues to provide an important safeguard notwithstanding the more recent
developments in the case law which I have discussed above. An individual must be allowed to make use
of all available remedies. The existence of another claim, for example a claim for damages in the
competent courts, cannot justify the obstruction of a repayment or entitlement claim which an individual
was seeking to exercise.

The present case

89.
With reference to the national courts seventh question, I therefore take the view that a member state
is entitled to rely upon a reasonable limitation period laid down by national law in order to resist claims
based on a Community directive notwithstanding the absence of proper implementation. The five-year
limitation period laid down by Danish law for challenges to decisions of the office seems wholly
reasonable and does not appear to make reliance upon the directive impossible or unduly difficult. Nor do
there appear to be any special circumstances in the present case such as those in Emmott.
90.
Apparently more objectionable however is the so-called forest fees principle referred to in the national
courts sixth question, a principle which was developed by the Hjesteret (the Danish Supreme Court) in
its judgment of 17 January 1899 and which takes its name from the fees in issue in the case. The
principle, as described in the order for reference (the parties disagree about its scope and effect), would
require the national court, in determining claims for
25
recovery of charges made without the requisite authority, to have regard to the fact that the charge
was made in pursuance of rules which had been in force over a long period without either the authorities
or other parties having been aware that the charge was unauthorised.
91.
In so far as the effect of the principle is to negate the otherwise reasonable period allowed by Danish
law for contesting charges, it renders the exercise of rights conferred by the directive virtually impossible
or excessively difficult contrary to the requirements laid down by the court in Rewe and subsequent
cases. The principle is analogous to the UK rule which the court held to be unlawful in FMC plc v
Intervention Board for Agricultura Produce Case C-212/94 [1996] ECR I-389. There the court held that a
rule by virtue of which a sum paid to a public authority under a mistake of law could be recovered only if it
was paid under protest was liable to prejudice effective protection of the rights conferred by Community
law (see [1996] ECR I-389 (para 72)).

QUESTION 8
92.
The final question is whether art 10(c), in conjunction with art 12(1)(e), of Directive 69/335 confers
rights on which individuals can rely before the national courts. That question must clearly be given an
affirmative answer, the provisions in question being unconditional and sufficiently precise. That answer is
implicit in the judgment in Ponente Carni8.
8
See my opinion in Ponente Carni [1993] ECR I-1915 (para 38), and on art 4(2)(b) of the directive, see Waldrich Siegen
Werkzeugmaschinen GmbH v Finanzamt Hagen Case C-38/88 [1990] ECR I-1447.

CONCLUSION
93.
Accordingly, I am of the opinion that the questions referred by the stre Landsret should be answered
as follows:
(1) The expression fees or dues in art 12(1)(e) of Council Directive (EEC) 69/335 concerning
indirect taxes on the raising of capital, means fees or dues charged to defray the cost of specific
services supplied to companies by the public authorities of a member state, including certain
mandatory services performed in the public interest. Such services include maintenance of a file for
the company in the statutory companies register and verification of compliance with filing and
disclosure requirements laid down by Community and national law. As regards more particularly
examination of accounts and bookkeeping, a member state is entitled to impose charges for
verifying that the accounts filed comply with statutory or stock exchange requirements but is not
entitled to charge for further work duplicating that performed by the statutory auditor. The services
for which a charge may be made do not include more general activities, such as preparatory legal
work in the field of company law.
(2) In fixing the fees or dues to be charged for such services the member state is entitled to take
into account all costs, including overheads, that are directly related to the services. The allocation
of costs should be made in accordance with the normal principles of commercial cost and
management accounting. In particular, where costs relate only partly to the services in question, a
reasonable apportionment must be made on the basis of suitable criteria. Costs relating to interest
and depreciation, official journeys and external dissemination of information may be taken into
account only in so far as they are directly related to the above-mentioned services. A member
26
 state is entitled to limit its charges to major transactions and pass on in those charges the cost
of comparatively minor services.
(3) A member state is entitled to fix standardised charges where individual costing of services is
not practicable. It must however periodically review its charges, whether fixed on a flat-rate basis or
containing a proportional element, in order to ensure that they do not exceed the average costs of
the services provided. A member state is not entitled to levy, in addition to a flat-rate basic charge,
a charge with no upper limit increasing in direct proportion to capital raised where that results in
total average charges which exceed the average cost of the services provided and in a
disproportionately high level of fees for certain companies.
(4) It is appropriate to leave it to the national court to arrive at the best estimate of any
repayment due in the light of the figures available to it. The national court could base its calculation
either on the actual cost to the registration authority of the specific services supplied to each
company or, if that is not possible, on the average cost of the services at or about the relevant time
or, if necessary, over a longer period. If the repayment is based on average costs, it may be
appropriate for the national court, if it is able to do so, to adjust the figures to take account of the
different costs of larger and smaller transactions.
(5) Community law precludes dismissal of an action for recovery of charges levied contrary to
Directive 69/335 on the ground that the charge was made in pursuance of rules which have been in
force over a long period without either the authorities or other parties having been aware that the
charge was unauthorised.
(6) Community law does not prevent a limitation period laid down by national law from beginning
to run before a directive has been properly implemented by a member state.
(7) Article 10(c), in conjunction with art 12(1)(e) of Directive 69/335, confers rights on which
individuals may rely before their national courts in the absence of proper implementation of the
directive.

2 December 1997.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 8 June 1995, received at the Court of Justice of the European Communities on 15 June
1995, the stre Landsret (the Eastern Regional Court) referred to the Court of Justice for a preliminary
ruling under art 177 of the EC Treaty eight questions on the interpretation of Council Directive (EEC)
69/335 concerning indirect taxes on the raising of capital, as amended (OJ S Edn 1969 (II) p 412).
2.
Those questions were raised in actions brought by Fantask A/S and a number of other companies or
groups of companies against the Industriministeriet (Erhvervsministeriet) (the Danish Ministry of Trade
and Industry) relating to charges levied on registration of new public and private limited companies and on
the capital of such companies being increased.
3.
Law No 468 of 29 September 1917, the First Law on Public Limited Companies (Lovtidende A 1917, p
1117), made it compulsory for public limited companies and increases in their capital to be entered in a
companies register. Entries in the register were subject to a charge at a rate to be determined by the
competent minister. Substantially recast for the first time in 1930, the Law was subject to general
amendment by Law No 370 of 13 June 1973 on Public Limited
27
Companies (Lovtidende A 1973, p 1025). On the same day Law No 371 on Private Limited Companies
(Lovtidende A 1973, p 1063) was adopted, which lays down, in relation to such companies, registration
formalities analogous to those applicable to public limited companies.
4.
Article 154(3) of the Law on Public Limited Companies and art 124(3) of the Law on Private Limited
Companies initially gave the competent minister the power to determine the rates of the registration
charges for those two categories of company.
5.
From the adoption of the First Law on Public Limited Companies until 1992, there was no change in
the charging structure for the registration of new companies and of increases in their capital. It consisted
of a fixed basic charge and a supplementary charge calculated in proportion to the nominal value of the
capital raised. The rates, on the other hand, were amended on several occasions.
6.
Between 1 January 1974 and 1 May 1992, the basic charge ranged from D Kr 500 to D Kr 1,700 for
the registration of new public and private limited companies and from D Kr 200 to D Kr 900 for the
registration of an increase in the capital of either category of company. During that period, the
supplementary charge was D Kr 4 per D Kr 1000 of the subscribed capital on registration of a new
company and the same percentage of the capital raised on registration of an increase in capital.
7.
 The registry of public limited companies set up by Law No 468 constituted a directorate of the Ministry
of Trade and was responsible for the registration of entries relating to public limited companies and, from
1974, to private limited companies. By Law No 851 of 23 December 1987 amending, in particular, the Law
on Public Limited Companies and the Law on Private Limited Companies (Lovtidende A 1987, p 3229),
the registry became the Erhvervs- og Selskabs- styrelsen (the Trade and Companies Office). Apart from
carrying out its registration duties and setting and collecting the related charges, the Trade and
Companies Office is involved in the drafting of legislation in the fields of company and business law and
ensures its application. It also performs various functions involving the provision of advice and
information.
8.
Following a report from the Danish Court of Auditors, which found that the Trade and Companies
Office had enjoyed significant surpluses of income over expenditure as a result of the levying of the
supplementary charge and questioned whether that charge was allowed under Danish law, the
supplementary charge was abolished by Order No 301 of 30 April 1992 (Lovtidende A 1992, p 1149) with
effect from 1 May 1992. At the same time the basic charge was increased to D Kr 2,500 for the
registration of a new public limited company and to D Kr 1,800 for that of a new private limited company.
The fee for registration of an increase in the capital of either category of company was raised to D Kr 600.
9.
Fantask and a number of other companies or groups of companies then asked the Trade and
Companies Office to refund the supplementary charges which they had been obliged to pay to that
directorate between 1983 and 1992. Only Fantask also claimed repayment of the basic charge.
10.
Since their requests for a refund were rejected, the companies in question commenced proceedings in
the stre Landsret against the Ministry of Industry. In their actions they submitted inter alia that, in the
light, in particular, of the judgment in Ponente Carni SpA v Amministrazione delle Finanze dello Stato
Joined cases C-71/91 and C-178/91 [1993] ECR I-1915, the supplementary chargeand
28
in Fantasks case the basic charge toowas contrary to arts 10 and 12 of the directive.
11.
In those circumstances the stre Landsret stayed proceedings and referred the following eight
questions to the Court of Justice for a preliminary ruling:
(1) Does Community law impose requirements upon the Member States delimitation of the
concept of fees or dues in Article 12(1)(e) of Directive 69/335/EEC or are the individual member
states free to decide what may be regarded as fees or dues for a specific service?
(2) May the basis for the calculation of duties charged under Article 12(1)(e) of Directive
69/335/EEC by a Member State for registration of formation or increase in capital of a public limited
company or a private limited company include the following types of costs or some of them: the
cost of salaries and pension contributions for officials not involved in effecting the registration, such
as the registration authoritys administrative staff or staff of the registration authority or other
authorities who are engaged on preparatory legal work in the field of company law; the cost of
effecting registration of other matters relating to companies, in respect of which the Member State
has determined that no specific consideration is to be paid; the cost of performing duties, other than
registration, required of the registration authority in pursuance of company legislation and
legislation related thereto, such as examination of companies accounts and supervision of
companies bookkeeping; payment of interest and depreciation of all capital costs which are
regarded by the registration authority as concerning the field of company law and related fields of
law; the cost of official journeys not connected with the specific work of registration; the cost of the
registration authoritys external dissemination of information and guidance not connected with the
specific work of registration, such as lecturing, preparation of articles and brochures and holding of
meetings with trade organizations and other interested groups.
(3)(a) Is Article 12(1)(e) of Directive 69/335/EEC to be interpreted as meaning that a Member
State is precluded from fixing standardized charges by rules valid without limitation of time? (b) If
that is not possible, is a Member State required to adjust its scale of charges every year or at other
fixed intervals? (c) Is it of any significance for the answer whether charges are fixed in proportion to
the amount of the capital to be raised, as notified for registration?
(4) Is Article 12(1)(e) in conjunction with Article 10(1) of Directive 69/335/EEC to be interpreted
as meaning that the amount charged as consideration for a specific servicesuch as, for example,
registration of the formation or increase in capital of a public limited company or a private limited
companyis to be calculated on the basis of the actual cost of the specific serviceregistration
or can the duty for the individual registration be fixed at, for example, a basic charge together with
D Kr 4 per D Kr 1 000 of the nominal value of the capital subscribed, so that the amount of the duty
is independent of the registration authoritys time used and other costs necessary for effecting the
registration?
(5) Is Article 12(1)(e) in conjunction with Article 10(1) of Directive 69/335/EEC to be interpreted
as meaning that the Member State in calculating any amount to be recovered must work on the
basis that the duty must reflect the cost of the specific service at the time at which the service is
29
 performed, or is the member state entitled to make a comprehensive assessment over a longer
period, for example an accounting year or within the period in which it will be possible under
national law to assert a claim for recovery?
(6) If national law contains a general principle that, in determining claims for recovery of charges
made without the requisite authority, importance should be attached to the fact that the charge was
made in pursuance of rules which have been in force over a long period without either the
authorities or other parties having been aware that the charge was unauthorized, will Community
law preclude dismissal on those grounds of an action for recovery of charges levied contrary to
Directive 69/335/EEC?
(7) Does Community law make it impossible under national law for the authorities of a Member
State, in cases of claims for recovery concerning charges made contrary to Directive 69/335/EEC,
to contend and establish that national limitation periods start to run from a time at which an unlawful
implementation of Directive 69/335/EEC occurred?
(8) Does Article 10(1) in conjunction with Article 12(1)(e) of Directive 69/335/EEC as interpreted
in the foregoing questions result in rights on which citizens in the individual Member States may
rely before the national courts?
12.
First, the objectives and the content of the directive, as set out in the judgment in Ponente Carni,
should be noted.
13.
As the recitals in its preamble indicate, the directive aims at encouraging the free movement of capital
which is regarded as essential for the creation of an economic union whose characteristics are similar to
those of a domestic market. As far as concerns taxes on the raising of capital, the pursuit of such an
objective presupposes the abolition of indirect taxes in force in the member states until then and imposing
in place of them a duty charged only once in the common market and at the same level in all the member
states.
14.
The directive thus provides for charging a capital duty on the raising of capital, which, according to the
sixth and seventh recitals in the preamble, should be harmonised with regard both to its structures and to
its rates, so as not to interfere with the movement of capital (see the judgment in Advokatrdet as
representative of P Conradsen A/S v Ministeriet for Skatter og Afgifter Case 161/78 [1979] ECR 2221
(para 11)). That capital duty is governed by arts 2 to 9 of the directive.
15.
Article 3 defines the capital companies to which the directive applies and they include, in particular,
public and private limited companies under Danish law.
16.
Articles 4, 8 and 9 list, subject to the provisions of art 7, the transactions subject to capital duty and
those which the member states may exempt. Under art 4(1)(a) and (c) transactions subject to capital duty
include the formation of a capital company and an increase in the capital of such a company by
contribution of assets of any kind.
17.
According to the last recital in its preamble, the directive also provides for the abolition of other indirect
taxes with the same characteristics as the capital duty or the stamp duty on securities, whose retention
might frustrate the purposes of the legislation. Those indirect taxes, the levying of which is prohibited, are
listed in arts 10 and 11 of the directive. Article 10 provides:
Apart from capital duty, Member States shall not charge, with regard to companies, firms,
associations or legal persons operating for profit, any taxes
30
whatsoever (c) in respect of registration or any other formality required before the
commencement of business to which a company, firm, association or legal person operating for
profit may be subject by reason of its legal form.
18.
Article 12(1) of the directive lays down an exhaustive list of taxes and duties other than capital duty
which, in derogation from arts 10 and 11, may be imposed on capital companies in connection with the
transactions referred to in those latter provisions (see, to that effect, Ministeriet for Skatter og Afgifter v
Investeringforeningen Dansk Sparinvest Case 36/86 [1988] ECR 409 (para 9)). Article 12(1)(e) of the
directive covers duties paid by way of fees or dues.

Questions 1 to 5
19.
In its first five questions, which should be answered together, the national court essentially asks
whether, on a sound construction of art 12(1)(e) of the directive, in order for charges levied on registration
of public and private limited companies and on their capital being increased to be by way of fees or dues,
their amount must be calculated solely on the basis of the cost of the formalities in question, or whether it
may be set so as to cover the whole or part of the costs of the authority responsible for registrations.
20.
 Since art 12 of the directive derogates, in particular, from the prohibitions laid down in art 10, it is
necessary to consider at the outset whether the charges at issue fall under any of those prohibitions.
21.
Article 10 of the directive, read in the light of the last recital in the preamble, prohibits in particular
indirect taxes with the same characteristics as capital duty. It thus applies, inter alia, to taxes in any form
which are payable in respect of the formation of a capital company or an increase in its capital (art 10(a)),
or in respect of registration or any other formality required before the commencement of business, to
which a company may be subject by reason of its legal form (art 10(c)). That latter prohibition is justified
by the fact that, even though the taxes in question are not imposed on capital contributions as such, they
are nevertheless imposed on account of formalities connected with the companys legal form, in other
words on account of the instrument employed for raising capital, so that their continued existence would
similarly risk frustrating the aims of the directive (see Denkavit Internationaal BV v Kamer van
Koophandel en Fabrieken voor Midden-Gelderland Case C-2/94 [1996] ECR I-2827 (para 23)).
22.
In this case, in so far as the basic charge and the supplementary charge are paid on the registration of
new public and private limited companies, they are directly referred to in the prohibition laid down by art
10(c) of the directive. A similar conclusion must also be reached where those charges are payable on the
registration of increases in the capital of such companies, since they too are imposed on account of an
essential formality connected with the legal form of the companies in question. While registration of an
increase in capital does not formally amount to a procedure which is required before a capital company
commences business, it is none the less necessary for the carrying on of that business.
23.
The Danish and Swedish governments maintain that the term duties paid by way of fees or dues in
art 12 of the directive also covers charges whose amount is calculated so as to offset not only the
registration costs directly at issue but also all the expenses of the charging authority which are linked, in
particular, to the drafting and application of legislation in the field of company law.
31
24.
The Danish government points out in particular that the directive did not harmonise the laws of the
member states concerning the duties paid by way of fees or dues referred to in art 12(1)(e) and that their
definition continues to be a matter for national law. However, the discretion granted to the member states
is not unlimited inasmuch as the assessment of the costs borne by the authority responsible for
registrations must, according to the judgment in Ponente Carni, be fixed in a reasonable manner. It
therefore considers that, unlike the position in that case, a member state may not, when calculating the
charges, take account of expenditure which has no link whatsoever with the administration of company
law.
25.
According to Fantask, the other applicants in the main proceedings which lodged observations and the
European Commission, it is, on the contrary, clear from Ponente Carni that the term duties paid by way of
fees or dues is one of Community law and that such charges must be calculated solely on the basis of
the cost of effecting the registration in respect of which they are paid. Thus, a charge set as a proportion
of the subscribed capital, such as the supplementary charge, cannot, by its very nature, fall within the
derogation provided for in art 12(1)(e) of the directive. While a member state is entitled to set charges
paid by way of fees or dues in advance, without limitation in time and on the basis of a flat-rate
assessment of the cost of effecting registrations, it must review them periodically, for example once a
year, so as to ensure that they continue not to exceed the costs incurred.
26.
It should be noted in that regard that the term duties paid by way of fees or dues is contained in a
provision of Community law which does not refer to the law of the member states in order to determine
the terms meaning and scope. Furthermore, the objectives of the directive would be undermined if the
member states were entirely free to retain taxes with the same characteristics as capital duty by
categorising them as duties paid by way of fees or dues. It follows that the interpretation of the term at
issue, considered in its entirety, cannot be left to the discretion of each member state (see Felicitas
Rickmers-Linie KG & Co v Finanzamt fr Verkehrsteuern, Hamburg Case 270/81 [1982] ECR 2771 (para
14)).
27.
Moreover, the court has already held, in its judgment in Ponente Carni [1993] ECR I-1915 (paras 41
42), that the distinction between taxes prohibited by art 10 of the directive and duties paid by way of fees
or dues implies that the latter cover only payments collected on registration whose amount is calculated
on the basis of the cost of the service rendered. A payment the amount of which had no link with the cost
of the particular service or was calculated not on the basis of the cost of the transaction for which it is
consideration but on the basis of all the running and capital costs of the department responsible for that
transaction would have to be regarded as a tax falling solely under the prohibition of art 10 of the
directive.
28.
It follows that charges levied on registration of public and private limited companies and on their
capital being increased cannot be by way of fees or dues within the meaning of art 12(1)(e) of the
directive if their amount is calculated so as to cover costs of the kind specified by the national court in the
first three indents of its second question. The costs in question are in fact unrelated to the registrations in
respect of which the contested charges are paid. However, for the reasons given by Advocate General
Jacobs in paras 37 and 45 of his opinion, above, a member state may impose charges for major
transactions only and pass on in those charges the costs of minor services performed without charge.
32
29.
As regards the setting of the amount of duties paid by way of fees or dues, the court stated in Ponente
Carni [1993] ECR I-1915 (para 43), that it may be difficult to determine the cost of certain transactions
such as the registration of a company. In such a case the assessment of the cost can only be on a flat-
rate basis and must be fixed in a reasonable manner, taking account, in particular, of the number and
qualification of the officials, the time taken by them and the various material costs necessary for carrying
out the transaction.
30.
It must be stated in that regard that, in calculating the amount of duties paid by way of fees or dues,
the member states are entitled to take account not only of the material and salary costs which are directly
related to the effecting of the registrations in respect of which they are incurred, but also, in the
circumstances indicated by Advocate General Jacobs in para 43 of his opinion, above, of the proportion of
the overheads of the competent authority which can be attributed to those registrations. To that extent
only, the costs specified by the national court in the first three indents of its second question may form
part of the basis for calculating the charges.
31.
Charges with no upper limit which increase directly in proportion to the nominal value of the capital
raised cannot, by their very nature, amount to duties paid by way of fees or dues within the meaning of
the directive. Even if there may be a link in some cases between the complexity of a registration and the
amount of capital raised, the amount of such charges will generally bear no relation to the costs actually
incurred by the authority on the registration formalities.
32.
Finally, as is evident from the judgment in Ponente Carni [1993] ECR I-1915 (para 43), the amount of
duties paid by way of fees or dues does not necessarily have to vary in accordance with the costs actually
incurred by the authority in effecting each registration and a member state is entitled to prescribe in
advance, on the basis of the projected average registration costs, standard charges for carrying out
registration formalities in relation to capital companies. Furthermore, there is nothing to prevent those
charges from being set for an indefinite period, provided that the member state checks at regular
intervals, for example once a year, that they continue not to exceed the registration costs.
33.
It is for the national court to review, on the basis of the above considerations, the extent to which the
charges at issue are paid by way of fees or dues and, where appropriate, to order a refund on that basis.
34.
The reply to the first five questions should therefore be that, on a sound construction of art 12(1)(e) of
the directive, in order for charges levied on registration of public and private limited companies and on
their capital being increased to be by way of fees or dues, their amount must be calculated solely on the
basis of the cost of the formalities in question. It may, however, also cover the costs of minor services
performed without charge. In calculating their amount, a member state is entitled to take account of all the
costs related to the effecting of registration, including the proportion of the overheads which may be
attributed thereto. Furthermore, a member state may impose flat-rate charges and fix their amount for an
indefinite period, provided that it checks at regular intervals that they continue not to exceed the average
cost of the registrations at issue.

Question 6
35.
By its sixth question, the national court seeks to ascertain whether Community law precludes actions
for the recovery of charges levied in breach of the directive from being dismissed on the ground that those
charges were
33
imposed as a result of an excusable error by the authorities of the member state inasmuch as they
were levied over a long period without either those authorities or the persons liable to them having been
aware that they were unlawful.
36.
It is settled case law that the interpretation which, in the exercise of the jurisdiction conferred upon it
by art 177 of the Treaty, the Court of Justice gives to a rule of Community law clarifies and defines where
necessary the meaning and scope of that rule as it must be or ought to have been understood and
applied from the time of its entry into force.
37.
It follows that the rule as so interpreted may, and must, be applied by the courts to legal relationships
arising and established before the judgment ruling on the request for interpretation, provided that in other
respects the conditions enabling an action relating to the application of that rule to be brought before the
courts having jurisdiction are satisfied (see the judgments in Amministrazione delle Finanze dello Stato v
Denkavit Italiana Srl Case 61/79 [1980] ECR 1205 (para 16) and Bautiaa Srl v Directeur des Services
Fiscaux, Socit Franaise Maritime SA v Directeur des Services Fiscaux Joined cases C-197/94 and C-
252/94 [1996] ECR I-505 (para 47)).
38.
 It is also settled case law that entitlement to the recovery of sums levied by a member state in breach
of Community law is a consequence of, and an adjunct to, the rights conferred on individuals by the
Community provisions as interpreted by the court (see the judgment in Amministrazione delle Finanze
dello Stato v SpA San Giorgio Case 199/82 [1983] ECR 3595 (para 12)). The member state is therefore in
principle required to repay charges levied in breach of Community law (see the judgment in Socit
Comateb v Directeur Gnral des Douanes et Droits Indirects Joined cases C-192218/95 [1997] ECR I-
165 (para 20)).
39.
Accordingly, while the recovery of such charges may, in the absence of Community rules governing
the matter, be sought only under the substantive and procedural conditions laid down by the national law
of the member states, those conditions must nevertheless be no less favourable than those governing
similar domestic claims nor render virtually impossible or excessively difficult the exercise of rights
conferred by Community law (see eg the judgment in SCS Peterbroeck Van Campenhout & Cie v
Belgium Case C-312/93 [1996] All ER (EC) 242, [1995] ECR I-4599 (para 12)).
40.
A general principle of national law under which the courts of a member state should dismiss claims for
the recovery of charges levied over a long period in breach of Community law without either the
authorities of that state or the persons liable to pay the charges having been aware that they were
unlawful, does not satisfy the above conditions. Application of such a principle in the circumstances
described would make it excessively difficult to obtain recovery of charges which are contrary to
Community law. It would, moreover, have the effect of encouraging infringements of Community law
which have been committed over a long period.
41.
The reply to the sixth question should therefore be that Community law precludes actions for the
recovery of charges levied in breach of the directive from being dismissed on the ground that those
charges were imposed as a result of an excusable error by the authorities of the member state inasmuch
as they were levied over a long period without either those authorities or the persons liable to them having
been aware that they were unlawful.
34

Question 7
42.
By its seventh question, the national court essentially asks whether Community law prevents a
member state from relying on a limitation period under national law to resist actions for the recovery of
charges levied in breach of the directive as long as that member state has not properly transposed the
directive.
43.
It is clear from the order for reference that under Danish law the right to recovery of a whole range of
debts becomes statute-barred after five years and that that period generally runs from the date on which
the debt became payable. On the expiry of that period the debt is normally no longer exigible, unless the
debtor has in the meantime acknowledged the debt or the creditor has commenced legal proceedings.
44.
When a number of the applicants in the main proceedings brought their applications for repayment,
the relevant time limit for at least some of their claims had expired.
45.
The applicants and the Commission consider, on the basis of Emmott v Minister for Social Welfare
Case C-208/90 [1991] ECR I-4269, that a member state may not rely on a limitation period under national
law as long as the directive, in breach of which charges have been wrongly levied, has not been properly
transposed into national law. According to them, until that date individuals are unable to ascertain the full
extent of their rights under the directive. A limitation period under national law thus does not begin to run
until the directive has been properly transposed.
46.
The Danish, French and UK governments consider that a member state is entitled to rely on a
limitation period under national law such as the period at issue, since it complies with the two conditions,
of equivalence and of effectiveness, laid down by the courts case law (see esp the judgments in San
Giorgio and Peterbroeck). In their view, the judgment in Emmott must be confined to the quite particular
circumstances of that case, as the court has, moreover, confirmed in its subsequent case law.
47.
As the court has pointed out in para 39 of this judgment, it is settled case law that, in the absence of
Community rules governing the matter, it is for the domestic legal system of each member state to lay
down the detailed procedural rules for actions seeking the recovery of sums wrongly paid, provided that
those rules are not less favourable than those governing similar domestic actions and do not render
virtually impossible or excessively difficult the exercise of rights conferred by Community law.
48.
The court has thus acknowledged, in the interests of legal certainty which protects both the taxpayer
and the authority concerned, that the setting of reasonable limitation periods for bringing proceedings is
compatible with Community law. Such periods cannot be regarded as rendering virtually impossible or
excessively difficult the exercise of rights conferred by Community law, even if the expiry of those periods
necessarily entails the dismissal, in whole or in part, of the action brought (see esp the judgments in
Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989 (para
5), Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043 (paras 1718) and
Palmisani v Istituto Nazionale della Previdenza Sociale (INPS) Case C-261/95 (1997) ECJ Transcript, 10
July 1997 (para 28)).
49.
The five-year limitation period under Danish law must be considered to be reasonable (see the
judgment in Haahr Petroleum Ltd v benr Havn Case C-90/94 (1997) ECJ Transcript, 17 July (para 49)).
Furthermore, it is apparent that that
35
period applies without distinction to actions based on Community law and those based on national law.
50.
It is true that the court held in Emmott [1991] ECR I-4269 (para 23), that until such time as a directive
has been properly transposed, a defaulting member state may not rely on an individuals delay in initiating
proceedings against it in order to protect rights conferred upon him by the provisions of the directive and
that a period laid down by national law within which proceedings must be initiated cannot begin to run
before that time.
51.
However, as was confirmed by the judgment in Johnson v Chief Adjudication Officer (No 2) Case
410/92 [1995] All ER (EC) 258, [1994] ECR I-5483 (para 26), it is clear from Steenhorst-Neerings v
Bestuur van de Bedrijfsvereniging voor Detailhandel, Ambachten en Huisvrouwen Case C-338/91 [1993]
ECR I-5475 that the solution adopted in Emmott was justified by the particular circumstances of that case,
in which the time-bar had the result of depriving the applicant of any opportunity whatever to rely on her
right to equal treatment under a Community directive (see also the judgments in Haahr Petroleum (1997)
ECJ Transcript, 17 July (para 52) and Texaco A/S v Middelfart Havn, Olieselskabet Danmark amba v
Trafikministeriet Joined cases C-114115/95 (1997) ECJ Transcript, 7 July 1997 (para 48)).
52.
The reply to the seventh question must therefore be that Community law, as it now stands, does not
prevent a member state which has not properly transposed the directive from resisting actions for the
repayment of charges levied in breach thereof by relying on a limitation period under national law which
runs from the date on which the charges in question became payable, provided that such a period is not
less favourable for actions based on Community law than for actions based on national law and does not
render virtually impossible or excessively difficult the exercise of rights conferred by Community law.

Question 8
53.
By its eighth question, the national court asks whether art 10 of the directive in conjunction with art
12(1)(e) thereof gives rise to rights on which individuals may rely before national courts.
54.
It is settled case law that where the provisions of a directive appear, as far as their subject matter is
concerned, to be unconditional and sufficiently precise, those provisions may be relied upon in national
courts by individuals against the state where the state fails to implement the directive in national law by
the end of the period prescribed or where it fails to implement the directive correctly (see esp the
judgment in Comitato di Coordinamento per la Difesa della Cava v Regione Lombardia Case C-236/92
[1994] ECR I-483 (para 8)).
55.
In this case, it is sufficient to observe that the prohibition laid down in art 10 of the directive and the
derogation from that prohibition in art 12(1)(e) are expressed in sufficiently precise and unconditional
terms to be invoked by individuals in their national courts in order to contest a provision of national law
which infringes the directive.
56.
The reply to the eighth question must therefore be that art 10 of the directive in conjunction with art
12(1)(e) thereof gives rise to rights on which individuals may rely before national courts.
36

Costs
57.
The costs incurred by the Danish, French, Italian, Swedish and UK governments and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the proceedings pending before the
national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the stre Landsret
by order of 8 June 1995, hereby rules: (1) On a sound construction of art 12(1)(e) of Council Directive
(EEC) 69/335 concerning indirect taxes on the raising of capital, as amended, in order for charges levied
on registration of public and private limited companies and on their capital being increased to be by way
of fees or dues, their amount must be calculated solely on the basis of the cost of the formalities in
question. It may, however, also cover the costs of minor services performed without charge. In calculating
their amount, a member state is entitled to take account of all the costs related to the effecting of
registration, including the proportion of the overheads which may be attributed thereto. Furthermore, a
member state may impose flat-rate charges and fix their amount for an indefinite period, provided that it
checks at regular intervals that they continue not to exceed the average cost of the registrations at issue.
(2) Community law precludes actions for the recovery of charges levied in breach of Directive 69/335,
from being dismissed on the ground that those charges were imposed as a result of an excusable error by
the authorities of the member state inasmuch as they were levied over a long period without either those
authorities or the persons liable to them having been aware that they were unlawful. (3) Community law,
as it now stands, does not prevent a member state which has not properly transposed Directive 69/335,
from resisting actions for the repayment of charges levied in breach thereof by relying on a limitation
period under national law which runs from the date on which the charges in question became payable,
provided that such a period is not less favourable for actions based on Community law than for actions
based on national law and does not render virtually impossible or excessively difficult the exercise of
rights conferred by Community law. (4) Article 10 of Directive 69/335, in conjunction with art 12(1)(e) gives
rise to rights on which individuals may rely before national courts.

37

[1998] All ER (EC) 38

Magorrian and another v Eastern Health and Social Services Board

and another
(Case C-246/96)

EUROPEAN COMMUNITY; Social Policy: EMPLOYMENT; Discrimination: PENSIONS

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES SCHINTGEN (PRESIDENT OF THE SECOND CHAMBER, ACTING FOR THE PRESIDENT
OF THE SIXTH CHAMBER), MANCINI (RAPPORTEUR), KAPTEYN, MURRAYAND HIRSCHADVOCATE
GENERAL COSMAS
5 JUNE, 10 JULY, 11 DECEMBER 1997
European Community Equality of treatment for men and women Equal pay for equal work Pension
Voluntary contracted-out occupational pension scheme Exclusion of part-time workers from status
conferring entitlement to additional pension benefits Female part-time workers claiming indirect
discrimination on grounds of sex National rules restricting successful claims for retroactive award of
benefit to two years benefits Date from which additional benefits to be calculated Whether restriction
compatible with Community law Occupational Pension Schemes (Equal Access to Membership)
Regulations (Northern Ireland) 1976, reg 12 EC Treaty, art 119, Protocol No 2.

The applicants were female psychiatric nurses employed by a public-sector health board in Northern
Ireland who contributed to a voluntary contracted-out pension scheme. They started their careers working
full-time with the status of mental health officers, but when their family responsibilities increased they
began working part-time with the result that they lost that status along with the entitlement to additional
pension benefits which it conferred. Shortly before retiring, both applicants brought proceedings against
their employer, seeking equal treatment with persons having mental health officer status in the calculation
of their future pensions, in accordance with art 119 1 of the EC Treaty. They contended that they were
entitled to the additional benefits, the calculation of which should be based on their length of service since
8 April 1976, the date on which art 119 was first ruled to have direct effect, and that the two-year
restriction on entitlement imposed by reg 122 of the Occupational Pension Schemes (Equal Access to
Membership) Regulations (Northern Ireland) 1976 was incompatible with Community law. The national
court found that the exclusion of part-time nurses from mental health officer status constituted indirect
discrimination on grounds of sex, since considerably more women than men were affected, and that the
discrimination was unjustified. The court stayed the proceedings and referred to the Court of Justice of
the European Communities for a preliminary ruling questions concerning: (i) the date from which the
periods of service of part-time workers who had suffered indirect discrimination based on sex should be
taken into account for the purpose of calculating the additional benefits to which they were entitled; and
(ii) whether Community law precluded the application of a national rule to a successful art 119 claim under
which 38 entitlement was limited to a period which started to run two years prior to commencement of
proceedings.
1
Article 119, so far as material, is set out at p 41 e f, post
2
Regulation 12, so far as material, is set out at p 54 d, post

Held (1) The direct effect of art 119 could be relied on as from 8 April 1976 in order retroactively to claim
equal treatment in relation to the right to join an occupational pension scheme. Moreover, entitlement to a
retirement pension under an occupational scheme was indissolubly linked to the right to join such a
scheme, and the same was true where the discrimination suffered by part-time workers stemmed from
discrimination concerning access to a special scheme which conferred entitlement to additional benefits.
It followed that periods of service completed by part-time workers who had suffered indirect discrimination
based on sex had to be taken into account as from 8 April 1976 for the purposes of calculating the
additional pension benefits to which they were entitled (see p 58 c to h and p 60 h, post); Defrenne v
Sabena Case 43/75 [1981] 1 All ER 122 and Dietz v Stichting Thuiszorg Rotterdam Case C-435/93 [1996]
ECR I-5223 applied.
 (2) The application in proceedings concerning access to membership of occupational pension
schemes of a national rule whereby the right to be admitted to a scheme, in the event of a successful art
119 claim, was linked to a period which started to run from a point in time two years prior to the institution
of proceedings would, in the instant case, prevent the applicants entire record of service after 8 April
1976 until 1990 from being taken into account for the purposes of calculating the additional benefits which
would be payable even after the date of the claim. Such a rule would render any action by individuals
relying on Community law impossible in practice. Moreover, the effect of that rule would be to limit in time
the direct effect of art 119 in cases in which no such limitation had been laid down either in the courts
case law or in EC Treaty Protocol No 2c. It followed that Community law precluded such a rule (see p 59
a to c g h and p 60 a b d h, post); Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland
Case 33/76 [1976] ECR 1989 applied; Steenhorst-Neerings v Bestuur van de Bedrijfsvereniging voor
Detailhandel, Ambachten en Huisvrouwen Case C-338/91 [1993] ECR I-5475 and Johnson v Chief
Adjudication Officer (No 2) Case 410/92 [1995] All ER (EC) 258 distinguished.
3
Protocol No 2, so far as material, is set out at p 57 h, post

Notes
For the principle of equal pay for equal work, see 52 Halsburys Laws (4th edn) paras 21112116. For
relevant cases see 21 Digest (2nd Reissue) 333, 342, 977, 989.
For the EC Treaty, art 119, see 50 Halsburys Statutes (4th edn) 306. For EC Treaty Protocol No 2
concerning art 119 (inserted by the Treaty on European Union), see ibid, Current Service, 142.

Cases cited
Barber v Guardian Royal Exchange Assurance Group Case C-262/88 [1990] 2 All ER 660, [1991] QB
344, [1991] 2 WLR 72, [1990] ECR I-1889, ECJ.
Bestuur van het Algemeen Burgerlijk Pensioenfonds v Beune Case C-7/93 [1995] All ER (EC) 97, [1994]
ECR I-4471, ECJ.
Bilka-Kaufhaus GmbH v Weber von Hartz Case 170/84 [1986] ECR 1607.
Coloroll Pension Trustees Ltd v Russell Case C-200/91 [1995] All ER (EC) 23, [1994] ECR I-4389, ECJ.
39
Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043.
Defrenne v Sabena Case 43/75 [1981] 1 All ER 122, [1976] ECR 455, ECJ.
Denkavit Internationaal BV v Kamer van Koophandel en Fabrieken voor Midden- Gelderland Case C-2/94
[1996] ECR I-2827.
Dietz v Stichting Thuiszorg Rotterdam Case C-435/93 [1996] ECR I-5223.
Draehmpaehl v Urania Immobilienservice ohG Case C-180/95 [1997] All ER (EC) 719, ECJ.
Emmott v Minister for Social Welfare Case C-208/90 [1991] ECR I-4269.
Fisscher v Voorhuis Hengelo BV Case C-128/93 [1995] All ER (EC) 193, [1994] ECR I-4583, ECJ.
Jenkins v Kingsgate (Clothing Productions) Ltd Case 96/80 [1981] 1 WLR 972, [1981] ECR 911, ECJ.
Johnson v Chief Adjudication Officer (No 2) Case 410/92 [1995] All ER (EC) 258, [1994] ECR I-5483,
ECJ.
Marshall v Southampton and South West Hampshire Area Health Authority (No 2) Case C-271/91 [1993]
4 All ER 586, [1994] QB 126, [1993] 3 WLR 1054, [1993] ECR I-4367, ECJ.
Moroni v Firma Collo GmbH Case C-110/91 [1993] ECR I-6591.
Neath v Hugh Steeper Ltd Case C-152/91 [1994] 1 All ER 929, [1993] ECR I-6935, ECJ.
Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989.
Rinner-Khn v FWW Spezial Gebudereinigung GmbH Case 171/88 [1989] ECR 2743.
Steenhorst-Neerings v Bestuur van de Bedrijfsvereniging voor Detailhandel, Ambachten en Huisvrouwen
Case C-338/91 [1993] ECR I-5475.
Ten Oever v Stichting Bedrijfspensioenfonds voor het Glazenwassersen Schoonmaakbedrijf Case C-
109/91 [1993] ECR I-4879.
Vroege v NCIV Instituut voor Volkshuisvesting BV Case C-57/93 [1995] All ER (EC) 193, [1994] ECR I-
4551, ECJ.

Reference
By order of 9 July 1996, the Office of the Industrial Tribunals and the Fair Employment Tribunal, Belfast,
referred to the Court of Justice of the European Communities for a preliminary ruling under art 177 of the
EC Treaty a number of questions (set out at p 56 b to g, post) on the interpretation of art 119 of the EC
Treaty and of Protocol No 2, concerning that provision. Those questions were raised in proceedings
between Mrs Magorrian and Mrs Cunningham and the Eastern Health and Social Services Board and the
Department of Health and Social Security in relation to certain additional benefits under a contracted-out
retirement pension scheme. Written observations were submitted on behalf of: Mrs Magorrian and Mrs
Cunningham, by J OHara, Barrister-at-law, and E McCaffrey, Solicitor; the UK government, by S Ridley,
of the Treasury Solicitors Department, acting as agent, and by R Weatherup QC, and N Paines, Barrister;
and the European Commission, by C Bury, M Wolfcarius and C Docksey, of its Legal Service, acting as
agents. Oral observations were made by Mrs Magorrian and Mrs Cunningham, the UK government and
the Commission. The language of the case was English. The facts of the case are set out in the opinion of
the Advocate General.

40
10 July 1997.

The Advocate General (C Cosmas)

delivered the following opinion (translated from the Greek).
1.
In the present case the Court of Justice of the European Communities is requested, by order for
reference from the Office of the Industrial Tribunals and the Fair Employment Tribunal, Belfast, Northern
Ireland, to reply to two questions submitted for a preliminary ruling which the Industrial Tribunal considers
are necessary in order to resolve the dispute pending before it.
2.
Those questions involve the interpretation of art 119 of the EC Treaty and of EC Treaty Protocol No 2
concerning art 119. More specifically, they concern, on the one hand, determination of the date as from
which account is to be taken of the period of service of two nurses who, in the national courts view, were
subjected to unfair discrimination on the ground of their sex, for the purposes of calculating certain
additional retirement benefits. On the other hand, the questions concern the compatibility with Community
law of certain national provisions which, if the action is successful, limit the years of service which may
retrospectively be taken into account to two years.

I LEGAL FRAMEWORK

A. Community legislation and case law

3.
Article 119 of the EC Treaty enshrines the principle of equal pay for men and women for equal work. It
provides as follows:
Each Member State shall during the first stage ensure and subsequently maintain the
application of the principle that men and women should receive equal pay for equal work.
For the purpose of this Article, pay means the ordinary basic or minimum wage or salary and
any other consideration, whether in cash or in kind, which the worker receives, directly or indirectly,
in respect of his employment from his employer.
Equal pay without discrimination based on sex means: (a) that pay for the same work at piece
rates shall be calculated on the basis of the same unit of measurement; (b) that pay for work at
time rates shall be the same for the same job.
4.
In Defrenne v Sabena Case 43/75 [1981] 1 All ER 122, [1976] ECR 455 (point 5, operative part) the
court held that the direct effect of art 119 may be relied on, that is to say the article in question produced
direct effect from the date of judgment in that case, that is as from 8 April 1976. In Bilka-Kaufhaus GmbH
v Weber von Hartz Case 170/84 [1986] ECR 1607 (confirmed in Barber v Guardian Royal Exchange
Assurance Group Case C-262/88 [1990] 2 All ER 660, [1990] ECR I-1889 and in Ten Oever v Stichting
Bedrijfspensioenfonds voor het Glazenwassersen Schoonmaakbedrijf Case C-109/91 [1993] ECR I-4879)
the court held that art 119 covered both the right to belong to occupational pension schemes and the right
to receive benefits under those schemes.
5.
In its judgment of 17 May 1990 in Barber [1990] 2 All ER 660, [1990] ECR I-1889 (point 5, operative
part), the court limited in time the effects of its judgment, by holding that
The direct effect of art 119 of the Treaty may not be relied upon in order to claim entitlement to
a pension, with effect from a date prior to that of this judgment, except in the case of workers or
those claiming under them, who 41 have before that date initiated legal proceedings or raised an
equivalent claim under the applicable national law.
In further elucidation of that limitation, the court held in Ten Oever [1993] ECR I-4879 (point 2, operative
part) that
By virtue of the judgment of 17 May 1990 in Barber the direct effect of Article 119 of the
EEC Treaty may be relied upon, for the purpose of claiming equal treatment in the matter of
occupational pensions, only in relation to benefits payable in respect of periods of employment
subsequent to 17 May 1990, subject to the exception in favour of workers or those claiming under
them who have, before that date, initiated legal proceedings or raised an equivalent claim under the
applicable national law.4

4
That decision has since been reaffirmed in the judgments in Moroni v Firma Collo GmbH Case C-110/91 [1993] ECR I-
6591 (point 3, operative part) and Neath v Hugh Steeper Ltd Case C-152/91 [1994] 1 All ER 929, [1993] ECR I-
6935 (point 1, operative part).
6.
Elucidation of the temporal effects of the direct effect of art 119 in the field of benefits granted under
occupational pension schemes was also provided by Protocol No 2, incorporated in the EC Treaty with
effect from 1 November 1993. Protocol No 2 is as follows:
For the purposes of Article 119 of this Treaty, benefits under occupational social security
schemes shall not be considered as remuneration if and in so far as they are attributable to periods
of employment prior to 17 May 1990, except in the case of workers or those claiming under them
who have before that date initiated legal proceedings or introduced an equivalent claim under the
applicable national law.
7.
Thus, Protocol No 2 essentially adopted the same interpretation as the judgment in Barber. Indeed, as
regards its content, the court has repeatedly held:
It is clear that the Protocol is linked to the Barber judgment, since it refers to the date of that
judgment, 17 May 1990. That judgment declares unlawful discrimination as between men and
women resulting from an age condition that varies according to sex for the purposes of entitlement
to a retirement pension following dismissal for economic reasons. There have been divergent
interpretations of the Barber judgment which limits, with effect from the date of the judgment,
namely 17 May 1990, the effect of its interpretation of art 119 of the Treaty. Those divergences
were removed by the judgment in Ten Oever, which was delivered before the entry into force of the
[Treaty on European Union (the TEU) (Maastricht, 7 February 1992; TS 12 (1994); Cmnd 2485; OJ
1992 C191)]. While extending it to all benefits payable under social security schemes and
incorporating it in the Treaty, the Protocol essentially adopted the same interpretation of the Barber
judgment as did the Ten Oever judgment. It did not, on the other hand, any more than the Barber
judgment, deal with or make any provision for, the conditions of membership of such occupational
schemes. (See the judgment in Beune [1995] All ER (EC) 97, [1994] ECR I-4471 (para 61).)
And it was made clear that that continued to be governed by the judgment in Bilka, which does not limit in
time the validity of the interpretation which it gives to art 119 of the Treaty 5.

5
See the three judgments in Bestuur van het Algemeen Burgerlijk Pensioenfonds v Beune Case C-7/93 [1995] All ER
(EC) 97, [1994] ECR I-4471 (paras 6162), Vroege v NCIV Instituut voor Volkshuisvesting BV Case C-57/93 [1995]
All ER (EC) 193, [1994] ECR I-4551 (paras 4142) and Fisscher v Voorhuis Hengelo BV Case C-128/93 [1995] All
ER (EC), [1994] ECR I-4583 (paras 4950).

42

B. National legislation
8.
Regulation 3 of the Health and Personal Social Services (Superannuation) Regulations (Northern
Ireland) 1984, SI 1984/3366 defines a mental health officer (MHO) as a whole-time officer on the medical
or nursing staff of a hospital used wholly or partly for the treatment of persons suffering from mental
disorder who devotes the whole or substantially the whole of his time to the treatment of such persons 7.
6
The 1984 regulations were issued on 18 September 1984 and their commencement date was 29 October 1984.
7
Moreover, the same provision of reg 3 of the 1984 regulations defines an MHO as a specialist who devotes nearly all his
time to the cure and care of mentally disturbed persons and, if the ministry so approves in a given case, any other
health officer who, having continued as an MHO without interruption in service and without acquiring the right to any
benefit under the 1984 regulations, subsequently works reduced hours in any of the above occupations, and such other
category of health workers employed in hospitals as above as the ministry may determine.

9.
As provided for in reg 50(2) of the 1984 regulations, any person who has worked as an MHO for a total
period of at least 20 years (a) obtains the right to retire at the age of 55 years instead of 60 years and (b)
every year of service as an MHO after the age of 50 years, or after completion of 20 years service as an
MHO, whichever shall be the later, is reckoned at twice its length (double-time service).
10.
Under s 2(4) of the Equal Pay Act (Northern Ireland) 1970, claims concerning the application of the
equal pay clause in connection with the employment of a woman must be brought before the competent
industrial tribunal within six months of the end of the relevant period of employment. Section 2(5) of the
1970 Act provides that in proceedings brought in respect of a failure to comply with an equal pay clause a
woman shall not be entitled to be awarded any payment by way of arrears of remuneration or damages in
respect of a time earlier than two years before the date on which the proceedings were instituted.
11.
Regulation 12 of the Occupational Pension Schemes (Equal Access to Membership) Regulations
(Northern Ireland) 1976, SI 1976/2388, amending the 1970 Act9, provides that in proceedings concerning
access to membership of occupational pension schemes the right to be admitted to the scheme, in
accordance with equal access requirements, is to have effect from a date no earlier than two years before
the institution of proceedings.
8
The 1976 regulations were issued on 9 August 1976 and their commencement date was 6 April 1978.
9
Regulation 12 concerns the power of the ordinary courts or of the Industrial Tribunal to uphold the fight to be admitted to
the occupational pension scheme and to require the employer to make additional contributions.
II FACTS
12.
Mrs Magorrian and Mrs Cunningham, the applicants in the main proceedings, were employed as
qualified nurses in the mental health sector by a public sector health board. They began their careers
working full-time with MHO status. When their family responsibilities increased, they began working part-
time10 and lost that status.
10
As is apparent from the order for reference, when the applicants in the main proceedings began working part-time, they
worked for 27 hours and 25 minutes per week. On reorganisation of the shift patterns in 1981, the part-time hours were
increased by the employer to 31 hours and five minutes per week. At the same time full-time nursing officers hours were
reduced from 40 hours to 37 hours and 30 minutes per week. Each applicant worked night shifts, which was better
suited to their domestic arrangements. Both applicants had been in charge of wards and of full-time nurses who had
MHO status.

13.
Both of the applicants in the main proceedings were affiliated to the Health and Personal Social
Services Scheme, a voluntary contracted-out scheme to 43 which both employer and employee
contribute. That scheme was open to part-time workers working a specified number of hours.
14.
The applicants in the main proceedings paid contributions to the scheme in order to qualify for
benefits, including a lump-sum payment on retirement and monthly payments thereafter.
15.
Mrs Magorrian retired on 18 October 1992, having completed 9 years and 111 days of full-time work as
an MHO between 1951 and 1963, and the equivalent of 11 years and 25 days of full-time service as a
part-time worker between 1979 and 1992. She also worked part-time between 1969 and 1979 but for
hours not reckonable for pension purposes.
16.
Mrs Cunningham retired in April 1994, having completed 15 years and 175 days of full-time service as
an MHO between 1956 and 1974, and the equivalent of 11 years and 105 days full-time service as a part-
time worker between 1980 and 1994. Between 1974 and 1980 she also worked part-time on a non-
superannuable basis and she elected not to make pension contributions during that period.
17.
When the applicants in the main proceedings retired, they received the lump sums to which they were
entitled, together with their basic retirement pensions, but not the additional benefits to which they would
have been entitled, if they had had MHO status at the time of their retirement.
18.
The applicants in the main proceedings emphasise that they had already initiated proceedings before
ceasing to be employed. Their claims against the Eastern Health and Social Services Board and the
Department of Health and Social Services (respondents in the main proceedings) were lodged on 22
September 1992. Mrs Magorrian retired on 18 October 1992, whilst Mrs Cunningham retired on 31 March
1994.
19.
The applicants in the main proceedings sought equal treatment with persons having MHO status in the
calculation of the amounts of their future pensions. Specifically, they argued that they were entitled to the
additional benefits and that the calculation thereof should be based on their length of service starting
either on 8 April 1976, the date of the judgment in Defrenne, or on 13 May 1986, the date of the judgment
in Bilka. Finally, it is contended on their behalf, there is no justification for applying the two-year restriction
laid down in the 1970 Act in calculating their length of service, or for applying the cut-off date of 17 May
1990, the date of the judgment in Barber since to do so would be to deprive them of an effective legal
remedy.
20.
The respondents in the main proceedings acknowledged a limited right on the part of the applicants to
the additional benefits, in accordance with the provisions concerning MHO status. In their view, it would
be necessary to take account of periods of service subsequent either to 22 September 1990, that is to say
two years prior to lodgement of their claim, as provided for by the 1970 Act, or to 17 May 1990, the date of
the judgment in Barber and the date referred to in Protocol No 2, until the date when they retired from the
service.
21.
The national court, in its order for reference dated 12 September 1995, found that it was common
ground between the parties to the main proceedings 44 that the payment of such benefits constitutes
pay within the meaning of art 119 of the EC Treaty and of Council Directive (EEC) 75/117 on the
approximation of the laws of the member states relating to the application of the principle of equal pay for
men and women (OJ 1975 L45 p 19). The national court also found that a considerably smaller proportion
of women than men employed in the mental health sector in Northern Ireland could comply with the
requirement of full-time working as MHOs, and accepted that the exclusion of the applicants in the main
proceedings from MHO status discriminates against them on the ground of sex and is not justified.

III THE PRELIMINARY QUESTIONS

22.
 In the context of the proceedings pending before it, the Industrial Tribunal referred the following
questions to the Court of Justice for a preliminary ruling:
In circumstances where: (a) a worker has been employed by a Health Board which is part of the
State, in employment concerned with the care of the mentally ill to which an occupational pension
scheme applies; (b) the worker has at all material times either been a member or been eligible to
be a member of the pension scheme; (c) the pension scheme contains a term according to which
those who work full-time and devote all or substantially all their working hours to the care of the
mentally ill (who are described as Mental Health Officers) are entitled to additional benefits not
available to those doing the same work part-time, as follows: where a person has reached or
passed the age of 50 and has worked as a Mental Health Officer for 20 years (here referred to as
the qualifying service) and continues to work as a Mental Health Officer, then (i) their subsequent
service is reckoned for pension purposes at twice its length (here referred to as the double-time
service); and (ii) they have the right to a pension at the age of 55 instead of the normal age of 60;
(d) the worker is deprived of the status of Mental Health Officer and the additional benefits attached
thereto solely on the ground that her employment was part-time; (e) the national tribunal has held
that the provisions described at (c) and (d) constitute discrimination on grounds of sex against
women working part-time in the care of the mentally ill; (f) the pension which the worker receives
and the additional benefits which they claim, are only payable to them as and from their respective
retirements in 1992 and 1994, after their claims have been brought before the national tribunal; and
(g) the calculation of the additional benefits from their respective retirement dates in 1992/94 would
involve counting their years of service prior to 1992,
Question 1: From what date should the service of the workers be counted for the purpose of
calculating the additional benefits to which they are entitled: (i) 8 April 1976, (ii) 7 May 1990, (iii)
some other and if so, what date?
Question 2: Where the relevant national legislation restricts back-dating entitlement in the event
of a successful claim to a period of 2 years prior to the date on which the claim was made, does
this amount to the denial of an effective remedy under Community law and is the Industrial Tribunal
obliged to disregard such provision in domestic law if it feels it necessary to do so?
45

IV REPLIES TO THE PRELIMINARY QUESTIONS

A. Reply to the first preliminary question

23.
The first question concerns the determination of the date from the point of view of Community law for
the purposes of calculating the additional retirement benefits to which Mrs Magorrian and Mrs
Cunningham are entitled, who have suffered discrimination on the ground of sex. The national court is
thereby essentially asking whether the temporal limitation of the effect of the Barber judgment, adopted in
Protocol No 2, applies to the action brought by the applicants.
24.
Therefore, since the first question concerns part-time workers not entitled to the additional benefits
provided for, the matter exercising the national court is whether the dispute in question concerns
admission to a pension scheme in which case, in accordance with the judgment in Bilka, the temporal
limitation imposed by the court in the Barber judgment, and adopted by Protocol No 2, does not apply or,
conversely, whether it concerns benefits granted under that scheme in which case the temporal limitation
does apply.
25.
The applicants in the main proceedings, who lodged their claims before they retired, emphasise that
they are not seeking retroactive payment of the additional benefits provided for in the case of MHOs but
equal treatment as regards future benefits according to their period of service on the same basis as
MHOs who under the law must be full-time employees. They are therefore asking for the whole of their
period of service to be taken into account for the purposes of calculating those benefits. At the hearing the
two applicants pointed out that no additional deduction or contribution is required of full-time employees
with MHO status in relation to part-time employees who did not have that status. In reply to a question by
the court, they emphasised that if they had enjoyed MHO status, they would have received those
additional benefits without being required to make additional contributions.
26.
The UK government underlines that the present case does not concern the right of membership to an
occupational social assurance scheme, to which Protocol No 2 does not apply, following the judgments in
Vroege and Fisscher, but the right to receive certain benefits the extent of which is to be determined in
accordance with art 119 of the Treaty. Moreover, relying on the judgment in Beune [1995] All ER (EC) 97,
[1994] ECR I-4471 (para 59), it considers that Protocol No 2, in view of the general terms in which it is
couched, covers benefits granted under an occupational pension scheme such as that in the present
case. But since the applicants in the main proceedings lodged their claims after 17 May 1990, they cannot
rely on art 119 in order to seek payment of equal benefits in respect of periods of employment prior to 17
May 1990, the date of the judgment in Barber [1990] 2 All ER 660, [1990] ECR I-1889.
27.
The first question to be clarified is indeed whether the temporal limitation provided for in Protocol No 2
may be extended also to cover indirect discrimination of part-time workers, as in the present case.
28.
 At the outset, it must be recalled that in the judgment in Bilka [1986] ECR 1607 (point 1, operative
part)11 there was held to be an infringement of art 119 by an undertaking which excluded part-time
workers from the undertakings occupational pension scheme, where that exclusion affected a far greater
number 46 of women than men, and the undertaking did not show that the exclusion was based on
objectively justified factors unrelated to any discrimination on the ground of sex 12.
11
See also judgment in Rinner-Khn v FWW Spezial Gebudereinigung GmbH Case 171/88 [1989] ECR 2743.
12
In the judgment in Vroege [1995] All ER (EC) 193, [1994] ECR I-4551 (para 16), the court held: The reasoning
behind that finding is that if, as can be seen from the judgment of 31 March 1981 in Jenkins v Kingsgate (Clothing
Productions) Ltd Case 96/80 [1981] 1 WLR 972, [1981] ECR 911, a pay policy which consists of setting a lower hourly
rate for part-time work than for full-time work may in certain cases entail discrimination between men and women, the
same applies where part-time workers are refused a company pension. Since such a pension falls within the concept of
pay, within the meaning of the second paragraph of art 119, it follows that, hour for hour, the total remuneration paid by
the employer to full-time workers is higher than that paid to part-time workers.

29.
In particular, as regards the right of membership of an occupational scheme, the court found that
there is no reason to suppose that the professional groups concerned could have been mistaken about
the applicability of art 119 (see Vroege [1995] All ER (EC) 193, [1994] ECR I-4551 (para 28)).
30.
Moreover, in Barber [1990] 2 All ER 660, [1990] ECR I-1889 (para 28), the court held that a pension
paid under a contracted-out scheme constitutes consideration paid by the employer to the worker in
respect of his employment and consequently falls within the scope of art 119.
31.
Yet, in Barber, the court, adjudicating for the first time on the question whether the determination of a
different pensionable age depending on sex constitutes, in the context of occupational pension schemes,
unlawful discrimination13, and answering that question in the affirmative, deemed it necessary to limit in
time the effects of its judgment. It based that limitation, on the one hand, on the fact that the member
states and the parties concerned were reasonably entitled, in light of the derogations provided for in art
7(1)(a) of Council Directive (EEC) 79/7 on the progressive application of equal treatment for men and
women in matters of social security (OJ 1979 L6 p 24) and art 9(a) of Council Directive (EEC) 86/378 on
the implementation of the principle of equal treatment for men and women in occupational social security
schemes (OJ 1986 L283 p 27), with regard to pensionable age, to take for granted that the principle of
equality as between male and female employees did not apply to the case in question and, on the other
hand, on the finding that the retroactive validity of the judgment would risk disturbing the financial
equilibrium of many occupational pension schemes.
13
See eg the judgment in Moroni [1993] ECR I-6591 (para 16).

32.
Consequently, the temporal limitation of the possibility of reliance on the direct effect of art 119 of the
Treaty concerns solely discrimination which might be justified on the basis of exceptions provided for in
Community provisions14, such as art 9 of Directive 86/37815. The court has repeatedly held that
14
I emphasised that in my opinion in Dietz v Stichting Thuiszorg Rotterdam Case C-435/93 [1996] ECR I-5223 (para 23).
Judgment was given in that case on 24 October 1996.
15
The exceptions provided for in that article concern (a) determination of a different retirement age for men and women,
(b) survivors pensions and (c) determination of different levels of contributions by workers.

the limitation of the effects in time of the Barber judgment concerned only those kinds of
discrimination which employers and pension schemes could reasonably have considered to be
permissible owing to the transitional derogations for which Community law provided and which
were capable of 47 being applied to occupational pensions (See the judgment in Dietz [1996]
ECR I-5223 (para 19).)16

16
See also the judgments in Vroege [1995] All ER (EC) 193, [1994] ECR I-4551 (paras 2027) and Fisscher [1995]
All ER (EC) 193, [1994] ECR I-4583 (paras 1724) and my opinion in Dietz [1996] ECR I-5223 (paras 2829).

It further held that the limitation of the effects in time of the Barber judgment did not apply to the right
to join an occupational pension scheme, inasmuch as the judgment in Bilka did not provide for any such
limitation (see Dietz v Stichting Thuiszorg Rotterdam Case C-435/93 [1996] ECR I-5223 (para 22),
Vroege [1995] All ER (EC) 193, [1994] ECR I-4551 (para 32) and Fisscher [1995] All ER (EC) 193, [1994]
ECR I-4583 (para 28)).
33.
According to the UK government, in order for the applicants in the main proceedings to be entitled to
the additional benefits provided for in the case of MHOs, they need to rely on art 119 of the Treaty at two
different stages. First, they have to rely on it in order to have their part-time service recognised as
satisfying the 20-year requirement, and secondly, to have their subsequent service counted double in
order to acquire the right to a pension at 55 years of age instead of 60. Particularly in regard to
recognition in the case of the two applicants of the whole or part of their period of service for the purposes
of completion of the 20-year period, the UK government maintains that to take into account any periods
prior to 17 May 1990 would run counter to Protocol No 2.
34.
 In the present case, it appears from the pleadings that the contracted-out occupational pension
scheme in the main proceedings is essentially a twofold scheme. On the one hand, it makes provision for
persons merely having the duties of psychiatric nurses (mental health officers), but not MHO status and,
consequently, not having the right to additional benefits because they were part-time workers. On the
other hand, it contains provisions concerning persons having MHO status and the right to additional
benefits because they were full-time employees.
35.
It is thus clear from the pleadings that, under the national legislation, the discrimination is directed
essentially against part-time employees to whom the more favourable provisions in favour of MHOs do
not apply.
36.
In accordance with the settled case law of the court 17, I do not consider that any provision of Directive
86/378 entitled the parties concerned reasonably to regard as permissible, owing to the transitional
derogations contained in that directive, the kind of discrimination, such as that in the present case,
directed against part-time workers depriving them on retirement of the additional benefits to which MHOs
are entitled. Thus, a person may very well invoke with retroactive effect rights in his favour which are not
covered by the exceptions contained in art 9(a) of the directive.

17
See eg the judgments in Barber [1990] 2 All ER 660, [1990] ECR I-1889 (paras 4243), Vroege [1995] All ER (EC)
193, [1994] ECR I-4551 (paras 2027), Fisscher [1995] All ER (EC) 193, [1994] ECR I-4583 (paras 1734) and
Dietz [1996] ECR I-5223 (para 19). See also the judgment in Coloroll Pension Trustees Ltd v Russell Case C-200/91
[1995] All ER (EC) 23, [1994] ECR I-4389 (para 53).

37.
Accordingly, there can be retroactive reliance on the direct effect of art 119 of the Treaty by the
applicants in the main proceedings who suffered discrimination, so that the additional benefits provided
for in the case of MHOs under the existing scheme may be made available to them, with effect from 8
April 1976, the date of the judgment in Defrenne.
48
38.
I now come to an examination of the question whether the present dispute concerns admission to a
pension scheme in which case, following the judgment in Bilka, the temporal limitation imposed by the
court in Barber and adopted in Protocol No 2 does not apply or, conversely, whether it concerns the
calculation of benefits granted under such a scheme in which case the temporal limitation does apply.
39.
Certainly, inasmuch as this case involves discrimination not covered by the Barber judgment and
Protocol No 2, it would be possible to avoid the distinction between the right to membership of an
occupational pension scheme and the right to receive benefits under that scheme. Thus, in the present
case, it would be possible to obviate the need for examination of the question whether the discrimination
suffered by the applicants constitutes discrimination in regard to membership of an occupational pension
scheme and not simply in regard to the claim of entitlement to certain benefits.
40.
Nevertheless, for the sake of completeness in the analysis of the issue raised in the first preliminary
question and in view of the issue arising in the second preliminary question, I would add the following
considerations. On the basis of the foregoing analysis, according to which the discrimination against the
applicants in the main proceedings could not reasonably be regarded by the parties concerned as
permissible, I consider that, as a condition precedent to a successful claim to additional benefits provided
for by the national legislation in the case of MHOs, retroactive admission to the status of MHO, of which
the two applicants were deprived, would be necessary and, with it, recognition of entitlement in that
connection to those benefits. Thus, in my opinion, the fact that account was not taken of the period of
part-time service in calculating the period of service as MHOs of the two applicants in the main
proceedings constitutes unfair treatment in connection with admission to the special MHO status which
provides for the grant of additional benefits, and not in connection with the calculation in itself of the
benefits, which would result in the application of the temporal limitation of the effects of the Barber
judgment. Consequently, it will be necessary to apply the solution adopted in Bilka, in connection with the
right of admission to an occupational pension scheme, and to exclude the possibility of a temporal
limitation pursuant to the Barber judgment18.

18
Consequently, the solution adopted in Beune [1995] All ER (EC) 97, [1994] ECR I-4471, cannot be applied,
contrary to the assertions of the United Kingdom. That is because in that case the national legislation at issue laid down
a rule for calculating the amount of the civil service pension for male married former civil servants which was different
from that applicable to female married former civil servants. The court held, first, that art 119 precluded such a national
provision, and that married men placed at a disadvantage by discrimination must be treated in the same way and have
the same rules applied to them as married women and, secondly, that the temporal limitation under Protocol No 2
applies in the case of the persons who may rely on the direct effect of art 119 in order to claim equal treatment as
regards the payment of benefits under a pension scheme, such as the scheme in the main proceedings in that case
(see [1995] All ER (EC) 97, [1994] ECR I-4471 (points 23, operative part)).

41.
That solution is warranted by the case law of the court. In its judgment in Dietz [1996] ECR I-5223
(paras 23ff)19, which concerned the old-age pension entitlement under an occupational scheme of a part-
time worker who was originally excluded from her pension scheme, it was held that the preconditions
which led to the temporal limitation in Barber are not fulfilled in the case of discrimination against part-time
workers. The court stressed that
19
See also my opinion in Dietz [1996] ECR I-5223 (paras 27ff).

49
Entitlement to a retirement pension under an occupational scheme is indissolubly linked to the
right to join such a scheme. Membership would be of no interest to employees if it did not confer
entitlement to the benefits provided for by the scheme. (See the judgment in Dietz [1996] ECR I-
5223 (para 23).)
It was further held that
it should be noted that since the judgment in Bilka imposed no time restriction the direct effect
of Article 119 may be relied upon retroactively as from 8 April 1976, the date of the Defrenne
judgment, by victims of discrimination regarding the right to join an occupational pension scheme
who seek payment of benefits under that scheme. (See [1996] ECR I-5223 (para 27).)
And the judgment concluded (point 2, operative part):
The limitation of the effects in time of the Barber judgment does not apply to the right to join
an occupational pension scheme, such as that at issue in the main proceedings, or to the right to
payment of a retirement pension where the worker was excluded from membership of the scheme
in breach of Article 119 of the Treaty.
42.
In light of the foregoing analysis, I consider that art 119 of the Treaty and related Protocol No 2 are to
be interpreted as meaning that in the case of discrimination against part-time workers, as described in the
order for reference, their period of work, for the purpose of conferring entitlement to benefits under an
occupational pension scheme, must be calculated as a matter of principle as from 8 April 1976, the date
of delivery of the judgment in Defrenne.

B. Reply to the second preliminary question

43.
The second question raises the issue of national rules governing remedies in respect of rights under
Community law20.
20
Since, as is apparent from the pleadings, the actions brought by Mrs Magorrian and Mrs Cunningham were lodged on
22 September 1992, that is to say before their retirement on 18 October 1992 and 31 March 1994 respectively, the reply
by the court to that question would have practical relevance only if the court were to take the view that the temporal
limitation on the effects of the judgment in Barber, and Protocol No 2, do not apply to the present case, as the
Commission and the two applicants correctly point out. If it did, the applicants would be almost entirely deprived of the
advantages of admission to MHO status, from which they were excluded, given that, as stated in the order for reference,
the defendants in the main proceedings maintained that the applicants periods of service should be calculated only as
from 17 May 1990.

44.
More specifically, the question raised here concerns the case where the national legislation applied
imposes a limitation, if a claim is successful, on the right to have years of service taken into account, to
the two-year period preceding commencement of proceedings. When the question is asked whether the
limitation imposed in this case by national law on the right of the persons concerned to receive the
additional benefits provided for in the case of MHOs, a right conferred by Community law, leads to a
denial of effective protection, national courts are obliged to disregard that provision of national law, if they
deem it necessary.
45.
The applicants in the main proceedings maintain that such a limitation rule in national law may not be
applied by the national court. For, as they also pointed out at the hearing, if it were otherwise, the
paradoxical result would be a 50 finding of unequal treatment contrary to art 119 of the Treaty, as the
national court has already found, yet nevertheless that finding of unequal treatment would have no
practical consequence owing to the application of the national rule in question. Moreover, they
maintained, the removal of that discriminatory treatment would cost the civil service scheme very little 21.
21
According to the applicants in the main proceedings, payment of the additional benefits provided for in the case of
MHOs would cost the civil service scheme about 70,000 to 100,000 pa, as opposed to an annual budget amounting
to 355m.

46.
The UK government submits that a national rule which limits to two years prior to commencement of
proceedings the period in respect of which the additional benefits awarded in the case of MHOs may be
backdated is perfectly compatible with Community law. Relying on the case law of the court 22, it considers
that a reasonable time limit may be applied in respect of claims based on art 119 in the context of
occupational schemes. Time restrictions on the retroactive effect of a judicial decision do not render
impossible the enforcement of rights under Community law but contribute to legal certainty by
encouraging the persons concerned to show due diligence.
22
See the judgments in Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989,
Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043, Fisscher [1995] All ER (EC), [1994] ECR
 I-4583 (para 40) and Johnson v Chief Adjudication Officer (No 2) Case 410/92 [1995] All ER (EC) 258, [1994] ECR
I-5483.

47.
It should first of all be remembered that, as the court has consistently held 23, in the absence of
Community rules in the matter, the national rules relating to time limits for bringing actions are also
applicable to actions based on Community law, provided that they satisfy two requirements. First, they
must be no less favourable for such actions than for similar actions of a domestic nature, that is to say
claims brought under national law and, secondly, they must not render the exercise of rights conferred by
Community law impossible in practice. Thus, the national court has discretion in assessing whether the
national rules setting time limits on the bringing of actions are no less favourable for that type of action
than for similar actions of a domestic nature24 and do not render the exercise of the right conferred by
Community law impossible in practice.
23
See eg the judgments in Fisscher [1995] All ER (EC), [1994] ECR I-4583 (paras 3940, point 5 operative part) and
Rewe-Zentralfinanz [1976] ECR 2043 (paras 56).
24
In the present case that issue is not examined at all by the national court.

48.
In the present case, as I have already indicated in earlier paragraphs, the limitation is one which
affects the right of an employee to be admitted to a special status under an occupational pension scheme
in accordance with the requirements concerning equal access, and not that persons right retroactively to
seek certain benefits.
49.
Regulation 12 of the 1976 regulations, provides that, in the context of proceedings concerning access
to membership of an occupational pension scheme, the right to join the scheme in question cannot be
backdated more than two years prior to commencement of the proceedings. However, I consider that the
above limitation under national law has more far-reaching effects than the limitation of the employees
right to claim a retroactive award of benefits. The reason is that it affects the number of years in respect of
which the persons concerned may claim MHO status and, consequently, their ability to claim the
additional benefits which flow from that status25, since the minimum period of 20 years of service can
never be completed with the result that exercise of the 51 right to additional benefits is rendered
impossible in practice26, as the European Commission, moreover, rightly points out.
25
See also the problems expounded in the opinion of Advocate General Jacobs in Denkavit Internationaal BV v Kamer
van Koophandel en Fabrieken voor Midden-Gelderland Case C-2/94 [1996] ECR I-2827 (para 67, and esp para 69), in
connection with the terms virtually impossible or excessively difficult in relation to the exercise of rights conferred by
Community law, owing to the limitation period imposed by national law for bringing proceedings. However, the court,
which gave judgment on 11 June 1996, did not express a view on this issue.
26
The present case differs from that which exercised the court in its judgment in Marshall v Southampton and South West
Hampshire Area Health Authority (No 2) Case C-271/91 [1993] 4 All ER 586, [1993] ECR I-4367, which concerned
the imposition of an upper limit fixed a priori for the reparation of the loss sustained by a person as a result of
discriminatory dismissal, and the exclusion of an award of interest to compensate for the loss sustained by the recipient
of the compensation as a result of the effluxion of time until the capital sum awarded was actually paid. That case differs
from the present case because the latter concerns a temporal limitation on the right to retrospective admission to a
special status for pension purposes (that of MHO), in the event that the action is successful.

50.
On the first preliminary question I reached the conclusion that, in the event of discrimination of part-
time employees, such as that described by the national court in the order for reference, their period of
service, for the purpose of awarding the additional benefits to which they are entitled under the
occupational pension scheme, must in principle be calculated with effect from 8 April 1976, the date of
delivery of the judgment in Defrenne. Thus, the temporal limitation imposed by Barber and Protocol No 2
does not apply in the present case. Consequently, I do not believe it is possible to invoke the principles of
legal certainty and avoidance of retroactive disturbance to the financial equilibrium of a contracted-out
occupational pension scheme relied on by the court in Barber [1990] 2 All ER 660, [1990] ECR I-1889
(para 44) in connection with recognition of the right of the applicants in the main proceedings to the
additional benefits. That is because the unfair treatment of part-time employees sanctioned for breach of
art 119 of the Treaty was known about long before the judgment in Barber.
51.
Nevertheless, a provision of national law such as reg 12 of the 1976 regulations, would substantially
reduce the practical effectiveness of art 119 of the Treaty, and could have the consequence of deterring
workers suffering discrimination from vindicating their rights under that article and, finally, of depriving
them of actual and effective judicial protection27.

27
See also analogous reasoning of the court in Draehmpaehl v Urania Immobilienservice ohG Case C-180/95 [1997] All
ER (EC) 719 (para 40). In that case the question arose, inter alia, whether Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to employment, vocational
training and promotion, and working conditions, precludes provisions of domestic law imposing a ceiling on the
aggregate amount of compensation payable to several applicants discriminated against on the grounds of their sex in
the making of an appointment. The court held that the directive precludes such provisions of domestic law (see [1997]
All ER (EC) 719 (para 43)).

52.
Consequently, I consider that national legislation which does not permit retroactive admission, beyond
the period of two years from commencement of proceedings, of part-time employees to the special
occupational status provided for in the case of MHOs is rendering impossible in practice the exercise of
their right to belong to the status in question and to receive the additional benefits provided for by it, which
is a right flowing from the application of art 119 of the Treaty, and the national court is required not to
apply such a provision28.
28
Since this is a case of unfair discrimination in breach of art 119 of the Treaty, I do not consider it possible for the
judgment in Emmott v Minister for Social Welfare Case C-208/90 [1991] ECR I-4269, relied on by the applicants in the
main proceedings, to be applied, since that case concerned inadequate transposition of Directive 79/7. According to the
operative part of that judgment, so long as a member state has not properly transposed that directive into its legal
system: Community law precludes the competent authorities of a Member State from relying, in proceedings brought
against them by an individual before the national courts in order to protect rights directly conferred upon him by Article
4(1) of Directive 79/7 on national procedural rules relating to time limits for bringing proceedings. See also the
opinion of Advocate General Van Gerven in Vroege [1995] All ER (EC) 193, [1994] ECR I-4551 (para 31).

52
53.
If it were otherwise, the application of such a national provision would lead to a situation in which
judgments of the court are not observed. On the one hand, it the court may, as happened in Defrenne and
Barber impose temporal limitations on the effects of its judgments and, on the other, may review the
criteria for the application of national rules which impose time limits for bringing actions, including those
based on Community law (see eg esp Fisscher [1995] All ER (EC) 193, [1994] ECR I-4583 (paras 3940,
point 5 operative part) and Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case
33/76 [1976] ECR 1989 (paras 56)).

V CONCLUSION
54.
In light of the foregoing analysis, I propose that the Court of Justice should reply as follows to the
questions referred to it for a preliminary ruling:
(1) Article 119 of the EC Treaty and related Protocol No 2 are to be interpreted as meaning that,
in the case of discrimination against part-time workers, as described in the order for reference, their
period of work, for the purpose of conferring entitlement to benefits under an occupational pension
scheme, must be calculated as a matter of principle as from 8 April 1976, the date of delivery of the
judgment in Defrenne.
(2) National courts are required not to apply a rule of national law which, in the event of a
successful claim, does not permit retroactive admission, beyond the period of two years from
commencement of proceedings, of part-time employees to the special occupational status provided
for in the case of MHOs, thus rendering impossible in practice the exercise of their right to belong
to the status in question and to receive the additional benefits provided for by it, which is a right
flowing from art 119.

11 December 1997.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By order of 9 July 1996, received at the Court of Justice of the European Communities on 17 July
1996, the Office of the Industrial Tribunals and the Fair Employment Tribunal, Belfast, referred to the
Court of Justice for a preliminary ruling under art 177 of the EC Treaty a number of questions on the
interpretation of art 119 of the Treaty and of Protocol No 2, concerning that provision.
2.
Those questions have been raised in proceedings between Mrs Magorrian and Mrs Cunningham and
the Eastern Health and Social Services Board and the Department of Health and Social Services in
relation to certain additional benefits under a contracted-out retirement pension scheme.
53

The national provisions

3.
Under s 2(4) of the Equal Pay Act (Northern Ireland) 1970, claims concerning equal pay must be
brought within six months of the end of the relevant period of employment. Section 2(5) of the 1970 Act
provides that, in proceedings brought in respect of a failure to comply with an equal pay clause, a woman
is not to be entitled to be awarded any payment by way of arrears of remuneration or damages in respect
of a period earlier than two years before the date on which the proceedings were instituted.
4.
Section 56 of the Social Security Pensions (Northern Ireland) Order 1975, SI 1975/1503, provides that,
in the case of a public service pension scheme, the minister, government department or other person or
body concerned with its administration are to take such steps as are open to them for bringing the rules of
the scheme into conformity with the requirement of equal access.
5.
Regulation 12 of the Occupational Pension Schemes (Equal Access to Membership) Regulations
(Northern Ireland) 1976, SI 1976/238, amending the 1970 Act, provides that, in proceedings concerning
access to membership of occupational pension schemes, the right to be admitted to the scheme is to
have effect from a date no earlier than two years before the institution of proceedings.
6.
Regulation 3 of the Health and Personal Social Services (Superannuation) Regulations (Northern
Ireland) 1984, SI 1984/336 defines a mental health officer (MHO) as a whole-time officer on the medical
or nursing staff of a hospital used wholly or partly for the treatment of persons suffering from mental
disorder who devotes the whole or substantially the whole of his time to the treatment of such persons.
7.
Under reg 50(2) of the 1984 regulations, where a person has reached or passed the age of 50 and has
worked as a mental health worker for a period of 20 years and continues to work as a mental health
officer, his or her subsequent service is reckoned for pension purposes at twice its length and the right to
a pension is obtained at the age of 55 instead of the normal age of 60.

The main proceedings

8.
Mrs Magorrian and Mrs Cunningham were employed as qualified nurses in the mental health sector by
a public-sector health board responsible for supplying medical and other services in a region of Northern
Ireland.
9.
They commenced their careers working full-time with MHO status. When their family responsibilities
increased, they both commenced working part-time and therefore lost that status. Each of them was
nevertheless in charge of a hospital ward and as such responsible for full-time nurses.
10.
The difference between part-time work and full-time work was very small. In fact, as a result of a
reorganisation of work in 1981, the working hours for part-time nurses were set at 31 hours 5 minutes per
week, whilst full-time hours were reduced from 40 hours to 37 hours and 30 minutes per week.
11.
Both of the applicants in the main proceedings were affiliated to and contributed to the Health and
Personal Social Services Superannuation Scheme, a voluntary contracted-out pension scheme to which
both the employer and the employee contribute. Since 1973 this scheme has been open to part-time
workers working a specified number of hours and, since 1991, all part-time workers have been entitled to
join the scheme regardless of the hours worked. Under the scheme, persons affiliated receive a lump-
sum payment on retirement, followed by monthly payments.
54
12.
On 18 October 1992, Mrs Magorrian retired at the age of 59 years and 355 days, having completed 9
years and 111 days of full-time work as an MHO between 1951 and 1963, and the equivalent of 11 years
and 25 days of part-time service between 1979 and 1992. She had also worked part-time between 1969
and 1979, but for hours not reckonable for pension purposes.
13.
Mrs Cunningham retired in April 1994 at the age of 56 years and 80 days, having completed 15 years
and 175 days of full-time service as an MHO between 1956 and 1974, and the equivalent of 11 years and
105 days of part-time service between 1980 and 1994. She had also worked part-time between 1974 and
1980 for hours not reckonable for pension purposes and had elected not to make pension contributions
during that period.
14.
Mrs Magorrian therefore interrupted her career between her full-time employment and her part-time
work, whilst Mrs Cunningham went directly from full-time work to part-time work without interruption.
15.
On their retirement, the applicants in the main proceedings received the lump sums to which they were
entitled, together with their basic retirement pensions, but were not given certain additional benefits to
which they would have been entitled under reg 50(2) of the 1984 regulations if they had had the status of
mental health officers at the time of their retirement. It was made clear at the hearing that, if the applicants
had had that status, they would have obtained the additional benefits without having to pay additional
contributions.
16.
By application dated 22 September 1992 the applicants in the main proceedings brought the matter
before the national court, relying on art 119 of the Treaty in support of their claim for additional benefits on
the basis of their length of service calculated from 8 April 1976, the date of the judgment in Defrenne v
Sabena Case 43/75 [1981] 1 All ER 122, [1976] ECR 455 or, in the alternative, from 13 May 1986, the
date of the judgment in Bilka-Kaufhaus GmbH v Weber von Hartz Case 170/84 [1986] ECR 1607. They
claim that there is no justification for limiting the calculation of their length of service to the period of two
years laid down in the 1970 Act, or to 17 May 1990, the date of the judgment in Barber v Guardian Royal
Exchange Assurance Group Case C-262/88 [1990] 2 All ER 660, [1990] ECR I-1889, since to do so would
deprive them of any effective legal remedy.
17.
It is clear from the order for reference that all the parties agree that payment of those benefits under
the applicants occupational pension scheme constitutes pay for the purposes of art 119 of the Treaty
and Council Directive (EEC) 75/117 on the approximation of the laws of the member states relating to the
application of the principle of equal pay for men and women (OJ 1975 L45 p 19). The national file also
shows that the applicants both lodged their claims before leaving their employment.
18.
In its interlocutory judgment of 12 September 1995 the national court found that the exclusion of part-
time psychiatric nurses from MHO status constituted indirect discrimination based on sex, since a
considerably smaller proportion of women than men working in the mental-health sector in Northern
Ireland were in a position to satisfy the requirements imposed by full-time working. It further found that
discrimination to be unjustified.

The preliminary questions

19.
Taking the view that the outcome of the dispute depended on the interpretation of Community law, the
national court decided to stay the 55 proceedings and to refer the following questions to the Court of
Justice for a preliminary ruling:
In circumstances where: (a) a worker has been employed by a Health Board which is part of the
State, in employment concerned with the care of the mentally ill to which an occupational pension
scheme applies; (b) the worker has at all material times either been a member or been eligible to
be a member of the pension scheme; (c) the pension scheme contains a term according to which
those who work full-time and devote all or substantially all their working hours to the care of the
mentally ill (who are described as Mental Health Officers) are entitled to additional benefits not
available to those doing the same work part-time, as follows: where a person has reached or
passed the age of 50 and has worked as a Mental Health Officer for 20 years (here referred to as
the qualifying service) and continues to work as a Mental Health Officer, then (i) their subsequent
service is reckoned for pension purposes at twice its length (here referred to as the double-time
service); and (ii) they have the right to a pension at the age of 55 instead of the normal age of 60;
(d) the worker is deprived of the status of Mental Health Officer and the additional benefits attached
thereto solely on the ground that her employment was part-time; (e) the national tribunal has held
that the provisions described at (c) and (d) constitute discrimination on grounds of sex against
women working part-time in the care of the mentally ill; (f) the pension which the worker receives
and the additional benefits which they claim, are only payable to them as from their respective
retirements in 1992 and 1994, after their claims have been brought before the national tribunal; and
(g) the calculation of the additional benefits from their respective retirement dates in 1992 and 1994
would involve counting their years of service prior to 1992
Question 1: From what date should the service of the workers be counted for the purpose of
calculating the additional benefits to which they are entitled: (i) 8 April 1976, (ii) 17 May 1990, (iii)
some other and, if so, what date?
Question 2: Where the relevant national legislation restricts back-dating entitlement in the event
of a successful claim to a period of 2 years prior to the date on which the claim was made, does
this amount to the denial of an effective remedy under Community law and is the Industrial Tribunal
obliged to disregard such provision in domestic law if it feels it necessary to do so?

The first question

20.
By its first question the national court essentially seeks to ascertain the date as from which the periods
of service of part-time workers who have suffered indirect discrimination based on sex must be taken into
account for the purpose of calculating the additional benefits to which they are entitled.
21.
It must be emphasised first of all that it is common ground that payment of additional benefits under an
occupational scheme such as that under consideration in the main proceedings falls, as a matter of
principle, within the concept of pay for the purposes of art 119 of the Treaty.
22.
In Defrenne [1981] 1 All ER 122, [1976] ECR 455, the court held that the principle of equal pay under
art 119 may be relied on before national courts and that those courts have a duty to ensure the protection
of the rights which that 56 provision vests in individuals. However, the court also stated in that judgment
that important considerations of legal certainty affecting all the interests involved, both public and private,
meant that the direct effect of art 119 could not be relied on in order to support claims concerning pay
periods prior to the date of that judgment, 8 April 1976, except as regards workers who had already
brought legal proceedings or made an equivalent claim (see [1981] 1 All ER 122, [1976] ECR 455 (paras
7475)).
23.
On the other hand, in Bilka [1986] ECR 1607 (paras 20, 22), the court found that, where an
occupational pension scheme, although adopted in accordance with the provisions laid down by national
legislation, was based on an agreement with workers or their representatives and the public authorities
did not contribute to its financing, such a scheme did not constitute a social security scheme governed
directly by statute and therefore fell outside the scope of art 119, and that the benefits paid to employees
under such a scheme constituted consideration received by the employee from the employer in respect of
his employment, as referred to in the second paragraph of art 119.
24.
 Although those principles were upheld in the judgment Barber [1990] 2 All ER 660, [1990] ECR I-1889
in relation to contracted-out occupational pension schemes, the court also stated in that judgment, that
overriding considerations of legal certainty precluded reliance being placed on the direct effect of art 119
in order to claim entitlement to a pension with effect from a date prior to delivery of the judgment in that
case, except in the case of persons who had in the meantime taken steps to safeguard their rights (see
[1990] 2 All ER 660, [1990] ECR I-1889 (paras 4445)).
25.
As the court stated in its judgment in Ten Oever v Stichting Bedrijfspensioenfonds voor het
Glazenwassersen Schoonmaakbedrijf Case C-109/91 [1993] ECR I-4879 (paras 1920), by virtue of the
judgment in Barber, the direct effect of art 119 may be relied on, for the purpose of claiming equal
treatment in the matter of occupational pensions, only in relation to benefits payable in respect of periods
of employment subsequent to 17 May 1990, the date of the judgment in that case, subject to the
exception in favour of workers or those claiming under them who have before that date initiated legal
proceedings or raised an equivalent claim under the applicable national law.
26.
That limitation is also to be found in EC Treaty Protocol No 2, which provides that:
For the purposes of Article 119 of this treaty, benefits under occupational social security
schemes shall not be considered as remuneration if and in so far as they are attributable to periods
of employment prior to 17 May 1990, except in the case of workers or those claiming under them
who have before that date initiated legal proceedings or introduced an equivalent claim under the
applicable national law.
27.
However, in Vroege v NCIV Instituut voor Volkshuisvesting BV Case C-57/93 [1995] All ER (EC) 193,
[1994] ECR I-4551 (paras 2027) and Fisscher v Voorhuis Hengelo BV Case C-128/93 [1995] All ER (EC)
193, [1994] ECR I-4583 (paras 1724), the court took the view that the limitation of the effects in time of
the Barber judgment concerned only those kinds of discrimination which, owing to the transitional
derogations for which Community law provided and which were capable of being applied to occupational
pensions, employers and pension schemes could reasonably have considered to be permissible (see the
judgment 57 in Dietz v Stichting Thuiszorg Rotterdam Case C-435/93 [1996] ECR I-5223 (para 19)).
28.
As far as the right to join an occupational scheme was concerned, it also stated that there was no
reason to suppose that those concerned could have been mistaken as to the applicability of art 119.
29.
In fact, it has been clear since the judgment in Bilka that any discrimination, based on sex, in the
recognition of that right infringes art 119 (see the judgments in Vroege [1995] All ER (EC) 193, [1994]
ECR I-4551 (para 29), Fisscher [1995] All ER (EC) 193, [1994] ECR I-4583 (para 26) and Dietz [1996]
ECR I-5223 (para 20)).
30.
Therefore, as the judgment in Bilka included no limitation of its effects in time, the direct effect of art
119 may be relied on, as from 8 April 1976, the date of the judgment in Defrenne, in which that article was
first held to have direct effect, in order retroactively to claim equal treatment in relation to the right to join
an occupational pension scheme (see the judgment in Dietz [1996] ECR I-5223 (para 21)).
31.
According to the UK government, the main proceedings concern the amount of a retirement pension
paid under an occupational social security scheme and not the right to belong to such a scheme:
consequently, art 119 applies only in order to alter the level of benefits to which a person in the situation
of the applicants in the main proceedings is entitled and only periods after 17 May 1990 should be taken
into account for the purposes of that calculation.
32.
As regards the right to receive benefits additional to a retirement pension under an occupational
scheme such as that involved in the main proceedings, the court finds that, even if the persons concerned
have always been entitled to a retirement pension under the Superannuation Scheme, nevertheless they
were not fully admitted to that contributory scheme. Solely on account of the fact that they worked part-
time, they were specifically excluded from MHO status which gives access to a special scheme under the
Superannuation Scheme.
33.
It is sufficient to recall in this regard that, in its judgment in Dietz [1996] ECR I-5223 (para 23), the
court has already stated that membership of a scheme would be of no interest to employees if it did not
confer entitlement to the benefits provided by the scheme in question. In a situation such as that involved
in that case, the court took the view that entitlement to a retirement pension under an occupational
scheme was indissolubly linked to the right to join such a scheme.
34.
The same is true where the discrimination suffered by part-time workers stems from discrimination
concerning access to a special scheme which confers entitlement to additional benefits.
35.
In light of those considerations, the reply to be given to the first question must be that periods of
service completed by part-time workers who have suffered indirect discrimination based on sex must be
taken into account as from 8 April 1976, the date of the judgment in Defrenne, for the purposes of
calculating the additional benefits to which they are entitled.

The second question

36.
By its second question the national court is asking essentially whether Community law precludes the
application to a claim based on art 119 of the Treaty of a national rule under which entitlement, in the
event of a successful claim, is limited to a period which starts to run from a point in time two years prior to
commencement of proceedings in connection with the claim.
58
37.
The court has consistently held that, in the absence of relevant Community rules, it is for the national
legal order of each member state to designate the competent courts and to lay down the procedural rules
for proceedings designed to ensure the protection of the rights which individuals acquire through the
direct effect of Community law, provided that such rules are not less favourable than those governing
similar domestic actions and are not framed in such a way as to render impossible in practice the exercise
of rights conferred by Community law (see, to that effect, the judgments in Rewe-Zentralfinanz eG v
Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989 (paras 56), Comet BV v
Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043 (para 13), Fisscher [1995] All ER (EC)
193, [1994] ECR I-4583 (para 39) and Johnson v Chief Adjudication Officer (No 2) Case 410/92 [1995] All
ER (EC) 258, [1994] ECR I-5483 (para 21)).
38.
According to the applicants in the main proceedings, there is nothing in Fisscher to justify restricting
the extent of the benefits which may be awarded to them, at least in relation to the period subsequent to
1976. In fact, there would appear to be little point in holding that persons in the applicants situation have
the right to become members of an occupational social security scheme if the benefits which they derive
under it can be calculated only by reference to their employment from 1990 onwards.
39.
At the hearing the European Commission maintained that reg 12 of the 1976 regulations in fact
prevents the applicants in the main proceedings from vindicating their rights under art 119 of the Treaty
and that, consequently, the application of that rule runs counter to the principle of legal protection.
40.
The UK government, on the other hand, argues that a restrictive national rule of the type at issue in
the main proceedings has the effect of limiting the scope of a retroactive claim relating to a period
preceding commencement of the proceedings and that it is therefore comparable to the rule at issue in
Johnson [1995] All ER (EC) 258, [1994] ECR I-5483.
41.
As far as this issue is concerned, it must be stated that application of a procedural rule such as reg 12
of the 1976 regulations whereby, in proceedings concerning access to membership of occupational
pension schemes, the right to be admitted to a scheme may have effect from a date no earlier than two
years before the institution of proceedings would deprive the applicants in the main proceedings of the
additional benefits under the scheme to which they are entitled to be affiliated, since those benefits could
be calculated only by reference to periods of service completed by them as from 1990, that is to say two
years prior to commencement of proceedings by them.
42.
However, it should be noted that, in such a case, the claim is not for the retroactive award of certain
additional benefits but for recognition of entitlement to full membership of an occupational scheme
through acquisition of MHO status which confers entitlement to the additional benefits.
43.
Thus, whereas the rules at issue in Steenhorst-Neerings v Bestuur van de Bedrijfsvereniging voor
Detailhandel, Ambachten en Huisvrouwen Case C-338/91 [1993] ECR I-5475 and in Johnson merely
limited the period, prior to commencement of proceedings, in respect of which backdated benefits could
be obtained, the rule at issue in the main proceedings in this case prevents the entire record of service
completed by those concerned after 8 April 1976 until 1990 from being taken into account for the
purposes of calculating the additional benefits which would be payable even after the date of the claim.
59
44.
Consequently, unlike the rules at issue in the judgments cited above, which in the interests of legal
certainty merely limited the retroactive scope of a claim for certain benefits and did not therefore strike at
the very essence of the rights conferred by the Community legal order, a rule such as that before the
national court in this case is such as to render any action by individuals relying on Community law
impossible in practice.
45.
Moreover, the effect of that national rule is to limit in time the direct effect of art 119 in cases in which
no such limitation has been laid down either in the courts case law or in Protocol No 2.
46.
Finally, the argument put forward by the UK government that the purpose of a temporal limitation is to
contribute to legal certainty by encouraging claimants to be diligent cannot invalidate that conclusion.
Suffice it to state in that regard that the national rules in question apply even to persons who, like Mrs
Magorrian and Mrs Cunningham, brought proceedings before leaving their employment and being
admitted to the occupational retirement pension scheme in question.
47.
Accordingly, the reply to be given to the second question must be that Community law precludes the
application, to a claim based on art 119 of the EC Treaty for recognition of the claimants entitlement to
join an occupational pension scheme, of a national rule under which such entitlement, in the event of a
successful claim, is limited to a period which starts to run from a point in time two years prior to
commencement of proceedings in connection with the claim.

Costs
48.
The costs incurred by the UK government and the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the proceedings pending before the national court, the decision on costs
is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the questions referred to it by
the Office of the Industrial Tribunals and the Fair Employment Tribunal, Belfast, by order of 9 July 1996,
hereby rules: (1) Periods of service completed by part-time workers who have suffered indirect
discrimination based on sex must be taken into account as from 8 April 1976, the date of the judgment in
Defrenne v Sabena Case 43/75 [1981] 1 All ER 122, [1976] ECR 455, for the purposes of calculating the
additional benefits to which they are entitled. (2) Community law precludes the application, to a claim
based on art 119 of the EC Treaty for recognition of the claimants entitlement to join an occupational
pension scheme, of a national rule under which such entitlement, in the event of a successful claim, is
limited to a period which starts to run from a point in time two years prior to commencement of
proceedings in connection with the claim.

60

[1998] All ER (EC) 61

Garage Molenheide BVBA and others v Belgium

(Joined cases C-286/94, C-340/95, C-401/95 and C-47/96)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES GULMANN (PRESIDENT OF THE CHAMBER), WATHELET, MOITINHO DE ALMEIDA, JANN
AND SEVN ADVOCATE GENERAL FENNELLY
30 JANUARY, 20 MARCH, 18 DECEMBER 1997
European Community Value added tax Input tax Disallowance of input tax Whether national
measures for protective retention of balance of value added tax claimed by taxpayer precluded by
Community law Effect of principle of proportionality Council Directive (EEC) 77/388, art 18(4).

The four plaintiffs claimed refunds of VAT from the Belgian fiscal authorities. However, on examination of
the plaintiffs respective tax returns, the authorities found either that there were grounds for presuming tax
evasion or that there was a value added tax debt owing to them which was not apparent from the tax
returns and which the taxable person contested. In such circumstances Belgian legislation allowed the
fiscal authorities to retain the amount of refund claimed as a protective measure and in each case the
amounts claimed were so retained. All four plaintiffs instituted proceedings against the Belgian state
claiming that the Belgian legislation was incompatible with Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment. In each case, the national courts referred to the Court of Justice
of the European Communities for a preliminary ruling questions seeking to determine (i) whether
measures such as those in the Belgian law were precluded by art 18(4) 1 of Directive 77/388 which
permitted member states, where authorised deductions exceeded the amount of tax due, either to make a
refund or to carry the excess forward according to the conditions which they were to determine, and, if
not, (ii) what effect the principle of proportionality might have in such circumstances.
1
Article 18, so far as material, is set out at p 86 g h, post

Held (1) Directive 77/388 was intended to establish a uniform basis so as to guarantee the neutrality of
the value added tax system and to harmonise the rules governing deductions to the extent that they
affected the actual amounts collected and to ensure that the deductible proportion was calculated in a
similar manner in all the member states. Accordingly, Title XI of the directive which dealt with deductions,
and in particular art 18, related to the normal functioning of the common system of value added tax and
did not concern measures involving the retention by the fiscal authorities as a protective measure of
refundable amounts of value added tax on grounds of suspected incorrect particulars in returns or a
suspected value added tax debt to the state. It followed that art 18(4) did not preclude such measures
(see p 93 c d and p 96 e, post).
 (2) Measures such as those at issue were liable to have an impact on the national authorities
obligation to make an immediate refund under art 18(4). 61Thus, in accordance with the principle of
proportionality, the member states had to employ means which, whilst enabling them effectively to attain
the objective pursued by their domestic laws, were the least detrimental to the objectives and the
principles laid down by the relevant Community legislation. Accordingly, whilst it was legitimate for
measures adopted by the member states to seek to preserve the rights of the treasury as effectively as
possible, such measures could not be used in such a way that they would have the effect of
systematically undermining the right to deduct value added tax, which was a fundamental principle of the
common system of value added tax established by Community legislation. It followed that the principle of
proportionality was applicable to national measures which were adopted by a member state in the
exercise of its powers relating to value added tax (see p 93 g to j and p 96 f, post).
(3) It was for the national court to examine whether or not the measures in question and the manner in
which they were applied by the competent administrative authority were proportionate. In the context of
that examination, if the national provisions or a particular construction of them would constitute a bar to
effective judicial review, in particular, review of the urgency and necessity of retaining the refundable
value added tax balance, and would prevent the taxable person from applying to a court for replacement
of the retention by another guarantee sufficient to protect the interests of the treasury but less onerous for
the taxable person, or would prevent an order from being made, at any stage of the procedure, for the
total or partial lifting of the retention, the national court should disapply those provisions or refrain from
placing such a construction on them. Moreover, in the event of the retention being lifted, calculation of the
interest payable by the treasury which did not take as its starting point the date on which the value added
tax balance in question would have had to be repaid in the normal course of events would be contrary to
the principle of proportionality (see p 95 j to p 96 b g to j, post); Gebhard v Consiglio dellOrdine degli
Avvocati e Procuratori di Milano Case C-55/94 [1996] All ER (EC) 189 applied.

Notes
For an introduction to Community provisions on value added tax, see 52 Halsburys Laws (4th edn) paras
20132014.

Cases cited
Balocchi v Ministero delle Finanze dello Stato Case C-10/92 [1997] STC 640, [1993] ECR I-5105, ECJ.
Becker v Finanzamt Mnster-Innenstadt Case 8/81 [1982] ECR 53.
BP Supergas Anonimos Etairia Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece Case C-62/93
[1995] All ER (EC) 684, [1995] ECR I-1883, ECJ.
Drexl (Criminal Proceedings against) Case 299/86 [1988] ECR 1213.
EC Commission v Belgium Case 324/82 [1984] ECR 1861.
EC Commission v France Case 196/85 [1987] ECR 1597.
EC Commission v France Case 50/87 [1988] ECR 4797.
EC Commission v France Case C-276/91 [1997] STC 584, [1993] ECR I-4413, ECJ.
Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325, ECJ.
Finanzamt Bergisch Gladbach v Skripalle Case C-63/96 [1997] STC 1035, ECJ.
Francovich v Italy Joined cases C-6/90 and C-9/90 [1991] ECR I-5357.
Fritz Werner Industrie-Ausrstungen GmbH v Germany Case C-70/94 [1995] ECR I-3189.
62
Gaston Schul Douane Expediteur BV v Inspecteur der Inverroechten en Accijnzen Case 15/81 [1982]
ECR 1409.
Gebhard v Consiglio dellOrdine degli Avvocati e Procuratori di Milano Case C-55/94 [1996] All ER (EC)
189, [1995] ECR I-4165, ECJ.
Jeunehomme v Belgium Joined cases 123/87 and 330/87 [1988] ECR 451.
Leifer (Criminal proceedings against) Case C-83/94 [1995] ECR I-3231.
Lennartz v Finanzamt Mnchen III Case C-97/90 [1995] STC 514, [1991] ECR I-3795, ECJ.
McDermott v Minister for Social Welfare Case 286/85 [1987] ECR 1453.
Netherlands v Federatie Nederlandse Vakbeweging Case 71/85 [1986] ECR 3855.
Pastoors v Belgium Case C-29/95 [1997] ECR I-285.
Peterbroeck, Van Campenhout & Cie (SCS) v Belgium Case C-312/93 [1996] All ER (EC) 242, [1995]
ECR I-4599, ECJ.
Pubblico Ministero v Ratti Case 148/78 [1979] ECR 1629.
R v Customs and Excise Comrs, ex p Strangewood Ltd [1987] STC 502.
R v Secretary of State for Transport, ex p Factortame Ltd Case C-213/89 [1991] 1 All ER 70, [1990] 3
WLR 818, [1990] ECR I-2433, ECJ.
Reisdorf v Finanzamt Kln-West Case C-85/95 [1997] STC 180, [1996] ECR I-6257, ECJ.
Rompelman v Minister van Financin Case 268/83 [1985] ECR 655.
Union nationale des entraneurs et Cadres techniques professionnels du football (Unectef) v Heylens
Case 222/86 [1987] ECR 4097.

Reference
The Hof van Beroep te Antwerpen (the Court of Appeal, Antwerp), 13th and 3rd Chambers (Cases C-
286/94 and C-340/95), the Rechtbank van Eerste Aanleg te Brussel (the Court of First Instance, Brussels)
(Case C-401/95) and the Rechtbank van Eerste Aanleg te Brugge (the Court of First Instance, Bruges)
(Case C-47/96) referred to the Court of Justice of the European Communities for preliminary rulings a
number of questions (set out at p 89 d e, p 90 g to j, p 91 e f and p 92 c to e, post) on the interpretation of
art 18(4) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states
relating to turnover taxescommon system of value added tax: uniform basis of assessment. Those
questions concerned the compatibility with Community law of a form of preventive attachment exercised
by a national fiscal authority over the reimbursement of amounts of value added tax ostensibly overpaid
by a taxable person for certain declared tax periods. Those questions were raised in four actions brought
against the Belgian state by Garage Molenheide BVBA (Molenheide), Peter Schepens, Bureau Rik
Decan-Business Research & Development NV (BRD) (Decan) and Sanders BVBA (Sanders). Written
observations were submitted on behalf of: Molenheide, by V Dauginet, of the Antwerp Bar; Decan and
Sanders, by L Vandenberghe and R Tournicourt, of the Brussels Bar; the Belgian government, by J
Devadder, Counsellor General, Ministry of Foreign Affairs, External Trade and Development Co-operation,
acting as agent; the Greek government (Cases C-340/95, C-401/95 and C-47/96), by F Georgakopoulos,
Deputy Legal Adviser, Legal Council of State, and A Rokophyllou, Special Adviser to the Deputy Minister
of Foreign Affairs, acting as agents; the Italian government (Cases C-286/94, C-340/95 and C-401/95), by
Professor Umberto Leanza, head of the Department of Contentious Diplomatic Affairs, Ministry of Foreign
Affairs, acting as agent, assisted by Maurizio Fiorilli, Avvocato Dello Stato; the Swedish government
(Case C-401/95), by E Brattgrd, 63Departmental Adviser, Department of Foreign Trade, Ministry of
Foreign Affairs, acting as agent; and the European Commission, by B J Drijber, of its legal service, acting
as agent. Oral observations were made on behalf of Molenheide, represented by M Vanden Broeck, of
the Antwerp Bar, Decan and Sanders, represented by L Vandenberghe, the Belgian government,
represented by B van de Walle de Ghelcke and G de Wit, of the Brussels Bar, the Greek government,
represented by F Georgakopoulos, the Italian government, represented by G De Bellis, Avvocato Dello
Stato, and the Commission, represented by B J Drijber. The language of the cases was Dutch. The facts
are set out in the opinion of the Advocate General.

20 March 1997.

The Advocate General (N Fennelly)

delivered the following opinion.
1.
The Court of Justice of the European Communities is asked in these joined preliminary references to
determine the compatibility with Community law, and particularly with Council Directive (EEC) 77/388 on
the harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment (OJ 1977 L145 p 1) (the Sixth Directive), of a form of preventive
attachment exercised by a national fiscal authority over the reimbursement of amounts of value added tax
(VAT) ostensibly overpaid by a taxable person for certain declared VAT periods. Does such an attachment
come within the scope of the Sixth Directive, or should it instead fall to be classified as a tax collection
measure within the exclusive competence of the member states? Central to all of the references is the
potential application of the Community law principle of proportionality.

I LEGAL BACKGROUND

A. Community legislation
2.
The relevant provisions of the Sixth Directive are contained essentially in Title XI, entitled Deductions,
comprising arts 17 to 20. Under art 17 a taxable person is entitled to deduct from the VAT which he is
liable to pay on his taxable supplies the VAT due or paid in respect of the goods and services supplied or
to be supplied to him by another taxable person. The rules governing the exercise of the right to deduct,
which I shall quote as far as relevant, are set out in art 18. Article 18(1) deals with the formal evidential
requirements, such as the obligation to hold the relevant invoices. Article 18(2) provides:
The taxable person shall effect the deduction by subtracting from the total amount of value
added tax due for a given tax period the total amount of the tax in respect of which, during the
same period, the right to deduct has arisen and can be exercised under the provisions of paragraph
1.
Where the requirements of art 18(1) and (2) are not respected, art 18(3) permits the member states to
determine the conditions and procedures whereby a taxable person may be authorised to make a
deduction. Article 18(4), which is central to the instant cases, provides:
Where for a given tax period the amount of authorized deductions exceeds the amount of tax
due, the Member States may either make a refund or carry the excess forward to the following
period according to conditions which they shall determine.
64
 However, Member States may refuse to refund or carry forward if the amount of the excess is
insignificant.
3.
Title XIII of the Sixth Directive concerns the Obligations of Persons Liable for Payment. Article 22
deals with taxable persons Obligations under the internal system, and includes obligations regarding
record-keeping, tax returns and accounting. The first sub-paragraph of art 22(4) obliges taxable persons
to submit a return within an interval to be determined by each Member State, which may not exceed two
months following the end of each tax period. The second sub-paragraph of art 22(4) provides:
The return must set out all the information needed to calculate the tax that has become
chargeable and the deductions to be made, including, where appropriate, and in so far as it seems
necessary for the establishment of the tax basis, the total amount of the transactions relative to
such tax and deductions, and the total amount of the exempted supplies.
Under art 22(5) the taxable person is obliged in principle to pay the net amount of value added tax
when submitting the return, unless the member state has specified a different payment date. Article 22(6)
permits member states to require taxable persons
to submit a statement, including the information specified in paragraph 4 concerning all
transactions carried out the preceding year [and which] must provide all the information necessary
for any adjustments.
Article 22(8), which is of particular relevance as regards national tax collection measures, provides:
Without prejudice to the provisions to be adopted pursuant to Article 17(4), Member States may
impose other obligations which they deem necessary for the correct levying and collection of the
tax and for the prevention of fraud. (My emphasis.)
4.
Article 27, the sole article of Title XV, entitled Simplification Procedures, of the Sixth Directive,
provides a procedure whereby member states may apply for derogations from the provisions of the
directive. Article 27(1) is worded as follows:
The Council, acting unanimously on a proposal from the Commission, may authorize any
Member State to introduce special measures for derogation from the provisions of this Directive, in
order to simplify the procedure for charging the tax or prevent certain types of tax evasion or
avoidance. Measures intended to simplify the procedure for charging the tax, except to a negligible
extent, may not affect the amount of tax due at the final consumption stage.

B. Belgian legislation
5.
Article 18(4) of the Sixth Directive is implemented in Belgian law primarily by art 47 of the BTW-
Wetboek (the VAT code), which provides that the excess of authorised deductions over VAT due for a
particular tax period shall be carried forward to the following tax period. Article 76(1)(1) of the code, as
amended by the Law of 28 December 19922, provides for the refund, where it is requested by 65 the
taxable person, of any excesses which are outstanding at the end of the calendar year in accordance with
the conditions to be established by royal decree. Under sub-para (2), the Crown may permit, subject to
such conditions as it sees fit, the grant of refunds even before the end of the calendar year. Sub-
paragraph (3) is central to the disputes involved in the present references and is worded as follows:
2
See Belgisch Staatsblad, 31 December 1992.

With respect to the requirements laid down in the first and second subparagraphs, provision
may be made by Royal Decree for a withholding in favour of the VAT, Registration and Property,
Authority3, with the effect of a preventive attachment within the meaning of Article 1445 of the
Judicial Code.

3
The Administratie van de BTW, Registratie en Domeinen.

6.
The new form of attachment was introduced by art 7 of the Koninklijk Besluit (Royal Decree) of 29
December 19924, which inserted, inter alia, a new art 8/(1)(3) into what is known in Belgium as Koninklijk
Besluit nr 4 of 20 December 1969 on refunds in respect of VAT (Royal Decree No 4) 5. For the purposes of
the present references, the two most important amendments concerned sub-paras (4) and (5) of art 8/(1)
(3).
4
See Belgisch Staatsblad (vierde uitgave), 30 December 1992. The French-language text was subsequently amended
and is now set out in the version inserted by art 6 of the Royal Decree of 14 April 1993 (see Moniteur Belge, 30 April
1992). The amended Royal Decree No 4 became effective on 1 January 1993 in accordance with art 13 of the Royal
Decree of 30 December 1992.
5
See Belgisch Staatsblad, 31 December 1969.

7.
However, it is important to note that, under sub-para (1) of art 8/(1)(3) of Royal Decree No 4, any
excess claimed as being due by a taxable person based upon his return is deemed to be claimed only
subject to payment of any unpaid tax debts. Article 8/(1)(3), sub-para (4) provides:
 If the tax debt referred to in the first paragraph does not constitute, in favour of the
administration, a claim which is, in whole or in part, certain, payable and definite, which is inter alia
the case where it is disputed or has given rise to a distress warrant within the meaning of Article 85
of the Code execution of which is opposed by an objection within the meaning of Article 89 of the
Code, the tax credit shall be retained by the Administration up to the amount of the tax claimed.
That retention shall take effect as a preventive attachment until the dispute has been definitively
resolved, either in the administrative procedure or by a final court judgment. The condition laid
down by Article 1413 of the Judicial Code shall be deemed to have been satisfied as regards the
implementation of that retention.6

6
Article 1413 of the Judicial Code would appear to concern the onus of establishing the urgency of interim measures
when they are sought.

Article 8/(1)(3), sub-para (5) provides:

If, with regard to the balance refundable resulting from the declaration referred to in Article
55(1)(3) of the Code, and in respect of which the taxable person has or has not opted for a refund,
there are either serious grounds for suspecting or evidence that the aforesaid declaration or
declarations concerning previous periods contain inaccurate information and if those grounds or
evidence point to the existence of a tax debt the actual existence of which cannot, however, be
established before the time for the payment 66 order or for the operation equivalent to payment, no
payment order shall be made in respect of the balance nor shall the balance be carried forward to
the following tax period, and the tax credit shall be retained in order to permit the administration to
verify the accuracy of the information.
The effect of the application of a retention or withholding (inhouding) under either sub-para (4) or (5) is to
attach provisionally the VAT debt which would otherwise either be carried forward to the next VAT period
or refunded to the taxable person in accordance with art 8/(1)(2).
8.
Under sub-para (6), the serious presumption or evidence referred to in sub-paras (5) must be
established by processen-verbaal (official minutes)7 drawn up in accordance with art 59(1) of the code 8.
Sub-paragraph (7) provides that the retention referred to in sub-paras (4) and (5) shall have the effect of a
preventive attachment and shall continue until the proof contained in the minutes referred to in sub-para
(6) is refuted or until the accuracy of the relevant transactions emerges from information obtained from
other member states under the co-operation mechanisms established by Community law. Sub-
paragraphs (8) and (9) deal with the notification and entry into effect of a preventive attachment. The
affected taxable person may contest the attachment in accordance with art 1420 of the Judicial Code
before a beslagrechter (a judge hearing attachment proceedings)9. Nevertheless, a beslagrechter is not
permitted under sub-para (10) to order the lifting of such an attachment where the evidence contained in
the relevant official minutes has not been refuted, where information requested from other member states
has not been obtained, or during an investigation by either the Parket (the office of the Public Prosecutor)
or an onderzoeksrechter (an examining magistrate). Under sub-para (11), when the attachment of the tax
credit is lifted, it may be set off in accordance with sub-para (2), without any formalities being necessary,
against a certain, payable and definite tax debt due to the administration.
7
There is no precise English translation of the notion of a process-verbaal (procs-verbal being the equivalent term in
the French text of the Belgian legislation). It may variously be translated as an official report, record or minute, but it is
important to note its evedential value which is described in footnote 8 below. I have chosen for the purposes of this
opinion to adopt the translation official minute(s), since minutes is the term used in the translation of procs-verbal
that appears in the English version of art 47 of the Rules of Procedure of the Court of Justice.
8
Under art 59 of the code a proces-verbaal is presumed to be authentic until the contrary is proved
9
The equivalent expression in the French version of the decree is juge des saisies. For convenience, and having regard
to the fact that Dutch is the language of procedure, the Dutch expression will hereinafter be employed. The plural form,
beslagrechters, will, where necessary, be used.

II FACTUAL BACKGROUND AND PROCEDURE

A. The facts
9.
The factual circumstances of each of the references are different. Hence, in the interests of clarity, I
shall set out briefly an individual description of each case.
(i) Molenheide (Case C-286/94)
10.
The plaintiff company operates a garage business in Antwerp. It submitted a VAT declaration for the
period 1 January 1993 to 31 March 1993 claiming the right to a deduction in the sum of BF 2,598,398.
However, an inspection carried out at the plaintiffs premises led the local VAT administration to raise
serious doubts about the truthfulness of the VAT declaration. The 67 plaintiff was informed by an official
minute sent to it by registered letter on 15 June 1993 that, in view of the doubts which had arisen, the tax
collector was going to retain any ostensible VAT refund due on the basis of the plaintiffs declaration. On
16 June 1993 the plaintiff received a notice of retention from the relevant tax collector, alluding to serious
grounds for suspecting the accuracy of one or more of the declarations submitted by the plaintiff.
Essentially, the VAT administration suspects the plaintiff of having engaged in fictitious circular sales
which artificially created an apparent credit for the first quarter of 1993. The retention affecting the sum
claimed as a credit by the plaintiff was stated to be protective. In effect, it froze the sum pending a
definitive administrative or judicial determination of whether the supposed credit was genuine. The legal
basis of the retention was the amended version of art 8/(1)(3)(5) of Royal Decree No 4.
11.
On 23 July 1993, the plaintiff brought an unsuccessful appeal against the retention to the Rechtbank
van Eerste Aanleg te Antwerpen (the Court of First Instance, Antwerp) with a view to setting aside the
attachment. It then appealed to the Hof van Beroep te Antwerpen (the Court of Appeal, 13th Chamber,
hearing civil matters (the national court)). The plaintiff submitted that the attachment prescribed in art
76(1)(3) of the code and in Royal Decree No 4 was contrary to arts 18(4) and 27 of the Sixth Directive.
Having heard the opposing arguments of the Belgian state advocate, the national court decided to refer
the following question to the Court of Justice for a preliminary ruling:
On a proper construction of Article 18(4) of the Sixth Directive, may a Member State refrain
from refunding substantial VAT credits of its residents or carrying them forward to a following tax
period, and instead attach them as a protective measure under national rules owing to the
existence of serious grounds for suspecting tax evasion, without creating a definitive legal title in
that respect and without the Member States having received any authorisation under Article 27 of
the Sixth Directive?
(ii) Schepens (Case C-340/95)
12.
The plaintiff owns a garage business. He submitted a VAT return in respect of the period 1 January
1993 to 31 March 1993 claiming a refund of an alleged credit of BF 3,311,438. Inspections of the plaintiffs
accounts effected in May 1993 by a chief inspector and an auditor from the VAT administration resulted in
the administration forming the view that there were serious grounds for doubting the accuracy of the
plaintiffs return and evidence that the return contained false and/or incomplete particulars. Accordingly,
on 15 June 1993 an official minute containing the findings of the inspections was drawn up. In a
statement sent by registered letter on 16 June 1993, the plaintiff was informed of the administrations
findings. He was also given a copy of the official minute and informed that, under the fifth sub-paragraph
of art 8/(1)(3) of Royal Decree No 4, as amended, it was intended to proceed with a retention of any
possible refund due to the plaintiff. The relevant withholding notice was drawn up and sent to the plaintiff
on 18 June 1993. A similar procedure was followed concerning the plaintiffs VAT return for the second
quarter of 1993, which, on its face, indicated a VAT credit of BF 2,419,078. Following an inspection
carried out on 15 September 1993, an official minute was prepared on 20 September 1993 which again
indicated that there were serious reasons for suspecting that the return was based on false particulars.
The plaintiff was informed of this second official 68 minute by registered letter on 22 September 1993 and
a withholding notice based on art 8/(1)(3)(5) of Royal Decree No 4 was sent to him on the same day 10.
10
A further official minute concerning the plaintiffs second quarterly return, containing additional findings, was drawn up
on 26 October 1993.

13.
The VAT administration essentially suspects the plaintiff of having participated in fraudulent circular
sales. In his case, this allegedly comprised buying various (expensive) cars on foot of invoices showing
considerable amounts of VAT to be payable as part of the purchase price paid by the plaintiffbut in
respect of which the administration had no record of having ever received the relevant sums of VAT from
any of the plaintiffs suppliersand selling the same cars to buyers located in other member states on
foot of invoices showing that no VAT was paid. The result was that the plaintiffs VAT returns showed
substantial tax credits due to him, although he was unable to prove, for example, that any of the cars sold
to non-Belgian purchasers ever, in fact, left Belgium.
14.
The plaintiff appealed to the beslagrechter at the Rechtbank van Eerste Aanleg te Antwerpen, who, by
decision of 8 March 1994, refused to lift the attachment. The plaintiff then appealed to the Hof van Beroep
te Antwerpen (Third Chamber, hearing civil appeals), which, in the light of the conflict between the plaintiff
and the Belgian state advocate regarding the compatibility of the Belgian legislation with arts 18(4) and 27
of the Sixth Directive, referred the following questions to the Court of Justice:
(1) Do Articles 18(4) and 27 of [the Sixth Directive] have direct effect in the national legal
systems of the Member States and thus in Belgian law?
(2) If so, does Article 18(4) preclude a Member State from refusing to refund to a taxable person
a VAT credit in relation to a specific period or periods during which that credit arose or to carry it
over to a subsequent tax period, and instead withholding it by means of the Belgian withholding
procedure, which has the effect of a preventive attachment within the meaning of Article 1445 of the
Belgian Judicial Code, as long as no definitive entitlement has arisen in that regard and only up to
the amount of the demand relating to that tax period or earlier periods, where the demand is
disputed by the taxable person?
(3) Is Article 18(4) applicable, given that, according to the Belgian State, such withholding is a
debt-recovery procedure? If so, is Article 27 applicable if such withholding were to form part of the
conditions (modalits)? If not, is Article 27 applicable, on the assumption that such withholding is
a debt-recovery procedure?
(4) If Article 18(4) is applicable to the Belgian withholding procedure, does that procedure
infringe the principle of proportionality as defined by the Court of Justice?
(iii) Decan (Case C-401/95)
15.
Unlike the Molenheide and Schepens cases, this case concerns an attachment effected on the basis
of art 8/(1)(3)(4) of Royal Decree No 4. The order for reference indicates that by registered letter of 26
September 1995 the VAT administration informed the company Decan (the plaintiff) that, pursuant to art 8/
(1)(3)(4), it was going to retain the supposed VAT credit of BF 705,404 owing to the plaintiff on foot of its
VAT return for the period 1 June to 30 June 1995. The plaintiff applied to the Rechtbank van Eerste
Aanleg te Brussel to set aside the 69 attachment. According to the national court, the attachment related
to a sum which is disputed by the plaintiff, noted in an official minute of 26 May 1994, in respect of which
the VAT administration served a distress warrant on 10 October 1995 for an amount of BF 784,305 plus
BF 130,500 by way of a penalty and BF 232,064 being interest calculated to 20 October 1995.
16.
It appears from the written observations of the plaintiff and the Belgian state that the plaintiffs earlier
alleged VAT debt related to the period 1 September 1990 to 30 August 1992. The state alleges that the
plaintiff used an inflated pro rata method for calculating the deductions claimed. A regularisation notice
asserting that the plaintiff in fact owed the sum of BF 784,306 (and a fine and interest) was issued on 30
August 1993 by the VAT administration. However, it was formally contested by the plaintiff on 16
September 1993. The administration, consequently, drew up on 26 May 1994 the official minute referred
to by the national court. An attachment order concerning the sum of BF 705,404 was ultimately adopted
pursuant to art 8/(1)(3)(4) of Royal Decree No 4 on 26 September 1995 (and apparently notified the same
day)11.
11
The written observations of the Belgian state refer also to another (first) art 8/(1)(3)(4) attachment order of 15 June
1994, which was notified by registered letter to the plaintiff on 16 June 1994, and which concerned a sum of BF 118,984
standing to the credit of the plaintiff in a special account for the period 31 January 1992 until 30 April 1992.

17.
On 13 October 1995, the plaintiff brought its application before the national court which, in the light of
the Community law arguments, decided to refer the following questions to the Court of Justice:
(1) Must Article 18(4) of the Sixth Directive be interpreted as permitting a Member State to
refuse to refund a VAT credit from a specific tax period or to carry it forward to a following period,
yet to retain it on the ground that, and for so long as, it has a claim against the taxpayer in question
relating to a previous tax period, if that claim is disputed by the taxpayer and thus does not yet
constitute a definitive title, where the Member State has not received any authorization under
Article 27 of the directive?
(2) If question 1 is to be answered in the affirmative, must Article 18(4), in conjunction with the
principle of proportionality, be interpreted as permitting the Member State to lay down that the
necessity or urgency of the retention may not be contested in any way and that the retention may in
no way be replaced by a guarantee or annulled so long as the disputed VAT claim has not been
made the subject-matter of a final judicial decision?
(iv) Sanders (Case C-47/96)
18.
Notwithstanding an initial allegation of fraud, this case also concerns an attachment effected under art
8/(1)(3)(4) of Royal Decree No 412. According to the order for reference, an official minute established by
the bijzondere belastingsinspectie (special tax inspectorate) on 30 January 1992 alleges that the plaintiff
company is liable to the Belgian state for the sum of BF 370,791 in VAT (plus a fine of BF 741,582 and
interest from 21 January 1988). This relates, firstly, to the purchase from another company without
invoices of 227,000 kg of flour, and, secondly, to having acted as agent in the supply by that other
company of 403,710 kg of flour to third parties. These activities allegedly took place in 1987. 70The
plaintiff was notified by registered letter of 23 November 1994 that the relevant tax collector was about to
effect a retention concerning a credit balance of BF 236,215 in favour of the plaintiff on 31 October 1994.
It is clear from the written observations of the plaintiff and the Belgian stateto which copies of the
relevant notice and attachment are annexedthat the state effected this retention on the basis of art 8/(1)
(3)(4)13. On 5 January 1995 the plaintiff applied, inter alia, to have the attachment set aside. The
application was referred to the beslagrechter at the Rechtbank van Eerste Aanleg te Brugge, who, having
regard to the parties differing arguments regarding Community law and noting the pending reference in
Molenheide, decided to refer the following questions to the Court of Justice:
12
The plaintiff states that criminal proceedings were initially instituted against the managers of Sanders. However, the
prosecuting authorities later decided to seek a non-suit, which was granted by court order of 10 December 1991. It was
only then that attachment proceedings were initiated.
13
The plaintiff points out in its written observations that on 7 March 1996 the VAT administration also retained (apparently
on the basis of art 8/(1)(3)(4) of Royal Decree No 4) the sum of BF 121,106, which was the sum standing to the credit of
the plaintiff in respect of December 1995, in respect of the same alleged debt as is at issue in the main proceedings in
Sanders.

(1) Must Article 18(4) of [the Sixth Directive] be interpreted as permitting a Member State,
instead of refunding to a taxable person a VAT credit for a given tax period, or carrying it forward to
a subsequent tax period, to withhold the same by way of protective attachment on the basis of an
additional demand in respect of an earlier tax period, where that additional demand is contested in
law and is thus not based on any definitive entitlement, and where the Member State has not
obtained authorization pursuant to Article 27 of [the directive]?
 (2) In the event that question 1 is answered in the affirmative: Do the principle of proportionality
enshrined in Community law, and Article 18(4) of [the Sixth Directive] permit the Member State to
provide: (1) that the taxable person may contest the attachment (as validated by the withholding
measure) only by adducing evidence rebutting the allegations made by the Treasury in the official
report, and not by challenging the actual need for, and urgency of, that measure; and (2) that
withholding may not be replaced by another form of security nor lifted pending the delivery of final
judgment on the contested demand for payment made by the Treasury?

B. Procedure before the court

19.
Since the written procedure in Molenheide was already at an advanced stage when the reference in
Schepens was made and since the community law issues raised in both cases were very similar, the
procedure in Schepens was suspended pending judgment in Molenheide. However, during the oral
hearing in Molenheide (the first hearing), counsel representing the plaintiff contended that the practice of
the Belgian VAT administration and the lack of discretion which beslagrechters perceived themselves as
having with regard to setting aside attachments, coupled with the imminent arrival of two further
preliminary references concerning another type of retention operated in Belgium by the VAT
administration, cast doubt on the proportionality of the entire system of retention under the new version of
art 8/(1)(3) of Royal Decree No 4. The plaintiff stated that it would have no objection to the delay which a
joinder of the various references would entail. Consequently, by orders of 22 March 1996, the court
ordered the re-opening of the oral procedure in Molenheide, while the President ordered, firstly, the
joining of the three other references for the 71 purposes of the written and oral procedures and judgment,
and, secondly, the joining of Molenheide to the other cases in respect of the oral procedure and
judgment14. For convenience, where the context so requires, the various national courts and plaintiffs in
the four references will collectively be referred to respectively as the national courts and the plaintiffs.
14
These orders were adopted, respectively, on the basis of arts 61 and 43 of the Rules of Procedure.

III SYNTHESIS OF THE QUESTIONS REFERRED

20.
Not surprisingly, there is a considerable overlap between the questions referred in the four cases. In
my opinion, the essential Community law issues raised are threefold. They concern the direct effect of
arts 18(4) and 27 of the Sixth Directive; the availability and scope of the preventive retention in the
circumstances provided for in art 8/(1)(3) of Royal Decree No 4 and its compatibility with the right of
deduction conferred on taxable persons by art 18(4) of the Sixth Directive; the extent of such a retention,
and particularly whether the apparently limited interim bases upon which it can be suspended are
compatible with the Community law principle of proportionality.

IV OBSERVATIONS
21.
Written and oral observations were submitted on behalf of all of the plaintiffs, with the exception of Mr
Schepens. Written observations were submitted by the Kingdoms of Belgium 15 and Sweden, the Hellenic
and Italian Republics and the European Commission, all of whom, save Sweden, also submitted oral
observations16. Those observations may, for convenience, essentially be summarised as follows.
15
The Belgian state is actually formally the defendant in each of the references. However, for convenience, I shall simply
hereinafter refer to the defendant as Belgium.
16
Greece and Sweden presented written observations on the Schepens, Decan and Sanders cases, whereas Italy, which
had initially submitted written observations on the Molenheide case, also submitted written observations on Schepens
and Decan.

A. The plaintiffs
22.
The plaintiffs contend that the provisions of art 18(4) of the Sixth Directive are sufficiently clear, precise
and unconditional to be capable of direct effect within the national legal orders of the member states.
Furthermore, they contend, essentially, that the retention provided under the Belgian legislation is
incompatible with the right conferred by art 18(4). Once the right to a deduction arises in accordance with
the provisions of arts 17 and 18(1) to (3), national VAT administrations may not subject it to any other
conditions but must either make a refund or carry the excess forward to the next VAT period.
23.
Decan and Sanders also contend that under art 18(2) of the Sixth Directive the balance of deductible
VAT must be calculated by reference to a specific deduction period. Hence the administration cannot
maintain that no reimbursement is due because, in relation to a wholly separate period, the taxable
person is allegedly indebted to it. They point out that the administration is not obliged to accept the
amount of the excess indicated unilaterally in the taxable persons declaration but, on the contrary, may
examine whether that declaration is correct17. In addition, all the plaintiffs submit that, although the first
paragraph of art 18(4) permits the member states to determine the conditions concerning 72 the refund,
they are not permitted to subject the right to that refund to supplemental substantive conditions. They
contend that the English text, which uses the word conditions, digresses from the other language
versions where words corresponding to the notion in English of details or arrangements are used 18. In
the light of the need to adopt a harmonious Community law interpretation of the provision, they contend
that, notwithstanding the possible ambiguity of the English text, member states have no competence to
lay down substantive conditions governing the refund, the latter being laid down exhaustively in arts 17
and 18(1) to (3). Accordingly, the plaintiffs submit that the retention system operated in Belgium
constitutes an impermissible additional condition. Moreover, even if the Belgian withholding system falls
within the powers of the member states concerning the administration of the VAT system, the plaintiffs
submit that it must, none the less, respect the Community law principle of proportionality 19. They contend
that the system constitutes a disproportionate limitation on the right to deduct guaranteed by the Sixth
Directive. Firstly, this is so because the authorities are permitted to attach, almost automatically,
outstanding balances in favour of the taxable person without having to prove necessity and without being
obliged to accept alternative securities or undertakings from the taxable person in circumstances where
no damages are paid if the attachment is ultimately found to be unjustified, and because, in that
eventuality, interest is only paid with effect from 1 April of the year following that when attachment was
exercised. Secondly, the taxable person has no effective remedy against the attachment, because the
beslagrechter has no competence to make a genuine assessment of either necessity or its merits. At the
second oral hearing (the second hearing), the plaintiffs contended that even if, contrary to the view
adopted in most, if not all, of the relevant Belgian case law and based on the apparently unambiguous
wording of Royal Decree No 4, the beslagrechter could, as contended by Belgium, suspend the
attachment pending judgment on its merits, the uncertainty of such a remedy would infringe the principle
of legal certainty.
17
See R v Customs and Excise Comrs, ex p Strangewood Ltd [1987] STC 502 which was cited in support of the
proposition that the VAT authorities must carry out this examination within a reasonable period.
18
Reference is made to the Dutch (regeling), German (einzelheiten), French (modalits), Italian (modalit) and Spanish
(modalidades) texts. The English version uses the same word, conditions, in both art 18(3) and (4), whereas several of
the other language versions use different words in each of these provisions.
19
Reference is made, amongst many others, to the judgments concerning VAT in EC Commission v France Case C-
276/91 [1997] STC 584, [1993] ECR I-4413 and Balocchi v Ministero delle Finanze dello Stato Case C-10/92 [1997]
STC 640, [1993] ECR I-5105.

B. Belgium and the other member states

24.
Belgium, with the support of Greece, Italy and Sweden, submits that, although art 18(4) deals with VAT
refunds, a retention of tax credits merely relates to the collection of taxes and falls outside the scope of
the Sixth Directive. It states that art 76(1) of the code deals separately with the different legal concepts of
refund and retention of tax credits. Whereas a tax refund is regulated by art 76(1)(1) and (1)(2) and must
satisfy certain substantive and formal conditions, the retention of tax credits is governed by art 76(1)(3) of
the code and by Royal Decree No 4 and constitutes an autonomous legal concept. In its view, the Sixth
Directive has two essential purposes: (i) the harmonisation of taxes; (ii) the mutual integration of
economies in order to achieve and expedite the common market for the free movement of persons,
goods, services and capital. That harmonisation is achieved by determining a common scope for the
application of 73 VAT throughout the Community. Each member state may establish its own collection
system while ensuring that those objectives are achieved.
25.
Member states are free to regulate the collection of VAT, including measures to combat tax evasion or
fraud. Belgium relies on art 22(8) of the Sixth Directive, which permits member states to adopt necessary
VAT levying and collection measures. This interpretation accords with the principle of subsidiarity
expressed in art 3B of the EC Treaty. In the alternative, if the Sixth Directive were to be interpreted as
covering matters relating to collection, Belgium submits that the retention at issue in the present
proceedings falls within the scope of art 18(4), which gives member states a wide discretion when
determining the conditions governing refunds20.
20
In Molenheide, Belgium maintained that it was this discretion which precludes art 18(4) from having direct effect. In its
observations on the three subsequent references, this contention was not (at least expressly) maintained.

26.
The retention procedure does not operate so as to render practically impossible the exercise by
taxpayers of the rights conferred by art 18(4) of the Sixth Directive, but, rather, merely postpones the
repayment of overpaid VAT until the veracity of the declaration upon which the putative overpayment is
based can be verified. It does not infringe art 18(4), because it merely attaches provisionally a sum which
is still recognised as being an asset of the taxpayer and which remains subject to restitution, unless the
taxpayers declaration is ultimately shown to be false or inaccurate. Accordingly, Belgium submits that art
27 of the Sixth Directive is of no relevance since the collection of VAT does not fall within the scope of the
directive.
27.
Belgium developed the above observations and its explanations of art 8/(1)(3) in its written
observations on the later references. The traditional attachment procedure, since it does not permit
provisional attachments, did not offer the administration sufficient protection in circumstances where
putative tax credits had either to be refunded or carried forward within three months of the relevant VAT
declaration, even where there was a disputed debt owing to the VAT administration in respect of an earlier
period. Without actually disputing the veracity of the alleged tax credit, the new preventive attachment
renders that credit provisionally unavailable. Its purpose is, thus, clearly to ensure that VAT, which may be
owing by the taxable person, is in fact paid. At the second hearing, Belgium, supported by Italy and
Greece, asserted that, to ensure the proper collection of VAT, fiscal authorities should be entitled to view
their relationship with the taxable person as constituting a form of current account, since to view each tax
period separately would ignore the fact that a current (even undisputed) credit might only have arisen
because of earlier false or inaccurate declarations.
28.
Belgium accepts that tax collection measures must not effectively undermine rights conferred by the
Sixth Directive but, with the support of Italy, Greece and Sweden, contends that the impugned Belgian
measures respect the principle of proportionality. Responding to questions posed during the first hearing,
Belgium maintained that, whatever might have been the prevailing practice, beslagrechters could,
pursuant to Royal Decree No 4, set aside, on an interim basis, even attachments based on art 8/(1)(3)(5)
if they were not satisfied about the allegations concerning fraud or substantive irregularity. Belgium
subsequently elaborated upon those initial observations. It submits that, under sub-para (4) of art 8/(1)(3),
not only is the beslagrechter obliged to verify 74 compliance with all the applicable procedural
safeguards21, but he may also assess the prima facie reliability of the allegations grounding the retention.
He has, as Belgium put it, a right to engage in a marginal review of the attachment. As to attachments
based on serious presumptions of fraud under sub-para (5) of art 8/(1)(3), Belgium accepted that the
power of a beslagrechter to lift such a retention is considerably more circumscribed; in effect, no lifting
may be ordered unless the beslagrechter, having apprised himself of all the relevant information
concerning the case, is satisfied that its maintenance is no longer justified 22. At the second hearing,
Belgium insisted, however, that this did not preclude the grant of a provisional lifting of a retention granted
under sub-para (5). It contended, generally, that, despite the perhaps misleading wording of sub-para (7)
of art 8/(1)(3), an attachment could be lifted provisionally by a beslagrechter.
21
eg the investigation and ultimate decision to attach are not taken by the taxable persons local VAT office, but by
specialised outside officials of the VAT administration.
22
Although sub-para (5) does not use the word fraud, counsel for Belgium submitted at the second hearing that acts of
retention based on sub-para (5) were, in practice, limited to cases of fraud. They concern suspected inaccurate
information resulting in the possibility of a tax debt, whose extent cannot readily or expeditiously be determined. This,
the counsel contended, would not be the case with potential debts based on mere inaccuracies or calculation oversights
in previous tax returns, the level of which could be calculated with relative ease.

C. The Commission
29.
The Commission contends that the Sixth Directive entrusts the fight against fraud to the member
states, which is evidenced by several references to their power to adopt such measures 23. Regarding art
18(4), and, in particular, the interpretation of the expression the amount of authorised deductions, the
Commission submits that the exercise of the right to deduct may be governed only by, the conditions laid
down by the Sixth Directive; it cannot be subject to approbation by national VAT authorities. However, the
Commission contends that the Belgian retention measures constitute tax collection measures and, thus,
fall outside the scope of the Sixth Directive. Consequently, Belgium was not obliged to seek an
authorisation pursuant to art 27(1).
23
It refers esp to arts 22(8), 13A(1), 13B(1), 14(1) and 15.

30.
Nevertheless, member states tax collection measures must not undermine the principles underpinning
the common Community system of VAT. The Commission draws an analogy with the case law of the court
concerning art 27 of the Sixth Directive24. Under that procedure, the Commission enjoys the advantage of
being fully apprised in advance of the proposed measures and is thus in a position to ensure that they
respect the principle of proportionality, which, it says, also applies to the collection measures involved in
this case25. In the Commissions opinion, the court should limit itself to establishing certain criteria whose
concrete application must be for the national court. In the case of attachments effected to secure an
earlier disputed tax debt, the measures may be presumed to be disproportionate. This should not,
however, be the case with attachments based on serious doubts or evidence of 75 fraud. In its written
observations, the Commission enumerates a number of criteria which, in its view, the national court
should consider when applying the proportionality principle. They are, essentially, the existence of
procedural guarantees at the pre-retention stage; the existence of an effective judicial remedy against the
retention; a relationship between the sum attached and the alleged tax debt; the possibility of the
acceptance by the authorities of alternative guarantees offering comparable security. In the light of the
contradictory observations submitted regarding the competence of the beslagrechter, the Commission
submitted at the second hearing that a relevant factor should be whether, in imposing the retention, the
VAT administration is obliged to take a substantive legal action regarding the alleged tax debts. However,
it argued that it would be overly formalistic to require national fiscal authorities to treat each VAT period
separately.
24
The Commission refers esp to EC Commission v Belgium Case 324/82 [1984] ECR 1861 and EC Commission v France
Case 196/85 [1987] ECR 1597.
25
Reference is also made to the case law concerning the proportionality of sanctions for the infringement of VAT
regulations (see Criminal Proceedings against Drexl Case 299/86 [1988] ECR 1213 and EC Commission v France Case
C-276/91 [1997] STC 584, [1993] ECR I-4413. At the second oral hearing, reliance was also placed on the recent
judgment in Pastoors v Belgium Case C-29/95 [1997] ECR I-285.

V ANALYSIS

A. Direct effect
31.
 In Schepens the court is asked expressly whether arts 18(4) and 27 of the Sixth Directive have direct
effect. In the light of the possible conflict between the system of preventive retention operated in Belgium
and the right conferred on taxable persons by art 18(4), the relevance of this question is clear and should,
in my opinion, be answered by the court26. However, the relevance of the direct effect or otherwise of art
27, which has not been invoked by Belgium, to the resolution of the Schepens case is less clear. I do not
think that the court need answer that question.
26
Although art 76(1) and (2) of the code appear to implement the right conferred by art 18(4) of the Sixth Directive in
Belgium, the scope of those provisions is limited by art 76(1)(3) of the code and by Royal Decree No 4.

32.
The court has consistently held that, even where the implementation period has expired, a non-
implemented or incorrectly implemented directive must be both sufficiently precise and unconditional
before its provisions can be directly invoked by an individual before the courts of a member state 27. The
right of individuals to rely upon the Sixth Directive has been consistently upheld in the case law of the
court28. In BP Supergas Anonimos Etairia Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece
Case C-62/93 [1995] All ER (EC) 684, [1995] ECR I-1883 (para 34) the court stated:
27
See, inter alia, Pubblico Ministero v Ratti Case 148/78 [1979] ECR 1629 and, more recently, Faccini Dori v Recreb Srl
Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325
28
See eg Becker v Finanzamt Mnster-Innenstadt Case 8/81 [1982] ECR 53 and BP Supergas Anonimos Etairia Geniki
Emporiki-Viomichaniki kai Antiprossopeion v Greece Case C-62/93 [1995] All ER (EC) 684, [1995] ECR I-1883.

It follows from that case law that, despite the relatively wide discretion enjoyed by member
states in implementing certain provisions of the Sixth Directive, individuals may effectively plead
before national courts the provisions of the directive which are sufficiently clear, precise and
unconditional.
The court ruled, inter alia, that the provisions of art 17(1) and (2) dealing with the right to deduct
satisfy the above-mentioned criteria and therefore confer rights on individuals which they may invoke
before a national court in order to challenge national rules which are incompatible with those provisions
(see 76[1995] All ER (EC) 684, [1995] ECR I-1883 (para 35)). Those provisions are closely linked with art
18(4), which, in my view, for the reasons stated below, satisfies the same criteria.
33.
Where a taxable persons authorised deductions exceed his tax liability, art 18(4) of the Sixth Directive
requires member states either to make a refund or carry the excess forward to the following period,
unless the amount of the excess is insignificant. In its written observations, Molenheide submits that the
text is absolutely clear. Once the sum involved is not insignificant, the tax authority is obliged to do one of
two things. Although Belgium initially contested the direct effect of art 18(4) before the national court in
Molenheide, that position was not maintained in Belgiums observations to this court. Responding to the
express question referred in Schepens, the Commission, quite correctly, submits that art 18(4) is
sufficiently clear and precise to have direct effect and may be invoked directly before national courts.
34.
Differences of opinion as to what may constitute an insignificant overpayment cannot, in my opinion,
prevent a national court from being able to identify in most cases a significant tax excess which must be
refunded or carried forward. Moreover, since the objective of the Sixth Directive in respect of deductions
is to ensure that the right to deduct shall arise at a time when the deductible tax becomes chargeable
(see art 17(1) of the Sixth Directive), it is clear that the member states freedom to refuse a refund of
insignificant excesses is very limited. Accordingly, in my view, to invoke the second sub-paragraph of art
18(4) as a basis for denying direct effect to that article would confuse the issue of direct effect with that of
the discretion available to Member States in transposing the directive into national law (see the opinion of
Advocate General Mancini in Netherlands v Federatie Nederlandse Vakbeweging Case 71/85 [1986] ECR
3855 (para 3)).
35.
All of the plaintiffs submit that the discretion afforded to member states by the first sub-paragraph of art
18(4) of the Sixth Directive does not permit them to prescribe substantive conditions for the exercise of
the right to a refund. I think that the member states are merely permitted under the first sentence of art
18(4) to establish the necessary procedures or detailed arrangements concerning such refund 29 .. Indeed,
even the fact that a multiplicity of alternatives30 may be available for the purpose of implementing an
obligation imposed by a directive does not prevent it from having direct effect, once its content can be
determined sufficiently precisely on the basis of the provisions of the directive alone 31. Consequently, I
am satisfied that the obligation imposed by art 18(4) is clear, precise and unconditional and capable of
direct effect.
29
In Netherlands v Federatie Nederlandse Vakbeweging the court was asked whether art 4(1) of Council Directive (EEC)
79/7 on the progressive implementation of the principle of equal treatment for men and women in matters of social
security (OJ 1979 L6 p 24) was directly effective. Doubt centred on the directives non-discrimination principle, since
there were (at least) four different ways in which the Netherlands alone could have implemented it. The court, however,
rejected the view that such options detracted from the unconditionality of the obligation imposed by the relevant
provision
30
The phrase used by the court to describe Irelands objection to the direct effect of Directive 79/7 in McDermott v Minister
for Social Welfare Case 286/85 [1987] ECR 1453 (para 15).
31
See the judgment in Francovich v Italy Joined cases C-6/90 and C-9/90 [1991] ECR I-5357 (para 17)). In BP Supergas
[1995] All ER (EC) 684, [1995] ECR I-1883 the court held that arts 11A(1), 11B(1)(2) and 17(1)(2) of the Sixth
Directive were directly effective, notwithstanding the discretion accorded to the member states by art 11B(2) to adopt as
the taxable amount for imports the value defined in Council Regulation (EEC) 803/68 because they do not leave the
 member states any discretion as regards their implementation (see the opinion of Advocate General Jacobs: [1995]
All ER (EC) 684, [1995] ECR I-1883 (para 48); and the judgment: [1995] All ER (EC) 684, [1995] ECR I-1883
(para 35) respectively).

77

B. The scope of art 18(4) of the Sixth Directive

(i) Introduction
36.
The quality of neutrality has frequently been attributed to the VAT system (see eg the judgments in
Rompelman v Minister van Financin Case 268/83 [1985] ECR 655 (para 16) and EC Commission v
France Case 50/87 [1988] ECR 4797 (para 15)). Only the final consumer bears the total cost of the tax,
and the basis of the system is that VAT is chargeable on each transaction only after deduction of the
amount of value-added tax borne directly by the cost of the various price components (see the judgment
in Gaston Schul Douane Expediteur BV v Inspecteur der Inverroechten en Accijnzen Case 15/81 [1982]
ECR 1409 (para 10)). Each intervening trader passes on the VAT element in his purchases to the next
trader (or the consumer). In this system, the right to deduct VAT inputs is pivotal 32. In the long run, each
trader neutralises these inputs by deduction and recovers the tax on the added value attributed to himself
by adding it to the price charged to his customers. Cash flow may operate favourably or unfavourably for
the trader but, essentially, only temporarily. Where, for a given taxable period, his inputs exceed his
outputs, cash flow is adversely affected, since he is, albeit temporarily, required to bear the burden of the
VAT on his supplies until he can recover it in part from his customers and in part from the fiscal
authorities. Claims for repayment should be infrequent; they imply trading at a loss or, at least, on the
basis of a negative balance of input and output invoices, for a particular accounting period, with possible
serious cash flow implications. Article 18(4) of the Sixth Directive essentially allows member states two
options, once the right to a refund is established for a given taxable period. They may either make an
immediate repayment or they may postpone it until the end of the following taxable period. In many cases,
this postponement will eliminate the need for any repayment because the trader returns to a profit in that
subsequent period. The right conferred on taxable persons by art 18(4) prohibits member states,
however, from withholding repayment for more than one period and certainly from doing so indefinitely, in
the hope or mere expectation of the return of profitable trading.
32
It is described as a basic element of the system in the judgment in Rompelman v Minister van Financin Case 268/83
[1985] ECR 655 (para 16).

(ii) Opinion
37.
I do not think that the direct effect of art 18(4) of the Sixth Directive, taken on its own, is sufficient to
establish the incompatibility of the impugned Belgian measures. The plaintiffs rightly submit that the
foundation of the right to authorised deductions, within the meaning of art 18(4), is contained in arts 17
and 18(1) to (3) and that, in the absence of any provision empowering the member states to limit the right
of deduction granted to taxable persons, the taxpayer must be permitted to exercise that right
immediately in respect of all the taxes charged on transactions relating to inputs (see the judgment in
Lennartz 78 v Finanzamt Mnchen III Case C-97/90 [1995] STC 514, [1991] ECR I-3795 (para 27)).
Member states are, therefore, only authorised to limit the right of deduction where they may rely on one
of the derogations provided for in the Sixth Directive (see [1995] STC 514, [1991] ECR I-3795 (para 29)).
In case of an excess of authorised deductions over tax due, the neutrality of the Community VAT system
means that the taxable person has the right to a refund. However, member states are not precluded from
adopting precautionary measures designed to ensure the veracity of the apparent excess of deductions
arising from the information contained in the underlying declaration made by the taxable person. A system
of control designed to verify authorised deductions within the meaning of art 18(4) before making
payment is not a repudiation of the taxable persons right to deduct.
38.
In Jeunehomme v Belgium Joined cases 123/87 and 330/87 [1988] ECR 4517 (para 16) the court,
referring to art 22(8), held that member states, when requiring an invoice to contain certain information,
other than that required by arts 18(1)(a) and 22(3)(b), are not required to have recourse to the procedure
laid down in art 27 in order to ensure the correct levying of value-added tax and permit supervision by the
tax authorities33. If member states are entitled to specify the information to be contained in invoices
grounding a right to deduct, they should also be entitled to adopt precautionary measures designed to
ensure the proper collection of taxes.
33
Advocate General Sir Gordon Slynn expressly recognised (at 4533) that the Sixth Directive was not a complete
harmonisation of all rules relating to the administration of the VAT scheme and that the Council left it to Member States
to require other matters to be stated in the invoice which were necessary for the administration of the scheme, so long
as the provisions adopted pursued the aims of the Directive, did not create exceptions to the Community scheme and
did not limit its scope.

39.
This view finds further support in the recent judgment of the court in Reisdorf v Finanzamt Kln-West
Case C-85/95 [1997] STC 180, [1996] ECR I-6257. In that case, which again concerned invoices, the
court confirmed the consistency with the Sixth Directive of the power of the member states of ensuring
that VAT is levied and collected, under the supervision of the tax authorities (see [1997] STC 180, [1996]
ECR I-6257 (para 24)). Moreover, it also held that art 18, in accordance with its heading, deals only with
the exercise of the right of deduction and does not govern proof of that right after it has been exercised by
a taxable person (see para 26). I am satisfied that member states are competent, within the framework of
their national VAT collection systems, to adopt measures to protect themselves against the risk of making
repayments where no genuine VAT credit exists. The adoption of precautionary, control measures
enables the interests of the national treasury to be balanced against the interest of a taxable person in
obtaining expeditious payment of any amount to which he appears, prima facie, to be entitled under art
18(4). I think that this interpretation of art 18(4) is consistent with its reference to the member states
obligation to effect the deduction according to conditions which they shall determine. Alternatively, art
22(8) permits member states to impose other obligations which they deem necessary for the correct
levying and collection of the tax and for the prevention of fraud. It does not seem to me to matter whether
the Belgian rules fall outside the scope of the matters harmonised by the Sixth Directive, or fall in principle
within its scope but are authorised by art 22(8) thereof. In either event, they must be submitted to the test
of proportionality to the extent to which they are capable of infringing the directive (see paras 42 to 54,
below).
79
40.
At the second hearing, counsel representing Decan and Sanders submitted that art 18(4) must be
interpreted in the light of art 18(2), and that, accordingly, once a member state has verified that a tax
credit is genuinely owing in respect of a particular tax period, the member state is obliged to permit the
taxable person immediately to exercise the right to deduct in respect of that credit. I agree with the
Commission, whose agent described this approach as too formalistic. It is more realistic to regard the
relationship between the taxable person and the national VAT authorities as being akin to a current
account. The plaintiffs refer, inter alia, to EC Commission v France Case 50/87 [1988] ECR 4797 (para
16), where the court held that
in the absence of any provision empowering the Member States to limit the right of deduction
granted to taxable persons, that right must be exercised immediately in respect of all the taxes
charged on transactions relating to inputs.
In that case, the impugned French decree permitted only a fraction of the VAT charged on the purchase or
construction of immovable property to be deducted where the property was let for a rent which was less
than one-fifteenth of its value. The Belgian retention rules, by contrast, do not restrict the right to deduct;
they merely provide that payment of the refund may, provisionally, be postponed. If member states were
entirely precluded from postponing refunding even undisputed tax credits arising in one tax period
because of a serious dispute regarding an earlier period, their financial interests could often be
significantly jeopardised. Nor do I think that art 18(4) should be interpreted as limiting member states
right to oppose a refund to cases where they have obtained a final judgment in respect of the taxable
persons indebtedness regarding an earlier period, or of the fraudulent nature of the supposed VAT credit.
Moreover, I do not think that the compatibility of a provisional retention of a payment due under art 18(4)
should depend on proof of a causal link with a debt or underpayment for an earlier period.
41.
As the court acknowledged in BP Supergas, member states enjoy a wide discretion in implementing
the provisions of the VAT system. There are two aspects to this. Firstly, the routine details of provisions
relating to the collection of the tax, which are typically to be found in national law, are not regulated by the
Sixth Directive at all. I would say, almost as an aside, for example, that there could be circumstances in
which a member state, without putting in question the application of art 18(4), would not strictly speaking
repay an established credit. If there is a clearly established counter-debt for VAT due from the taxable
person with the result that there are clear and certain amounts due both from and to the taxpayer, it would
be pointless to make a repayment, when one can be set off against the other. This might arise if there
was an underpayment from an otherwise uncontested earlier VAT return either by accident or by reason
of shortage of funds. More generally, the responsibility for managing the entire VAT system is left to
member states. As already stated at para 39, above, it is not necessary to decide whether this arises
because those matters fall outside the scope of the Sixth Directive or are expressly authorised, for
example, by art 22(8). However, measures coming within the scope allowed by art 22(8) may inevitably
have an impact on the obligation of the member state to make an immediate repayment under art 18(4),
in which event those measures will (for the reasons set out below) need to be assessed for their
proportionality in accordance with 80 Community law; in other words, for their appropriateness in view of
the aim which they seek to achieve.

C. The application of the principle of proportionality

42.
If the operation of a retention of the type impugned in the present references does not, in itself, infringe
art 18(4) of the Sixth Directive, it does not necessarily follow that it is compatible with Community law.
Although member states may remain competent to determine their own systems of VAT collection, they
are nevertheless required to operate those systems in conformity with the Sixth Directive, and especially
its fundamental provisions such as the right of deduction.
(i) The aim of retention measures
43.
I think that it is clear that retention measures of the sort at issue in the present references pursue a
genuine aim. The member states who have submitted observations rightly observe that, since the benefit
of VAT receipts accrues almost entirely to the member states, they clearly have a legitimate interest in
taking appropriate steps to protect their financial interests. This, in my opinion, is essentially also what art
22(8) of the Sixth Directive recognises. As I stated at para 38, above, art 22(8) does not permit member
states to contravene, in the interests of preventing fraud or ensuring effective tax collection, other
provisions of the Sixth Directive. In this respect, art 27 expressly provides a procedure whereby member
states may apply to the EU Council for an authorisation to apply, inter alia, national measures designed to
counteract tax evasion or avoidance which conflict with the Sixth Directive. The court has consistently
held that such measures must respect the principle of proportionality. Thus, for example, in EC
Commission v Belgium Case 324/82 [1984] ECR 1861 (para 29) it held that measures notified pursuant to
art 27 must be of such a nature
as to prevent tax evasion or avoidance and that in principle they may not derogate from the
basis for charging VAT laid down in Article 11, except within the limits strictly necessary for
achieving that aim.34

34
It should be noted that the measures at issue in that case were in existence on 1 January 1977 and, thus, fell within the
scope of the transitional arrangement set out in art 27(5). There is no reason, however, to suppose that the court would
have adopted a different view had art 27(5) not been applicable (see, in this respect, my opinion in Finanzamt Bergisch
Gladbach v Skripalle Case C-63/96 [1997] STC 1035 (esp paras 2528)).

I think that the Commission is, therefore, correct in contending that national measures which, though not
based on an art 27 authorisation, may affect rights granted by the Sixth Directive must comply with the
principle of proportionality. In brief, where it is necessary to reconcile potential conflicts between the
application of national tax collection measures, such as preventive retention in Belgium, and the
fundamental right to deduct excess VAT guaranteed by the directive, the relevant national measures must
not go beyond what is necessary to achieve their aim.
44.
This view is supported by the case law of the court. Thus, for example, in Balocchi v Ministero delle
Finanze dello Stato Case C-10/92 [1997] STC 640, [1993] ECR I-5105, it held that the facility granted to
member states under art 22(5) to oblige taxable persons to make interim payments of VAT during the
course of a tax period did not permit them to require the payment of a fixed percentage 81(65%) of the
amount payable for an unexpired period. It follows that member states do not have unlimited discretion in
operating their tax collection systems.
(ii) Effective judicial control
45.
In such a situation, as a first step, the effectiveness of judicial control required by Community law
imposes an obligation on the authorities of the member states to give reasons for their decisions (see the
judgment in Union nationale des entraneurs et Cadres techniques professionnels du football (Unectef) v
Heylens Case 222/86 [1987] ECR 4097 (paras 1415)). In the present case, this function is performed by
the official minute which must be drawn up and duly notified to the taxable person in order that the VAT
authority may exercise its right to withhold its payment. In so far as Belgian law, in that respect, provides
for adequate notification and statement of reasons to be given to a taxable person whose right to refund is
delayed, it complies with that aspect of the requirements of Community law. At the second hearing, the
adequacy in practice of the pre-retention procedural safeguards was questioned by the counsel
representing Decan and Sanders. This, however, is a matter for the national court, which alone has the
jurisdiction and the capacity to interpret the relevant Belgian rules.
46.
The member states obligation to ensure effectiveness of judicial control requires, in addition, that
national courts be provided with an adequate range of powers to ensure a proper balance between the
rights of the taxable person and the rights and interests of the VAT authority. Of course, the nature and
extent of the judicial remedy is conditioned by the context and the type of danger against which a
measure is intended to protect. For example, in his joint opinion in Fritz Werner Industrie-Ausrstungen
GmbH v Germany Case C-70/94 [1995] ECR I-3189 (para 65) and Criminal proceedings against Leifer
Case C-83/94 [1995] ECR I-3231, Advocate General Jacobs, treating of an altogether different subject,
expressed the opinion that the nature of the issues that arise where the external security of a Member
State is at stake will usually prevent a court of law from adopting a strict proportionality test. In view of the
fact that assessments were made on the basis of intelligence information which could not be checked by
the court, he thought it difficult to ensure full judicial review. The present cases do not fall within that
exceptional category.
47.
Sub-paragraphs (4) and (5) of art 8/(1)(3) of Royal Decree No 4 cover a range of cases and practical
situations. At one extreme, covered by sub-para (5), there are serious allegations of fraud supported by
credible objective evidence. At the other, there will be cases of genuine disputed issues of fact and legal
interpretation leading, at most, to a disputed claim by the fiscal authorities. In the present sort of case,
where the court is, essentially, called upon to provide criteria to permit the national court, in performing its
task of judicial control, to assess whether the application of the impugned national measures is liable, in
practice, to undermine the effectiveness of the right conferred by art 18(4) of the Sixth Directive, the court
should not, in my opinion, lay down detailed rules. It should, instead, give broad guidance as to the nature
of the discretion which should be conferred on the relevant national courts.
(iii) Opinion
48.
Firstly, I would emphasise that Community law should not impose an inflexible need for the provision
of a full substantive review of the facts and merits of the claim of the fiscal authority at the stage of judicial
control of the withholding decision. As is apparent, in the present cases, from the terms, in 82 particular,
of sub-para (4) of art 8/(1)(3) of Royal Decree No 4, final determination of the disputed fiscal claim must
await a definitive court judgment. The latter implies a detailed examination of the contested factual and
legal issues with the possibility of appeals within the hierarchy of the national legal system. This
necessarily takes time and, indeed, some emphasis was laid at the second hearing, on behalf of
Molenheide, on the fact that many of these claims remain outstanding for years. That is not surprising.
The present case is, however, concerned with provisions for withholding of payment on a provisional
basis as a conservation measure. In its nature, this type of measure does not lend itself to final
determination of issues. I do not think that the beslagrechter can be expected to determine finally, but
rather only on an interim basis and as a matter of urgency, disputed and complex issues of fact and law.
49.
The beslagrechter should, however, in principle have power to order the lifting of the retention. In the
observations submitted, particularly at the second hearing, there was considerable disagreement between
counsel for the plaintiffs, on the one hand, and for Belgium, on the other, as to the extent or existence of
any discretionary power residing in the beslagrechter. Indeed, this situation of uncertainty seems to be
reflected to some extent in the varying behaviour of different beslagrechters. Some were described as
being very strict in the interpretation of their powers, while others appear to have been more liberal.
These matters of Belgian law cannot, of course, be determined by this court, which can merely indicate,
as stated above, the nature of the discretion, which must exist, and be known generally to exist, if due
protection is to be accorded to rights conferred by Community law.
50.
The fundamental test is, of course, that the measure taken be proportionate to the end to be achieved.
In other words, it should not go further than is strictly necessary in the pursuit of that purpose, namely the
protection of the VAT authority itself in its duty to ensure the financial integrity of the system and to collect
the tax. I would apply, by analogy, the principle enunciated by the court in EC Commission v Belgium
Case 324/82 [1984] ECR 1861 (para 29), in the case of a derogation pursuant to art 27 of the Sixth
Directive, that measures adopted to prevent tax evasion or avoidance may not derogate from the basis
for charging VAT laid down in art 11, except within the limits strictly necessary for achieving that aim. This
approach enables some broad principles of guidance to be adopted. I agree with the approach suggested
by the Commission at the second hearing; namely, that in cases such as those covered by sub-para (4) of
art 8/(1)(3) of Royal Decree No 4broadly speaking cases of disputed debtsthe presumption would be
against proportionality and that, in the same way, in cases such as those covered by sub-para (5),
especially those establishing serious presumptions of fraud, the presumption would operate in the
opposite direction.
51.
The decision to be made by a national court regarding the maintenance or lifting of a withholding
measure must, like all provisional measures, depend on the particular circumstances of the case. I shall
try to indicate the correct approach to a number of those circumstances. I think that the overriding interest
must be the need to protect VAT revenues. It would appear that the withholding provision found in art 8/(1)
(3) of Royal Decree No 4 reverses the normal requirement in Belgian law for the existence of urgency, ie
the urgent necessity to protect the revenue. In my view, the VAT authority must in all cases be able to
justify the measures taken on the basis of necessity; that is to say a genuine and urgent necessity to
protect tax revenues. In cases falling within sub-para (5), the official minute must establish, prima facie,
the existence of serious presumptions 83 of fraud or, alternatively, the existence of evidence of serious
underdeclarations for earlier periods to the effect that a debt in favour of the VAT authority can be
foreseen. It goes without saying that such an envisaged debt must be of such a size as to justify
withholding payment. Subject to that, the existence of evidence based on fraud or serious
underdeclaration (even if not fraud) could reasonably lead a national court to consider that fiscal interests
are in need of protection. That need must also, however, take account of any other relevant
circumstances. For example, the scale of the enterprise conducted by the taxable person and its financial
stability may be such as to satisfy a court that there is no need for the retention. If, on the other hand, that
enterprise is shown to be in danger of becoming insolvent or becoming bankrupt, it may reasonably be
said that, in an uncertain situation, a debt should not be paid to an enterprise which may enure to the
general body of creditors and not be available to satisfy the tax authorities claim.
52.
At the second hearing, the plaintiffs repeatedly contradicted Belgiums assertion that a beslagrechter
may set aside a retention, particularly in sub-para (5) type cases. It goes without saying that,
notwithstanding the apparent ambiguities of Belgian law, the Community law principle of proportionality
requires the subsistence of a genuine opportunity of judicial control for the taxable person. Thus, under
the principle of co-operation laid down in Article 5 of the Treaty, it is for the Member States to ensure the
legal protection which individuals derive from the direct effect of Community law (see the judgment in
SCS Peterbroeck, Van Campenhout & Cie v Belgium Case C-312/93 [1995] ECR I-4599 (para 12)). Since
there are no Community rules governing the collection of VAT, it is for the domestic legal systems
to designate the courts and tribunals having jurisdiction and to lay down the detailed procedural
rules governing actions for safeguarding rights which individuals derive from the direct effect of
Community law. (see para 12.)
In the circumstances of the present reference, Belgian courts and, in particular, beslagrechters, must
ensure that in balancing the rights of taxable persons under art 18(4) against the treasury, they do not
render virtually impossible or excessively difficult the exercise of rights conferred by Community law (see
para 12).
53.
Where the beslagrechter is concerned with sub-para (4) type cases, I am of the opinion that the
retention should not be allowed to stand unless there is convincing evidence of the need to protect the
national treasury. This would not be limited to cases of impending insolvency, though this would be an
important matter. Furthermore, I do not think that treasury interests require the adoption of withholding
orders that seek, in addition to protecting the payment of alleged tax debts (and interest thereon), also to
secure administrative penalties imposed in respect of such supposed debts. A retention of that sort would
constitute an unjustifiable interference with the exercise of the right to deduct.
54.
In my opinion, the beslagrechter must, on the basis of his appreciation of the manifold elements of fact
which he may expect to find, be enabled to make a balanced judgment as to the necessity for the
withholding measure in the interests of the VAT administration, taking into account: the apparent strength
and seriousness of the evidence of fraud or underdeclaration; the likelihood of the due recovery of the
amount of the claim after complete determination of the dispute; the financial stability of the taxable
person; the length of time likely to elapse before resolution of the dispute; and giving due weight to the
need of the 84 taxable person for payment of the sum due to him for the continued conduct of his
business. As the court stated in the judgment in R v Secretary of State for Transport, ex p Factortame Ltd
Case C-213/89 [1991] 1 All ER 70, [1990] ECR I-2433 (para 21):
the full effectiveness of Community law would be just as much impaired if a rule of national
law could prevent a court seised of a dispute governed by Community law from granting interim
relief in order to ensure the full effectiveness of the judgment to be given on the existence of the
rights claimed under Community law.
Furthermore, a taxable person, who ultimately succeeds in the main action against the VAT authority,
must be entitled to effective compensation from Belgium, such as interest from the date the retained
refund would otherwise (ie but for the retention) have been paid.

VI CONCLUSION
55.
In the light of the foregoing, I recommend that the Court of Justice answer the various questions
referred by the three national courts, which have submitted questions in these four joined cases, as
follows:
(1) Article 18(4) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the
member states relating to turnover taxescommon system of value added tax: uniform basis of
assessment (the Sixth Directive) is sufficiently clear, precise and unconditional to be capable of
being invoked directly by taxable persons in the internal legal systems of the member states.
(2) National provisions that permit the preventive retention of VAT credits supposedly due to a
taxable person in respect of a particular tax period are not, in principle, incompatible with art 18(4)
of the Sixth Directive.
(3) VAT retention measures must, however, not go beyond what is strictly necessary to protect
the fiscal interests of the member states in ensuring the effective collection of VAT and national
courts must ensure that a taxable person whose exercise of the right to deduct conferred by art
18(4) of the Sixth Directive is affected by such retention measures is guaranteed effective
procedural safeguards prior to the adoption of the measure, which must include, inter alia,
reasonable notice of the reasons for the proposed retention and the opportunity of seeking effective
judicial control of the measure after its adoption by the VAT administration. In the case of a
retention of an undisputed VAT credit based on a disputed earlier VAT debt, national legal
provisions which presume the urgency and/or need for the withholding measure must be set aside
by the national court before whom an interim application to lift the retention is brought; such a court
must have the power to determine for itself on the basis of all the evidence available and all the
circumstances of the case, including the availability of effective alternative forms of protection for
the VAT administration, whether there is an urgent necessity for the retention. However, such
urgent necessity may not extend beyond securing the amount of the disputed debt (plus interest).
In the case of a retention based on serious suspicions of fraud or other serious irregularities in the
affairs of the taxable person claiming entitlement to a VAT credit, the court or judge hearing an
interim application for its suspension must be entitled, where it/he is not satisfied, on the basis of
contradictory evidence which must be produced by the taxable person, of the genuineness of the
VAT administrations doubts, to lift the retention on 85 such terms as it/he sees fit. In all preventive
retention cases, the VAT administration must undertake, in the event of the taxable person being
ultimately successful in the main action concerning the retention, to pay interest on the sum
retained from the moment when, in accordance with the normal deduction rules applied in that
member state in the implementation of the Sixth Directive, the sum would have been paid to that
taxable person.

18 December 1997.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By orders of 17 October 1994 (Case C-286/94), 25 October 1995 (Case C-340/95), 12 December
1995 (Case C-401/95) and 6 February 1996 (Case C-47/96), received on 21 October 1994, 30 October
1995, 21 December 1995 and 16 February 1996 respectively, the Hof van Beroep te Antwerpen (the
Court of Appeal, Antwerp), 13th and 3rd Chambers (Cases C-286/94 and C-340/95), the Rechtbank van
Eerste Aanleg te Brussel (the Court of First Instance, Brussels) (Case C-401/95) and the Rechtbank van
Eerste Aanleg te Brugge (the Court of First Instance, Bruges) (Case C-47/96) referred to the Court of
Justice of the European Communities for preliminary rulings under art 177 of the EC Treaty a number of
questions on the interpretation of art 18(4) of Council Directive (EEC) 77/388 on the harmonisation of the
laws of the member states relating to turnover taxescommon system of value added tax: uniform basis
of assessment (OJ 1977 L145 p 1) (the Sixth Directive).
2.
Those questions were raised in four actions brought against the Belgian state by Garage Molenheide
BVBA (Molenheide), Peter Schepens, Bureau Rik Decan-Business Research & Development NV (Decan)
and Sanders BVBA (Sanders).

The Community legislation

3.
Article 18(2) and (4) of the Sixth Directive, concerning procedures relating to the right of deduction,
provide:
(2) The taxable person shall effect the deduction by subtracting from the total amount of value
added tax due for a given tax period the total amount of the tax in respect of which, during the
same period, the right to deduct has arisen and can be exercised under the provisions of paragraph
1
(4) Where for a given tax period the amount of authorized deductions exceeds the amount of tax
due, the Member States may either make a refund or carry the excess forward to the following
period according to conditions which they shall determine.
However, Member States may refuse to refund or carry forward if the amount of the excess is
insignificant.

The Belgian legislation

4.
In Belgian law, art 18(4) of the Sixth Directive was implemented in particular by art 47 of the Value
Added Tax Code, which provides that, where the authorised deductions exceed the tax due for a
particular period, the excess is to be carried forward to the following tax period.
5.
The first sub-paragraph of art 76(1) of that code, as amended by the Law of 28 December 1992, adds
that any excess outstanding at the end of the calendar year is to be refunded in accordance with the
conditions to be established by the 86 King, on application by the taxable person. Pursuant to the second
sub-paragraph, the King may permit the grant of refunds even before the end of the calendar year. Finally,
according to the third sub-paragraph
with respect to the requirements laid down in the first and second sub-paragraphs, provision
may be made by Royal Decree for a retention in favour of the VAT, Registration and Property
Authority, having the effect of a preventive attachment within the meaning of Article 1445 of the
Judicial Code.
6.
That provision was implemented by art 7 of the Royal Decree of 29 December 1992, which inserted
into Royal Decree No 4 of 29 December 1969 on refunds in respect of VAT (Royal Decree No 4) an art
8/1(3) which is worded as follows:
If the tax debt referred to in the first paragraph does not constitute, in favour of the
administration, a debt which is, in whole or in part, certain, definite and due for payment, which is
inter alia the case where it is disputed or has given rise to an order for recovery within the meaning
of Article 85 of the Code, execution of which is opposed by an objection within the meaning of
Article 89 of the Code, the tax credit shall be retained by the administration up to the amount of the
tax claimed. That retention shall take effect as a preventive attachment until the dispute has been
definitively resolved, either in the administrative procedure or by a final court judgment. The
condition laid down by Article 1413 of the Judicial Code shall be deemed to have been satisfied as
regards the implementation of that retention [fourth sub-paragraph].
If, with regard to the balance refundable resulting from the return referred to in Article 55(1)(3) of
the Code, and in respect of which the taxable person has or has not opted for a refund, either there
are serious grounds for presuming or there is evidence that the aforesaid return or returns
concerning previous periods contain inaccurate information and if such grounds for presumption or
evidence point to the existence of a tax debt the actual existence of which cannot, however, be
established before the time for the payment order or for the operation equivalent to payment, no
payment order shall be made in respect of the balance nor shall the balance be carried forward to
the following tax period, and the tax credit shall be retained in order to permit the administration to
verify the accuracy of the information [fifth sub-paragraph]
The serious grounds for presumption or the evidence referred to in the foregoing sub-
paragraph, proving or indicating the tax debt, must be established by an official report drawn up in
accordance with Article 59(1) of the Code. The report shall be brought to the notice of the taxable
person by registered letter [sixth sub-paragraph].
The retention referred to in sub-paragraphs (4) and (5) shall have the effect of a preventive
attachment until the evidence contained in the report referred to in the foregoing sub-paragraph is
refuted or until the accuracy of the relevant transactions emerges from information obtained under
 the co-operation mechanisms established by the European Communities on exchange of
information between member states of the Community [seventh sub-paragraph]
87
The taxable person may only contest the attachment referred to in sub-paragraphs (4) and (5) in
accordance with Article 1420 of the Judicial Code. However, the court having jurisdiction in the
matter of attachments may not order the attachment to be lifted for so long as the evidence
contained in the report referred to in sub-paragraph (6) has not been refuted, particulars have not
been obtained by way of exchange of information between member states of the Community or an
investigation by either the Office of the Public Prosecutor or an examining magistrate is pending.
The retention shall cease when the attachment is lifted by the administration or by judicial decision.
If it is lifted by the administration, the taxable person shall be informed by registered letter indicating
the date on which it was lifted [tenth sub-paragraph].
Where the tax credit ceases to be retained, the tax debt constituting a debt in favour of the
administration which is certain, definite and due for payment shall if appropriate be discharged in
accordance with sub-paragraph (2), without any formality having to be completed [eleventh sub-
paragraph].
7.
Pursuant to art 1413 of the Judicial Code, to which the fourth sub-paragraph of art 8/1(3) of Royal
Decree No 4 refers, a preventive attachment may be carried out only in cases where prompt action is
required.
8.
According to the national courts, the retention provided for in the fifth sub-paragraph of art 8/1(3) of
Royal Decree No 4, which operates as a preventive attachment within the meaning of art 1445 of the
Belgian Judicial Code, is designed to block by way of a precautionary measure the refundable value
added tax (VAT) balance until proceedings concerning any sum which may be payable by the taxable
person in respect of VAT are concluded, either by administrative decision or by a judicial decision which
has become final or until the evidence or the serious grounds for presumption, referred to in the official
report, have been refuted or until the veracity of the transactions emerges from the information obtained
under the procedures laid down by the rules adopted by the European Communities concerning the
exchange of information between member states or from an investigation by either the Office of the Public
Prosecutor or an examining magistrate. The mechanism is essentially the same with regard to the other
retention provided for in the fourth sub-paragraph of that provision.

Molenheide (Case C-286/94)

9.
Molenheide runs a garage in Antwerp (Belgium). That company filed, for the period from 1 January
1993 to 31 December 1993, a VAT return in which it claimed entitlement to a deduction in the sum of BF
2,598,398.
10.
However, during a check carried out at its premises, the VAT authority discovered circumstances
giving rise to serious grounds for presuming that the return in question contained incorrect and incomplete
particulars.
11.
An official report was drawn up by the chief inspector of the main Wijnegem VAT office on the basis of
those findings and was notified to Molenheide by registered letter of 15 June 1993. The official report also
indicated that the relevant collector would effect a retention on the basis of it.
12.
On 16 June 1993 a retention notice was served on Molenheide by registered letter. In that notice, the
tax authority stated that there were serious grounds for presuming, and indeed evidence, that the above-
mentioned return contained incorrect particulars and that those grounds or evidence were 88 indicative of
a tax debt, the amount of which could not be properly determined at that time.
13.
The retention, which corresponded to the refundable amount arrived at on the basis of the VAT return
filed by Molenheide, was based on the fifth sub-paragraph of art 8/1(3) of Royal Decree No 4.
14.
On 23 July 1993 Molenheide contested the retention decision before the judge hearing attachment
proceedings in the Rechtbank van Eerste Aanleg te Antwerpen, maintaining that the fifth sub-paragraph of
art 8/1(3) of Royal Decree No 4 was invalid.
15.
By order of 4 November 1993, the judge hearing attachment proceedings declared the action
unfounded.
16.
On 24 December 1993 Molenheide appealed against that order to the Hof van Beroep te Antwerpen.
In those proceedings Molenheide claimed that the retention of tax credits, as provided for by the third sub-
paragraph of art 76(1) of the Belgian VAT Code and by the fifth sub-paragraph of art 8/1(3) of Royal
Decree No 4, was contrary to arts 18(4) and 27 of the Sixth Directive.
17.
 Uncertain as to how the latter provisions should be interpreted, the Hof van Beroep te Antwerpen
considered it appropriate to seek a preliminary ruling from the Court of Justice on the following question:
On a proper construction of Article 18(4) of [the Sixth Directive], may a Member State refrain
from refunding substantial VAT credits of its residents or carrying them forward to a following tax
period, and instead attach them as a protective measure under national rules owing to the
existence of serious grounds for suspecting tax evasion, without creating a definitive legal title in
that respect and without the Member State having received any authorisation under Article 27 of
[the Sixth Directive]?

Schepens (Case C-340/95)

18.
This case too is concerned with a retention under the fifth sub-paragraph of art 8/1(3) of Royal Decree
No 4, prompted by serious grounds for presumption of tax evasion.
19.
Mr Schepens owns a garage. He filed a VAT return for the period from 1 January 1993 to 31 March
1993 in which he claimed the right to refund of the sum of BF 3,311,438.
20.
Following a check carried out in May 1993 a chief inspector and an auditor from the VAT authority
drew up an official report on 15 June 1993 to the effect that there were serious grounds for presuming
that the VAT returns for the first quarter of 1993 contained incorrect particulars and gave grounds for
concluding that tax was payable. On 16 June 1993 the plaintiff was informed of the conclusions of the
inspection by registered letter. He also received a copy of the official report and the tax authority informed
him that it intended to retain the amounts that had been refundable. The retention notice was sent to him
on 18 June 1993.
21.
The tax authority followed the same procedure for the tax return for the second quarter of 1993, which
showed a credit of BF 2,419,078. After carrying out a check on 15 September 1993 it drew up an official
report on 20 September 1993, which it notified to the person concerned by registered letter of 22
September 1993, followed on the same date by a retention notice.
22.
Those serious grounds for presumption related in particular to a type of fraud known as circular sales,
not involving evasion of VAT but creating 89 fictitious VAT excesses, in particular on intra-Community
transactions. Thus, according to the Belgian administrative authorities, Mr Schepens sought to recover
amounts of VAT which he claimed to have paid when purchasing a number of vehicles. However, the
findings of the tax authority established that eight of his suppliers had not filed VAT returns for the first
quarter of 1993 or paid any VAT. Moreover, Mr Schepens had likewise not proved that he had paid them
the VAT, all the transactions having been conducted in cash or by cheque. Most of the vehicles had been
delivered outside Belgium but within the Community and at least some of them had been purchased more
than once in Belgium. For each transaction, the VAT indicated on the Belgian purchase invoice had not
been paid by the persons who issued the invoices and Mr Schepens had been unable to establish, by
evidence of the kind prescribed in art 3 of Royal Decree No 52, that he had in fact delivered the vehicles
outside Belgium but within the Community. For February and March 1993, the intra-Community
transactions carried out represented an amount of BF 11,625,000.
23.
In the case of the intra-Community deliveries, the VAT had not been accounted for on the outgoing
invoices and, under the VAT mechanism, the right to refund of the VAT mentioned on the corresponding
purchase invoices came into being. Moreover, there were grounds for presuming that those vehicles had
never left Belgium.
24.
Mr Schepens then applied for the lifting of the retentions or the preventive attachments carried out.
25.
His application was refused by the competent court of first instance, whereupon he appealed to the
Hof van Beroep te Antwerpen, claiming, on the basis of legal arguments similar to those advanced by
Molenheide, that art 18(4) of the Sixth Directive allowed a choice only between carrying the excess
forward to the following period and refunding it. Consequently, if it wished to follow another course, the
Belgian state should, pursuant to art 27 of the Sixth Directive, have sought authorisation from the EU
Council. Mr Schepens also invoked the principle of proportionality.
26.
The Hof van Beroep te Antwerpen then referred the following questions to the Court of Justice for a
preliminary ruling:
(1) Do Articles 18(4) and 27 of [the Sixth Directive] have direct effect in the national legal
systems of the Member States and thus in Belgian law?
(2) If so, does Article 18(4) of [the directive] preclude a Member State from refusing to refund to
a taxable person a VAT credit in relation to a specific period or periods during which that credit
arose or to carry it over to a subsequent tax period, and instead withholding it by means of the
Belgian withholding procedure, which has the effect of a preventive attachment within the meaning
of Article 1445 of the Belgian Judicial Code, as long as no definitive entitlement has arisen in that
regard and only up to the amount of the demand relating to that tax period or earlier periods, where
the demand is disputed by the taxable person?
 (3) Is Article 18(4) of [the directive] applicable, given that, according to the Belgian State, such
withholding is a debt-recovery procedure? If so, is Article 27 of the directive applicable if such
withholding were to form part of the conditions (modalits)? If not, is Article 27 applicable, on the
assumption that such withholding is a debt-recovery procedure?
(4) If Article 18(4) of the directive is applicable to the Belgian withholding procedure, does that
procedure infringe the principle of proportionality as defined by the Court of Justice?
90

Decan (Case C-401/95)

27.
In this case the retention was made on the basis not of the fifth sub-paragraph, but of the fourth sub-
paragraph, of art 8/1(3) of Royal Decree No 4.
28.
By registered letter of 26 September 1995, the tax authority informed Decan that on that date it was
effecting a retention or preventive attachment of the VAT credit of BF 705,404 resulting from its VAT return
for the period from 1 to 30 June 1995. That retention was made because of a VAT debt claimed by the
Belgian state for a period covered by an earlier return. Without giving further particulars of the debt
claimed, the national court states that it was recorded in an official report of 26 May 1994 and that it was
the subject of an order for recovery served on 10 October 1995 in respect of the sum of BF 784,305,
together with fines of BF 130,500 and interest of BF 232,064.
29.
Before the Rechtbank van Eerste Aanleg te Brussel, the parties put forward the same arguments as
those exchanged in the two other cases described above, and the national court has merely referred to
the orders relating to those cases. It adds, however, that whilst in the Molenheide case there were serious
grounds for presumption of tax evasion, the position is different in the Decan case.
30.
The Rechtbank van Eerste Aanleg te Brussel therefore referred the following questions to the Court of
Justice for a preliminary ruling:
(1) Must Article 18(4) of [the Sixth Directive] be interpreted as permitting a Member State to
refuse to refund a VAT credit from a specific tax period or to carry it forward to a following period,
yet to retain it on the ground that, and for so long as it has a claim against the taxpayer in question
relating to a previous tax period, if that claim is disputed by the taxpayer and thus does not yet
constitute a definitive title, where the Member State has not received any authorization under
Article 27 of the directive?
(2) If question 1 is to be answered in the affirmative, must Article 18(4) of [the Sixth Directive], in
conjunction with the principle of proportionality, be interpreted as permitting the Member State to
lay down that the necessity or urgency of the retention may not be contested in any way and that
the retention may in no way be replaced by a guarantee or annulled so long as the disputed VAT
claim has not been made the subject-matter of a final judicial decision?

Sanders (Case C-47/96)

31.
As in the Decan case, the retention was made pursuant to the fourth sub-paragraph of art 8/1(3) of
Royal Decree No 4.
32.
According to an official report of 30 January 1992, Sanders owes the Belgian state VAT in the sum of
BF 370,791 (together with a fine of BF 741,582 and interest as from 21 January 1988) for the purchase
without an invoice of 227,000 kg of flour from CERES NV and for involvement in the delivery of 403,710
kg of flour by the latter company to a third party. Those transactions were carried out in 1987.
33.
Sanders contested that debt, which is thus not certain, definite and due for payment within the
meaning of the fourth sub-paragraph of art 8/1(3) of Royal Decree No 4, whereupon the Roeselare VAT
collector, by registered letter of 23 November 1994, gave notice that it was retaining, by way of preventive
attachment in respect of the above-mentioned debt, the balance of the current account relating to its
periodical VAT return made up to 31 October 1994, namely BF 236,215.
91
34.
On 5 January 1995 Sanders instituted proceedings against the Belgian state for lifting of the
preventive attachment before the judge hearing attachment proceedings in the Rechtbank van Eerste
Aanleg te Brugge, relying on the same arguments as those put forward in the other cases, and on the
principle of proportionality, since in its view the retention was neither necessary nor the only measure
available.
35.
Uncertain as to how to interpret the Community provisions relied on, the judge hearing attachment
proceedings also decided to seek a preliminary ruling on the following two questions:
(1) Must Article 18(4) of [the Sixth Directive] be interpreted as permitting a Member State,
instead of refunding to a taxable person a VAT credit for a given tax period, or carrying it forward to
 a subsequent tax period, to withhold the same by way of protective attachment on the basis of an
additional demand in respect of an earlier tax period, where that additional demand is contested in
law and is thus not based on any definitive entitlement, and where the Member State has not
obtained authorization pursuant to Article 27 of [the Sixth Directive]?
(2) In the event that question 1 is answered in the affirmative: Do the principle of proportionality
enshrined in Community law, and Article 18(4) of [the Sixth Directive] permit the Member State to
provide: (1) that the taxable person may contest the attachment (as validated by the withholding
measure) only by adducing evidence rebutting the allegations made by the treasury in the official
report, and not by challenging the actual need for, and urgency of, that measure; (2) that
withholding may not be replaced by another form of security nor lifted pending the delivery of final
judgment on the contested demand for payment made by the treasury?

The preliminary questions

36.
In these four cases the national courts essentially wish to ascertain whether art 18(4) of the Sixth
Directive precludes measures such as those at issue in the main proceedings and, if not, what effect the
principle of proportionality might have in such circumstances.
37.
With regard, first, to art 18(4) of the Sixth Directive, the national courts ask essentially whether that
provision precludes national measures providing for the preventive attachment of a refundable VAT credit
where either there are serious grounds for presumption of tax evasion or there is a VAT debt claimed by
the tax authority, that debt being contested by the taxable person.
38.
The applicants consider that the retentions provided for in the fourth and fifth sub-paragraphs of art
8/1(3) of Royal Decree No 4 are incompatible with art 18(4) of the Sixth Directive since, where the VAT
excess is not insignificant, the national administrative authority may only choose either to make a refund
or to carry the excess forward to the period covered by the next return. Retention of the balance, which is
not covered by that choice, constitutes an outright negation of the taxable persons right to deduct VAT.
39.
The applicants also maintain that art 18(2) and (4) of the Sixth Directive refer to periods covered by
returns and infer that the Belgian authority may retain a VAT balance relating to a period other than the
period to which the dispute relates, an approach which, moreover, is consistent with the requirement for a
reasonable time limit.
92
40.
On the other hand, the Belgian, Greek, Italian and Swedish governments and the European
Commission maintain that the retentions provided for by the Belgian legislation constitute measures of
recovery and, as such, are not governed by the Sixth Directive or by the applicable Community legislation
but fall within the exclusive competence of the member states.
41.
Measures such as those at issue in the main proceedings are designed to enable the competent fiscal
authorities to retain, as a protective measure, refundable amounts of VAT where there are grounds for
presumption of tax evasion or where those authorities claim that there is a VAT debt owing to them which
is not apparent from the taxable persons returns and which the taxable person contests.
42.
It is clear from the Sixth Directive as a whole that it is intended to establish a uniform basis so as to
guarantee the neutrality of the system and, as indicated in the twelfth recital in its preamble, to harmonise
the rules governing deductions to the extent that they affect the actual amounts collected and to ensure
that the deductible proportion [is] calculated in a similar manner in all the Member States.
43.
It follows that Title XI of the Sixth Directive, which deals with deductions, and in particular art 18,
relates to the normal functioning of the common system of VAT and does not in principle concern
measures such as those described in para 41, above.
44.
The answer to be given must therefore be that art 18(4) of the Sixth Directive does not in principle
preclude measures of the kind at issue in the main proceedings.
45.
As regards, next, the effects which the principle of proportionality may have in this context, it must be
emphasised that whilst the member states may, in principle, adopt such measures, it is nevertheless the
case that those measures are liable to have an impact on the national authorities obligation to make an
immediate refund under art 18(4) of the Sixth Directive.
46.
Thus, in accordance with the principle of proportionality, the member states must employ means
which, whilst enabling them effectively to attain the objective pursued by their domestic laws, are the least
detrimental to the objectives and the principles laid down by the relevant Community legislation.
47.
Accordingly, whilst it is legitimate for the measures adopted by the member states to seek to preserve
the rights of the treasury as effectively as possible, they must not go further than is necessary for that
purpose. They may not therefore be used in such a way that they would have the effect of systematically
undermining the right to deduct VAT, which is a fundamental principle of the common system of VAT
established by the relevant Community legislation.
48.
The answer to be given in that regard must therefore be that the principle of proportionality is
applicable to national measures which, like those at issue in the main proceedings, are adopted by a
member state in the exercise of its powers relating to VAT, since, if those measures go further than
necessary in order to attain their objective, they would undermine the principles of the common system of
VAT and in particular the rules governing deductions which constitute an essential component of that
system.
49.
As regards the specific application of that principle, it is for the national court to determine whether the
national measures are compatible with Community law, the competence of the Court of Justice being
limited to providing the national court with all the criteria for the interpretation of Community law which
may enable it to make such a determination (see in 93 particular Gebhard v Consiglio dellOrdine degli
Avvocati e Procuratori di Milano Case C-55/94 [1996] All ER (EC) 189, [1995] ECR I-4165).
50.
In that connection, the plaintiffs submit, first, that the retention is absolute and is effected automatically
as soon as there is a dispute between the administrative authority and the taxable person. In their view,
the retention provided for by the fourth sub-paragraph of art 8/1(3) of Royal Decree No 4 is, by virtue of
the actual wording of that provision, compulsory whenever a tax debt is contested, this being a rule to
which there are no exceptions, and the court seised of the matter is not required to consider whether such
a retention is necessary or whether the matter is urgent, those conditions being irrebuttably presumed to
be satisfied. The same applies in their view to the retention provided for by the fifth sub-paragraph of the
same provision.
51.
It must be held that, where a preventive attachment procedure constitutes a derogation from the
ordinary law applicable to preventive attachments, in that necessity and urgency are irrebuttably
presumed, doubts may legitimately be entertained as to whether it is an indispensable instrument for
ensuring recovery of the sums due.
52.
It must therefore be held that an irrebuttable presumption, as opposed to an ordinary presumption,
would go further than is necessary in order to ensure effective recovery and would be contrary to the
principle of proportionality in that it would not enable the taxable person to adduce evidence in rebuttal for
consideration by the judge hearing attachment proceedings.
53.
Second, the applicants draw attention to the lack of any effective remedies both before the judge
hearing attachment proceedings and in the proceedings on the substance of the case. Without the
consent of the VAT authority, the judge hearing attachment proceedings is, according to the applicants,
never permitted, save where a formal requirement has been infringed, to lift in whole or in part the
retention of the refundable balance. That situation derives from the combined effect of various legal
provisions, for the most part derogating from the ordinary law relating to preventive attachments, which
provide that the judge hearing attachment proceedings may not order such a measure until such time as
the evidence contained in the official reports of the tax authority is refuted or the genuineness of the
transactions emerges from the particulars obtained through the Community procedures for exchange of
information between member states. The judge hearing attachment proceedings is thus concerned only
with the formal propriety of the preventive attachment procedure, not with the substantive conditions
governing the attachment.
54.
For the same reasons, where there is an appeal by the administrative authority against a decision
favourable to the taxable person, it is impossible to have the attachment lifted, even partially (eg in
respect of the fines), since the decision does not definitively dispose of the substantive issues. The
retention operates as a preventive attachment until the dispute has been finally determined, either by
administrative measure or by a judgment which has become definitive.
55.
In that connection, it must be observed that, in considering whether the adverse effect on the right of
deduction is proportionate, the availability of effective judicial review is necessary both in the proceedings
on the substance of the case and in those before the judge hearing attachment proceedings.
56.
Consequently, provisions of laws or regulations which would prevent the judge hearing attachment
proceedings from lifting in whole or in part the retention of the refundable VAT balance, even though there
is evidence before him which would prima facie justify the conclusion that the findings of the official 94
reports drawn up by the administrative authority were incorrect, should be regarded as going further than
is necessary in order to ensure effective recovery and would adversely affect to a disproportionate extent
the right of deduction.
57.
Similarly, provisions of laws or regulations which would make it impossible for the court adjudicating on
the substance of the case to lift in whole or in part the retention of the refundable VAT balance before the
decision on the substance of the case becomes definitive would be disproportionate.
58.
Third, the plaintiffs observe that it is impossible for the taxable person to request a court to adopt in
place of the retention a different protective measure which is sufficient to protect the interests of the
treasury but is less onerous for the taxable person, such as, for example, provision of a bond or a bank
guarantee. Such a possibility is open only to the tax authority and is entirely a matter for its discretion.
59.
It must be pointed out that such impossibility, if proved, would also exceed the bounds of what is
necessary to guarantee recovery of any sums due, in that the substitution in question might mitigate the
adverse effect on the right of deduction and the grant of such a measure should be amenable to review
by a court.
60.
Fourth, the plaintiffs observe that the retention is not limited to the principal amount due in respect of
VAT but also covers interest on it, procedural costs and penalties which may amount to as much as 200%
of the principal amount. That measure is thus disproportionate to the objective which it pursues, in
particular where the dispute concerns a question of pure law and not tax evasion in the strict sense.
61.
In that regard, it must be observed that the exercise of effective judicial review of the kind described
above, in particular if both the court adjudicating on the substance of the case and the judge hearing
attachment proceedings were able to grant the taxable person, at his request and at any stage of the
procedure, a total or partial lifting of the retention, would suffice to eliminate any lack of proportionality in
the calculation of the amounts retained, in particular as far as penalties are concerned.
62.
Fifth, the plaintiffs state that, under Belgian law, in the event of release of the retained VAT balances,
interest is not payable by the treasury unless the sums retained have not been duly returned by 31 March
of the year following that in which the refundable balances came into being and unless the amount
refundable is at least BF 10,000, the last VAT return for the calendar year in which the VAT credit arose
was signed at the place on the form indicated for that purpose and all the VAT returns have been filed
within the prescribed time limits.
63.
In that regard, it must be observed that it is not necessary, in order to attain the objective pursued by
legislation such as that at issue in the main proceedings, namely to ensure recovery of the amounts due,
for interest to be calculated from a date other than that on which the retained VAT balance would normally
have been paid under the Sixth Directive, and therefore that the principle of proportionality precludes the
application of such legislation. The same applies to the other conditions mentioned above: in particular,
lateness in filing returns can be penalised in a manner unconnected with the retention procedure and
without affecting the right to refund of the VAT balance.
64.
The answer to be given must therefore be that it is for the national court to examine whether or not the
measures in question and the manner in which they are applied by the competent administrative authority
are proportionate. In the context of that examination, if the national provisions or a particular 95
construction of them would constitute a bar to effective judicial review, in particular review of the urgency
and necessity of retaining the refundable VAT balance, and would prevent the taxable person from
applying to a court for replacement of the retention by another guarantee sufficient to protect the interests
of the treasury but less onerous for the taxable person, or would prevent an order from being made, at
any stage of the procedure, for the total or partial lifting of the retention, the national court should disapply
those provisions or refrain from placing such a construction on them. Moreover, in the event of the
retention being lifted, calculation of the interest payable by the treasury which did not take as its starting
point the date on which the VAT balance in question would have had to be repaid in the normal course of
events would be contrary to the principle of proportionality.

Costs
65.
The costs incurred by the Belgian, Greek, Italian and Swedish governments and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the proceedings pending before the
national courts, the decision on costs is a matter for those courts.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the questions referred to it by the
Hof van Beroep te Antwerpen, the Rechtbank van Eerste Aanleg te Brussel and the Rechtbank van
Eerste Aanleg te Brugge by orders of 17 October 1994, 25 October 1995, 12 December 1995 and 6
February 1996, hereby rules: (1) Article 18(4) of the Sixth Directive does not in principle preclude
measures of the kind at issue in the main proceedings. (2) However, the principle of proportionality is
applicable to national measures which, like those at issue in the main proceedings, are adopted by a
member state in the exercise of its powers relating to VAT, in that, if they went further than was necessary
in order to attain their objective, they would undermine the principles of the common system of VAT, in
particular the conditions governing deductions, which are an essential component of that system. It is for
the national court to examine whether or not the measures in question and the manner in which they are
applied by the competent administrative authority are proportionate. In the context of that examination, if
the national provisions or a particular construction of them would constitute a bar to effective judicial
review, in particular review of the urgency and necessity of retaining the refundable VAT balance, and
would prevent the taxable person from applying to a court for replacement of the retention by another
guarantee sufficient to protect the interests of the treasury but less onerous for the taxable person, or
would prevent an order from being made, at any stage of the procedure, for the total or partial lifting of the
retention, the national court should disapply those provisions or refrain from placing such a construction
on them. Moreover, in the event of the retention being lifted, calculation of the interest payable by the
treasury which did not take as its starting point the date on which the VAT balance in question would have
had to be repaid in the normal course of events would be contrary to the principle of proportionality.

96

[1998] All ER (EC) 97

Schning-Kougebetopoulou v Freie und Hansestadt Hamburg

(Case C-15/96)
EUROPEAN COMMUNITY; Free movement of workers
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES GULMANN (PRESIDENT OF THE THIRD AND FIFTH CHAMBERS) ACTING FOR THE
PRESIDENT), RAGNEMALM, WATHELET AND SCHINTGEN (PRESIDENTS OF CHAMBERS),
MANCINI, MOITINHO DE ALMEIDA, KAPTEYN, MURRAY, EDWARD (RAPPORTEUR), PUISSOCHET,
HIRSCH, JANN AND SEVN ADVOCATE GENERAL JACOBS
13 MAY, 17 JULY 1997, 15 JANUARY 1998
European Community Freedom of movement Workers Discrimination on grounds of nationality
Collective wage agreement applicable to public sector employees Promotion on grounds of seniority
conditional on time spent working under agreement Professional experience acquired in another
member state excluded when calculating seniority Plaintiffs practice in Greece not considered when
calculating salary classification Compatibility with Community provisions Council Regulation (EEC)
1612/68, art 7 EC Treaty, art 48.

The plaintiff was a Greek national who had worked in the Greek civil service for a number of years as a
specialist doctor. She subsequently moved to Germany where she was employed as a specialist doctor
by the City of Hamburg. Under the relevant collective wage agreement, specialist doctors were promoted
to a higher salary group after eight years practice. Since the eight years practice had to be completed as
an employee classified in a lower salary group specific to that agreement, any periods of employment
completed outside the agreement were excluded. The plaintiffs seniority was therefore calculated without
taking into account her practice in Greece. She subsequently brought an action before the Labour Court
seeking classification in the higher salary group, contending that she had suffered indirect discrimination
on grounds of nationality. The German court stayed the proceedings and referred to the Court of Justice
of the European Communities for a preliminary ruling the questions (i) whether art 48 1 of the EC Treaty
and art 7(1) and (4)2 of Council Regulation (EEC) 1612/68 on freedom of movement for workers within the
Community, prohibited a collective wage agreement (applicable to public service employees), which made
promotion on grounds of seniority conditional on eight years employment in a salary group determined by
that agreement, without regard to previous periods of comparable employment completed in the public
service of another member state and (ii) if so, what consequences ensued.
1
Article 48, so far as material, provides: freedom of movement shall entail the abolition of any discrimination based on
nationality between workers of the Member States as regards employment, remuneration and other conditions of work
and employment.
2
Article 7, so far as material, is set out at p 100 f, post

Held (1) The conditions for promotion on grounds of seniority laid down in the agreement in question
manifestly worked to the detriment of migrant workers, by failing to take into account comparable periods
of services completed in the public service of another member state, and therefore contravened the
principle of 97 non-discrimination set out in art 48 of the Treaty and art 7(1) and (4) of Regulation
1612/68. The fact that public service employees in other areas of Germany not covered by the collective
wage agreement were equally excluded for the purposes of calculating periods of service did not remedy
the breach of Community law; nor did the fact that seniority in the public service was governed by
different organisational and operational rules in each member state. Moreover, there was no question of
justification on the grounds of rewarding employee loyalty, since the wage agreement covered the
majority of German public institutions and undertakings performing public interest tasks and therefore
afforded a high degree of mobility to employees covered by the agreement. Further, the activities of a
specialist doctor did not fall within the scope of the derogation on freedom of movement set out in art
48(4)3 of the Treaty for public sector employees (see p 108 j, p 110 b to g and p 111 h, post).
3
Article 48(4), so far as material, provides: The provisions of this Article shall not apply to employment in the public
sector.

(2) A clause such as that at issue was null and void under art 7(4) of Regulation 1612/68 in so far as it
laid down or authorised discriminatory conditions in relation to workers who were nationals of other
member states. In order to determine the consequences which ensued from art 7(4), pending the
adoption by the parties to the collective agreement of the amendments necessary to eliminate the
discrimination, it was appropriate to apply the case law on the principle of equal pay for men and women.
Accordingly, where a provision discriminated against women, the members of the disadvantaged group
were to be treated in the same way and to have applied to them the same rules as the other workers and,
failing correct implementation of art 119 of the Treaty in national law, those rules remained the only valid
point of reference. It followed that, without requiring or waiting for the clause to be abolished by collective
negotiation or some other procedure, the national court had to apply the same rules to the members of
the group disadvantaged by that discrimination as those applicable to the other workers (see p 110 j to p
111 e j, post); van Cant v Rijksdienst voor Pensionen Case C-154/92 [1993] ECR I-3811 applied.

Notes
For the principle of non-discrimination between workers on grounds of nationality, and the public service
derogation from that principle, see 52 Halsburys Laws (4th edn) paras 1513, 1523.
For the EC Treaty, art 48, see 50 Halsburys Statutes (4th edn) 283.

Cases cited
Allu v Universita degli Studi di Venezia Case 33/88 [1989] ECR 1591.
Allu v Universita degli Studi di Venezia Joined cases C-259/91, C-221/91 and C-332/91 [1993] ECR I-
4309.
Bachmann v Belgium Case C-204/90 [1994] STC 855, [1992] ECR I-249, ECJ.
Biehl v Administration des Contributions du Grand-Duch de Luxembourg Case C-175/88 [1991] STC
575, [1990] ECR I-1779, ECJ.
EC Commission v Belgium Case 149/79 [1980] ECR 3881.
EC Commission v France Case 307/84 [1986] ECR 1725.
EC Commission v Luxembourg Case C-111/91 [1993] ECR I-817.
EC Commission v UK Case C-279/89 [1992] ECR I-5785.
98
Eurico Italia Srl v Ente Nazionale Risi Joined cases C-332333/92 and C-335/92 [1994] ECR I-711.
Grahame v Bestuur van de Nieuwe Algemene Bedrijfsvereniging Case C-248/96 (1997) Transcript, 13
November, ECJ.
Kowalska v Freie und Hansestadt Hamburg Case C-33/89 [1990] ECR I-2591.
McDermott v Minister for Social Welfare Case 286/85 [1987] ECR 1453.
Nimz v Freie und Hansestadt Hamburg Case C-184/89 [1991] ECR I-297.
OFlynn v Adjudication Officer Case C-237/94 [1996] All ER (EC) 541, [1996] ECR I-2617, ECJ.
Paraschi v Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501.
Pinna v Caisse dAllocations Familiales de la Savoie Case 41/84 [1986] ECR 1.
Scholz v Opera Universitaria di Cagliari Case C-419/92 [1994] ECR I-505.
Sotgiu v Deutsche Bundespost Case 152/73 [1974] ECR 153.
Spotti v Freistaat Bayern Case C-272/92 [1993] ECR I-5185.
Union de Recouvrement des Cotisations de Scurit Sociale et dAllocations Familiales de la Savoie
(URSSAf) v Hostellerie Le Manoir Srl Case C-27/91 [1991] ECR I-5531.
van Cant v Rijksdienst voor Pensionen Case C-154/92 [1993] ECR I-3811.

Reference
By order of 1 December 1995, the Arbeitsgericht (the Labour Court) Hamburg referred to the Court of
Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty two
questions (set out at p 107 j to p 108 c, post) on the interpretation of art 48 of that Treaty and art 7(1) and
(4) of Council Regulation (EEC) 1612/68 on freedom of movement for workers within the Community.
Those questions arose in proceedings between Mrs Schning- Kougebetopoulou, of Greek nationality,
and the Freie und Hansestadt Hamburg concerning her classification in a higher salary group under the
applicable collective wage agreement. Written observations were submitted on behalf of: Mrs Schning-
Kougebetopoulou, by K Bertelsmann, Rechtsanwalt, Hamburg; the German government, by E Rder,
Ministerialrat in the Federal Ministry of Economic Affairs, and S Maass, Regierungsrtin zur Anstellung in
the same ministry, acting as agents; the French government, by C Chavance, Attach Principal
dAdministration Centrale in the Legal Affairs Directorate of the Ministry of Foreign Affairs, and C de
Salins, Deputy Director in the same directorate, acting as agents; and the European Commission, by P
Hillenkamp, Legal Adviser, and P van Nuffel, of its Legal Service, acting as agents. Oral observations
were made by: Mrs Schning-Kougebetopoulou, represented by K Bertelsmann; the German
government, represented by E Rder; the Spanish government, represented by S Ortiz Vaamonde,
Abogado del Estado, acting as agent; the French government, represented by C Chavance; and the
Commission, represented by B Jansen, Legal Adviser, acting as agent. The language of the case was
German. The facts are set out in the opinion of the Advocate General.

17 July 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
In the present case the Arbeitsgericht (the Labour Court) Hamburg has asked the Court of Justice of
the European Communities to give a preliminary ruling on the compatibility of a clause of a collective
wage agreement for public servants with art 48 of the EC Treaty and art 7(1) of Council Regulation (EEC)
1612/68 on freedom of movement for workers within the Community (OJ S Edn 1968 (II) p 475).
99

The facts
2.
Dr Schning-Kougebetopoulou, a Greek national, has been employed since 1 August 1993 as a
specialist doctor by the City of Hamburg. She occupies the post of an Angestellte, that is to say, a
contractual employee not having the status of Beamte (an established official). As such her employment
relationship is governed, under the terms of her contract of employment, by the
Bundesangestelltentarifvertrag (the BAT) (the Federal Collective Wage Agreement for Contractual
Employees).
3.
As a specialist doctor employed as such Dr Schning-Kougebetopoulou is graded under the BAT in
Salary Group Ib. She considers however that she should be graded in a higher salary group, namely
Group Ia. Under the BAT a specialist doctor employed as such is classified in Salary Group Ia after eight
years practice as a doctor (either generalist or specialist) in Salary Group Ib. By stipulating that the eight
years practice must be completed in Salary Group Ib the BAT leaves out of account periods of
employment which a national or non-national may have completed abroad, or indeed in Germany with a
private employer, as an established official in the German public service or as a contractual employee not
covered by the BAT or by Salary Group Ib of the BAT.
4.
Before taking up employment with the City of Hamburg Dr Schning- Kougebetopoulou spent the
period from 1 October 1986 to 31 August 1992 working as a specialist doctor in the Greek public service.
She considers that, by disregarding that experience for the purposes of her grading, the City of Hamburg
contravened art 48 of the Treaty and art 7(1) and (4) of Regulation 1612/68.
5.
Article 7(1) and (4) provides:
(1) A worker who is a national of a Member State may not, in the territory of another Member
State, be treated differently from national workers by reason of his nationality in respect of any
conditions of employment and work, in particular as regards remuneration, dismissal, and should
he become unemployed, reinstatement or re-employment (4) Any clause of a collective or
individual agreement or of any other collective regulation concerning eligibility for employment,
employment, remuneration and other conditions of work or dismissal shall be null and void in so far
as it lays down or authorises discriminatory conditions in respect of workers who are nationals of
the other Member States.
6.
The national court notes that, although the BAT does not discriminate directly on grounds of
nationality, the condition for access to Salary Group Ia cannot be fulfilled abroad but only by nationals and
foreigners working in Germany. The national court considers that the proportion of foreigners among
those adversely affected by the rule is likely to be significantly higher than the proportion of foreigners
among those favoured by the rule. However, in the national courts view it is unnecessary to consider
whether the figures are sufficiently significant because the requirement for access to the higher grade is
objectively justified. Subsequent promotion to the higher grade represents the reward for loyalty and also
provides motivation by allowing an employee to improve his financial circumstances without changing his
job. It therefore serves to tie specialist staff to their public employer and is comparable to a private sector
bonus scheme based on years of service with an employer. The national court points out finally that, if the
opposite view were taken, there would be discrimination against German and foreign doctors whose
experience was gained in Germany but not under the relevant salary group of the BAT.
100
7.
Although apparently in little doubt about the correctness of its view the national court decided to put
the following questions to the Court of Justice:
(1) Is there an infringement of Article 48 of the EC Treaty and Article 7(1) and (4) of Regulation
(EEC) No 1612/68 of the Council on freedom of movement for workers within the Community
where a collective agreement for the public service provides for promotion on grounds of seniority
after eight years service only in a particular salary bracket provided for by the collective wage
agreement in force for all employees in the public service of the Federal Republic of Germany (the
BAT) and therefore does not take account of comparable activities carried out in the public service
of another member state of the EC?
(2) If the reply to question 1 is in the affirmative: Does Article 48 together with Regulation (EEC)
No 1612/68 of the Council on freedom of movement for workers within the Community require that,
where doctors have worked as such in the public service of another Member State of the EC, the
time spent in such employment should likewise be taken into account for the purposes of promotion
on grounds of seniority as provided for in the BAT or should the court instead take no such decision
 and leave this matter to the parties to the collective agreement, having regard to their freedom to
agree terms?
8.
The court has received written and oral argument from Dr Schning- Kougebetopoulou, the French
and German governments and the European Commission. The Spanish government, although not
presenting written observations, was represented at the hearing. In addition the court, in response to a
written question put to the member states, received replies from the Austrian, Danish, Finnish, French,
German, Greek, Irish, Luxembourg, Netherlands, Spanish, Swedish and UK governments.

Question 1
9.
Dr Schning-Kougebetopoulou contends that the BAT rule infringes art 48 of the Treaty and art 7 of
Regulation 1612/68. It discriminates indirectly on grounds of nationality and is not objectively justified. Dr
Schning-Kougebetopoulou is supported in her view by the Commission.
10.
The German government, on the other hand, considers that the rule is neither directly nor indirectly
discriminatory and, in any event, is objectively justified. It is supported in that view by the French and
Spanish governments.
11.
It is common ground that the BAT rule does not discriminate directly on grounds of nationality. It
remains to be considered, however, whether it does so indirectly. A comprehensive summary of the
courts case law on indirect discrimination in the context of art 48 of the Treaty is to be found in the recent
judgment in OFlynn v Adjudication Officer Case C-237/94 [1996] All ER (EC) 541, [1996] ECR I-2617
(paras 1819), where the court held:
(18) Accordingly, conditions imposed by national law must be regarded as indirectly
discriminatory where, although applicable irrespective of nationality, they affect essentially migrant
workers (see Pinna v Caisse dAllocations Familiales de la Savoie Case 41/84 [1986] ECR 1 (para
24), Allu v Universita degli Studi di Venezia Case 33/88 [1989] ECR 1591 (para 12) and Union de
Recouvrement des Cotisations de Scurit Sociale et dAllocations Familiales de la Savoie
(URSSAF) v Hostellerie Le Manoir Srl Case C-27/91 [1991] ECR I-5531 (para 11)) or the great
majority of those affected are migrant workers (see EC Commission v UK Case 101 C-279/89
[1992] ECR I-5785 (para 42) and Spotti v Freistaat Bayern Case C-272/92 [1993] ECR I-5185 (para
18)), where they are indistinctly applicable but can more easily be satisfied by national workers
than by migrant workers (see EC Commission v Luxembourg Case C-111/91 [1993] ECR I-817
(para 10) and Paraschi v Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-
4501 (para 23)) or where there is a risk that they may operate to the particular detriment of migrant
workers (see Biehl v Administration des Contributions du Grand-Duch de Luxembourg Case C-
175/88 [1991] STC 575, [1990] ECR I-1779 (para 14) and Bachmann v Belgium Case C-204/90
[1994] STC 855, [1992] ECR I-249 (para 9)).
(19) It is otherwise only if those provisions are justified by objective considerations independent
of the nationality of the workers concerned, and if they are proportionate to the legitimate aim
pursued by the national law (see, to that effect, Bachmann [1994] STC 855, [1992] ECR I-249 (para
27), EC Commission v Luxembourg [1993] ECR I-817 (para 12) and Allu v Universita degli Studi
di Venezia Joined cases C-259/91, C-221/91 and C-332/91 [1993] ECR I-4309 (para 15)).
12.
In my view it is clear that the BAT arrangements work to the particular detriment of migrant workers.
Any specialist doctor who has spent part of his career in the public service of another member state
incurs the disadvantage that his employment in that member state can never be taken into account for
grading purposes. By contrast any specialist doctor who has been continuously employed under the BAT
is entitled to have the whole period of his employment as a doctor taken into account.
13.
Of the public servants who spend their careers in Germany, it is only those who move from non-BAT to
BAT employment who are liable to incur the same disadvantage as migrant workers. Such persons are
likely to represent a small proportion of German public servants. The broad scope of the BAT allows
public servants extensive mobility between German public service employers without any pay
disadvantage. Moreover, it cannot be suggested that all doctors who are public servants in non-BAT
employment are actually disadvantaged by the BAT rule. Most employees not covered by the BAT will be
subject to their ownin some cases superiorcareer arrangements, such as the regulations applicable
to officials or to other collective agreements or contractual arrangements specific to their employer 4. By
contrast the BAT rule works to the disadvantage of every migrant worker joining the German public
service.
4
Eg the German governments reply to the courts written questions states that the Land Berlin, although no longer
subject to the BAT, has concluded a specific collective agreement adopting the collective conventions concluded under
the BAT by the body representing the other Lnder, the Tarifgemeinschaft deutscher Lnder.

14.
The fact that owing to the hiatuses between the arrangements applicable to the different categories of
staff some German public servants suffer the same disadvantage as migrant workers scarcely justifies
denying all migrant workers who take up employment in the German public sector the advantages
enjoyed by an entireand apparently the largestcategory of German public servants.
15.
The untenability of the German governments position is particularly clear if one considers what the
position would be if the BAT rule were repeated in the arrangements applicable in the case of other public
service employers or categories of staff. On the German governments analysis none of the arrangements
considered individually would be discriminatory owing to the existence of other 102 categories of staff and
public service employers not covered by the arrangements; yet the cumulative effect would be to place
migrant workers at a disadvantage for the purposes of grading in the entire German public sector.
Whether the rules are objectively justified
16.
In the present proceedings twoconflictinglines of argument have been put forward in order to
show that the BAT arrangements are objectively justified. On the one hand, the French and Spanish
governments, developing the arguments of the national court (set out at para 6, above), consider that
restrictions on mobility are justified by considerations specific to public service. The German government,
on the other hand, contends that the BAT can be compared to collective agreements in the private sector
and even denies that it is exclusively a public service arrangement. Its purpose is to promote a stable
workforce by rewarding loyalty to BAT employers viewed collectively. I shall consider those arguments in
turn.
Public service justification
17.
It is first of all common ground that the post of specialist doctor is outside the scope of art 48(4) of the
Treaty. The court has held that that provision applies solely to posts which presume on the part of those
occupying them the existence of a special relationship of allegiance to the state and reciprocity of rights
and duties which form the foundation of the bond of nationality (see the judgments in EC Commission v
Belgium Case 149/79 [1980] ECR 3881 (para 10) and EC Commission v France Case 307/84 [1986] ECR
1725 (para 12)). In any event art 48(4) cannot permit discrimination on the grounds of nationality, for
instance in pay or working conditions, if in fact a non-national is appointed to the post in question (see the
judgment in Sotgiu v Deutsche Bundespost Case 152/73 [1974] ECR 153 (para 4)).
18.
One cannot therefore in the present case speak of allegiance to the state in the above sense. There
nevertheless remains the question whether the BAT rule can be justified by the particular features of
public service employment. The starting point in considering this issue is the courts ruling in Scholz v
Opera Universitaria di Cagliari Case C-419/92 [1994] ECR I-505. There a public body, namely an Italian
university, based its selection criteria for appointment to a post as a university canteen assistant partly on
length of previous employment in the public service. The court held that the failure by the university to
take account of experience gained by the applicant in the public service of another member state (the
German post office) constituted unjustified indirect discrimination.
19.
The French and Spanish governments consider, however, that the ruling in Scholz relating to
appointment to a post in the public service cannot be transposed to the present case, which concerns
grading. They argue that under systems such as theirs public service entails the undertaking of reciprocal
obligations by employee and employer. In the absence of harmonised regulations on career structures
and pay, recognition of public service completed in another member state would disrupt their public
service structures.
20.
At the hearing the Spanish government gave the example of two teachers or doctors employed in
different member states. The employee in member state A must wait patiently for a number of years for
his pay to increase automatically according to length of service; the employee in member state B is paid
more highly from the start by virtue of more flexible grading rules. According to the Spanish government,
to require member state A to take account of the latter employees experience in member state B would
be to assimilate years of service that were not comparable.
103
21.
The French government nevertheless suggested at the hearing that the interest in free movement and
the integrity of national public services might be reconciled in cases such as the present, first, by requiring
that the employment of the migrant worker should have been in a public service or a body pursuing public
service aims and, secondly, by requiring a comparison of the functions performed, the level of
qualifications required and the level of remuneration.
22.
As I stated in my opinion in Scholz [1994] ECR I-505 (paras 2728), I am prepared to accept that
public servants share a special ethos which may be relevant in assessing whether periods of employment
should be taken into account for the purposes of recruitment and promotion in the public service. Public
servants may be motivated by factors that differ essentially from those obtaining in the private sector.
Typically, employment in the public service implies a willingness to accept relatively modest financial
rewards in return for greater long-term security, together, perhaps, with the satisfaction of rendering
service to the collectivity. Promotion may be viewed in part as a reward for such service.
23.
 Notwithstanding the potential difficulty in defining the limits of public service, I therefore accept the
French governments suggestion that a member state may restrict periods of qualifying employment to
those completed in public service. However, as I noted in my opinion in Scholz, I do not think the special
ethos of public service can justify the disregarding of periods of public service completed in another
member state.
24.
As regards the French and Spanish governments argument concerning differences in career
structures, I accept that there are significant differences between the member states in the rules
governing the employment of public servants. In its above-mentioned written question, the court asked
the member states to state whether periods of service completed in another member state were taken
into account for the purposes of determining seniority in their public services. The replies reveal
considerable diversity in the public service structures of the member states, ranging from a highly flexible
system based on individually determined remuneration in Sweden to the rather more rigid systems
applied in France, Spain and Germany. It appears that in a number of member states, including Austria,
Denmark, the Netherlands, Sweden and the United Kingdom, relevant previous experience, including
experience gained in another member state, is taken into account, at least to some extent, upon initial
appointment but length of service plays no automatic role in future grading. In other member states, in
particular France, Germany, Luxembourg and Spain, length of service plays a formal role in subsequent
grading and salary progression. Moreover, it appears that at present no or little account is taken of
previous employment in another member state5; an apparent exception is Greece which, although also
applying a system of automatic advancement, recently adopted a law expressly providing for periods
completed in the public service of another member state to be taken into account for the purposes of
advancement (see Law No 2470 of 21 March 1997).
5
It appears that under a law of 16 December 1996 France now takes account of compulsory national service completed
in another member state or European Economic Agreement country for the purposes of calculating seniority.

25.
Despite the differences in public service structure public servants in the various member states
perform broadly similar functions and doubtless share similar types of motivation. The fact that a public
servant was subject to different conditions in the public service of another member state cannot justify
non-recognition of periods of employment in the service of that state. To accept 104 that would be to allow
a member state to place a migrant worker at a permanent disadvantage by comparison with national
public servants simply because he benefited, or may have benefited, from a more flexible career structure
in the past. It would serve to compartmentalise national public services and constitute a significant
obstacle to public sector mobility in the Community.
26.
In any event, it seems to me that the French and Spanish governments overlook part of the equation.
It may be true that the more flexible career structures in the public services of some member states may
allow higher initial salaries and more rapid promotion for some public servants. However, greater flexibility
brings with it greater demands on public servants: promotion and even continued employment may
depend to some degree on performance. The argument that it would be unfair to a migrant workers
colleagues to recognise his public service in another member state ignores the comparative
disadvantages to which he may have been subject.
27.
It has moreover not been suggested that the different career structures would of themselves induce
public servants to switch from the public service of one member state to that of another. In any event,
significant differences in career structures may exist between member states even in the private sector,
particularly in the professions.
Loyalty to an employer or group of employers
28.
The German government has not sought to rely on considerations specific to public sector
employment. It argues that, by rewarding loyalty for continuous service, the BAT rule does no more than a
private employer or a number of private employers collectively would do. Indeed the German government
has sought to show that the BAT is also in part a private sector arrangement.
29.
I agree with the Commission that arrangements adopted by an undertaking (or group of undertakings
in common ownership) in order to reward employees for long service do not necessarily entail unjustified
discrimination. However, as the Commission points out, it is difficult to see how the arrangements under
the BAT can be equated with the rewarding of loyalty by a specific private employer when service with
any BAT employer is treated as qualifying employment. It is apparent that the BAT applies to an array of
legally separate employers, sometimes with conflicting interests and policies and with only limited
common structures, operating methods or institutional goals.
30.
The German government replies that it is conceivable that a number of private employers might
collectively agree on arrangements such as those of the BAT. At the hearing it gave the example of two
car manufacturers who enter into a collective agreement with the relevant unions under which they each
agree to provide, in their loyalty schemes, for recognition of service under the other company.
31.
 However, such an agreementon the assumption that it were able to withstand scrutiny under art 85
of the Treatywould not, in the absence of a single employer, involve the rewarding of loyalty in any
normal sense of the word. Service with a competitor can hardly be thought of as demonstrating loyalty to
ones employer. At the very least its sole purpose could not be to reward loyalty since that purpose would
be better served by agreements specific to each employer.
32.
The same applies to the BAT. If the aim of the BAT were solely to reward an employees loyalty to his
employer it would be sufficient for each employer to adopt its own arrangements or, if uniform conditions
were sought, to agree collectively on conditions to be applied independently by each employer without
any provision for 105 mutual recognition of periods of service. That would however presumably be
unacceptable to the parties to the BAT for the very reason which causes migrant workers to be placed at
a disadvantage under the present arrangements. It would preclude mobility, not only for migrant workers
(and a small number of German public servants), but for the majority of German public servants by
compartmentalising the German public sector. Thus the BAT goes beyond the mere rewarding of loyalty
by allowing in addition for extensive mobility for the majority of public servants.
33.
Finally, the German governments assertion that the BAT is in part a private sector arrangement casts
doubt on whether it would even be lawful to restrict qualifying periods of employment to those completed
in the public sector. However, the national courts questions are posited on the basis that the BAT is a
public sector agreement, and in any event it seems clear that Dr Schning- Kougebetopoulous previous
employment as a doctor was in the Greek public service.

Question 2
34.
By its second question the national court asks whether, on the assumption that the BAT rule is
unlawful, it should require the periods of service in another member state to be taken into account or refer
the matter back to the parties to the collective agreement, having regard to their freedom to contract.
35.
On this question I share the Commissions view that guidance may be sought from the courts case
law on sex discrimination, in particular the judgments in McDermott v Minister for Social Welfare Case
286/85 [1987] ECR 1453 (para 19) and van Cant v Rijksdienst voor Pensionen Case C-154/92 [1993]
ECR I-3811 (para 20). By virtue of art 7(4) of Regulation 1612/68 the BAT rule is null and void in so far as
it lays down or authorises discriminatory conditions for workers who are nationals of other member states.
It will be for the parties to the BAT to make the necessary amendments to it in order to eliminate the
discrimination. However, until such amendments are made the rules applied to the advantaged group
provide the only appropriate point of reference.
36.
As the Commission points out, it will be necessary for the national court in such circumstances to
examine the equivalence of the periods of service completed in the public service of the other member
state. However, there appears to be no dispute on that issue in relation to Dr Schning-Kougebetopoulou.
The BAT requires eight years employment as a doctor for promotion to Group Ia, and it appears to be
common ground that the years of service for which she seeks recognition were spent as a doctor in the
Greek public service.

Conclusion
37.
Accordingly, I am of the opinion that the questions referred by the Arbeitsgericht Hamburg should be
answered as follows:
(1) A clause of a collective agreement for the public service of a member state which provides
for promotion on grounds of seniority after eight years service in a particular career bracket under
the agreement and does not take account of comparable employment in the public service of
another member state infringes art 48 of the EC Treaty and art 7(1) of Council Regulation (EEC)
1612/68 on freedom of movement for workers within the Community and is null and void under art
7(4) of that regulation in so far as it lays down or authorises discriminatory conditions for workers
who are nationals of other member states.
106
(2) Until such time as the terms of the collective agreement are amended so as to eliminate the
discrimination, art 48 of the Treaty, together with art 7(1) of Regulation 1612/68, requires that the
advantages conferred by the agreement should be extended to a migrant worker, so that
comparable employment in the public service of another member state is given equivalent
recognition.

15 January 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 1 December 1995, received at the Court of Justice of the European Community on 19
January 1996, the Arbeitsgericht (the Labour court) Hamburg referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty two questions on the interpretation of art 48 of that
Treaty and art 7(1) and (4) of Council Regulation (EEC) 1612/68 on freedom of movement for workers
within the Community (OJ S Edn 1968 (II) p 475).
2.
Those questions have been raised in proceedings between Mrs Schning- Kougebetopoulou, of Greek
nationality, and the Freie und Hansestadt Hamburg (the Free Hanseatic City of Hamburg) concerning her
classification in a higher salary group under the Bundes-Angestelltentarifvertrag (the BAT) (the Federal
Collective Wage Agreement for Contractual Employees).
3.
Annex 1a to the BAT lays down salary scales. Thus, specialist doctors employed as such after eight
years practice as a doctor in Salary Group Ib are to be classified in Salary Group Ia, sub-group 4.
4.
Since 1 August 1993 Mrs Schning-Kougebetopoulou has been employed under a contract of
employment as a specialist doctor in the public service of the Freie und Hansestadt Hamburg in Germany.
In her contract of employment, drawn up on the basis of the BAT, she is classified in Salary Group Ib, sub-
group 7, as a specialist doctor employed as such.
5.
From 1 October 1986 until 31 August 1992 Mrs Schning-Kougebetopoulou worked in the Greek
public service as a specialist doctor under the staff regulations applicable to civil servants of that state.
6.
Since that period was not taken into account for the purposes of calculating her seniority, she brought
an action on 22 June 1995 before the Arbeitsgericht Hamburg seeking classification in a higher salary
group under the BAT. In support of that claim she submits that she has suffered indirect discrimination
contrary to art 48 of the Treaty and to art 7(1) and (4) of Regulation 1612/68.
7.
Article 7(1) and (4) of Regulation 1612/68 provides:
(1) A worker who is a national of a Member State may not, in the territory of another Member
State, be treated differently from national workers by reason of his nationality in respect of any
conditions of employment and work, in particular as regards remuneration, dismissal, and should
he become unemployed, reinstatement or re-employment (4) Any clause of a collective or
individual agreement or of any other collective regulation concerning eligibility for employment,
employment, remuneration and other conditions of work or dismissal shall be null and void in so far
as it lays down or authorises discriminatory conditions in respect of workers who are nationals of
the other Member State.
8.
The Arbeitsgericht Hamburg decided to stay proceedings and to refer the following questions to the
Court of Justice for a preliminary ruling:
(1) Is there an infringement of Article 48 of the EC Treaty and Article 7(1) and (4) of Regulation
(EEC) No 1612/68 of the Council, on freedom of 107 movement for workers within the Community,
where a collective agreement for the public service provides for promotion on grounds of seniority
after eight years service only in a particular salary bracket provided for by the collective wage
agreement in force for all employees in the public service of the Federal Republic of Germany (the
BAT) and therefore does not take account of comparable activities carried out in the public service
of another Member State of the EC?
(2) If the reply to Question 1 is in the affirmative: Does Article 48 together with Regulation (EEC)
No 1612/68 of the Council on freedom of movement for workers within the Community require that,
where doctors have worked as such in the public service of another Member State of the EC, the
time spent in such employment should likewise be taken into account for the purposes of promotion
on grounds of seniority as provided for in the BAT or should the court take no such decision and
leave this matter instead to the parties to the collective agreement, having regard to their freedom
to agree terms?

The first question

9.
The court has consistently held that, in the context of the application of art 177 of the Treaty, it has no
jurisdiction to decide whether a national provision is compatible with Community law. The court may,
however, extract from the wording of the questions formulated by the national court, having regard to the
facts stated by it, those elements which concern the interpretation of Community law for the purpose of
enabling that court to resolve the legal problem before it (see, inter alia, the judgment in Eurico Italia Srl v
Ente Nazionale Risi Joined cases C-332333/92 and C-335/92 [1994] ECR I-711 (para 19)).
10.
In the present case, the person concerned claims only that periods during which she worked as a
specialist doctor in the public service of another member state should be taken into account.
11.
Second, it is clear from the first question that her activity as a specialist doctor in the public service of
her member state of origin and her activity as a specialised doctor in the public service of the host
member state must be regarded as comparable. The profession concerned is, moreover, one which is
regulated at Community level.
12.
Third, art 48 of the Treaty lays down the fundamental principle of freedom of movement for workers.
Article 7(4) of Regulation 1612/68, which merely clarifies and gives effect to rights already conferred by
art 48 of the Treaty (see Scholz v Opera Universitaria di Cagliari Case C-419/92 [1994] ECR I-505 (para
6)), guarantees equal treatment of workers who are nationals of other member states in regard to any
clause of a collective or individual agreement or any other collective regulation concerning, in particular,
pay.
13.
Fourth, the derogation in art 48(4) of the Treaty, according to which the provisions on freedom of
movement for workers are not to apply to employment in the public service, concerns only access for
nationals of other member states to certain posts in the civil service (see the judgment in Grahame v
Bestuur van de Nieuwe Algemene Bedrijfsvereniging Case C-248/96 (1997) ECJ Transcript, 13
November (para 32)). It does not concern the activities of a specialist doctor, which do not involve direct
or indirect participation in the exercise of powers conferred by public law and duties designed to
safeguard the general interests of the state or of public authorities (see, in that regard, EC Commission v
Belgium Case 149/79 [1980] ECR 3881 (para 10)).
108
14.
In those circumstances, the first question asked by the Arbeitsgericht must be read as seeking to
ascertain whether art 48 of the Treaty and art 7(1) and (4) of Regulation 1612/68 prohibit a clause of a
collective agreement applicable to the public service of a member state, such as the clause in question,
which provides for promotion on grounds of seniority of employees of that public service after eight years
employment in a salary group determined by that agreement, without regard to previous periods of
comparable employment completed in the public service of another member state.
15.
The German government argues that the BAT clause at issue has neither the object nor the effect of
treating only, or mainly, nationals of other member states less favourably than German nationals. It points
out that the clause at issue not only takes no account of periods of employment completed abroad but
also takes no account of periods completed in Germany that are not covered by the BAT or of periods
completed in a Salary Group other than Group Ib.
16.
According to the Spanish government, the clause at issue cannot be characterised as discriminatory.
Years of seniority acquired in the German and Greek public administrations are based on different rules
and are not comparable. Clauses which treat such situations differently cannot be regarded as contrary to
the principle of equal treatment.
17.
The German, Spanish and French governments consider in any event that the clause at issue is
based on objectively justified factors unconnected with any discrimination. In that regard, they put forward
two arguments.
18.
First, the French and Spanish governments submit that the conditions for promotion on grounds of
seniority laid down by the BAT may be justified by characteristics specific to employment in the public
service. In the absence of harmonisation or even coordination of national organisational and operating
rules applicable to the public service, the recognition of service completed in the public service of another
member state would disrupt the application of the various schemes applicable to public service posts in
the different member states, in particular as regards the rules for taking seniority into account for the
purposes of internal promotion and career progression.
19.
Second, although the national courts questions are based on the premiss that the BAT is a public
sector agreement which aims to engender loyalty amongst qualified staff throughout the sector, the
German government submits that the aim of promotion on grounds of seniority provided for by that
agreement is, like that of collective agreements in the private sector, to reward an employees loyalty to
the whole of a given group of employers and to motivate him by the prospect of improvement in his
financial situation. Community law does not preclude an employees loyalty to a private sector employer
from being rewarded in that way.
20.
The French and Spanish governments also point out the difficulties in comparing the rules on
promotion on grounds of seniority in the public sector with those in the private sector.
21.
The questions for consideration are therefore whether a clause of a collective agreement applicable to
the public service of a member state, such as the clause in question, is such as to infringe the principle of
non-discrimination laid down in art 48 of the Treaty and art 7(1) and (4) of Regulation 1612/68 and, if so,
whether such rules are justified by objective considerations independent of the nationality of the
employees concerned and whether they are proportionate to the legitimate aim of the national provisions
(see, inter alia, the judgment in OFlynn v Adjudication Officer Case C-237/94 [1996] All ER (EC) 541,
[1996] ECR I-2617).
109
The principle of non-discrimination
22.
It is common ground that the BAT does not allow periods of employment completed in the public
service of another member state to be taken into account.
23.
As is explained in paras 12 to 14 of the Advocate Generals opinion, the conditions for promotion on
grounds of seniority laid down in the BAT thus manifestly work to the detriment of migrant workers who
have spent part of their careers in the public service of another member state. For that reason they are
such as to contravene the principle of non-discrimination laid down by art 48 of the Treaty and art 7(1)
and (4) of Regulation 1612/68.
24.
That finding is not called into question in the circumstances of the present case either by the fact that
some employees of the German public service might encounter the same situation as migrant workers or
by the fact that the public service is governed by different organisational and operational rules in the
member states.
Justification
25.
As regards the argument based on the particular characteristics of employment in public service, it is
sufficient to point out, as is explained in para 13 of this judgment, that the dispute before the national
court concerns only the occupation of specialist doctor which does not fall within the scope of art 48(4) of
the Treaty.
26.
As regards the argument claiming, as a ground of justification, that one purpose of the BAT is to
reward an employees loyalty to his employer and to motivate him by the prospect of improvement in his
financial situation, the German government explained at the hearing that the BAT covers not only the
majority of German public institutions but also undertakings performing public interest tasks.
27.
However, if that is the case, to take into account periods of employment completed with one of those
institutions or undertakings in determining seniority for the purposes of promotion cannot, given the
multiplicity of employers, be justified by the desire to reward employee loyalty. On the contrary, the system
affords employees covered by the BAT considerable mobility within a group of legally separate employers.
28.
Consequently, the answer to be given to the first question must be that art 48 of the Treaty and art 7(1)
and (4) of Regulation No 1612/68 preclude a clause in a collective agreement applicable to the public
service of a member state which provides for promotion on grounds of seniority for employees of that
service after eight years employment in a salary group determined by that agreement without taking any
account of previous periods of comparable employment completed in the public service of another
member state.

The second question

29.
The second question concerns the consequences which, given the freedom of contract of the parties
to a collective agreement, would arise from the national courts finding that a clause in a collective
agreement, such as that at issue in the main proceedings, is incompatible with art 48 of the Treaty and art
7(1) and (4) of Regulation 1612/68.
30.
A clause in a collective agreement applicable to the public service of a member state which provides
for promotion on grounds of seniority after eight years employment in a salary group determined by that
agreement but which takes no account of periods of comparable employment previously completed in the
public service of another member state is null and void under art 7(4) of Regulation 1101612/68 in so far
as it lays down or authorises discriminatory conditions in relation to workers who are nationals of other
member states.
31.
Therefore, having regard to the answer given to the first question, it is necessary to determine the
consequences which ensue from art 7(4) of Regulation 1612/68 pending the adoption by the parties to the
collective agreement of the amendments necessary to eliminate the discrimination.
32.
As Mrs Schning-Kougebetopoulou and the Commission have submitted, it is appropriate to apply
here the courts case law on the principle of equal pay for men and women.
33.
According to that case law, where a provision discriminates against women, the members of the
disadvantaged group are to be treated in the same way and to have applied to them the same rules as
the other workers and, failing correct implementation of art 119 of the Treaty in national law, those rules
remain the only valid point of reference (see the judgments in van Cant v Rijksdienst voor Pensionen
Case C-154/92 [1993] ECR I-3811 (para 20), Nimz v Freie und Hansestadt Hamburg Case C-184/89
[1991] ECR I-297 (para 18), Kowalska v Freie und Hansestadt Hamburg Case C-33/89 [1990] ECR I-
2591 (para 20) and McDermott [1987] ECR 1453 (para 19)).
34.
 As was found in para 11 of this judgment, above, the activities of specialist doctor pursued by Mrs
Schning-Kougebetopoulou in this case, in the public service of the member state of origin and in that of
the host member state, must be regarded as comparable.
35.
In reply to the second question, it is therefore sufficient to state that a clause in a collective agreement
entailing discrimination contrary to art 48 of the Treaty and to art 7(1) of Regulation 1612/68 is null and
void by virtue of art 7(4) of that regulation. Without requiring or waiting for that clause to be abolished by
collective negotiation or by some other procedure, the national court must therefore apply the same rules
to the members of the group disadvantaged by that discrimination as those applicable to the other
workers.

Costs
36.
The costs incurred by the French, German and Spanish governments and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the proceedings pending before the
national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Arbeitsgericht
Hamburg by order of 1 December 1995, hereby rules: (1) Article 48 of the EC Treaty and art 7(1) and (4)
of Council Regulation (EEC) 1612/68 on freedom of movement for workers within the Community
preclude a clause in a collective agreement applicable to the public service of a member state which
provides for promotion on grounds of seniority for employees of that service after eight years employment
in a salary group determined by that agreement without taking any account of previous periods of
comparable employment completed in the public service of another member state. (2) A clause in a
collective agreement entailing discrimination contrary to art 48 of the Treaty and to art 7(1) of Regulation
1612/68 is null and void by virtue of art 7(4) of that regulation. Without requiring or waiting for that clause
to be abolished by collective negotiation or by some other procedure, the national court must therefore
apply the same rules to the members of the group disadvantaged by that discrimination as those
applicable to the other workers.

111

[1998] All ER (EC) 112

Danmarks Aktive Handelsrejsende, acting on behalf of Mosbk v

Lnmodtagernes Garantifond
(Case C-117/96)
EUROPEAN COMMUNITY; Social policy, Other European Community: COMPANY; Insolvency
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES MOITINHO DE ALMEIDA (PRESIDENT OF THE CHAMBER), SEVN, EDWARD, JANN AND
WATHELET (RAPPORTEUR) ADVOCATE GENERAL COSMAS
24 APRIL, 29 MAY, 17 SEPTEMBER 1997
European Community Employment Protection of employees on employers insolvency Plaintiff
residing, working and employed in Denmark Employer established in United Kingdom Employer
becoming insolvent Plaintiff claiming outstanding salary, expenses and commission from Danish
guarantee institution Danish institution refusing claim on grounds that UK guarantee institution
responsible Which guarantee institution responsible for settling claim Council Directive (EEC) 80/987,
arts 1, 2, 3.

The plaintiff lived and worked in Denmark where she was employed by an English company which was
neither established nor registered in Denmark. The company paid the plaintiff directly without making
deductions for tax, pension or other social security contributions under Danish law. In 1994 the company
was declared insolvent and its employees, including the plaintiff, were dismissed. In accordance with art
31 of Council Directive (EEC) 80/987 on the approximation of the laws of the member states relating to the
protection of employees in the event of the insolvency of their employer, which required member states to
take the measures necessary to ensure that guarantee institutions guaranteed payment of employees
outstanding claims resulting from contracts of employment or employment relationships, the plaintiff
claimed a sum representing unpaid salary, commissions and expenses from the Danish institution
responsible for settling outstanding claims under the directive. The defendant institution refused to settle
the claim, contending that responsibility lay with the equivalent guarantee institution of the state where the
employer was established. The plaintiff brought proceedings against the defendant institution in the stre
Landsret, which stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling a question seeking to determine which guarantee institution was
responsible in such circumstances.
1
Article 3, so far as material, is set out at p 114 g, post
Held Although Directive 80/987 contained no provisions expressly relating to the claims of employees
residing and employed, or having been employed, in a member state other than that in which their
employer was established, it could not be inferred that the directive did not apply to such claims,
particularly since art 1(1) did not restrict its scope in that regard. In order for the directive to apply, art 2(1)
required that a request for the opening of proceedings to satisfy collectively the claims of creditors should
be lodged with the competent national 112 authority, and that there should be either a decision to open
those proceedings, or a finding that the business had been closed down where the available assets were
insufficient. It followed that where the employer was established in a member state other than that in
which the employee resided and was employed, the guarantee institution responsible under art 3 of the
directive for the payment of that employees claims in the event of the employers insolvency was the
institution of the state in which either of the events set out in art 2(1) had occurred (see p 133 b to d and p
134 b e, post).
Bonifaci v Instituto Nazionale della Previdenza Sociale (INPS) Joined cases C-9495/95 (1997) ECJ
Transcript, 10 July, and Maso v Instituto Nazionale della Previdenza Sociale (INPS) Case C-373/95
(1997) ECJ Transcript, 10 July, applied.

Notes
For protection of employees on employers insolvency, see 52 Halsburys Laws (4th edn) para 2021.

Cases cited
Abels v Administrative Board of the Bedrijfsvereniging voor de Metaalindustrie en de Electrotechnische
Industrie Case 135/83 [1985] ECR 469.
Berkholz v Finunzamt Hamburg-Mitte-Altstadt Case 168/84 [1985] ECR 2251.
Bonifaci v Instituto Nazionale della Previdenza Sociale (INPS) Joined cases C-9495/95 (1997)
Transcript, 10 July, ECJ.
de Jaeck v Staatssecretaris van Financin Case C-340/94 [1997] ECR I-461.
EC Commission v France Case C-30/89 [1990] ECR I-691.
EC Commission v Greece Case C-53/88 [1990] ECR I-3917.
Francovich v Italy Joined cases C-6/90 and C-9/90 [1991] ECR I-5357.
Francovich v Italy Case C-479/93 [1995] ECR I-3843.
Maso v Instituto Nazionale della Previdenza Sociale (INPS) Case C-373/95 (1997) Transcript, 10 July,
ECJ.
Suifritti v Instituto Nazionale della Previdenza Sociale (INPS) Joined cases C-140141/91, C-278279/91
[1992] ECR I-6337.
Trans Tirreno Express SpA v Ufficio Provinciale IVA Case 283/84 [1986] ECR 231.
van Schaik (Criminal proceedings against) Case C-55/93 [1994] ECR I-4837.
Wagner Miret v Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911.

Reference
By order of 27 March 1996, the stre Landsret (the Eastern Regional Court, Denmark) referred to the
Court of Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty a
question (set out at p 131 h j, post) on the interpretation of art 3 of Council Directive (EEC) 80/987 on the
approximation of the laws of the member states relating to the protection of employees in the event of the
insolvency of their employer. The question arose in a dispute between Mrs Mosbk and the
Lnmodtagernes Garantifond, the guarantee institution responsible pursuant to the directive for settling
outstanding claims in Denmark, following the insolvency of Mrs Mosbks employer. Written observations
were submitted on behalf of: Lnmodtagernes Garantifond, by U Andersen and A Rubach-Larsen, both of
the Copenhagen Bar; the German government, by E Rder, Ministerialrat in the Federal Ministry of the
Economy, and B Kloke, Oberregierungsrat in the same ministry, acting as agents; the French government,
by C de Salins, Head of Sub-directorate in the Legal Directorate of the Ministry of Foreign Affairs, and C
Chavance, Foreign 113 Affairs Secretary in the same directorate, acting as agents; the UK government,
by S Ridley, of the Treasury Solicitors Department, acting as agent, assisted by N Green, Barrister; and
the European Commission, by H Stvlbk and M Patakia, of its Legal Service, acting as agents. Oral
observations were made by: Danmarks Aktive Handelsrejsende, acting on behalf of Carina Mosbk,
represented by C Elmquist-Clausen, of the Copenhagen Bar; Lnmodtagernes Garantifond, represented
by U Andersen; the French government, represented by C Chavance; the UK government, represented
by S Ridley, assisted by N Green; and the Commission, represented by H Stvlbk. The language of the
case was Danish. The facts of the case are set out in the opinion of the Advocate General.

29 May 1997.

The Advocate General (G Cosmas)

delivered the following opinion (translated from the Greek).

I PRELIMINARY OBSERVATIONS
1.
In the present case the stre Landsret (the Eastern Regional Court) has referred to the Court of
Justice for a preliminary ruling under art 177 of the EC Treaty a question on the interpretation of Council
Directive (EEC) 80/987 on the approximation of the laws of the member states relating to the protection of
employees in the event of the insolvency of their employer (OJ 1980 L283 p 23).

II LEGAL FRAMEWORK
2.
The purpose of Directive 80/987 is to guarantee employees a minimum Community protection in the
event of the insolvency of their employer, but without preventing member states from establishing more
favourable systems. To that end it requires that the member states create bodies which will ensure that
employees are paid part of their outstanding claims against employers who have become insolvent.
3.
More particularly, art 1 of the directive provides:
(1) This Directive shall apply to employees claims arising from contracts of employment or
employment relationships and existing against employers who are in a state of insolvency within
the meaning of Article 2(1).
(2) Member States may, by way of exception, exclude claims by certain categories of employee
from the scope of this Directive, by virtue of the special nature of the employees contract of
employment or employment relationship or of the existence of other forms of guarantee offering the
employee protection equivalent to that resulting from this Directive. The categories of employee
referred to in the first subparagraph are listed in the Annex.
4.
Article 2 provides:
(1) For the purposes of this Directive, an employer shall be deemed to be in a state of
insolvency: (a) where a request has been made for the opening of proceedings involving the
employers assets, as provided for under the laws, regulations and administrative provisions of the
Member State concerned, to satisfy collectively the claims of creditors and which make it possible
to take into consideration the claims referred to in Article 1(1), and (b) where the authority which is
competent pursuant to the said laws, regulations and administrative provisions has: either decided
to open the proceedings, or established that the employers undertaking or business has 114 been
definitively closed down and that the available assets are insufficient to warrant the opening of the
proceedings.
(2) This Directive is without prejudice to national law as regards the definition of the terms
employee, employer, pay, right conferring immediate entitlement and right conferring
prospective entitlement.
5.
Article 3 provides:
(1) Member States shall take the measures necessary to ensure that guarantee institutions
guarantee, subject to Article 4, payment of employees outstanding claims resulting from contracts
of employment or employment relationships and relating to pay for the period prior to a given date.
(2) At the choice of the Member States, the date referred to in paragraph 1 shall be: either that
of the onset of the employers insolvency; or that of the notice of dismissal issued to the employee
concerned on account of the employers insolvency; or that of the onset of the employers
insolvency or that on which the contract of employment or the employment relationship with the
employee concerned was discontinued on account of the employers insolvency.
6.
Article 4 gives member states the option to limit the liability of the guarantee institutions referred to in
art 3(1), according to the rules laid down in art 3(2), while art 4(3) provides that in order to avoid the
payment of sums going beyond the social objective of the directive member states may set a ceiling to the
liability for employees outstanding claims.
7.
Article 5 of the directive provides:
Member States shall lay down detailed rules for the organization, financing and operation of the
guarantee institutions, complying with the following principles in particular: (a) the assets of the
institutions shall be independent of the employers operating capital and be inaccessible to
proceedings for insolvency; (b) employers shall contribute to financing, unless it is fully covered by
the public authorities; (c) the institutions liabilities shall not depend on whether or not obligations to
contribute to financing have been fulfilled.
8.
Finally, art 9 provides: This Directive shall not affect the option of Member States to apply or introduce
laws, regulations or administrative provisions which are more favourable to employees.

III FACTS
According to the order for reference, the dispute which gave rise to the question referred to the court
may be described as follows. In May 1993 the English company Colorgen Ltd, whose registered office is
in Warrington, Cheshire, England, engaged Mrs Carina Mosbk, who resides in Denmark, with effect
from 1 July 1993 as commercial manager with responsibility for promoting sales in Denmark, Norway,
Sweden, Finland and, subsequently, Germany.
9.
Throughout the employment relationship Colorgen was never registered as an undertaking in
Denmark, where it was neither established nor represented other than by Mrs Carina Mosbk. Under the
contract of employment Mrs Mosbk was paid a fixed salary plus commission. Throughout the
employment relationship she was paid directly by Colorgen, from England, without deduction of Danish
tax or social security contributions, which, according to Danish law, 115are normally deducted at source
by employers in Denmark, including foreign employers which have set up a branch or other permanent
establishment there.
During the initial months Mrs Mosbk worked at home; from 6 September 1993, however, Colorgen
rented an office for her in the premises of another Danish undertaking.
10.
On 1 July 1994 Colorgen was declared insolvent. All its employees were therefore dismissed,
including Mrs Mosbaek, whose claims against the company for wages, commission and expenses came
to D Kr 471,996.
11.
In the same month Mrs Mosbk declared that claim to the Lnmodtagernes Garantifond, the
guarantee institution in Denmark responsible for ensuring payment of the claims covered by the directive,
and also to Colorgens receivers in England.
Furthermore, by letter of 22 August 1994 Mrs Mosbaek declared the same claim to the National
Insurance Fund, the UK guarantee body, which had not reached a definitive decision by the date of the
order for reference2.
2
It should be pointed out that the UK government stated in its written observations that no application had been received
by the UK institution. At the hearing, however, the representative of the UK government stated that Mrs Mosbaek s
application had been received by the UK institution a few days earlier.

12.
The Danish body refused to settle Mrs Mosbaeks claim on the ground that payment was the
responsibility of the UK guarantee institution. Consequently, on 19 December 1994 Mrs Mosbaek brought
proceedings against the defendant before the Hillerd Court, which, because of the importance of the
principle involved, referred the case to the stre Landsret.
13.
Before the stre Landsret the plaintiff, relying on the purpose of the directive, maintained that the
institution responsible for settling her claim was the Danish guarantee institution, owing, in particular, to
the fact that her place of residence and the place where she had provided services were Denmark, where
her employer had also rented premises to enable her to perform her work, but also owing to the difficulties
which she would have in pursuing her claim before the guarantee institution and the courts in the United
Kingdom.
14.
The Danish institution, on the other hand, contended that responsibility for the guarantee lay with the
institution of the state in which the employer was established and whose legislation governed the
employers insolvency, that is in the present case the UK guarantee institution. Furthermore, pursuant to
art 5 of the directive only the state in which the employer is established can require the employer to
contribute to the financing of the guarantee institutions so that those institutions are in a position to
compensate employees in the event of the employers insolvency. Finally, the Danish institution pointed
out that it could not, even in the interest of economy, compensate the plaintiff and then seek
reimbursement from the English guarantee institution, since the directive made no provision for such an
arrangement.
15.
It is apparent from the order for reference that Danish law does not regulate, either directly or
indirectly, the question of the competence of the Danish guarantee institution to settle claims such as the
plaintiffs. It is for that reason that the stre Landsret has asked the court to provide it with the
interpretation of the directive which will enable it to resolve the issue before it. To that end, it has referred
the following question to the Court of Justice for a preliminary ruling:
116
In a situation where the employer is not established in the Member State in which the employee
is resident and is solely represented in the state of the employees residence by way of the said
employees work, which inter alia is carried out in office premises rented by the employer for the
employees use, is it the guarantee institution in the country where the employer is established or
the guarantee institution in the country where the employee is resident which, on the employers
insolvency, according to Article 3 of Directive 80/987/EEC is to guarantee payment of the
employees outstanding claims resulting from the employment relationship in question?

IV SUBSTANCE
16.
 The directive does not say which guarantee institution is required, where the case arises, to settle
outstanding claims of employees residing and working in one member state against an employer
established and declared insolvent in another member state.
17.
In those circumstances, the parties to the main proceedings, the German, French and UK
governments and the European Commission maintained in their written and oral observations that an
answer to the question referred to the court based on Community law can and must be determined on the
basis of the directive as a whole, taking account also of its purpose.
18.
However, the parties differ in their views as to which guarantee institution is eventually required to
settle employees claims. More particularly, the French and UK governments and the Commission
maintain that in the particular circumstances it is the guarantee institution of the state in which the
employer is established that is competent, or in the present case the UK institution. The plaintiff in the
main proceedings and the German government, on the other hand, claim that the competent institution is
the institution of the member state in which the employee resides and/or works, or in the present case the
Danish institution.
19.
To my mind, the principal arguments in support of the first view are as follows: (a) The directive is
designed to reduce the differences between national laws as regards the protection of employees in the
event of the employers insolvency. If the solution to the question in issue were a matter for the member
states employees would have no protection in the event of a negative conflict of laws 3, whilst other
difficulties would arise in the event of competing rules4. Thus the differences would be exacerbated.
Accordingly, the problem calls for a uniform Community solution. (b) For the purpose of determining
whether an employer is insolvent, art 2 of the directive refers to the legislation and the competent
authority of the member state concerned, that is, normally those of the state where the employer is
established. Since the guarantee institution assumes the obligations of a defaulting employer it is
normally the guarantee institution of the member state where the employer is established which must
assume that obligation. Furthermore, that institution is better placed to ascertain the existence and extent
of the employers liabilities. (c) It follows from art 5(c) of the directive that a member state can only impose
an obligation to contribute to the financing of the national guarantee institution on an employer covered by
its legislation, that is to say, an employer established in its territory. Consequently, only that institution is
required and in a position to fulfil the 117 employers obligations where the employer becomes insolvent.
If the guarantee fell to be paid by the institution of the employees state of residence and/or employment
that institution would be unprepared for liabilities in respect of which it would have been unable to require
a statutory contribution on the part of the employer. That would upset the financial equilibrium of the
guarantee organisations, however, which cannot have been the Community legislatures intention.
Furthermore, the directive does not provide for an action for indemnity between national guarantee
institutions. (d) In the case of migrant workers the body responsible for paying their claims is not known,
which leads to legal uncertainty and even provides the opportunity for abuse on the part of those workers,
again owing to the above-mentioned absence of an action for indemnity between national institutions.
3
That is, where no institution was required to provide the guarantee.
4
That is, where a number of institutions were required to provide the guarantee under national law.

20.
As regards the contrary argument: (a) The German government contends that the guarantee
institutions in the directive are social security institutions in the wide sense, like those referred to in
Regulation 1408/715. It therefore considers that, if art 13(2)(a) or art 14(2)(b) of that regulation is applied
by analogy, the guarantee should in this case be provided by the Danish institution. (b) At the hearing the
plaintiff in the main proceedings, who did not submit written observations, repeated the arguments which
she had put forward before the national court. She further stated that under a clause in her contract of
employment Danish labour law was also to be taken into account when interpreting that contract.
Furthermore, the plaintiff argued that the Danish institution was competent as the institution of the place
where the work was carried out and the place whose law was applicable, on the basis of Regulation
1408/71 and arts 4(1) and 3(3) of the Convention on the law applicable to contractual obligations (Rome,
19 June 1980; OJ 1980 L266 p 1)6 and also art 5(1) of the Convention on Jurisdiction and the
Enforcement of Judgments in Civil and Commercial Matters 1968.
5
As amended and updated by Council Regulation (EEC) 2001/83 (OJ 1983 L230 p 6).
6
See the Contracts (Applicable Law) Act 1990, Sch 1.

21.
None of the above points of view has convinced me to the extent that I am able to propose that it be
adopted unconditionally. The arguments put forward are certainly not without logic, but they are implicitly
based on certain premises concerning the purpose and meaning of the provisions of the directive which, I
fear, go beyond the letter and the spirit of the law and belong rather to the realm of what ought to be.
22.
The premises on which all the above arguments are based are in my view the following. First, the
premise that the concepts of employer and employee derive from the directive and are valid for all
member states. Secondly, the concept that the scope ratione personae of the directive includes all
employees, within the above meaning, of all insolvent employers, within the meaning of the directive,
within the Community, irrespective of their attachment to a member state. Thirdly, the concept that the
provisions of the directive are sufficiently clear, as regards the determination of the national institution
required to provide the guarantee and the scope of that obligation, for employees to be able to rely on
them directly, even in the absence of national implementing provisions, indeed in spite of any national
provisions to the contrary. Fourthly, and finally, the assumption that, in that context, the Community
legislature intended, in substance, to harmonise national laws relating to the protection of employees in
the event of the insolvency of their employer.
118
23.
I shall argue that none of those premises corresponds with the letter and spirit of the directive. The
lacuna found in the directive as regards the question before the national court is intentional and cannot be
filled by the interpretation proposed by either of the parties. In a case such as that before the national
court it is in the final analysis for the competent authorities and the competent courts of each of the
member states concerned to determine, on the basis of the national provisions, whether the employee is
entitled to the protection of the directive. The resultant lacuna in the protection of employees, or any other
unfavourable consequence for them, is the necessary consequence of the limited objectives of the
gradual harmonisation which the Community legislature has chosen at the current stage in the
development of Community law.

The purpose of the directive

24.
I shall begin my analysis by determining the purpose of the directive. In order to do so, it is appropriate
to turn to the drafting history and compare the initial proposal for a directive submitted by the Commission
on 13 April 1978 with the version eventually adopted by the EC Council (see OJ 1978 C135 p 2).
25.
After referring to art 100 of the Treaty in the citations, the proposal observes in the sixth recital of the
preamble that
increasing economic interdependence across national boundaries imposes the requirement that
employees claims arising from the employment relationship in the event of the insolvency of their
employer should receive equal protection in all Member States.
In the following recital it is stated that it is necessary to promote the approximation of the laws, regulations
and administrative provisions of the member states while maintaining the improvement described in art
117 of the Treaty7.
7
It should be pointed out that the preamble to the proposal for the directive was accepted without amendment by the
European Parliament (see its opinion, published in OJ 1979 C39). Nor did the Economic and Social Committee raise
any objections to the preamble.

As regards the scope of the directive, moreover, art 1 of the proposal provided that the directive was to
apply to claims arising from employment or training relationships against insolvent employers whose
undertaking or business is situated within the territorial jurisdiction of the Treaty8.
8
The Parliament also adopted the proposal on this point.

Article 2 of the proposal provided that an employer was to be deemed insolvent 9 where proceedings
have been opened under by the laws, regulations and administrative provision of the Member States to
satisfy jointly the claims of creditors, etc (art 2(a)), or where an application for the opening of such
proceedings has been rejected on the grounds of lack of assets (art 2(b)) or where his business has
been closed down due to insolvency10.
9
The expression deemed insolvent was replaced in the actual directive by in a state of insolvency.
10
It will be observed that the proposal failed to establish precisely which authority was to determine, and according to what
rules, that the relevant proceedings had been opened or that an application to open them had been rejected, etc.

As regards the content of the guarantee, the proposal provided in art 3 for payment of salary claims
arising before the onset of the employers insolvency (art 3(a)) and also for payment of certain social
benefits (art 3(b)). It also allowed the member states to limit the guarantee to three months remuneration,
irrespective of any reference period (art 4(a)), while the payment of claims for 119 social benefits could be
limited to those which had arisen during the 12 months preceding the onset of insolvency 11.
11
The Parliament regret[ed] in particular that the Commission had merely adopted minimum rules of protection (para 4 of
the preamble to an opinion) and considered it wholly unacceptable that the guarantee was not extended to claims
arising after the employers insolvency (para 5); finally, taking the view that it was quite unreasonable that the
guarantee should be limited to three months wages (para 6), it proposed a guarantee of six months (see art 4(a) of the
proposal as amended by the Parliament).

It should be particularly noted that the proposal for the directive provided (art 5(b)) that the guarantee
institutions must not be financed by contributions solely from employees; in other words, employees
were to be primarily responsible for financing those institutions 12.
12
In para 7 of the preamble to its resolution, the Parliament considered that under no circumstances [could] there he any
question of asking employees to contribute to the financing of a guarantee fund to cover their legally justified claims
against their employer and, accordingly, it proposed that employers should pay the necessary contributions to cover the
expenditure of the fund, including administrative expenditure (see the amended version of art 5(b)). The Council agreed
to a certain extent, since the directive provides that the guarantee institutions are to be financed by employers and by
the member states (art 5(b)).
26.
Despite being somewhat vague and general, the wording of those provisions, in particular the
preamble and art 1, none the less permits the conclusion, in my opinion, that, in accordance with the
approach adopted when it was being drawn up, the directive sought to apply to all employees,
irrespective of whether they resided in the same member state as the employer, provided that the
employer was established in the territory of the Community.
Furthermore, it follows from the overall wording of the proposal, read in conjunction with the opinion of
the Parliament, that at the preparatory stage the objectives of the directive were just as ambitious as
regards the extent of the protection guaranteed to employees, which, it was envisaged, would be the
same in all member states.
27.
The Council eventually made radical amendments to the proposal.
28.
The amendments are evident even in the preamble. First, while the first recital states that it is
necessary to provide for the protection of employees in the event of the insolvency of their employer, the
second recital states first that differences still remain between the Member States as regards the extent
of the protection of employees in this respect13, before further stating merely that efforts should be
directed towards reducing these differences. For that reason, instead of the harmonization of the
national provisions to which the Commissions proposal referred, the third recital of the preamble to the
directive merely states that the approximation of laws in this field should, therefore, be promoted.
13
The Commission had already pointed that out when it referred to the absence of appropriate institutions in certain
member states (see the fifth recital in the preamble to the proposal for a directive).

As the court has observed (see the judgment in Francovich v Italy Case C-479/93 [1995] ECR I-3843
(para 20) (Francovich II)):
Whilst the legislature considered, in general, that it was necessary to provide for the protection
of employees in the event of the insolvency of their employer, it also limited the specific purpose of
its action to reducing the remaining differences between the Member States as regards the
protection of employees in that respect.
120
29.
As the court said in the same judgment, the restriction which the Community legislature placed on
itself was clearly due both to the intrinsic difficulties generally presented by any effort at harmonising
different national laws undertaken on the basis of art 100 and to the specific difficulties in drawing up
common rules in the particular field concerned. The latter difficulties are explained by the absence of a
definition of insolvency common to the insolvency proceedings of the member states, which, precisely
because of the existing differences14, have thus far proved impossible to harmonise throughout the
Community (see [1995] ECR I-3843 (para 28)).
14
The court had already held in Abels v Administrative Board of the Bedrijfsvereniging voor de Metaalindustrie en de
Electrotechnische Industrie Case 135/83 [1985] ECR 469 (paras 1617) that the specificity of insolvency law,
encountered in all the legal systems of the Member States, is confirmed in Community law [and] was reflected in
[inter alia] the adoption of Council Directive No 80/987, before going on to observe that the rules on liquidation
proceedings and analogous proceedings are very different in the various Member States.

30.
Consequently, the Community legislature sought, by adopting the directive, to promote a partial
harmonisation, or a gradual harmonization, of national provisions relating to the protection of employees
in the event of the insolvency of their employer (see [1995] ECR I-3843 (paras 20, 27)). In other words,
the directive represents the first step towards the harmonisation of national laws in this field.
31.
That the Community legislature opted in favour of partial or gradual harmonisation is confirmed by the
wording of the directive on two fundamental points, namely its scope and the extent to which employees
are to be protected, that is to say, the content of the guarantee.

The content of the guarantee

32.
To take the second point first, it will be observed that the Council reversed the balance of the proposal
submitted by the Commission. Thus instead of a reinforced guarantee ensured by a fund financed by (or
mainly by) the employees themselves the Council decided in favour of a limited guarantee ensured by an
institution jointly funded by employers and the member states.
33.
In the light, evidently, of the financial burden for employers and states resulting from the creation and
functioning of such institutions, the directive leaves member states the option, after evaluating domestic
economic and social conditions, of substantially restricting the scope of the guarantee. Thus member
states may: (a) choose a date before which they will guarantee employees outstanding claims (art 3 of
the directive); (b) establish, in relation to that date, a reference period by determining that cover is to
apply to the proportion of outstanding claims relating to that period (art 4(2)); and (c) set a ceiling where
that guarantee, already limited as indicated above, may involve payment of sums going beyond the social
objective of the directive (art 4(3)); none of which, of course, is to affect the option of member states to
apply measures which are more favourable to employees (art 9 of the directive).
34.
It is therefore a minimum level of protection under Community law that the directive is intended to
guarantee to employees, as the court has pointed out on a number of occasions (see the judgments in
EC Commission v Greece Case C-53/88 [1990] ECR I-3917 (para 19), Francovich v Italy Joined cases C-
6/90 and C-9/90 [1991] ECR I-5357 (Francovich I) and Suifritti v Instituto Nazionale della Previdenza
Sociale (INPS) Joined cases C-140141/91, C-278279/91 [1992] ECR I-6337 (para 3)).
121

The scope of the directive

35.
The partial or gradual harmonisation sought by the directive is even clearer, however, if reference is
made to the manner in which its scope is defined. The scope ratione personae of the directive and, as we
shall see in what follows, its scope ratione territoriae, which is what interests us most here, are
incompletely defined in the directive by direct or indirect reference to the national provisions of each
member state.
36.
The scope ratione personae of the directive, that is to say, the circle of persons entitled to the
guarantee, is defined in arts 1 and 2. More precisely, as the court held in Francovich I [1991] ECR I-5357
(para 13):
With regard to the identity of the persons entitled to the guarantee, it is to be noted that,
according to Article 1(1), the directive applies to employees claims arising from contracts of
employment or employment relationships and existing against employers who are in a state of
insolvency within the meaning of Article 2(1), the latter provision defining the circumstances in
which an employer must be deemed to be in a state of insolvency. Article 2(2) refers to the national
law for the definition of the concepts of employee and employer. Finally, Article 1(2) provides
that the Member States may, by way of exception and under certain conditions, exclude claims by
certain categories of employees listed in the annex to the directive. (My emphasis.)
37.
In the following passage in the same judgment the court defines the conditions under which national
courts are to consider that a person comes within the scope of the directive. More particularly, the court
states that in order to determine whether or not a person is intended to benefit from the guarantee a
national court must verify whether the person concerned is an employed person under national law and
whether he is excluded from the scope of the directive in accordance with art 1(2) and the Annex, and
then ascertain whether a state of insolvency as provided for in art 2 of the directive exists (see Francovich
I [1991] ECR I-5357 (para 14) and Francovich II [1995] ECR I-3843 (para 17)).
38.
The first of those conditions was defined in the judgment in Wagner Miret v Fondo de Garanta
Salarial Case C-334/92 [1993] ECR I-6911. In that case a Spanish court asked whether a member of the
higher management staff of an undertaking, who was not regarded as an employee and was therefore
excluded from the guarantee under the provisions relating to the guarantee institution, came within the
scope of the directive where he could be regarded as an employee within the meaning of the general
provisions of national law15.
15
In his opinion in that case, Advocate General Lenz pointed out that, according to the information supplied by the court
making the reference, under Spanish law higher management staff were regarded as employees.

After pointing out that under Article 2(2) of the directive the definition of employee is a matter of
national law ([1993] ECR I-6911 (para 11)), the court observed (para 12):
It follows that the directive on the insolvency of employers is intended to apply to all categories
of employee defined as such by the national law of a Member State, with the exception of those
listed in the Annex to the directive.
122
On the basis of those considerations, the court gave the following reply to the national court ([1993]
ECR I-6911 (para 14)):
higher management staff may not be excluded from the scope of Directive 80/987/EEC
where they are classified under national law as employees and they are not listed in section I of the
Annex to the directive.
39.
According the judgment in Francovich II [1995] ECR I-3843 (para 14), however, a finding to that effect
is not sufficient to bring a person within the category of persons entitled to the guarantee. As I have
already had occasion to point out, it constitutes a first step by the national court in bringing the person
concerned within the scope of the directive (see my opinion [1995] ECR I-3843 (para 22)). To follow its
reasoning to a conclusion the court will also have to take account of the employers situation.
Therefore, since Article 2(2) [of the directive] refers to national law for the definition of the concepts of
employee and employer, the national court will also have to determine, first, whether the employer
actually referred to is regarded as an employer in national law and, secondly, whether the employer is
insolvent within the meaning of the directive (see the judgment in Francovich I [1991] ECR I-5357 (para
13)).
40.
As regards the latter condition, the court, when defining art 2 of the directive, held as follows (see
Francovich II [1995] ECR I-3843 (paras 1819)):
(18) It is clear from the terms of Article 2 that in order for an employer to be deemed to be in a
state of insolvency, it is necessary, first, that the laws, regulations and administrative provisions of
the Member State concerned provide for proceedings involving the employers assets to satisfy
collectively the claims of creditors; secondly, that employees claims resulting from contracts of
employment or employment relationships may be taken into consideration in such proceedings;
thirdly, that a request has been made for the proceedings to be opened; and, fourthly, that the
authority competent under the said national provisions has either decided to open the proceedings
or established that the employers undertaking or business has been definitively closed down and
that the available assets are insufficient to warrant the opening of the proceedings.
(19) It thus appears that the Community legislature has expressly limited the scope of the
directive so that the rights which it introduces cannot be relied upon by employees whose contract
of employment or employment relationship is with an employer who cannot, under the provisions in
force in the Member State concerned, be subject to proceedings to satisfy collectively the claims of
creditors. Such an employer cannot be in a state of insolvency within the specific meaning of that
phrase as used in the Directive.
41.
On that basis, the court rejected an argument advanced by the Commission, among others, to the
effect that the directive was intended to protect all employees with the sole exception of those referred to
in the annex (see my opinion in Francovich II [1995] ECR I-3843 (para 21)). More precisely, the court held
that
(20) Although the literal interpretation of Article 2 of the Directive may mean that the protection
afforded by the Directive varies from one Member State to another as a result of differences
between the various national rules 123 governing proceedings to satisfy collectively the claims of
creditors, it cannot be rebutted by arguments based on the aim set out in the first recital in the
preamble. Whilst the legislature considered, in general, that it was necessary to provide for the
protection of employees in the event of the insolvency of their employer, it also limited the specific
purpose of its action to reducing the remaining differences between the Member States as regards
the protection of employees in that respect. That literal interpretation is thus consistent with the
partial harmonization pursued by the Directive.
(21) the Directive is to be interpreted as applying to all employees, other than those in the
categories listed in the annex thereto, whose employers may, under the applicable national law, be
made subject to proceedings involving their assets in order to satisfy collectively the claims of
creditors. (See the judgment in Francovich II [1995] ECR I-3843 (paras 2021).)
42.
It may be inferred from the foregoing, first, that the directive does not contain a Community concept of
employer and employee which can be applied uniformly in all member states. The content of those
concepts expressly falls to be determined by the member states. For the purpose of applying the directive
those concepts must be taken in the sense which they have in each national legal order 16.
16
This technique is common in secondary Community legislation. See de Jaeck v Staatssecretaris van Financin Case C-
340/94 [1997] ECR I-461 (paras 19, 23 and 24ff), where the court held that the concepts of employed person and paid
employment as used in Council Regulation (EEC) 1408/71 on social security for migrant workers (OJ S Edn 1971(II) p
416) are to be regarded as referring to the definitions given them by member states social security legislation, that they
are independent of the nature of the activity for the purposes of employment law and, furthermore, that they do not refer
to the Community concept of worker as used in art 48 of the Treaty.

Secondly, it follows from what I have said above that the aim of the directive is not to protect all
employees without exception, as the parties see in, wrongly, to believe, but only those who (a) are
regarded as employees in national law and (b) are not expressly excluded from the benefit of the
directive, provided that they have a relationship with employers who (c) are regarded as such in national
law and (d) are insolvent within the meaning of the directive.
Thirdly, it is impossible to determine solely on the basis of the directive whether or not a particular
person comes within its scope. That falls to be determined by the national court in the context of national
law.

The guarantee institutions obligation to pay

43.
Is the fact that an individual employee comes within the scope of the directive sufficient to allow him to
rely on the guarantee as against the guarantee institution or, at least, the member state?
44.
The answer is that it is not. For the employee to be able to do so, it is at least necessary, according to
the case law of the court, that the directive has been transposed into national law.
The question arises in the present case whether the directive refers to a particular branch of national
law for the definition of the words employer and employee. The directive concerns the protection of
employees in the event of the insolvency of the employer. Consequently, the rules on insolvency must be
taken into consideration. Owing to the specificity of the corresponding provisions however, it is probable
that insolvency law has not established its own 124 particular concepts of employer and employee but
has borrowed them from labour law. The national court will therefore also have to refer to labour law.
Finally, in arts 6 and 7 the directive contains provisions which relate to employees statutory and
supplementary social security schemes. Consequently, the corresponding provisions of social security law
will also have to be taken into account. In those circumstances, I consider that it is for the national court,
which has an overall picture of the various branches of domestic law, to define the concepts of employer
and employee, taking into account the aims of the directive.
As the court stated in Francovich I [1991] ECR I-5357 (paras 2526):
(25) It follows from the terms of the directive that the Member State is required to organize an
appropriate institutional guarantee system The fact that the directive envisages as one
possibility among others that such a system may be financed entirely by the public authorities
cannot mean that the State can be identified as the person liable for unpaid claims.
(26) Accordingly, even though the provisions of the directive in question are sufficiently precise
and unconditional as regards the determination of the persons entitled to the guarantee and as
regards the content of that guarantee, those elements are not sufficient to enable individuals to rely
on those provisions before the national courts. Those provisions do not identify the person liable to
provide that guarantee, and the State cannot be considered liable on the sole ground that it has
failed to take transposition measures within the prescribed period.
45.
Following those considerations, the court held that where the directive had not been transposed into
domestic law the injured employee may under certain conditions be able to rely on a right to reparation as
against the state as a result of its failure to transpose the directive (see paras 38ff).
46.
That being so, the following question arises: where the directive has been transposed into national law
and the guarantee institution has been set up, is the person concerned able to demand payment of the
guarantee from that institution? In other words, does an obligation for the institution to pay the guarantee
derive from the directive?
47.
I would observe in that regard that in Wagner Miret v Fondo de Garanta Salarial Case C-334/92
[1993] ECR I-6911, the Spanish court asked whether the person concerned can, on the basis of the
directive, take direct action against the national guarantee institution, or indeed claim reparation from the
state, where he is an employee within the meaning of national law but not within the meaning of the
provisions governing the guarantee institution.
After referring to member states discretion with regard to the general organisation of the guarantee
institutions and after observing that art 3(1) leaves it to the member states to adopt the measures
necessary to ensure that guarantee institutions guarantee payment of employees outstanding claims
(see [1993] ECR I-6911 (paras 1719)), the court replied that
(a) higher management staff are not entitled, under Directive 80/987, to request payment of
amounts owing to them by way of salary from the guarantee institution established by national law
for the other categories of employee, and (b) in the event that, even when interpreted in the light of
that directive, national law does not enable higher management staff to obtain the benefit of the
guarantees for which it provides, such staff are entitled to 125 request the State concerned to make
good the loss and damage sustained as a result of the failure to implement the directive in their
respect. (See [1993] ECR I-6911 (para 23).)17

17
In his opinion in that case, Advocate General Lenz observed that in Francovich I [1991] ECR I-5357 the court took the
view that despite being sufficiently precise and unconditional as regards determining the categories of persons entitled
to the guarantee and as regards the content of the guarantee, the provisions of the directive did not identify the person
liable to provide the guarantee and, accordingly, the persons concerned could not rely directly on those provisions in the
absence of transposition measures. The Advocate General therefore considered whether the existence of a guarantee
institution in Spain conferred direct effect on the provisions of the directive. He came to the conclusion that it did not,
since direct effect must derive from the rule itself, having regard to its context, and not from the law of a member state
(see [1991] ECR I-5357 (paras 1216). I support that argument. I also believe that in the Wagner Miret judgment the
court clearly supported it, albeit by implication, since it did not consider that the national provision which excluded the
person concerned from the protection could or should be set aside on the basis of the directive.

48.
The Wagner Miret case, like the Francovich cases, concerned claims by employers who were subject
to the jurisdiction of the same member state as their employers. The judgments delivered by the court,
especially in the Wagner Miret case, indicate that the directive does not itself determine the obligations of
the guarantee institutions, but requires the member states to do so in their implementing measures.
Consequently, it is impossible to determine on the basis of the provisions of the directive in themselves
whether and to what extent a national guarantee institution is required to ensure payment of outstanding
claims, whether for a certain category of employees or for certain employees. Even where, in objective
terms, they come within the scope of the directive, such employees cannot rely on its provisions to require
an institution to provide the guarantee to them where the measures adopted to transpose the directive do
not so provide; where the directive has not been properly transposed, however, they are entitled to
request reparation from the negligent member state.
49.
Consequently, where it is impossible to determine on the basis of the directive alone whether an
institution (which I shall presume has already been constituted) is obliged to ensure the guarantee where
both employer and employee come within the jurisdiction of the same member state, it is a fortiori
impossible to determine whether that institution is required to provide the guarantee where the
competence of the member state extends only to the employer or only to the employee. Accordingly,
where employer and employee are subject to the jurisdiction of different member states, as appears to be
the case here, it is also impossible to say which of the institutions involved is required to provide the
guarantee provided for in the directive.
50.
On the basis of those considerations, if the question referred for a preliminary ruling were to be strictly
interpreted, it would be necessary to give a reply such as: In a situation such as that in the main
proceedings, the employees claim must be paid by the guarantee institution of the member state on
which the obligation to do so is imposed by national law.
51.
Clearly, such a reply would be neither useful nor satisfactory to the national court, since it would
immediately beg the following question: Where there is no body which is competent, which member state
must, to apply the directive properly, ensure payment of the claim in circumstances such as those in the
main proceedings? Consequently, the question referred by the national court is in substance as follows:
126
In a situation such as that in the main proceedings, where the employer is established and has
been declared insolvent in one Member State, while the employee resides and works in another
Member State, does the directive require one of the two Member Statesand if so, which to
guarantee, by means of the measures adopted to transpose it, payment of the employees
claims?18

18
In all logic, this question consists of two parts: the first concerns Danish law and the answer must be such that the court
making the reference is able to determine the main dispute, and the other concerns English law and there is no need to
answer it from that point of view. Since, however, the question in its entirely lends itself better to a complete
interpretation of the directive, I shall consider both parts together.

52.
As thus reformulated, the question, if I interpret it properly, raises the problem of the scope rationae
territoriae of the directive.
According to the provisions of the directive, in the light of its purpose, its scope ratione territoriae to my
mind corresponds with the limits of the jurisdiction of each member state. In other words, the directive is
addressed to each member state and requires it to guarantee, by means of institutions envisaged for that
purpose, payment of employees claims against insolvent employers in so far as employees and
employers come within its jurisdiction19. Anything going beyond this is not a requirement but is authorised
by the directive as a measure which is more favourable to employees, within the meaning of art 9.
19
That is to say, in principle, but not necessarily, employers and employees established in the territory of the member
state.

53.
It should be pointed out at the outset that the reference to increasing economic interdependence
across national boundaries in the preamble to the proposal disappeared from the wording of the directive,
as did the criterion in art 1 of the proposal, which provided that the scope of the directive was to be
determined by reference to whether the undertaking or business was situated within the territorial
jurisdiction of the Treaty (see para 25, above). If it may therefore be inferred from its wording that the
proposal sought also to cover claims arising under employment relationships of an international nature, as
I maintained in para 26, above, then the removal or amendment of the corresponding points necessarily
means that the Community legislature abandoned that intention.
54.
Secondly, it should be observed that in arts 1 and 2 of the directive the scope ratione materiae
(employees outstanding claims, etc) is defined in conjunction with its scope ratione personae. Moreover,
to determine whether an employee comes within the scope ratione personae of the directive, the national
court must examine, as we have seen, first, whether the employee is regarded as such by national law
and does not conic within one of the categories excluded by the Annex to the directive and, secondly,
whether there is an employer within the meaning of national law and whether that employer is in a state of
insolvency within the meaning of the directive (see para 42, above). In so far as both these questions fall
to be resolved on the basis of the same national law, it is logical to conclude that the Community
legislature had in mind employees and employers subject to the same national law.
55.
It should be observed, thirdly, that the member states option, pursuant to arts 4 and 10 of the
directive, to limit the guarantee institutions obligations to pay in various ways presupposes an evaluation
of the economic and social conditions and the employment relationships which as a rule develop within a
single state. 127If the directive had intended to require member states also to cover claims arising from
employment relationships with transfrontier characteristics it would undoubtedly have established a
system for the co-ordination of the corresponding provisions of the member states, and in particular a
scheme for the adaptation of the guarantee which took account of the conditions in force in the place
where the guarantee is actually to be provided.
56.
 Fourthly, tile method of financing the guarantee institutions, within the meaning of art 5, leads to the
same conclusion. There is no doubt that the Community legislature had in mind financially balanced
organisations, that is to say organisations whose income would cover expenditure, so that they would be
in a position to meet their obligations. Income also includes financing by employers and/or the member
state, while the principal expenditure and obligation consist in providing the guarantee to employees.
Furthermore, the financial balance of any system of this type depends to a large extent on being able to
foresee the date and amount of income and expenditure and also, of course, their implementation.
Clearly, such a balance can only be struck in the context of one and the same national law. In fact, in the
absence of express provisions in the directive, in a case such as this employers subject to the jurisdiction
of another state cannot be obliged to contribute to the financing of the guarantee institution of the state
under whose jurisdiction the employee comes, nor can the institution of the first state be required to settle
the claims of employees coming within the jurisdiction of another state, which it therefore did not foresee.
No support for the opposite argument is to be found in art 5(c) of the directive, since that provision in
reality severs the link between the guarantee institutions obligation to make payment and the actual
payment of employers contributions and does not refer to the employers statutory obligation to contribute
to the financing of the institutions as imposed by the state.
57.
Finally, it is necessary to consider arts 6 to 8 of the directive, which form part of Section III, entitled
Provisions concerning social security. Those articles establish a link between the guarantee institutions
obligation to pay and the national social security or provident schemes. They thus require coordination
between the guarantee institutions and the national social security schemes which is only practicable in
the context of the national law of each member state. It follows from art 6, for example, that the obligation
imposed on the guarantee institutions under arts 3 and 5 of the directive also generally extends to the
contributions payable by employees to national social security or provident schemes. All that
presupposes, logically, that employers and employees are subject to the same national social security
scheme and to the same guarantee institution. Furthermore, the guarantee institution and the social
security scheme must come within the jurisdiction of the same member state and, accordingly, employers
and employees must also be subject to the jurisdiction of that state.
58.
To my mind, it must be inferred from the foregoing that the scope ratione territoriae of the directive
coincides with the limits of the jurisdiction of each member state. That solution is consistent with both the
letter and the purpose of the directive, which was limited to a partial harmonisation of national laws
applicable in the event of the insolvency of the employer.
Were a contrary solution, that is one or other of those proposed by the parties, to be adopted, even
adapted to the true meaning of the question referred as reformulated, that would presuppose a
coordination of member states insolvency laws, labour laws and social security laws, not to mention
private 128 international law, which is completely extraneous to the purposes of the directive.
59.
That, moreover, is the solution which the court has adopted in the case of other directives designed to
approximate national laws.
When interpreting the Council Directive (EEC) 77/388 on the harmonisation of the laws of the member
states relating to turnover taxesCommon system of value added tax: uniform basis of assessment (OJ
1977 L145 p 1), the court has held that the territorial scope of the directive coincides, in the case of each
member state, with the scope of its value added tax legislation 20, but that
20
See the judgment in Berkholz v Finunzamt Hamburg-Mitte-Altstadt Case 168/84 [1985] ECR 2251 (para 16).

[the directive] in no way restricts the freedom of the Member States to extend the scope of their
tax legislation beyond their normal territorial limits, so long as they do not encroach on the
jurisdiction of other States. (See the judgment in Trans Tirreno Express SpA v Ufficio Provinciale
IVA Case 283/84 [1986] ECR 231 (para 20).)
Therefore, whilst the directive obliges member states to tax services which are provided in their
territory it does not oblige them to tax services provided on a ship in international waters, even where the
ship is travelling between two points in the national territory 21, although it is true that it does not prevent
them from doing so (see [1986] ECR 231 (para 21)).
21
See the judgment in EC Commission v France Case C-30/89 [1990] ECR I-691.

Similarly, in a judgment on the interpretation of Council Directive (EEC) 77/143 on the approximation of
the laws of the member states relating to roadworthiness tests for motor vehicles and their trailers (OJ
1977 L47 p 47), the court considered that that directive is based on the premiss that a member state can
only undertake direct supervision of testing establishments which are situated on its own territory and
that, because of the incomplete harmonisation of the criteria for testing, it is required under certain
conditions to recognise test certificates issued in other member states (see the judgment in Criminal
proceedings against van Schaik Case C-55/93 [1994] ECR I-4837 (paras 2022)).
60.
Lastly, it should be pointed out that the directive governs a particular question, so that it cannot be
interpreted by analogy, in the light of other particular provisions, such as the measures on which the
plaintiff and the German government relied. Moreover, those measures predate the directive, and if the
Community legislature had intended to refer to them it would have expressly said so in the directive.
Furthermore, the obligations which the directive imposes on member states are mandatory and cannot be
altered by private law agreements such as the contract of employment to which the plaintiff refers.
61.
I shall now examine the consequences of the solution which I propose to the court in the present case.
That solution implies that every member state must ensure the guarantee to all employees coming within
its jurisdiction, provided that they come within the scope of the directive. The directive does not require
states to extend the guarantee to other categories of employees over and above those who, strictly
speaking, are subject to its jurisdiction, but it does not prevent them from doing so. The national courts of
each member state are responsible, where necessary, for determining whether the person concerned
satisfies the conditions which bring him within the scope of the directive from the point of view described
above.
129
Where the court finds that the person concerned is covered by the directive it will then consider
whether the national institution ensures him the guarantee owing to him. Where the provisions relating to
the national institution do not provide for the cover of his outstanding claims 22 the employee has a right to
reparation.
22
It is possible, of course, that in a case such as the present one the national provisions of a member state will not regard
as employees persons employed in another member state in situations such as that of the plaintiff, while the provisions
of the state in which the employee carries out his activity will not regard an undertaking such as Colorgen as an
employer. In that case the employee will not benefit from the guarantee, since in each of the legal orders involved he
will fail to satisfy a substantive condition of the directive. That should come as no surprise. Such an eventuality may
arise even in the context of a single national legal order (as in Francovich II) and is explained by the limited objectives of
the directive. It may also happen that both states offer to provide the guarantee. That eventuality falls outside the
framework of the directive, however, and where necessary must be examined in the context of the national legislation of
each state.

62.
The order for reference does not say clearly whether Mrs Mosbk is subject to Danish law and
whether she is to be regarded as an employee for the purposes of Danish law. It follows from the terms
of the order, however, read in conjunction with the written and oral submissions of the representative of
the Danish institution, which have not been challenged, that Colorgen, the plaintiffs employer, does not
come under Danish law and cannot therefore be regarded as an employer for the purposes of Danish
law. If that is so the Kingdom of Denmark cannot be regarded as having infringed the directive on the
ground that it did not make provision for Mrs Mosbaeks claim to be settled by the Danish guarantee
institution. Consequently, the plaintiff has no right under Danish law to the guarantee or to reparation.
63.
On the other hand, the UK government stated in its written observations and at the hearing that Mrs
Mosbaek is entitled to the guarantee under English law which is not contrary to the directive.
I do not know to what extent that statement is binding on the UK guarantee institution and the courts in
the United Kingdom which may be called upon to resolve the plaintiffs case. In any event, the answer to
that question is not necessary for the solution to be provided to the dispute before the national court and,
consequently, the court need not concern itself with it.

VI CONCLUSION
64.
I therefore propose that the question referred to the Court of Justice be answered as follows:
Council Directive (EEC) 80/987 on the approximation of the laws of the member states relating
to the protection of employees in the event of the insolvency of their employer is to interpreted as
meaning that it does not require a member state to provide an employee with the guarantee
provided for in the directive where the employee or his employer is not subject to its jurisdiction or
where the employer or the employee is not recognised as such by national law. In a case such as
that in the main proceedings, where the employer is established and has been declared insolvent in
one member state, while the employee resides and works in another member state, it is for the
national courts of each member state concerned to determine whether, in those circumstances, the
employee must be regarded under the corresponding national law as coming within the scope of
the directive.

130
17 September 1997.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By order of 27 March 1996, received at the Court of Justice of the European Communities on 12 April
1996, the stre Landsret (the Eastern Regional Court, Denmark) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty a question on the interpretation of art 3 of Council
Directive (EEC) 80/987 on the approximation of the laws of the member states relating to the protection of
employees in the event of the insolvency of their employer (OJ 1980 L283 p 23).
2.
 The question arose in a dispute between Mrs Mosbk and the Lnmodtagernes Garantifond, the
guarantee institution responsible pursuant to the directive for settling outstanding claims in Denmark (the
fund), following the insolvency of Mrs Mosbks employer.
3.
Mrs Mosbk, who lives in Denmark, was employed from 1 June 1993 by the English company
Colorgen Ltd as commercial manager for Denmark, Norway, Sweden, Finland and, subsequently,
Germany.
4.
Colorgen, whose registered office was at Warrington, England, was neither established nor registered
in Denmark either as an undertaking or in any other capacity, and in particular was not registered with the
tax or customs authorities. It was represented there only by Mrs Mosbk. For the performance of her
activities, Colorgen had rented an office.
5.
During the whole of the employment relationship, Mrs Mosbks remuneration was paid directly to her
by Colorgen, with no deductions for tax, pension or other social security contributions under Danish law.
6.
In July 1994, Colorgen was declared insolvent, and its employees, including Mrs Mosbk, were
dismissed. Pursuant to art 3 of the directive, Mrs Mosbk declared, both to the fund and to the
companys English receiver, an outstanding claim of D Kr 471,996, representing her salary, commissions
and expenses.
7.
The fund refused to settle the claim on the ground that responsibility lay with the guarantee institution
of the state where the employer was established, in this case the National Insurance Fund.
8.
On 19 December 1994, Mrs Mosbk brought court proceedings against the fund in Hillerd,
Denmark.
9.
The case was then referred, at the funds request, to the stre Landsret, which considered it
necessary to refer the following question to the Court of Justice:
In a situation where the employer is not established in the Member State in which the employee
is resident and is solely represented in the State of the employees residence by way of the said
employees work, which inter alia is carried out in office premises rented by the employer for the
employees use, is it the guarantee institution in the country where the employer is established or
the guarantee institution in the country where the employee is resident which, on the employers
insolvency, according to Article 3 of Directive 80/987/EEC is to guarantee payment of the
employees outstanding claims resulting from the employment relationship in question?
10.
As a preliminary point, it must be borne in mind that the directive is intended to guarantee to
employees a minimum level of protection under Community law in the event of the insolvency of their
employer, without prejudice to more favourable provisions existing in the member states. To that 131 end
it provides in particular for specific guarantees of payment of outstanding claims.
11.
Article 1(1) of the directive provides:
This Directive shall apply to employees claims arising from contracts of employment or
employment relationships and existing against employers who are in a state of insolvency within
the meaning of Article 2(1).
12.
Article 2(1) provides:
For the purposes of this Directive, an employer shall be deemed to be in a state of insolvency:
(a) where a request has been made for the opening of proceedings involving the employers
assets, as provided for under the laws, regulations and administrative provisions of the Member
State concerned, to satisfy collectively the claims of creditors and which make it possible to take
into consideration the claims referred to in Article 1(1), and (b) where the authority which is
competent pursuant to the said laws, regulations and administrative provisions has: either decided
to open the proceedings, or established that the employers undertaking or business has been
definitively closed down and that the available assets are insufficient to warrant the opening of the
proceedings.
13.
Article 3 of the directive requires the member states to take the measures necessary to ensure that
guarantee institutions guarantee payment of employees outstanding claims resulting from contracts of
employment or employment relationships and relating to pay for the period prior to a given date.
14.
Finally, art 5 provides:
Member States shall lay down detailed rules for the organization, financing and operation of the
guarantee institutions, complying with the following principles in particular: (a) the assets of the
institutions shall be independent of the employers operating capital and be inaccessible to
 proceedings for insolvency; (b) employers shall contribute to financing, unless it is fully covered by
the public authorities; (c) the institutions liabilities shall not depend on whether or not obligations to
contribute to financing have been fulfilled.
15.
What the national court wishes to know is, essentially, which guarantee institution is responsible under
art 3 of the directive for guaranteeing payment of an employees claims on the employers insolvency,
where that employer is established in a member state other than that in which the employee resides and
was employed.
16.
Whilst it is true that the directive contains no provisions expressly envisaging the circumstances
described in the order for reference, it cannot be inferred that it does not apply to the claims of employees
residing and employed, or having been employed, in a member state other than that in which their
employer is established.
17.
The purpose of the directive is to guarantee a minimum level of protection for employees who have
suffered as a result of their employers insolvency, there being no restriction imposed, in particular in art
1(1) of the directive which defines its scope, as regards cases where the employees place of residence
or employment does not coincide with the employers place of establishment.
18.
In order to be effective, Community lawwhich guarantees freedom of movement for persons within
the Community and thus encourages situations 132 involving, as in the present case, foreign elements
requires such an interpretation of the directive, the second recital in the preamble to which states that it is
intended to reduce existing differences between member states which can have a direct effect on the
functioning of the common market.
19.
It is therefore necessary to determine the guarantee institution responsible for paying claims where the
employer is established in a member state other than that of the employees place of residence or
employment.
20.
That institution, it is clear from the scheme of the directive, must be the guarantee institution of the
state in which, in accordance with art 2(1) of the directive, either it is decided to open the proceedings for
the collective satisfaction of creditors claims, or it has been established that the employers undertaking
or business has been definitively closed down.
21.
In the first place, for the directive to apply, art 2(1) requires two events to have occurred: first, a
request for proceedings to be opened to satisfy collectively the claims of creditors must have been lodged
with the competent national authority; secondly, there must have been either a decision to open those
proceedings, or a finding that the business has been closed down where the available assets are
insufficient (see the judgments in Bonifaci v v Instituto Nazionale della Previdenza Sociale (INPS) Joined
cases C-9495/95 (1997) ECJ Transcript, 10 July (para 35) and Maso v Instituto Nazionale della
Previdenza Sociale (INPS) Case C-373/95 (1997) ECJ Transcript, 10 July (para 45)).
22.
As the fund, the French and UK governments and the European Commission have pointed out, it thus
appears that the operation of the guarantee system established by the directive, and hence the
intervention of the guarantee institution, are conditional, first and foremost, upon a request being made for
the opening of proceedings to satisfy creditors claims collectively, thus making it possible for the salary
claims in question to be taken into consideration.
23.
In practice, the opening of such proceedings is most often requested in the state in which the
employer is established. That general tendency should be reinforced by the entry into force of the
Convention on Insolvency Proceedings (Brussels, 23 November 1995), art 3(1) of which uses as the main
criterion for jurisdiction the centre of a debtors main interests.
24.
In the second place, art 5(b) of the directive provides that the guarantee system which it is designed to
establish is to be financed by employers, unless it is fully covered by the public authorities. It accords with
the scheme of the directive, in the absence of any contrary indication therein, for the guarantee institution
responsible for employees outstanding claims to be the one which levied, or at all events should have
levied, the insolvent employers contributions. That cannot be the case of the institution of the member
state in which the employee resides and was employed without the employer having some establishment
or commercial presence there.
25.
Thus, art 5(b) of the directive confirms the link between the obligation of the guarantee institution to
pay and the place of establishment of the employer who, as a general rule, contributes to the financing of
the institution. As has already been emphasised in para 23, above, the state of the employers
establishment is most often that in which the request for the opening of the proceedings is made.
26.
Finally, the fact that the directive has not provided for a system of set-off or reimbursement of
payments between the guarantee institutions of the various member states tends to confirm that the
Community legislature intended, in the event of an employers insolvency, that the guarantee institution of
only one 133 member state should become involved, in order to prevent unnecessary entanglements
between national systems and, in particular, situations in which a worker might claim the benefit of the
directive in several member states.
27.
The answer to the question must therefore be that, where the employer is established in a member
state other than that in which the employee resides and was employed, the guarantee institution
responsible, under art 3 of the directive, for the payment of that employees claims in the event of the
employers insolvency is the institution of the state in which, in accordance with art 2(1) of the directive,
either it is decided to open the proceedings for the collective satisfaction of creditors claims or it has been
established that the employers undertaking or business has been closed down.

Costs
28.
The costs incurred by the German, French and UK governments and by the European Commission,
which have submitted observations to the Court of Justice, are not recoverable. Since these proceedings
are, for the parties to the main proceedings, a step in the action pending before the national court, the
decision on costs is a matter for that court.
On those grounds, Court of Justice (Fifth Chamber), in answer to the question referred to it by the
stre Landsret by order of 27 March 1996, hereby rules: where the employer is established in a member
state other than that in which the employee resides and was employed, the guarantee institution
responsible, under art 3 of Council Directive (EEC) 80/987 on the approximation of the laws of the
member states relating to the protection of employees in the event of the insolvency of their employer, for
the payment of that employees claims in the event of the employers insolvency is the institution of the
state in which, in accordance with art 2(1) of the directive, either it is decided to open the proceedings for
the collective satisfaction of creditors claims or it has been established that the employers undertaking or
business has been closed down.

134

[1998] All ER (EC) 135

Benincasa v Dentalkit Srl

(Case C-269/95)
CONFLICT OF LAWS: ADMINISTRATION OF JUSTICE; Courts: EUROPEAN COMMUNITY; Other
European Community
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES MANCINI (PRESIDENT OF THE CHAMBER), MURRAY, KAKOURIS (RAPPORTEUR),
KAPTEYN AND RAGNEMALM ADVOCATE GENERAL RUIZ-JARABO COLOMER
22 JANUARY, 20 FEBRUARY, 3 JULY 1997
Conflict of laws Civil and commercial matters Jurisdiction Concept of consumer Exclusive
jurisdiction clause Plaintiff concluding franchise contract but never trading under it Contract conferring
jurisdiction for any dispute on Italian courts Plaintiff attempting to have whole contract declared void by
German court German court declining jurisdiction Whether plaintiff a consumer Whether exclusive
jurisdiction clause operative where validity of whole contract disputed Convention on Jurisdiction and
the Enforcement of Judgments in Civil and Commercial Matters 1968, art 13, 14, 17.

The plaintiff, an Italian national, concluded a franchising contract with the defendant, an Italian company,
with a view to setting up and operating a shop in Mnchen to sell the latters dental hygiene products. In
return for exploiting the defendants trademark the plaintiff undertook, inter alia, to equip the shop at his
own expense, stock only the defendants products and pay a fee of Lit 8m. A clause of the contract
provided that the courts of Florence were to have jurisdiction to entertain any dispute relating the
interpretation, performance or other aspects of the contract. The plaintiff set up the shop and paid the
initial fee, but failed to pay for goods purchased from the defendant and subsequently ceased trading.
Thereafter, the plaintiff brought proceedings in the Landgericht, Mnchen I seeking to have the
franchising contract declared vodkas a whole under German law. He contended that the German court
had jurisdiction under art 5(1)1 of the Convention on Jurisdiction and the Enforcement of Judgments in
Civil and Commercial Matters 1968, since (a) his action sought to have the whole of the contract,
including the clause conferring jurisdiction exclusively on the Italian courts, declared void, and (b) the fact
that he not yet started trading meant that he should be regarded as a consumer for the purposes of arts
132 and 14 of the 1968 convention, and could therefore choose whether to bring the action in the German
or Italian courts. The Landgericht declined jurisdiction on the grounds that the exclusive jurisdiction clause
was valid under art 17c of the 1968 convention, which set out the conditions for concluding such clauses.
The plaintiff appealed to the Oberlandesgericht, Mnchen, which stayed the proceedings and referred to
the Court of Justice of the European Communities for a preliminary ruling the questions: (i) whether,
under arts 13 and 14, a person who had concluded a contract with a view to pursuing a trade or
profession in the future could be regarded as a consumer; and (ii) whether the courts of a 135 contracting
state which had been designated in a jurisdiction clause validly concluded under art 17 also had exclusive
jurisdiction where the action sought a declaration that the entire contract containing that clause was void.
 1
Article 5(1), so far as material, provides: A person domiciled in a Contracting State may, in another Contracting State,
be sued: (1) in matters relating to a contract, in the courts for the place of performance of the obligation in question
2
Article s 13 and 14, so far as material, are set out at p 150 g h, post
3
Article 17, so far as material, is set out at p 153 c, post

Held (1) The definition of consumer in art 13 of the 1968 convention only applied to a private final
consumer, who was not engaged in trade or professional activities. Accordingly, only contracts concluded
for the purpose of satisfying an individuals own private consumption came within the scope of those
provisions, which were designed to protect the consumer as the economically weaker party. Moreover,
that protection applied only to contracts concluded outside and independently of any trade or professional
activity or purpose, whether present or future. It followed that a plaintiff who had concluded a contract with
a view to pursuing a trade or profession in the future, could not be regarded as a consumer for the
purposes of arts 13 and 14 (see p 152 c f h and p 154 g, post); Shearson Lehman Hutton Inc v Treuhand
fr Vermgensverwaltung und Beteiligungen (TVB) mbH Case C-89/91 [1993] ECR I-139 applied.
(2) A jurisdiction clause which served a procedural purpose was governed by the provisions of the
1968 convention establishing uniform rules of international jurisdiction. Article 17 set out to designate,
clearly and precisely, a court in a contracting state which was to have exclusive jurisdiction in accordance
with the consensus formed between the parties to a contract, which was to be expressed in accordance
with the strict requirements as to form laid down therein. The legal certainty which that provision sought to
secure could easily be jeopardised if one party to a contract could frustrate that rule by claiming that the
whole of the contract was void under the applicable substantive law. It followed that the courts of a
contracting state which had been designated in a jurisdiction clause validly concluded under art 17 also
had exclusive jurisdiction where the action sought, in particular, a declaration that the entire contract
containing that clause was void (see p 153 d j and p 154 d h, post); Effer SpA v Kantner Case 38/81
[1982] ECR 825 considered.
Per curiam. It is for the national court to determine whether the clause invoked before it, which refers
to any dispute relating to the interpretation, performance or other aspects of the contract, also covers
any dispute relating to the validity of the contract (see p 154 c, post); Powell Duffryn plc v Petereit Case
C-214/92 [1992] ECR I-1745 applied.

Notes
For the Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters
1968, arts 13, 14, 17 (as set out in Sch 1 of the Civil Jurisdiction and Judgments Act 1982), see 11
Halsburys Statutes (4th edn, 1991 reissue) 1142, 1143, 1145.

Cases cited
Bertrand (Socit) v Paul Ott KG Case 150/77 [1978] ECR 1431.
Dumez France v Hessische Landesbank (Helaba) Case C-220/88 [1990] ECR I-49.
Effer SpA v Kantner Case 38/81 [1982] ECR 825.
El Corte Ingls SA v Blzquez Rivero Case C-192/94 [1996] I-1281.
Handte & Cie GmbH v Traitements mcano-chimiques des surfaces (TMCS) Case C-26/91 [1992] ECR I-
3967.
Mainschiffahrts-Genossenschaft eG (MSG) v Les Gravires Rhnanes SARL Case C-106/95 [1997] All
ER(EC) 385, [1997] ECR I-911, ECJ.
136
Mulox IBC Ltd v Geels Case C-125/92 [1993] ECR I-4075.
Powell Duffryn plc v Petereit Case C-214/92 [1992] ECR I-1745.
Pronuptia de Paris GmbH v Pronuptia de Paris Irmgard Schillgalis Case 161/84 [1986] ECR 353.
Sanders v van der Putte Case 73/77 [1977] ECR 2383.
Shearson Lehman Hutton Inc v Treuhand fr Vermgensverwaltung und Beteiligungen (TVB) mbH Case
C-89/91 [1993] ECR I-139.

Reference
By order of 5 May 1995, the Oberlandesgericht (the Higher Regional Court), Mnchen referred to the
Court of Justice of the European Communities for a preliminary ruling under the Protocol of 3 June 1971
on the interpretation by the Court of Justice of the Convention on Jurisdiction and the Enforcement of
Judgments in Civil and Commercial Matters 1968, as amended, three questions (set out at p 151 a to d,
post) on the interpretation of arts 13, 14 and 17 of that convention. Those questions were raised in
proceedings between Dentalkit Srl, having its seat in Florence, and Mr Benincasa, an Italian national,
relating to the validity of a franchising contract concluded between them. Written observations were
submitted on behalf of: Mr Benincasa, by R Bhner, Rechtsanwalt, Mnchen; Dentalkit, by A von
Kuhlberg, Rechtsanwalt, Mnchen; the German government, by J Pirrung, Ministerialrat in the Federal
Ministry of Justice, acting as agent; and the European Commission, by P van Nuffel, of its Legal Service,
acting as agent; and H-J Rabe, Rechtsanwalt, Hamburg. Oral observations were made by: Mr Benincasa,
represented by R Bhner; and the Commission, represented by M Nez-Mller, Rechtsanwalt,
Hamburg. The language of the case was German. The facts are set out in the opinion of the Advocate
General.

20 February 1997.

The Advocate General (D Ruiz-Jarabo Colomer)

delivered the following opinion (translated from the Spanish).
1.
In this case the Oberlandesgericht (the Higher Regional Court), Mnchen, requests a preliminary
ruling on three questions concerning the interpretation of arts 13, 14 and 17 of the Convention on
Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968 (the Brussels
Convention), as amended by the 1978 Accession Convention (Denmark, Ireland, United Kingdom).
2.
The questions, which have been referred to the Court of Justice of the European Communities
pursuant to the Protocol of 3 June 1971 on the interpretation of the Brussels Convention by the Court of
Justice relate to the concepts of consumer contract and sale of goods on instalment credit terms, both
used in art 13 of the convention, and identification of the competent court to take cognisance of a dispute
concerning an agreement conferring jurisdiction of the kind referred to in art 17 of that convention.

Facts and main proceedings according to the order for reference

3.
Dentalkit Srl, a company whose registered office is in Florence (Italy), is the promoter of a chain of
franchise shops specialising in the sale of dental hygiene products.
4.
On 28 September 1992 Dentalkit and Francesco Benincasa, an Italian national, concluded a franchise
agreement in Florence for the opening and operation of a shop in Mnchen, where Mr Benincasa stated
that he resided.
137
5.
Under clause 2 of the agreement Dentalkit undertook, inter alia, (a) to allow use of the Dentalkit trade
mark for naming the sales outlet; (b) to grant the exclusive right to use the mark within a specified area;
(c) to give the necessary assistance in opening the establishment; (d) to supply a list of products; (e) to
supply the goods; (f) to give assistance in the presentation of the products; (g) to pass on the technical
and commercial expertise and know-how in its possession; (h) to give assistance in planning advertising
and local sales promotion schemes; (i) to supply certain printed matter; (j) to provide theoretical and
practical training; (m) to carry out an advertising and sales promotion campaign at national level; and (n)
not to open any other sales outlet within the exclusive territory.
6.
Under clause 3 of the agreement, Mr Benincasa for his part undertook (a) to arrange for registration in
the commercial register and to obtain the requisite authorisations; (b) to keep the business premises
available for the term of the agreement; (c) to fit out the premises in the same manner as existing
Dentalkit shops; (d) to sell only products supplied by Dentalkit and to keep suitable stocks of them; (e) to
suggest new products which could be introduced to fit in with the range; (f) to keep the premises in proper
condition and to offer customers an efficient, skilled service; (g) to use the business name and logo in
accordance with Dentalkits instructions, without alteration; (h) to maintain secrecy with regard to
information and documentation relating to the Dentalkit system; and (i) to carry out advertising
campaigns and local sales promotion at his own expense, by prior agreement with Dentalkit.
7.
Finally, Mr Benincasa undertook to pay the defendant Lit 8m to cover the cost of technical and
commercial assistance in opening the shop and, after the first 12 months of trading, 3% of the annual
turnover in consideration of the use of the business names and marks granted exclusively for the
specified territory.
8.
When concluding the agreement, the original term of which was three years and could be extended
tacitly, the two parties signed a document in Italian, which is generally used by Dentalkit for that purpose.
9.
Mr Benincasa opened the shop, paid the initial sum of Lit 8m and made various purchases, for which
he did not pay. In the meantime he ceased trading. Subsequently he brought an action against the
defendant before the Landgericht (the Regional Court) Mnchen I, seeking: (a) repayment of Lit 8m,
together with 12% interest from the date of service of the writ (27 December 1993), (b) a declaration that
the franchise agreement of 28 September 1992 is void and that, consequently, the contracts of sale
concluded thereunder are also void.
10.
Mr Benincasa contends that the franchise agreement is void on the grounds, first, that it is contrary to
para 138 of the Brgerliches Gesetzbuch (the BGB) (the German Civil Code) and, second, that the
agreement lays down an obligation for more than two years without complying with para 11(1 2)(a), in
conjunction with para 6, of the Gesetz ber die Algemeinen Geschftsbedingungen (the Law on standard
business conditions). He also challenges the agreement on the grounds of mistake pursuant to para 119
of the BGB and fraudulent misrepresentation pursuant to para 123 of the BGB.
11.
In its defence, Dentalkit asked the court to dismiss Mr Benincasas claim, but first raised the objection
that the German court before which the action had been brought had no international or local jurisdiction.
In Dentalkits opinion, the courts of Florence had exclusive jurisdiction by virtue of the jurisdiction clause
included by the parties in the franchise agreement (clause 12).
138
12.
Mr Benincasas observations in reply to that procedural objection were in essence as follows: (a) the
Landgericht Mnchen was the court for the place of performance of the obligation in question, within the
meaning of art 5(1) of the Brussels Convention; (b) the clause conferring jurisdiction on the courts of
Florence did not have the effect of excluding jurisdiction to hear his action because he sought the
annulment of the entire agreement, including that clause; (c) application of the jurisdiction clause was also
excluded by point 1 of the first paragraph of art 13 and the first paragraph of art 14 of the Brussels
Convention so that, in accordance with the third paragraph of art 17 and art 15 of the convention, that
clause could not take effect.
13.
In support of the last mentioned submission, Mr Benincasa states that when he concluded the
franchise agreement he was not carrying on a business and that therefore he should be regarded as a
consumer within the meaning of the first paragraph of art 13 of the Brussels Convention. He bases that
view on a teleological interpretation of that provision in the light of an objective of the EC Treaty, namely
to ensure a high level of protection for the final consumer.
14.
By judgment of 19 July 1993 the Landgericht Mnchen I upheld Dentalkits objection and therefore
declared the action inadmissible for lack of international jurisdiction.
15.
Consequently the national court took the view that the jurisdiction clause in the franchise agreement
was valid and, in accordance with art 17 of the Brussels Convention, found that the courts of Florence
had jurisdiction.
16.
The judgment found that the agreement was not a consumer contract, which meant that in this case
art 13 of the Brussels Convention could not be relied on to defeat the jurisdiction clause. It was clear from
the actual wording, as well as the purpose, of the first paragraph of art 13 of the convention that an
agreement intended to establish a trade or profession must be deemed to have been concluded for the
purpose of a trade or profession.
17.
In the opinion of the Landgericht Mnchen I, the other conclusions reached by applying the German
Verbraucherkreditgesetz (the Law on consumer credit) are not relevant to the interpretation of art 13 of
the Brussels Convention, which must be construed independently. Finally, the agreement in question did
not fulfil the other requirements for a consumer contract.
18.
The plaintiff lodged an appeal against the judgment given at first instance, which Dentalkit considers
should be dismissed. In the appeal proceedings both parties in essence repeat their opposing arguments
concerning the international jurisdiction of the German courts.
19.
In view of the doubts raised regarding interpretation of the Brussels Convention, the appeal court has
referred the following questions to the Court of Justice:
(1) Is a plaintiff to be regarded as a consumer within the meaning of the first paragraph of
Article 13 and the first paragraph of Article 14 of the Brussels Convention even if his action relates
to a contract which he concluded not for the purpose of a trade which he was already pursuing but
a trade to be taken up only at a future date (here: a franchise agreement concluded for the purpose
of setting up a business)?
(2) If Question 1 is to be answered in the affirmative: Does point 1 of the first paragraph of
Article 13 of the Convention (contract for the sale of goods on instalment credit terms) cover a
franchising agreement which obliges the plaintiff to buy from the other party to the agreement, over
a period of 139 several (three) years, the articles and goods required to equip and operate a
business (without instalment credit terms having been agreed) and to pay an initial fee and, as from
the second year of the business, a licence fee of 3% of turnover?
(3) Does the court of a Member State specified in an agreement conferring jurisdiction have
exclusive jurisdiction pursuant to the first sentence of the first paragraph of Article 17 of the
Convention even when an action is inter alia for a declaration of the invalidity of a franchising
agreement containing the jurisdiction clause itself, which is worded: The courts at Florence shall
have jurisdiction to entertain any dispute relating to the interpretation, performance or other aspects
of the present contract, that clause having been specifically approved within the meaning of
Articles 1341 and 1342 of the Italian Codice Civile?

The provisions of the Brussels Convention to be construed

20.
Section 4 of the Brussels Convention is entitled Jurisdiction over consumer contracts.
21.
Article 13, which forms part of that section, provides as follows:
In proceedings concerning a contract concluded by a person for a purpose which can be
regarded as being outside his trade or profession, hereinafter called the consumer, jurisdiction
shall be determined by this Section, without prejudice to the provisions of Articles 4 and 5(5), if it is:
(1) a contract for the sale of goods on instalment credit terms, or (2) a contract for a loan repayable
by instalments, or for any other form of credit, made to finance the sale of goods, or (3) any other
contract for the supply of goods or a contract for the supply of services, and (a) in the state of the
consumers domicile the conclusion of the contract was preceded by a specific invitation addressed
to him or by advertising, and (b) the consumer took in that state the steps necessary for the
conclusion of the contract.
Where a consumer enters into a contract with a party who is not domiciled in a Contracting
State but has a branch, agency or other establishment in one of the Contracting States, that party
shall, in disputes arising out of the operation of the branch, agency or establishment, be deemed to
be domiciled in that State.
22.
Article 14 provides as follows:
A consumer may bring proceedings against the other party to a contract either in the courts of
the Contracting State in which that party is domiciled or in the courts of the Contracting State in
which he himself is domiciled.
Proceedings may be brought against a consumer by the other party to the contract only in the
courts of the Contracting State in which the consumer is domiciled.
These provisions shall not affect the right to bring a counterclaim in the court in which, in
accordance with this Section, the original claim is pending.
23.
Finally, art 17 of the convention, which forms part of s 6 entitled Prorogation of jurisdiction, is worded
as follows:
If the parties, one or more of whom is domiciled in a Contracting State, have agreed that a court
or the courts of a Contracting State are to have 140 jurisdiction to settle any disputes which have
arisen or which may arise in connection with a particular legal relationship, that court or those
courts shall have exclusive jurisdiction. Such an agreement conferring jurisdiction shall be either:
(a) in writing or evidenced in writing, or (b) in a form which accords with practices which the parties
have established between themselves, or (c) in international trade or commerce, in a form which
accords with a usage of which the parties are or ought to have been aware and which in such trade
or commerce is widely known to, and regularly observed by, parties to contracts of the type
involved in the particular trade or commerce concerned. Where such an agreement is concluded by
parties, none of whom is domiciled in a Contracting State, the courts of other Contracting States
shall have no jurisdiction over their disputes unless the court or courts chosen have declined
jurisdiction
Agreements conferring jurisdiction shall have no legal force if they are contrary to the
provisions of Article 12 or 15, or if the courts whose jurisdiction they purport to exclude have
exclusive jurisdiction by virtue of Article 16. If an agreement conferring jurisdiction was concluded
for the benefit of only one of the parties, that party shall retain the right to bring proceedings in any
other court which has jurisdiction by virtue of this Convention.

Question 1
24.
To answer the first question from the national court, the Court of Justice must decide whether a
franchise agreement signed by a person who has not previously carried on a business should be
classified as a contract for a purpose which can be regarded as being outside his trade or profession,
within the meaning of art 13 of the Brussels Convention.
25.
I shall begin by examining franchise agreements and go on to consider the courts case law relating to
the term consumer contracts in art 13 of the convention. This will lead to my conclusion that the term is
not applicable to agreements of that kind.
Franchise agreements
26.
Franchise agreements are a widespread form of business arrangement whereby one undertaking, the
franchiser, grants to another, the franchisee, the right to exploit its own system for marketing goods or
services.
27.
Natural or legal persons who act as franchisers normally establish a network of franchises in a
particular field of commerce. They offer to include future franchisees in the network, by concluding a
contract setting out the principal stipulations of the bilateral agreement. In most cases this a standard-
form contract.
28.
The franchisee is legally independent. He is a genuinely independent trader carrying on his own
business and engaging in commercial transactions (he purchases goods from his supplier for resale to his
customers).
29.
The Court of Justice considered this form of business from the viewpoint of freedom of competition in
its judgment in Pronuptia de Paris GmbH v Pronuptia de Paris Irmgard Schillgalis Case 161/84 [1986]
ECR 353, which relates to atypical distribution franchise agreements under which the franchisee merely
sells certain products in a shop which bears the franchisers business name.
141
30.
In that judgment the court outlined the main features of the relationship between franchisers and
franchisees, stressing that the latter are independent traders ([1986] ECR 353 (para 15)):
In a system of distribution franchises of that kind, an undertaking which has established itself as
a distributor on a given market and thus developed certain business methods grants independent
traders, for a fee, the right to establish themselves in other markets, using its business name and
the business methods which have made it successful. Rather than a method of distribution, it is a
way for an undertaking to derive financial benefit from its expertise without investing its own capital.
Moreover, the system gives traders who do not have the necessary experience access to methods
which they could not have learned without considerable effort and allows them to benefit from the
reputation of the franchisers business name. (My emphasis.)
31.
Article 1 of Commission Regulation (EEC) 4087/88 on the application of art 85(3) of the Treaty to
categories of franchise agreements (OJ 1988 L359 p 46) also presumes that a franchise agreement is
entered into by two undertakings, that is to say two economic entities operating on a commercial basis 4.
4
Article 1 defines a franchise as a package of industrial or intellectual property rights relating to trade marks, trade
names, shop signs, utility models, designs, copyrights, know-how or patents, to be exploited for the resale of goods or
the provision of services to end users. A franchise agreement is defined as an agreement whereby one undertaking,
the franchiser, grants the other, the franchisee, in exchange for direct or indirect financial consideration, the right to
exploit a franchise for the purposes of marketing specified types of goods and/or services.

(ii) The courts case law relating to consumer contracts

32.
The judgment in Shearson Lehman Hutton Inc v Treuhand fr Vermgensverwaltung und
Beteiligungen (TVB) mbH Case C-89/91 [1993] ECR I-139 clearly set out the requirements for contracts
of this kind and the position to be taken by national courts when construing art 13 of the Brussels
Convention.
33.
That judgment begins by referring to
the principle, established by case law [the judgments in Socit Bertrand v Paul Ott KG Case
150/77 [1978] ECR 1431 (paras 1416, 19), Handte & Cie GmbH v Traitements mcano-chimiques
des surfaces (TMCS) Case C-26/91 [1992] ECR I-3967 (para 10)], according to which the concepts
used in the Convention, which may have a different content depending on the national law of the
Contracting States, must be interpreted independently, by reference principally to the system and
objectives of the Convention, in order to ensure that the Convention is uniformly applied in all the
Contracting States. This rule must apply in particular to the concept of consumer within the
meaning of Article 13 et seq of the Convention, in so far as that concept is the principal factor in the
determination of rules governing jurisdiction. (See [1993] ECR I-139 (para 13).)
34.
The judgment goes on to describe the interaction between the general and the special rules
concerning jurisdiction: under the system of the convention, the general principle, stated in the first
paragraph of art 2, is that the national courts of the contracting state in which the defendant is domiciled
are to have jurisdiction; it is only by way of derogation from that general principle that the convention
provides for the cases, exhaustively listed in ss 2 to 6 of Title II, in 142 which a defendant domiciled or
established in a contracting state may, where the situation comes under a rule of exclusive jurisdiction or
of prorogation of jurisdiction, be sued in the courts of another contracting state. Consequently, the rules of
jurisdiction which derogate from that general principle cannot give rise to an interpretation going beyond
the cases envisaged by the convention.
35.
In the judgment the court observes that
such an interpretation must apply a fortiori with respect to a rule of jurisdiction such as that
contained in Article 14 of the Convention, which allows a consumer, within the meaning of Article 13
of the Convention, to sue the defendant in the courts of the Contracting State in which the plaintiff
is domiciled. Apart from the cases expressly provided for, the convention appears clearly hostile
towards the attribution of jurisdiction to the courts of the plaintiffs domicile (See [1993] ECR I-
139 (para 17).)
36.
After those detailed observations concerning the principles for construing the rules governing
jurisdiction, the judgment clarifies the definition of consumer for the purpose of arts 13 and 14 of the
convention (see [1993] ECR I-139 (para 22)):
It follows from the wording and the function of those provisions that they affect only a private
final consumer, not engaged in trade or professional activities (see the judgment in Bertrand and
the Expert Report drawn up when the Kingdom of Denmark, Ireland and the United Kingdom of
Great Britain and Northern Ireland acceded to the Convention, OJ 1979 C59 p 71, paragraph 153),
who is bound by one of the contracts listed in Article 13 and who is a party to the action in
accordance with Article 14.
(iii) Application of those principles to franchise agreements
37.
In view of those considerations, it cannot be said that the parties to a franchise agreement are
consumers for the purpose of art 13 of the Brussels Convention.
38.
The status of consumer referred to by art 13 is not determined by a pre-existing subjective situation:
the same natural person may be a consumer for certain purposes and an entrepreneur for others. The
decisive factor is, therefore, not the personal circumstances of the individual but rather his position under
a particular contract, having regard to its scope and purpose.
39.
Where contracts such as franchise agreements are concerned, which are clearly of a commercial
nature and necessarily relate to a trade or profession 5 of the parties, the latters personal circumstances
before the formation of the contract are irrelevant for the purpose of art 13.
5
The term actividad profesional used in the Spanish version must, of course, be construed widely so as to include trade.
The English version of art 13 is more concrete as it refers to contracts concluded by a person for a purpose which can
be regarded as being outside his trade or profession. The German version also refers to both kinds of activity
berufliche oder gewerbliche Tatigkeit.

40.
Contrary to the appellants viewwhich was expounded at greater length during the oral procedureI
certainly do not consider that the court should disregard or water down the traditional principle of
autonomous interpretation of the terms, including consumer, used in the convention.
41.
In my opinion, the autonomous interpretation of consumer to which I referred in connection with the
judgment in Shearson Lehman Hutton is preferable 143 to an interpretation which relies on national law,
and there are two reasons for that view: (a) national legislation need not coincide from one state to
another and may differ slightly in certain respects, depending on the particular case. To rely on one body
of legislation rather than another (and what would be the criterion for making the choice?) would prejudice
the legal certainty which the Brussels Convention aims to ensure; (b) the same national legislation may
contain different definitions consumer, depending on the field of law in which they occur.
42.
In the opinion of counsel for the appellant, the court should give primacy to the German definition of
consumer deriving from the German Law on consumer credit (the Verbraucherkreditgesetz) 6, which
confers the status of consumers upon persons applying for credit in order to pursue an activity which they
had not previously taken up.
6
One of the grounds of the appellants action in the German court of first instance was that that law applied by analogy to
franchise agreements.

43.
I do not agree with that argument: it is also opposed by the German government itself which, in its
written observations, points out that that wider definition of consumer was expressly and intentionally
formulated by the national legislature to go beyond the minimum standard laid down by Council Directive
(EEC) 87/102 for the approximation of the laws, regulations and administrative provisions of the member
states concerning consumer credit (OJ 1987 L42 p 48), which the Law on consumer credit was intended
to implement so as to offer consumers a higher level of protection than that provided for by the
Community measure.
44.
According to the German government, the Community definition of consumer in Directive 87/102 7
excludes not only persons who are parties to contracts relating to a trade or profession already taken up
(in the actual words of the Law on consumer credit), but also generally persons who are parties to
contracts which are concluded for the purpose of a trade or profession.
7
This is also the criterion used in Council Directive (EEC) 93/13 on unfair terms in consumer contracts (OJ 1993 L95 p
29), art 2 of which defines consumer as any natural person who, in contracts covered by this Directive, is acting for
purposes which are outside his trade, business or profession, whereas dealer or supplier means any natural or legal
person who, in contracts covered by this Directive, is acting for purposes relating to his trade, business or profession,
whether publicly or privately owned.

45.
 In the same observations, the German government adds that in its legal system a narrower definition
of consumer is used in other consumer protection measures: for example, the Law on the cancellation of
doorstep sales (the Haustrwiderrufsgesetz).
46.
All this merely confirms the necessity to adhere to the autonomous interpretation of the definition of
consumer in art 13 of the Brussels Convention, which need not be linked to the definitions used in
individual cases in the respective national legal systems.
47.
Finally, the view I have just put forward is not contradicted by the inclusion in the EC Treaty of a new
Title XI8 on consumer protection, in which art 129a lays down a high level of consumer protection as an
objective of the Community. First, the legal scope of that provision is limited 9 and, secondly, art 129a(3)
expressly permits member states to maintain or introduce more stringent protective measures. It follows
logically that the Community level of 144 protection need not be identified with the level obtaining in one
or more of the member states.
8
Added by art G.38 of the Treaty on European Union (Maastricht, 7 February 1992; TS 12 (1994); Cmnd 2485; OJ 1992
C191 p 1).
9
According to para 19 of the judgment in El Corte Ingls SA v Blzquez Rivero Case C-192/94 [1996] I-1281. In para 20
of the same judgment the court stated that art 129a merely assigns an objective to the Community and confers powers
on it to that end without also laying down any obligation on member states or individuals.

48.
To sum up, it is in my view necessary to uphold the autonomous interpretation of the term consumer
used in the Brussels Convention, as the court did in the judgment in Shearson Lehman Hutton, which
means that the term must be limited to private final consumers not acting in the capacity of parties to
contracts concerning their trade or professional activities.
49.
Sometimes, no doubt, franchisees do not have previous business experience, but this does not justify
describing the activity covered by the franchise agreement as being outside a trade or profession. It is
precisely the activity in question and not, I emphasise, the existing personal circumstances of the party to
the agreement which was the factor taken into account when special rules of jurisdiction in relation to
certain contracts were laid down in art 13 of the convention.
50.
Therefore, the wording of art 13 does not permit it to be extended to cover any contract, irrespective of
its subject matter and purpose, in which an economically weaker party is faced by a party in a position
which is objectively superior or superior by reason of the circumstances.
51.
Contracting parties are not normally in a position of equality in the area of business relationships, but
that does not mean that contracts of that kind, including standard-form contracts, entered into by
businessmen benefit from the special rule in art 13. Although that provision aims to protect the weaker
party in a contractual relationship, its scope is limited to contracts in which one party is acting for
purposes unrelated to a business activity, that is to say as a private final consumer, not engaged in trade
or professional activities.
52.
In other words, the mere fact that one of the parties to a contract concluded with a view to the pursuit
of a trade or professional activity or in the course of such activities is in an inferior position, as in the case
of franchise agreements, is not regarded by the Brussels Convention as requiring special protection in
relation to the attribution of jurisdiction.
53.
Therefore I consider that the courts reply to the first question should be that art 13 of the Brussels
Convention does not apply to a contract such as that in the present case.

Question 2
54.
The national courts second question falls to be answered only if the courts reply to the first is that art
13 of the Brussels Convention does apply to a contract such as that in the present case. In view of my
proposed reply to the first question, it is unnecessary, in my opinion, to answer the second.
55.
Should the court find that art 13 is applicable, it will have to reply to the second question from the
Oberlandesgericht Mnchen concerning the interpretation of point 1 of the first paragraph of art 13,
relating to the sale of goods on instalment credit terms.
56.
Specifically, the national court asks whether or not the legal concept contract for the sale of goods on
instalment credit terms covers a franchise agreement under which one party undertakes to buy from the
franchiser, over a 145 period of three years, the articles and goods required to equip and operate a
business, without instalment credit terms having been agreed. In that connection the franchisees only
obligations are to pay an initial fee and, as from the second year, a licence fee of 3% of his turnover.
57.
The significance of the second question lies in the fact that art 13 not only requires the contract to
have been concluded for a purpose which can be regarded as being outside any trade or profession, but
also requires it to belong to one of the three categories described in points 1, 2 and 3 of the first
paragraph, point 1 being contracts for the sale of goods on instalment credit terms.
58.
In my view, the reply to this question must also be in the negative. Contracts for the sale of goods on
instalment credit terms cannot be confused with other contracts which, although entailing successive
obligations, possess features quite different from those of a sale on instalment credit terms.
59.
In the present case, the national court itself observes that the instalment credit rules do not apply to
the purchases which the franchisee undertakes to make from the franchiser during the three-year term of
the agreement, which makes no provision for the payment of instalments for the goods on successive,
predetermined dates.
60.
The fact that a contract provides for successive obligations to be performed by one or both parties
does not in itself justify treating it as a sale of goods on instalment credit terms.
61.
The periodical purchases which the franchisee has to make under a franchise agreement are the
consequence or effect of the initial contract which, as such, bears no resemblance at all to a sale on
instalment credit terms for the purpose of art 13.
62.
Furthermore, those periodical purchases are not in this case even subject to provisions relating to the
sale of goods on instalment credit terms.
63.
Still less can it be concluded from the fact that the franchisee has to pay at regular intervals a sum
equal to 3% of his annual turnover for the use of the franchisers business names that there is a sale of
goods on instalment credit terms. It is quite clear that in such circumstances no bilateral contract for the
sale of goods exists, whether by instalments or for cash.
64.
Consequently, if a reply is required to the second question, in my opinion the court should answer in
the negative.

Question 3
65.
The third question is wider in scope. In essence, the national court asks whether, pursuant to the first
paragraph of art 17 of the Brussels Convention, the court with exclusive jurisdiction designated by the
parties in a jurisdiction clause 16 is also competent to adjudicate in an action for the annulment of the
contract containing the jurisdiction clause.
66.
Let me begin with two points which seem to me important: (a) it is not disputed that the jurisdiction
clause in the agreement in question meets the formal requirements of art 17 of the convention 10; (b) the
jurisdiction clause is in the most general terms possible, as it relates to any dispute concerning the 146
interpretation, performance or other aspect of this contract, and it is stipulated that any such dispute must
be brought before the courts of Florence.
10
Nor is it disputed that the clause complies with the formal requirements of Italian law as it is a contractual stipulation
separately approved in accordance with arts 1341 and 1342 of the Italian Codice Civile. In any case, there is no
question here of applying national law, whether Italian or German: the issue is whether the clause complies with the
Brussels Convention.

67.
In my view, an agreement conferring jurisdiction such as that in the present case, which is formally
valid under the convention and has been stipulated by the parties in order to settle future differences of
any kind concerning any aspect of the contract, is applicable to any legal disputes which may arise,
including those relating to the requirements for the validity of the contract containing the jurisdiction
clause.
68.
In its judgment in Effer SpA v Kantner Case 38/81 [1982] ECR 825 the court examined a similar
problem relating to the ambit of art 5 of the convention. The question in that case was whether the courts
for the place of performance of the contract had jurisdiction when the dispute between the parties
concerned the very existence of the main contract or the conditions for its formation.
69.
The court found that the national courts jurisdiction to determine questions relating to a contract
included the power to consider the existence of the constituent parts of the contract itself since that was
indispensable in order to enable the national court to examine whether it had jurisdiction under art 5 of the
convention.
70.
That view was strengthened by consideration of the harmful effects on legal certainty 11 if that were not
the case. The provisions of the Brussels Convention would be liable to be deprived of their legal effect if it
were accepted that, in order to defeat them, it would be sufficient for one of the parties merely to claim
that the contract did not exist.
 11
The objective of the Brussels Convention is precisely to ensure legal certaintyand, specifically, certainty or
foreseeability in designating a competent court: the objectives [of the Conventionj include unification of the rules on
jurisdiction of the contracting states, so as to avoid as far as possible the multiplication of the bases of jurisdiction in
relation to one and the same legal relationship and to reinforce the legal protection available to persons established in
the Community by, at the same time, enabling the plaintiff easily to identify the court before which he may bring an
action and the defendant reasonably to foresee the court before which he may be sued (see the judgment in Mulox IBC
Ltd v Geels Case C-125/92 [1993] ECR I-4075).

71.
The court added that, on the contrary, respect for the aims and spirit of the convention demanded that
its provisions should be construed as meaning that the court called upon to decide a dispute arising out of
a contract may examine, even of its own motion, the essential preconditions for its jurisdiction, having
regard to conclusive and relevant evidence adduced by the party concerned, establishing the existence or
the non-existence of the contract.
72.
The same arguments are applicable by analogy to the present case, where the issue is the validity, not
the existence, of the contract. The difference between this situation and that considered in the judgment
in Effer is that territorial jurisdiction was determined in the latter case by a legal criterion (the place of
performance of the obligation), and not by a jurisdiction clause in the contract. In my opinion, however, the
legal reasoning used in the earlier case applies equally to both situations.
73.
It is possible to reach the same conclusion by analysing the nature of the jurisdiction clauses provided
for by art 17 of the convention. In my opinion, it must be recognised that they are to some extent
independent of the contract of which they form part.
74.
This question is the subject of a well known academic controversy 12 which has not been settled.
However, I consider that, so far as art 17 is concerned, the court should take the same approach as in the
Effer judgment and incline towards 147 the view which promotes legal certainty and which, specifically,
means recognising the court chosen in a jurisdiction clause (provided, of course, that the latter fulfils the
requirements of art 17) as the competent court, even when it is alleged that the contract containing the
clause is void.
12
See the recent works of Blanchin Lautonomie de la clause compromissoire: Un modle pour la clause attributive
dejuridiction? (1995) and Rodrguez Benot Los acuerdos atributivos de competencia judicial internacional en Derecho
conzunitano europeo (1994).

75.
There are a number of reasons for taking this approach. First, jurisdiction clauses do not depend on
the economic or legal factors forming the basis of the contract, nor is the cause of the contract the same
as that of the jurisdiction clause, the purpose of which is merely procedural (to locate within a particular
forum proceedings to deal with any future disputes). Consequently, any grounds of nullity which might
affect the substantive elements of a contract should not have any impact on jurisdiction clauses.
76.
Secondly, if one party alleges that there was no consensus ad idemfor example, that there was a
mistake as to essential aspects of the subject matter, rendering the mutual obligations voidthe
jurisdiction clause is not necessarily affected because the mistake does not extend to the express choice
of the competent court. This applies a fortiori where the alleged grounds of nullity of the contract relate to
its compatibility or otherwise with the substantive law of a particular national system.
77.
Thirdly, if a court other than that designated by the parties in a jurisdiction clause were allowed to
decide on the validity of the contract in general, the practical consequences would be rather
disconcerting. If, for example, such proceedings led to a decision that the contract in general was valid,
the court in question would immediately have to decline jurisdiction in favour of the court chosen by the
parties, which has exclusive jurisdiction to settle their differences. It would be difficult to deny that the
latter court in turn would not be competent to rule that the contract or any of its essential stipulations was
invalid, even if by so doing it contradicted the judgment of the first court.
78.
Finally, the approach that I advocate has the advantage of avoiding concurrent litigation and
preventing circumvention of the very system whereby only one forum should have jurisdiction, which
forms the basis of the Brussels Convention. Merely by claiming that the contract containing the jurisdiction
clause was void, either party would displace the jurisdiction rules and thereby render art 17 ineffective.
That would certainly detract from certainty and foreseeability in designating the competent court.
79.
In that connection I should point out that neither party to this action has questioned the validity of the
jurisdiction clause itself on substantive13 or on formal grounds. Mr Benincasa merely claims that the
franchise agreement is void in general on grounds based on substantive German law (alleged
infringement of the BGB and the German Law on standard business conditions) 14.
13
The legal system of a particular state may lay down certain substantive conditions for the validity of jurisdiction clauses.
It is not clear whether such provisions would be consistent with art 17 of the Brussels Convention. So far as formal
requirements are concerned, it is clear that art 17 is the only permissible point of reference.
14
See para 10 of this opinion, above.
80.
The response to those allegations, that is to say the decision on the validity of the franchise
agreement, will depend on the substantive law found to be applicable to it. However, I consider that the
competent court to give a ruling on 148 that pointsince the parties have stipulated a jurisdiction clause
in such general termsmust be precisely the court designated by them beforehand.
81.
The parties intention expressed in the jurisdiction clause is clear: any dispute concerning any aspect
of the contract (which must include disputes on Validity) falls within the jurisdiction of the courts of
Florence.
82.
On this point it is inevitable that the courts reply cannot be limited to a hypothetical interpretation of art
17 of the Brussels Convention, unconnected with the dispute with which the question from the national
court is concerned. Without any intention to take the place of the competent national court in interpreting
the agreement between the parties, a helpful reply from the Court of Justice, under the preliminary ruling
procedure, necessitates an analysis of the terms of the jurisdiction clause in order, by reference to its
characteristics, to provide the national court with an interpretation of the Brussels Convention, as
requested.

Conclusion
83.
I therefore propose that the Court of Justice reply as follows to the questions from the
Oberlandesgericht Mnchen:
(1) The parties to a franchise agreement for the forthcoming opening of a commercial
establishment cannot be regarded as consumers within the meaning of the first paragraph of art 13
and the first paragraph of art 14 of the Convention on Jurisdiction and the Enforcement of
Judgments in Civil and Commercial Matters 1968.
(2) The court designated in a jurisdiction clause as having jurisdiction to entertain any dispute
relating to the interpretation, performance or other aspect of this contract has exclusive jurisdiction,
in accordance with the first sentence of the first paragraph of art 17 of the 1968 convention, even
where the action seeks, inter alia, annulment of the contract containing the jurisdiction clause.

3 July 1997.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By order of 5 May 1995, received at the Court of Justice of the European Communities on 9 August
1995, the Oberlandesgericht (the Higher Regional Court), Mnchen, referred to the Court of Justice for a
preliminary ruling under the Protocol of 3 June 1971 on the interpretation by the Court of Justice of the
Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968, as
amended (the Brussels Convention), three questions on the interpretation of the first paragraphs of arts
13, 14 and 17 of that convention.
2.
Those questions were raised in proceedings between Dentalkit Srl, having its seat in Florence, and Mr
Benincasa, an Italian national, relating to the validity of a franchising contract concluded between them.
3.
According to the case file relating to the main proceedings, in 1987 Dentalkit developed a chain of
franchised shops in Italy specialising in the sale of dental hygiene products. In 1992 Mr Benincasa
concluded a franchising contract with Dentalkit with a view to setting up and operating a shop in
Mnchen. In that contract Dentalkit authorised Mr Benincasa to exploit the exclusive right to use the
Dentalkit trade mark within a particular geographical area. Dentalkit further undertook to supply goods
bearing that trade mark, to support him in 149 various spheres, to carry out the requisite training and
promotion and advertising activities and not to open any shop within the geographical area covered by the
exclusive right.
4.
For his part, Mr Benincasa undertook to equip business premises at his own cost, to stock exclusively
Dentalkits products, not to disclose any information or documents concerning Dentalkit and to pay it a
sum of Lit 8m as payment for the cost of technical and commercial assistance provided when opening the
shop and 3% of his annual turnover. By reference to arts 1341 and 1342 of the Italian Civil Code, the
parties specifically approved a clause of the contract reading The courts at Florence shall have
jurisdiction to entertain any dispute relating to the interpretation, performance or other aspects of the
present contract by separately signing it.
5.
Mr Benincasa set up his shop, paid the initial sum of Lit 8m and made several purchases, for which,
however, he never paid. In the meantime, he has ceased trading altogether.
6.
 Mr Benincasa brought proceedings in the Landgericht (the Regional Court), Mnchen I, where he
sought to have the franchising contract declared void on the ground that the contract as a whole was void
under German law. He also claimed that the sales contracts concluded subsequently pursuant to the
basic franchising contract were void.
7.
Mr Benincasa argued that the Landgericht Mnchen I had jurisdiction as the court for the place of
performance of the obligation in question within the meaning of art 5(1) of the Brussels Convention. He
argued that the clause of the franchising contract conferring jurisdiction on the courts at Florence did not
have the effect of derogating from art 5(1) as regards his action to avoid the contract because that action
sought to have the whole franchising agreement declared void and, therefore, also the jurisdiction clause.
Mr Benincasa further argued that, since he had not yet started trading, he should be regarded as a
consumer within the meaning of the first paragraphs of arts 13 and 14 of the convention.
8.
The relevant provisions of the convention read as follows:
Article 13
In proceedings concerning a contract concluded by a person for a purpose which can be
regarded as being outside his trade or profession, hereinafter called the consumer, jurisdiction
shall be determined by this Section, without prejudice to the provisions of point 5 of Articles 4 and
5, if it is: (1) a contract for the sale of goods on instalment credit terms
Article 14
A consumer may bring proceedings against the other party to a contract either in the courts of
the Contracting State in which that party is domiciled or in the courts of the Contracting State in
which he is himself domiciled.
9.
The Landgericht Mnchen I declined jurisdiction on the ground that the jurisdiction clause contained in
the franchising contract was valid and that the contract was not a contract concluded by a consumer.
10.
Mr Benincasa appealed against that decision to the Oberlandesgericht Mnchen, which stayed
proceedings and referred the following questions to the Court of Justice for a preliminary ruling:
150
(1) Is a plaintiff to be regarded as a consumer within the meaning of the first paragraph of
Article 13 and the first paragraph of Article 14 of the Convention even if his action relates to a
contract which he concluded not for the purpose of a trade which he was already pursuing but a
trade to be taken up only at a future date (here: a franchising agreement concluded for the purpose
of setting up a business)?
(2) If Question 1 is to be answered in the affirmative: Does point 1 of the first paragraph of
Article 13 of the Convention (contract for the sale of goods on instalment credit terms) cover a
franchising agreement which obliges the plaintiff to buy from the other party to the agreement, over
a period of several (three) years, the articles and goods required to equip and operate a business
(without instalment credit terms having been agreed) and to pay an initial fee and, as from the
second year of the business, a licence fee of 3% of turnover?
(3) Does the court of a Member State specified in an agreement conferring jurisdiction have
exclusive jurisdiction pursuant to the first paragraph of Article 17 of the Convention even when the
action is inter alia for a declaration of the invalidity of a franchising agreement containing the
jurisdiction clause itself, which is worded The courts at Florence shall have jurisdiction to entertain
any dispute relating to the interpretation, performance or other aspects of the present contract, that
clause having been specifically approved within the meaning of Articles 1341 and 1342 of the
Italian Civil Code?

The first question

11.
The point sought to be clarified by the national courts first question is whether the first paragraphs of
arts 13 and 14 of the Brussels Convention must be interpreted as meaning that a plaintiff who has
concluded a contract with a view to pursuing a trade or profession, not at the present time but in the
future, may be regarded as a consumer.
12.
In this connection, regard should be had to the principle laid down by the case law according to which
the concepts used in the convention, which may have a different content depending on the national law of
the contracting states, must be interpreted independently, by reference principally to the system and
objectives of the convention, in order to ensure that the convention is uniformly applied in all the
contracting states (see esp the judgments in Socit Bertrand v Paul Ott KG Case 150/77 [1978] ECR
1431 (paras 1416, 19) and Shearson Lehman Hutton Inc v Treuhand fr Vermgensverwaltung und
Beteiligungen (TVB) mbH Case C-89/91 [1993] ECR I-139 (para 13)). This must apply in particular to the
concept of consumer within the meaning of arts 13ff of the convention, in so far as it determines the rules
governing jurisdiction.
13.
 It must next be observed that, as the court has consistently held, under the system of the convention
the general principle is that the courts of the contracting state in which the defendant is domiciled are to
have jurisdiction and that it is only by way of derogation from that principle that the convention provides
for cases, which are exhaustively listed, in which the defendant may or must, depending on the case, be
sued in the courts of another contracting state. Consequently, the rules of jurisdiction which derogate from
that general principle cannot give rise to an interpretation going beyond the cases envisaged 151 by the
convention (see the judgment in Shearson Lehman Hutton [1993] ECR I-139 (paras 1416)).
14.
Such an interpretation must apply a fortiori with respect to a rule of jurisdiction, such as that contained
in art 14 of the convention, which allows a consumer, within the meaning of art 13, to sue the defendant in
the courts of the contracting state in which the plaintiff is domiciled. Apart from the cases expressly
provided for, the convention appears hostile towards the attribution of jurisdiction to the courts of the
plaintiffs domicile (see the judgments in Dumez France v Hessische Landesbank (Helaba) Case C-
220/88 [1990] ECR I-49 (paras 16, 19) and Shearson Lehman Hutton [1993] ECR I-139 (para 17)).
15.
As far as the concept of consumer is concerned, the first paragraph of art 13 of the convention
defines a consumer as a person acting for a purpose which can be regarded as being outside his trade
or profession. According to settled case law, it follows from the wording and the function of that provision
that it affects only a private final consumer, not engaged in trade or professional activities (see the
judgment in Shearson Lehman Hutton [1993] ECR I-139 (paras 20, 22)).
16.
It follows from the foregoing that, in order to determine whether a person has the capacity of a
consumer, a concept which must be strictly construed, reference must be made to the position of the
person concerned in a particular contract, having regard to the nature and aim of that contract, and not to
the subjective situation of the person concerned. As the Advocate General rightly observed in para 38 of
his opinion, above, the self-same person may be regarded as a consumer in relation to certain
transactions and as an economic operator in relation to others.
17.
Consequently, only contracts concluded for the purpose of satisfying an individuals own needs in
terms of private consumption come under the provisions designed to protect the consumer as the party
deemed to be the weaker party economically. The specific protection sought to be afforded by those
provisions is unwarranted in the case of contracts for the purpose of trade or professional activity, even if
that activity is only planned for the future, since the fact that an activity is in the nature of a future activity
does not divest it in any way of its trade or professional character.
18.
Accordingly, it is consistent with the wording, the spirit and the aim of the provisions concerned to
consider that the specific protective rules enshrined in them apply only to contracts concluded outside and
independently of any trade or professional activity or purpose, whether present or future.
19.
The answer to the national courts first question must therefore be that the first paragraphs of arts 13
and 14 of the convention must be interpreted as meaning that a plaintiff who has concluded a contract
with a view to pursuing a trade or profession, not at the present time but in the future, may not be
regarded as a consumer.

The second question

20.
In view of the answer given to the first question, there is no need to answer the second.

The third question

21.
The point sought to be clarified by the national courts third question is whether the courts of a
contracting state which have been designated in a jurisdiction clause validly concluded under the first
paragraph of art 17 of the 152 convention also have exclusive jurisdiction where the action seeks in
particular a declaration that the contract containing that clause is void.
22.
The national court also raises the question whether a jurisdiction clause validly concluded under the
rules of the convention and contained in the main contract must be considered on its own, independently
of any allegation as to the validity of the remainder of the contract.
23.
The first paragraph of art 17 provides as follows:
If the parties, one or more of whom is domiciled in a Contracting State, have agreed that a court
or the courts of a Contracting State are to have jurisdiction to settle any disputes which have arisen
or which may arise in connection with a particular legal relationship, that court or those courts shall
have exclusive jurisdiction. Such an agreement conferring jurisdiction shall be either in writing
24.
A distinction must first be drawn between a jurisdiction clause and the substantive provisions of the
contract in which it is incorporated.
25.
 A jurisdiction clause, which serves a procedural purpose, is governed by the provisions of the
convention, whose aim is to establish uniform rules of international jurisdiction. In contrast, the
substantive provisions of the main contract in which that clause is incorporated, and likewise any dispute
as to the validity of that contract, are governed by the lex causae determined by the private international
law of the state of the court having jurisdiction.
26.
Next, as the court has consistently held, the objectives of the convention include unification of the
rules on jurisdiction of the contracting statess courts, so as to avoid as far as possible the multiplication of
the bases of jurisdiction in relation to one and the same legal relationship and to reinforce the legal
protection available to persons established in the Community by, at the same time, allowing the plaintiff
easily to identify the court before which he may bring an action and the defendant reasonably to foresee
the court before which he may be sued (see the judgments in Effer SpA v Kantner Case 38/81 [1982]
ECR 825 (para 6) and Mulox IBC Ltd v Geels Case C-125/92 [1993] ECR I-4075 (para 11)).
27.
It is also consonant with that aim of legal certainty that the court seised should be able readily to
decide whether it has jurisdiction on the basis of the rules of the convention, without having to consider
the substance of the case.
28.
The aim of securing legal certainty by making it possible reliably to foresee which court will have
jurisdiction has been interpreted in connection with art 17 of the convention, which accords with the
intentions of the parties to the contract and provides for exclusive jurisdiction by dispensing with any
objective connection between the relationship in dispute and the court designated, by fixing strict
conditions as to form (see, in this regard, Mainschiffahrts- Genossenschaft eG (MSG) v Les Gravires
Rhnanes SARL Case C-106/95 [1997] All ER (EC) 385, [1997] ECR I-911 (para 34)).
29.
Article 17 of the convention sets out to designate, clearly and precisely, a court in a contracting state
which is to have exclusive jurisdiction in accordance with the consensus formed between the parties,
which is to be expressed in accordance with the strict requirements as to form laid down therein. The
legal certainty which that provision seeks to secure could easily be jeopardised if one party to the contract
could frustrate that rule of the convention simply by claiming that the whole of the contract was void on
grounds derived from the applicable substantive law.
153
30.
That solution is consistent not only with the approach taken by the court in Effer, in which it ruled that
the plaintiff may invoke the jurisdiction of the courts of the place of performance in accordance with art
5(1) of the convention even when the existence of the contract on which the claim is based is in dispute
between the parties, but also with the judgment in Sanders v van der Putte Case 73/77 [1977] ECR 2383
(para 15), in which the court held, in connection with art 16(1) of the convention, that, in the matter of
tenancies of immovable property, the courts of the state in which the immovable property is situated
continue to have jurisdiction even where the dispute is concerned with the existence of the lease.
31.
It must be added that, as the court has held, it is for the national court to interpret the clause conferring
jurisdiction invoked before it in order to determine which disputes fall within its scope (see the judgment in
Powell Duffryn plc v Petereit Case C-214/92 [1992] ECR I-1745 (para 37)). Consequently, in the instant
case it is for the national court to determine whether the clause invoked before it, which refers to any
dispute relating to the interpretation, performance or other aspects of the contract, also covers any
dispute relating to the validity of the contract.
32.
The answer to the national courts third question must therefore be that the courts of a contracting
state which have been designated in a jurisdiction clause validly concluded under the first paragraph of
art 17 of the convention also have exclusive jurisdiction where the action seeks in particular a declaration
that the contract containing that clause is void.

Costs
33.
The costs incurred by the German government and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the proceedings pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the questions referred to it by
the Oberlandesgericht Mnchen by order of 5 May 1995, hereby rules: (1) The first paragraphs of arts 13
and 14 of the Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial
Matters 1968, as amended, must be interpreted as meaning that a plaintiff who has concluded a contract
with a view to pursuing a trade or profession, not at the present time but in the future, may not be
regarded as a consumer. (2) The courts of a contracting state which have been designated in a
jurisdiction clause validly concluded under the first paragraph of art 17 of the 1968 convention also have
exclusive jurisdiction where the action seeks in particular a declaration that the contract containing that
clause is void.

154
 [1998] All ER (EC) 155

Inter-Environnement Wallonie ASBL v Rgion Wallonie

(Case C-129/96)
EUROPEAN COMMUNITY; Environment
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, RAGNEMALM AND SCHINTGEN
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, KAPTEYN, MURRAY, EDWARD,
PUISSOCHET, HIRSCH, JANN AND SEVN (RAPPORTEUR) ADVOCATE GENERAL JACOBS
24 APRIL, 5 FEBRUARY, 18 DECEMBER 1997
European Community Environment Waste Whether concept of waste excluding substances
forming integral part of industrial production process Council Directive (EEC) 75/442, art 1(a).
European Community Directives Period for transposition Member states obligations during
transposition period Council Directive (EEC) 91/156 EC Treaty, arts 5, 189.

A Belgian regional council adopted an order implementing Community provisions on waste before the
period laid down in those provisions for their transposition into national law had expired. The Community
provisions required any undertaking carrying out waste disposal or recovery to obtain a permit, but
provided exemptions for companies carrying out waste disposal on their own premises at the place of
production and companies carrying out waste recovery on condition, inter alia, that the waste was non-
hazardous. The plaintiff environmental group brought proceedings against the regional council, seeking
annulment of the order on the grounds that it infringed the relevant Community provisions by excluding
from the permit system installations intended for the collection, pre-treatment, disposal or recovery of
toxic or dangerous waste, where they formed an integral part of an industrial production process. The
national court stayed the proceedings and referred to the Court of Justice of the European Communities
for a preliminary ruling the questions: (i) whether a substance was excluded from the definition of waste
in art 1(a)1 of Council Directive (EEC) 75/442 on waste, as amended by Council Directive (EEC) 91/156,
merely because it directly or indirectly formed an integral part of an industrial production process; and (ii)
whether arts 52 and 1893 of the EC Treaty precluded member states from adopting measures contrary to
Directive 91/156 during the period prescribed for its transposition.
1
Article 1(a), so far as material, is set out at p 172 c, post
2
Article 5, so far as material, provides: Member States shall take all appropriate measures to ensure fulfilment of the
obligations arising out of action taken by the institutions
3
Article 189, so far as material, provides: A Directive shall be binding, as to the result to be achieved, upon each Member
State to which it is addressed, but shall leave to the national authorities the choice of form and methods.

Held (1) On a proper construction of Directive 72/442, the scope of the term waste in art 1(a) turned on
the meaning of discard, which covered both disposal and recovery of a substance or object. The concept
of waste did not therefore exclude any kind of residue, industrial by-product or other substance arising
from production processes and was not to be understood as excluding substances and 155 objects
capable of economic reutilisation. It followed that a substance was not excluded from the definition of
waste in art 1(a) of the directive by the mere fact that it directly or indirectly formed an integral part of an
industrial production process. That conclusion did not, however, undermine the distinction which had to be
drawn between waste recovery within the meaning of the directive and the normal industrial treatment of
products which were not waste (see p 175 b to e j and p 177 h, post); Criminal proceedings against
Tombesi Joined cases C-304/94, C-330/94, C-342/94 and C-224/95 [1997] All ER (EC) 639 applied.
(2) In accordance with art 191 of the EC Treaty, a directive had legal effect with respect to the member
state to which it was addressed from the moment of its notification. During the subsequent transposition
period, member states had to take the measures necessary to ensure that the result prescribed by the
directive was achieved at the end of that period and although they were not obliged to adopt those
measures before the end of the transposition period, arts 5 and 189 of the Treaty and Directive 91/156
required that they should not take any measures which could seriously compromise the result during that
period. It followed from the Treaty that, during the transposition period, member states had to refrain from
adopting measures liable to compromise seriously the aims of the directive and, in assessing whether that
had happened, the national court had to consider whether the national measure constituted full
transposition of the directive as well as the effects in practice of applying those provisions and of their
duration in time (see p 176 h and p 177 a to c e j, post).

Notes
For a general outline of Community provisions on waste disposal, see 51 Halsburys Laws (4th edn) paras
822826.
For the EC Treaty, art 5, see 50 Halsburys Statutes (4th edn) 267 and for art 189 (as amended by art
G.60 of the Treaty on European Union), see ibid, Current Service, 101.
Cases cited
Aannemersbedrijf P K Kraaijeveld BV v Gedeputeerde Staten van Zuid-Holland Case C-72/95 [1997] All
ER (EC) 134, [1996] ECR I-5403, ECJ.
EC Commission v UK Case 804/79 [1981] ECR 1045.
European Commission v Germany Case C-422/92 [1995] ECR I-1097.
France v EC Commission Joined cases 1516/76 [1979] ECR 321.
Hansa Fleisch Ernst Munde GmbH & Co KG v Landrat des Kreises Schleswig-Flensburg Case C-156/91
[1992] ECR I-5567.
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135.
Marshall v Southampton and South West Hampshire Health Authority (Teaching) Case 152/84 [1986] 2 All
ER 584, [1986] QB 401, [1986] 2 WLR 780, [1986] ECR 723, ECJ.
Opel Austria GmbH v EU Council Case T-115/94 [1997] All ER (EC) 97, ECJ.
Peskeloglou v Bundesanstalt fr Arbeit Case 77/82 [1983] ECR 1085.
Pubblico Ministero v Ratti Case 148/78 [1979] ECR 1629.
Teuling v Bedrijfsvereniging voor de Chemische Industrie Case 30/85 [1987] ECR 2497.
Tombesi (Criminal proceedings against) Joined cases C-304/94, C-330/94, C-342/94 and C-224/95
[1997] All ER (EC) 639, [1997] ECR I-3561, ECJ.
Vaneetveld v La Foyer SA Case C-316/93 [1994] ECR I-763.
156
Verbond van Nederlandse Ondernemingen v Inspecteur der Invoerrechten en Accijnzen Case 51/76
[1977] ECR 113.
Zanetti (Criminal proceedings against) Case C-359/88 [1990] ECR I-1509.

Reference
By judgment of 29 March 1996, the Belgian Conseil dtat referred to the Court of Justice of the European
Communities for a preliminary ruling under art 177 of the EC Treaty two questions (set out at p 174 h j,
post) on the interpretation of arts 5 and 189 of the Treaty and art 1(a) of Council Directive (EEC) 75/442
on waste, as amended by Council Directive (EEC) 91/156. Those questions were raised in proceedings
brought by Inter-Environnement Wallonie ASBL, a non-profit-making association, for annulment of the
order of the Walloon Regional Executive of 9 April 1992 on toxic or hazardous waste. Written
observations were submitted on behalf of: Inter-Environnement Wallonie, by J Sambon, of the Brussels
Bar; the Belgian government, by J Devadder, Senior Adviser in the Ministry of Foreign Affairs, External
Trade and Development Cooperation, acting as agent; the German government, by E Rder,
Ministerialrat in the Federal Ministry of Economic Affairs, and B Kloke, Oberregierungsrat in the same
ministry, acting as agents; the French government, by J-F Dobelle, Deputy Director in the Legal
Directorate of the Ministry of Foreign Affairs, and R Nadal, Assistant Foreign Affairs Secretary in that
directorate, acting as agents; the Netherlands government, by A Bos, Legal Adviser in the Ministry of
Foreign Affairs, acting as agent; the UK government, by J E Collins, Assistant Treasury Solicitor, acting as
agent, and D Wyatt QC; and the European Commission, by M Condou-Durande, of its Legal Service,
acting as agent. Oral observations were made by: Inter-Environnement Wallonie ASBL, represented by J
Sambon; the French government, represented by J-F Dobelle and R Nadal; the Netherlands government,
represented by J Steven van den Oosterkamp, Deputy Legal Adviser in the Ministry of Foreign Affairs,
acting as agent; the UK government, represented by D Wyatt QC; and the Commission, represented by M
Condou Durande. The language of the case was French. The facts are set out in the opinion of the
Advocate General.

24 April 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
In the main proceedings Inter-Environnement Wallonie ASBL (Inter- Environnement) is asking the
Belgian Conseil dtat to annul, in whole or in part, the regulation of the Walloon Regional Executive of 9
April 1992 on toxic or hazardous waste. In its judgment of 29 March 1996 seeking a preliminary ruling the
Conseil dtat has already adjudicated on five of Inter-Environnements six pleas and annulled certain
provisions of the regulation. Inter-Environnements remaining plea has prompted the Conseil dtat to
seek a ruling on two points of Community law concerning, first, its power to review the legality of a
national measure adopted before the final date for the implementation of a directive and, secondly, the
scope of the notion of waste in the Community legislation on waste, with particular reference to materials
produced or used in industrial processes.
2.
Inter-Environnements plea is concerned specifically with art 5(1) of the national regulation, which
provides:
Authorization is required for the setting up and running of an installation intended specifically for
the collection, pre-treatment, disposal or recovery of toxic or dangerous waste which is not an
 integral part of an industrial production process and which processes waste originating from third
parties
157
3.
In Inter-Environnements view that provision wrongly excludes from the authorisation requirement toxic
or dangerous waste forming an integral part of an industrial process. Its plea is in two parts.
4.
Inter-Environnement contends first that art 5(1) of the regulation is contrary to art 11 of Council
Directive (EEC) 75/442 on waste (OJ 1975 L194 p 39), as amended by Council Directive (EEC) 91/156
(OJ 1991 L8 p 32), and to art 3 of Council Directive (EEC) 91/689 on hazardous waste (OJ 1991 L377 p
20).
5.
In its judgment the Conseil dtat concludes that art 5(1) of the regulation does not comply with those
provisions. Articles 9 and 10 of Directive 75/442 lay down permit requirements for establishments or
undertakings which carry out the waste disposal or recovery operations covered by the directive. Article
11 of the directive allows member states to exempt from the permit requirement establishments or
undertakings carrying out their own waste disposal at the place of production (art 11(1)(a)) and
establishments or undertakings that carry out waste recovery (art 11(1)(b)). However, that exemption may
apply only if certain conditions are met: the competent authorities must have adopted general rules for
each type of activity laying down certain matters; moreover, the types or quantities of waste and the
methods of disposal or recovery must be such that the basic aims of the directive of preventing danger to
human health and harm to the environment as laid down in art 4 of the directive are met. The Conseil
dtat notes that those conditions have not been implemented in Belgian law. Moreover, art 3(1) of
Directive 91/689 provides that the power, provided for in art 11(1)(a) of Directive 75/442, to exempt
establishments or undertakings carrying out waste disposal does not apply to hazardous waste covered
by that directive.
6.
The Conseil dtat notes however that the contested regulation was adopted before the final date for
implementing Directive 91/156, namely 1 April 1993; the applicants plea therefore appears to fall foul of
the rule of Belgian administrative law according to which the validity of a measure is to be assessed at the
moment of its adoption.
7.
In the second part of its plea Inter-Environnement argues that art 5(1) of the regulation infringes
certain provisions of the Decree of the Walloon Regional Council of 5 July 1985 on waste, in particular art
3(1). That provision, as amended by the Decree of 25 July 1991, defines waste as: all substances or
objects in the categories set out in Annex I which the holder discards or intends or is required to discard.
8.
That provision is intended to give effect to the similarly worded art 1 of Directive 75/442, as amended.
The Conseil dtat considers that Inter- Environnements plea raises the issue whether a substance or
object which, directly or indirectly, forms part of an industrial production process is waste within the
meaning of the Community and national provisions.
9.
The Conseil dtat therefore seeks a preliminary ruling from the Court of Justice on the following
questions:
(1) Do Articles 5 and 189 of the EEC Treaty preclude Member States from adopting a provision
contrary to Directive 75/442/EEC of 15 July 1975 on waste, as amended by Directive 91/156/EEC
of 18 March 1991, before the period for transposing the latter has expired? Do those same Treaty
articles preclude Member States from adopting and bringing into force legislation which purports to
transpose the abovementioned directive but whose provisions appear to be contrary to the
requirements of that directive?
158
(2) Is a substance referred to in Annex I to Council Directive 91/156/EEC of 18 March 1991
amending Directive 75/442/EEC on waste and which directly or indirectly forms an integral part of
an industrial production process to be considered waste within the meaning of Article 1(a) of that
directive?
10.
The court has received written observations from Inter-Environnement, the Belgian, French, German,
Netherlands and UK governments and the European Commission. With the exception of the Belgian and
German governments, the foregoing were also represented at the hearing.
11.
Before I examine the national courts first question it may be helpful to set out certain basic principles
concerning directives which flow from the Treaty and from the existing case law.
12.
Under art 189 of the Treaty a directive is binding, as to the result to be achieved, upon each member
state to which it is addressed, but leaves to the national authorities the choice of form and methods. By
virtue of art 191, as amended by the Treaty on European Union (Maastricht, 7 February 1992; TS 12
(1994); Cmnd 2485; OJ 1992 C191 p 1) with effect from 1 November 1993, directives addressed to all
member states enter into force on the date specified therein or, if no date is specified, on the 20th day
following their publication. Other directives take effect from the date of their notification to the addressees.
Prior to its amendment, art 191 provided that all directives were to take effect from the date of their
notification to their addressees.
13.
Although entering into force or taking effect on the above dates, directives which unlike regulations are
not directly applicable, invariably allow member states a period of time in which to adopt the laws,
regulations and administrative provisions necessary to ensure implementation of the directive in national
law.
14.
In its judgment in Pubblico Ministero v Ratti Case 148/78 [1979] ECR 1629 (paras 22, 24) the court
held that
(14) a Member State which has not adopted the implementing measures required by the
directive in the prescribed periods may not rely, as against individuals, on its own failure to perform
the obligations which the directive entails
(24) after the expiration of the period fixed for the implementation of a directive a Member
State may not apply its internal law which has not yet been adapted in compliance with the
directive, to a person who has complied with the requirements of the directive.
15.
In the same case the court was also asked whether a directive was immediately and directly
applicable with regard to the obligations imposed on member states as from the date of notification of the
directive in a case where a person had complied with the provisions of the directive before the expiry of
the period laid down for implementation. The court replied that it was only at the end of that period and in
the event of the member states default that the directive could have effects in relation to individuals and
that until that date was reached member states retained their freedom of action (see [1979] ECR 1629
(paras 4344 and point 5, operative part)) 4.
4
See also the judgment in Vaneetveld v La Foyer SA Case C-316/93 [1994] ECR I-763 (para 16).

159
16.
Moreover, in its many judgments in proceedings under art 169 of the Treaty concerning non-
implementation of directives the court, in declaring member states to be in breach of their obligations
under Community law, has consistently defined the breach as a failure to adopt the necessary
implementing measures within the prescribed period.
17.
In his opinion in Teuling v Bedrijfsvereniging voor de Chemische Industrie Case 30/85 [1987] ECR
2497 (para 7) Advocate General Mancini nevertheless suggested that the legislative freedom of member
states following the adoption of a directive was subject to certain limits:
such freedom does not include the power to aggravate the defect which the directive is
intended to remedy. Indeed it may be that measures adopted during the prescribed period must of
necessity be measures intended to transpose the Community provisions. Such measures must at
least not conflict with the requirements laid down in those provisions.
18.
Advocate General Mancinis remarks were made against the background of national rules, adopted
during the period for the implementation of Council Directive (EEC) 79/7 on the progressive
implementation of the principle of equal treatment for men and women in matters of social security (OJ
1979 L6 p 24), which were alleged to increase discrimination on grounds of sex in the award of benefits in
respect of incapacity for work. In its judgment the court did not however find it necessary to rule on the
issue.
19.
In the present proceedings Inter-Environnement and the Commission propose an affirmative answer to
both parts of the national courts first question.
20.
Inter-Environnement emphasises that it does not seek to challenge the principle that an individual can
rely upon the provisions of a directive before a national court only from the expiry of the implementation
period. In its action before the Conseil dtat it is not seeking to assert individual rights. Its action seeks
the annulment of the national regulation on the basis that it conflicts with a superior rule, namely the
Community directive which it is intended to implement. Under Belgian law such an action must be brought
within 60 days of the publication of the measure concerned. In the absence of an obligation on the
Walloon executive of the kind canvassed in the first question, an implementing measure adopted more
than 60 days before the expiry of the implementation period would be unassailable. Challenges based on
Community law would therefore be treated less favourably than challenges based on national law.
21.
Inter-Environnement considers that the legality of national implementing measures can be reviewed
even before the expiry of the implementation period. In the course of such review, account must be taken
of the duty of co-operation under art 5 of the Treaty, which imposes an obligation on member states to
implement the directive properly. The position is particularly clear in the present case since the contested
regulation expressly purports to be an implementing measure.
22.
The Commission argues that the adoption of a directive entails a standstill obligation based on arts 5
and 189 of the Treaty. A member state is not permitted to adopt any new measure which would increase
the disparity between the national and Community rules. Such a measure would be liable to jeopardise
the attainment of one or more of the objectives of the directive and lead to legal uncertainty for
individuals. It is therefore immaterial whether a measure is specifically intended to implement a directive.
23.
In its written observations the Commission stated that the failure by a state to fulfil its obligations under
the directive may be established only after the 160 implementation period. At the hearing it suggested that
an action could nevertheless be brought against a member state under art 169 of the Treaty to establish
an infringement of the standstill obligation itself. The Commission accepts however that such obligation
would not give rise to rights for individuals before the end of the implementation period.
24.
The Belgian, French, Netherlands and UK governments all propose a negative reply to the first
question. The German government did not submit observations on that question.
25.
The Netherlands government accepts that the adoption of a directive entails a form of standstill
obligation. However, it considers that a member state cannot be in breach of arts 5 and 189 where, as
here, it is unclear whether the provisions in question are in breach of the directive. Moreover, it considers
that no action may be brought under art 169 of the Treaty before the end of the implementation period.
26.
The Belgian, French and UK governments all take the view that pending the final date for
implementation of a directive member states remain free to enact rules which are contrary to it. The
United Kingdom adds the qualification that a member state is precluded by arts 5 and 189 of the Treaty
from adopting national measures which would have the effect of making it impossible or excessively
difficult for the state to give proper effect to the directive when it subsequently introduces measures to
transpose the directive into national law.
27.
It seems to me that, in considering the Conseil dtats first question, it is important to bear in mind the
reasons which have led it to put the question to the court. Although the question is phrased in terms of the
power of member states to adopt rules contrary to a directive during the period for its implementation,
Inter-Environnements action is not specifically concerned with that period. The question arises because
of the principle in Belgian administrative law that the validity of a measure must be assessed by reference
to the circumstances obtaining at the moment of its adoption. The Conseil dtats question is therefore
designed to discover whether at the moment of the adoption of the contested regulation Community law
precluded Belgium from enacting a measure which was contrary to the directive. The Conseil dtat
assumes that, unless at the moment of the adoption of the contested regulation the directive imposed
such an obligation on Belgium, the regulation cannot be challenged by way of an action for annulment
before the Conseil dtat.
28.
Inter-Environnement also points out in its written observations that an action for annulment must be
brought before the Conseil dtat within 60 days of the publication of the contested measure. Thus in
order to be admissible its action had to be brought before the final date for implementation of Directive
91/156. However, it is not clear whether, even if an applicant were able to bring an action after the final
date for implementation of a directive because the contested measure was adopted less than 60 days
before that date, its action could succeed unless an obligation existed under Community law at the
moment of the adoption of the measure.
29.
Against that background it seems to me that the real issue in the present case is not whether Directive
91/156 precludes the application of the contested regulation during the period for the implementation of
the directive but whether, despite the fact that that period had not yet expired at the moment when the
contested regulation was adopted, the Conseil dtat is obliged to take account of the directive in
assessing the validity of the regulation. In my view that question must plainly be given an affirmative reply
even on the basis of existing principles.
161
30.
The Treaty, in particular the third para of art 189, imposes on member states an obligation to achieve
the result required by a directive no later than the final date fixed for its implementation. However, that
obligation actually arises not from the final date for implementation but on the day on which the directive
enters into force or takes effect as laid down in art 191. Directive 91/156 took effect, by virtue of the
second paragraph of art 191 of the Treaty, on the date on which it was notified to Belgium. From that date
onwards and throughout the period for the implementation of the directive Belgium was under an
obligation to bring its national provisions into line with the directive by 1 April 1993. In other words, it was
under an obligation to ensure that there were no national rules contrary to the directive after that date.
That obligation flowed from a legal instrument which had already taken effect at the moment of the
adoption of the contested national regulation and must clearly be taken into account by the Conseil dtat
in assessing the validity of the regulation. That is so notwithstanding the fact that performance of the
obligation to implement the directive was not required until a later date; as I shall explain below, that may
however be relevant in determining the scope of the remedy to be granted by the Conseil dtat (see para
34, below).
31.
Such a conclusion is therefore consistent with the rule, applicable also under Community law, that the
validity of a measure must be assessed on the basis of the situation at the moment of its adoption (see eg
the judgment in France v EC Commission Joined cases 1516/76 [1979] ECR 321). The rationale for that
rule is that the legislature can reasonably be expected to take account only of legal and factual
circumstances obtaining at that moment. In the present case however the Community directive had
already been adopted and had taken effect at the moment of the adoption of the contested national
regulation. Its existence was therefore a legal circumstance of which the Walloon executive must be taken
to have been aware.
32.
In that connection reference may also be made to the recent judgment of the Court of First Instance in
Opel Austria GmbH v EU Council Case T-115/94 [1997] All ER (EC) 97 (paras 87ff). The Court of First
Instance, while acknowledging that the legality of a measure contested under art 173 of the Treaty had to
be assessed in the light of the elements of fact and law existing at the time when the measure was
adopted, held that, in reviewing the legality of a Council regulation, it was entitled to take account of the
Agreement on the European Economic Area between the European Communities, their Member States
and the Republic of Austria, the Republic of Finland, the Republic of Iceland, the Principality of
Liechtenstein, the Kingdom of Norway, the Kingdom of Sweden and the Swiss Confederation (the EEA
Agreement) (Oporto, 2 May 1992; TS 26 (1995); Cm 2847; OJ 1994 L1, p 3) which had been concluded
by the Communities seven days before the adoption of the regulation but entered into force shortly
afterwards. The Court of First Instance based that conclusion inter alia on the principle of good faith as
embodied in art 18 of the Vienna Convention on the Law of Treaties (the First Vienna Convention)
(Vienna, 23 May 1969; TS 58 (1980); Cmnd 7964).
33.
It seems to me to be even clearer that, by virtue of the duty of co-operation imposed on member states
by art 5 of the Treaty and their specific obligation to implement under art 189, account must be taken of a
directive which has already taken effect in assessing the legality of national measures.
34.
In such a case it is for a national court, if it concludes that a national measure is incompatible with the
obligations flowing from a directive, to determine the appropriate remedy in accordance with the relevant
national rules. It may be that in certain cases, for example where a national measure was adopted under
delegated 162 powers and the enactment of a measure contrary to the directive would be ultra vires, the
measure would fall to be annulled ab initio. More generally, in the case of a measure specifically intended
as an implementing measure that result may be thought to be consistent with the purpose of an action,
such as the action for annulment in the present case, which is designed to allow rapid review of new
regulations in the interests of legal certainty.
35.
However, that is a matter for national law. As a matter of Community law it would in my view be
sufficient in the present case for the Conseil dtat to annul the contested national regulation, if it were
found to contravene the directive, solely with effect from the final date for implementation of the directive,
namely 1 April 1993. It is from that date that the regulation would begin to produce unlawful effects. I
should emphasise that the foregoing is of course without prejudice to the rules applicable under national
law with respect to locus standi, time limits and other procedural conditions.
36.
The above solution may at first sight seem somewhat curious in so far as the moment by reference to
which the legality of a national measure must be appraised does not coincide with the moment from which
it produces unlawful effects. However, that anomaly is the consequence of the special features of
directives. If, owing to the particular method by which directives are transposed into national law, national
implementing measures adopted before the final date for implementation of a directive were immune from
challenge by way of an action for annulment before the Conseil dtat, the result would be that actions
based on Community law would be placed at a disadvantage by comparison with actions based on
national law. For example, an action based on a superior rule of national law, such as a Decree of the
Walloon Regional Council, would not encounter the same obstacle. Implementing measures would then
at most be susceptible to indirect challenge by way of a plea of illegality raised in proceedings brought
against individual decisions or other provisions adopted pursuant thereto.
37.
The above conclusion based on existing principles should be sufficient to resolve the anomaly which
prevents the Conseil dtat from reviewing the legality of implementing measures adopted before the final
date for implementation of directives. I do not therefore think it is necessary in the present case for the
court to rule on the views put forward by Inter-Environnement and the Commission, namely that the entry
into force of a directive has a blocking effect in the sense that it confines the power of member states to
the adoption of implementing measures in conformity with the directive, at least where such measures
expressly purport to implement the directive, or precludes the adoption of measures which increase the
disparity between national and Community rules (a standstill obligation). Those arguments raise more
difficult issues which would be better resolved in a case in which they genuinely arose. I shall
nevertheless consider them briefly.
38.
 It is clear that Directive 91/156 does not expressly impose limitations of the kind suggested on the
powers of member states. It would have been open to the Community legislature to do so, yet it did not.
Like other directives, it merely imposes an obligation to bring into force the laws, regulations and
administrative provisions necessary to comply with the directive not later than a certain date.
39.
The question arises whether such limitations may nevertheless be inferred from arts 5 and 189 of the
Treaty. I certainly accept that a member state would be in breach of its duty of co-operation under art 5 of
the Treaty if it were to adopt measures during the period for implementation of a directive which were
liable to jeopardise the attainment of the aims of a directive by the appointed date. It would also be in
breach of its more specific duty to implement under art 189 of the Treaty.
163
40.
As the United Kingdom suggests, that would be so if a member state adopted measures which
rendered the attainment of the objectives of the directive impossible or unduly difficult. For example,
suppose that the Community adopted a directive imposing limits on the quantity of nuclear waste
produced by member states. It seems to me that a member state would plainly be in breach of its
obligations under arts 5 and 189 of the Treaty if, after the adoption of the directive, it were to invest in a
nuclear programme which would of necessity make compliance with the directive impossible or unduly
difficult.
41.
In such a case, in which measures adopted by a member state were liable to create a situation which
would make implementation of a directive impossible or unduly difficult, there would be sound reasons for
recognising that the member state was in breach of its obligations under Community law even before the
implementation period had expired. The Commission would then be able to bring the matter before the
court as a matter of urgency, with a view to forestalling the application of the measures.
42.
I do not rule out the possibility that a member state might in some circumstances also be considered to
be in breach of its duty of faithful co-operation under art 5 of the Treaty (although possibly not its duty to
implement under art 189) if it were to enact without justification, after the adoption of a directive,
measures which were wholly contrary to the spirit and tenor of a directive, especially one which conferred
rights on individuals. That might particularly be so where the measures, although repealed before the final
date for implementation, continued to produce practical effects after that date. For example, as Advocate
General Mancini suggested in Teuling, a member state might be in breach of its duty of co-operation if,
following the adoption of a directive prohibiting discrimination on grounds of sex in certain areas, it
adopted measures introducing discrimination in those areas.
43.
However, such cases would be exceptional. Contrary to the views expressed by Inter-Environnement
and the Commission, I do not think it would be appropriate to interpret arts 5 and 189 as entailing a
general blocking effect in any of the senses mentioned in para 37, above.
44.
First, the somewhat extreme view that the power of a member state is limited, following the adoption of
a directive, solely to the adoption of implementing measures ignores the fact that, during the
implementation period, member states are permitted to continue applying their existing arrangements and
may well need to make adjustments to those arrangements during that period. It seems clear that
member states must in principle retain the power freely to amend individual aspects of such
arrangements, which may be based on methods and even objectives which differ materially from those
underlying the directive, in order to preserve their coherence and effectiveness pending their replacement
by the Community system.
45.
Even the imposition of a standstill obligation, precluding a member state from exacerbating any
disparities between the national and Community rules, would in my view be inappropriate other than in
special circumstances such as those described above. Such a restriction would, as the French
government suggests, unduly curtail the freedom of choice of member states in implementing a directive;
in some cases it may even make member states reluctant to agree to the adoption of a directive in the
first place.
46.
Suppose, for example, that during 1997 the Council adopts a directive introducing a maximum rate of
VAT with effect from 31 December 2000. A member state, in agreeing to the adoption of the directive,
calculates that for the years 1999 and 2000 it will require additional revenue in order to meet its public
164 expenditure which can be met by increasing the rate of VAT which it levies above the maximum rate
for those years. At the same time, however, it plans to use those years to draft and adopt new legislation
introducing a wealth tax from the beginning of the year 2001 in order to meet the shortfall in tax revenue,
allowing it to comply with the maximum rate of VAT from that year onwards. It seems to me that in that
example the member state, far from being in breach of its duty of co-operation, has been co-operative
both in the adoption and implementation of the directive.
47.
An implied standstill obligation would also be inappropriate in areas such as the environment, where a
member state by the adoption of measures seeks to ensure the attainment of certain physical parameters
or values required by a directive. A member state may be reluctant to guarantee that some worsening of
the situation will not occur, and need to be reflected in its national rules, before the measures it puts in
place for the attainment of the aims of the directive are effective.
48.
Moreover, numerous directives designed to ensure the free movement of goods, services or persons
provide for the replacement of regulation or supervision in the importing or host state by regulation or
supervision in the state of origin. Pending the introduction of uniform rules permitting the transfer of
competence to the state of origin it may be necessary, in the light of commercial or other developments,
for an importing or host state to introduce additional arrangements or procedures which, once the
directive is implemented, will become the exclusive responsibility of the state of origin. Yet the introduction
by the importing or host state of any new arrangements would by definition be contrary to the directive
and in breach of a standstill obligation.
49.
Contrary to the Commissions view, I do not think any support for the view that a standstill obligation
arises from arts 5 and 189 of the Treaty can be derived from the courts judgment in Peskeloglou v
Bundesanstalt fr Arbeit Case 77/82 [1983] ECR 1085 or EC Commission v UK Case 804/79 [1981] ECR
1045. The courts finding in Peskeloglou that, during the transitional period provided for in the second
sub-paragraph of art 45(1) of the Act of Accession (1979): EC 18 (1979); Cmnd 7650; OJ 1979; L291 p 17
(Greece), national provisions concerning the first grant of a work permit to a Greek national could not be
made more restrictive after the entry into force of the Act was based on a restrictive interpretation of a
provision of the Act itself, there is no equivalent provision in the directive in issue here (see, to the same
effect, Prechal Directives in European Community Law (1995) p 26).
50.
The Commission refers to the passage in EC Commission v UK [1981] ECR 1045 (para 28) where the
court stated that art 5 of the Treaty
imposes on Member States special duties of action and abstention in a situation in which the
Commission, in order to meet urgent needs of conservation, has submitted to the Council
proposals which, although they have not been adopted by the Council, represent the point of
departure for concerted Community action.
However, those remarks concerning the duties of member states were made in the context of art 102 of
the Act of Accession (1972); TS 17 (1979); Cmnd 7463 (UK, Denmark, Ireland), under which the exclusive
power to adopt measures relating to the conservation of the resources of the sea had been expressly
transferred to the Community as from 1 January 1979.
51.
Inter-Environnements narrower view that, where a member state adopts measures purporting to
implement a directive, its power is limited to the adoption of measures which are in conformity with the
directive is less objectionable at first 165 sight. As I have already noted, it is certainly possible that as a
matter of national law the power of the legislator is so limited.
52.
However, other than in special circumstances I see no compelling reason why as a matter of
Community law an implementing measure should be considered unlawful before the obligation to
implement has crystallised. In proceedings brought by individuals based on the direct effect of directives
national courts are obliged to set aside conflicting national rules only after expiry of the implementation
period. It is more consistent with that principle that in the present case the Conseil dtat should, as a
matter of Community law, be obliged to annul the regulation (if found to be contrary to the directive) only
with effect from the final date for implementation.
53.
Finally, I see no inconsistency between the above result and the view which I took in my opinion in
Hansa Fleisch Ernst Munde GmbH & Co KG v Landrat des Kreises Schleswig-Flensburg Case C-156/91
[1992] ECR I-5567 the effect that, even before the expiry of the period for implementation, a national court
was obliged under Community law to interpret national legislation intended to implement a directive in
accordance with the directive. As I noted in that opinion, the duty to interpret implementing provisions
consistently with the directive arises, not from the expiry of that period, but from the duty of the national
court under art 5 of the Treaty to cooperate with other national authorities in their endeavour to implement
the directive. It would plainly be absurd if a national court were permitted to frustrate the intention of the
national legislature by refusing to interpret implementing provisions in conformity with the directive where
they were capable of being so interpreted.
54.
The purpose of the national courts second question is to ascertain whether the fact that the collection,
pre-treatment, disposal or recovery of a substance forms part of an industrial process removes it from the
scope of Directive 75/442.
55.
In what follows I shall refer only to those provisions of the Community legislation which are directly
relevant to the present case. For a fuller account of that legislation I would refer to my opinions in
European Commission v Germany Case C-422/92 [1995] ECR I-1097 (esp paras 211) and Criminal
proceedings against Tombesi Joined cases C-304/94, C-330/94, C-342/94 and C-224/95 [1997] All ER
(EC) 639, [1997] ECR I-3561 (esp paras 218)).
56.
 Inter-Environnement, the Commission and the Belgian, German, Netherlands and UK governments all
consider that it is immaterial whether the operation in question forms part of an industrial process. I share
that view.
57.
Article 1(a) of Directive 75/442 defines waste as: any substance or object in the categories set out in
Annex I which the holder discards or intends or is required to discard.
58.
Annex I to Directive 75/442, to which art 1(a) of the directive refers and which is entitled Categories of
waste, lists different categories of waste, including
(Q1) Production or consumption residues not otherwise specified below
(Q2) Off-specification products
(Q7) Substances which no longer perform satisfactorily (eg contaminated acids, contaminated
solvents, exhausted tempering salts)
(Q8) Residues of industrial processes
(Q11) Residues from raw material extraction and processing (eg mining residues, oil field slops,
etc)
(Q12) Adulterated materials
166
The broad definition of waste is reinforced by the final category: (Q16) any materials, substances or
products which are not contained in the above categories.
59.
A detailed list of waste known as the European Waste Catalogue (the EWC) was adopted by the
Commission pursuant to art 1(a) of the directive by Commission Decision (EC) 94/3 (OJ 1994 L5 p 15).
Introductory Note 3 of the EWC states that:
the EWC is an harmonized, non-exhaustive list of wastes, that is to say, a list which will be
periodically reviewed and if necessary revised in accordance with the committee procedure.
However, the inclusion of a material in the EWC does not mean that the material is a waste in all
circumstances. The entry is only relevant when the definition of waste has been satisfied.
60.
Thus a substance, in particular those listed in Annex I or in the EWC, constitutes waste within the
meaning of Directive 75/442 where the holder discards or intends to discard or is required to discard it.
The scope of the term waste therefore turns on the meaning of the term discard. As I noted in my
opinion in Tombesi [1997] All ER (EC) 639, [1997] ECR I-3561 (para 50), it is clear from the provisions of
the directive, in particular art 4, arts 8 to 12 and Annexes IIA and B, that the term discard employed in the
definition of waste in art 1(a) has a special meaning encompassing not only the disposal of waste but also
its consignment to a recovery operation.
61.
There is nothing in the directive to suggest that its scope is limited to disposal or recovery operations
not forming an integral part of an industrial process. On the contrary, it is clear from the list of categories
of waste in Annex I, in particular the items mentioned at para 58 above, and from the disposal and
recovery operations listed in Annexes IIA and IIB that the notion of waste is sufficiently broad to cover all
kinds of industrial residues, by-products and other materials resulting from production processes.
62.
Moreover, it is clear from arts 9,10 and 11 of the directive, summarised at para 5 above, that the permit
requirement laid down by those provisions applies not only to undertakings specialising in waste disposal
and recovery for third parties but also to undertakings which carry out their own waste disposal and
recovery. The broad scope of the directive is confirmed by the power of member states to lay down
certain exemptions. Article 11 allows member states to exempt undertakings carrying out their own waste
disposal or recovery from the permit requirement on certain closely defined conditions. That power is
subject to further limits in the case of hazardous waste. Under art 3(1) of Directive 91/689 member states
are not permitted to exempt undertakings carrying out their own disposal operations in so far as the
substance is classified as hazardous waste within the meaning of that directive. Moreover, art 3(2) of the
directive lays down more specific conditions for the exemption of undertakings carrying out their own
recovery of hazardous waste going beyond those applicable to general waste under art 11 of Directive
75/442.
63.
It is clear therefore that a substance which is disposed of or recovered within the meaning of the
directive constitutes waste even where such disposal or recovery forms part of an industrial process. That
conclusion is sufficient to provide the national court with the guidance it seeks.
64.
The member states which have submitted observations on the second question have however gone
further and explained the criteria which they use for distinguishing between waste and non-waste
materials with particular reference to materials produced or used in industrial processes.
65.
 The Belgian government considers that a substance referred to in Annex I to the directive integrated
into an industrial production process constitutes waste only 167 if it is used in a manner which is not in
conformity with the nature or function assigned to it under a natural process or assigned to it intentionally
under a production or other process.
66.
The German government takes the view that it is necessary, in the absence of criteria laid down by the
directive, to consider each case on its merits in the light of the view generally held in the trade. The
German government produces by way of an annex to its written observations a recent OECD discussion
document providing guidance on the factors which may be relevant in distinguishing between waste and
non-waste materials in individual cases (see Discussion Paper on guidance for distinguishing waste from
non-waste issued by the OECD Waste Management Policy Group, ENV/EPOC/WMP(96)1). I shall
consider that document in some detail below.
67.
The Netherlands and UK governments put forward similar views to those which they expressed in
Tombesi [1997] All ER (EC) 639, [1997] ECR I-3561 (see paras 4748 of the opinion). The Netherlands
government distinguishes between waste and secondary raw materials. The environmental objectives
underlying the waste directives mean that very stringent conditions must be fulfilled before a substance
can be categorised as a secondary raw material. In the Netherlands the following criteria have been laid
down for that purpose: the substance must be transported directly from the producer to the person who
will make further use of it; it must be used 100% in a production process, for example as a substitute for a
primary raw material; and it must not be subject to any process comparable to a current means of waste
disposal or recovery.
68.
The United Kingdom considers that a substance constitutes waste if it is consigned to a disposal
operation within the meaning of Annex IIA or to an operation which falls within Annex IIB by virtue of the
fact that it is unequivocally associated with the recovery of waste. Production residues, secondary raw
materials and useful by-products used in industrial production processes in the same way as any other
raw materials of non-waste origin do not constitute waste provided that they are not subjected to an
operation unequivocally associated with the recovery of waste. Production residues which, by virtue of
adulteration or other like reasons associated with their character as secondary raw materials, require
processing in order to protect human health or the environment under conditions different from those
necessary in the case of other raw materials of non-waste origin, are to be regarded as destined for
recovery processes within the meaning of Annex IIB and constitute waste.
69.
It seems to me that it would be possibleand may be desirable in the interests of legal certaintyfor
the court, in the light of the observations presented to it in the present case, to offer some general
guidance on the distinction between the discarding of waste within the meaning of the directive and
normal industrial processing of non-waste products. As I suggested in my opinion in Tombesi [1997] All
ER (EC) 639, [1997] ECR I-3561 (para 56), however, as the directive stands at present it must to some
extent be left to member states to develop more detailed criteria to apply the rules of the directive, as
interpreted by the court, to the various situations which may occur in practice.
70.
Such an approach is moreover consistent with the division of responsibilities between the court and
the national courts under the Treaty. It is for the national courts to verify that national authorities have
properly applied the directive in individual cases.
71.
Although the governments represented in these proceedings have put forward different criteria for
applying the directive, it seems to me that there is 168 nevertheless considerable common ground.
Moreover, the OECD document produced by the German government suggests that the same is true
among OECD countries. Although prepared for a different purpose (namely the application of the notion of
waste in the context of an OECD decision), the document contains a useful comparative survey of the
relevant law and practice of OECD countries.
72.
The document notes that OECD countries commonly distinguish between primary raw materials,
residues and secondary raw materials. Primary raw materials are defined as materials obtained from
natural sources for use in manufacturing or production processes (eg mined or quarried minerals, crude
oil, harvested crops). Such materials are not regarded as waste because they have been intentionally
obtained (although I presume that even primary raw materials could become waste if the holder formed
an intention to dispose of them).
73.
Residual materials or residues are defined as materials which unavoidably arise during the
manufacture or use of a product. Such a material may be used directly as an effective substitute for a
product or as an ingredient in another manufacturing process to create a different product or may have no
direct use without being subjected to further processing. Some countries define the term by-product in
the same way. Such materials are generally regarded as waste unless they are capable of being used as
an effective substitute for another product or as an ingredient in another process other than a recovery
operation. In both cases any direct use should be environmentally sound, i.e. comply with the same
standards, regulations and specifications as the product or ingredient that the material is replacing.
74.
 According to the OECD document its member countries use the term secondary raw material in three
different ways: (a) a material which may no longer be used for its originally intended purpose but is in a
form which allows it to be used directly in a production process as a substitute for a primary raw material
(in which case it is unlikely to constitute waste); (b) a material which can be so used only after being
consigned to a recovery operation (and hence is likely to constitute waste); (c) a material which has
undergone a recovery operation and is now ready for use in a production process (and is therefore likely
to have ceased being waste).
75.
Although that appears to be the limit of the international consensus on the subject, the OECD
document notes that its member countries employ a range of other criteria for determining whether a
material constitutes waste in a given case. I think it is helpful to set out those criteria in full (para 17 of the
document):
(1) Is the material produced intentionally? (2) Is the production of the material subject to quality
control? (3) Does the material meet well developed nationally and internationally recognised
specifications/standards? (4) Do these standards include environmental considerations, in addition
to technical or economic considerations? (5) Is the material made in response to market demand?
(6) Is the overall economic value of the material negative? (7) Is further processing required before
the material can be directly used in a manufacturing/commercial application? (8) Is this processing
limited to minor repair? (9) Is the material still suitable for its originally intended purpose? (10). Can
the material be used for another purpose as a substitute material? (11) Is the use of the material as
environmentally sound as that of a primary product? (12) Will the material actually be used in a
production process? (13) Does the material have an identified use? (14) Does use of the material in
a production process cause any increased risks to human health or the environment greater than
the use of the corresponding raw material? (15) Is the material no longer part of the normal
commercial cycle or chain of utility? 169(16) Can the material be used in its present form or in the
same way as a raw material without being subjected to a recovery operation? (17) Can the material
be used only after it has been subjected to a recovery operation?
76.
The document then continues (para 18):
In presenting these questions, it should be noted that no particular weighting can be assigned
to individual questions nor any judgment made on their application given the variations in the way
that they are used in different countries. Some of these questions overlap nor is the list exhaustive.
In order to fully evaluate the status of the material, all of these questions may be considered when
making a determination.
77.
In my opinion in Tombesi [1997] All ER (EC) 639, [1997] ECR I-3561 (paras 5354) I suggested that
underlying the Community directive was an implicit distinction between non-waste materials which are put
to continued use in their existing form and waste materials which are subject to a recovery operation.
Thus a by-product or residual product would not constitute waste if it was destined for direct use in a
further process in its existing form, in other words if it was not destined for disposal or consignment to a
recovery operation prior to its continued use. I pointed out however that a particular problem arose in
distinguishing between direct use and recovery.
78.
It seems to me that the foregoing comparative survey provides further guidance on how that difficulty
may be resolved, as well as being consistent with the basic conclusions which I reached in Tombesi. The
OECD document suggests that there is general consensus that, where a secondary raw material or
residue can be used directly in a further process, possibly as a substitute for a primary raw material, it is
unlikely to constitute waste. It will on the other hand constitute waste if it must first undergo a recovery
process. Similar considerations appear to underlie a number of the detailed criteria applied by individual
countries (see esp points 7 to 10, 12 to 13 and 15 to 17 at para 75, above).
79.
As regards the sometimes difficult distinction between recovery of waste and direct use of non-waste
materials, as noted at para 73, above, it appears that there is general consensus among OECD countries
that it is relevant to consider whether the use of a residual product or by-product as a substitute for
another material or ingredient is as environmentally sound as that of the material or ingredient which it is
replacing; in other words whether it complies with the same standards, regulations and specifications as
those applicable to that product. Once again similar considerations appear to underlie a number of the
more detailed criteria (see esp points 3, 4, 11 and 14 at para 75, above).
80.
It seems to me that a similar approach would be appropriate in interpreting the term waste in the
Community legislation. The directive seeks to ensure that waste is recovered or disposed of without
endangering human health and without using processes or methods which could harm the environment
(see art 4). The notion of waste must therefore be interpreted sufficiently broadly to ensure that any
processing of a substance that is undertaken by reason of its nature as waste falls within the regulatory
system of the directive. Thus where, owing to the fact that it is a residue, by-product, secondary raw
material or other material resulting from an industrial process, a materialor the process which it
undergoesdoes not meet normal health or environmental requirements or standards, it must be
regarded as waste and subject to special regulation under the directive. In so far as a material is wholly
interchangeable with another product and requires no additional regulation 170 or supervision beyond
that applicable to the product it is replacing, it is unnecessary for it to be classified as waste.
81.
I conclude therefore that the mere fact that a disposal or recovery operation within the meaning of the
directive is carried out as part of an industrial process does not remove it from the scope of the directive.
For the purpose of distinguishing between waste recovery and processing of non-waste materials it is
relevant to consider whether a substance is destined to be put directly to continued use in its existing
form. In the case of residues, by-products, secondary raw materials or other materials resulting from
industrial processes, that condition is fulfilled where the material, or the process to which it is destined to
be put, meets normal health and environmental requirements applicable to non-waste products or
processes.
82.
As the Community legislation stands at present, it is for the member states to lay down further detailed
criteria for the application of the rules of the directive to individual cases. For example, there is the
problem that the mere possibility of a material being put to direct use in another process without
undergoing recovery does not of itself guarantee that it will be so used; a number of the additional criteria
mentioned in the OECD document may well be relevant in determining the intention of the holder to put
the material to such use (eg points 3, 5, 6 and 12 at para 75, above).
83.
It would therefore not be appropriate to attempt to go further at this stage in defining the notion of
waste for the purposes of Community legislation.
84.
Accordingly, the questions referred by the Belgian Conseil dtat should in my opinion be answered as
follows:
(1) In reviewing the legality of a national measure enacted after the adoption of a directive but
before the final date for its implementation, a national court must take account of the obligations
imposed by the directive on the member state concerned. It may therefore be called upon to annul
the national measure with effect from that final date.
(2) A substance which would otherwise constitute waste within the meaning of art 1(a) of
Council Directive (EEC) 75/442 on waste, as amended by Council Directive (EEC) 91/156, does
not cease to be waste merely because the disposal or recovery operation to which it is subject
forms part of an industrial production process. For the purpose of distinguishing between waste
recovery and processing of non-waste materials it is relevant to consider whether a substance is
destined to be put directly to continued use in its existing form. In the case of residues, by-products,
secondary raw materials or other materials resulting from industrial processes, that condition is
fulfilled where the material, or the process to which it is destined to be put, meets normal health
and environmental requirements applicable to non-waste products or processes.

18 December 1997.

The COURT OF JUSTICE

delivered the following judgment.
1.
By judgment of 29 March 1996, received at the Court of Justice of the European Communities on 23
April 1996, the Belgian Conseil dtat referred to the Court of Justice for a preliminary ruling under art 177
of the EC Treaty two questions on the interpretation of arts 5 and 189 of the Treaty and art 1(a) of Council
Directive (EEC) 75/442 on waste (OJ 1975 L194 p 39), as amended by Council Directive (EEC) 91/156
(OJ 1991 L8 p 32).
171
2.
Those questions were raised in proceedings brought by Inter-Environnement Wallonie, a non-profit-
making association, for annulment of the order of the Walloon Regional Executive of 9 April 1992 on toxic
or hazardous waste (the order).
3.
The object of Directive 75/442 is to approximate the laws of the member states on the disposal of
waste. It has been amended by Directive 91/156.
4.
Directive 75/442, as amended, defines waste in art 1(a) as follows:
For the purposes of this Directive: (a) waste shall mean any substance or object in the
categories set out in Annex I which the holder discards or intends or is required to discard. The
Commission, acting in accordance with the procedure laid down in Article 18, will draw up, not later
than 1 April 1993, a list of wastes belonging to the categories listed in Annex I. This list will be
periodically reviewed and, if necessary, revised by the same procedure.
5.
The list mentioned in that last provision was laid down in Commission Decision (EC) 94/3 establishing
a list of wastes pursuant to art 1(a) of Directive 75/442 (OJ 1994 L5 p 15). Paragraph 3 of the introductory
note to that list states, first, that the list is not exhaustive and, second, that the fact that a material appears
in it is only relevant when the definition of waste has been satisfied.
6.
 Articles 9(1) and 10 of Directive 75/442, as amended, provide that any establishment or undertaking
which carries out the operations specified in Annex IIA or Annex IIB must obtain a permit from the
competent authority. Annex IIA concerns disposal operations whilst Annex IIB lists the operations which
may lead to recovery.
7.
Article 11 of Directive 75/442, as amended, provides an exception to the requirement of a permit:
(1) Without prejudice to Council Directive 78/319/EEC of 20 March 1978 on toxic and
dangerous waste [OJ 1978 L84 p 43], as last amended by the Act of Accession of Spain and
Portugal, the following may be exempted from the permit requirement imposed in Article 9 or Article
10: (a) establishments or undertakings carrying out their own waste disposal at the place of
production; and (b) establishments or undertakings that carry out waste recovery. This exemption
may apply only: if the competent authorities have adopted general rules for each type of activity
laying down the types and quantities of waste and the conditions under which the activity in
question may be exempted from the permit requirements, and if the types or quantities of waste
and methods of disposal or recovery are such that the conditions imposed in Article 4 are complied
with. (2) The establishments or undertakings referred to in paragraph 1 shall be registered with the
competent authorities
8.
Article 4 of Directive 75/442, as amended, provides:
Member States shall take the necessary measures to ensure that waste is recovered or
disposed of without endangering human health and without using processes or methods which
could harm the environment, and in particular: without risk to water, air, soil and plants and animals,
without causing a nuisance through noise or odours, without adversely affecting the countryside or
places of special interest
9.
According to the first indent of art 2(1) of Directive 91/156, the member states were to bring into force
the laws, regulations and administrative provisions 172 necessary to comply with that directive not later
than 1 April 1993 and forthwith to inform the Commission thereof. The second indent provides:
When Member States adopt these measures, the measures shall contain a reference to this
Directive or shall be accompanied by such reference on the occasion of their official publication.
The methods of making such a reference shall be laid down by the Member States.
10.
Article 1(3) of Council Directive (EEC) 91/689 on hazardous waste refers (OJ 1991 L377 p 20), for the
definition of waste, to Directive 75/442. Article 1(4) defines hazardous waste.
11.
Article 3 of Directive 91/689 provides:
(1) The derogation referred to in Article 11(1)(a) of Directive 75/442/EEC from the permit
requirement for establishments or undertakings which carry out their own waste disposal shall not
apply to hazardous waste covered by this Directive. (2) In accordance with Article 11(1)(b) of
Directive 75/442/EEC, a Member State may waive Article 10 of that Directive for establishments or
undertakings which recover waste covered by this Directive: if the Member State adopts general
rules listing the type and quantity of waste and laying down specific conditions (limit values for the
content of hazardous substances in the waste, emission limit values, type of activity) and other
necessary requirements for carrying out different forms of recovery, and if the types or quantities of
waste and methods of recovery are such that the conditions laid down in Article 4 of Directive
75/442/EEC are complied with.
12.
Article 11 of Directive 91/689 repealed Council Directive (EEC) 78/319 on toxic and dangerous waste
(OJ 1978 L84 p 43) with effect from 12 December 1993. However, art 1 of Council Directive (EC) 94/31
amending Directive 91/689 (OJ 1994 L168 p 28) deferred the repeal of Directive 78/319 until 27 June
1995.
13.
The Decree of the Walloon Regional Council of 5 July 1985 on waste, as amended by the Decree of
25 July 1991 (the decree), defines waste at art 3(1) as follows: waste: all substances or objects in the
categories set out in Annex I which the holder discards or intends or is required to discard.
14.
Article 5(1) of the order provides:
Authorization is required for the setting-up and running of an installation intended specifically for
the collection, pre-treatment, disposal or recovery of toxic or dangerous waste which is not an
integral part of an industrial production process
15.
The preamble to the order makes particular reference to the decree, Directive 75/442, as amended,
and to Directives 78/319 and 91/689. Article 86 of the order states that it is to come into force on the day
of its publication in the Moniteur Belge. Publication took place on 23 June 1992.
16.
 By application lodged on 21 August 1992, Inter-Environnement Wallonie requested the Belgian
Conseil dtat to annul the order in its entirety or, in the alternative, certain of its provisions.
173
17.
In its order for reference, the Conseil dtat has already ruled on five of the six pleas raised by Inter-
Environnement Wallonie and has annulled various provisions in the order.
18.
In its remaining plea, Inter-Environnement Wallonie maintains that art 5(1) of the order infringes, in
particular, art 11 of Directive 75/442, as amended, and art 3 of Directive 91/689, inasmuch as it excludes
from the permit system the operations of setting up and running an installation intended specifically for the
collection, pre-treatment, disposal or recovery of toxic or dangerous waste, where that installation forms
an integral part of an industrial production process.
19.
In the first part of that plea, Inter-Environnement Wallonie claims that art 11 of Directive 75/442, as
amended, in conjunction with art 3 of Directive 91/689, allows exemptions from the permit requirement for
undertakings carrying out waste recovery only on the conditions laid down by those provisions and only
where those undertakings are registered with the competent authorities.
20.
On that point, the Conseil dtat considers that art 5(1) of the order is indeed contrary to art 11 of
Directive 75/442, as amended, in conjunction with art 3 of Directive 91/689.
21.
Finding that the order was adopted at a time when the period allowed by the directive for its
transposition had not yet expired, the Conseil dtat questions to what extent a member state may, during
that period, adopt a measure contrary to the directive. It adds that a negative reply to that question, as
proposed by Inter-Environnement Wallonie, would be incompatible with the rule that the validity of a
measure is to be assessed at the time of its adoption.
22.
In the second part of its plea, Inter-Environnement Wallonie claims that the exception in art 5(1) of the
order is contrary to the decree which, it states, does not provide for any derogation for operations forming
part of an industrial process.
23.
On that point, the Conseil dtat finds that art 3(1) of the decree and the annex to which it refers are
intended to be a faithful transposition of Directive 75/442, as amended. While the case law of the court
makes it clear that waste means any substances and objects which the holder discards or is required to
discard without intending thereby to exclude their economic reutilisation by other persons, it does not
make it possible to establish whether a substance or object referred to in art 1 of Directive 75/442, as
amended, which directly or indirectly forms an integral part of an industrial production process is waste
within the meaning of art 1(a) of that directive.
24.
In those circumstances, the Conseil dtat has referred the following questions to the Court of Justice
for a preliminary ruling:
(1) Do Articles 5 and 189 of the EEC Treaty preclude Member States from adopting a provision
contrary to Directive 75/442/EEC of 15 July 1975 on waste, as amended by Directive 91/156/EEC
of 18 March 1991, before the period for transposing the latter has expired? Do those same Treaty
articles preclude Member States from adopting and bringing into force legislation which purports to
transpose the abovementioned directive but whose provisions appear to be contrary to the
requirements of that directive?
(2) Is a substance referred to in Annex I to Council Directive 91/156/EEC of 18 March 1991
amending Directive 75/442/EEC on waste and which directly or indirectly forms an integral part of
an industrial production process to be considered waste within the meaning of Article 1(a) of that
directive?
174
25.
By its second question, which it is appropriate to consider first, the national court is in essence asking
whether a substance is excluded from the definition of waste in art 1(a) of Directive 75/442, as amended,
merely because it directly or indirectly forms an integral part of an industrial production process.
26.
First of all, it follows from the wording of art 1(a), that the scope of the term waste turns on the
meaning of the term discard.
27.
It is also clear from the provisions of Directive 75/442, as amended, in particular from art 4, arts 8 to 12
and Annexes IIA and IIB, that the term discard covers both disposal and recovery of a substance or
object.
28.
As the Advocate General has pointed out in paras 58 to 61 of his opinion, above, the list of categories
of waste in Annex I to Directive 75/442, as amended, and the disposal and recovery operations listed in
Annexes IIA and IIB to that directive demonstrate that the concept of waste does not in principle exclude
any kind of residue, industrial by-product or other substance arising from production processes. This
finding is further supported by the list of waste drawn up by the Commission in Decision 94/3.
29.
First, Directive 75/442, as amended, applies, as is apparent in particular from arts 9 to 11, not only to
disposal and recovery of waste by specialist undertakings, but also to disposal and recovery of waste by
the undertaking which produced them, at the place of production.
30.
Second, while art 4 provides that waste is to be recovered or disposed of without endangering human
health or using processes or methods which could harm the environment, there is nothing in that directive
to indicate that it does not apply to disposal or recovery operations forming part of an industrial process
where they do not appear to constitute a danger to human health or the environment.
31.
Finally, it should be borne in mind that the court has already held that the definition of waste in art 1 is
not to be understood as excluding substances and objects which were capable of economic reutilisation
(se the judgments in Criminal proceedings against Zanetti Case C-359/88 [1990] ECR I-1509 (paras 12
13), European Commission v Germany Case C-422/92 [1995] ECR I-1097 (paras 2223) and Criminal
proceedings against Tombesi Joined cases C-304/94, C-330/94, C-342/94 and C-224/95 [1997] All ER
(EC) 639, [1997] ECR I-3561 (paras 4748)).
32.
It follows from all those considerations that substances forming part of an industrial process may
constitute waste within the meaning of art 1(a) of Directive 75/442, as amended.
33.
That conclusion does not undermine the distinction which must be drawn, as the Belgian, German,
Netherlands and UK governments have correctly submitted, between waste recovery within the meaning
of Directive 75/442, as amended, and normal industrial treatment of products which are not waste, no
matter how difficult that distinction may be.
34.
The answer to the second question must therefore be that a substance is not excluded from the
definition of waste in art 1(a) of Council Directive 75/442, as amended, by the mere fact that it directly or
indirectly forms an integral part of an industrial production process.
35.
By its first question, the national court is in substance asking whether arts 5 and 189 of the Treaty
preclude the member states from adopting measures contrary to Directive 91/156 during the period
prescribed for its transposition.
175
36.
According to Inter-Environnement Wallonie, it follows from the primacy of Community law and from art
5 of the Treaty that, even where a member state decides to transpose a Community directive before the
end of the period prescribed therein, such transposition must be consistent with the directive.
Consequently, since it chose to transpose Directive 91/156 on 9 April 1992, the Rgion Wallonie should
have complied with that directive.
37.
The Commission endorses that position and maintains that arts 5 and 189 of the Treaty preclude
member states from adopting a provision contrary to Directive 91/156 during the period prescribed for its
transposition. It states that in this respect it is irrelevant whether or not a particular measure is specifically
intended to transpose the directive.
38.
On the other hand, the Belgian, French and UK governments consider that until the period prescribed
for transposition of a directive has expired, the member states remain free to adopt national rules which
are at variance with it. The UK government adds, however, that it would be contrary to arts 5 and 189 of
the Treaty for a member state to adopt measures which would have the effect of making it impossible or
excessively difficult for that state to transpose the directive correctly into national law.
39.
The Netherlands government is of the opinion that the adoption of a directive means that the member
states are no longer free to undertake anything which might make it more difficult to achieve the result
prescribed. None the less, it considers that a member state cannot be regarded as being in breach of arts
5 and 189 of the Treaty where, as in the present case, it is not certain that the national provisions are
inconsistent with the directive concerned.
40.
It should be recalled at the outset that the obligation of a member state to take all the measures
necessary to achieve the result prescribed by a directive is a binding obligation imposed by the third
paragraph of art 189 of the Treaty and by the directive itself (see the judgments in Verbond van
Nederlandse Ondernemingen v Inspecteur der Invoerrechten en Accijnzen Case 51/76 [1977] ECR 113
(para 22), Marshall v Southampton and South West Hampshire Health Authority (Teaching) Case 152/84
[1986] 2 All ER 584, [1986] QB 401, [1986] 2 WLR 780, [1986] ECR 723 (para 48) and Aannemersbedrijf
P K Kraaijeveld BV v Gedeputeerde Staten van Zuid-Holland Case C-72/95 [1997] All ER (EC) 134,
[1996] ECR I-5403 (para 55)). That duty to take all appropriate measures, whether general or particular, is
binding on all the authorities of member states including, for matters within their jurisdiction, the courts
(see the judgments in Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89
[1990] ECR I-4135 (para 8) and Kraaijeveld [1997] All ER (EC) 134, [1996] ECR I-5403 (para 55)).
41.
The next point to note is that, in accordance with the second para of art 191 of the Treaty, applicable at
the material time, Directives and decisions shall be notified to those to whom they are addressed and
shall take effect upon such notification. It follows from that provision that a directive has legal effect with
respect to the member state to which it is addressed from the moment of its notification.
42.
Here, and in accordance with current practice, Directive 91/156 itself laid down a period by the end of
which the laws, regulations and administrative provisions necessary for compliance are to have been
brought into force.
43.
Since the purpose of such a period is, in particular, to give member states the necessary time to adopt
transposition measures, they cannot be faulted for not having transposed the directive into their internal
legal order before expiry of that period.
176
44.
Nevertheless, it is during the transposition period that the member states must take the measures
necessary to ensure that the result prescribed by the directive is achieved at the end of that period.
45.
Although the member states are not obliged to adopt those measures before the end of the period
prescribed for transposition, it follows from the second paragraph of art 5 in conjunction with the third
paragraph of art 189 of the Treaty and from the directive itself that during that period they must refrain
from taking any measures liable seriously to compromise the result prescribed.
46.
It is for the national court to assess whether that is the case as regards the national provisions whose
legality it is called upon to consider.
47.
In making that assessment, the national court must consider, in particular, whether the provisions in
issue purport to constitute full transposition of the directive, as well as the effects in practice of applying
those incompatible provisions and of their duration in time.
48.
For example, if the provisions in issue are intended to constitute full and definitive transposition of the
directive, their incompatibility with the directive might give rise to the presumption that the result
prescribed by the directive will not be achieved within the period prescribed if it is impossible to amend
them in time.
49.
Conversely, the national court could take into account the right of a member state to adopt transitional
measures or to implement the directive in stages. In such cases, the incompatibility of the transitional
national measures with the directive, or the non-transposition of certain of its provisions, would not
necessarily compromise the result prescribed.
50.
The answer to the first question must therefore be that the second paragraph of art 5 and the third
paragraph of art 189 of the Treaty, and Directive 91/156, require the member states to which that directive
is addressed to refrain, during the period laid down therein for its implementation, from adopting
measures liable seriously to compromise the result prescribed.
51.
The costs incurred by the Belgian, German, French, Netherlands and UK governments and by the
European Commission, which have submitted observations to the Court of Justice, are not recoverable.
Since these proceedings are, for the parties to the main proceedings, a step in the proceedings pending
before the national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Belgian Conseil
dtat by judgment of 29 March 1996, hereby rules:
(1) A substance is not excluded from the definition of waste in art 1(a) of Council Directive (EEC)
75/442 on waste, as amended by Council Directive (EEC) 91/156, merely because it directly or indirectly
forms an integral part of an industrial production process. (2) The second paragraph of art 5 and the third
paragraph of art 189 of the EC Treaty, and Directive 91/156, require the member states to which that
directive is addressed to refrain, during the period laid down therein for its implementation, from adopting
measures liable seriously to compromise the result prescribed.

177

[1998] All ER (EC) 178

Finanzamt Osnabrck-Land v Langhorst

(Case C-141/96)
TAXATION; VAT and Customs and Excise: EUROPEAN COMMUNITY; Taxation
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), MANCINI, MOITINHO DE ALMEIDA, MURRAY,
SEVN (PRESIDENTS OF CHAMBER), KAPTEYN, GULMANN, EDWARD, PUISSOCHET, HIRSCH,
JANN, RAGNEMALM AND WATHELET
ADVOCATE GENERAL LGER,
15 APRIL, 27 MAY, 17 SEPTEMBER 1997
European Community Value added tax Invoice Document serving as invoice Credit note issued by
recipient of supply Credit note drawn up with agreement of supplier Supplier able to contest amount
of value added tax mentioned Whether credit note capable of being regarded as document serving as
invoice Council Directive (EEC) 77/388, art 22(3)(c).
European Community Value added tax Liability for tax Person who mentioned value added tax on
invoice Credit note issued by recipient of supply serving as invoice Whether supplier to be regarded
as person who mentioned tax on invoice Whether supplier liable to pay value added tax shown if
incorrect credit note not contested Council Directive (EEC) 77/388, art 21(1)(c).

The respondent farmer elected under a German law to be charged to value added tax at the rate of 7%
rather than 13%, but did not notify that fact to livestock dealers to whom he had supplied fat pigs. The
dealers therefore issued him with credit notes, which he did not contest, on which value added tax was
calculated at 13%. Under German law a credit note by which a trader settled up for a taxable supply was
deemed to be an invoice and as a result, the tax office made a demand based on the 13% rate in relation
to those supplies. Subsequently, the respondent successfully brought proceedings in the Finanzgericht for
a reduction in the amount of tax demanded. The tax office appealed to the Bundesfinanzhof, which stayed
the proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling,
inter alia, the questions: (i) whether art 22(3)(c) 1 of Council (EEC) Directive 77/388 on the harmonisation
of the laws of the member states relating to turnover taxescommon system of value added tax: uniform
basis of assessment, authorised a member state to regard a credit note issued by the recipient of goods
or services as a document serving as an invoice; and (ii) whether a taxable person who had not
contested an incorrect amount of value added tax mentioned in a credit note serving as an invoice was to
be regarded as the person who had so mentioned the value added tax, and therefore as the person liable
for the amount stated pursuant to art 21(1)(c) 2.
1
Article 22(3)(c), so far as material, is set out at p 188 h, post
2
Article 21(1)(c), so far as material, is set out at p 188 g, post

Held (1) Since the purpose of art 22 of the directive was to ensure correct collection of the tax and the
avoidance of fraud, there was no reason why an invoice or document serving as such should not be
drawn up by the recipient of the goods or services, provided that it included the information prescribed for
an invoice and the taxable person had been given the opportunity, if necessary, to correct that 178
information. In such a case, the taxable person could be regarded as the author of the document, the
drawing up of which he had effectively delegated to his customer. It followed that a credit note fulfilled the
function of documenting the taxable persons rights and obligations with respect to value added tax, as it
contained the same information as a traditional invoice and the taxable person was free to approve its
content. Accordingly, art 22(3)(c) authorised member states to regard a credit note issued by the recipient
of goods or services as a document serving as an invoice where it included the information prescribed
for invoices, it was drawn up with the agreement of the taxable person, and the latter was able to contest
the amount of value added tax mentioned (see p 191 e f j and p 192 h, post).
(2) Where a credit note served as an invoice, the taxable person was to be regarded for the purposes
of art 21(1)(c) as the person who had in fact mentioned value added tax in the credit note, and he was
consequently liable to pay the amount stated. Otherwise, part of the value added tax appearing in the
document serving as invoice would not have to be paid by the taxable person, even though that value
added tax might have been deducted in full by the recipient of the goods and services, thus giving scope
for possible fraud or collusion. It followed that a taxable person who had not contested the mention, in a
credit note serving as an invoice, of an amount of value added tax greater than that owed by reason of
taxable transactions could be regarded as the person who had mentioned that amount, and was therefore
liable to pay the amount shown pursuant to art 21(1)(c) (see p 192 c to e j, post).

Notes
For persons liable for payment of tax and the obligations of those persons, see 52 Halsburys Laws
(4th edn) paras 20342035.

Cases cited
Genius Holding BV v Staatssecretaris van Financin Case C-342/87 [1991] STC 239, [1989] ECR 4227,
ECJ.
Jeunehomme v Belgium Joined cases 123/87 and 330/87 [1988] ECR 4517.
Reisdorf v Finanzamt Kln-West Case C-85/95 [1997] STC 180, [1996] ECR I-6257, ECJ.
Reference
By order of 14 March 1996, the Bundesfinanzhof (the Federal Finance Court) referred to the Court of
Justice of the European Communities for a preliminary ruling three questions (set out at p 190 e to g,
post) on the interpretation of arts 21(1)(c) and 22(3)(c) of Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment. Those questions were raised in proceedings between the
Finanzamt (the tax office) Osnabrck-Land and Mr Langhorst concerning the question whether Mr
Langhorst was liable to pay the amount of value added tax mentioned on a credit note issued by a
customer, which he had not contested, even though the amount was higher than that owed by reason of
the taxable transactions. Written observations were submitted on behalf of: the German government, by E
Rder, Ministerialrat in the Federal Ministry of Economic Affairs, acting as agent; the UK government, by
S Ridley, of the Treasury Solicitors Department, acting as agent, and S Lee, Barrister; and the European
Commission, by J Sack, Legal Adviser, acting as agent. Oral observations were submitted on behalf of:
the German government, represented by E Rder; the Greek government, represented by V Kontolaimos,
Deputy Legal Adviser in the 179 State Legal Service, and A Rokofyllou, Special Adviser to the Deputy
Minister of Foreign Affairs, acting as agents; the UK government, represented by S Ridley and S
Richards, Barrister; and the Commission, represented by J Sack. The language of the case was German.
The facts are set out in the opinion of the Advocate General.

27 May 1997.

The Advocate General (P Lger)

delivered the following opinion (translated from the French).
1.
The Bundesfinanzhof (the Federal Finance Court) brings before the Court of Justice of the European
Communities for the first time the practice of self-billing, in which the buyer of goods or the recipient of
services himself draws up the invoice for the economic transaction for which he is to pay.
2.
In view of the legal consequences attached to the concept of an invoice under the Community
legislation relating to value added tax (VAT), the recognition of that procedure by the laws of certain
member states logically led to national courts having to consider the legal effects of the documents drawn
up and issued under the conditions defined by the various member states 3.
3
The UK government, for example, stated at the hearing that the practice of self-billing existed in that country before the
introduction of VAT in 1973, when UK legislation expressly acknowledged the principle of self-billing.

3.
It is thus appropriate to determine whether the characteristics of a document drawn up by the debtor
rather than the creditor are close enough to those of a traditional invoice for it to be acknowledged as
having an identical function in the common system of VAT, even though certain provisions of Council
Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxes
common system of value added tax: uniform basis of assessment (the Sixth Directive) appear to
preclude such an equation.

I LEGAL AND FACTUAL BACKGROUND

A. Facts and national proceedings

4.
Mr Bernhard Langhorst, the plaintiff in the main proceedings, declared the turnover for 1985 of his
agricultural business after having elected, as permitted by para 24(4) of the Umsatzsteuergesetz in its
1980 version (the UStG)4, to be taxed at the rate of 7%, under the general provisions of the UStG, rather
than at the rate of 13% provided for in the first sentence of para 24(1) of that law.
4
The Law on Turnover Tax of 26 November 1979 (BGBl 1979 I, p 1953), as amended by the Law of 18 August 1980
(BGBl 1980 I, p 1537, 1543).

5.
Not being aware of that election, livestock dealers to whom Mr Langhorst had supplied fat pigs issued
him with credit notes mentioning separately VAT calculated at the rate of 13%. Mr Langhorst did not
initially contest the amount of VAT mentioned in the credit notes.
6.
He then brought proceedings in the Finanzgericht (the Finance Court), which gave a judgment
reducing the amount of tax. The Finanzamt (the tax office) appealed on a point of law to the
Bundesfinanzhof.

B. National legislation
7.
 The Bundesfinanzhof considers that the reduction was correctly determined by the Finanzgericht, but
that the appeal by the Finanzamt might nevertheless succeed under the first sentence of para 14(2) of the
UStG. Paragraph 14(2) provides as follows:
180
If the trader has in an invoice for a supply or other service shown separately a higher amount of
tax than he owes under this Law in respect of the transaction, then he shall also owe the additional
amount. If he corrects the amount of tax as against the recipient, then Paragraph 17(1) shall apply
correspondingly.
8.
The UStG equates credit notes with invoices, under certain conditions. Thus para 14(5) provides:
A credit note by which a trader settles up for a taxable supply or other service made to him shall
also be deemed to be an invoice. A credit note shall be recognised if the following conditions are
met: (1) The trader providing the service (the recipient of the credit note) must be entitled under
subparagraph 1 to show the tax separately in an invoice. (2) There must be agreement between
the issuer and the recipient of the credit note that the supply or other service is to be settled by a
credit note. (3) The credit note must include the information prescribed in the second sentence of
subparagraph 1 above5. (4) The credit note must have been delivered to the trader providing the
service. Sentences 1 and 2 above shall apply by analogy to credit notes which the trader issues as
payment or partial payment for a taxable supply or other service which has not yet been carried
out. The credit note shall cease to have effect as an invoice in so far as the recipient contests the
amount of tax shown therein.

5
The second sentence of para 14(1) prescribes: Such invoices must include the following information: (5) the
consideration for the supply or other service and (6) the amount of tax due on the consideration

9.
The Bundesfinanzhof considers it necessary to interpret para 14(2) of the German UStG in
accordance with Community law.

C. The national courts questions and the relevant provisions of Community law
10.
It consequently refers the following three questions to the Court of Justice:
(1) Is it permissible under Article 22(3)(c) of the Sixth Council Directive for a credit note
within the meaning of Paragraph 14(5) of the Umsatzsteuergesetz 1980 to be regarded as an
invoice or other document serving as an invoice (Article 21(1)(c) of the Sixth Directive)?
(2) If so, is it permissible under Article 21(1)(c) of the Sixth Directive for a person who accepts a
credit note showing a higher amount of tax than that owed by reason of taxable transactions, and
does not contest in that respect the amount of tax mentioned in the credit note, to be regarded as a
person who mentions value added tax in an invoice or other document serving as an invoice and is
therefore liable to pay that value added tax?
(3) Can the recipient of a credit note, in the circumstances set out in Question 2, rely on Article
21(1)(c) of the Sixth Directive if the value added tax mentioned in the credit note is claimed from
him as a tax debt to the extent of the difference between the tax mentioned and the tax owed by
reason of taxable transactions?
11.
The national court states that the third question arises only if the answer to the second question is
negative.
181
12.
Article 21(1) of the Sixth Directive lists the persons liable to pay VAT under the internal system. Sub-
paragraphs (a) and (c) specify that
(a) taxable persons who carry out taxable transactions other than those referred to in Article
9(2)(e) and carried out by a taxable person resident abroad (c) any person who mentions the
value added tax on an invoice or other document serving as invoice
are liable to pay VAT.
13.
Article 22 of the Sixth Directive principally determines the obligations of persons liable to pay VAT
under the internal system. Article 22(3) deals with invoices and their content. Sub-paras (a) and (c) state:
(a) Every taxable person shall issue an invoice, or other document serving as invoice in respect
of all goods and services supplied by him to another taxable person (c) The Member States shall
determine the criteria for considering whether a document serves as an invoice.

II ANSWERS TO THE NATIONAL COURTS QUESTIONS

14.
By its questions the Bundesfinanzhof essentially seeks to know: whether a credit note drawn up by the
beneficiary of a supply of goods or provision of services (the recipient of the goods or services) may be
equated with an invoice within the meaning of the Sixth Directive; whether the trader who has carried out
the economic transaction (the trader or the recipient of the credit note) and has not challenged the
excessive amount of VAT shown in the credit note may for that reason be regarded as the person who
has mentioned it and is consequently liable to pay it; and in the event that, under the relevant provisions
of the Sixth Directive cited above, the trader is not liable to pay the excess amount of VAT because he
may not be regarded as the person who has mentioned it, whether he may rely on that provision to
contest the obligation to pay under German law.

A. The first question: whether a credit note issued by the recipient of a supply of goods or
services may be equated with an invoice within the meaning of the Sixth Directive
15.
A precise delimitation of the concept of an invoice within the meaning of the Sixth Directive is of use in
view of the significant part played by that document in the Community legislation on VAT.
16.
The invoice constitutes the documentary evidence of the amount of VAT owed by the trader, thus
serving both the payment of that tax and the deduction of the tax paid by the previous trader (see art
18(1)(a) of the Sixth Directive).
17.
The Sixth Directive contains no definition of the terms invoice and document serving as invoice used
in arts 21(1)(c) and 22(3)(a). After listing the minimum conditions relating to the information an invoice
must by virtue of its very purpose contain6, the directive leaves it to the member states to determine the
criteria for considering whether a document serves as an invoice (see art 22(3)(c)).
6
Article 22(3)(b) states: The invoice shall state clearly the price exclusive of tax and the corresponding tax at each rate
as well as any exemptions.

18.
The Federal Republic of Germany made use of that power by enacting para 14(5) of the UStG, which
introduces the credit note at issue in these proceedings.
182
19.
The court has ruled on two occasions on the extent of the power thus conferred on the member states
(see the judgments in Jeunehomme v Belgium Joined cases 123/87 and 330/87 [1988] ECR 4517, and
Reisdorf v Finanzamt Kln-West Case C-85/95 [1997] STC 180, [1996] ECR I-6257). In both cases it
adopted a position favourable to the member states and allowed them a considerable discretion.
20.
In Jeunehomme, on the basis also of art 22(8) of the Sixth Directive, which allows member states to
impose other obligations which they deem necessary for the correct levying and collection of the tax and
for the prevention of fraud, it was held that a member state could require invoices to include additional
particulars (see Jeunehomme [1988] ECR 4517 (para 16))7.
7
The judgment in Jeunehomme [1988] ECR 4517 (para 17) also states, however, that such particulars must not, by
reason of their number or technical nature, render the exercise of the right to deduction practically impossible or
excessively difficult.

21.
In Reisdorf [1997] STC 180, [1996] ECR I-6257 (para 31), the court held that member states were
permitted to regard as an invoice not only the original but also any other document serving as an invoice
that fulfils the criteria determined by the member states themselves.
22.
The court was careful to state, however, that that power must be exercised consistently with one of
the aims of the Sixth Directive, that of ensuring that VAT is levied and collected, under the supervision of
the tax authorities (see [1997] STC 180, [1996] ECR I-6257 (para 24)) 8.
8
See also the judgment in Jeunehomme [1988] ECR 4517 (para 17).

23.
The aim thus described reflects the concern which, in my opinion, the court should continue to be
guided by in determining requirements relating to invoicing, although, unlike in the cases cited, the
relevant criterion for classification of the credit note is linked not only to its content but also to the person
who issues it.
24.
The national court refers in its first question, inter alia, to art 21(1)(c) of the Sixth Directive. That
provision forms the logical reference point for the present case, in view of the fact that the proceedings
pending in the German court concern an amount of VAT which differs from that owed by reason of the
taxable transaction alone. The mistake affecting the amount shown in the credit note would thus no longer
make the trader liable in his capacity as a taxable person within the meaning of art 21(1)(a), but in his
capacity as a person who mentions VAT on an invoice or document serving as an invoice.
25.
 That provision, however, cited by the national court in order to define the legal context of the case, is
of no use in answering the first question, since it gives no indication of whether or not the decisive factor
in deciding whether a credit note is a document serving as an invoice is the identity of the person who
issues or delivers it.
26.
The present question seeks to establish whether the power given to member states by art 22(3)(c) of
the Sixth Directive permits a document drawn up not by the trader but by the recipient of the goods or
services to be regarded as an invoice.
27.
A literal interpretation of art 22(3)(a), which describes the taxable person as the person who is to issue
the invoice, argues against the conclusion that the recipient of the goods or services, who cannot claim to
be the taxable person with respect to the VAT at issue, may fulfil the obligation of issuing an invoice and
that the trader may be allowed to escape that obligation. The credit note would then not be capable of
taking the place of the invoice.
183
28.
I do not support such a reading, however, which I consider too formal. The aim pursued by the
legislature when drawing up the relevant provision must be considered when interpreting it 9.
9
If the Community legislature in 1977, like the German legislature in 1980, took no account of the practice of self-billing in
the wording of the provision, that was no doubt because that practice was as yet little developed.

29.
Since its aim, as stated at para 22, above, is to ensure that the tax is correctly levied and to avoid
fraud, there appears to be no valid reason why the document drawn up by the recipient of the goods or
services should not serve as an invoice, where that document contains the information prescribed for
invoices by the Sixth Directive and its addressee is able to correct it if necessary (see para 17, above).
30.
In those circumstances, by exercising his power to check and correct, the trader remains the person
who issues the credit note, the drawing-up of which he has merely delegated, as it were, to his customer.
The credit note does not lose its function of documenting the traders fiscal rights and obligations, since it
contains the same information as a traditional invoice and the trader is free to approve its contents. He
thus keeps responsibility for drawing up the invoices, whatever their form, and ultimately remains their
true author.
31.
I consider, as moreover do all the intervening governments and the European Commission, that self-
billing as regulated by para 14(5) of the UStG fulfils the conditions under which uncontested credit notes
may be equated with invoices issued by the trader.
32.
Under para 14(5)(3), the credit note must include the same information as that prescribed for invoices,
including the consideration for the supply or other service and the amount of tax due on the
consideration.
33.
A power of control for the trader is ensured by para 14(5)(2), which states that the parties to the
contract must agree that a credit note is to be used. Again, under para 14(5)(4), the credit note must have
been delivered to the trader providing the service, and the second sentence of the second indent
provides that the credit note shall cease to have effect as an invoice in so far as the recipient contests the
amount of tax shown therein.
34.
The trader admittedly does not have an express power of rectification, but his right to deprive a credit
note of its status as an invoice by contesting its content nevertheless gives him complete control of the
classification in law of that document, which suffices to accept the equation of that document with an
invoice within the meaning of the Sixth Directive.

B. The second question

35.
Two readings of the question are possible.
36.
One reading of the question itself, in the light of the grounds stated in the order for reference, is that
the Bundesfinanzhof is asking the court to state whether the fact that a trader has accepted, without
contestation, an incorrect credit note issued by the recipient of the goods or services means that he may
be regarded as a person who has mentioned VAT within the meaning of art 21(1)(c) of the Sixth Directive.
If the answer is affirmative, the national court concludes automatically that the trader is liable for the VAT
mentioned in the credit note, without raising questions on that point, which it regards as not in doubt.
37.
The other reading is that the national courts question refers also to the precise amount of VATthat
stated in the credit note or that corresponding to the taxable transactionwhich the trader ultimately has
to pay.
184
38.
In my opinion, the first reading is correct. The grounds of the order for reference show that the
Bundesfinanzhof did not make the reference in order to establish whether the first sentence of para 14(2)
of the UStG may, having regard to Community law, require payment of the amount of VAT shown in the
invoice rather than the amount corresponding to the taxable transaction. The question concerns the effect
of art 21(1)(c) of the Sixth Directive on the applicability to credit notes of the first sentence of para 14(2).
Common to those two provisions is that they both designate the person who has mentioned the VAT as
liable. That point, which distinguishes a credit note from an invoice, is at the heart of the order for
reference.
39.
Thus according to the national court
it would be doubtful who was to be regarded as responsible for the excessive separate mention
of tax the livestock dealers (recipients of supplies) because of issuing the credit notes or the
plaintiff (provider of supplies) in view of the fact that he neither took steps to have the credit notes
corrected by the dealers nor, with the aid of the information in the credit notes, issued the dealers
with a separate tax mention in the correct amount.
40.
The Bundesfinanzhof also states that it is
of decisive importance whether the first sentence of Paragraph 14(2) of the UStG 1980 is
applicable with respect to the plaintiff, and that it is important to proceed from the correct
Community law framework for the national legislature in interpreting and applying the provision.
41.
I consider, in other words, that the Court of Justice is being asked to rule whether the above provisions
of the Sixth Directive allow a trader to be made liable for the excess amount of VAT mentioned in a credit
note if he has not contested that excess amount, in the same way that he would be liable for excess tax
mentioned in an invoice issued by him.
42.
However, in case the court should not accept that interpretation of the reference, I shall examine the
scope of art 21(1)(c) of the Sixth Directive independently of the nature of the document with the incorrect
mention of VAT.
1. Whether the recipient of a credit note who has not contested the amount stated in that note is
liable
43.
A trader who carries out a taxable transaction is liable in principle for the corresponding VAT, and as
such must issue an invoice which mentions it. As I have concluded at paras 29 to 31, above, the
obligation to issue an invoice is satisfied in the case of self-billing if the conditions under which a credit
note may be equated with an invoice are met, so that the special nature of that situation makes no
difference to the liability of the trader.
44.
It might be different, however, if for various reasonsfraud or mistakethe amount of VAT appearing
in the credit note does not correspond to the VAT due. National law and Community law appear, on that
hypothesis, to limit the extent to which a credit note is equated with an invoice.
45.
While the UStG clearly states the principle of payment by the trader of the amount of VAT mentioned
by him in an invoice where that amount is greater than that for which he was originally liable, it is silent
where the same difference relates, as in the present case, to the content of a credit note.
46.
Moreover, the wording of the Sixth Directive raises a doubt as to the national courts discretion to
require the trader, as the recipient of the credit note, to pay the 185 excess. Article 21(1)(c), which refers
to the case when, as here, the amount stated does not correspond to the taxable transaction, imposes
liability for VAT on any person who mentions the tax in an invoice or document serving as invoice (see
para 24, above).
47.
The recipient of the credit note has not, strictly speaking, himself mentioned the excess amount of VAT
in that document.
48.
It would then follow that while the wording of art 22(3)(a) of the Sixth Directive does not preclude the
trader from remaining the issuer of the credit note, the wording of the provisions of art 21 requires, by
contrast, the recipient of the goods or services to be liable for the VAT.
49.
Here too I prefer a more flexible interpretation of the text. By including among those liable for VAT,
besides the taxable person, the person who mentions VAT in the invoice, the provision aims at
discouraging tax fraud. It places the liability for payment of the VAT on persons who issue invoices which
are incorrect or correspond to fictitious economic operations (see the opinion of Advocate General Mischo
in Genius Holding v Staatssecretaris van Financin Case C-342/87 [1991] STC 239, [1989] ECR 4227
(paras 10, 14)).
50.
The aim pursued does not justify a different interpretation from that I have suggested for the answer to
the first question. A trader faced with an incorrect credit note, who is entitled to check its content and
correct it, must if he does not contest it be regarded as the person who has mentioned the tax. He is
therefore liable for the errors which have been made, which may be indicative of fraud.
51.
I consider, moreover, in common with all the intervening governments and the Commission, that the
answer cannot vary according to whether the mention of VAT is in a traditional invoice or a credit note. If
under the above conditions a credit note is equated with an invoice, on the ground that the trader remains
the real issuer of the credit note, it is natural that the legal consequences attached to the two documents
should be the same. The equation would otherwise be purely formal and the different rules would create
unjustified discrimination between traders who issue their invoices themselves and traders who use the
self-billing system.
2. The obligation to pay the VAT mentioned in the invoice or credit note
52.
If the amount invoiced does not, or not completely, correspond to the taxable economic transaction,
does the obligation to pay the VAT extend to the entire amount shown?
53.
That question, the answer to which must be the same in Community law whether the document in
which the incorrect amount is mentioned is an invoice or a credit note, arises in similar terms to those in
the Genius Holding case.
54.
In that case the court was asked whether the amount of VAT which is owed solely because it is
mentioned in an invoice may be taken into account for the exercise of the right to deduct provided for in
the Sixth Directive. The court held that it could not, considering that the right to deduct may be exercised
only in respect of taxes actually due, that is to say, the taxes corresponding to a transaction subject to
VAT or paid in so far as they were due (see the judgment in Genius Holding [1991] STC 239, [1989] ECR
4227 (para 13)).
55.
The court, anxious to combat tax evasion, thus decided that it was necessary to determine the amount
of the right to deduct by reference to the taxable transaction, which amounts to depriving invoices of their
function of evidencing the right to deduct where they do not correspond to any transaction or are based
on an overvalued tax base (see [1991] STC 239, [1989] ECR 4227 (para 17)). If there is a contradiction
between them, the tax corresponding to the taxable transaction must 186 take priority over the tax
invoiced, so as not to allow an undue deduction to be made.
56.
According to the judgment in Genius Holding [1991] STC 239, [1989] ECR 4227 (para 15), the
information in the invoice does not have as much weight as the reality of the taxable transaction, at least
with regard to exercise of the right to deduct. The same question may arise in the present case with
reference to the taxable persons obligation to pay VAT 10.
10
The Bundesfinanzhof states that the rule of payment in full of the amount invoiced, laid down in the first sentence of
para 14(2), is based on the consideration that the recipient is entitled to deduct as input tax the tax which is invoiced to
him separately and that the legislative plan thus clearly amounted to balancing the deduction of input tax on the basis
of an excessive separate mention of tax, to the extent not justified by supplies, by establishing a corresponding tax
liability. That provision thus derives from the right, recognised in the German case law, of the recipient of the goods or
services to deduct input tax to the full extent of the amount of tax mentioned separately in an invoice, even if the
amount exceed[s] the turnover tax owed on the basis of the taxable supply, which clashes with the judgment Genius
Holding.

57.
On this point, my view is that the invoice should preserve its function of documentation. My reason for
proposing that the court should thus limit the scope of its previous case law is no different from that which
inspired the solution adopted in that case: to discourage tax evasion.
58.
It would be an incitement to fraud if, in the event of a difference between the amount invoiced and the
amount following from the taxable transaction, the incorrect part of the amount shown in the invoice could
fall outside the obligation to pay on the part of the author of the invoice.
59.
Above all, that would make the invoice purposeless and deprive the supervisory authorities of a
reference document describing the economic operation, on the basis of which checks may be made.
60.
This, moreover, is the solution the court implicitly adopted in the judgment in Genius Holding [1991]
STC 239, [1989] ECR 4227 (para 18), when it observed that the member states may provide in their
internal legal systems for the possibility of correcting any tax improperly invoiced where the person who
issued the invoice shows that he acted in good faith. In other words, the principle is that the person who
issues an invoice mentioning excess tax must pay it, unless he is able to prove that there was no
fraudulent intent behind the amount stated.
61.
 Moreover, the priority given to what is mentioned in the invoice is dictated by the relevant provisions of
the Sixth Directive. The Genius Holding judgment is based largely on art 17(2)(a) of the Sixth Directive,
concerning the right to deduct11. In the present case, the provision whose scope is decisive is art 21(1)(c)
of the Sixth Directive. That provision admittedly defines the person liable for VAT and gives no direct
information as to the amount to be paid. However, the fact that it is intended to apply to cases where VAT
is not legally due, and where the economic operation cited might not even exist, shows that the mention in
the invoice is all that matters, since it constitutes the only means of referring to a specific figure.
11
That provision provides: the taxable person shall be entitled to deduct from the tax which he is liable to pay: (a) value
added tax due or paid (my emphasis).

62.
In view of the answer given to the second question, there is no need to express an opinion on the
national courts third question.

CONCLUSION
63.
In the light of the above considerations, I propose that the Court of Justice should answer as follows:
187
(1) Article 22(3)(c) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the
member states relating to turnover taxescommon system of value added tax: uniform basis of
assessment allows member states to regard a credit note issued by the recipient of a supply of
goods or services as a document serving as an invoice, where it includes the information
prescribed for invoices by the directive and its content may be corrected or contested by the trader
who has carried out the economic operation.
(2) Article 21(1)(c) of the directive allows a person who accepts a credit note mentioning a
greater amount of VAT than that due, without contesting the amount thus mentioned, to be
regarded as a person who has mentioned VAT in an invoice or document serving as an invoice and
is consequently liable to pay it.

17 September 1997.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 14 March 1996, received at the Court of Justice of the European Communities on 29 April
1996, the Bundesfinanzhof (the Federal Finance Court) referred to the Court of Justice for a preliminary
ruling under art 177 of the EC Treaty three questions on the interpretation of arts 21(1)(c) and 22(3)(c) of
Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover
taxescommon system of value added tax: uniform basis of assessment (the Sixth Directive).
2.
Those questions were raised in proceedings between the Finanzamt (the tax office) Osnabrck-Land
and Mr Langhorst concerning the question whether Mr Langhorst is liable to pay the amount of value
added tax (VAT) mentioned on a credit note issued by a customer, which he did not contest, even though
the amount is higher than that owed by reason of the taxable transactions.

The Sixth Directive

3.
Article 21(1)(a) and (c) of the Sixth Directive provides:
The following shall be liable to pay value added tax: (1) under the internal system: (a) taxable
persons who carry out taxable transactions other than those referred to in Article 9(2)(e) and
carried out by a taxable person resident abroad (c) any person who mentions the value added
tax on an invoice or other document serving as invoice
4.
Article 22(3) of the Sixth Directive further provides:
(a) Every taxable person shall issue an invoice, or other document serving as invoice in respect
of all goods and services supplied by him to another taxable person, and shall keep a copy thereof.
Every taxable person shall likewise issue an invoice in respect of payments on account made to
him by another taxable person before the supply of goods or services is effected or completed. (b)
The invoice shall state clearly the price exclusive of tax and the corresponding tax at each rate as
well as any exemptions. (c) The Member States shall determine the criteria for considering whether
a document serves as an invoice.

German law
5.
Paragraph 14(1) of the Umsatzsteuergesetz (the Law on Turnover Tax) (the UStG) of 26 November
1979, as amended by the Law of 18 August 1980, provides:
188
 If a trader carries out taxable supplies or other services under Paragraph 1(1)(1), he is entitled
and, where he carries out the transactions to another trader for the latters undertaking, obliged on
request by the latter, to issue invoices in which the tax is shown separately. Such invoices must
include the following information: (1) the name and address of the trader providing the services, (2)
the name and address of the recipient of the services, (3) the quantity and the usual commercial
description of the subject-matter of the supply or the type and extent of the other service, (4) the
date of the supply or other service, (5) the consideration for the supply or other service (Paragraph
10 above) and (6) the amount of tax due on the consideration (point 5 above).
6.
Paragraph 14(2) of the UStG then provides:
If the trader has in an invoice for a supply or other service shown separately a higher amount of
tax than he owes under this Law in respect of the transaction, then he shall also owe the additional
amount. If he corrects the amount of tax as against the recipient, then Paragraph 17(1) shall apply
correspondingly.
7.
Under para 14(5) of the UStG:
A credit note by which a trader settles up for a taxable supply or other service made to him shall
also be deemed to be an invoice. A credit note shall be recognised if the following conditions are
met: (1) The trader providing the service (the recipient of the credit note) must be entitled under
sub-paragraph (1) to show the tax separately in an invoice. (2) There must be agreement between
the issuer and the recipient of the credit note that the supply or other service is to be settled by a
credit note. (3) The credit note must include the information prescribed in the second sentence of
sub-paragraph (1) above. (4) The credit note must have been delivered to the trader providing the
service. Sentences 1 and 2 above shall apply by analogy to credit notes which the trader issues as
payment or partial payment for a taxable supply or other service which has not yet been carried
out. The credit note shall cease to have effect as an invoice in so far as the recipient contests the
amount of tax shown therein.

Facts of the main proceedings

8.
Mr Langhorst, a farmer, declared his turnover for 1985, the year at issue in the present proceedings,
after having elected under para 24(4) of the UStG to be taxed at the rate of 7% instead of the rate of 13%
provided for in the first sentence of para 24(1) of the UStG.
9.
Unaware of that fact, livestock dealers whom Mr Langhorst had supplied with fat pigs issued him with
credit notes which mentioned separately VAT calculated at 13%. Mr Langhorst did not at first contest the
amount of VAT mentioned in the credit notes.
10.
Mr Langhorst subsequently brought proceedings in the Niederschsisches Finanzgericht (the Finance
Court of Lower Saxony), which gave judgment on 10 October 1989 reducing the amount of VAT. By
application of 27 January 1992, supplemented on 26 March 1992, the Finanzamt appealed on a point of
law to the Bundesfinanzhof against that decision.
11.
In its order for reference, the Bundesfinanzhof considers that the reduction of VAT was calculated
correctly by the Niederschsisches Finanzgericht, but that the 189 Finanzamts appeal could nevertheless
be upheld if, under the first sentence of para 14(2) of the UStG, Mr Langhorst had to pay the total amount
of VAT mentioned separately in the credit notes, including the additional amounts which were not justified
by the taxable transactions.
12.
The Bundesfinanzhof observes, however, that the wording of para 14(2) of the UStG expressly
describes as the person solely liable for the higher amount of tax the trader who has mentioned the higher
amount of VAT separately in an invoice. In the present case, however, it was not the supplier, Mr
Langhorst, who issued the document which indicated a higher amount of VAT but his customers, the
livestock merchants. The Bundesfinanzhof also states that in so far as the credit notes issued by the
livestock merchants are deemed to be invoices under para 14(5) of the UStG, thus allowing them to
deduct an amount of VAT which is not in fact justified, it could be argued that Mr Langhorst is liable for
that amount of VAT, since he did not contest it.
13.
The Bundesfinanzhof is uncertain whether such an interpretation of para 14(2) of the UStG is
consistent with Community law, in particular art 21(1)(c) of the Sixth Directive, under which any person
who mentions VAT on an invoice or other document serving as an invoice is liable to pay VAT, and art
22(3)(c) of that directive, which gives the member states power to determine the criteria under which a
document may serve as an invoice, but does not expressly state that credit notes issued by customers
may be treated as invoices issued by the taxable person. The Bundesfinanzhof accordingly stayed the
proceedings and referred the following questions to the Court of Justice for a preliminary ruling:
(1) Is it permissible under Article 22(3)(c) of the Sixth Council Directive for a credit note
within the meaning of Paragraph 14(5) of the Umsatzsteuergesetz 1980 to be regarded as an
invoice or other document serving as an invoice (Article 21(1)(c) of the Sixth Directive)?
 (2) If so, is it permissible under Article 21(1)(c) of the Sixth Directive for a person who accepts a
credit note showing a higher amount of tax than that owed by reason of taxable transactions, and
does not contest in that respect the amount of tax mentioned in the credit note, to be regarded as a
person who mentions value added tax in an invoice or other document serving as an invoice and is
therefore liable to pay that value added tax?
(3) Can the recipient of a credit note, in the circumstances set out in Question 2, rely on Article
21(1)(c) of the Sixth Directive if the value added tax mentioned in the credit note is claimed from
him as a tax debt to the extent of the difference between the tax mentioned and the tax owed by
reason of taxable transactions?
14.
The order for reference explains that question 3 arises only if the answer to question 2 is negative.

Question 1
15.
By its first question the national court essentially asks whether art 22(3)(c) of the Sixth Directive
authorises member states to regard a credit note issued by the recipient of goods or services as a
document serving as an invoice.
16.
Under art 22(3)(a) of the Sixth Directive, an invoice or document serving as such must be issued by
every taxable person in respect of all goods and services supplied by him to another taxable person and,
under art 22(3)(b), that that invoice or document serving as such must state clearly the price exclusive of
tax and the corresponding tax at each rate as well as any exemptions.
190
17.
The minimum conditions regarding the information which must appear on the invoice or document
serving as such having thus been laid down, art 22(3)(c) leaves the member states free to determine the
criteria for considering whether such a document serves as an invoice. That power must, however, be
exercised consistently with one of the objectives of the Sixth Directive, namely to ensure that VAT is
levied and collected under the supervision of the tax authorities (see, to this effect, the judgments in
Jeunehomme v Belgium Joined cases 123/87 and 330/87 [1988] ECR 4517 (paras 16, 17) and Reisdorf v
Finanzamt Kln-West Case C-85/95 [1997] STC 180, [1996] ECR I-6257 (para 24)).
18.
It must therefore be considered whether a credit note which includes, as in the case in the main
proceedings, the compulsory information referred to in art 22(3)(b) of the Sixth Directive may be regarded
as serving as an invoice even though it is issued by the recipient of the goods or services, where under
the relevant national provisions the issuer and the recipient of the credit note must have agreed that the
supply or service is to be settled by a credit note, the credit note must have been delivered to the trader
providing the service, and the credit note may no longer serve as an invoice if its recipient contests the
amount of tax stated therein.
19.
Article 22(3)(a) of the Sixth Directive refers to the normal situation where an invoice or document
serving as an invoice is issued by the taxable person who supplies the goods or services.
20.
However, as Advocate General Lger observes in paras 29ff of his opinion, above, since the purpose
of that provision is to ensure correct collection of the tax and to avoid fraud, there is no reason why the
document in question should not be drawn up by the recipient of the goods or services, provided that it
includes the information prescribed for an invoice and the taxable person who supplies the goods or
services has been given the opportunity to ask, if necessary, for the information to be corrected.
21.
In such a case, since he has that power of control, the taxable person may be regarded as the author
of the document, the drawing-up of which he has, as it were, delegated to his customer. The credit note
thus fulfils the function of documenting the taxable persons rights and obligations with respect to VAT,
since it contains the same information as a traditional invoice and the taxable person is free to approve its
content.
22.
As the German and UK governments have rightly observed, a credit note issued by the recipient of the
goods or services is in many cases the best means of accounting for the supplies effected, in that it is
only the recipient of the goods or services who is in a position to check that they comply with the terms of
the contract.
23.
In those circumstances, it cannot be deduced from the fact that art 22(3)(a) of the Sixth Directive
provides only for the issuing by the taxable person who supplies the goods or services of an invoice or
document serving as such that it is not possible for the member states to regard a document as serving
as an invoice solely because it has been issued by the recipient of the goods or services.
24.
The answer to question 1 must therefore be that art 22(3)(c) of the Sixth Directive authorises member
states to regard a credit note issued by the recipient of the goods or services as a document serving as
an invoice, where it includes the information prescribed for invoices by that directive, it is drawn up with
the agreement of the taxable person who supplies the goods or services, and the latter is able to contest
the amount of VAT mentioned.
191

Question 2
25.
By its second question the national court seeks to establish whether a taxable person who has not
contested the mention, in a credit note serving as an invoice, of an amount of VAT greater than that owed
by reason of taxable transactions may be regarded as the person who has mentioned that amount, and is
consequently liable for the amount stated, within the meaning of art 21(1)(c) of the Sixth Directive.
26.
Article 21(1)(c) of the Sixth Directive provides that any person who mentions the VAT on an invoice or
other document serving as invoice is liable to pay VAT under the internal system.
27.
Where, as in the case in the main proceedings, a credit note serves as an invoice, the taxable person
must be regarded as the person who has in fact mentioned VAT in the credit note, within the meaning of
art 21(1)(c) of the Sixth Directive, and is consequently liable to pay the amount stated.
28.
Were it otherwise, part of the VAT appearing in the document serving as an invoice would not have to
be paid by the taxable person, even though, as the order for reference observes, that VAT might have
been deducted in full by the recipient of the goods or services, thus giving scope for possible fraud or
collusion, contrary to the proper functioning of the common system of VAT established by the Sixth
Directive and to its objective of ensuring that VAT is levied and collected under the supervision of the tax
authorities.
29.
The answer to question 2 must therefore be that a taxable person who has not contested the mention,
in a credit note serving as an invoice, of an amount of VAT greater than that owed by reason of taxable
transactions may be regarded as the person who has mentioned that amount, and is consequently liable
to pay the amount shown, within the meaning of art 21(1)(c) of the Sixth Directive.

Question 3
30.
In view of the answer to question 2, there is no need to answer question 3.

Costs
31.
The costs incurred by the German, Greek and UK governments and by the European Commission,
which have submitted observations to the Court of Justice, are not recoverable. Since these proceedings
are, for the parties to the main proceedings, a step in the action pending before the national court, the
decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the
Bundesfinanzhof by order of 14 March 1996, hereby rules:
(1) Article 22(3)(c) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member
states relating to turnover taxescommon system of value added tax: uniform basis of assessment
authorises member states to regard a credit note issued by the recipient of the goods or services as a
document serving as an invoice, where it includes the information prescribed for invoices by that
directive, it is drawn up with the agreement of the taxable person who supplies the goods or services, and
the latter is able to contest the amount of VAT mentioned. (2) A taxable person who has not contested the
mention, in a credit note serving as an invoice, of an amount of VAT greater than that owed by reason of
taxable transactions may be regarded as the person who has mentioned that amount, and is
consequently liable to pay the amount shown, within the meaning of art 21(1)(c) of the directive.

192

[1998] All ER (EC) 193

Grant v South-West Trains Ltd

(Case C-249/96)

EUROPEAN COMMUNITY; Social policy: EMPLOYMENT; Discrimination

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, RAGNEMALM, WATHELET
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, KAPTEYN, MURRAY, EDWARD,
PUISSOCHET (RAPPORTEUR), HIRSCH, JANN AND SEVN ADVOCATE GENERAL ELMER
9 JULY, 30 SEPTEMBER 1997, 17 FEBRUARY 1998
European Community Equality of treatment for men and women Equal pay for equal work Employer
only granting family travel concessions to unmarried employees where partner of opposite sex Whether
Community provisions prohibiting discrimination on grounds of sex covering discrimination on grounds of
sexual orientation Council Directive (EEC) 75/117 EC Treaty, art 119.

The plaintiff applied to her employer for travel concessions for her female partner, with whom she
declared she had had a meaningful relationship for over two years. Her employer refused the request on
the ground that, under company regulations, travel concessions for unmarried persons could only be
granted for a partner of the opposite sex. The plaintiff subsequently brought proceedings before an
industrial tribunal, claiming that the refusal constituted discrimination on grounds of sex and was contrary
to Community provisions on equal pay. The tribunal stayed the proceedings and referred to the Court of
Justice of the European Communities for a preliminary ruling questions as to whether an employers
refusal to grant travel concessions to the person of the same sex with whom an employee had a stable
relationship constituted discrimination prohibited by art 119 1 of the EC Treaty and Council Directive (EEC)
75/117 on the approximation of the laws of the member states relating to the application of the principle of
equal pay for men and women, where such concessions were granted to an employees spouse or a
person of the opposite sex with whom an employee had a stable relationship outside marriage.
1
Article 119, so far as material, is set out at p 198 h, post

Held The condition imposed by the employers regulations applied in the same way to female and male
workers and did not therefore constitute discrimination based directly on sex for the purposes of art 119 of
the Treaty. Moreover, in the present state of the law within the Community, stable relationships between
two persons of the same sex were not regarded as equivalent to marriage or stable relationships outside
marriage between two persons of the opposite sex. That could be inferred from, inter alia, the absence of
any Community rule providing for such equivalence, from the treatment of such relationships by member
states, where for the most part they were either treated as equivalent only with respect to a limited
number of rights, or else not recognised in any particular way, and from the fact that national provisions
which accorded more favourable treatment to married persons and stable relationships between two
persons of the opposite sex, for the purpose of protecting the family, were not considered to be contrary
to art 14 of the Convention for the Protection of Human Rights and Fundamental Freedoms 1950. It
followed that an employers refusal to grant travel concessions 193 to an employees partner of the same
sex, where such concessions were available to a spouse or the partner of the opposite sex with whom the
employee had a stable relationship outside marriage, did not constitute discrimination prohibited by art
119 of the Treaty or Directive 75/117 (see p 207 h, p 208 a to f and p 210 b d e, post).
S v UK (1986) 47 D & R 274, B v UK (1990) 64 D & R 278, Cossey v UK (1990) 13 EHRR 622
considered.
Per curiam. Although the court has previously argued against a restrictive interpretation of art 119, that
reasoning only applies to discrimination based essentially, if not exclusively, on the sex of the person
concerned. It is therefore limited to the case of a workers gender reassignment and does not apply to
differences of treatment based on a persons sexual orientation (see p 209 b, post); P v S Case C-13/94
[1996] All ER (EC) 397 distinguished.

Notes
For the principle of equal pay for men and women, see 52 Halsburys Laws (4th edn) paras 2111, and
2112, and for cases on the subject, see ibid paras 21162117.
For the EC Treaty, art 119, see 50 Halsburys Statutes (4th edn) 306.

Cases cited
B v UK (1990) 64 D & R 278, E Com HR.
C v UK App No 14753/89 (9 October 1989, unreported) E Com HR.
Coloroll Pension Trustees Ltd v Russell Case C-200/91 [1995] All ER (EC) 23, [1994] ECR I-4389, ECJ.
Cossey v UK (1990) 13 EHRR 622, E Ct HR.
Davis v Johnson [1978] 1 All ER 1132, [1979] AC 264, [1978] 2 WLR 553, HL.
Defrenne v Sabena Case 43/75 [1981] 1 All ER 122, [1976] ECR 455, ECJ.
Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990]
ECR I-3941.
Dzodzi v Belgium Joined cases C-297/88 and C-197/89 [1990] ECR I-3763.
Enderby v Frenchay Health Authority Case C-127/92 [1993] ECR I-5535.
Garland v British Rail Engineering Ltd Case 12/81 [1982] 2 All ER 402, [1983] AC 751, [1982] 2 WLR 918,
[1982] ECR 359, ECJ.
Gillespie v Northern Health and Social Services Board Case C-342/93 [1996] All ER (EC) 284, [1996]
ECR I-475, ECJ.
Kerkhoven v the Netherlands App No 15666/89 (19 May 1992, unreported), E Com HR.
McDermott v Minister for Social Welfare Case 286/85 [1987] ECR 1453.
Neath v Hugh Steeper Ltd Case C-152/91 [1994] 1 All ER 929, [1993] ECR I-6935, ECJ.
Nimz v Freie und Hansestadt Hamburg Case C-184/89 [1991] ECR I-297.
Opinion 2/94 [1996] ECR I-1759.
Orkem v EC Commission Case 374/87 [1989] ECR 3283.
P v S Case C-13/94 [1996] All ER (EC) 397, [1996] ECR I-2143, ECJ.
Rees v UK (1986) 9 EHRR 56, E Ct HR.
S v UK (1986) 47 D & R 274, E Com HR.
Society for the Protection of Unborn Children Ireland Ltd v Grogan Case C-159/90 [1991] ECR I-4685.
Toonen v Australia Communication [1994] 1-3 IHRR 97, HR Committee.
Van Cant v Rijksdienst voor Pensioenen Case C-154/92 [1993] ECR I-3811.
X v UK (1983) 32 D & R 220, E Com HR.
194

Reference
By decision of 19 July 1996, the Industrial Tribunal, Southampton, referred to the Court of Justice of the
European Communities for a preliminary ruling under art 177 of the EC Treaty six questions (set out at p
205 d to g, post) on the interpretation of art 119 of that Treaty, Council Directive (EEC) 75/117 on the
approximation of the laws of the member states relating to the application of the principle of equal pay for
men and women, and Council Directive (EEC) 76/207 on the implementation of the principle of equal
treatment for men and women as regards access to employment, vocational training and promotion, and
working conditions. Those questions were raised in proceedings between Ms Grant and her employer
South-West Trains Ltd (SWT) concerning the refusal by SWT of travel concessions for Ms Grants female
partner. Written observations were submitted on behalf of: Ms Grant, by C Booth QC, and by P Duffy and
M Demetriou, Barristers; SWT, by N Underhill QC and M Shanks, Barrister; the UK government, by J E
Collins, of the Treasury Solicitors Department, acting as agent; and S Richards and D Anderson,
Barristers; the French government, by C de Salins, Deputy Director in the Legal Affairs Department of the
Ministry of Foreign Affairs, and A de Bourgoing, Charg de Mission in that department, acting as agents;
and the European Commission, by C Docksey, M Wolfcarius and C OReilly, of its Legal Service, acting
as agents. Oral observations were made by: Ms Grant, represented by C Booth QC, P Duffy QC and M
Demetriou; SWT, represented by N Underhill QC and M Shanks; the UK government, represented by J E
Collins, D Anderson and P Elias QC; and the Commission, represented by C OReilly and M Wolfcarius.
The language of the case was English. The facts are set out in the opinion of the Advocate General.

30 September 1997.

The Advocate General (M B Elmer)

delivered the following opinion (translated from the Danish).
1.
Does a provision in an employers pay regulations according to which the employee is to be granted a
pay benefit in the form of travel concessions for a cohabitee of the opposite gender to the employee, but
is denied such concessions for a cohabitee of the same gender as the employee, constitute gender
discrimination in breach of art 119 of the EC Treaty?

The case before the national court and the questions referred for a preliminary ruling
2.
On 4 June 1993 Lisa Grant was engaged as a clerical officer by the British Railways Board. On 31
March 1995 the employment relationship was transferred to South-West Trains Ltd (SWT), a wholly-
owned subsidiary, which was privatised on 4 February 1996. Clause 18 of her contract of employment,
entitled Travel facilities, states:
You will be granted such free and reduced rate travel concessions as are applicable to a
member of your grade. Your spouse and dependants will also be granted travel concessions. Travel
concessions are granted at the discretion of [the employer] and will be withdrawn in the event of
their misuse.
3.
Those travel concessions are further regulated in the Staff Travel Facilities Privilege Ticket Regulations
(the ticket regulations), issued by the British Railways Board and adopted by SWT after privatisation.
Clause 8 of the ticket regulations, entitled Spouses, provides inter alia:
195
Privilege tickets are granted for one common law opposite sex spouse of staff subject to a
statutory declaration being made that a meaningful relationship has existed for a period of two
years or more
4.
Under cls 10 and 11 of the ticket regulations, the employee is also entitled to concessions for
unmarried children living at home. Under cl 12 it is further stated thatPrivilege tickets may be issued
for a relative acting as a bona fide permanent resident housekeeper to and entirely dependent upon the
applicant if the employee is either living alone or with an invalid spouse. Under that clause relative is
defined as a mother, father, brother sister, daughter or son.
5.
Mr Potter, who was Lisa Grants predecessor in post, had in his time made a statutory declaration that
a meaningful relationship had existed between him and his female cohabitee for a period of two years or
more, and on that basis had obtained travel concessions for her.
6.
On 9 January 1995 Lisa Grant similarly applied for travel concessions for her female cohabitee, Jillian
Percey, at the same time making a declaration that she lived together with
the individual described as my common law spouse on my application for concessionary travel
facilities in a Common law relationship and that I have so lived for a continuous period of two
years or more
Lisa Grants application was rejected on the ground that, under cl 8 of the ticket regulations, travel
concessions were not granted for cohabitees of the same sex.
7.
Lisa Grant then brought a case against SWT before the Industrial Tribunal, Southampton, United
Kingdom, claiming that art 119 of the EC Treaty precluded her being denied part of her pay consisting in
obtaining travel concessions for her female cohabitee, when a male employee in the same circumstances
would obtain travel concessions for his female cohabitee.
8.
By order registered at the Court of Justice of the European Communities on 22 July 1996 the Industrial
Tribunal stayed the proceedings and referred the following questions to the Court of Justice:
(1) Is it (subject to (6) below) contrary to the principle of equal pay for men and women
established by Article 119 of the Treaty establishing the European Community and by Article 1 of
Council Directive 75/117 for an employee to be refused travel concessions for an unmarried
cohabiting same-sex partner where such concessions are available for spouses or unmarried
opposite-sex cohabiting partners of such an employee?
(2) For the purposes of Article 119 does discrimination based on sex include discrimination
based on the employees sexual orientation?
(3) For the purposes of Article 119, does discrimination based on sex include discrimination
based on the sex of that employees partner?
(4) If the answer to Question (1) is yes, does an employee, to whom such concessions are
refused, enjoy a directly enforceable Community right against his employer?
(5) Is such a refusal contrary to the provisions of Council Directive 76/207?
(6) Is it open to an employer to justify such refusal if he can show (a) that the purpose of the
concessions in question is to confer benefits on married partners or partners in an equivalent
position to married partners and (b) that relationships between same-sex cohabiting partners have
not traditionally been, and are not generally, regarded by society as equivalent to marriage;
196rather than on the basis of an economic or organizational reason relating to the employment in
question?

Which rules are relevant?

9.
In Garland v British Rail Engineering Ltd Case 12/81 [1982] 2 All ER 402, [1982] ECR 359 the court
held that travel concessions for employees and members of their family should be treated as pay within
the meaning of art 119 of the EC Treaty.
10.
Council Directive (EEC) 75/117 on the approximation of the laws of the member states relating to the
application of the principle of equal pay for men and women (OJ 1975 L45 p 19), which on certain points
clarifies the content of art 119 of the Treaty, has no independent significance where a pay benefit falling
within the scope of art 119 is involved (see the judgment in Defrenne v Sabena Case 43/75 [1981] 1 All
ER 122, [1976] ECR 455 (para 54)). Directive 75/117 is therefore not relevant to this case (see the
judgment in Garland [1982] 2 All ER 402, [1982] ECR 359 (para 12)).
11.
Council Directive (EEC) 76/207 on the implementation of the principle of equal treatment for men and
women as regards access to employment, vocational training and promotion, and working conditions (OJ
1976 L39 p 40) does not cover pay benefits, and is therefore also irrelevant to the present case.
12.
The questions referred to the court must therefore be answered on the basis of art 119 of the Treaty
alone. Since the above directives supplement and develop the basic principle of equal treatment
contained in art 119 of the Treaty, the courts case law concerning those directives is nevertheless of great
importance in this case.

General remarks concerning gender discrimination

13.
In its judgment in P v S Case C-13/94 [1996] All ER (EC) 397, [1996] ECR I-2143, the court had an
opportunity to give a ruling that dealt more fundamentally with the scope of the Community rules
prohibiting discrimination based on sex. The case concerned an employee who was dismissed after
informing his employer that he intended to undergo gender reassignment. The court held that the
dismissal constituted discrimination based on sex that was contrary to Directive 76/207. In para 21 the
court attributed great weight to the fact that such discrimination was based, essentially if not exclusively,
on the sex of the person concerned. It was therefore irrelevant as far as the court was concerned that
there was discrimination because of Ps transsexuality.
14.
I would in particular refer to paras 19 to 22 of the judgment, where the court stated as follows:
(19) Moreover, as the court has repeatedly held, the right not to be discriminated against on
grounds of sex is one of the fundamental human rights whose observance the court has a duty to
ensure
(20) Accordingly, the scope of the directive cannot be confined simply to discrimination based on
the fact that a person is of one or other sex. In view of its purpose and the nature of the rights
which it seeks to safeguard, the scope of the directive is also such as to apply to discrimination
arising, as in this case, from the gender reassignment of the person concerned.
(21) Such discrimination is based, essentially if not exclusively, on the sex of the person
concerned. Where a person is dismissed on the ground that he or she intends to undergo, or has
undergone, gender reassignment, he or she 197 is treated unfavourably by comparison with
persons of the sex to which he or she was deemed to belong before undergoing gender
reassignment.
(22) To tolerate such discrimination would be tantamount, as regards such a person, to a failure
to respect the dignity and freedom to which he or she is entitled, and which the court has a duty to
safeguard.
15.
In that judgment the court, in my view, took a decisive step away from an interpretation of the principle
of equal treatment based on the traditional comparison between a female and a male employee. The
court thus held that it did not matter that there was no reason to think that a woman who wished to
undergo gender reassignment would have been treated more favourably than a man who wished to do
so. I would refer in particular to para 20 of the judgment, where the court refused to confine the principle
of equal treatment simply to discrimination based on the fact that a person is of one or other sex. The
essential point was that the discrimination was based exclusively, or essentially, on gender. The court
thereby, in my view, interpreted the Community principle of equal treatment in a way that renders the
principle appropriate for dealing with the cases of gender discrimination that come before the courts in
present-day society.
16.
The courts judgment in P v S technically concerned Directive 76/207, but because of its general
character it has corresponding significance for art 119 of the EC Treaty which sets out the basic principle
prohibiting discrimination based on sex. In order to give full effect to that principle, in my view it is, in the
same way as in the P v S case, appropriate to construe art 119 of the Treaty as precluding forms of
discrimination against employees based exclusively, or essentially, on gender. The provision must further,
in order to be effective, be understood as prohibiting discrimination against employees not solely on the
basis of the employees own gender but also on the basis of the gender of the employees child, parent or
other dependent. The provision must therefore also be regarded as precluding an employer from, for
instance, denying a household allowance to an employee for sons under 18 living at home when such an
allowance in otherwise equivalent circumstances was given for daughters living at home. Such a
construction, which generally attaches weight to the fact that gender is the factor giving rise to
discrimination, would also appear to accord with the formulation of art 119; which in its first paragraph
does indeed refer to the principle that men and women should receive equal pay for equal work, but in its
third paragraph expands on that principle by speaking more generally of equal pay without discrimination
based on sex. Article 119 of the Treaty must therefore be construed as covering all cases where gender
is objectively the factor causing an employee to be paid less.
17.
It is important to bear in mind that, in examining whether there is gender discrimination, a purely
objective assessment must be made. The decisive point is whether de jure or de facto there is objective
gender discrimination, not, however, what the subjective motivation of the employers discriminatory
conduct may be. The delimitation of the scope of art 119 must be kept free of conceptions of morality
which may vary from member state to member state and change with time. Only a purely objective
assessment will ensure the clarity and foreseeability which are of crucial significance for legal certainty. In
P v S, conceptions of morality in connection with transsexuality were thus irrelevant to the courts
decision. The court has thus confirmed that the Treaty cannot be interpreted on the basis of the moral
conceptions of a member state (in this 198 respect see also Society for the Protection of Unborn Children
Ireland Ltd v Grogan Case C-159/90 [1991] ECR I-4685).
18.
In summary, I therefore consider that art 119 of the Treaty covers all cases where, on an objective
assessment, there is de jure or de facto discrimination based exclusively or essentially on gender.

Is there gender discrimination in this case?

19.
SWT, the French government and the United Kingdom contend that in this case there is discrimination
based, not on sex, but rather on sexual orientation.
20.
Examination of the question whether, in this case, there is de jure gender discrimination must start with
a more detailed analysis of cl 8 of the ticket regulations. Under that clause, an employee is entitled to
travel concessions for a cohabitee of the opposite sex, but not for a cohabitee of the same sex. The issue
is therefore whether the different treatment of the two cases, viewed objectively, is exclusively or
essentially gender-based.
21.
I would point out more generally that the travel concessions under the ticket regulations are in reality
household benefits. Under cls 10 to 12 of the ticket regulations employees are entitled to travel
concessions for children and close relatives maintained by the employee and under those clauses the
sexual orientation of the employees or their relatives is irrelevant. If, therefore, Lisa Grant was maintaining
a child or a mother or father, she would, regardless of her sexual orientation, have received travel
concessions for them.
22.
Where, under cl 8 of the ticket regulations, an employee is also entitled to travel concessions for a
cohabitee, under the wording of the provision the same applies as for the concessions under cls 10 to 12.
The concessions are in fact a household benefit. Clause 8, like cls 10 to 12, makes no mention of the
sexual orientation of the employee or cohabitee, and the question of sexual orientation is thus, under the
objective content of that clause, irrelevant as far as entitlement to the concessions is concerned. Nor does
SWT investigate the question of the employees sexual orientation either by requiring the employee to
provide such information in the declaration to be produced or by inspecting the joint home.
23.
Clause 8 of the ticket regulations makes the concessions conditional, however, on the cohabitees
being of the opposite sex to the employee. The discrimination is therefore, under the objective content of
the provision, exclusively gender-based. Gender is simply the only decisive criterion in the provision. If the
rule had been gender-neutral so that the concessions were given, without discrimination, to all employees
who submitted a declaration that for at least the last two years they had been living in a permanent
relationship, Lisa Grant would have obtained the pay benefit in question which, according to her
undisputed evidence, is worth 1000 pa (corresponding to ECU 1,500). Gender is thus, objectively, the
factor that leads to discrimination relating to pay against a particular group of employees.
24.
Thus cl 8 of the ticket regulations makes the grant of the pay benefit in question dependent on the
gender of the employee, inasmuch as employees must be of the opposite sex to their cohabitees. At the
same time the clause contains a requirement that the cohabitee must be of the opposite sex to the
employee. Whether the requirement for obtaining the concessions is satisfied accordingly depends on the
gender both of the employee and of the cohabitee. Travel concessions for a male cohabitee may only be
obtained if the employee is a 199 woman. Travel concessions for a female cohabitee may only be
obtained if the employee is a man.
25.
The fact that cl 8 of the ticket regulations does not refer to a specific sex as the criterion for
discrimination, but lays down a more abstract criterion (opposite sex) can, in my view, make no
difference, since the decisive point, as laid down in P v S is whether discrimination is exclusively or
essentially based on sex, whereas the fact that the discrimination is, de jure or de facto, on the basis of a
specific sex cannot be decisive.
26.
In the light of the foregoing, it is my view that a provision in an employers pay regulations under which
the employee is granted travel concessions for a cohabitee of the opposite sex to the employee but
refused such concessions for a cohabitee of the same sex as the employee constitutes discrimination on
the basis of gender which falls within the scope of art 119 of the Treaty.

Does the case concern a family law issue falling outside the EC Treaty?
27.
A further question that must, however, be examined is whether that discrimination is a consequence of
the family law legislation in the member state in question. The European Commission has thus stated that
the case concerns the definition of a common law spouse and is thus a family law issue which does not
fall under the EC Treaty.
28.
Had cl 8 of the ticket regulations specified, as the determinant criterion, that the employee and the
cohabitee must have contracted marriage, that would, in my opinion, have been a restriction on the travel
concessions which was not contrary to Community law, because it would be by reference to a family law
concept, the content of which is laid down by the member states.
29.
There would, nevertheless, have been a precondition that male and female employees and their
spouses be treated in the same way. If an employer, on the basis of his private moral views, wished to
combat the breaking-down of the traditional sex roles by giving employees whose wives stayed at home a
special benefit but refusing to give employees whose husbands stayed at home a corresponding benefit,
such a rule would be contrary to art 119 of the Treaty, since that would involve discrimination based on
both the employees gender and the gender of their spouses, and would not simply refer to family law
status.
30.
 Clause 8 of the ticket regulations does not, however, refer to a concept which in English law confers a
family law status, but rather uses the expression common law spouse. However, neither in statute law
nor common law does that expression have any legal significance in England. English law has put
unmarried cohabitees on the same footing as married couples only in limited circumstances, for example
under rent legislation, and here a more precise formulation is used such as, for example a man and a
woman who lived with each other as husband and wife. Such provisions are assumed in general to
require that the couple in question have shared finances, share a social life and have sexual relations,
although the absence of the last is not decisive (see, further, Bromley and Lowe Family Law (1992, 8th
edn) pp 510).
31.
The term common law spouse and similar terms are thus not used in English family law legislation
and, in a case in the social law area, the members of the House of Lords expressed a certain reluctance
to employ the term in a legal context when stating inter alia that the case in question concerned an
unmarried woman commonly but not very appropriately referred to as a common law wife (see Davis v
Johnson [1978] 1 All ER 1132 at 1149, [1979] AC 264 at 340). The expression common law spouse or
similar expressions must therefore be 200 regarded simply as an expression used in everyday language
with no specifically defined content which is liable to change in accordance with changes in the general
view, so that in principle there is nothing to prevent the expression common law spouse extending to
cohabitees of the same gender. In English contract law the expression agency of cohabitation is,
moreover, used to mean that in certain circumstances cohabitation entitles a person to enter into certain
legal transactions which will bind his or her cohabitee (see Cheshire, Fifoot & Furmstons Law of Contract
(1996, 13th edn) pp 4912). Such entitlement is based on societys view of the permanency and
character of the cohabitation, and therefore there does not seem to be anything, in principle, to prevent
cohabitees of the same sex from satisfying the conditions for an agency of cohabitation.
32.
The fact that two persons of the same sex may, in the United Kingdom, be regarded as common law
spouses would also appear to be the case on an a contrario construction of cl 8 of SWTs own ticket
regulations. If the expression common law spouses referred exclusively to persons of different sexes,
there would be no reason to refer to a common law opposite sex spouse .
33.
It was SWT itself which introduced that restriction, leading to gender discrimination, into the term.
34.
Gender discrimination is accordingly, in this case, not the result of family law legislation in the member
state in question and for that reason outside the scope of Community law.

May gender discrimination be justified by reference to an employers conception of

morality?
35.
SWTs reason for restricting employees entitlement to travel concessions for cohabitees to entitlement
for persons of the opposite sex is, according to the evidence, its intention to benefit only persons who are
married or living in a heterosexual relationship, but not persons living in a homosexual relationship, since
cohabitation with a person of the same sex is not traditionally regarded as equivalent to a heterosexual
relationship. As the national court implies in its sixth question, consideration must be given to the question
of whether gender discrimination can be justified on the basis of conceptions of morality.
36.
Lisa Grant submits that discrimination under art 119 of the Treaty may only be justified if it is on
essential economic or business grounds or is required by law. On the other hand, an employer cannot
justify gender discrimination which is prohibited under the Treaty by reference to his private conception of
morality, regardless of the fact that it might correspond to the prevalent conception of morality in the
member state in question.
37.
I would begin my examination of that question by pointing out that in its assessment of whether
discrimination based on sex might be justified, the court has traditionally drawn a distinction between
direct and indirect discrimination (see the judgments in Dekker v Stichting Vormingscentrum voor Jong
Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990] ECR I-3941 (para 13), Nimz v Freie und
Hansestadt Hamburg Case C-184/89 [1991] ECR I-297 (para 15) and Enderby v Frenchay Health
Authority Case C-127/92 [1993] ECR I-5535 (para 14)). Whether discrimination is direct or indirect
depends on whether it follows directly from the legal criteria applied (direct discrimination) or whether,
without being evident under the legal criteria, in practice it proves detrimental to one sex (indirect
discrimination). Only where discrimination is indirect does the court appear to accept the possibility that it
might be justified by reference to objective 201 circumstances (see Dekker [1990] ECR I-3941 (para 13)
as compared with Enderby [1993] ECR I-5535 (para 14)).
38.
In the present case gender discrimination results directly from the legal criterion laid down in cl 8 of the
ticket regulations, and according to the courts case law such direct discrimination cannot be justified by
reference to objective circumstances.
39.
Even were it assumed, however, that this was a case of indirect discrimination that might be justified
by reference to objective circumstances, I find it difficult to see how discrimination such as this could be
so justified. An employees households expenses in travelling by train must, all things being equal, be just
as high regardless of whether the household consists of cohabitees of different sexes or of the same sex.
SWTs justification consists in reality simply in a reference to the purely subjective circumstance that its
intention is to treat homosexuals differently from heterosexuals. That constitutes therefore, in my view, a
purely subjective reason as opposed to objective circumstances such as, for instance, actuarial
calculations relating to the value of contributions paid in under certain forms of pension arrangements in
relation to average life expectancy for men and women (see the judgments in Coloroll Pension Trustees
Ltd v Russell Case C-200/91 [1995] All ER (EC) 23, [1994] ECR I-4389 (paras 7688) and Neath v Hugh
Steeper Ltd Case C-152/91 [1994] 1 All ER 929, [1993] ECR I-6935 (paras 2833)).
40.
South-West Trains justification amounts, in reality, to nothing more than saying that on the basis of its
own private conceptions of morality that employer wishes to set aside a fundamental principle of
Community law in relation to some people because it does not care for their life style.
41.
Whether the private conceptions of morality held by the employer in question correspond to those
prevalent in the United Kingdom or not must be irrelevant in this connection. Under the Treaty it is the rule
of law in the Community that the court must safeguard; it is not its task to watch over questions of morality
either in the individual member states or in the Community, nor does it have any practical possibility of or
political mandate for doing so. If a choice should have to be made in the Community between various
views of morality that must be a task for the Communitys political institutions, and hence it is for the
legislature to make such choices by way of treaty or Community legislation.
42.
There is nothing in either the Treaty on European Union (Maastricht, 7 February 1992; TS 12 (1994);
Cmnd 2485) or the EC Treaty to indicate that the rights and duties which result from the EC Treaty,
including the right not to be discriminated against on the basis of gender, should not apply to
homosexuals, to the handicapped, to persons of a particular ethnic origin or to persons holding particular
religious views. Equality before the law is a fundamental principle in every community governed by the
rule of law and accordingly in the Community as well. The rights and duties which result from Community
law apply to all without discrimination and therefore also to the approximately 35m citizens of the
Community, depending on the method of calculation used, who are homosexual (see in this respect the
judgment in P v S [1996] All ER (EC) 397, [1996] ECR I-2143 (para 22)).
43.
To summarise, I consider that the answer to that part of the questions referred should therefore be that
the said gender discrimination cannot be justified by reference to the fact that the employers intention is
to confer benefits on heterosexual couples as opposed to homosexual couples.
202

Direct applicability
44.
By its fourth question, the national court wishes to ascertain whether art 119 of the Treaty is directly
applicant and therefore, in a case such as this, may be enforced directly before national courts.
45.
The court held in the judgment in Defrenne [1981] 1 All ER 122, [1976] ECR 455 (paras 1822) that art
119 of the Treaty is directly applicable if direct discrimination is involved, in other words discrimination that
can be ascertained by applying the criteria laid down in the provision. As an example, the court mentions
discrimination which has its origin in legislative provisions or in collective labour agreements, as well as
discrimination in relation to pay, operating in one and the same private or public undertaking or service,
inasmuch as such discrimination can be detected on the basis of a purely legal analysis of the situation.
46.
In this case discrimination against employees whose cohabitees are of the same sex in relation to
employees whose cohabitees are of the opposite sex can be detected in one and the same undertaking
and one and the same set of regulations, namely SWT and cl 8 of the ticket regulations. The precondition
for art 119 having direct effect must, therefore, be satisfied.
47.
The national court is accordingly required to ensure that the disadvantaged group is treated in the
same way as the favoured group2, and thereby to ensure that employees of SWT who have a cohabitee
of the same gender are put on the same footing as employees who have a cohabitee of the opposite
gender and accordingly receive travel concessions for their cohabitee.

2
See the judgment in Coloroll [1995] All ER (EC) 23, [1994] ECR I-4389 (para 32), Van Cant v Rijksdienst voor
Pensioenen Case C-154/92 [1993] ECR I-3811 (para 22) and McDermott v Minister for Social Welfare Case 286/85
[1987] ECR 1453 (para 19).

48.
The answer to that question must therefore, in my opinion, be that art 119 of the Treaty is directly
applicable and it is for the national courts to ensure that the disadvantaged group of employees is treated
in the same way as the favoured group.

Temporal effect
49.
 In its observations the United Kingdom requested the court to limit the temporal effect of the judgment
if the questions referred to it are answered in the affirmative. The United Kingdom did not repeat that
request at the hearing and has not produced evidence to show that in this case there is a need for
temporal effect to be restricted. In my view a judgment that followed my opinion would not depart from the
courts case law hitherto and would also be founded on its facts. I do not see any reason for laying down
any temporal restriction on the effect of the judgment.

Conclusion
50.
In the light of the foregoing, I would suggest that the Court of Justice reply to the questions referred by
the Industrial Tribunal, Southampton, as follows:
(1) A provision in an employers pay regulations under which the employee is granted a pay
benefit in the form of travel concessions for a cohabitee of the opposite gender to the employee,
but refused such concessions for a cohabitee of the same gender as the employee, constitutes
discrimination on the basis of gender, which is contrary to art 119 of the EC Treaty.
203
(2) Such discrimination on the basis of gender cannot be justified by reference to the fact that
the employers intention is to confer benefits on heterosexual couples as opposed to homosexual
couples.
(3) Article 119 is directly applicable and it is for the national courts to ensure that the
disadvantaged group of employees is treated in the same way as the favoured group.

17 February 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By decision of 19 July 1996, received at the Court of Justice of the European Communities on 22 July
1996, the Industrial Tribunal, Southampton, referred to the Court of Justice for a preliminary ruling under
art 177 of the EC Treaty six questions on the interpretation of art 119 of that Treaty, Council Directive
(EEC) 75/117 on the approximation of the laws of the member states relating to the application of the
principle of equal pay for men and women (OJ 1975 L45 p 19), and Council Directive (EEC) 76/207 on
the implementation of the principle of equal treatment for men and women as regards access to
employment, vocational training and promotion, and working conditions (OJ 1976 L39 p 40).
2.
Those questions were raised in proceedings between Ms Grant and her employer South-West Trains
Ltd (SWT) concerning the refusal by SWT of travel concessions for Ms Grants female partner.
3.
Ms Grant is employed by SWT, a company which operates railways in the Southampton region.
4.
Clause 18 of her contract of employment, entitled Travel facilities, states:
You will be granted such free and reduced rate travel concessions as are applicable to a
member of your grade. Your spouse and depend[a]nts will also be granted travel concessions.
Travel concessions are granted at the discretion of [the employer] and will be withdrawn in the
event of their misuse.
5.
At the material time, the regulations adopted by the employer for the application of those provisions,
the Staff Travel Facilities Privilege Ticket Regulations, provided in cl 8 (Spouses) that:
Privilege tickets are granted to a married member of staff for one legal spouse but not for a
spouse legally separated from the employee Privilege tickets are granted for one common law
opposite sex spouse of staff subject to a statutory declaration being made that a meaningful
relationship has existed for a period of two years or more.
6.
The regulations also defined the conditions under which travel concessions could be granted to
current employees (cls 1 to 4), employees having provisionally or definitively ceased working (cls 5 to 7),
surviving spouses of employees (cl 9), children of employees (cls 10 and 11) and dependent members of
employees families (cl 12).
7.
On the basis of those provisions Ms Grant applied on 9 January 1995 for travel concessions for her
female partner, with whom she declared she had had a meaningful relationship for over two years.
8.
SWT refused to allow the benefit sought, on the ground that for unmarried persons travel concessions
could be granted only for a partner of the opposite sex.
9.
 Ms Grant thereupon made an application against SWT to the Industrial Tribunal, Southampton,
arguing that that refusal constituted discrimination 204 based on sex, contrary to the Equal Pay Act 1970,
art 119 of the EC Treaty and/or Directive 76/207. She submitted in particular that her predecessor in the
post, a man who had declared that he had had a meaningful relationship with a woman for over two
years, had enjoyed the benefit which had been refused her.
10.
The Industrial Tribunal considered that the problem facing it was whether refusal of the benefit at issue
on the ground of the employees sexual orientation was discrimination based on sex within the meaning
of art 119 of the Treaty and the directives on equal treatment of men and women. It observed that while
some UK courts had held that that was not the case, the judgment of the Court of Justice in P v S Case
C-13/94 [1996] All ER (EC) 397, [1996] ECR I-2143 was, on the other hand, persuasive authority for the
proposition that discrimination on the ground of sexual orientation [was] unlawful.
11.
For those reasons the industrial tribunal referred the following questions to the Court of Justice for a
preliminary ruling:
(1) Is it (subject to (6) below) contrary to the principle of equal pay for men and women
established by Article 119 of the Treaty establishing the European Community and by Article 1 of
Council Directive 75/117 for an employee to be refused travel concessions for an unmarried
cohabiting same-sex partner where such concessions are available for spouses or unmarried
opposite-sex cohabiting partners of such an employee?
(2) For the purposes of Article 119 does discrimination based on sex include discrimination
based on the employees sexual orientation?
(3) For the purposes of Article 119, does discrimination based on sex include discrimination
based on the sex of that employees partner?
(4) If the answer to Question (1) is yes, does an employee, to whom such concessions are
refused, enjoy a directly enforceable Community right against his employer?
(5) Is such a refusal contrary to the provisions of Council Directive 76/207?
(6) Is it open to an employer to justify such refusal if he can show (a) that the purpose of the
concessions in question is to confer benefits on married partners or partners in an equivalent
position to married partners and (b) that relationships between same-sex cohabiting partners have
not traditionally been, and are not generally, regarded by society as equivalent to marriage; rather
than on the basis of an economic or organisational reason relating to the employment in question?
12.
In view of the close links between the questions, they should be considered together.
13.
As a preliminary point, it should be observed that the court has already held that travel concessions
granted by an employer to former employees, their spouses or dependants, in respect of their
employment are pay within the meaning of art 119 of the Treaty (see, to that effect, the judgment in
Garland v British Rail Engineering Ltd Case 12/81 [1982] 2 All ER 402, [1982] ECR 359 (para 9)).
14.
In the present case it is common ground that a travel concession granted by an employer, on the basis
of the contract of employment, to the employees spouse or the person of the opposite sex with whom the
employee has a stable relationship outside marriage falls within art 119 of the Treaty. Such a benefit is
therefore not covered by Directive 76/207, referred to in the national tribunals 205 Question 5 (see the
judgment in Gillespie v Northern Health and Social Services Board Case C-342/93 [1996] All ER (EC)
284, [1996] ECR I-475 (para 24)).
15.
In view of the wording of the other questions and the grounds of the decision making the reference,
the essential point raised by the national tribunal is whether an employers refusal to grant travel
concessions to the person of the same sex with whom an employee has a stable relationship constitutes
discrimination prohibited by art 119 of the Treaty and Directive 75/117, where such concessions are
granted to an employees spouse or the person of the opposite sex with whom an employee has a stable
relationship outside marriage.
16.
Ms Grant submits, first, that such a refusal constitutes discrimination directly based on sex. She
submits that her employers decision would have been different if the benefits in issue in the main
proceedings had been claimed by a man living with a woman, and not by a woman living with a woman.
17.
Ms Grant argues that the mere fact that the male worker who previously occupied her post had
obtained travel concessions for his female partner, without being married to her, is enough to identify
direct discrimination based on sex. In her submission, if a female worker does not receive the same
benefits as a male worker, all other things being equal, she is the victim of discrimination based on sex
(the but for test).
18.
Ms Grant contends, next, that such a refusal constitutes discrimination based on sexual orientation,
which is included in the concept of discrimination based on sex in art 119 of the Treaty. In her opinion,
differences in treatment based on sexual orientation originate in prejudices regarding the sexual and
emotional behaviour of persons of a particular sex, and are in fact based on those persons sex. She
submits that such an interpretation follows from the judgment in P v S and corresponds both to the
resolutions and recommendations adopted by the Community institutions and to the development of
international human rights standards and national rules on equal treatment.
19.
Ms Grant claims, finally, that the refusal to allow her the benefit is not objectively justified.
20.
SWT and the United Kingdom and French governments consider that the refusal of a benefit such as
that in issue in the main proceedings is not contrary to art 119 of the Treaty. They submit, first, that the
judgment in P v S, which is limited to cases of gender reassignment, does no more than treat
discrimination based on a persons change of sex as equivalent to discrimination based on a persons
belonging to a particular sex.
21.
They submit, next, that the difference in treatment of which Ms Grant complains is based not on her
sexual orientation or preference but on the fact that she does not satisfy the conditions laid down in the
undertakings regulations.
22.
Finally, in their opinion, discrimination based on sexual orientation is not discrimination based on sex
within the meaning of art 119 of the Treaty or Directive 75/117. They refer on this point in particular to the
wording and objectives of art 119, the lack of consensus among member states as to whether stable
relationships between persons of the same sex may be regarded as equivalent to stable relationships
between persons of opposite sex, the fact that those relationships are not protected by arts 8 or 12 of the
Convention for the Protection of Human Rights and Fundamental Freedoms (the European Human Rights
Convention) (Rome, 4 November 1950; TS 71 (1953); Cmd 8969), and the consequent absence of
discrimination within the meaning of art 14 of that convention.
206
23.
The European Commission likewise considers that the refusal of the benefits to Ms Grant is not
contrary to art 119 of the Treaty or Directive 75/117. In its opinion, discrimination based on the sexual
orientation of workers may be regarded as discrimination based on sex for the purposes of art 119. It
submits, however, that the discrimination of which Ms Grant complains is based not on her sexual
orientation but on the fact that she is not living as a couple or with a spouse, as those terms are
understood in the laws of most of the member states, in Community law and in the law of the European
Human Rights Convention. It considers that in those circumstances the difference of treatment applied by
the regulations in force in the undertaking in which Ms Grant works is not contrary to art 119.
24.
In the light of all the material in the case, the first question to answer is whether a condition in the
regulations of an undertaking such as that in issue in the main proceedings constitutes discrimination
based directly on the sex of the worker. If it does not, the next point to examine will be whether
Community law requires that stable relationships between two persons of the same sex should be
regarded by all employers as equivalent to marriages or stable relationships outside marriage between
two persons of opposite sex. Finally, it will have to be considered whether discrimination based on sexual
orientation constitutes discrimination based on the sex of the worker.
25.
First, it should be observed that the regulations of the undertaking in which Ms Grant works provide for
travel concessions for the worker, for the workers spouse, that is, the person to whom he or she is
married and from whom he or she is not legally separated, or the person of the opposite sex with whom
he or she has had a meaningful relationship for at least two years, and for the children, dependent
members of the family, and surviving spouse of the worker.
26.
The refusal to allow Ms Grant the concessions is based on the fact that she does not satisfy the
conditions prescribed in those regulations, more particularly on the fact that she does not live with a
spouse or a person of the opposite sex with whom she has had a meaningful relationship for at least
two years.
27.
That condition, the effect of which is that the worker must live in a stable relationship with a person of
the opposite sex in order to benefit from the travel concessions, is, like the other alternative conditions
prescribed in the undertakings regulations, applied regardless of the sex of the worker concerned. Thus
travel concessions are refused to a male worker if he is living with a person of the same sex, just as they
are to a female worker if she is living with a person of the same sex.
28.
Since the condition imposed by the undertakings regulations applies in the same way to female and
male workers, it cannot be regarded as constituting discrimination directly based on sex.
29.
Second, the court must consider whether, with respect to the application of a condition such as that in
issue in the main proceedings, persons who have a stable relationship with a partner of the same sex are
in the same situation as those who are married or have a stable relationship outside marriage with a
partner of the opposite sex.
30.
Ms Grant submits in particular that the laws of the member states, as well as those of the Community
and other international organisations, increasingly treat the two situations as equivalent.
31.
While the European Parliament, as Ms Grant observes, has indeed declared that it deplores all forms
of discrimination based on an individuals 207 sexual orientation, it is nevertheless the case that the
Community has not as yet adopted rules providing for such equivalence.
32.
As for the laws of the member states, while in some of them cohabitation by two persons of the same
sex is treated as equivalent to marriage, although not completely, in most of them it is treated as
equivalent to a stable heterosexual relationship outside marriage only with respect to a limited number of
rights, or else is not recognised in any particular way.
33.
The European Commission on Human Rights for its part considers that despite the modern evolution
of attitudes towards homosexuality, stable homosexual relationships do not fall within the scope of the
right to respect for family life under art 8 of the European Human Rights Convention (see especially the
decisions in X v UK (1983) 32 D & R 220, S v UK (1986) 47 D & R 274 (para 2) and Kerkhoven v the
Netherlands App No 15666/89 (19 May 1992, unreported) (para 1)), and that national provisions which,
for the purpose of protecting the family, accord more favourable treatment to married persons and
persons of opposite sex living together as man and wife than to persons of the same sex in a stable
relationship are not contrary to art 14 of the convention, which prohibits inter alia discrimination on the
ground of sex (see the decisions in S v UK (1986) 47 D & R 274 (para 7); C v UK App No 14753/89 (9
October 1989, unreported) (para 2) and B v UK (1990) 64 D & R 278 (para 2)).
34.
In another context, the European Court of Human Rights has interpreted art 12 of the the European
Human Rights Convention as applying only to the traditional marriage between two persons of opposite
biological sex (see the judgments in Rees v UK (1986) 9 EHRR 56 (para 49), and Cossey v UK (1990) 13
EHRR 622 (para 43).
35.
It follows that, in the present state of the law within the Community, stable relationships between two
persons of the same sex are not regarded as equivalent to marriages or stable relationships outside
marriage between persons of opposite sex. Consequently, an employer is not required by Community law
to treat the situation of a person who has a stable relationship with a partner of the same sex as
equivalent to that of a person who is married to or has a stable relationship outside marriage with a
partner of the opposite sex.
36.
In those circumstances, it is for the legislature alone to adopt, if appropriate, measures which may
affect that position.
37.
Finally, Ms Grant submits that it follows from P v S that differences of treatment based on sexual
orientation are included in the discrimination based on sex prohibited by art 119 of the Treaty.
38.
In P v S the court was asked whether a dismissal based on the change of sex of the worker concerned
was to be regarded as discrimination on grounds of sex within the meaning of Directive 76/207.
39.
The national court was uncertain whether the scope of that directive was wider than that of the Sex
Discrimination Act 1975, which it had to apply and which in its view applied only to discrimination based
on the workers belonging to one or other of the sexes.
40.
In their observations to the court the UK government and the Commission submitted that the directive
prohibited only discrimination based on the fact that the worker concerned belonged to one sex or the
other, not discrimination based on the workers gender reassignment.
41.
In reply to that argument, the court stated that the provisions of the directive prohibiting discrimination
between men and women were simply the expression, in their limited field of application, of the principle
of equality, which 208 is one of the fundamental principles of Community law. It considered that that
circumstance argued against a restrictive interpretation of the scope of those provisions and in favour of
applying them to discrimination based on the workers gender reassignment.
42.
The court considered that such discrimination was in fact based, essentially if not exclusively, on the
sex of the person concerned. That reasoning, which leads to the conclusion that such discrimination is to
be prohibited just as is discrimination based on the fact that a person belongs to a particular sex, is limited
to the case of a workers gender reassignment and does not therefore apply to differences of treatment
based on a persons sexual orientation.
43.
Ms Grant submits, however, that, like certain provisions of national law or of international conventions,
the Community provisions on equal treatment of men and women should be interpreted as covering
discrimination based on sexual orientation. She refers in particular to the International Covenant on Civil
and Political Rights (New York, 16 December 1966; TS 6 (1977); Cmnd 6702), in which, in the view of the
Human Rights Committee established under art 28 of the covenant, the term sex is to be taken as
including sexual orientation (see Toonen v Australia Communication [1994] 1-3 IHRR 97).
44.
The covenant is one of the international instruments relating to the protection of human rights of which
the court takes account in applying the fundamental principles of Community law (see eg the judgments in
Orkem v EC Commission Case 374/87 [1989] ECR 3283 (para 31) and Dzodzi v Belgium Joined cases
C-297/88 and C-197/89 [1990] ECR I-3763 (para 68)).
45.
However, although respect for the fundamental rights which form an integral part of those general
principles of law is a condition of the legality of Community acts, those rights cannot in themselves have
the effect of extending the scope of the Treaty provisions beyond the competence of the Community (see,
inter alia, on the scope of art 235 of the EC Treaty as regards respect for human rights, Opinion 2/94
[1996] ECR I-1759 (paras 34 and 35)).
46.
Furthermore, in the communication referred to by Ms Grant, the Human Rights Committee, which is
not a judicial institution and whose findings have no binding force in law, confined itself, as it stated itself
without giving specific reasons, to noting that in its view the reference to sex in Articles 2, paragraph
1, and 26 is to be taken as including sexual orientation.
47.
Such an observation, which does not in any event appear to reflect the interpretation so far generally
accepted of the concept of discrimination based on sex which appears in various international instruments
concerning the protection of fundamental rights, cannot in any case constitute a basis for the court to
extend the scope of art 119 of the Treaty. That being so, the scope of that article, as of any provision of
Community law, is to be determined only by having regard to its wording and purpose, its place in the
scheme of the Treaty and its legal context. It follows from the considerations set out above that
Community law as it stands at present does not cover discrimination based on sexual orientation, such as
that in issue in the main proceedings.
48.
It should be observed, however, that the Treaty of Amsterdam (OJ 1997 C340) provides for the
insertion in the EC Treaty of an art 6a which, once the Treaty of Amsterdam has entered into force, will
allow the EU Council under certain conditions (a unanimous vote on a proposal from the Commission
after consulting the European Parliament) to take appropriate action to eliminate various forms of
discrimination, including discrimination based on sexual orientation.
209
49.
Finally, in the light of the foregoing, there is no need to consider Ms Grants argument that a refusal
such as that which she encountered is not objectively justified.
50.
Accordingly, the answer to the national tribunal must be that the refusal by an employer to allow travel
concessions to the person of the same sex with whom a worker has a stable relationship, where such
concessions are allowed to a workers spouse or to the person of the opposite sex with whom a worker
has a stable relationship outside marriage, does not constitute discrimination prohibited by art 119 of the
Treaty or Directive 75/117.

Costs
51.
The costs incurred by the United Kingdom and French governments and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the action pending before the national
tribunal, the decision on costs is a matter for that tribunal.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Industrial
Tribunal, Southampton, by decision of 19 July 1996, hereby rules: the refusal by an employer to allow
travel concessions to the person of the same sex with whom a worker has a stable relationship, where
such concessions are allowed to a workers spouse or to the person of the opposite sex with whom a
worker has a stable relationship outside marriage, does not constitute discrimination prohibited by art 119
of the EC Treaty or Council Directive (EEC) 75/117 on the approximation of the laws of the member states
relating to the application of the principle of equal pay for men and women.

210

[1998] All ER (EC) 211

Blasi v Finanzamt Mnchen I

(Case C-346/95)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES GULMANN (PRESIDENT OF THE CHAMBER), WATHELET, MOITINHO DE ALMEIDA,
EDWARD AND PUISSOCHET
ADVOCATE GENERAL JACOBS
5 JUNE, 25 SEPTEMBER 1997, 12 FEBRUARY 1998
European Community Value added tax Exemptions Leasing or letting of immovable property
Exclusion from exemption of accommodation provided in hotel sector and sectors with a similar function
German law excluding short-term accommodation from exemption Criterion of sixth months duration
used to determine whether short-term Whether criterion appropriate Whether letting agreement
determinative of duration Council Directive (EEC) 77/388, art 13B(b)(1).

The plaintiff let accommodation to refugees referred to her by the municipal social services department,
which issued her with certificates attesting that it would pay the costs of that accommodation. The
certificates were usually valid for one month but could be extended if necessary. In most cases the
average actual duration of a refugees stay was greater than six months; however, the letting agreements
concluded with the refugees were always signed for less than six months, notwithstanding the plaintiffs
stated wish to let accomodation only to those who would stay for at least six months. Under German law
traders letting short-term (defined as less than six months) living and sleeping accommodation to guests
were liable to value added tax on those transactions and that the criterion of short-term accommodation
referred not to the actual length of stay but to the traders intention as expressed in a letting agreement.
On receiving an assessment to tax in respect of the refugees lettings, the plaintiff brought proceedings
before the Finanzgericht, Mnchen contending that she was not liable. That court stayed the proceedings
and referred to the Court of Justice of the European Communities for a preliminary ruling the question
whether short-term accommodation, as defined by the German legislation, constituted the provision of
accommodation in the hotel sector or in sectors with a similar function for the purposes of art 13B(b)(1) 1
of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to
turnover taxescommon system of value added tax: uniform basis of assessment, which provided that
such accommodation was not exempt from value added tax.
1
Article 13B(b)(1), so far as material, is set out at p 219 j, post

Held The phrase sectors with a similar function in art 13B(b)(1) had to be given a broad construction,
since its purpose was to ensure that the provision of temporary accommodation which was similar to and
hence in potential competition with that provided in the hotel sector was subject to tax. Member states
enjoyed a margin of discretion in defining the provision of accommodation to be taxed by virtue of art
13B(b)(1), and it was for them to introduce appropriate criteria to distinguish between the taxable
provision of accommodation in the hotel and similar sectors and the exempted transactions of leasing and
letting of immovable property. The criterion of duration was an appropriate one for the 211 purpose of
making the necessary distinction and the use of a criterion of less than six months duration was a
reasonable means of ensuring that the transactions of taxable persons whose business was similar to the
essential function of a hotel were subject to tax. Moreover, the German law whereby qualification for
exemption depended on proving an intention, evidenced by a letting agreement, to let for a minimum of
six months, provided a criterion that was easy to apply and appropriate to ensure a correct and
straightforward application of the exemptions for which art 13B provided. Accordingly, art 13B(b)(1) could
be construed as meaning that the provision of short-term accommodation for guests was taxable, as
constituting the provision of accommodation in sectors with a function similar to that of the hotel sector. In
that regard, art 13B(b)(1) did not preclude taxation in respect of agreements concluded for a period of
less than six months, if that duration was deemed to reflect the parties intention. However, it was for the
national court to determine whether, in a case before it, certain factors, such as the automatic renewal of
the letting agreement, suggested that the duration stated in the letting agreement did not reflect the
parties true intention, in which case the actual total duration of the accommodation, rather than that
specified in the letting agreement, would have to be taken into consideration (see p 221 b to p 222 c e to
g, post).

Notes
For exemptions within the territory of a country, see 52 Halsburys Laws (4th edn) para 2028.

Cases cited
Bulthuis-Griffioen v Inspector der Omzetbelasting Case C-453/93 [1995] ECR I-2341, [1995] STC 954,
ECJ.
Skatteministeriet v Henriksen Case 173/88 [1989] ECR 2763, [1990] STC 768, ECJ.
Stichting Uitvoering Financile Acties v Staatssecretaris van Financile Case 348/87 [1989] ECR 1737.

Reference
By order of 20 September 1995 the Finanzgericht Mnchen (the Finance Court, Munich) referred to the
Court of Justice of the European Communities for a preliminary ruling three questions (set out at p 220 c
to e, post) concerning the interpretation of art 13B(b)(1) of Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment. Those questions arose in a dispute between Mrs Blasi and the
Finanzamt Mnchen I (Munich I tax office) concerning Mrs Blasis liability to pay value added tax in
respect of the provision of accommodation treated under German law as being short-term because the
letting agreements were entered into for a period of less than six months. Written observations were
submitted on behalf of: the German government, by E Rder, Ministerialrat in the Federal Ministry of
Economic Affairs, and B Kloke, Oberregierungsrat in that ministry, acting as agents; and the European
Commission by J Grunwald, Legal Adviser, acting as agent. Oral observations were submitted by: Mrs
Blasi, represented by H Weindl, Rechtsanwalt, Munich; the German government, represented by E Rder
and C Quassowski, Regierungsdirektor in the Federal Ministry of Economic Affairs, acting as agent; and
the Commission, represented by J Grunwald. The language of the case was German. The facts are set
out in the opinion of the Advocate General.

212
25 September 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
The present case, referred to the Court of Justice of the European Communities by the Finanzgericht
Mnchen (the Finance Court of Munich), raises the question whether art 13B(b) of Council Directive
(EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxescommon
system of value added tax: uniform basis of assessment (the Sixth Directive) (OJ 1977 L145 p 1) permits
a member state to impose value added tax (VAT) on the provision of temporary accommodation for
asylum seekers and emigrants.

Relevant Community and national provisions

2.
In so far as is relevant art 13B of the Sixth Directive provides:
Without prejudice to other Community provisions, Member States shall exempt the following
under conditions which they shall lay down for the purpose of ensuring the correct and
straightforward application of the exemptions and of preventing any possible evasion, avoidance or
abuse (b) the leasing or letting of immovable property excluding: 1. the provisions of
accommodation, as defined in the laws of the Member States, in the hotel sector or in sectors with
a similar function, including the provision of accommodation in holiday camps or on sites developed
for use as camping sites Member States may apply further exclusions to the scope of this
exemption.
3.
It is clear from the ruling in Skatteministeriet v Henriksen Case 173/88 [1989] ECR 2763, [1990] STC
768 (para 21), that, while a member state may restrict the scope of the exemption laid down in art 13B(b)
by applying further exclusions, it may not exempt from tax the transactions which are excluded from
exemption. The transactions mentioned in art 13B(b)(1) are therefore compulsorily taxable.
4.
The relevant provisions of German law are to be found in the Umsatzsteuergesetz 1980 (the Law on
Turnover Tax) (the UStG). Under para 4(12), first sentence, letter (a), of the UStG, as interpreted by the
German courts, the leasing of immovable property, including individual parts of a property such as living
and sleeping rooms, is exempt from tax. However, under para 4(12), second sentence, of the UStG the
letting of living and sleeping accommodation which a trader keeps available for the short-term
accommodation of guests is taxable. The referring court states that, according to the case law of the
Bundesfinanzhof (the Federal Finance Court), the criterion of short-term accommodation refers not to the
actual length of stay but to the traders intention. If a trader lets rooms to a public authority which uses
them for the accommodation of third parties, what is decisive is whether the trader intends that the rooms
be used for longer term letting and has implemented that intention by entering into a long-term agreement
with the public authority. According to the case law of the Bundesfinanzhof, the practice of the tax
authorities and academic opinion, a letting is to be regarded as short-term if it is for a period of less than
six months.

The facts and the national courts questions

5.
Since 1984 Mrs Blasi has provided accommodation in three buildings situated in Munich for emigrant
families from Eastern European countries. The rooms in the buildings are fully furnished and equipped
with cooking facilities 213 and fridges. They are cleaned by the emigrants themselves, who in some
cases install telephones at their own expense.
6.
Every 14 days Mrs Blasi provides fresh bedding and cleans the landings, staircases, baths and
lavatories. The emigrants are not supplied with meals or drinks. There are no lounges or common rooms
in any of the buildings, nor is there a hotel reception.
7.
The emigrants are put in contact with Mrs Blasi by the City of Munich, which at the material time paid
her DM 25 per day for each guest. It appears that Mrs Blasi informed the City of Munich that she was not
interested in taking emigrants who wanted to stay for short periods only and that, according to Mrs Blasis
calculations, the average length of stay is 144 months. However, there is no contractual provision
requiring the lettings to be of a minimum length.
8.
Having received an assessment to VAT in respect of the lettings for the year 1984 Mrs Blasi brought
proceedings before the referring court, claiming that she had no liability to VAT. In its order for reference,
the referring court raises doubts as to whether para 4(12), second sentence, of the UStG is compatible
with art 13B(b)(1) of the directive. It notes that, whereas the directive excludes from the exemption
applicable to the leasing or letting of immovable property, the provision of accommodation in the hotel
sector or in sectors with a similar function, the UStG refers only to the short-term provision of
accommodation. The referring court has therefore put the following questions to this court:
(1) Is Article 13B(b)(1) of [the Sixth Directive] to be interpreted as meaning that the provision of
accommodation in the hotel sector or in sectors with a similar function means no more than the
short-term accommodation of guests?
(2) If question 1 is answered in the affirmative. (a) What period of accommodation can properly
be regarded as short-term? Is it no longer provision of accommodation in the hotel sector if the
operator keeps the rooms ready for long-term accommodation and this finds expression in the
conclusion of a long-term letting agreement (longer than six months)? (b) Is a tax exemption under
Article 13B(b)(1) for a proportion of the time possible if it transpires that all the accommodation can
be let on a short or long-term basis according to choice?
(3) If question 1 is answered in the negative. On the basis of what temporal, spatial and
conceptual criteria must the phrase provision of accommodation in the hotel sector or in sectors
with a similar function be defined and which of them must necessarily be present?

Appraisal or the issues

9.
The referring courts questions, which may be considered together, are posited on the basis that, if art
13B(b)(1) is inapplicable to Mrs Blasis services, there would be no basis in the directive for taxing them.
In particular, the German provision could not be based on the last sentence of art 13B(b), which allows
member states to apply further exclusions to the scope of this exemption. That finding is based on the
courts statement in Henriksen that the last sentence of art 13B(b) merely permits a member state to add
to the exclusions from exemption and not to restrict such exclusions; the referring court concludes that
Germany cannot rely on that provision in order to restrict the exclusion from exemption in art 13B(b)(1) to
short stays.
214
10.
However, that reasoning is plainly erroneous. It may be that, by failing to tax long stays in the hotel
and similar sectors, the German provision is narrower than art 13B(b)(1). If that were so, the provision
would exempt from taxation transactions that are compulsorily taxable under the directive, contrary to the
courts ruling in Henriksen.
11.
However, that has no relevance to the present case. The issue here is not whether Germany may
exempt Mrs Blasis services but whether it may tax them. In so far as her services were not compulsorily
taxable under art 13B(b)(1), the German provision, by taxing them, would be adding to, rather than
limiting, the exclusions from exemption.
12.
The last sentence of art 13B(b) is broadly worded so as to allow the member states a large degree of
discretion in placing limits on the scope of the exemption in art 13B(b). As the court stated in Henriksen
[1989] ECR 2763, [1990] STC 768 (para 21) Member States are free to limit the scope of the exemption
by providing for additional exclusions. Unlike exemptions, which generally fall to be construed narrowly
because they constitute exceptions to the general principle that turnover tax is levied on all supplies for
consideration made by a taxable person, the exclusion of transactions from exemption is in conformity
with that general principle (see eg Stichting Uitvoering Financile Acties v Staatssecretaris van Financile
Case 348/87 [1989] ECR 1737 (para 13)). I see no Community interest in seeking to interpret narrowly
the discretion granted to member states by that provision to bring further transactions within the scope of
the charge to tax.
13.
It might therefore have been possible for Germany to justify the provision on the basis of the last
sentence of art 13B(b). It argues, however, that the provision can be based on art 13B(b)(1). I shall
therefore turn to the referring courts questions, which concern that provision.
14.
By its first and second questions the referring court asks in effect whether the short-term provision of
accommodation is a sufficient criterion for the purposes of art 13B(b)(1), and, if so, what period is to be
regarded as short-term.
15.
 Under the directive the supply and leasing of immovable property are in principle exempt from VAT
(see art 13B(b), (g) and (h)). Those exemptions reflect the particular difficulties in applying VAT to such
goods. Unlike ordinary goods, land is not the result of a production process; moreover, buildings, once
constructed, may change hands many times during their life, often without being subject to further
economic activity2. Under the Sixth Directive the charge to VAT is therefore limited in principle to the
supply of building land or of new buildings and the land on which they stand 3. The preparation of land for
development entails economic activity enhancing the value of the land; and the supply of a new building
marks the end of a production process. Thereafter repeated taxation of immovable property each time it is
sold would not be justified. The same applies to the letting of such property, which is normally a 215
comparatively passive activity not entailing significant added value; although an economic activity for the
purposes of art 4 of the Sixth Directive, the letting of immovable property is therefore in principle exempt
from tax4.
2
Supplies consisting in the alteration or repair of a building are in any event separately taxable as supplies of services
3
Member states are permitted to grant taxable persons the right to opt to tax the supply or letting of immovable property
under art 13C of the directive. The option is provided for with commercial property in mind. A vendor or lessor may
prefer taxation of the supply or letting of commercial property to a taxable person who uses the property for the
purposes of his economic activity and therefore has the right to deduct the tax charged. The vendor or lessor will then
himself have the right to deduct any VAT incurred on the purchase, leasing, alteration or refurbishment of the property.
The incurring of irrecoverable VAT by taxable persons in relation to the property is thereby avoided.
4
It may be noted however that it was thought necessary to include an express provision in art 4(2) to make it clear that
the exploitation of tangible property for the purpose of obtaining income therefrom on a continuing basis shall also be
considered an economic activity within the meaning of the Sixth Directive.

16.
However, while generally exempting the leasing or letting of immovable property, art 13B(b) also
provides for exclusion of certain transactions from exemption. The common feature of those transactions
is that they entail more active exploitation of the immovable property justifying further taxation in addition
to that levied upon its initial sale.
17.
With more particular reference to art 13B(b)(1), it may be noted, first, that its terms, in particular the
phrases accommodation, as defined in the laws of the Member States and sectors with a similar
function, are somewhat imprecise. It seems to me that the intention was to leave the member states
some latitude in defining the precise limits of the exclusion.
18.
Secondly, as already noted, art 13B(b)(1) lays down an exclusion from the exemption and therefore
does not fall to be construed strictly. Indeed it seems to me that the words sectors with a similar function
should be given a broad construction since their purpose is to ensure that the provision of temporary
accommodation similar to, and hence in potential competition with, that provided in the hotel sector is
subject to tax.
19.
As regards the German provision, it is true that the short-term letting of residential property may not
entail all of the additional supplies of goods and services, such as provision of meals and drinks, cleaning
of rooms, provision of bed linen etc, normally provided in hotels. Nevertheless, there can be no doubt that
a taxable person offering, for example, short-term holiday lets of residential property fulfils essentially the
same function asand is in a competitive relationship witha taxable person in the hotel sector. The
essential distinction between such lettings and exempt lettings of residential property is the temporary
nature of the accommodation. In any event, short-term lets are more likely to involve additional services
such as provision of linen and cleaning of common parts of buildings or even of the accommodation itself
(indeed a number of such services are provided by Mrs Blasi); moreover, they involve more active
exploitation of the property than long-term lets in so far as greater supervision and management is
required.
20.
Against that background it seems to me that, althoughunlike art 13B(b)(1)the German provision
does not expressly focus on the nature of the establishment providing the accommodation or the sector in
which it is provided, the criterion of the provision of short-term accommodation to guests which it employs
represents a reasonable means of achieving the underlying aim of the provision. It ensures that the
transactions of taxable persons whose business is similar to the essential function performed by a hotel,
namely the provision of temporary accommodation on a commercial basis, are subject to tax. While it may
be true that it is not unknown for persons to stay for long periods in hotels, any inadequacies of the
German provision in that respect are not material to the present case and may in any event be of minimal
practical significance.
216
21.
Moreover, it seems to me that the requirement flowing from the case law of the Bundesfinanzhof that,
in order for the letting of an immovable property to qualify for exemption, there must be an intention,
evidenced by a lease or other agreement, to let the property for a minimum period of six months is not
unreasonable. It provides a workable and legally certain means of distinguishing between short-term
accommodation similar to that provided in the hotel sector and the longer term letting of residential
property for which the directive provides exemption. A hotel or hostel will be willing to accept guests for
potentially short stays, whereas a landlord interested in more passive longer term lets will require an
agreement providing confirmation of the tenants intention to stay for a longer period. I see no reason to
interpret the directive as imposing a maximum of three months as the European Commission suggests.
22.
It may be noted that the opening words of art 13B require member states to lay down conditions for
ensuring the correct and straightforward application of the exemptions and of preventing any possible
evasion, avoidance or abuse. Germany is in my view entitled to consider that the other criteria suggested
by Mrs Blasi, such as whether the accommodation is the centre of interests of the persons concerned or
whether additional services are provided, would be too uncertain and difficult to apply. For example, a
residential caravan on a camping site in Spain might be regarded as the centre of interests of a retired
person who sells his house and goes to live there throughout the year; it would be difficult for the camping
site owner to apply such a criterion. Moreover, the level of services and facilities provided by hotels,
hostels and camping sites varies considerably. There are hotels which offer no more than a room and
camping sites which provide little more than a camping field. Moreover, the German government might
reasonably consider that such criteria would be less likely than one based on the period of stay to achieve
the aim of competitive neutrality.
23.
In so far as Mrs Blasis genuine intention is to provide longer-term lets of residential property it would
be open to her to enter into an agreement to that effect. If, on the other hand, the inherently temporary
nature of the stays prevents her from doing so, then it is not unreasonable that the tax authorities should
take the view that what is at issue is short-term commercial exploitation of immovable property and
equate the accommodation which she provides with the taxable accommodation provided by a hostel or
cheaper hotel. Indeed it is conceivable that the City of Munich might equally use such establishments for
the temporary accommodation of asylum seekers and emigrants.
24.
With reference to question 2(b) put by the national court, I do not think the German authorities are
obliged to grant exemption in respect of part of the period of a letting where a letting happens to exceed
six months. The German authorities are in my view justified in considering that a letting is of a long-term
nature and qualifies for exemption only if there is evidence of an intention from the beginning that
accommodation is to be provided for a period of at least six months.
25.
Finally, as I have already explained, in so far as the German rule was considered to go beyond the
terms of art 13B(b)(1) by taxing accommodation such as that provided by Mrs Blasi, it could in any event
be based on the last sentence of art 13B(b).

Conclusion
26.
Accordingly, I am of the opinion that the questions referred by the Finanzgericht Mnchen should be
answered as follows:
217
A national rule subjecting to VAT the provision of short-term accommodation to guests, that is to
say, the provision of accommodation otherwise than under an agreement providing for a minimum
stay of six months, is compatible with art 13B(b)(1) of Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of
value added tax: uniform basis of assessment.

12 February 1998.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By order of 20 September 1995, received at the Court of Justice of the European Communities on 9
November 1995, the Finanzgericht Mnchen (the Finance Court, Munich) referred for a preliminary ruling
under art 177 of the EC Treaty three questions concerning the interpretation of art 13B(b)(1) of Council
Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxes
common system of value added tax: uniform basis of assessment (the Sixth Directive) (OJ 1977 L145 p
1).
2.
Those questions arose in a dispute between Mrs Blasi and the Finanzamt Mnchen I (the Munich I tax
office) concerning Mrs Blasis liability to pay value added tax (VAT) in respect of the provision of
accommodation treated under German law as being short-term because the letting agreements were
entered into for periods of less than six months.
3.
According to the order for reference, the provisions of German legislation applicable to the facts in the
main proceedings which correspond to art 13B(b)(1) of the Sixth Directive are contained in the
Umsatzsteuergesetz 1980 (the Law on Turnover Tax) (the UStG).
4.
Under para 4(12), first sentence, letter (a), and second sentence, of the UStG:
Among the transactions referred to in Paragraph 1(1)(1) to (3), the following are exempted
12. (a) leasing and letting of immovable property, rights governed by provisions of civil law relating
to immovable property, and sovereign rights of the State in regard to immovable goods and
 property (c) The letting of living and sleeping accommodation which a trader keeps available
for the short-term accommodation of guests (Fremden) shall not be exempted
5.
Mrs Blasi let a number of buildings in Munich and, from 1984, provided accommodation in those
buildings for refugee families from countries of central and eastern Europe sent to her by the municipal
social services department.
6.
It appears from the order for reference that the buildings used for accommodation were normal
residential buildings each containing several dwellings. Fully furnished rooms equipped with cooking
facilities were made available to the families. The rooms were cleaned by the refugees themselves. Mrs
Blasi supplied and washed the bedlinen and also saw to the cleaning of the landings, staircases,
bathrooms and lavatories. Occupants were not supplied with meals. There was no reception area in the
buildings, nor any lounges or other common amenity rooms.
7.
The City of Munich did not have any right to allocate accommodation, but acted as an intermediary
and sent persons to Mrs Blasi when it learned that she had dwellings available.
8.
For the accommodation costs, which were met by the City of Munich, the municipal social services
department issued Mrs Blasi with certificates attesting 218 that it would pay the costs incurred, which
were usually valid for one month and could be extended if necessary. Payment was made on the basis of
a daily charge, fixed by agreement between Mrs Blasi and the City of Munich. That charge was DM 25
per person per day in the year at issue, 1984.
9.
Letting agreements were concluded on an individual basis between Mrs Blasi and the tenants sent by
the City of Munich social services department. In most cases, the actual duration of the refugees stay
was greater than six months. It appears from the order for reference that, in a number of letters sent to
the City of Munich, Mrs Blasi had refused to accept refugees who wished to stay with her for only a short
period and expressed her desire to let only to those who wished to stay with her for at least six months.
However, the letting agreements concluded between Mrs Blasi and her tenants were always signed for a
period of less than six months.
10.
By a tax assessment notice of 7 April 1987, Mrs Blasi was assessed on the basis of her tax return and
VAT was determined in the amount of DM 82,043. Pursuant to the findings of an audit report of 28
December 1990, the Finanzamt, by an amending decision of 13 July 1993, fixed the VAT at DM 151,278,
on the ground that Mrs Blasi had failed to declare all her taxable transactions for the period from 1
January 1984 to 30 June 1984.
11.
After unsuccessfully lodging an objection against that decision, Mrs Blasi instituted proceedings before
the Finanzgericht Mnchen.
12.
Mrs Blasi argued that the tax assessment notice, the amending decision and the decision dismissing
her objection should be amended and the claim against her for VAT for 1984 consequently set aside. The
Finanzamt submitted that the action should be dismissed.
13.
The Finanzgericht Mnchen first of all points out that, according to the case law of the
Bundesfinanzhof (the Federal Finance Court) on para 4(12) of the UStG, the provision of hotel
accommodation will be treated as taxable if the operators intention is to make the premises available for
temporary accommodation only. The actual duration of the let is of little importance, since, according to
well-established case law, where accommodation provided for emigrants or asylum-seekers is
guaranteed by public authorities, consideration must be given to the duration provided for under the
letting agreement. The let is deemed to be short-term if the agreement is for a period of less than six
months.
14.
The Finanzgericht Mnchen points out that, if those criteria are applied in the case before it, it would
appear that the transactions carried out by Mrs Blasi could not be exempted from VAT because, even if
she intended the dwellings in question to be for long-term lets, no long-term agreement was ever
concluded. Only in retrospect is it possible to determine that the letting of dwellings through the
intermediary of the City of Munich related to several years, the actual average length of a let being,
according to Mrs Blasis calculations, 144 months.
15.
The Finanzgericht Mnchen also points out that the exception to the exemption introduced by the
second sentence of para 4(12) of the UStG is based on art 13B(b)(1) of the Sixth Directive, which
provides as follows:
Without prejudice to other Community provisions, Member States shall exempt the following
under conditions which they shall lay down for the purpose of ensuring the correct and
straightforward application of the exemptions and of preventing any possible evasion, avoidance or
abuse (b) the leasing or letting of immovable property excluding: (1) the provisions of
accommodation, as defined in the laws of the Member States, in the hotel sector or in sectors with
 a similar function, including the provision of 219 accommodation in holiday camps or on sites
developed for use as camping sites
16.
Since art 13B(b)(1) of the Sixth Directive merely states that the provision of accommodation in sectors
with a function similar to that of the hotel sector is excluded from the exemption enjoyed by the leasing or
letting of immovable property, without referring specifically to the provision of short-term accommodation,
the Finanzgericht Mnchen is uncertain whether the second sentence of para 4(12) of the UStG is
compatible with that exclusion. It has accordingly referred the following questions to the Court of Justice
for a preliminary ruling:
(1) Is Article 13B(b)(1) of [the Sixth Directive] to be interpreted as meaning that the provision of
accommodation in the hotel sector or in sectors with a similar function means no more than the
short-term accommodation of guests?
(2) If question 1 is answered in the affirmative. (a) What period of accommodation can properly
be regarded as short-term? Is it no longer provision of accommodation in the hotel sector if the
operator keeps rooms available for long-term accommodation and this finds expression in the
conclusion of a long-term letting agreement (longer than six months)? (b) Is a tax exemption under
Article 13B(b)(1) for a proportion of the time possible if it transpires that all the accommodation was
let on either a short or a long-term basis?
(3) If question 1 is answered in the negative. On the basis of what temporal, spatial and
conceptual criteria must the phrase provision of accommodation in the hotel sector or in sectors
with a similar function be defined and which of them must necessarily be present?
17.
In its questions, which should be examined together, the national court is in substance asking whether
art 13B(b)(1) of the Sixth Directive may be construed as meaning that what is defined in German law as
the provision of short-term accommodation for guests constitutes, within the meaning of Community law,
the provision of accommodation in sectors with a function similar to that of the hotel sector, thus being
subject to VAT. The national court is also asking whether it is compatible with art 13B(b)(1) of the Sixth
Directive to draw a distinction between taxable transactions and transactions that are exempted on the
basis of the duration of the accommodation, such exemption being reserved for those letting transactions
that involve the conclusion of a letting agreement for more than six months, irrespective of the actual total
duration of the let.
18.
It must first be noted that the court has consistently held that the terms used to specify the exemptions
provided for by art 13 of the Sixth Directive are to be interpreted strictly, since they constitute exceptions
to the general principle that turnover tax is to be levied on all services supplied for consideration by a
taxable person (see the judgments in Stichting Uitvoering Financile Acties v Staatssecretaris van
Financile Case 348/87 [1989] ECR 1737 (para 13) and Bulthuis-Griffioen v Inspector der
Omzetbelasting Case C-453/93 [1995] ECR I-2341, [1995] STC 954 (para 19)).
19.
The phrase excluding the provision of accommodation, as defined in the laws of the Member
States, in the hotel sector or in sectors with a similar function in art 13B(b)(1) of the Sixth Directive
introduces an exception to the exemption which art 13B provides for the leasing or letting of immovable
property. It thus subjects the transactions to which it refers to the general rule 220 laid down in the
directive, namely that VAT is to be charged on all taxable transactions, except in the case of derogations
expressly provided for. That phrase cannot therefore be interpreted strictly.
20.
It should be added that, as Advocate General Jacobs has noted at para 18 of his opinion, above, the
words sectors with a similar function should be given a broad construction since their purpose is to
ensure that the provision of temporary accommodation similar to, and hence in potential competition with,
that provided in the hotel sector is subject to tax.
21.
In defining the classes of provision of accommodation which are to be taxed by derogation from the
exemption for the leasing or letting of immovable property, in accordance with art 13B(b)(1) of the Sixth
Directive, the member states enjoy a margin of discretion. That discretion is circumscribed by the purpose
of the derogation, which, in regard to making dwelling accommodation available, is that thetaxable
provision of accommodation in the hotel sector or in sectors with a similar function must be distinguished
from the exempted transactions of leasing and letting of immovable property.
22.
It is consequently a matter for the member states, when transposing art 13B(b)(1) of the Sixth
Directive, to introduce those criteria which seem to them appropriate in order to draw that distinction.
23.
Where accommodation in the hotel sector (as a taxable transaction) is distinguished from the letting of
dwelling accommodation (as an exempted transaction) on the basis of its duration, that constitutes an
appropriate criterion of distinction, since one of the ways in which hotel accommodation specifically differs
from the letting of dwelling accommodation is the duration of the stay. In general, a stay in a hotel tends to
be rather short and that in a rented flat fairly long.
24.
 In this connection, as Advocate General Jacobs has stated at para 20 of his opinion, above, the use of
the criterion of the provision of short-term accommodation, being defined as less than six months,
appears to be a reasonable means by which to ensure that the transactions of taxable persons whose
business is similar to the essential function performed by a hotel, namely the provision of temporary
accommodation on a commercial basis, are subject to tax.
25.
Furthermore, in regard to the definition of the concept of short-term in German law, the requirement
derived from the case law of the Bundesfinanzhof, to the effect that, in order to qualify for the exemption,
it is necessary to prove the intention, evidenced by a letting agreement or other contract, to let property
for a minimum period of six months, appears to be a criterion that is easily applied and appropriate to
attain the objective sought by art 13B of the Sixth Directive, which is to ensure a correct and
straightforward application of the exemptions for which it provides.
26.
It must, however, be recognised that, in certain circumstances, which will probably be exceptional, it is
possible that some clauses in a letting agreement, including that relating to duration, will not fully reflect
the reality of the contractual relations, the taxable person being, for instance, unable to fix the duration of
the let freely with his tenants where it depends on certificates from the public authorities attesting that
they will pay the costs incurred. In such circumstances, it is ultimately for the national court to determine
whether, for those reasons, reflected in particular in the automatic renewal of letting agreements, it might
not be appropriate to take into consideration the actual total duration of the accommodation rather than
that specified in the letting agreement.
221
27.
In the light of the foregoing, the answer to be given to the national court must be that art 13B(b)(1) of
the Sixth Directive may be construed as meaning that the provision of short-term accommodation for
guests is taxable, as constituting the provision of accommodation in sectors with a function similar to that
of the hotel sector. In that regard, art 13B(b)(1) does not preclude taxation in respect of agreements
concluded for a period of less than six months, if that duration is deemed to reflect the parties intention. It
is, however, for the national court to determine whether, in a case before it, certain factors (such as the
automatic renewal of the letting agreement) suggest that the duration stated in the letting agreement does
not reflect the parties true intention, in which case the actual total duration of the accommodation, rather
than that specified in the letting agreement, would have to be taken into consideration.

Costs
28.
The costs incurred by the German government and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the action pending before the national court, the decision on
costs is a matter for that court.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the questions referred to it by the
Finanzgericht Mnchen by order of 20 September 1995, hereby rules: art 13B(b)(1) of Council Directive
(EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxescommon
system of value added tax: uniform basis of assessment, may be construed as meaning that the provision
of short-term accommodation for guests is taxable, as constituting the provision of accommodation in
sectors with a function similar to that of the hotel sector. In that regard, art 13B(b)(1) does not preclude
taxation in respect of agreements concluded for a period of less than six months, if that duration is
deemed to reflect the parties intention. It is, however, for the national court to determine whether, in a
case before it, certain factors (such as the automatic renewal of the letting agreement) suggest that the
duration stated in the letting agreement does not reflect the parties true intention, in which case the
actual total duration of the accommodation, rather than that specified in the letting agreement, would have
to be taken into consideration.

222

[1998] All ER (EC) 223

Belgium v Ghent Coal Terminal NV

(Case C-37/95)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SECOND CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE SIXTH CHAMBER, ACTING AS PRESIDENT OF THE
SECOND CHAMBER) MANCINI AND HIRSCH
ADVOCATE GENERAL RUIZ-JARABO COLOMER
11 JULY 1996, 15 JANUARY 1998
European Community Value added tax Input tax Deductibility Supplies intended but not ultimately
used for taxable transactions for reasons beyond taxable persons control Whether right to deduct
arising on supplies Whether right could be withdrawn retroactively by reason of non-use for taxable
transactions Council Directive (EEC) 77/388, art 17(2).

The taxpayer company carried out investment work on land owned by it in the harbour area of Ghent and
deducted the value added tax paid on the goods and services relating to that work for the tax period from
January 1981 to December 1983. In March 1983 the taxpayer was required by the City of Ghent to
exchange that land for other land in the harbour area and consequently never used the land in respect of
which it had carried out the investment work giving rise to the tax deduction. Although it was not disputed
that the investment would have been used in taxable transactions but for the exchange and that the
exchange had neither been foreseen nor planned in advance by the taxpayer, the tax authorities sought
repayment of the tax deducted on the grounds that the taxpayer had not ultimately used the land in
question for that purpose. In proceedings in which the taxpayer resisted the repayment demand, the
Belgian Court of Cassation stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling the questions: (i) whether art 17 1 of Council Directive (EEC) 77/388
(on the harmonisation of the laws of the member states relating to turnover taxescommon system of
value added tax: uniform basis of assessment) allowed a taxable person to deduct value added tax which
he was liable to pay on goods and services supplied to him for the purposes of investment work intended
to be used in connection with taxable transactions; and (ii) if so, whether the right to deduct was retained
where, due to circumstances beyond his control, the taxable person had never made use of that
investment work in order to carry out taxable transactions.
1
Article 17, so far as material, is set out at p 234 g, post

Held The right of deduction granted to taxable persons under art 17 of Directive 77/388 had to be
exercised immediately in respect of all the taxes charged on transactions relating to inputs. On a proper
construction of art 17, therefore, a taxable person acting as such was entitled to deduct the VAT payable
or paid for goods or services supplied to him for the purpose of investment work intended to be used in
connection with taxable transactions. Moreover, in the absence of fraud or abuse, the right to deduct,
once it had arisen, remained acquired even if the planned economic activity did not give rise to taxable
transactions. When circumstances beyond the control of the taxable person had prevented him from 223
using the goods or services giving rise to deduction for the needs of his taxable transactions, there was
no risk of fraud or abuse justifying subsequent repayment (see p 236 e j to p 237 d g h, post).
Intercommunale voor Zeewaterontzilting (INZO) (in liq) v Belgium Case C-110/94 [1996] ECR I-857
applied.
Per curiam. It is possible under the provisions of the directive for the original deduction to be adjusted
where circumstances which affect the factors to be taken into account when calculating the original
deduction subsequently arise (see p 237 c, post).

Notes
For the right to deduct, see 52 Halsburys Laws (4th edn) para 2033.

Cases cited
BP Supergas Anonimos Etaira Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece Case C-62/93
[1995] All ER (EC) 684, [1995] ECR I-1883, ECJ.
EC Commission v France Case 50/87 [1988] ECR 4797.
Intercommunale voor Zeewaterontzilting (INZO) (in liq) v Belgium Case C-110/94 [1996] ECR I-857,
[1996] STC 569, ECJ.
Lennartz v Finanzamt Mnchen III Case C-97/90 [1991] ECR I-3795, [1995] STC 514, ECJ.
Rompelman v Minister van Financin Case 268/83 [1985] ECR 655.

Reference
By decision of 10 February 1995, the Belgian Hof van Cassatie (the Court of Cassation) referred for a
preliminary ruling a question (set out at p 236 b, post) concerning the interpretation of art 17 of Council
Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxes
common system of value added tax: uniform basis of assessment. That question arose in a dispute
between the Belgian State and Ghent Coal Terminal NV concerning payment of an amount of value
added tax which it deducted in respect of certain investment work which it had carried out. Written
observations were submitted on behalf of: Belgium, by J Devadder, General Adviser in the Legal Service
of the Ministry of Foreign Affairs, External Trade and Development Co-operation, acting as agent,
assisted by I C Bouaert, Advocate with right of audience before the Belgian Hof van Cassatie, and B van
de Walle de Ghelcke, of the Brussels Bar; Ghent Coal Terminal NV, by P Van Ommeslaghe, Advocate
with right of audience before the Belgian Hof van Cassatie; the German government, by E Rder,
Ministerialrat in the Federal Ministry of Economic Affairs, and G Thiele, Assessor in that ministry, acting as
agents; the Greek government, by M Apessos, Deputy Legal Adviser in the State Legal Council, Maria
Basdeki, Agent for Legal Proceedings in the State Legal Council, and A Rokofyllou, Special Adviser to the
Deputy Minister for Foreign Affairs, acting as agents; and the European Commission, by B J Drijber, of its
legal service, acting as agent. Oral observations were submitted by: the Belgian State, represented by B
van de Walle de Ghelcke; Ghent Coal Terminal NV, represented by M Lebbe, of the Brussels Bar; the
Greek government, represented by M Apessos and A Rokofyllou; and the Commission, represented by B
J Drijber. The language of the case was Dutch. The facts are set out in the opinion of the Advocate
General.

224
11 July 1996.

The Advocate General (Ruiz-Jarabo Colomer)

delivered the following opinion (translated from the Spanish).
1.
The Hof van Cassatie, Belgium, has referred to the Court of Justice of the European Communities for
a preliminary ruling a question on the interpretation of art 17(2) of Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment (OJ 1977 L145 p 1) (the Sixth Directive).
2.
The purpose of the reference to the Court of Justice is to determine the content and scope of the right
to deduct the value added tax (VAT) borne by an industrial undertaking in connection with certain
investments made for the development of land which was not, in the final event, used for the purpose
initially envisaged.
3.
The question referred to the Court of Justice is worded as follows:
Does Article 17 of [the Sixth Directive] mean that the right to deduct remains in existence for
value added tax on investments which were originally intended for use in the undertaking but
which, for reasons beyond its control, were never in fact put into use by the undertaking?

Facts and procedure in the main proceedings

4.
The order for reference merely sets out, very succinctly, only the three following facts as relevant to
the dispute: (a) in 1980 Ghent Coal Terminal NV (Ghent Coal) bought land in the harbour area of Ghent;
(b) the undertaking carried out investment work in respect of that land and immediately deducted the VAT
in its return for the period 1 January 1981 to 31 December 1983; (c) on the initiative of the City of Ghent,
Ghent Coal exchanged the purchased ground on 1 March 1983 and, as a result of that exchange, never
used the investment work which it had carried out.
5.
The order for reference acknowledges that it is common ground between the parties thatas found in
the judgment appealed against
the invested goods had been in the normal course of events intended for use in taxable
transactions, that the exchange had not been foreseen or planned in advance by the respondent,
and that it could not have been avoided by the respondent in the normal course of its business and
even constituted economic force majeure for it.
6.
The various documents submitted by the parties in the main proceedings also reveal other facts and
the progress of the procedure before the national courts, which are of interest for a better understanding
of the dispute and which I shall now briefly describe.
7.
In 1980 Ghent Coal, which had decided to extend its port installations, purchased various plots of land
at Imsakkerlaan, alongside Ghent tanker quay, to build a coal terminal and a coal packing plant.
8.
Ghent Coal commenced the necessary work for this construction, to which end it made certain
investments for, in particular, a preliminary survey carried out by a construction undertaking, another
survey of the physical characteristics of the ground, and the installation of a high-voltage cable and
operations to level the land. The cost of that work came to more than BF 50m.
9.
On 1 March 1983 the municipal authorities of Ghent required Ghent Coal to exchange the above-
mentioned land, which was already partly developed, for 225 other land belonging to the town. Ghent
Coal also received additional compensation.
10.
Ghent Coal, which during the 1981, 1982 and 1983 tax years had deducted BF 9,354,677 as VAT paid
in respect of the expenditure incurred for the development of the land, was required by the Belgian tax
authorities to repay the amount of those deductions to which, in the authorities opinion, it was not
entitled.
11.
In September 1984 Ghent Coal and the tax authorities concluded an agreement or arrangement under
which Ghent Coal undertook to pay the sum of BF 9,379,000 by way of VAT (plus interest and a fine).
Payment was effected, in the form of a set-off against other credits in favour of Ghent Coal, on 31 January
1985.
12.
Ghent Coal subsequently formed the view that both the repayment and the agreement concluded with
the tax authorities had been unlawful. Consequently, on 10 March 1986 it claimed repayment from the
Belgian authorities of BF 2,751,085, the amount which in its opinion it was entitled to deduct for the
investment expenditure after the corresponding adjustment had been made.
13.
When the tax authorities refused to accede, Ghent Coal claimed that amount before the Rechtbank
van Eerste Aanleg (the Court of First Instance), Ghent, which, by judgment of 4 April 1990, dismissed its
claim on the ground that the parties were validly bound by the agreement concluded between them in
September 1984.
14.
Ghent Coal appealed against the judgment at first instance to the Hof van Beroep (the Court of
Appeal), Ghent, which set it aside in a judgment of 26 October 1992 on the ground that the agreement
concluded was not valid in law, since it concerned tax debts which could only be determined by applying
the statutory rules. The appeal court took the view that the right of deduction was properly exercised and
that Ghent Coals claim should be granted. It therefore ordered the tax authorities to pay Ghent Coal the
amount of BF 2,751,085.
15.
On 23 February 1993 the Belgian State sought to have the judgment of the Hof van Beroep set aside
on a point of law. In the course of the proceedings the Hof van Cassatie (the Court of Cassation) decided
to refer the question to this court for a preliminary ruling.

Applicable Community legislation

16.
As a tax on supplies of goods or services, VAT seeks to be a general tax on consumption exactly
proportional to the price of the goods and services, whatever the number of transactions which take place
in the production and distribution process before the stage when the tax is charged.
17.
On each transaction, VAT is chargeable at the applicable rate on the taxable amount (the price of the
taxable goods or services), after deduction of the amount of VAT borne directly by the various cost
components.
18.
The principle of deduction thus allows the taxable person to deduct from the VAT borne by the
transactions which he has carried out the VAT which he paid when he acquired goods or received
services in connection with the pursuit of an economic activity. The rules governing the deduction
mechanism are laid down in Title XI (arts 17 to 20) of the Sixth Directive 2.
2
The original rules were initially laid down by the Council Directive (EEC) 67/228 on the harmonisation of legislation of
member states concerning turnover taxesstructure and procedures for application of the common system of value
added tax (OJ S edn 1967 p 16), in particular art 11.

19.
Article 17(1) of the Sixth Directive provides: The right to deduct shall arise at the time when the
deductible tax becomes chargeable.
226
20.
Article 10(2) provides: The chargeable event shall occur and the tax shall become chargeable when
the goods are delivered or the services are performed.
21.
Article 17(2) provides as follows:
In so far as the goods and services are used for the purposes of his taxable transactions, the
taxable person shall be entitled to deduct from the tax which he is liable to pay: (a) value added tax
due or paid in respect of goods or services supplied or to be supplied to him by another taxable
person
22.
Article 20(1) refers to the adjustment of deductions, in the following terms:
The initial deduction shall be adjusted according to the procedures laid down by the Member
States, in particular: (a) where that deduction was higher or lower than that to which the taxable
person was entitled; (b) where after the return is made some change occurs in the factors used to
determine the amount to be deducted, in particular where purchases are cancelled or price
reductions are obtained
23.
Article 20(2) lays down special rules for the adjustment of deductions in respect of capital goods:
In the case of capital goods, adjustments shall be spread over five years including that in which
the goods were acquired or manufactured. The annual adjustment shall be made only in respect of
 one-fifth of the tax imposed on the goods. The adjustment shall be made on the basis of the
variations in the deduction entitlement in subsequent years in relation to that for the year in which
the goods were acquired or manufactured.
By way of derogation from the preceding sub-paragraph, Member States may base the
adjustment on a period of five full years starting from the time at which the goods are first used.
In the case of immovable property acquired as capital goods, the adjustment period may be
extended up to 10 years.
24.
Finally, art 20(3) provides for the situation of capital goods which have been transferred during the
period of adjustment.

The wording of the question referred to the court

25.
The order for reference hinges on a key phrase (whether the right to deduct the VAT remains in
existence), which, applied to the present case, contains a certain measure of ambiguity. By asking
whether the right to deduct remains in existence in the case of investments intended for goods which are
not subsequently used, the Hof van Cassatie seems by implication to accept that such a right had already
arisen and expresses doubt only as to its possible continuity in time 3.
3
In its various senses, the expression remains in existence implies that something or someone which or who previously
existed remains, endures, subsists, continues its or his life, in spite of any adverse circumstances which may arise or
the passing of time.

26.
The parties submissions differ, however, concerning the content of the question referred to the court.
In Ghent Coals view the way in which the question is worded by the Hof van Cassatie implies that the
right to deduct had 227 already arisen, as demonstrated by the fact that the Hof van Cassatie did not
adopt the wording suggested by the Belgian State in its appeal 4. The question therefore draws a
distinction between the origin of the right and its subsequent continued existence.
4
The suggested wording was: Does art 17 mean that the right to deduct arises and remains in existence when the
investments have not in fact been put into use by the undertaking?

27.
The Belgian government, however, takes the view that the failure to use the investment work carried
out by Ghent Coal means that the deduction must be rejected ab initio, outright and in full. In other
words, the right to deduct had never legally arisen.
28.
The problem is accentuated when it is combined with the system of review or adjustment 5 of
deductions. Adjustment is the mechanism whereby the Sixth Directive (art 20) allows subsequent
changes to be made to deductions.
5
The word normally used is adjustment, in preference to review.

29.
Article 20 provides that the initial deduction is to be adjusted where that deduction proves to be higher
or lower than that to which the taxable person was entitled or where some change subsequently occurs in
the factors used to determine the amount to be deducted.
30.
In might be thought, in principle, that the Hof van Cassatie, by asking whether the right to deduct
remains in existence, seeks to ascertain whether or not it is possible to apply the procedure for adjusting
deductions to the present case in order to amend the deductions already effected, since it appears to be
the specific procedure provided for by the Sixth Directive. Neither of the parties to the main proceedings
supports that approach, however.
31.
According to the Belgian government, rejection of the right to deduct ab initio is not to be confused
with adjustment of a deduction of VAT and, since the right to deduct did not arise in this case, it is not
correct to speak of the adjustment of such a deduction.
32.
Ghent Coal, on the other hand, maintains that the question of adjustment was not in dispute between
the parties and was not raised before the Hof van Cassatie. It therefore falls outside the scope of the
proceedings between those parties. On the basis that the original deduction was lawful, however, and that
subsequent events deprived the investments of their intended purpose, Ghent Coal suggests that the
court should answer the Hof van Cassatie by confirming that, in principle, the deduction may be adjusted
subject to the limits and conditions laid down in the Sixth Directive 6.
6
That does not prevent Ghent Coal from maintaining that, as it was a case of economic force majeure, comparable to
the destruction or loss of the goods (a situation provided for in the Sixth Directive as an exceptional case in which the
deduction is not to be adjusted), the original deduction had become definitive, which allowed Ghent Coal, inter alia, to
claim full reimbursement, in separate proceedings, of the VAT which it had deducted in respect of the investments in
question and had subsequently repaid on demand by the Belgian State.
33.
I consider that the Court of Justices answer to the Hof van Cassatie should essentially be confined to
the actual terms of the question, which do not refer directly toalthough they do not excludeproblems
of adjustment.
34.
I shall therefore analyse, first of all, the requirements necessary to give rise to the right to deduct the
business expenditure incurred with a view to setting up Ghent Coals project. That analysis will extend to
the possible effect on the right to deduct of the fact that the project initially envisaged was abandoned.
35.
Secondly, if, as seems likely at first sight, that analysis should favour the existence of the right to
deduct, I shall consider to what extent the usefulness of 228 the courts answer would be enhanced by
going on to examine the problems of the adjustment of the deductions.

The right to deduct

36.
In the dynamics of VAT, deduction, governed by art 17 of the Sixth Directive, becomes a key part of the
system. As a result of the way in which it is regulated, the VAT paid by undertakings does not entail any
fiscal burden whatsoever for them and the underlying principle of the neutrality of VAT, a tax on final
consumption and not on the earlier economic stages, is respected. If the right to deduct VAT on inputs did
not exist that VAT would become an extra fiscal cost for undertakings and distort the principle of neutrality.
37.
Deduction of VAT on inputs is possible in so far as the corresponding goods or services (those whose
acquisition or use determines the right to deduct) are acquired and used by the taxable person to carry
out, in turn, transactions which fall within the scope of his economic activity 7.
7
This statement must be qualified where the taxable person acquires and uses the goods or services for exempt
transactions, in which case the right to deduct does not arise and the taxable person becomes, so to speak, the final
consumer and is unable to deduct the VAT. In such a case the taxable person must therefore bear all the VAT which has
been passed on to him by the previous economic agents (those who have supplied the products or services) and cannot
deduct it, in strictly legal terms. He is therefore in the same position as the final consumers, the true payers of VAT.

38.
It is sufficient, then, that the goods or services are acquired and used by an undertaking within the
framework of an economic activity for the VAT paid or due to be deductible. Where art 17(2) of the Sixth
Directive speaks of goods and services used for the purposes of his taxable transactions, it seeks to
emphasise that the use must be specifically aimed at the business activity and not at other activities of a
different kind.
39.
That does not mean, however, that the purpose or objective for which the goods acquired or services
received are to be used in the normal course must always be achieved in every case. On the contrary, it
is perfectly possible that certain business transactions for the realisation of which goods or services were
acquired may subsequently be frustrated. The right to deduct the VAT paid does not cease to exist for that
reason.
40.
I believe that the terms employed by the court in its decisions on the deduction of VAT and, specifically,
in its judgments in Rompelman v Minister van Financin Case 268/83 [1985] ECR 655, Lennartz v
Finanzamt Mnchen III Case C-97/90 [1991] ECR I-3795, [1995] STC 514, and Intercommunale voor
Zeewaterontzilting (INZO) (in liq) v Belgium Case C-110/94 [1996] ECR I-857, [1996] STC 569 are
sufficient to resolve the present case. For that reason I consider it necessary to set out some of those
considerations before analysing their application to this case.
41.
The judgment in Lennartz [1991] ECR I-3795, [1995] STC 514 (para 8) begins by stating:
Pursuant to Article 17(1) of the Sixth Directive, which is entitled Origin and scope of the right to
deduct, the right to deduct arises at the time when the deductible tax becomes chargeable.
Consequently, only the capacity in which a person is acting at that time can determine the
existence of the right to deduct. By virtue of Article 17(2), in so far as a taxable person, acting as
such, uses the goods for the purposes of his taxable transactions, he is entitled to deduct the tax
due or paid in respect of those goods.
229
42.
As regards the scope of the right to deduct the VAT due, the judgment in Lennartz [1991] ECR I-3795,
[1995] STC 514 (paras 1314) refers to the judgment in Rompelman and reiterates:
(13) the economic activities referred to in Article 4(1) may consist in several consecutive
transactions, as is indeed suggested by the wording of Article 4(2). Amongst such transactions
preparatory activities, such as the acquisition of operating assets, must be treated as constituting
economic activities within the meaning of that article.
(14) a person who acquires goods for the purposes of an economic activity within the
meaning of Article 4 does so as a taxable person, even if the goods are not used immediately for
such economic activities.
43.
Consequently, according to the judgment in Lennartz [1991] ECR I-3795, [1995] STC 514 (para 15), it
is the acquisition of the goods by a taxable person acting as such that gives rise to the application of the
VAT system and therefore of the deduction mechanism. The use to which the goods are put, or intended
to be put, merely determines the extent of the initial deduction to which the taxable person is entitled
under art 17 and the extent of any adjustments in the course of the following periods.
44.
More recently, in the INZO judgment the Court of Justice answered a question referred by another
Belgian court (the Rechtbank van Eerste Aanleg, Bruges) which presents great similarities to the question
referred in these proceedings.
45.
The question for the court in INZO was whether or not an undertaking which had acquired certain
capital goods and commissioned a study on the profitability of a project for the construction of a
desalination plant (in respect of which supply of goods and services the undertaking paid VAT) could
deduct the VAT paid, despite the fact that owing to profitability problems and the withdrawal of some
investors it subsequently abandoned the project without commencing the activity envisaged 8.
8
Rather than deduction stricto sensu, the case concerned the repayment of the VAT paid, which was initially agreed by
the tax authority pursuant to art 76 of the Belgian VAT code. On subsequently finding in the course of a tax inspection
that INZO had not carried out any taxable transaction, the tax authority claimed repayment of the VAT recovered by
INZO. INZO contested that claim before the Rechtbank van Eerste Aanleg, relying on the doctrine formulated by the
Court of Justice in Rompelman.

46.
In INZO the court observed that it had held (in Rompelman) that even the first investment expenditure
incurred for the purposes of a business may be regarded as an economic activity within the meaning of
art 4 of the Sixth Directive and that, in that context, the tax authority must take into account the declared
intention of the business.
47.
Next, the judgment stated that where the tax authority had accepted that a company which had
declared its intention to begin an economic activity giving rise to taxable transactions had the status of a
taxable person for the purposes of VAT, the carrying out of a study into the profitability of the activity
envisaged may be regarded as an economic activity within the meaning of art 4 of the Sixth Directive
even if the purpose of that study is to investigate the degree of profitability of the activity concerned.
48.
In the courts view, it followed that, if the same requirements were met, VAT paid in respect of such a
profitability study might in principle be deducted in accordance with art 17 of the Sixth Directive, even if it
had subsequently been 230 decided, in view of the results of that study, not to move to the operational
phase but to put the company into liquidation, with the result that the economic activity envisaged had not
given rise to taxed transactions9.
9
The judgment based that conclusion on two principles. (a) The principle of legal certainty, according to which the rights
and obligations of taxable persons cannot depend on facts, circumstances or events which occurred after they were
recognised by the tax authority. It follows that, as from the time when the tax authority accepted, on the basis of
information provided by a business, that it should be accorded the status of a taxable person, that status cannot, in
principle, subsequently be withdrawn retroactively on account of the fact that certain events have or have not occurred.
(b) The principle that VAT should be neutral as regards the tax burden on a business. Any other interpretation of the
directive, according to the court, would be liable to create, as regards the tax treatment of the same investment
activities, unjustified differences between businesses already carrying out taxable transactions and other businesses
seeking by investment to commence activities which will in future be a source of taxable transactions. Likewise, arbitrary
differences would be established between the latter businesses, in that final acceptance of the deductions would
depend on whether or not the investment resulted in taxable transactions.

49.
In accordance with those legal principles, the court gave the following answers to the questions
referred to it in INZO [1996] ECR I-857, [1996] STC 569 (operative part):
where the tax authority has accepted that a company which has declared an intention to
commence an economic activity giving rise to taxable transactions has the status of a taxable
person for the purposes of VAT, the commissioning of a profitability study in respect of the
envisaged activity may be regarded as an economic activity within the meaning of that article, even
if the purpose of that study is to investigate to what degree the activity envisaged is profitable, and
that except in cases of fraud or abuse, the status of taxable person for the purpose of VAT may
not be withdrawn from that company retroactively where, in view of the results of that study, it has
been decided not to move to the operational phase, but to put the company into liquidation with the
result that the economic activity envisaged has not given rise to taxable transactions.

The application of those decisions to the present case

50.
Transposing the decisions in the Rompelman, Lennartz and INZO judgments to the present case,
there is little doubt that Ghent Coals argument regarding the right to deductwhich is also supported by
the European Commission and, with certain qualifications, by the German governmentis better founded
than that of the Belgian government.
51.
 If the conditions determining the right to deduct are to be assessed at the time when the tax is
payable, Ghent Coal enjoyed that right when it acquired or received, for the purpose of its business
activities, certain goods and services which all bore VAT. At that time the corresponding economic agents
(providers, or sellers in general) charged Ghent Coal VAT, which, in order to guarantee the fiscal neutrality
of VAT provided for in the Sixth Directive, it was entitled to, and which it did in fact, deduct.
52.
In other words, when Ghent Coal acquired the land at Imsakkerlaan, engaged the services of other
undertakings to carry out a development survey and carried out certain investment works in connection
with the land for the purpose of constructing a coal terminal, it was entitled to deduct from the VAT paid on
those transactions the VAT which, according to the corresponding 231 invoices, had been passed on to it
in respect of the works carried out and the services received.
53.
For the right to deduct to arise, it was irrelevant that the necessarily slow process of preparing and
developing the land for the purpose of erecting the coal plant had not been completed when the Ghent
municipal authorities required Ghent Coal to exchange its land. What matters for the purposes of VAT is
that the VAT which was paid when the goods or services were received could also be deducted during the
corresponding period.
54.
The condition that determines whether the right to deduct VAT arises is that the goods acquired and
the services received are acquired and received in connection with the business activity of the taxable
person, that is to say for the purpose of being incorporated within its economic activity.
55.
In accordance with the principles expressed in the judgments cited, Ghent Coal could deduct the VAT
paid on purchasing the goods or paying for the services connected with the construction of the coal plant,
in so far as: (a) it is not necessary that the goods and services acquired in the course of operations
preparatory to carrying out an activity be immediately used for transactions subject to VAT (Rompelman),
when there is no doubt whatsoever as to Ghent Coals purpose in acquiring those goods and services,
which are directly linked to the pursuit of its business activity; (b) strictly speaking, it is not even necessary
that those goods and services be used to carry out subsequent taxable transactions where they have
been acquired in the course of the stages prior to the performance of an envisaged activity which
subsequently and for lawful reasons does not reach the operational phase (INZO); (c) in this case all
suspicion of fraud or abuse is precluded, since Ghent Coal was in fact unavoidably obliged to abandon
the construction project which it had begun by the requirements of a public administration acting in the
exercise of its functions.
56.
The conclusion from all the foregoing is that, in accordance with art 17(2) of the Sixth Directive, an
undertaking like Ghent Coal is entitled to deduct the VAT paid in connection with the acquisition of goods
and the receipt of services corresponding to investment works initially intended to be used for its business
activity but which, for subsequent reasons beyond its control, were never in fact put into use.

Possible adjustment of the deductions

57.
In paras 28 to 32 of this opinion, above, I set out the arguments of both parties to the dispute
regarding the treatment of the possible adjustment of the deductions, a question which, in Ghent Coals
view, is extraneous to this dispute in so far as it was neither submitted to the Hof van Cassatie nor raised
by that court in its reference for a preliminary ruling.
58.
It is true that the question referred simply seeks the interpretation of art 17(2) of the Sixth Directive,
without referring to art 20, which lays down the system of adjustments to initial deductions.
59.
However, the mechanism of co-operation established by art 177 of the EC Treaty allows the Court of
Justice to provide the national court with the matters relating to the interpretation of the rules of
Community law which it considers applicable to the case, even where the national court has not referred
expressly to any of them.
60.
It would be difficult to accept that adjustment is extraneous to the problem raised in the main
proceedings: in fact Ghent Coals request to the Belgian authorities for repayment of BF 2,751,085, and
also its claim of 27 March 1987 232before the court of first instance, quantify that figure as the amount
owed in respect of the investment expenditure effected, once the corresponding adjustment has been
made10.
10
The application at first instance, in addition to confirming the origin of the right to deduct, further states: in any event,
that deduction is subject to adjustment in pursuance of Article 48 of the VAT Code, which provides that the deduction
can be reviewed where there have been variations in the factors taken into consideration for the calculation of the VAT
deductible, and Article 10.4 of Royal Decree No 3.

61.
Nor can it be said that the problem is extraneous to the appeal to the Hof van Cassatie: indeed,
counsel for the Belgian State claimed that the appeal court had erred in law by stating, inter alia, that the
right to deduct had arisen and that the only available means of correcting the deduction was adjustment 11.
 11
The judgment of the appeal court stated: in so far as it is apparent that after the return has been made a change has
occurred in the factors used to determine the amount to be deducted, as happened in the present case, since as a
result of the exchange the purpose normally attributed to the goods in question could not be achieved, the only possible
way of correcting it is by adjustment, as provided for in Article 48 of the VAT Code and Articles 6 and 10 of Royal Decree
No 3

62.
Moreover, in their observations to the court both the Commission and Ghent Coal suggest answers to
the questions referred which expressly include a confirmation of the possibility of adjustment 12.
12
The Commission maintains that the right to deduct the VAT paid in respect of those investments originally intended to be
used in the undertaking remains in existence even where the undertaking, for reasons beyond its control, has
subsequently been unable to use them; it adds thatin any event, it is appropriate to adjust the deductions, to the
extent and subject to the conditions provided for by Article 20(3) of the Sixth Directive, in the ease of tax-exempt
deliveries of capital goods during the adjustment period. Ghent Coal suggests that the court should answer the question
referred to it by stating, first, that the deduction in respect of the investments intended for a business activity aimed at
taxable activities is valid. In its view the courts answer should further state thatwhere it transpires that such
investments subsequently became devoid of purpose and, consequently, were never actually used in the undertaking,
there should then, in principle, be an adjustment within the limits and subject to the conditions determined by the Sixth
Directive. The fact that the investments have become devoid of purpose and consequently have never actually been
used, for reasons beyond the control of the undertaking, cannot affect the lawfulness of the deduction already made,
except that, at the very most, it may be possible to adjust the deduction.

63.
That, in my view, is the more reasonable position. If the courts answer were simply to confirm the
applicability of the deduction, without further distinction, it might lead to confusion, since it would only
address part of the problem (the validity of the original deduction) but not the associated problem.
64.
The answer should therefore contain an express reference to the possibility of adjusting the original
deduction owing to the existence of subsequent circumstances which affected the factors taken into
account to establish that deduction.
65.
On that point, however, I do not believe that the courts answer should go much beyond a reference to
art 20 of the Sixth Directive. I do not in fact believe that the court should become involved in the argument
concerning the actual extent of the adjustment (the number of years to be taken into account, the possible
extension of the delivery of the land under Belgian law, the different schemes applicable to the delivery of
capital goods and to the services received, the rules applicable where capital goods are delivered during
the adjustment period, etc).
66.
An answer going into a detailed analysis of such questions would in my view go beyond the terms in
which the courts interpretation was sought in the question referred for a preliminary ruling. In order to
remain faithful to the question raised and at the same time provide further elements based on 233
Community rules to which the national court did not refer, it is sufficient, in this case, to indicate that art 20
of the Sixth Directive lays down the procedure for the adjustment of deductions validly made.

Conclusion
67.
I therefore propose that the Court of Justice should answer the question referred by the Hof van
Cassatie as follows:
Article 17(2) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member
states relating to turnover taxescommon system of value added tax: uniform basis of
assessment, allows an undertaking to deduct the VAT paid in connection with the acquisition of
goods and the receipt of services corresponding to investment works originally intended to be used
in its business activity but which, for reasons beyond its control, were never in fact put to use by the
undertaking. The adjustment of those deductions must be effected as provided for in art 20 of that
directive.

15 January 1998.

The COURT OF JUSTICE (Second Chamber)

delivered the following judgment.
1.
By decision of 10 February 1995, received at the Court of Justice of the European Communities on 16
February 1995, the Belgian Hof van Cassatie (the Court of Cassation) referred for a preliminary ruling
under art 177 of the EC Treaty a question concerning the interpretation of art 17 of Council Directive
(EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxescommon
system of value added tax: uniform basis of assessment (OJ 1977 L145 p 1) (the Sixth Directive).
2.
That question arose in a dispute between the Belgian State and Ghent Coal Terminal NV (Ghent Coal)
concerning payment of an amount of value added tax (VAT) which Ghent Coal deducted in respect of
certain investment work which it had carried out.
3.
 Article 17 of the Sixth Directive provides as follows:
(1) The right to deduct shall arise at the time when the deductible tax becomes chargeable. (2)
In so far as the goods and services are used for the purposes of his taxable transactions, the
taxable person shall be entitled to deduct from the tax which he is liable to pay: (a) value added tax
due or paid in respect of goods or services supplied or to be supplied to him by another taxable
person; (b) value added tax due or paid in respect of imported goods
4.
The adjustment of deductions is governed by art 20, which provides as follows:
(1) The initial deduction shall be adjusted according to the procedures laid down by the Member
States, in particular: (a) where that deduction was higher or lower than that to which the taxable
person was entitled; (b) where after the return is made some change occurs in the factors used to
determine the amount to be deducted, in particular where purchases are cancelled or price
reductions are obtained; however, adjustment shall not be made in cases of transactions remaining
totally or partially unpaid and of destruction, loss or theft of property duly proved or confirmed
234
(2) In the case of capital goods, adjustments shall be spread over five years including that in
which the goods were acquired or manufactured. The annual adjustment shall be made only in
respect of one-fifth of the tax imposed on the goods. The adjustment shall be made on the basis of
the variations in the deduction entitlement in subsequent years in relation to that for the year in
which the goods were acquired or manufactured
(3) In the case of supply during the period of adjustment capital goods shall be regarded as if
they had still been applied for business use by the taxable person until expiry of the period of
adjustment. Such business activities are presumed to be fully taxed in cases where the delivery of
the said goods is taxed; they are presumed to be fully exempt where the delivery is exempt. The
adjustment shall be made only once for the whole period of adjustment still to be covered
5.
In Belgium, the supply of land is exempt from VAT.
6.
In 1980, Ghent Coal purchased land in the harbour area of Ghent. It subsequently carried out
investment work and immediately deducted the VAT paid on the goods and services relating to that work
for the period between 1 January 1981 and 31 December 1983.
7.
On 1 March 1983, on the initiative of the City of Ghent, Ghent Coal exchanged the land in question for
other land situated elsewhere in the Ghent harbour area. Consequently, it never used the land in respect
of which it had carried out the investment work giving rise to the deduction.
8.
It is not disputed that the investment was in the normal course of events to have been used in taxable
transactions, or that the exchange had been neither foreseen nor planned in advance by Ghent Coal,
which was unable to avoid it from an economic point of view and for which it even constituted a case of
economic force majeure.
9.
Following investigations carried out in 1984, the tax authorities concluded that Ghent Coal had not
used the land in question for the purpose of carrying out taxable transactions and accordingly sought
repayment of the VAT deducted in connection with the investment work carried out on the land in
question, along with payment of a fine and default interest.
10.
Ghent Coal initially accepted the view taken by the tax authorities. On 27 March 1986, however, it
brought proceedings against the Belgian State before the Rechtbank van eerste Aanleg (the Court of First
Instance), Ghent, which, by judgment of 4 April 1990, dismissed its application. However, by judgment of
26 October 1992, the Hof van Beroep (the Court of Appeal), Ghent, upheld the appeal lodged by Ghent
Coal. The Belgian State thereupon sought to have that judgment set aside.
11.
The Belgian State takes the view that, when goods or services supplied have given rise to a deduction
but have never been used for the purpose of carrying out taxable transactions, the right to deduct must be
retroactively withdrawn and the deducted VAT repaid in full.
12.
Ghent Coal argues that the right to deduct VAT due or paid in respect of goods or services originally
intended to be used for the purpose of carrying out taxable transactions is an absolute right and cannot
therefore be called into question even if the person concerned has never actually made use of those
goods or services.
13.
Since it took the view that an interpretation of art 17 of the Sixth Directive was necessary to resolve
the dispute before it, the Belgian Hof van Cassatie 235 decided to stay the proceedings and refer the
following question to the Court of Justice for a preliminary ruling:
 Does Article 17 of [the Sixth Directive] mean that the right to deduct remains in existence for
value added tax on investments which were originally intended for use in the undertaking but
which, for reasons beyond its control, were never in fact put into use by the undertaking?
14.
By its question, the national court is essentially asking whether art 17 of the Sixth Directive must be
construed as allowing a taxable person acting as such to deduct VAT which he is liable to pay on goods or
services supplied to him for the purposes of investment work intended to be used in taxable transactions
and, if so, whether the right to deduct remains acquired where, by reason of circumstances beyond his
control, the taxable person has never made use of that investment work in order to carry out taxable
transactions.
15.
With regard to the first part of this question, the court has stated repeatedly that the deduction system
is meant to relieve the trader entirely of the burden of the VAT payable or paid in the course of all his
economic activities. The common system of value added tax consequently ensures that all economic
activities, whatever their purpose or results, provided that they are themselves subject to VAT, are taxed in
a wholly neutral way (see esp the judgments in Rompelman v Minister van Financin Case 268/83 [1985]
ECR 655 (para 19) and EC Commission v France Case 50/87 [1988] ECR 4797 (para 15)).
16.
In the absence of any provision empowering the member states to limit the right of deduction granted
to taxable persons, that right must be exercised immediately in respect of all the taxes charged on
transactions relating to inputs. Such limitations on the right of deduction must be applied in a similar
manner in all the member states and therefore derogations are permitted only in the cases expressly
provided for in the Sixth Directive (see esp EC Commission v France (paras 1617), Lennartz v
Finanzamt Mnchen III Case C-97/90 [1991] ECR I-3795, [1995] STC 514 (para 27) and BP Supergas
Anonimos Etaira Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece Case C-62/93 [1995] All ER
(EC) 684, [1995] ECR I-1883 (para 18)).
17.
It follows that a taxable person acting as such is entitled to deduct the VAT payable or paid for goods
or services supplied to him for the purpose of investment work intended to be used in connection with
taxable transactions.
18.
With regard, next, to the second part of the question, it follows from the judgment in Lennartz [1991]
ECR I-3795, [1995] STC 514 (para 15) that the use to which the goods and services are put merely
determines the extent of the initial deduction to which the taxable person is entitled under art 17 and the
extent of any adjustments in the course of the following periods.
19.
Furthermore, in its judgment in Intercommunale voor Zeewaterontzilting (INZO) (in liq) v Belgium Case
C-110/94 [1996] ECR I-857, [1996] STC 569 (paras 2021), concerning the position of an undertaking
which had never effected any taxable transaction, the court ruled that the right to deduct, once it has
arisen, remains acquired even if the planned economic activity has not given rise to taxable transactions.
20.
Likewise, the right to deduct remains acquired where the taxable person has been unable to use the
goods or services which gave rise to a deduction in the context of taxable transactions by reason of
circumstances beyond his control.
21.
It also follows from the judgment in INZO [1996] ECR I-857, [1996] STC 569 (para 24), that in cases of
fraud or abuse, in which the person concerned, on 236 the pretext of intending to pursue a particular
economic activity, in fact sought to acquire as his private assets goods in respect of which a deduction
could be made, the tax authority may claim repayment of the sums retroactively on the ground that those
deductions were made on the basis of false declarations.
22.
However, when circumstances beyond the control of the taxable person have prevented him from
using the goods or services giving rise to deduction for the needs of his taxable transactions, there is no
risk of fraud or abuse capable of justifying subsequent repayment.
23.
Finally, it must be pointed out that a supply of investment goods during the adjustment period, such as
occurred in the main proceedings in this case, may give rise to an adjustment of the deduction under the
conditions set out in art 20(3) of the Sixth Directive.
24.
The answer to the question submitted must therefore be that art 17 of the Sixth Directive must be
construed as allowing a taxable person acting as such to deduct the VAT payable by him on goods or
services supplied to him for the purpose of investment work intended to be used in connection with
taxable transactions. The right to deduct remains acquired where, by reason of circumstances beyond his
control, the taxable person has never made use of those goods or services for the purpose of carrying out
taxable transactions. A supply of investment goods during the adjustment period, where such occurs, may
give rise to an adjustment of the deduction under the conditions set out in art 20(3) of the Sixth Directive.

Costs
25.
 The costs incurred by the German and Greek governments and by the European Commission, which
have submitted observations to the Court of Justice, are not recoverable. Since these proceedings are,
for the parties to the main proceedings, a step in the action pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Second Chamber), in answer to the question referred to it by
the Belgian Hof van Cassatie by decision of 10 February 1995, hereby rules: art 17 of Council Directive
(EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxescommon
system of value added tax: uniform basis of assessment, must be construed as allowing a taxable person
acting as such to deduct the VAT payable by him on goods or services supplied to him for the purpose of
investment work intended to be used in connection with taxable transactions. The right to deduct remains
acquired where, by reason of circumstances beyond his control, the taxable person has never made use
of those goods or services for the purpose of carrying out taxable transactions. A supply of investment
goods during the adjustment period, where such occurs, may give rise to an adjustment of the deduction
under the conditions set out in art 20(3) of the directive.

237

[1998] All ER (EC) 238

Saldanha and another v Hiross Holding AG

(Case C-122/96)

EUROPEAN COMMUNITY; Other European Community

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES MANCINI (PRESIDENT OF THE CHAMBER), KAPTEYN (RAPPORTEUR) AND RAGNEMALM
ADVOCATE GENERAL LA PERGOLA
20 MARCH, 6 MAY, 2 OCTOBER 1997
European Community Costs Security for costs Discrimination on grounds of nationality
Shareholder in Austrian company having British and American nationality and resident in Florida
Shareholder seeking injunction restraining Austrian company from transferring certain shares without
agreement of shareholders general meeting Shareholder ordered by Austrian court to lodge security for
costs on grounds of non-Austrian nationality Austria acceding to Community during proceedings
Temporal and legal scope of relevant EC Treaty provision EC Treaty, art 6.

The first plaintiff was a national of both the United Kingdom and the United States of America but was
resident in Florida; the second plaintiff was a company established in the United States. Both plaintiffs
sought an injunction restraining the defendant, an Austrian company in which they were both
shareholders, from selling or transferring certain shares without the approval of a general meeting of
shareholders. The defendant thereupon applied for and was granted an order requiring the plaintiffs, as
foreign nationals, to provide security for the costs of the proceedings pursuant to para 57(1) of the
Austrian Code of Civil Procedure. Austria subsequently acceded to the European Union and to the
treaties on which it was founded. On the plaintiffs appeal, the Oberlandesgericht set aside the order for
security in so far as the first plaintiff was concerned, on the grounds that he was a British national and it
would be contrary to the prohibition of discrimination on grounds of nationality set out in art 6 1 of the EC
Treaty to require him to lodge security for costs when Austrian nationals in the same situation were not
required to do so. The defendant appealed to the Oberster Gerichtshof, which stayed the proceedings
and referred to the Court of Justice of the European Communities for a preliminary ruling the question
whether art 6 precluded a member state from requiring security for costs in such circumstances. As a
preliminary point, the defendant contended that the matter at issue was outside the scope of Community
law, since the order predated Austrias accession to the EC Treaty.
1
Article 6, so far as material, is set out at p 248 c, post

Held (1) Since the Act of Accession contained no specific provisions with regard to art 6 of the EC
Treaty, that provision had to be regarded as being immediately applicable and binding on Austria from the
date of its accession. It followed that art 6 applied to the future effects of situations arising prior to
Austrias accession to the Community with effect from the date of that accession. Accordingly, from that
date, nationals of other member states could no longer be made subject to a procedural rule which
discriminated on grounds of nationality provided that that rule came within the scope of the Treaty (see p
249 d e, post).
238
(2) The fact that a national of a member state was also a national of a non-member country in which
he was resident did not deprive him of the right, as a national of that member state, to rely on the
prohibition of discrimination on grounds of nationality set out in art 6. Moreover, national rules which
sought to protect the interests of shareholders fell within the scope of the Treaty and were therefore
subject to the art 6 prohibition. It followed that art 6 precluded a member state from requiring security for
costs from a national of another member state, who was also a national of a non-member country in
which he was resident, where he had brought proceedings before its civil courts, in his capacity as a
shareholder, against a company established in that state, if such a requirement was not imposed on its
own nationals who similarly were not resident and had no assets there (see p 249 f, p 250 g and p 251 f j,
post); Micheletti v Delegacin del Gobierno en Cantabria Case C-369/90 [1992] ECR I-4239 and Cowan
v Trsor Public Case 186/87 [1989] ECR 195 applied.

Notes
For the general principle of non-discrimination, see 51 Halsburys Laws (4th edn) para 140, and for cases
on the subject, see 21(1) Digest (2nd reissue) 251, 740746.
For the EC Treaty, art 6 (formerly the EEC Treaty, art 7), see 50 Halsburys Statutes (4th edn) 267, and
for the Treaty on European Union, art G.8, amending art 7 of the EEC Treaty, see Current Service, ibid
53.

Cases cited
Amministrazione delle Finanze dello Stato v Simmenthal SpA Case 106/77 [1978] ECR 629.
Cowan v Trsor Public Case 186/87 [1989] ECR 195.
Data Delecta Aktiebolag v MSL Dynamics Ltd Case C-43/95 [1996] All ER (EC) 961, [1996] ECR I-4661,
ECJ.
EC Commission v Italy Case 104/86 [1988] ECR 1799.
EC Commission v Italy Case 168/85 [1986] ECR 2945.
EC Commission v Netherlands Case 72/85 [1986] ECR 1219.
Giloy v Hauptzollamt Frankfurt am Main-Ost Case C-130/95 (1997) Transcript, 17 July, ECJ.
Hayes v Kronenberger GmbH (in liq) Case C-323/95 [1997] ECR I-1711.
Kleinwort Benson Ltd v Glasgow City Council Case C-346/93 [1995] All ER (EC) 514, [1995] ECR I-615,
ECJ.
Leur-Bloem v Inspecteur der Belastingdienst/Ondernemingen Amsterdam 2 Cases C-28/95 [1997] All ER
(EC) 739, ECJ.
Micheletti v Delegacin del Gobierno en Cantabria Case C-369/90 [1992] ECR I-4239.
Opinion 1/92 [1992] ECR I-2821.
Opinion 2/94 [1996] ECR I-1759.

Reference
By order of 11 March 1996, the Oberster Gerichtshof referred to the Court of Justice of the European
Communities for a preliminary ruling under art 177 of the EC Treaty a question (set out at p 248 g to j,
post) on the interpretation of the first paragraph of art 6 of that Treaty. That question arose in proceedings
brought by Mr Saldanha and MTS Securities Corp against Hiross Holding AG, an Austrian company in
which they are shareholders, seeking an injunction to restrain Hiross from selling or transferring certain
shares, without the approval of the general meeting of shareholders. Written observations were submitted
on behalf of: Mr 239 Saldanha and MTS Securities Corp, by P Lambert, Rechtsanwalt, Vienna; Hiross
Holding AG, by G Zeiler, Rechtsanwalt, Vienna; the Austrian government, by F Cede, Ambassador,
Federal Ministry of Foreign Affairs, acting as agent; the UK government, by S R Ridley, of the Treasury
Solicitors Department, acting as agent; and the European Commission, by U Wlker, of its Legal Service,
acting as agent. Oral observations were made by: Mr Saldanha and MTS Securities Corp, Hiross Holding
AG and the Commission. The language of the case was German. The facts are set out in the opinion of
the Advocate General.

6 May 1997.

The Advocate General (A La Pergola)

delivered the following opinion (translated from the Italian).

I INTRODUCTION
1.
In this case, the Court of Justice of the European Communities is called upon once again to rule on
whether certain national procedural provisions are compatible with the rules of Community law, in
particular the principle of non-discrimination laid down in art 6 of the EC Treaty. The national provisions in
question are those of the Austrian legal system which, in this case, require foreign nationals bringing legal
proceedings against nationals of the country in which the action has been brought to lodge a sum of
money as security for the costs of the legal proceedings (caulio judicalum solvi).

II THE FACTS OF THE CASE

2.
 Mr Saldanha, who is resident in Florida (USA) and holds both American and British nationality, and
MTS Securities Corp, which has its registered office in the United States of America, are shareholders in
Hiross Holding AG (Hiross), a company which has its registered office in Austria.
On 27 September 1994, Mr Saldanha, along with MTS, instituted proceedings against Hiross before
the Handelsgericht (the Commercial Court), Vienna, seeking to prevent internal restructuring transactions
within the group controlled by Hiross which would have involved the transfer of shares from certain
companies to others within the same group. At the request of the defendant, the Handelsgericht ordered
the plaintiffs to lodge the security for costs provided for under para 57 of the Zivilprozeordnung (the
Austrian Code of Civil Procedure) (the ZPO), since none of the grounds of exemption set out in para 57(2)
was applicable in the case. The appellate court before which Mr Saldanha contested the decision
delivered at first instance concerning provision of the security in question held that, under art 6 of the EC
Treaty, the Austrian procedural rule in question did constitute discrimination on grounds of nationality. For
the purpose of applying that Treaty provision, the appellate court pointed out, it was immaterial that the
plaintiff had dual British and American nationality. Nor was it relevant that the person concerned resided
outside the Community: the complete equality which the Treaty established between Austrian nationals
and Community nationals had, it ruled, the effect of prohibiting in every case the security at issue, since
national law exempted Austrian nationals, even if residing outside Austria, from the obligation to provide it.
3.
The Oberster Gerichtshof (the Supreme Court), to which a further appeal was made against the above
appellate decision and before which the matter is now pending, states that, under para 57(1) of the ZPO,
foreigners who bring proceedings before an Austrian court are required to provide, for the benefit of
defendants and at the latters request, security for the legal costs, except where otherwise provided by
international treaty or convention. That rule, the national 240 court states, is designed to protect
defendants in proceedings before national courts against abusive or vexatious claims made by foreign
plaintiffs. Under para 57(2) of the ZPO, however, there is, no obligation, to provide such security where
the plaintiff is normally resident in Austria or a judicial decision ordering the plaintiff to indemnify the
defendant for his legal costs would be enforceable in the, state in which the plaintiff is normally resident.
The Oberster Gerichtshof also points out that Austrian nationals who are normally, resident or have
their domicile abroad are not required to provide the security within the meaning of para 57 of the ZPO
according to court, the rule of procedural law in question does not draw, any distinction according to
whether a plaintiff has or has not within the national territory assets capable of forming the subject of
enforcement proceedings. According to the Oberster Gerichtshof, the disputed rule has its basis, in the
desire to safeguard Austrian nationals residing abroad as well as in the desire to take account, for
reasons of reciprocity, of the similar rule provided for in this regard by the majority of European legal
systems.
4.
So far as concerns the application ratione temporis of Community-law rules, the Oberster Gerichtshof
sets out a number of observations arguing in favour of the application of the Treaty rules to the present
dispute, notwithstanding the fact that the dispute arose before Austria acceded to the Community. The
Oberster Gerichtshof notes that appellate courts, among which, in this case, the Oberster Gerichtshof
itself must be included, are obliged, in the absence of transitional provisions, to take account of binding
ius superveniens even after the decision first instance, even the facts in issue relate to a period prior to
the entry into force of that new law. The Oberster Gerichtshof argues that the provisions of the Treaty, and
in particular art 6 thereof, are of such a binding nature. The decision to refer in this case must therefore, it
argues, be based on those provisions. Also for the reasons set out above, the Oberster Gerichtshof
discounts the possibility that, for the purposes of the decision which it is required to reach, the similar
prohibition of discrimination on grounds of nationality laid, down in art 4 of the Agreement on the
European Economic Area (the EEA Agreement) (Oporto, 2 May 1992; TS26 (1995); Cm 2847; OJ 1994
L1 p 3), to which Austria has been a party since 1 January 1994, may be of relevance.
5.
In view of the Community nature of the legal problems arising in the present dispute, the Oberster
Gerichtshof in Vienna has considered it necessary to refer the following question to the Court of Justice
for a preliminary ruling:
Where proceedings are brought before an Austrian civil court by a British national who is also a
national of the United states of America, who resides in that country (in Florida) and does not have
any residence or assets in Austria, against a limited company whose registered office is in Austria,
seeking to have that company restrained from selling or otherwise transferring shares in specified
subsidiary Companies to its Italian subsidiary company, or to subsidiaries of that company with
registered offices in Italy, without the approval of a qualified majority of three-quarters of the
general meeting of shareholders or, in the alternative, of a simple majority of the general meeting of
shareholders, does the fact that he has been ordered by the competent Austrian court (of first
instance), on application by the defendant company pursuant to Paragraph 57(1) of the Austrian
Code of Civil Procedure, to provide security for costs in a specified sum constitute discrimination on
grounds of nationality contrary to the first paragraph of Article 6 of the EC Treaty?
241

III EXAMINATION OF THE CASE

A. Admissibility
6.
 The question submitted to the Court of Justice in this case raises a delicate preliminary problem
concerning the application ratione temporis of the provisions of the EC Treaty. The court is being asked to
rule on the interpretation of a Treaty provision in relation to facts which occurred at a time when the
Republic of Austria, whose procedural provisions are in dispute in this case, did not yet belong to the
Community. It is for that reason necessary to consider how the above Community provisions are dealt
with in the domestic legal system and, more generally, to examine the criterion governing the application
ratione temporis of primary Community law.
7.
A situation quite similar in several respects to that here under examination was recently considered by
the court in Data Delecta Aktiebolag v MSL Dynamics Ltd Case C-43/95 [1996] All ER (EC) 961, [1996]
ECR I-4661. The facts giving rise to that dispute before the national court had occurred before the
Kingdom of Sweden acceded to the European Communities. The court, however, addressed the merits of
the question submitted to it without having previously examined whether and how the facts in issue came
within the scope ratione temporis of Community law; in that case, the court did not provide the national
court with any interpretative criteria regarding the exact definition of the temporal context in which
Community law produces its effects and did not, therefore, make it clear whether the case in question fell
to be resolved according to Community law or according to rules derived from some other legal system.
What actually happened in the case was that, in ruling on the merits, the Court of Justice set out the
principles of law with which the national court, in casu the Swedish Supreme Court, had to comply. This
latter court, however, subsequently ruled that the dispute did not come within the scope of the EC Treaty 2.
The result was that the ruling of the Court of Justice was set aside since it had become irrelevant in
regard to the dispute before the national court; it ought therefore to be said, if one reflects on the case,
that that question could not and ought not to have been submitted for a preliminary ruling.
2
See the judgment of the Hgsta Domstol of 13 November 1996.

8.
The facts of the dispute at present before the court also occurred prior to the entry into force in Austria
of the EC Treaty. In the Data Delecta case, the national court making the reference had not specified the
reasons in law which could have justified retroactive application of the rules of Community law to the
dispute. In the present case, however, the national court making the reference has itself indicated, albeit
with a certain amount of doubt, that Community law does apply to facts dating back to a period when
Austria did not yet belong to the Community. The reason for this retroactive application of Community law,
the national court states, lies in the mandatory nature of Community law, which, as ius superveniens, is
directly applicable, under Austrian procedural law, to all cases in which final decisions had not yet been
delivered by the date on which the EC Treaty entered into force in Austria.
9.
This line of argument regarding the application ratione temporis of Community law, however, gives rise
to perplexity on a number of counts. I for my part, doubt whether Community law can in itself govern
situations arising under different legislation which does not, unless in special circumstances such as
those allowing retroactive application of more favourable penal provisions, show 242 sufficient connecting
factors with the Community rules which could have been relied on at the period in which those situations
arose. In the absence of express provisions to the contrary, the principle tempus regit actum governs the
temporal effect of Community law. Situations arising at a time when the rules in question were not yet law
in relation to the facts underlying the dispute do not therefore come within the scope ratione temporis of
the Treaty. It follows that the above-mentioned principle tempus regit actum must be uniformly applied
within the Community. Otherwise, the regulation of the temporal effects of the Treaty rules could vary, with
obvious and unjustified different treatment of similar situations depending on the national legal system
called on each particular time to guarantee the provisions of the Community legal system.
10.
Another consequence which should be pointed out follows from the foregoing considerations. A
member state may, on the basis of its own independent choice, decide that certain rules of Community
law are to be retroactively applicable to situations arising at a period prior to that states accession to the
Community, for the specific purpose of conferring on the persons concerned rights and privileges to which
they would not otherwise be entitled under the Treaty. The Community legal system does not, in my view,
prohibit national legislatures from adopting such legislation. The problem is that these would, in any
event, be internal rules, which certainly do not change their nature simply through the fact of being
adopted by reference to Community rules. It is not the Community legal system that imposes these latter
rules within the domestic sphere of the state, but rather the national system that has taken them over by
conferring on them a temporal effect which they would be lacking in their original legal system. In my
view, the Court of Justice would therefore not be competent to rule, in proceedings for a preliminary
ruling, on the provisions thus introduced into the domestic legal system, precisely because they are not
rules which come within its jurisdiction under art 177 of the Treaty 3.

3
See in this regard the judgment and opinion in Kleinwort Benson Ltd v Glasgow City Council Case C-346/93 [1995] All
ER (EC) 514, [1995] ECR I-615. See also the opinions of Advocate General Jacobs in Leur-Bloem v Inspecteur der
Belastingdienst/Ondernemingen Amsterdam 2 Case C-28/95 [1997] All ER (EC) 739 and Giloy v Hauptzollamt
Frankfurt am Main-Ost Case C-130/95 (1997) ECJ Transcript, 17 July.

11.
The reasoning just outlined cannot in any way be altered by the possible application to this case of the
rules set out in the EEA Agreement, which entered into force on 1 January 1994 and to which Austria was
a party on that date. Even if the court wished to reformulate the question along these lines in order to
bring it within the ambit of the corresponding provision in the EEA Agreement (art 4), the court would in
my view lack jurisdiction to rule on the question submitted to it for a preliminary ruling. The question in this
case comes from a national court which, at the time when the facts in issue occurred or at the time
subsequently when the main proceedings were instituted, was not competent to address a question to the
Court of Justice pursuant to that agreement. It is true that, under art 107 of the EEA Agreement and
Protocol 34 thereto4, a European Free Trade Association (EFTA) state may allow a court or tribunal to ask
the Court of Justice of the European Communities to decide on the interpretation of an EEA rule. The
exercise of that option, however, is expressly made subject to prior performance of the obligation, set out
in art 2 of Protocol 34, to notify the 243 Depositary and the Court of Justice of the European Communities
to what extent and according to what modalities the protocol is to apply to the courts and tribunals of the
state in question. Austria, however, has not exercised that power or satisfied the obligation to which such
exercise is subject. Nor can it be argued that the court gained that jurisdiction as a result of Austrias
subsequent accession to the Community. The protocol referred to does not contain any provision to that
effect. There is also no reason to assume that the power in question was implicitly authorised by the
protocol but left dormant, so to speak, until the country concerned had acceded to the EC Treaty.
4
See Protocol 34 on the possibility for courts and tribunals of EFTA states to request the Court of Justice of the European
Communities to decide on the interpretation of EEA rules corresponding to EC rules (see OJ 1994 L1 p 204).

12.
Nor can it be argued that the Court of Justice may substitute itself for the EFTA Court in order to
interpret, in its place, provisions of the EEA Agreement which it had no jurisdiction to interpret at the time
when the contested facts occurred or when the proceedings were instituted. On this point, moreover, it
must be recalled that art 108 of the EEA Agreement provided for the establishment of an EFTA Court
equipped with a jurisdictional remit similar in some respects to that established by art 177 of the EC
Treaty, to which national courts of countries which are party to the EEA Agreement but are not Community
member states may have recourse5. At the time when Austria, Finland and Sweden acceded to the
Community, an ad hoc agreement6 laid down rules allowing the courts and tribunals of those three
member states, albeit for a short period, to refer cases of the type mentioned to the EFTA Court, even
following accession to the Community. In those cases in which the jurisdiction of the EFTA Court is
expressly provided for, the national court called on to apply the provisions of the EEA Agreement is
certainly not empowered alternatively to raise the question for a preliminary ruling before the Court of
Justice rather than before the EFTA Court. There is no such possibility of choice, nor could such a choice
have been provided for without running counter to the principle of exclusive jurisdiction which the Court of
Justice has, moreover, confirmed even in specific relation to the EEA Agreement (see Opinion 1/92 [1992]
ECR I-2821)7.
5
Article 108 was given effect by the EFTA states when they concluded the Agreement on the Establishment of a
Surveillance Authority and a Court of Justice. Article 34 of that agreement provides that: (1) The EFTA Court shall
have jurisdiction to give advisory opinions on the interpretation of the EEA Agreement. (2) Where such question is raised
before any court or tribunal in an EFTA state, that court or tribunal may, if it considers it necessary to enable it to give
judgment, request the EFTA Court to give such an opinion.
6
See the Agreement on Transitional Arrangements for a Period after the Accession of Certain States to the European
Union. See esp art 5(1), which provides that: After accession, new proceedings may only be instituted before the EFTA
Court in cases in which the events giving rise to an action under the EEA Agreement on the Surveillance and Court
Agreement occurred before accession and an application is lodged with the EFTA Court within three months after
accession.
7
The clear distinction in the areas of jurisdiction between the EC Court and the EFTA Court, introduced after the adoption
of the initial text of the Agreement, in fact made it possible to regard as compatible with the EC Treaty the jurisdictional
mechanism created by art 108 of the EEA Agreement (para 19). The need to comply with arts 164 and 219 of the Treaty
was further clearly underlined by the Court of Justice in Opinion 2/94 [1996] ECR I-1759 (para 20).

B. Merits
13.
In the event that the court should rule that it does have jurisdiction, I shall set out below some thoughts
on the merits of the question submitted for a preliminary ruling.
There is here, in my view, one single real problem, which arises on the basis of this courts case law
regarding the compatibility of the security for legal costs with 244 the principle of non-discrimination. It is
necessary to determine whether the proceedings brought by the plaintiff involve a substantive right
protected by the Community legal system or whether the court is faced with proceedings brought
exclusively by reference to the provisions of company law in force at the time within the Austrian legal
system.
14.
The European Commission wishes to place the present case in the context of the provisions of art
54(3)(g) and the fourth indent of art 220 of the EC Treaty. For his part, the plaintiff submits that the
proceedings relate to the rights provided for by arts 52 and 54(3)(g) of the Treaty. It seems to me,
however, that the proceedings in question cannot have any basis in the rules of Community law, precisely
because of the time at which they were instituted, which was prior to Austrias accession to the
Community. The only possibility which seems to me to be reasonably well founded is therefore that of
bringing the legal proceedings instituted by the plaintiff into relation with the provisions of the EEA
Agreement and specifically with those set out in arts 31 to 35 seeking to eliminate restrictions on freedom
of establishment, which also relate to companies, as well as art 77, dealing with company law, which
refers to Annex XXII, which imposes an obligation on EFTA states party to the Agreement to give effect
within the EEA to a series of Community directives, with a view to rendering equivalent the guarantees
designed to protect the interests of shareholders and others in the corporate context and with a view to
regulating specific matters relating to companies, such as secessions, mergers and so forth.
15.
In the light of the courts case law on the matter, there is also no basis to the arguments set out by the
defendant regarding the plaintiffs nationality and the alleged absence of discrimination in the treatment to
which he is subject under the Austrian legislation (see the judgment in Micheletti v Delegacin del
Gobierno en Cantabria Case C-369/90 [1992] ECR I-4239). The courts case law is clear on this point: the
key factor with regard to exercise of the rights conferred by the Treaty is the fact of being a national of a
Community member state. That criterion must also be valid with regard to the scope ratione personae of
the EEA Agreement: the fact that the person concerned may have another, non-Community (or non-EEA)
nationality neither adds to nor detracts from the rights to which he is entitled under the Community legal
order and that established by the EEA Agreement.
16.
The other issue in dispute concerns the nature and the effects of the rule of Austrian procedural law
underlying the dispute which has been referred to the court. On this point, for considerations of a general
nature regarding the legality under Community law of the security for costs and the criteria permitting
application thereof, I refer to the judgments given most recently by the court on this subject and the
opinions which I delivered in those cases (see the judgments in Data Delecta [1996] All ER (EC) 961,
[1996] ECR I-4661 and Hayes v Kronenberger GmbH (in liq) Case C-323/95 [1997] ECR I-1711).
The provision in question is also clearly based on a criterion involving discrimination between nationals
and non-nationals. An Austrian national is exempted from providing the security even if he is resident
outside Austria or does not own sufficient assets within the national territory to satisfy a defendants
claims for reimbursement of legal costs incurred by the latter. The rule is thus in no way designed to
protect a party against whom unfounded or vexatious proceedings have been brought, except in those
cases where such proceedings have been instituted by a non-national. No special consideration need
therefore 245 be paid to the objectivesof ensuring equality or providing guaranteeswhich the
defendant, in my view unjustifiably, attributes to it.
17.
In contrast, consideration should be paid to the reasoning set out by the Austrian government, which
argues that the disputed provision of national law is compatible with Community law because it provides
an exception for contrary provisions under international treaties or conventions, with the result that
Community nationals are exempt from having to provide the security. The same, it need hardly be
mentioned, applies with regard to the EEA. This interpretation, in conformity with Community law, put
forward by the Austrian government does not alter, but rather confirms, the conclusion which I have
reached: the only difference is that it shifts the obligation to comply with the Community or EEA rule from
the legislature, which would be required to repeal or amend the provision contrary to the principle of non-
discrimination, to the courts, which would also be required not to apply that provision in order to ensure
compliance with the rules in which that principle is set out. What matters, in any event, for the purpose of
proceedings which are the subject matter of a reference to the court, is that the rules with which it must
ensure compliance should be fully and directly applied within the national legal order. The member state,
thus, remains competent in regard to solutions at the constitutional level which it considers appropriate or
preferable for the purpose of achieving such a result, on condition that individuals can enjoy effectively
and with certainty the rights conferred on them by the Community legal order or that of the EEA (see the
judgments on EC Commission v Netherlands Case 72/85 [1986] ECR 1219, EC Commission v Italy Case
168/85 [1986] ECR 2945 and EC Commission v Italy Case 104/86 [1988] ECR 1799). In any event, the
national courts are required to take the view that the contested security cannot be applied to or held
against a Community or EEA national (see the judgment in Amministrazione delle Finanze dello Stato v
Simmenthal SpA Case 106/77 [1978] ECR 629).

IV CONCLUSION
18.
On the basis of the foregoing, I propose that the Court of Justice reply as follows to the question
referred by the Oberster Gerichtshof in Vienna:
The court does not have jurisdiction to rule on the question submitted for a preliminary ruling by
the Oberster Gerichtshof in Vienna, in so far as the facts underlying the dispute pending before that
court predate the Republic of Austrias accession to the European Community and thus fall outside
the scope ratione temporis of the EC Treaty.
In the alternative, should the court consider that it does have jurisdiction to rule on the question submitted,
I propose that it reply along the following lines:
Article 4 of the Agreement on the European Economic Area precludes nationals of Community
member states or of states party to the EEA Agreement from being required to provide security for
the costs of legal proceedings, as provided for under para 57 of the Austrian Code of Civil
Procedure, where Austrian nationals, in similar circumstances, are not required to provide such
security.

246
2 October 1997.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By order of 11 March 1996, received at the Court of Justice of the European Communities on 16 April
1996, the Oberster Gerichtshof (the Austrian Supreme Court) referred for a preliminary ruling under art
177 of the EC Treaty a question on the interpretation of the first paragraph of art 6 of that Treaty.
2.
That question has arisen in proceedings brought by Mr Saldanha and MTS Securities Corp against
Hiross Holding AG, an Austrian company (Hiross) in which they are shareholders, seeking an injunction to
restrain Hiross from selling or transferring shares which it holds in a number of its subsidiaries to its Italian
subsidiary, or subsidiaries of that company established in Italy, without the approval of the general
meeting of shareholders.
3.
Hiross thereupon applied to the Handelsgericht Wien (the Commercial Court, Vienna) for an order
requiring Mr Saldanha, a national of both the United States of America and the United Kingdom living in
Florida, and MTS Securities Corp, a company domiciled in the United States, to provide security for the
costs of the proceedings, pursuant to para 57(1) of the Zivilprozeordnung (the Austrian Code of Civil
Procedure) (the ZPO).
4.
Under that provision, foreign nationals who are plaintiffs in proceedings brought before Austrian courts
must, on application by the defendant, lodge a sum as security for the costs of the proceedings (cautio
judicatum solvi), except where otherwise provided by international treaty or convention. Paragraph 57(2)
of the ZPO, however, provides that that obligation does not apply where, in particular, the plaintiff is
normally resident in Austria or a judicial decision ordering the plaintiff to indemnify the defendant for his
legal costs is enforceable in the state in which the plaintiff is normally resident.
5.
In this connection, it appears from the order for reference that there is no convention or treaty between
the Republic of Austria and the United States of America or the State of Florida enabling an Austrian
decision on legal costs to be enforced in Florida (see art 37 of the Decree of 21 October 1986 on
international judicial assistance and other legal relations with foreign countries in civil matters; see (1986)
JABl 53). According to the Austrian government, even though it appears that some American courts have
recognised Austrian decisions on enforcement, recognition and enforcement of such decisions in the
United States of America cannot be guaranteed because in the absence of a convention or treaty it is not
possible to enforce American decisions in Austria. In any event, it appears from the order for reference
that the Oberster Gerichtshof has already ruled that a foreign plaintiff normally resident in Florida must in
principle be subject to the requirement of lodging security for legal costs owing to the absence of a
convention or treaty governing the matter.
6.
Although art 11 of the Convention between the United Kingdom and the Republic of Austria regarding
Legal Proceedings in Civil and Commercial Matters (London, 31 March 1931; TS4 (1932); Cmd 4007)
provides that nationals of the signatory states are to be exempt from the obligation to provide security for
costs, that exemption is limited to persons domiciled in one or other of those states. Under the
Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1988
(Lugano, 16 September 1988; OJ 1988 L319 p 9), which has been binding on both the Republic of Austria
and the United Kingdom since 1 September 1996, a decision delivered in one contracting state and
capable of being enforced there must in principle also be recognised in another 247 contracting state after
it has been declared enforceable. However, para 57(2)(1a) of the ZPO makes its application subject to the
possibility of enforcement in the state in which the plaintiff is normally resident, in casu the United States
of America.
7.
On 22 November 1994 the Handelsgericht Wien made an order requiring Mr Saldanha and MTS
Securities Corp jointly and severally to lodge a sum of S 500,000 as security for Hiross legal costs, on the
ground that they were not entitled to exemption under para 57(2) of the ZPO, and indicating that they
would be deemed, on application by Hiross, to have discontinued the proceedings if they failed to comply
with the time limit for meeting that obligation.
8.
The Republic of Austria acceded on 1 January 1995 to the European Union and to the Treaties on
which it is founded, including the EC Treaty. The first paragraph of art 6 of that treaty provides:
Within the scope of application of this Treaty, and without prejudice to any special provisions
contained therein, any discrimination on grounds of nationality shall be prohibited.
9.
The Oberlandesgericht Wien (the Higher Regional Court, Vienna), before which the matter came on
appeal, set aside the order of the Handelsgericht in so far as Mr Saldanha was concerned on the ground
that he was a British national and that it would therefore be contrary to the first paragraph of art 6 of the
Treaty to require him to provide security. The Oberlandesgericht took the view that neither his dual
nationality nor the fact that he was not normally resident in a member state could in any way alter that
conclusion.
10.
Hiross appealed against that decision on a point of law to the Oberster Gerichtshof. Since it formed
the view that art 6 of the Treaty was a public policy provision which must, by virtue of Austrian procedural
law, be taken into account by national courts, even in a case predating the Republic of Austrias accession
to the Communities, the Oberster Gerichtshof decided to stay the proceedings and refer the following
question to the Court of Justice for a preliminary ruling:
Where proceedings are brought before an Austrian civil court by a British national who is also a
national of the United States of America, who resides in that country (in Florida) and does not have
any residence or assets in Austria, against a limited company whose registered office is in Austria,
seeking to have that company restrained from selling or otherwise transferring shares in specified
subsidiary companies to its Italian subsidiary company, or to subsidiaries of that company with
registered offices in Italy, without the approval of a qualified majority of three-quarters of the
general meeting of shareholders or, in the alternative, of a simple majority of the general meeting of
shareholders, does the fact that he has been ordered by the competent Austrian court (of first
instance), on application by the defendant company pursuant to Paragraph 57(1) of the Austrian
Code of Civil Procedure, to provide security for costs in a specified sum constitute discrimination on
grounds of nationality contrary to the first paragraph of Article 6 of the EC Treaty?
11.
The question thus seeks to ascertain whether the first paragraph of art 6 of the Treaty precludes a
member state from requiring provision of security for costs by a national of another member state who is
also a national of a non-member country, in which he is domiciled, where that national, who is not resident
and has no assets in the first member state, has brought before one of its civil courts an 248 action in his
capacity as a shareholder against a company established in that member state, even though such a
requirement is not imposed on its own nationals who are not resident and have no assets there.

The application ratione temporis of the first paragraph of art 6 of the Treaty
12.
As a preliminary point, Hiross argues that the matter at issue in the main proceedings is outside the
scope ratione temporis of Community law, since the facts, including the order of the Handelsgericht
requiring Mr Saldanha to provide security, predate the Republic of Austrias accession to the European
Communities. From this Hiross concludes that there cannot be any discrimination contrary to art 6 of the
Treaty.
13.
Article 2 of the Act of Accession (1994); OJ 1994 C241 p 21 (Norway, Austria, Finland, Sweden)
provides that, from the date of accession, the provisions of the original Treaties on which the European
Union is founded are to be binding on the new member states and are to apply in those states under the
conditions laid down in those Treaties and in the Act of Accession.
14.
In view of the fact that the Act of Accession contains no specific conditions whatsoever with regard to
the application of art 6 of the Treaty, the latter provision must be regarded as being immediately
applicable and binding on the Republic of Austria from the date of its accession, with the result that it
applies to the future effects of situations arising prior to that new member states accession to the
Communities. From the date of accession, therefore, nationals of another member state can no longer be
made subject to a procedural rule which discriminates on grounds of nationality, provided that such a rule
comes within the scope ratione materiae of the EC Treaty.

The scope ratione materiae et personae of the first paragraph of art 6 of the Treaty
15.
It is first necessary to point out that the mere fact that a national of a member state is also a national of
a non-member country, in which he is resident, does not deprive him of the right, as a national of that
member state, to rely on the prohibition of discrimination on grounds of nationality enshrined in the first
paragraph of art 6 (see to that effect, with regard to art 52 of the Treaty, Micheletti v Delegacin del
Gobierno en Cantabria Case C-369/90 [1992] ECR I-4239 (para 15)).
16.
Since art 6 of the Treaty produces effects within the area covered by the Treaty, it is necessary to
consider next whether that article applies to a provision in a member state, such as that at issue in the
main proceedings, which requires nationals of another member state to provide security for costs where,
in their capacity as shareholders, they bring proceedings against a company established in that member
state, even though its own nationals are not subject to such a requirement.
17.
In this connection, the court has held, in Data Delecta Aktiebolag v MSL Dynamics Ltd Case C-43/95
[1996] All ER (EC) 961, [1996] ECR I-4661 (para 15) and in Hayes v Kronenberger GmbH (in liq) Case C-
323/95 [1997] ECR I-1711 (para 17), that such a rule of domestic procedure falls within the scope of
application of the Treaty within the meaning of the first paragraph of art 6, where the main proceedings
relate to the exercise of the fundamental freedoms guaranteed by Community law, such as, in those
cases, proceedings to recover payment for the supply of goods.
18.
Hiross argues that, in this case, the main proceedingswhich seek to prevent Hiross from selling or
transferring shares which it holds in a number of its subsidiaries to its Italian subsidiary, or subsidiaries of
that company established in 249 Italy, without the approval of the general meeting of shareholdersare in
no way connected with the exercise of a fundamental freedom guaranteed by Community law. Moreover,
it contends that the national provision at issue in the main proceedings does not come within the scope of
application of the Treaty by virtue of art 220 of the EC Treaty.
19.
It should be noted in this regard that, while a rule of procedure such as that at issue in the main
proceedings is in principle a matter for which the member states are responsible, the court has
consistently held that such a provision may not discriminate against persons to whom Community law
gives the right to equal treatment or restrict the fundamental freedoms guaranteed by Community law
(see the judgment in Cowan v Trsor Public Case 186/87 [1989] ECR 195 (para 19)).
20.
In the above-mentioned judgments in Data Delecta [1996] All ER (EC) 961, [1996] ECR I-4661 (para
15) and Hayes [1997] ECR I-1711 (para 17), the court held that a rule of domestic procedure requiring for
judicial proceedings, such as those at issue in those cases, the provision of security for costs was liable to
have an effect, even though indirect, on trade in goods and services between member states and
therefore fell within the scope of application of the Treaty.
21.
Without its being necessary to examine the argument of Hiross that, in view of the subject-matter of
the dispute in the main proceedings, the contested rule cannot in this case restrict, even indirectly, any
fundamental freedom guaranteed by Community law, it must be held that such a rule cannot, in any
event, discriminate against persons on whom Community law confers the right to equal treatment.
22.
The dispute in the main proceedings concerns the protection of interests relied on by a shareholder
who is a national of one member state against a company established in another member state.
23.
Article 54(3)(g) of the EC Treaty empowers the Council and the Commission, for the purpose of giving
effect to the freedom of establishment, to co-ordinate to the necessary extent the safeguards which, for
the protection of the interests of members and others, are required by member states of companies or
firms within the meaning of the second paragraph of art 58 of the EC Treaty with a view to making such
safeguards equivalent throughout the Community. It follows that rules which, in the area of company law,
seek to protect the interests of shareholders come within the scope of the Treaty and are for that reason
subject to the prohibition of all discrimination based on nationality.
24.
If Community law thus prohibits all discrimination based on nationality in regard to the safeguards
required, in the member states, of companies or firms within the meaning of the second paragraph of art
58 of the Treaty for the purpose of protecting the interests of shareholders, nationals of a member state
must also be in a position to seise the courts of another member state of disputes to which their interests
in companies there established may give rise, without being subject to discrimination vis--vis nationals of
that state.

Discrimination within the meaning of the first paragraph of art 6 of the Treaty
25.
By prohibiting any discrimination on grounds of nationality, art 6 of the Treaty requires, in the member
states, complete equality of treatment between persons in a situation governed by Community law and
nationals of the member state in question.
26.
It is clear that a provision such as that at issue in the main proceedings amounts to direct
discrimination on grounds of nationality. Under that provision, 250a member state does not require its
own nationals to provide security, even if they are not resident and have no assets in that state.
27.
Hiross, however, takes the view that the distinction based on nationality is justified on objective
grounds. In support of this view, it argues that the object of the provision in dispute is to ensure that a
defendant will be able to exercise his right to recover his costs if successful. It refers in particular to the
problems of enforcement which might arise where the plaintiff is not resident and has no assets in the
Community, as is the position in the main case.
28.
In this context, Hiross submits that the possession of Austrian nationality as a criterion permitting
exemption from the obligation to provide security is justified by the likelihood of obtaining enforcement,
within the national territory, of a right to reimbursement of costs awarded against a solvent national, a
likelihood attributable, in particular, to the assumed existence of assets linked to the national territory and
the tendency to comply with decisions of national courts.
29.
Suffice it in this regard to point out that, even though the object of a provision such as that at issue in
the main proceedings, namely that of ensuring enforcement of a decision on costs in favour of a
defendant who has been successful in proceedings, is not as such contrary to art 6 of the Treaty, the fact
remains that that provision does not require Austrian nationals to provide security for costs, even if they
are not resident and have no assets in Austria and are resident in a non-member country in which
enforcement of a decision on costs in favour of a defendant is not guaranteed.
30.
In those circumstances, the answer to the question submitted must be that the first paragraph of art 6
of the Treaty must be construed as precluding a member state from requiring provision of security for
costs by a national of another member state who is also a national of a non-member country, in which he
is resident, where that national, who is not resident and has no assets in the first member state, has
brought proceedings before one of its civil courts in his capacity as a shareholder against a company
established in that member state, if such a requirement is not imposed on its own nationals who are not
resident and have no assets there.

Costs
31.
The costs incurred by the Austrian and UK governments and by the European Commission, which
have submitted observations to the Court of Justice, are not recoverable. Since these proceedings are,
for the parties to the main proceedings, a step in the action pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the question referred to it by the
Oberster Gerichtshof by order of 11 March 1996, hereby rules: the first paragraph of art 6 of the EC
Treaty must be construed as precluding a member state from requiring provision of security for costs by a
national of another member state who is also a national of a non-member country, in which he is resident,
where that national, who is not resident and has no assets in the first member state, has brought
proceedings before one of its civil courts in his capacity as a shareholder against a company established
in that member state, if such a requirement is not imposed on its own nationals who are not resident and
have no assets there.

251

[1998] All ER (EC) 252

Provincia Autonoma di Trento and another v Dega di Depretto Gino

Snc
(Case C-83/96)
EUROPEAN COMMUNITY; Consumer
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIRST CHAMBER)
JUDGES SEVN (PRESIDENT OF THE CHAMBER), JANN AND WATHELET (RAPPORTEUR)
ADVOCATE GENERAL LGER
17 APRIL, 29 MAY, 17 SEPTEMBER 1997
European Community Consumer protection Packaging Labelling of foodstuffs Plaintiff marketing
food product in Community Labels bearing addresses of manufacturer and packager established
outside Community Whether labels should also bear details of at least one trader established within
Community Council Directive (EEC) 79/112, art 3.

The defendant undertaking marketed cans of pineapples in Italy. The labels on the cans stated only the
name and address of the producer and packager, which was an undertaking established outside the
Community. The administrative authorities of the Province of Trento subsequently fined the defendant on
the grounds that the labels did not bear the details of a trader established within the Community in
accordance with the Italian measures implementing art 3(1)(6) 1 of Council Directive (EEC) 79/112 on the
approximation of the laws of the member states relating to the labelling, presentation and advertising of
foodstuffs for sale to the ultimate consumer. That provision made compulsory the inclusion of the name
or business name and address of the manufacturer or packager, or of a seller established within the
Community on the labelling of foodstuffs. The defendant appealed against the fine and the Pretura
Rovereto allowed the appeal on the grounds that the expression established within the European
Economic Community in the relevant Italian provision referred only to sellers and that it was sufficient to
indicate name and address of the producer and packager established in a third country. The provincial
authorities appealed against that ruling, contending that the protection of the ultimate consumer was only
fully guaranteed if details of at least one trader established within the Community were indicated on the
label, whether of the manufacturer, the packager, or the seller. Since the Italian legislation reproduced art
3(1)(6) of the directive almost exactly, the Corte Suprema di Cassazione stayed the proceedings and
referred to the Court of Justice of the European Communities for a preliminary ruling a question on the
interpretation of art 3(1)(6).
1
Article 3, so far as material, is set out at p 259 j, post

Held Directive 70/112 was intended to ensure the information and protection of the ultimate consumer
of foodstuffs in relation to the nature, identity, properties, composition, quantity, durability, origin or
provenance and method of manufacture or production of those products. In particular, art 3(1)(6) was
intended to enable the consumer to contact a person responsible for the manufacture or packaging of the
foodstuffs with a view to expressing any positive or negative criticism about the product purchased. That
goal could only be achieved if the ultimate consumer could easily identify the person responsible for 252
the product. In that respect, producers and packagers were generally established, easily identifiable
traders, so that the fact that they might be situated outside the Community did not present a problem;
sellers, however, were generally much smaller traders and consequently more difficult to identify,
particularly if established outside the Community. It followed that a label on the packaging could indicate
details of either the manufacturer or the packager, whether established within the Community or not; with
regard to sellers, the label could only indicate details of a trader established within the Community.
Accordingly, art 3(1)(6) was to be interpreted as meaning that the expression established within the
Community referred only to the seller (see p 261 c to g j, post).

Cases cited
Aannemersbedrijf PK Kraaijeveld B v Gedeputeerde staten van Zuid-Holland Case C-72/95 [1996] ECR I-
5403.
Coperatieve Zuivelindustrie Twee Provincin WA (Criminal proceedings against) Case C-285/92 [1993]
ECR I-6045.
EC Commission v UK Case 100/84 [1985] ECR 1169.
R v Bouchereau Case 30/77 [1977] ECR 1999.
Rockfon A/S v Specialarbeijderforbundet i Danmark Case C-449/93 [1995] ECR I-4291.
Smanor SA (Proceedings for compulsory reconstruction against) Case 298/87 [1988] ECR 4489.

Reference
By order of 4 December 1995, the Corte Suprema di Cassazione (the Supreme Court of Cassation)
referred to the Court of Justice of the European Communities for a preliminary ruling under art 177 of the
EC Treaty a question (set out at p 260 d, post) on the interpretation of art 3(1)(6) of Council Directive
(EEC) 79/112 on the approximation of the laws of the member states relating to the labelling, presentation
and advertising of foodstuffs for sale to the ultimate consumer. That question was raised in proceedings
between, on the one hand, the Provincia Autonoma di Trento (the Autonomous Province of Trento) and
the Ufficio del Medico Provinciale di Trento (the Medical Office of the Province of Trento) and, on the
other, the partnership Dega di Depretto Gino (Dega) concerning the latters failure to comply with the
Italian legislation on labelling. Written observations were submitted on behalf of: the Italian government,
by Professor U Leanza, Head of the Legal Department of the Ministry of Foreign Affairs, acting as agent,
assisted by D Del Gaizo, Avvocato dello Stato; the Greek government, by K Vasileios and M Basdeki,
Assistant Legal Adviser and Legal Representative respectively in the State Legal Service, acting as
agents; and the European Commission, by A Aresu and P Stancanelli, of its Legal Service, acting as
agents. Oral observations were made by: the French government, represented by R Loosli-Surrans, Head
of Section in the Legal Directorate of the Ministry of Foreign Affairs, acting as agent; the Italian
government, represented by D Del Gaizo; the Greek government, represented by K Vasileios; and the
Commission, represented by P Stancanelli. The language of the case was Italian. The facts are set out in
the opinion of the Advocate General.

253
29 May 1997.

The Advocate General (P Lger)

delivered the following opinion (translated from the French).
1.
Il suffit du dplacement dune virgule pour denaturer le sens de ma pense (moving a single comma
is enough to distort my meaning), said Michelet2, a comment which indicates to perfection the profound
perplexity of the reader faced with two differently punctuated versions of the same text. Such perplexity
the Corte Suprema di Cassazione (the Supreme Court of Cassation) must have experienced in its
endeavour to determine the precise meaning of a provision of Community law.
2
See Berthier and Colignon Le franias practique p 192.

2.
It is in considering a reference for a preliminary ruling from that court, seeking clarification of one of the
requirements of the Community rules on the labelling and presentation of foodstuffs 3 concerning the
indication of details of a trader, that the finesses and nuances of good punctuation become apparent.
3
See Council Directive (EEC) 79/112 on the approximation of the laws of the member states relating to the labelling,
presentation and advertising of foodstuffs for sale to the ultimate consumer (OJ 1979 L33 p 1).

3.
Let us first consider the facts and procedure in this case.

Facts and procedure

4.
A fine was imposed on the company Dega di Depretto Gino Snc for having marketed in Italy cans of
pineapple in syrup, produced and packaged by a company established outside the Community, without
indicating the details required by the Italian legislation 4, under which the labelling on foodstuffs must
indicate, in particular:
4
See art 3(h) of the Presidential Decree No 322 of 18 May 1992, entitled implementation of Directive 79/112 on
foodstuffs for sale to the ultimate consumer and the advertising thereof, and Directive (EEC) 77/94 on foodstuffs for
particular nutritional uses.
 the name, business name or registered trademark and the registered office of the manufacturer
or packager or of a seller established within the European Economic Community. 5

5
The Italian text, as reproduced by the national court at point 1 of the order, is drafted as follows: il nome o la ragione
sociale o il marchio depositato e la sede del fabbricante o del confezionatore o di un venditore stabilito nella Communit
economica europea.

5.
The appeal against that administrative fine was upheld by the Pretura (the Magistrates Court),
Rovereto, which, by judgment of 20 November 1990, annulled the fine on the ground that it had been
imposed on the basis of an incorrect interpretation of the legal provision cited above. That court, ruling on
the substance, held that the expression established within the European Economic Community referred
only to sellers and that an indication of the name and address of the producer and packager established
in a non-member country, as in this case, was sufficient.
6.
The Provincia Autonoma di Trento (the Autonomous Province of Trento) and the Ufficio del Medico
Provinciale di Trento (the Medical Office of the Province of Trento) lodged an appeal seeking to have the
judgment set aside on a single ground: they challenged the interpretation given by the Pretura to the
national provision at issue. They claimed that protection of the ultimate consumer is only fully guaranteed
if at least one trader (producer, packager or seller) established within the Community is indicated on the
label.
7.
The Corte Suprema di Cassazione noted that the national provision which it was asked to interpret
reproduces almost exactly the wording of art 3(1)(6) of 254 Directive 79/112, the French version of which
provides that the labelling of foodstuffs must indicate le nom ou la raison sociale et ladresse du fabricant
ou du conditionneur,6 ou dun vendeur tablit lintrieur de la Communaut; the Italian version reads as
follows: il nome o la ragione sociale e lindirizzo del fabbricante o del condizionatore o di un venditore
stabilito nella Communit.
6
I would highlight the presence of this comma, to which I shall return later.

8.
It concluded that the judgment to be given entails, as a matter of necessity and priority, an
interpretation of that Community provision, since the Italian provision is obviously merely a re-enactment
of it, and has therefore asked the Court of Justice to rule on the following question:
Must Article 3(1)(6) of Council Directive 79/112/EEC (on the approximation of the laws of the
member states relating to the labelling, presentation and advertising of foodstuffs for sale to the
ultimate consumer) be interpreted as meaning that the expression established within the
Community used in it refers only to the seller or refers, in the absence of a seller established within
the Community, also to the manufacturer and/or packager? Must that provision therefore be taken
to mean that, in the absence of a seller established in the Community, the manufacturer and/or the
packager must be established in the Community?

Legal background
9.
The directive on the labelling of foodstuffs enacts Community rules of a general nature applicable
horizontally to all foodstuffs put on the market (third recital in the preamble to the directive). The
harmonisation of laws carried out by the directive is intended to prevent barriers to the free movement of
such products, which can lead to unequal conditions of competition, in order to contribute to the smooth
functioning of the common market (first and second recitals). The prime consideration, however, is the
need to inform and protect the consumer (sixth recital). In order to do so, the directive provides in
particular that the labelling must indicate certain information, to the exclusion of any other. That
information is set out in art 3(1), (1) to (8).

Answer to the question

10.
The question relating to the interpretation of art 3(1)(6) essentially asks the court to rule whether the
condition of Community establishment applies only to the seller or whether, on the contrary, it must be
satisfied by at least one of the three traders mentioned (manufacturer, packager and seller).
11.
A careful reader will have noted the slight difference in meaning between the Italian and the French
versions of the directive: in the former, no comma separates the expression or of a seller established
within the Community from the other persons mentioned.
12.
The position of the comma in the French version has the advantage of separating, in my view quite
clearly, the last person mentioned from the two others, and thus of identifying that person as being the
only one to whom the expression at issue could apply. The English version of that provision, the name or
business name and address of the manufacturer or packager, or of a seller established within the
Community, which also has the comma, seems to call for the same reading.
255
13.
The German (den Namen oder die Firma und die Anschrift des Herstellers, des Verpackers oder
eines in der Gemeinschaft niedergelassenen Verkufers) and Dutch (de naam of de handelsnaam en
het adres van de fabrikant of van de verpakker of van een in de Gemeenschap gevestigde verkoper)
versions of the text are also unambiguous. The syntax of those two languages makes it even clearer that
the expression established within the Community (in der Gemeinschaft niedergelassenen and in de
Gemeenschap gevestigde) applies only to the seller (Verkufer and verkoper), since that expression
directly precedes the word seller and it is thus not possible to interpret the text so as to apply the
condition at issue to all three traders listed.
14.
The Italian version, on the other hand, which gave rise to the national courts questions and which can
be compared to later translations7, could be interpreted as having an entirely different meaning.
Nonetheless, contrary to what was argued by the Italian government at the hearing, the interpretation to
be given to the provision at issue cannot depend on that language version alone. It is vain to argue that
the need to take account of other language versions constitutes discrimination against Italian traders, in
so far as the court has consistently held that
7
There is no sign of the comma at issue in the Spanish version: el nombre o la razn social y la direccin del fabricante
o del embalador o de un vendedor establecido dentro de la Communidad. The Danish and Greek versions seem to
follow the Italian example.

the different language versions of a Community text must be given a uniform interpretation and
in the case of divergence between the versions the provision in question must therefore be
interpreted by reference to the purpose and general scheme of the rules of which it forms part.
(See the judgment in Rockfon A/S v Specialarbeijderforbundet i Danmark Case C-449/93 [1995]
ECR I-4291 (para 28) (my emphasis), which refers to R v Bouchereau Case 30/77 [1977] ECR
1999 (para 14).8

8
See also eg EC Commission v UK Case 100/84 [1985] ECR 1169 (para 17) and Aannemersbedrijf PK Kraaijeveld B v
Gedeputeerde staten van Zuid-Holland Case C-72/95 [1996] ECR I-5403 (para 28).

15.
In view of the uncertainty which persists after reading the different language versions and since none
of them can be given precedence, the court must endeavour, in accordance with its case law, to interpret
the text referred for its consideration in a manner which is consistent with the purpose and general
scheme of the rules of which it forms a part.
16.
However, before doing so, I suggest account should be taken of one factor which appears to me to be
conclusive.
17.
The importance of the distinction in meaning resulting from the insertion of the comma, at least in the
French and English versions, did not escape the notice of the Economic and Social Committee which, in
its opinion on what was then merely the proposal for Directive 79/112 9, proposed that the punctuation be
amended specifically in order that the condition of Community establishment should apply to each of the
traders mentioned in that text.
9
See the opinion on the proposal for a Council Directive on the approximation of the laws of the member states relating
to the labelling, presentation and advertising of foodstuffs for sale to the ultimate consumer (OJ 1976 C285 p 3).

18.
Point 2.7.1 of the Committees specific comments in respect of art 3 of the proposed directive thus
reads as follows:
The Committee considers that an accountable party within the Community must be specified on
the labelling. The Committee, therefore, 256asks that Article 3(6) should be reworded to read: 6.
The name or business name of the manufacturer, packer or seller, together with his address, which
must be within the Community.10

10
The French version of the proposed wording (Le nom ou la raison sociale et ladresse du fabricant ou du conditionneur,
ou dun vendeur, tabli linteneur de la Communaut) includes two commas, separating the expression ou dun
vendeur from the rest of the sentence.

The Italian version of that proposal is even clearer:

Il Comitato ritiene che limballaggio debba comportare lindicazione della persona responsabile
nella Communit. Esso chiede quindi che il punto sia redatto nel modo seguente: il nome e
cognome o la ragione sociale e lindirizzo del fabbricante o del condizionatore o di un vendore
stabiliti nella Communit.11

11
It should be noted that, in the Italian version, the proposed amendment concerned the plural agreement of the past
participle of the verb to establish (stabiliti instead of stabilito), thus referring to all the traders mentioned and not merely
to one of them.

19.
 The fact that that proposal was not taken up in the final draft of the provision at issue 12 can, in my
opinion, mean only that the Community legislature considered that the requirement of Community
establishment should apply only to the seller.
12
The wording in Directive 79/112 is exactly the same as that originally proposed by the European Commission in its
proposal for a Council directive on the harmonisation of the laws of the member states relating to the labelling,
presentation and advertising of foodstuffs for sale to the ultimate consumer, submitted to the EC Council on 30 March
1976 (OJ 1976 C91 p 3), including the Italian version.

20.
A purposive interpretation of the provision at issue only confirms that view.
21.
In that respect, as the Italian and Greek governments have rightly recalled, the court has already
identified the purpose and general scheme of Directive 79/112 as follows:
it is clear from both the statement of the reasons on which the directive is based and the
terms of Article 2 thereof, that its object was to inform and protect the ultimate consumer of
foodstuffs, in particular as regards the nature, identity, properties, composition, quantity, durability,
origin or provenance, and the method of manufacture or production thereof. (See the judgment in
Proceedings for compulsory reconstruction against Smanor SA Case 298/87 [1988] ECR 4489
(para 30).)13

13
See also the judgment in Criminal proceedings against Coperatieve Zuivelindustrie Twee Provincin WA Case C-
285/92 [1993] ECR I-6045 (paras 1415).

22.
I do not consider that it would detract from the information and protection of consumers to permit
indication of the manufacturer or packager of the product established in a non-member country.
23.
More specifically, art 3(1)(6) of the directive
is mainly intended to enable the consumer to contact a person responsible for the manufacture
or packaging of the foodstuff with a view to expressing any positive or negative criticism about the
product purchased (see the European Commissions answer to written question E-2170/95; OJ
1995 C340 p 19).
24.
In order to ensure that that aim is achieved, as the Commission observes, the final consumer must be
able easily to identify the person responsible for the 257 product. That is why the Community legislature
chose expressly to separate the three main traders involved in bringing the foodstuff from production to
distribution, without giving more weight to one than to the others. However, that decision logically involves
drawing some distinction between those traders.
25.
On the one hand, manufacturers and packagers have in common the fact that they are generally
established, easily identifiable traders who can therefore be contacted without difficulty. Those
characteristics mean that they satisfy the conditions necessary for achieving the aim pursued.
26.
However, contrary to what has been argued by the Italian authorities and the Greek government,
subjecting those traders to the requirement of Community establishment contained in art 3(1)(6) would
have, at the very least, surprising consequences.
27.
The two categories of trader referred to by the directive play a unique role in the chain whereby the
foodstuff is made available to the final consumer 14. Therefore, if the labelling were required to indicate
their establishment within the Community, access to the Community market would effectively be restricted
to foodstuffs manufactured or packaged within the territory of the Community. That can most certainly not
have been the intention of the legislature.
14
Moreover, the French version of art 3(1)(6) uses the words du fabricant and du conditionneur (del in Italian; the in
English), but to un vendeur (un in Italian; a in English). Du is a definite article (more specifically, a contracted definite
article: a contraction of the preposition de and the definite article le), whereas un is an indefinite article.

28.
If, none the less, it were desired to follow that interpretation without entailing such a consequence, the
text at issue could be understood as systematically requiring reference to the establishment within the
Community of one of the three traders mentioned, so that reference to a seller established within the
Community would be required if the product were not manufactured or packaged within the Community.
But in that case how is it possible to explain the legislatures failure to take up the wording proposed by
the Economic and Social Committee in its opinion?
29.
When a seller is indicated, however, the requirement of Community establishment is justified with
regard to the aim pursued. Whilst the directive only refers to the manufacturer or to the packager, a
number of different sellers may be involved in marketing a product. Therefore, the requirement that the
seller indicated on the label must be established within the Community does not amount to a requirement
that only Community products may be distributed within the territory of the Union, in contrast to what
could be inferred if the same requirement were to apply to the producer or to the packager. Furthermore,
as the Commission notes, sellers are, by their nature, less established and less easily identifiable than
manufacturers or packagers and the requirement that they should be established within the Community
makes it possible to limit the drawbacks of such a circumstance.
30.
In the light of those considerations, the court should not accept that art 3(1)(6) of the directive requires
the expression established within the Community to refer to the manufacturer or to the packager.

Conclusion
31.
To conclude, I propose that the question referred by the Corte Suprema di Cassazione should be
answered as follows:
258
Article 3(1)(6) of Council Directive (EEC) 79/112 on the approximation of the laws of the
member states relating to the labelling, presentation and advertising of foodstuffs for sale to the
ultimate consumer is to be interpreted as meaning that the expression established within the
Community used in it refers only to the seller, and not to the manufacturer or packager, details of
whom may appear on the labelling, even if they are established outside the Communities.

17 September 1997.

The COURT OF JUSTICE (First Chamber)

delivered the following judgment.
1.
By order of 4 December 1995, received at the Court of Justice of the European Communities on 18
March 1996, the Corte Suprema di Cassazione (the Supreme Court of Cassation) referred to the Court of
Justice for a preliminary ruling under art 177 of the EC Treaty a question on the interpretation of art 3(1)
(6) of Council Directive (EEC) 79/112 on the approximation of the laws of the member states relating to
the labelling, presentation and advertising of foodstuffs for sale to the ultimate consumer (OJ 1979 L33 p
1).
2.
That question was raised in proceedings between, on the one hand, the Provincia Autonoma di Trento
(the Autonomous Province of Trento) and the Ufficio del Medico Provinciale di Trento (the Medical Office
of the Province of Trento) (the administrative authorities of the Province of Trento) and, on the other, the
partnership Dega di Depretto Gino (Dega) concerning the latters failure to comply with the Italian
legislation on labelling.
3.
Dega markets cans of pineapple in Italy, on which the labels state only the name and address of the
producer and packager, an undertaking established outside the Community.
4.
On 13 June 1988, an administrative fine was imposed on Dega on the ground that those labels did not
bear the details of a trader established within the Community.
5.
In support of that decision, the administrative authorities of the province of Trento relied on art 3(h) of
Decree No 322 of the President of the Republic of 18 May 1982 (GURI No 156 of 9 June 1982, p 4167)
according to which the labelling on foodstuffs must indicate, in particular
the name or business name or registered trademark and the registered office of the
manufacturer or packager or of a seller established within the European Community.
6.
That article transposes art 3(1)(6) of the directive into Italian law. The English version of that provision,
which agrees with the French version, reads as follows:
1. In accordance with Articles 4 to 14 and subject to the exceptions contained therein, indication
of the following particulars alone shall be compulsory on the labelling of foodstuffs (6) the name
or business name and address of the manufacturer or packager, or of a seller established within
the Community.
In contrast to the French and English texts, the Italian version of the directive does not include a comma
between the words packager and or of a seller established within the Community.
7.
By judgment of 20 November 1990, the Pretura (the Magistrates Court), Rovereto, upheld the appeal
brought by Dega against the decision of the 259 administrative authorities of the province of Trento and
annulled the administrative fine. It held that the expression established within the European Economic
Community in art 3(h) of Decree No 322 referred only to sellers, and that, as in this case, it was sufficient
to indicate the name and address of the producer and packager established in a third country.
8.
 By application of 3 June 1992, the administrative authorities of the Province of Trento lodged an
appeal seeking to have the Preturas judgment set aside. They claimed that protection of the ultimate
consumer is only fully guaranteed if details of at least one trader established within the Community are
indicated on the label, whether of the manufacturer, the packager, or the seller.
9.
As the national provision at issue reproduces almost exactly art 3(1)(6) of the directive, the Corte
Suprema di Cassazione considered it necessary to refer the following question to the Court of Justice for
a preliminary ruling:
Must Article 3(1)(6) of Council Directive 79/112/EEC (on the approximation of the laws of the
member states relating to the labelling, presentation and advertising of foodstuffs for sale to the
ultimate consumer) be interpreted as meaning that the expression established within the
Community used in it refers only to the seller or refers, in the absence of a seller established within
the Community, also to the manufacturer and/or packager? Must that provision therefore be taken
to mean that, in the absence of a seller established in the Community, the manufacturer and/or the
packager must be established in the Community?
10.
The Italian and Greek governments consider that the labels of products referred to by the directive
must always bear details of a trader established within the Community, whether of the producer, the
packager or the seller. In their view, the word established is capable of referring without distinction to all
the operators mentioned in the directive.
11.
Furthermore, they consider that interpretation to be in accordance with the aim of the directive, which
is to inform and protect consumers. Article 3(1) is intended to identify traders responsible for any
misrepresentation, food-borne infections or other injury caused by the product so that penalties can be
more easily imposed and compensation more easily sought. If the packaging of the products does not
bear details of a natural or legal person established in the Community, the protection of consumers in that
respect is severely jeopardised.
12.
That interpretation cannot be accepted.
13.
Firstly, it is apparent from the French, Danish, English, German and Dutch language versions of the
directive that the expression established within the Community applies only to the seller. In the French,
Danish and English versions, the position of the comma serves to separate the seller from the other two
traders. Furthermore, that separation is reinforced in the English version by the fact that the word seller
is preceded by the indefinite article a, whilst the words manufacturer and packager are preceded by
the definite article the. Finally, the nature of German and Dutch syntax makes it even more clear that the
expression established within the Community applies only to the seller (den Namen oder die Firma und
die Anschrift des Herstellers, des Verpackers oder eines in der Gemeinschaft niedergelassenen
Verkufers, de naam of de handelsnaam en het adres van de fabrikant of van de verpakker of van een in
de Gemeenschap gevestigde verkoper).
14.
Secondly, it should be noted that the final version of the directive does not take up the punctuation
proposed by the Economic and Social Committee in its 260 opinion on the draft directive, which was
precisely intended to apply the condition of Community establishment to all the traders mentioned in the
provision at issue (see OJ 1976 C285 p 3, point 2.7.1).
15.
Thirdly, art 3(1)(6) of the directive must be interpreted with reference to the purpose and general
scheme of the rules of which it forms a part (see the judgments in EC Commission v UK Case 100/84
[1985] ECR 1169 (para 17), Rockfon A/S v Specialarbeijderforbundet i Danmark Case C-449/93 [1995]
ECR I-4291 (para 28) and Aannemersbedrijf PK Kraaijeveld B v Gedeputeerde staten van Zuid-Holland
Case C-72/95 [1996] ECR I-5403 (para 28)).
16.
In that respect, it is apparent from both the sixth recital in the preamble and from art 2 that the directive
was intended to ensure the information and protection of the ultimate consumer of foodstuffs, in particular
as regards the nature, identity, properties, composition, quantity, durability, origin or provenance and
method of manufacture or production of those products.
17.
In particular, art 3(1)(6) of the directive
is mainly intended to enable the consumer to contact a person responsible for the manufacture
or packaging of the foodstuffs with a view to expressing any positive or negative criticism about the
product purchased. (See the Commissions answer to written question E-2170/95; OJ 1995 C340 p
19.)
18.
That goal can only be achieved if the ultimate consumer can easily identify the person responsible for
the product. In that respect, producers and packagers differ from sellers. The former are generally
established, easily identifiable traders, meaning that the fact that they might be situated outside the
Community does not present a problem. In contrast, sellers are generally much smaller traders and,
consequently, more difficult to identify, particularly if they are established outside the Community.
19.
That is why the Community legislature, for the purposes of the rules on labelling of foodstuffs, laid
down different rules in relation to traders depending on whether they are manufacturers or packagers, on
the one hand, or sellers, on the other. As regards manufacturers and packagers, the label on the
packaging may indicate details of either one or the other, whether established within the Community or
outside; as regards sellers, the label may only indicate details of a trader established within the
Community.
20.
The answer to the national courts question must therefore be that art 3(1)(6) of the directive is to be
interpreted as meaning that the expression established within the Community used in it refers only to the
seller.

Costs
21.
The costs incurred by the Italian, Greek and French governments and by the European Commission,
which have submitted observations to the Court of Justice, are not recoverable. Since these proceedings
are, for the parties to the main proceedings, a step in the action pending before the national court, the
decision on costs is a matter for that court.
On those grounds, the Court of Justice (First Chamber), in answer to the question referred to it by the
Corte Suprema di Cassazione by order of 4 December 1995, hereby rules: art 3(1)(6) of Council Directive
(EEC) 79/112 on the approximation of the laws of the member states relating to the labelling, presentation
and advertising of foodstuffs for sale to the ultimate consumer is to be interpreted as meaning that the
expression established within the Community used in it refers only to the seller.

261

[1998] All ER (EC) 262

Dorsch Consult Ingenieurgesellschaft mbH v Bundesbaugesellschaft

Berlin mbH
(Case C-54/96)
EUROPEAN COMMUNITY; Other European Community
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), MANCINI, MOITINHO DE ALMEIDA, MURRAY AND
SEVN (PRESIDENTS OF CHAMBERS), KAKOURIS, KAPTEYN, GULMANN, EDWARD,
PUISSOCHET, HIRSCH, JANN (RAPPORTEUR), RAGNEMALM, WATHELET AND SCHINTGEN
ADVOCATE GENERAL TESAURO
28 JANUARY, 15 MAY, 17 SEPTEMBER 1997
European Community Reference for a preliminary ruling Admissibility Meaning of national court or
tribunal Dispute concerning award of public service contract National public procurement awards
supervisory board referring question to Court of Justice for preliminary ruling Whether supervisory
board a national court or tribunal Whether reference admissible EC Treaty, art 177.

The plaintiff undertaking submitted a tender for a public contract requiring architectural and engineering
services. On learning that its tender had been unsuccessful, the plaintiff applied to the appropriate review
body in order to have the procedure stopped, contending that the awarding authority had breached the
relevant Community and domestic provisions coordinating public procurement award procedures. The
review body declined jurisdiction on the ground that domestic public procurement legislation had not yet
been extended to cover contracts for services as required under Council Directive (EEC) 90/50
coordinating procedures for the award of public service contracts. The plaintiff subsequently applied for a
determination to the Federal Public Procurement Awards Supervisory Board, contending that in so far as
the Community provisions concerning public service contracts had not been transposed into German
legislation, the Community provisions were directly applicable and had to be complied with by the review
body. The supervisory board stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling a question on the interpretation of Directive 90/50. However, a
preliminary issue arose as to whether the supervisory board which was set up by the
Haushaltsgrundstzgesetz (the Budget Principals Law), in which it was described as a quasi-judicial
body, was to be regarded as a court or tribunal within the meaning of art 177 1 of the Treaty and therefore
whether the reference was admissible.
1
Article 177, so far as material, provides: Where a question is raised before any court or tribunal of a Member State,
that court or tribunal may request the Court of Justice to give a ruling

Held The question whether a body making a reference was a court or tribunal for the purposes of art
177 of the Treaty was governed by Community law, which required that the court take account of various
factors including: whether the body was established by law; whether it was permanent; whether its
jurisdiction was compulsory; whether its procedure was inter partes; whether it applied rules of law and
whether it was independent. In the instant case, since the supervisory board was established by the
Haushaltsgrundstzgesetz, its establishment by law 262 and its permanence could not be disputed.
Moreover, under German legislation the board was the only body with the power to review the legality of
determinations made by a review body, which, following a ruling that a determination was unlawful, would
be directed to make a fresh determination in conformity with the boards findings; the legislation also
required the board to apply provisions of Community directives and the domestic provisions transposing
them, and provided, inter alia, that the principal rules on the appointment and independence of the
judiciary applied to the boards official and lay members. It followed that the board fulfilled all of the
required criteria, except that of an inter partes procedure, which was not in any case an absolute criterion.
Accordingly, in the procedure which led to the instant reference, the supervisory board was a court or
tribunal for the purposes of art 177, with the result that the question referred was admissible (see p 284 e
f j, p 285 b to j and p 286 c, post).
Vaassen (ne Gbbels) v Management of the Beambtenfonds voor het Mijnbedrejf Case 61/65 [1966]
ECR 261, Pretore di Sal v Persons Unknown Case 14/86 [1987] ECR 2545, Handels- og
Kontorfunktionrernes Forbund i Danmark v Dansk Arbejds- giverforening acting on behalf of Danfoss
Case 109/88 [1989] ECR 3199, Municipality of Almelo v Energiebedrijf IJsselmij NV Case C-393/92
[1994] ECR I-1477 and Non-contentious proceedings brought by Job Centre Coop arl Case C-111/94
[1995] ECR I-3361 applied.

Notes
For Community provisions on courts which may refer, see 51 Halsburys Laws (4th edn) para 2181.
For the EC Treaty, art 177 (as amended by art G.56 of the Treaty on European Union), see 50
Halsburys Statutes (4th edn), Current Service 98.

Cases cited
Almelo (Municipality of) v Energiebedrijf IJsselmij NV Case C-393/92 [1994] ECR I-1477.
Borker Case 138/80 [1980] ECR 1975.
Bozetti v Invernizzi SpA Case 179/84 [1985] ECR 2301.
Broekmeulen v Huisarts Registratie Commissie Case 246/80 [1981] ECR 2311.
Corbiau v Administration des Contributions du Grand-Duch de Luxembourg Case C-24/92 [1993] ECR I-
1277.
Dillenkofer v Germany Joined cases C-178179/94 and C-188190/94 [1996] All ER (EC) 917, [1996]
ECR I-4845, ECJ.
Direccin General de Defensa de la Competencia v Asociacin Espaola de Banca Privada (AEB) Case
C-67/91 [1992] ECR I-4785.
European Commission v Germany Case C-253/95 [1996] ECR I-2423.
European Commission v Germany Case C-433/93 [1995] ECR I-2303.
Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325, ECJ.
Greis Unterweger (Criminal proceedings against) Case 318/85 [1986] ECR 955.
Hagen OHG v Einfuhr- und Vorratsstelle fr Getreide und Futtermittel Case 49/71 [1972] ECR 23.
Handels- og Kontorfunktionrernes Forbund i Danmark v Dansk Arbejdsgiverforening acting on behalf of
Danfoss Case 109/88 [1989] ECR 3199.
Job Centre Coop arl (Non-contentious proceedings brought by) Case C-111/94 [1995] ECR I-3361.
263
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135.
Pretore di Sal v Persons Unknown Case 14/86 [1987] ECR 2545.
SEIMSociedade de Exporto e Importo de Materiais Ld v Subdirector-Geral das Alfndegas Case C-
446/93 [1996] ECR I-73.
Simmenthal SpA v Amministrazione delle Finanze dello Stato Case 70/77 [1978] ECR 1453.
Vaassen (ne Gbbels) v Management of the Beambtenfonds voor het Mijnbedrejf Case 61/65 [1966]
ECR 261.
Wagner Miret v Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911.

Reference
By order of 5 February 1996, the Vergabeberwachungsausschu des Bundes (the Federal Public
Procurement Awards Supervisory Board) referred to the Court of Justice of the European Communities
for a preliminary ruling under art 177 of the EC Treaty a question (set out at p 281 e, post) on the
interpretation of art 41 of Council Directive (EEC) 92/50 relating to the coordination of procedures for the
award of public service contracts. The question was raised in proceedings between Dorsch Consult
Ingenieurgesellschaft mbH (Dorsch Consult) and Bundesbaugesellschaft Berlin mbH concerning a
procedure for the award of a service contract. An initial question arose as to the admissibility of the
reference. Written observations were submitted on behalf of: Dorsch Consult, by F G Siebeck,
Rechsanwalt, Munich; the German government, by E Rder, Ministerialrat at the Federal Ministry for
Economic Affairs, and B Kloke, Oberregierungsrat at the same ministry, acting as agents; and the
European Commission, by H van Lier, Legal Adviser, and C Schmidt, of its Legal Service, acting as
agents. Oral observations were made by Dorsch Consult, the German government and the Commission.
The language of the case was German. The facts are set out in the opinion of the Advocate General.

15 May 1997.

The Advocate General (G Tesauro)

delivered the following opinion (translated from the Italian).
1.
The question referred to the Court of Justice of the European Communities for a preliminary ruling in
this case has been submitted by the Vergabeberwachungsausschu des Bundes (the Federal Public
Procurement Awards Supervisory Board) and concerns the interpretation of art 41 of Council Directive
(EEC) 92/50 relating to the co-ordination of procedures for the award of public service contracts (OJ 1992
L209 p 1) (the services directive).
The Federal Supervisory Board has asked the Court of Justice to determine whether that article
means that the bodies set up by the member states for the purposes of Council Directive (EEC) 89/665
on the co-ordination of the laws, regulations and administrative provisions relating to the application of
review procedures to the award of public supply and public works contracts (OJ 1989 L395 p 33) (the
review directive), are competent, as from the final date for transposition stipulated in the services directive
(and where no national implementing measures have been taken), also to review procedures for the
award of public service contracts where infringements of the relevant provisions of Community law are
alleged.
264

RELEVANT COMMUNITY AND NATIONAL LEGISLATION

2.
In order better to understand the point of this question, it is first necessary to place it in its proper legal
context, by briefly looking at the relevant provisions of both Community and national law.

Community legislation
3.
Article 36 of the services directive lays down the criteria which a contracting authority is required to
follow in awarding a contract. In particular, art 36(1)(a) provides that, where the award is made to the
economically most advantageous tender, the contracting authority must take into account
various criteria relating to the contract: for example, quality, technical merit, aesthetic and
functional characteristics, technical assistance and after-sales service, delivery date, delivery
period or period of completion, price
In other cases, sub-para (b) provides that the lowest price only is to be taken into account.
Article 44 of the services directive sets 1 July 1993 as the date by which member states are required
to adopt and communicate the necessary transposition measures.
4.
The review directive requires member states to take the measures necessary to ensure that award
procedures for public works contracts and public supply contracts governed by the relevant Community
directives2 may be reviewed rapidly and effectively where the grounds of alleged illegality involve (directly
or indirectly) Community law (art 1(1)).
2
These are, of course, contracts covered by Council Directives (EEC) 71/305 (OJ 1971 L185 p 5) and (EEC) 77/62 (OJ
1977 L13 p 1), now Council Directive (EEC) 93/37 (OJ 1993 L199 p 54) and Council Directive (EEC) 93/36 (OJ 1993
L199 p 1) respectively.

Following the entry into force of the services directive, the review directive also applies to procedures
for the award of service contracts; art 41 of the services directive, which the referring body is now asking
the court to interpret, amended the wording of art 1 of the review directive to extend its scope to include
the review of service contract awards.
5.
Article 2(8) of the review directive provides that, where bodies responsible for review procedures are
not judicial in character, written reasons for their decisions must be given. In such a case, the member
states must also guarantee that
any allegedly illegal measure taken by the review body or any alleged defect in the exercise of
the powers conferred on it can be the subject of judicial review or review by another body which is
a court or tribunal within the meaning of Article 177 of the EEC Treaty and independent of both the
contracting authority and the review body.
In order to ensure the independence of such bodies, the second sub-paragraph of art 2(8) further
requires that their members be subject to the same conditions as members of the judiciary as regards the
authority responsible for their appointment, their period of office and their removal; and that at least the
president shall have the same legal and professional qualifications as members of the judiciary. Finally,
the sub-paragraph provides that the body in question is to 265 adopt its decisions following a procedure in
which both sides are heard and that its decisions are to be legally binding.
 The national legislation
6.
The Community public procurement directives were transposed into German domestic law by means
of an amendment of the Haushalts- grundstzegesetz (the Budget Principles Law) (the HGrG). In
particular, the Second Law Amending the HGrG (BGBl 1993, I, pp 1928ff), which came into effect in 1993,
inserted new paras 57a to 57c, which were intended by the German legislature to give effect to the
directives on the award of public works contracts and public supply contracts, and to the review directive 3.
3
This legislative technique is commonly referred to as the budget solution by German academic writers.

The services directive, however, has not been transposed into German law, and this is confirmed by
the order for reference.
7.
Paragraph 57a of the HGrG constitutes the general legal basis for the adoption of the measures
implementing the Community directives on public procurement. It confers power on the Federal
government to issue regulations, with the assent of the Bundesrat, governing the award of public supply
contracts, public works contracts and public services contracts, which are put out to tender by the bodies
listed in the said paragraph.
The ensuing provisions (paras 57b and 57c), which deal with the remedies available in the case of
infringement of rules of Community law (or of national provisions transposing them), set up a two-tier
review procedure for this purpose. Initial recourse lies to the award review bodies (Vergabeprfstellen)
whilst their decisions may in turn be challenged before the supervisory boards
(Vergabeberwachungsausschsse) set up by each of the Lnder, or, in cases where the impugned
contract-award procedure falls within a federal authoritys province, before the Federal Supervisory
Board.
8.
Paragraph 57b of the HGrG regulates, in particular, the operation of the review bodies. According to
sub-para (2) thereof, the terms of reference of these bodies are to be determined by the federal
government by means of regulations and with the prior assent of the Bundesrat. Sub-paras (3), (4) and
(5) contain a series of provisions concerning the commencement of the review procedure, the suspension
of award procedures adjudged suspect and the information which applicants are required to provide to the
review bodies.
Sub-para (6) provides that the lodging of an application with a review body does not preclude the right
of the individual concerned to bring an action in the ordinary courts for damages for loss suffered as a
result of an award procedure.
9.
Paragraph 57c of the HGrG lays down a series of rules applicable to the supervisory boards 4; the
members of such boards, the composition of which is regulated in detail in sub-paras (2), (3) and (4) of
this provision, perform their functions independently and on their own responsibility. In particular, para
57c(3) provides that certain provisions of the Deutsches Richtergesetz (the Law on the Judiciary) (the
DRiG) are to apply by analogy to the official members of the supervisory boards as regards annulment or
withdrawal of their appointment and their independence and dismissal 5.
4
These rules also apply to the Federal Supervisory Board, save for the specific provisions contained in para 57c(7).
5
The relevant paragraphs of the DRiG are:para 18(1) and (2), which specifies the circumstances in which the
appointment of a judge is void (appointment by an authority acting outside its powers, appointment of a person who is
not a German national or is not qualified to hold public office); however, para 18(3) (which provides that the nullity of an
appointment may not be relied upon until declared in a court decision having the force of res judicata) is not applicable
and in its place para 57c(3) provides that the nullity of an appointment may not be relied upon until it has been declared
by the authority which made the appointment and that decision has become final; para 19(1) and (2), which deals with
the cases in which appointments can be withdrawn. Paragraph 19(3), which makes withdrawal of an appointment
subject to the consent of the party concerned or a court decision having the force of res judicata, does not apply,
however; para 26(1) and (2), which provides that administrative supervisory control over members of the judiciary
may not limit their independence;para 27(1), judges are assigned to a particular court); para 30(1) and (3), paras 31
to 33, and para 37, which lay down the conditions for removing a judge from his office or for transferring him; in general,
this may happen pursuant to formal disciplinary proceedings or on the grounds of court restructuring. It is to be noted
that para 30(2), which provides that removal from office or transfer other than for organisational reasons shall require a
court order having the force of res judicata, does not apply.

266
The supervisory boards review only the legality of the award procedures and do not examine the
findings of fact on which the determinations of review bodies are based; they may, where appropriate, set
aside the determinations of review bodies and direct them to make fresh determinations (para 57c(5)).
Any person claiming that provisions governing the award of contracts have been infringed may make an
application to a supervisory board (sub-para (6)).
10.
Finally, para 57c(7) lays down specific rules applicable to the Federal Supervisory Board. Its official
members are selected from the chairmen and assessors serving in the decision making departments of
the Bundeskartellamt (the national administrative authority responsible for competition matters). The
chairmen of the chambers of the supervisory board are appointed from the chairmen of the
Bundeskartellamt decision-making departments. The president of the Bundeskartellamt appoints lay
assessors6 on a proposal from the top public-law trade boards, decides on the formation and composition
of chambers and exercises administrative supervisory control on behalf of the Federal government.
 6
As well as the official members (Bundeskartellamt personnel), the Federal Supervisory Board also has outside or lay
assessors. At present it is composed of a single chamber, presided over by a Bundeskartellamt department chairman,
and has four official members, five lay members and five alternate lay members (see Stockmann Die
Vergabeberwachung des Bundes WUW (1995) pp 572ff; the author is the president of the Federal Supervisory Board).

This sub-para also provides that the Federal Supervisory Board is to adopt its own internal rules of
procedure to regulate the allocation and conduct of cases.
11.
On the basis of the provisions described above, the Federal government adopted two implementing
regulations on 22 February 1994, after having obtained the assent of the Bundesrat. The first regulation,
governing the award of public procurement contracts, expressly applies only to public works contracts and
public supply contracts and does not therefore apply to public service contracts (see BGBl 1994, I, pp
321ff).
The second regulation, however, is of general application and contains provisions fulfilling the
obligation to provide for appropriate review procedures in the field of public procurement (the review
regulation)7. Paragraph 1 of this regulation designates the bodies competent to conduct reviews in
respect of each of the awarding authorities listed in para 57a of the HGrG. Under para 2, the review body
has power to suspend the award procedure; its determinations are 267 to be given in writing, contain a
statement of reasons and be notified to the awarding authority and to the person claiming an infringement
of the procurement rules; the review body must draw the latters attention to the possibility of challenging
the determination before a supervisory board and specify the supervisory board competent to hear such a
challenge.
7
See BGBl 1994, I, pp 324ff

12.
Finally, para 3 of the review regulation deals with the operation of the supervisory boards. It provides
that the supervisory boards must make a reference to the Court of Justice under art 177 when they
consider that a ruling on the interpretation of the EC Treaty or on the validity or interpretation of legal acts
adopted on the basis of the Treaty is necessary.
It also provides that the supervisory boards are to adopt internal rules of procedure in the light of the
principles set out in the law amending the HGrG. They must issue reasoned determinations in writing
after having heard the parties. Finally, para 3(4) provides that the supervisory boards, unlike the review
bodies, shall not be empowered to suspend a procedure for the award of a contract.
13.
The rules of procedure of the Federal Supervisory Board came into effect on 1 August 1995 and have
not been published. The version produced by the German government in these proceedings consists of
five sections regulating the organisation and allocation of cases within the board, the conduct of
procedure, which includes a written stage and an oral hearing, and its decisions and other technical
matters, including formalities relating to final determinations.
This version of the internal rules of procedure seems to be an amended version of rules which came
into effect in June 1994 and was likewise not published. According to the European Commission, which
supplied this information without challenge at the hearing, there are a number of differences between the
original version and the one now in force, in particular as regards the openness and necessity of hearings
and procedural time limits8.
8
The Commission indicated in fact that it had doubts as to which version of the rules of procedure was to be regarded as
in force. It explains that the only version officially submitted by the German government which has not been followed by
any official revision or amendment is that of June 1994, which was produced as an official document in the course of an
enforcement action brought by the Commission against Germany under art 169.

14.
To complete this survey of the legal background it should be noted that the transposition of the
Community public procurement directives has been the subject of two recent judgments of this court, in
which it found that the German government had failed to comply with its obligations under the Treaty. The
first judgment concerned the incorrect transposition of the directives on the award of public works
contracts and public supply contracts9 and the second concerned the failure to transpose the services
directive (see the judgments in European Commission v Germany Case C-433/93 [1995] ECR I-2303 and
European Commission v Germany Case C-253/95 [1996] ECR I-2423 respectively).
9
The directives applicable at the time of the relevant facts were Council Directive (EEC) 88/295 in respect of public
supply contracts (OJ 1988 L127 p 1) and Council Directive (EEC) 89/440 in respect of public works contracts (OJ 1989
L210 p 1). In line with its established case law, the court confirmed the breach of obligations by reference to the legal
position existing at the expiry of the period set by the Commission in its reasoned opinion (in this case, 3 February
1993).

As regards transposition of the review directive, the Commission has commenced enforcement
proceedings which are still pending10. The 268 Commission contends, essentially, that in the national
measures transposing the directive the German government has provided individuals with less judicial
protection than the corresponding directive.
10
The letter of formal notice dated 31 December 1995 has been published in German, in Zeitschrift fr Wirtschaftsrecht
und Insolvenzpraxis, N 23/95, pp 1940ff. The reasoned opinion in the case was delivered on 29 July 1996 (see
Commission press release No IP/96/614).
FACTS
15.
Unlike the legislative background, the facts which led to the main proceedings are straightforward and
may be summarised as follows.
In 1995, the Bundesbaugesellschaft Berlin mbH (the contracting authority) issued an invitation to
tender for the award of a general planning services contract relating to new government buildings in
Berlin11. As the services in question were of an intellectual nature, the contracting authority opted to use a
negotiated procedure with prior publication of the contract notice, as permitted by art 11(2)(c) of the
services directive.
11
Published in the Amtsblatt Berlin of 23 June 1995 and in OJ 1995 S120 p 166.

16.
Dorsch Consult Ingenieurgesellschaft mbH (the applicant) took part in the tendering procedure and
submitted its tender on 25 August 1995.
The contracting authority examined the 18 tenders received and drew up a short list of seven. It then
decided to award the contract to two of the other firms which had submitted a tender, which were required
to form a working party to provide the services in question. The contract was signed on 12 January 1996,
after the working party had already commenced its work.
17.
The applicant took the view that the elimination of its tender constituted a breach of the services
directive and of the relevant national legislation and applied to the Federal Ministry for Regional Planning,
Building and Urban Planning (in its capacity as the competent review body) seeking, by way of interim
relief, to have the contract awarding procedure suspended and, by way of primary relief, to be awarded
the contract. In support of its claim, the applicant contended that it had been repeatedly informed by the
contracting authority that it was technically competent to perform the contract and that its tender was the
most attractive in terms of price.
By letter of 20 December 1995, the review body declined jurisdiction and dismissed the application
without consideration of the merits of the claim. The decision was based on the grounds that the federal
regulation provided for under the legislation, by which the Federal government was to have extended the
jurisdiction of the review bodies to include the hearing and determination of disputes concerning service
contracts, had still not been enacted.
18.
In its notification to the applicant of the outcome of its application, the review body also informed it of
its right to challenge the legality of the decision before the Federal Supervisory Board. The applicant
thereupon made an application to that Board for the setting aside of the review bodys decision to decline
jurisdiction, the suspension of the contract awarding procedure and the award of the contract to the
applicant; alternatively, it asked for a reference to be made to the Court of Justice for a preliminary ruling
on the point in issue.
The Federal Supervisory Board decided to stay the proceedings and to refer to the Court of Justice the
question whether art 41 of the services directive is to be interpreted as meaning that the bodies set up by
the member states for the purposes of the review directive are also competent, from the date by which the
services directive ought to have been transposed into national law, to review procedures for the award of
public service contracts.
269

ADMISSIBILITY
19.
Before considering the substance of the question referred, it is necessary to determine whether the
Court of Justice has jurisdiction to entertain the reference made by the Federal Supervisory Board. From
the foregoing survey of the legislation establishing the board and of the rules governing its procedure
serious doubts arise as to whether that body can be regarded as a court or tribunal within the meaning of
art 177 of the Treaty and, accordingly, whether the reference is admissible 12.
12
The sole purpose of this examination is, incidentally, to determine whether the subjective conditions for acceptance of
the reference are satisfied and not to establish whether the review directive was properly transposed, which may
eventually be the subject of separate proceedings. Of course, the enforcement action brought by the Commission
against the German government for failure correctly to transpose the review directive, which I mentioned earlier, has
points in common with the present case; but there are also many dissimilarities, so that a rigorous distinction between
the two cases should be maintained, with regard to both form and substance.

The issue was the subject of lively argument between the parties, both in their written observations
and at the hearing. It is noteworthy that the applicant itself, which had expressly requested (albeit as an
alternative relief) a reference to be made to the Court of Justice for a preliminary ruling on the point in
issue, accepted that the Federal Supervisory Board does not constitute a court or tribunal within the
meaning of art 177; it submitted none the less that the court should in any case answer the question
referred to it, butnot without self-contradictionon grounds relating to the effective protection of
individual rights by the courts.
The Commission considered that the question was so clearly inadmissible that it felt it unnecessary to
address its substance. The German government, for its part, argued that the body in question does
constitute a court or tribunal within the meaning of art 177. But it explicitly conceded in the course of the
hearing that it had begun the process of amending the relevant provisions to allow, inter alia,
determinations of the supervisory boards to be challenged in the ordinary courts in order to ensure the
effective protection by the courts of the rights of the persons concerned 13.
13
This point was raised at the hearing by the Commission, which saw it as further proof that the supervisory boards as
presently constituted are not courts; this was disputed by the German government, which at the hearing argued that the
amendments in question were aimed solely at making a number of improvements to a system already providing
satisfactory legal protection, and it pointed out that this information had been supplied in the course of other, separate
proceedings.

20.
My first observation in considering this issue is that, in German domestic law itself, the Federal
Supervisory Board (as well as the supervisory boards of the Lnder) are described as quasi-judicial
bodies (gerichtshnliche Einrichtungen) and not as courts or tribunals strictu sensu 14.
14
See the preamble to the Second Law Amending the HGrG, which inserted new paras 57a to 57c, discussed above (BT-
Drucksache 12/4636, p 12). The designation is not surprising if one has regard to the particularities of German
constitutional law. Were a fully-fledged Federal court to have been established by means of an ordinary law (such as the
HGrG) this would have been in breach of the relevant constitutional provisions (at least as regards the Federal
Supervisory Board); this is because the German Basic Law (arts 95 and 96) contains an exhaustive list of all Federal
courts, any addition to which would require a constitutional amendment.

This fact, whilst not being conclusive on its own since the concept of court or tribunal within the
meaning of art 177 is a term of Community law15 within the audit of which the Court of Justice has seen fit
(on occasion) to include bodies which were not so regarded in the eyes of their own national law, none
the less 270 calls for a detailed analysis of the nature of the body in question and the manner in which it is
required to carry out its functions, in order to ascertain whether it possesses those organisational and
functional characteristics which the court has in previous cases held to be necessary in order for a body
which is not a court to still be able to fall within the scope of art 177 (see eg Vaassen (ne Gbbels) v
Management of the Beambtenfonds voor het Mijnbedrejf Case 61/65 [1966] ECR 261).
15
The definition of which, in the obvious interests of the uniform application of Community law, cannot be left to the
discretion of the courts of the member states (see in general Hagen OHG v Einfuhr- und Vorratsstelle fr Getreide und
Futtermittel Case 49/71 [1972] ECR 23).

21.
I shall therefore begin by briefly reviewing those leading decisions of the court in this area which are
relevant here, but it should be borne in mind that the case law developed by reference to the individual
cases that have come before the court has not led to a general, exhaustive definition of the concept of
court or tribunal within the meaning of art 177.
22.
The first case concerned a Dutch industrial arbitration tribunal (Scheidsgericht) which made a
reference to the Court of Justice for a preliminary ruling even though it stated that it did not consider itself
to be a judicial body under Dutch law. In the now landmark judgment of Vaassen (ne Gbbels), the court
decided that it had jurisdiction to rule on the questions submitted to it, having found that the referring body
in question possessed the characteristics of a court or tribunal within the meaning of art 177.
The court expressly took the following factors into account: the Scheidsgericht was a body duly
established under Dutch law; it was permanent; it was charged with the settlement of disputes and had to
follow rules of inter partes procedure similar to those applying in the ordinary courts of law; it was required
to apply rules of law; furthermore, all those belonging to the relevant industry had to bring any dispute
with their insurer before the Scheidsgericht; finally, the members of the body in question were appointed
by the minister responsible, who also designated its chairman and laid down its rules of procedure 16.
16
In more recent decisions the court also confirmed that employment arbitration tribunals which satisfy the aforementioned
criteria are courts or tribunals within the meaning of art 177. See eg the judgment in Handels- og Kontorfunktionrernes
Forbund i Danmark v Dansk Arbejdsgiverforening acting on behalf of Danfoss Case 109/88 [1989] ECR 3199, in which
the court found that the body making the reference, a Danish industrial arbitration board, had been established by law
(which laid down detailed rules governing its composition, the number of members to be nominated by the parties and
the manner of appointment of the umpire), had exclusive and final jurisdiction over the relevant disputes and could hear
a case brought by either party irrespective of the objections of the other. Advocate General Lenz also pointed out that
the board was also required to apply rules of law, such as the provisions of the relevant collective agreements.

23.
By adopting this approach, the court thus made it clear from the outset that in deciding whether a
referring body is a court or tribunal for the purposes of the Treaty it does not attach importance to its
formal designation but considers its substantive characteristics (establishment by law, permanence,
compulsory jurisdiction, transparent rules of procedure and the application of rules of law). This approach
was entirely justified, especially in view of the historical context in which it evolved. For at the time of the
Vaassen (ne Gbbels) case the mechanism of co-operation between national courts and the Court of
Justice had only just begun to operate and the Court of Justice was very mindful of the need to encourage
the use of the mechanism in order to ensure the spread and uniform application of Community law, with
the aidif necessaryof a broad interpretation of the category of bodies entitled to make references to it.
24.
A number of subsequent judgments should also be read in this light, like that in, for example,
Broekmeulen v Huisarts Registratie Commissie Case 246/80 [1981] ECR 2311 (discussed in greater
detail in paras 39 and 40, below), in which 271 the court held to be admissible a question submitted to it
by a (Netherlands) appeals committee which heard appeals from medical practitioners who had been
refused authorisation to practise or enrolment on the medical register. In this case, too, the court found
that the appeals committee possessed a number of organisational and functional characteristics which
warranted it being treated as having a judicial function. The court took into account the fact that the
appeals committee concerned was permanent, that the public authorities were involved in deciding its
composition, that it had internal rules of procedure providing for inter partes procedure, that its jurisdiction
was exclusive and that its determinations were final. Given also the fact that it was called upon to apply
Community law, which had been pleaded by the applicants in the main proceedings, the court held that it
was necessary, in the interest of the practical effect of Community law, to answer the question submitted.
25.
The line of judgments beginning with Simmenthal SpA v Amministrazione delle Finanze dello Stato
Case 70/77 [1978] ECR 1453 is to be viewed in the same light. In that case, a reference was made to the
court by the Preture di Alessandria in proceedings for an interlocutory order. The Italian government
contested the jurisdiction of the court to reply to the questions submitted by the Preture on the grounds
that the procedure was not inter partes, pointing out that the judge in question had power in the course of
it to make a determination based solely on the plaintiffs submissions. Having found that the Preture was
exercising the functions of a court or tribunal within the meaning of Article 177, the court held that the
Pretures capacity to make a reference for a preliminary ruling could not depend on whether or not the
proceedings in which the reference was made were defended; it did, however, add that
it may where necessary prove to be in the interests of the proper administration of justice that a
question should be referred for a preliminary ruling only after both sides have been heard.
In other words, the court established that, whenever a referring body is unquestionably a court or tribunal,
the fact that a reference is made before any inter partes hearing does not render it inadmissible (see
[1978] ECR 1453 (paras 910)).
26.
While adopting this broad interpretation, the court has none the less set clear limits to the concept of
court or tribunal within the meaning of the Treaty. In its order in Borker Case 138/80 [1980] ECR 1975
(para 4), subsequently confirmed in Criminal proceedings against Greis Unterweger Case 318/85 [1986]
ECR 955, the court held that it had jurisdiction to give preliminary rulings only on questions submitted by a
court or tribunal called upon to give judgment in proceedings intended to lead to a decision of a judicial
nature. The referring bodies in question in those cases were, respectively, the Paris Bar Council (which
had been requested by a lawyer on its register to issue a declaration to be produced as evidence in legal
proceedings pending before the courts of another member state) and the Italian Consultative Committee
for Currency Offences (whose function was to give reasoned, non-binding opinions to the Italian
Treasury), and in both cases the court found that this condition was not satisfied 17.
17
See also, on this point, the recent judgment in Non-contentious proceedings brought by Job Centre Coop arl Case C-
111/94 [1995] ECR I-3361, which concerned voluntary proceedings involving an application for approval of a companys
memorandum of association with a view to its registration. The Court of Justice held that the Tribunale di Milano, which
made the reference, was in this instance performing the functions of an administrative authority rather than those of a
judicial body.

27.
In addition, the Court of Justice subsequently held that, in order to qualify as a court or tribunal within
the meaning of art 177, the body making the reference must be independent. This criterion, perhaps
because it goes to the very essence of the judicial function, was explicitly identified for the first time 272
only in Corbiau v Administration des Contributions du Grand-Duch de Luxembourg Case C-24/92 [1993]
ECR I-1277, in which the court declined jurisdiction on the ground that the body making the reference,
although a court under national law, did not, in the courts view, offer the necessary guarantees of
impartiality between parties to disputes which it was called upon to resolve 18. That case involved
Luxembourgs Director of Taxation and Excise who had jurisdiction under the law to hear at first instance
disputes between taxpayers and the departments (of which he was director) which had charged them to
tax.
18
There had, in fact, already been some fairly explicit references to the criterion of independence in previous decisions:
see eg the judgment in Pretore di Sal v Persons Unknown Case 14/86 [1987] ECR 2545 (para 7), in which the Court of
Justice had regard, among other factors, to the referring courts independence in reaching the conclusion that it
constituted a court or tribunal within the meaning of art 177.

The criterion of independence also appears to have been a key factor, albeit with the opposite result,
in Direccin General de Defensa de la Competencia v Asociacin Espaola de Banca Privada (AEB)
Case C-67/91 [1992] ECR I-4785. In that case the body making the reference was Spains Tribunal de
Defensa de la Competencia, which Advocate General Jacobs, in his opinion, found to present a number
of characteristics constitutive of a court or tribunal within the meaning of art 177; these included the
adversarial nature of the procedure which was clearly laid down by law, the independent exercise by its
members of their functions and the fact that its members could not be removed from office (see the
opinion in AEB [1992] ECR I-4785 (para 11)). In its judgment the court did not specifically address the
issue; but the fact that it replied to the questions submitted indicates that the court implicitly endorsed the
view of the Advocate General.
28.
These decisions therefore clearly show that, even in the absence of a general definition of the concept
of court or tribunal within the meaning of art 177, the court has developed a number of tests which must
be satisfied in order for a body to be entitled to make a reference for a preliminary ruling.
 These tests concern the manner of establishment of the body, which must have been established by
law and not by agreement between the parties; its connection to the exercise of public authority; its
permanent nature, in the sense that it must not exercise a judicial function only on an occasional basis; its
competence to resolve a dispute by a decision of a judicial nature; the conduct, before it, of a procedure
analogous to that which is followed in ordinary courts of law, involving (within the limits discussed above)
exchange of argument inter partes; the application by the body in question of rules of law (rather than
principles of fairness); compulsory jurisdiction, which means that alternative remedies are not available;
and finally independence, in the sense that it acts as a third party in relation to the parties to the dispute
and that its members may not be removed from office.
29.
To return to the case in hand, it is now therefore necessary to establish whether the Federal
Supervisory Board possesses the characteristics allowing it to be regarded as performing a judicial
function, as required by the Court of Justice for the purposes of art 177 of the Treaty.
273
As already mentioned, the Commission takes the view that the Federal Supervisory Board does not
satisfy any of the tests laid down by the court in the cases referred to above. Its main argument is that the
Federal Supervisory Board was established by a framework law (the Second Law amending the HGrG),
which does not impose obligations or confer rights on individuals and which must be supplemented by
regulations; moreover, the body in question could easily be deprived of its legal basisand thereby of its
capacity to give judgment in cases such as the present case where there is no competent review body at
first instance. The Commission also makes these points: that the referring body does not make its
determinations following an inter partes procedure, as is confirmed in the grounds of its first decision 19;
that its proceedings are governed by internal rules of procedure which have not been published and
which may be amended autonomously at any time; that there is no legislative provision for its
determinations to have binding legal effect; that it is not an independent body, since it is linked to the staff
and organisational structure of the Bundeskartellamt, which is itself an administrative rather than a judicial
body; and that the minimum term of office of its official members and of its chairman is not fixed by law.
19
See the decision of 2 August 1994 (published in EU Public Contract Law, No 3/94, pp 47ff), in which the Federal
Supervisory Board stated that it did not conduct an inter partes procedure and that the parties applications were only
requests for a particular determination.

30.
In view of the nature of the Federal Supervisory Board, the legislative technique by which it was
established and, above all, the provisions governing the way in which it functions, I must confess that I
agree with at least some of the Commissions observations, which I also consider to be particularly
important.
To begin with, I do not believe that the rules governing the review procedure before the supervisory
boards can be regarded as comparable to the rules governing procedure before ordinary courts of law.
On the contrary, the fact that, under the legislation, rules of procedure are to be adopted autonomously by
each supervisory board, which may subsequently amend them autonomously, and that in addition there is
no requirement that they be published, leads me to conclude that the degree of transparency and legal
certainty required in any judicial process is not guaranteed here.
31.
I am not only referring here to the absence of any inter partes procedure, which has now been proved
in practice: of far greater significance, to my mind, is the absence of the minimal functional requirements
which characterise judicial proceedings, as found in Vaassen (ne Gbbels) [1966] ECR 26120. In that
case, as I have explained, the rules of procedure governing proceedings before the referring body were
subject to the approval of the minister responsible, so that there could be no doubt as to the certainty,
transparency and ascertainability of the procedural rules applicable. That fact, which was indeed
expressly mentioned was taken into consideration both by the Advocate General and by the court in
arriving at the conclusion that the body in question in that case was bound by rules of adversary
procedure similar to those used by the ordinary courts of law (my emphasis)21.
20
Of course, in attaching less importance, in Simmenthal (within the limits referred to above), to the specific requirement
of inter partes procedure, the court certainly did not intend to dispense with the more general requirement for the
procedure to be of a judicial nature. It is in fact a fundamental requirement which played a decisive role not only in
Vaassen (ne Gbbels), where it was expressly addressed, but also, as we have seen, in the reasoning underlying the
other decisions of the court on this issue.
21
See Vaassen (ne Gbbels) [1966] ECR 261 (para 1), and the opinion of Advocate General Gand in the same case,
where he states: The procedure which is followed is of a judicial nature.

274
32.
In the present case, however, I find it difficult to see similar procedural safeguards; if they do exist,
they are subject to opaque autonomous amendment by the decision making body and this seems to me
to run counter to the most basic requirements of legal certainty. The point is borne out, indeed, by the
Commissions doubts, referred to above, as to the version of the rules of procedure now in force due to
the discrepancies between the version provided by the German government in the course of the
enforcement proceedings and the version produced in this case.
In these circumstances, I do not consider that the review proceedings conducted before the Federal
Supervisory Board can be regarded as having the character of judicial proceedings as required by the
court. In this regard, the present case falls clearly outside even the generously broad parameters laid
down in Vaasen (ne Gbbels).
33.
There are also serious doubts, in my opinion, regarding the independence of the Federal Supervisory
Board, at least as regards the question of unremoveability of its members from office.
Of significance in this regard is para 57c(7) of the HGrG, which I shall recapitulate for the sake of
convenience: the official members of the Federal Supervisory Board are Bundeskartellamt department
chairmen and assessors, with the former acting as chairmen of the chambers of the board. The president
of the Bundeskartellamt appoints the lay assessors, decides on the formation and composition of
chambers and exercises administrative supervisory control by delegation from the government 22. The
Federal Supervisory Board also uses the Bundeskartellamts facilities and services.
22
This control is limited to reproach (Vorhalt) and reprimand (Ermahnung) and consequently, according to legal writers,
should in no case concern the content of judicial determinations. The case law appears to confirm this view, although
there are exceptions in cases of purportedly manifest error. See also, in this regard, paras 26(1) and 26(2) of the DRiG,
which are applicable to the body in issue and which provide that administrative supervisory control over the actions of
members of the judiciary cannot limit their independence. Paragraph 26(3), which gives judges the right to challenge
administrative supervisory measures addressed to them, does not, however, apply to official board members (para
57c(3)).

34.
In other words the official members of the board are also members of the Bundeskartellamt and
formally remain on its staff. In practice, this means that they simultaneously perform the functions of
Federal Supervisory Board members and those of Bundeskartellamt members. Moreover, the legislation
establishing the body in question does not include any provision as to the term of office of its official
members, and the fact that the term of office of the lay members is fixed (at five years) 23 suggests that the
omission was not inadvertent.
23
See the HGrG, para 57c(2).

What all this amounts to, in effect, is that not only do the members of the Federal Supervisory Board
enjoy no guarantee against dismissal, but neither do they have the assurance of a fixed term of office,
which is an essential prerequisite of independence. On the contrary, they can be relieved of their
additional duties and reassigned to their ordinary duties at any moment and by means of purely internal
organisational measures. While it is true, as we have seen, that some of the provisions of the DRiG
regarding the permanence and independence of members of the judiciary apply by analogy to the
members of the board when acting in that capacity, it is also the case that the latter are not covered by
the 275 DRiG provisions which give members of the judiciary the right to challenge their removal from
office or reassignment, with the result that they may be freely dismissed at any time by the president of
the Bundeskartellamt.
35.
Nor do I believe that the petitio principii contained in para 57c(3), which provides that the members of
the board are to be independent and unremovable, is sufficient to justify taking a different view, since it is
contradicted by the fact that the board members belong to the administrative authority and continue to
belong to it, even from a functional point of view. Such a system, under which a limited number of
administrative officials are, temporarily and for the performance of specific functions, given the title of
judge and then made subject to an equally limited number of provisions applicable to members of the
judiciary, but excluding the safeguards which ordinarily apply to the judiciary in relation to removal from
office and reassignment, appears to me to be too complicated and too intransparent to guarantee in
practice the stability required to ensure the independence of those performing judicial functions.
36.
One can have further doubts about the specific question of the impartiality of the Federal Supervisory
Board in relation to disputes falling within its area of jurisdiction. This body is, as we have seen, part of the
Bundeskartellamt, which is part of the public administration, but it is given power, in spite of this, to
adjudicate in disputes involving public procurement awards, that is to say in disputes between the public
administration itself and citizens. This fact alone would make it impossible to regard the Federal
Supervisory Board as acting as a third party, thus independently. Unless, of course, one regards judicial
independence as a moral quality of the actual persons who sit on the bench (see the judgment in Corbiau
[1993] ECR I-1277)24.
24
It is true that in that case the referring body was linked to the very departments which had made the disputed tax
assessment: however, the rationale of the courts judgment (and of the Advocate Generals opinion) is not unlike the
approach which should prevail in the present case, since the Federal Supervisory Board is, after all, an integral part of
the public administration and thus is not a third party in relation to disputes between the administration and citizens.

37.
Lastly, the Commission put forward a further telling argument with which I would agree. The legislation
establishing the Federal Supervisory Board makes no provision concerning the legal effects of its
determinations, especially their binding force. Since the body in question is one which, under national law,
is not a court, the general principle that all judicial determinations are binding does not apply. So, in the
absence of express provisions, the fact that the body in question was established using the so-called
budget solution, with the declared aim of not creating individual rights for those taking part in public
tendering procedures25, gives rise to doubts as to the binding nature of its decisions 26.
25
See the explanatory memorandum to the Draft Amending Law to the HGrG (BT-Drucksache 12/4636, p 12).
26
See the doubts expressed by legal writers, in particular by Boesen, (1996) EuZW 586, who points out that the decisions
of the Federal Supervisory Board are not enforceable; see also, on the same point, the letter of formal notice and the
 reasoned opinion sent by the Commission to the German government in the enforcement proceedings referred to
above.

This is a factor which produces further doubts as to the judicial nature of the decisions which the board
is called upon to take and thus as to whether it can be regarded as having the attributes of a court or
tribunal within the meaning of art 177.
38.
So, in view of all of the points I have made, I consider that the Federal Supervisory Board does not
satisfy the requirements, certainly as far as 276 procedural safeguards and guarantees of independence
are concerned, for it to qualify as a court or tribunal within the meaning of art 177 and therefore that its
reference is inadmissible.
39.
It could be argued, on the other hand, that when a reference is made by a body which offers the only
legal remedy available to an individual relying on Community law, the court should accept the reference in
any event, in order to prevent the applicant from being deprived of an effective remedy and to ensure the
uniform application of Community law.
This is, in substance, the argument put forward by the applicant. As mentioned above, even though
the applicant submits that the Federal Supervisory Board is not a court or tribunal within the meaning of
art 177, it suggests that the court should turn a blind eye to this and nevertheless answer the question
submitted by the board on the ground that not to do so would be detrimental to the applicant.
This proposal might be supported, it claims, by the judgment in Broekmeulen [1981] ECR 2311 (para
17) in which, as explained above, the court accepted a reference from a professional body having the
power to hear appeals concerning the registration of members of the profession and held, inter alia:
in the absence, in practice, of any right of appeal to the ordinary courts, the Appeals
Committee, which operates with the consent of the public authorities and with their cooperation,
and which, after an adversarial procedure, delivers decisions which are in fact recognized as final,
must, in a matter involving the application of Community law, be considered as a court or tribunal of
a Member State within the meaning of Article 177 of the Treaty.
40.
I cannot subscribe to the applicants argument; the ratio of Broekmeulen should not be stretched too
far. The subjective and objective conditions for the functioning of the system of co-operation between
national courts and the Court of Justice, which was created by art 177 of the Treaty, cannot vary in
accordance with the particular circumstances of each case. If a body is not a judicial body, it does not
become one simply because there is no better solution. To hold otherwise and interpret Broekmeulen to
that effect would mean conferring crucial importance on an aspect of the procedural system of which the
body in question is part rather than on features of the body itself, so that it would no longer matter
whether the requirements expressly laid down by the court were satisfied or not.
41.
In any event, even with the best of will to make concessions, the circumstances of the present case
are altogether different in this respect. The decisive factor in Broekmeulen was, as quoted above, that in
practice there was no right of appeal to the ordinary courts for a citizen relying on a point of Community
law.
In the present case, however, a person who considers himself to have been unlawfully excluded from
a contract award procedure is expressly given the possibility of bringing an action in damages for any loss
suffered, which in itself could well provide, at least in principle, a satisfactory remedy.
42.
Moreover, even in the absence of any express legislative provision and despite the doubts expressed
by legal writers, ordinary German courts seem to have come round to the view that they have jurisdiction
to hear cases brought by participants in contract award procedures for alleged infringements of the
relevant provisions, including those of Community law. This is borne out by two recent decisions in which
the Kammergericht (the Appeal Court) of Berlin ruled 277 admissiblebefore dismissing them on the
meritsapplications for interlocutory relief by tenderers who had been excluded from public tendering
procedures and were seeking the suspension of the award procedures. 27

27
See Kammergericht Berlin, decisions of 10 April 1995 (KartU 7605/94 (1995) EuZW 645ff) and of 31 May 1995 (KartU
3259/95 (1996) NVwZ 415ff).

Furthermore, as I have already mentioned, the German government has notified the Commission that
it has commenced the process of amending its legislation to bring it into conformity with the review
directive; the new rules will make express provision for, inter alia, review by the ordinary courts of the
determinations of the supervisory boards.28
28
I would point out that the legislative changes which are being enacted could mean that the ruling which the court is now
called upon to give will have only historical significance, in relation, that is, to the issue of admissibility, concerning the
question whether or not the existing supervisory boards are entitled to make a reference for a preliminary ruling.

43.
So, as matters stand, it would not only be entirely in line with the case law of the Court of Justice but
also pose no problem for the effective judicial protection of individual rights, in the sense explained above,
if the Federal Supervisory Board were held not to be a court or tribunal within the meaning of art 177.
 On the contrary, I hold the view that the solution which I have advocated, that is to say that the court
should declare that it has no jurisdiction to rule on this reference, offers a wider perspective going beyond
the present case and affords a greater safeguard of individual rights, for which only a court of law can
provide effective protection. Underlying this conclusion is, quite clearly, the conviction that only those
bodies which are able to provide all the safeguards of individual rights developed by the Court of Justice
can be treated as courts or tribunals for the purposes of art 177, and no others.
44.
Lastly, one final consideration, based on the underlying purpose of the review directive, should not go
unmentioned. As is well known, this directive was adopted to meet a strongly felt need to raise and make
uniform the level of judicial protection of individual rights in the field of public procurement. In some
member states, neither the award itself nor the other related administrative acts were capable of being
challenged in a court of law, or if they were, then with unsatisfactory implications for subsequent contract
award procedures. It was in order to remedy these very defects that the review directive introduced the
obligation for member states to put in place a system capable of effectively ensuring the vindication of the
substantive rights conferred by the relevant Community instruments (the directives on public works
contracts, public supply contracts and public service contracts) on those taking part in public procurement
procedures. This is the light in which the provisions of the directive should be read and, according to the
Commission at least, in which the other member states have implemented them until now. At the hearing
the Commission produced a document providing an overview of the bodies to which the individual
member states have given jurisdiction in the matter of public procurement awards in order to transpose
the review directive: the majority of member states have designated the ordinary courts or, in those
member states where they exist, the administrative courts, subject to the appellate jurisdiction of the
Council of State29. This is a significant factor which, in my view, should be given due 278 weight, whilst
observing the distinction between this case and the enforcement action brought by the Commission under
art 169 of the Treaty.
29
The Commission has thus declared itself satisfied with the transposition measures adopted by all the member states
(with the exception, of course, of the Federal Republic of Germany, against which it has commenced art 169
proceedings).

45.
Having regard to all the considerations set forth above, I propose, in conclusion, that the reference for
a preliminary ruling should be declared inadmissible on the ground that the body which made it is not a
court or tribunal within the meaning of art 177 of the Treaty.

SUBSTANCE
46.
On the substance of the reference, which I shall consider solely for the sake of completeness, a few
remarks will suffice.
The question is, as I have stated, whether, after expiry of the period for transposition of the services
directive, the review bodies are also competent to review procedures for the award of public service
contracts in the absence of any express measure conferring such jurisdiction upon them (in this case, a
Federal regulation, although this is provided for by statute).
47.
Both the applicant and the German government argue that, since the relevant provisions of the
services directive must be regarded as having direct effect 30, the bodies set up for the purposes of the
review directive should also be able to adjudicate in disputes in relation to public service contracts.
30
Such direct effect, besides deriving from the sufficiently precise and unconditional character of the provisions in issue,
was, it is argued, confirmed by the court, albeit indirectly, in the judgment in European Commission v Germany Case C-
253/95 [1996] ECR I-2423, where it was held that the German government had failed in its obligation to transpose the
services directive.

Clearly, however, this is an issue which cannot be determined by the court in these proceedings. For
the court may not take the place of the national legislature, to which the relevant power has been
expressly reserved by law, and decide whether the review bodies should also review procedures for the
award of public service contracts.
48.
Relying on the direct effect of the provisions of the services directive does not change matters. Even if
the court were to find that the relevant provisions are indeed directly effective, this would merely mean
that an individual had the right to rely on those provisions before a court; under no circumstances could it
go so far as to indicate before which court that should be, for this would encroach on the domain of the
national legislature31.
31
See, on this very point, Bozetti v Invernizzi SpA Case 179/84 [1985] ECR 2301 (para 17) and the more recent judgment
in SEIMSociedade de Exporto e Importo de Materiais Ld v Subdirector-Geral das Alfndegas Case C-446/93
[1996] ECR I-73 (para 32), where it was held: it is for the legal system of each Member State to determine which
court has jurisdiction to hear disputes involving individual rights derived from Community law, but at the same time the
Member States are responsible for ensuring that those rights are effectively protected in each case. Subject to that
reservation, it is not for the court to intervene in order to resolve questions of jurisdiction to which the classification of
certain legal situations based on Community law may give rise in the national judicial system.

If an individual had no actual possibility of relying on a directly effective provision of Community law for
want of a court competent to hear his case, this would, of course, indicate the existence of a clear
violation of Community law32. Such a violation could, of course, be pursued by the competent authorities
using the procedures provided for in such cases and could also bring into play the remedies which the
court has established in the area of state liability towards individuals who have suffered material loss as a
result of the failure of the state in 279 question to fulfil its obligations under Community law. But, to repeat,
these are remedies which, both in form and in substance, are distinct from the procedure now in point and
therefore have no bearing on the solution which I have proposed in this case.
32
And, in all likelihood, a breach of art 6(1) of the Convention for the Protection of Human Rights and Fundamental
Freedoms (Rome, 4 November 1950; TS 71 (1953); Cmd 8969), which enshrines the right of access to a court of law.

49.
In view of the considerations set out above, I therefore propose that the court should declare the
reference for a preliminary ruling inadmissible, on the ground that the Federal Public Procurement Awards
Supervisory Board, which made the reference, is not a court or tribunal within the meaning of art 177 of
the Treaty.

17 September 1997.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 5 February 1996, received at the Court of Justice of the European Communities on 21
February 1996, the Vergabeberwachungs- ausschu des Bundes (the Federal Public Procurement
Awards Supervisory Board) referred to the Court of Justice for a preliminary ruling under art 177 of the EC
Treaty a question on the interpretation of art 41 of Council Directive (EEC) 92/50 relating to the co-
ordination of procedures for the award of public service contracts (OJ 1992 L209 p 1).
2.
The question has been raised in proceedings between Dorsch Consult Ingenieurgesellschaft mbH
(Dorsch Consult) and Bundesbaugesellschaft Berlin mbH (the awarding authority) concerning a
procedure for the award of a service contract.
3.
On 28 June 1995 the awarding authority published in the Official Journal of the European
Communities a notice advertising the award of a contract for architectural and construction engineering
services. On 25 August 1995 Dorsch Consult submitted its tender to the awarding authority. In all, 18
tenders were received, of which seven, including that of Dorsch Consult, were chosen for further
consideration. On 30 November 1995, two companies, together with an architect, were chosen to form a
working party to perform the services which were the subject of the contract. The contract itself was
signed on 12 January 1996. Dorsch Consult was informed on 25 January 1996 that its tender was not the
most advantageous economically.
4.
Having learned that the awarding authority had not chosen it for the contract but before its tender was
formally rejected, Dorsch Consult had applied, on 14 December 1995, to the Bundesministerium fr
Raumordnung, Bauwesen und Stdtebau (the Federal Ministry for Regional Planning, Building and Urban
Planning), as the body responsible for reviewing public procurement awards (Vergabeprfstelle), seeking
to have the contract awarding procedure stopped and the contract awarded to it. It considered that, in
concluding the contract with another undertaking, the awarding authority had acted in breach of both
Directive 92/50 and para 57a(1) of the Haushaltsgrundstzegesetz (the HGrG) (the Budget Principles
Law). By decision of 20 December 1995, the review body held that it had no competence in the matter on
the ground that, under paras 57a and 57b of the HGrG, it had no power to review the award of contracts
when they related to services.
5.
In those circumstances, on 27 December 1995 Dorsch Consult lodged an application for a
determination by the Federal Supervisory Board on the ground that the review body had wrongly declined
jurisdiction. It stated that, in so far as Council Directive (EEC) 89/665 on the coordination of the laws,
regulations and administrative provisions relating to the application of review procedures to the 280 award
of public supply and public works contracts (OJ 1989 L395 p 33) had not been transposed, it was directly
applicable and had to be complied with by the review bodies.
6.
The Federal Supervisory Board found that the Federal Republic of Germany had not yet transposed
Directive 92/50. Although a circular had been issued by the Federal Ministry for Economic Affairs on 11
June 1993 stating that the directive was directly applicable and that it had to be applied by the
administration, it could not be regarded as a proper transposition of the directive. According to the Federal
Supervisory Board, where public service contracts are concerned, German domestic law does not
empower a review body to determine whether the provisions governing public procurement have been
complied with. It is quite possible that the provisions of Directive 92/50 have direct effect. Finally, the
Federal Supervisory Board is unsure whether, by virtue of art 41 of Directive 92/50, the competence of
existing review bodies also applies directly to the award of public service contracts.
7.
The Federal Supervisory Board therefore suspended proceedings and referred the following question
to the Court of Justice:
 Is Article 41 of Council Directive 92/50/EEC of 18 June 1992 to be interpreted to the effect that,
after 30 June 1993, the bodies set up by the Member States which, under Council Directive
89/665/EEC of 21 December 1989, are competent to review procedures for the award of public
contracts falling within the scope of Directives 71/305/EEC and 77/62/EEC are also competent to
review procedures for the award of public service contracts within the meaning of Directive
92/50/EEC in order to determine whether alleged infringements of Community public procurement
law or of domestic rules enacted in implementation of that law have taken place?

Legal background
8.
The purpose of Directive 92/50 is to regulate the award of public service contracts. It applies to
contracts having a value above a certain limit. As far as the matter of legal protection is concerned, art 41
provides:
Article 1(1) of Council Directive 89/665/EEC shall be replaced by the following: 1. The
Member States shall take the measures necessary to ensure that, as regards contract award
procedures falling within the scope of Directives 71/305/EEC, 77/62/EEC, and 92/50/EEC,
decisions taken by the contracting authorities may be reviewed effectively and, in particular, as
rapidly as possible in accordance with the conditions set out in the following Articles and, in
particular, Article 2(7) on the grounds that such decisions have infringed Community law in the field
of public procurement or nation[al] rules implementing that law.
9.
In accordance with art 44(1), Directive 92/50 had to be transposed by the member states before 1
June 1993.
10.
Article 2(8) of Directive 89/665 provides:
Where bodies responsible for review procedures are not judicial in character, written reasons
for their decisions shall always be given. Furthermore, in such a case, provision must be made to
guarantee procedures whereby any allegedly illegal measure taken by the review body 281 or any
alleged defect in the exercise of the powers conferred on it can be the subject of judicial review or
review by another body which is a court or tribunal within the meaning of Article 177 of the EEC
Treaty and independent of both the contracting authority and the review body.
The members of such an independent body shall be appointed and leave office under the same
conditions as members of the judiciary as regards the authority responsible for their appointment,
their period of office, and their removal. At least the President of this independent body shall have
the same legal and professional qualifications as members of the judiciary. The independent body
shall take its decisions following a procedure in which both sides are heard, and these decisions
shall, by means determined by each Member State, be legally binding.
11.
Directive 89/665 was transposed into German law by a Law of 26 November 1993 (BGBl I, p 1928),
which supplemented the HGrG by adding paras 57a to 57c.
12.
Paragraph 57a(1) of the HGrG provides:
In order to meet obligations arising from directives of the European Communities, the Federal
government shall regulate, by means of regulations, with the assent of the Bundesrat, the award of
public supply contracts, public works contracts and public service contracts and the procedures for
awarding public service contracts
13.
Paragraph 57b(1) of the HGrG makes provision for the procedures for awarding public supply
contracts, public works contracts and public service contracts mentioned in para 57a(1) to be reviewed by
review bodies. Under para 57b(2), the Federal government is to adopt, in the form of regulations, with the
assent of the Bundesrat, the provisions governing the competence of those review bodies. According to
sub-para (3), a review body must initiate a review procedure if there is evidence of a breach of
procurement rules applicable under a regulation adopted pursuant to para 57a. In particular, it must
initiate that procedure where a person who has, or had, an interest in a particular contract claims that the
aforementioned provisions were contravened.
14.
According to para 57b(4) of the HGrG, the review body must determine whether the provisions
adopted pursuant to para 57a have been complied with. It may compel the awarding authority to annul
unlawful measures or decisions or to take lawful measures or decisions. It may also provisionally suspend
a procedure for the award of a contract. Under para 57b(5), a review body may require the awarding
authority to provide the information necessary for it to carry out its task. Sub-para (6) provides that actions
for damages in the event of breach of the provisions applicable in relation to the award of contracts are to
be brought before the ordinary courts.
15.
Paragraph 57c(1) of the HGrG provides that the Federation and the Lnder must each establish a
supervisory board, performing its functions independently and on its own responsibility, to supervise
procedures for the award of contracts in the fields concerning them. According to sub-paras (2), (3) and
(4) of that provision, each supervisory board is to sit in chambers composed of a chairman, an official
assessor and a lay assessor, who are to be independent and subject only to observance of the law. The
chairman and one of the assessors must be public officials. As regards annulment or withdrawal of their
appointment and their independence and dismissal, various provisions of the Richtergesetz (the Law on
282 the Judiciary) apply by analogy. As regards the annulment or withdrawal of a lay members
appointment, certain provisions of the Richtergesetz also apply by analogy. Where a lay member commits
a serious breach of his duties, his appointment must be annulled. The term of office of a supervisory
boards lay members is five years.
16.
Under sub-para (5), the supervisory board is to determine the legality of determinations made by
review bodies but it does not review the way in which they ascertain the facts. Where a determination is
found to be unlawful, the supervisory board directs the review body to make a fresh determination taking
account of its own legal findings. Paragraph 57c(6) of the HGrG provides that any person claiming that
the provisions governing the award of public contracts have been infringed may make application to the
supervisory board within a period of four weeks following the relevant determination of the review body.
17.
Paragraph 57c(7) of the HGrG establishes a Federal Supervisory Board
(Vergabeberwachungsausschu des Bundes). Its official members are the chairman and assessors from
the decision making departments of the Bundes- kartellamt (the Federal Cartel Office). The president of
the Bundeskartellamt decides on the composition of the Federal Supervisory Board and the formation and
composition of its chambers. He appoints lay assessors and their deputies on a proposal from the leading
public-law trade boards. He also exercises administrative supervisory control on behalf of the Federal
government. The Federal Supervisory Board adopts its own internal rules of procedure.
18.
Pursuant to para 57a of the HGrG the Federal government adopted a regulation on the award of
contracts. This regulation is, however, applicable only to supply contracts and works contracts and not to
contracts for services. Directive 92/50 has not yet been transposed by the Federal Republic of Germany.
19.
Pursuant to paras 57b and 57c of the HGrG, the Federal government has adopted a regulation on the
procedure for review of public procurement awards (the Verordnung ber das Nachprfungsvesfahren fr
ffentliche Vertrge; BGBl I 1994, p 324). Paragraph 2(3) of the regulation provides:
The review bodys determination regarding the awarding authority shall be given in writing,
contain a statement of reasons and be notified without delay. The review body shall send without
delay to the person claiming that there has been a breach of public procurement provisions the text
of its determination, shall draw attention to the possibility of making application for a determination
by the supervisory board within a period of four weeks and shall name the competent supervisory
board.
20.
Paragraph 3 provides:
(1) Procedure before the Public Procurement Awards Supervisory Board shall be governed by
Paragraph 57c of the Haushaltsgrundstzegesetz and by this regulation according to the rules of
procedure which the board shall adopt. (2) The Public Procurement Awards Supervisory Board
shall be obliged, under art 177 of the Treaty establishing the European Community, to make a
reference to the Court of Justice of the European Communities when it considers that a preliminary
ruling on a question relating to the interpretation of that Treaty or to the validity and interpretation of
a legal act adopted on that basis is necessary in order to enable it to make its determination. (3)
Before a chamber makes any determination, the parties to the procedure before the procurement
review body shall be heard. (4) A 283 chamber shall not be empowered to suspend a procedure for
the award of a contract or to give other directions concerning a procedure for the award of a
contract. (5) A chamber shall reach its determination by an absolute majority of votes.
Determinations shall be in writing, contain a statement of reasons and shall be sent to the parties
without delay.
21.
The rules of procedure of the Federal Supervisory Board regulate the organisation and allocation of
cases and the conduct of procedure, which consists of a hearing to which the persons concerned are
called, and the conditions governing determinations of the Federal Supervisory Board.

Admissibility
22.
Before the question submitted by the national court is addressed, it is necessary to examine whether
the Federal Supervisory Board, in the procedure which led to this reference for a preliminary ruling, is to
be regarded as a court or tribunal within the meaning of art 177 of the Treaty. That question must be
distinguished from the question whether the Federal Supervisory Board fulfils the conditions laid down in
art 2(8) of Directive 89/665, which is not in point in this case.
23.
In order to determine whether a body making a reference is a court or tribunal for the purposes of art
177 of the Treaty, which is a question governed by Community law alone, the court takes account of a
number of factors, such as whether the body is established by law, whether it is permanent, whether its
jurisdiction is compulsory, whether its procedure is inter partes, whether it applies rules of law and
whether it is independent (see esp the judgments in Vaassen (ne Gbbels) v Management of the
Beambtenfonds voor het Mijnbedrejf Case 61/65 [1966] ECR 261, Pretore di Sal v Persons Unknown
Case 14/86 [1987] ECR 2545 (para 7), Handels- og Kontorfunktionrernes Forbund i Danmark v Dansk
Arbejdsgiverforening acting on behalf of Danfoss Case 109/88 [1989] ECR 3199 (paras 78), Municipality
of Almelo v Energiebedrijf IJsselmij NV Case C-393/92 [1994] ECR I-1477 and Non-contentious
proceedings brought by Job Centre Coop arl Case C-111/94 [1995] ECR I-3361 (para 9)).
24.
As regards the question of establishment by law, the European Commission states that the HGrG is a
framework budgetary law which does not give rise to rights or obligations for citizens as legal persons. It
points out that the Federal Supervisory Boards action is confined to reviewing determinations made by
review bodies. However, in the field of public service contracts, there is, as yet, no competent review
body. The Commission therefore concludes that in such matters the Federal Supervisory Board has no
basis in law on which it can act.
25.
It is sufficient to note in this regard that the Federal Supervisory Board was established by para 57c(7)
of the HGrG. Its establishment by law cannot therefore be disputed. In determining establishment by law,
it is immaterial that domestic legislation has not conferred on the Federal Supervisory Board powers in
the specific area of public service contracts.
26.
Nor is there any doubt about the permanent existence of the Federal Supervisory Board.
27.
The Commission also submits that the Federal Supervisory Board does not have compulsory
jurisdiction, a condition which, in its view, may mean two things: either that the parties must be required to
apply to the relevant review body for settlement of their dispute or that determinations of that body are to
be 284 binding. The Commission, adopting the second interpretation, concludes that German legislation
does not provide for the determinations made by the Federal Supervisory Board to be enforceable.
28.
It must be stated first of all that para 57c of the HGrG establishes the supervisory board as the only
body for reviewing the legality of determinations made by review bodies. In order to establish a breach of
the provisions governing public procurement, application must be made to the supervisory board.
29.
Secondly, under para 57c(5) of the HGrG, when the supervisory board finds that determinations made
by a review body are unlawful, it directs that body to make a fresh determination, in conformity with the
supervisory boards findings on points of law. It follows that determinations of the supervisory board are
binding.
30.
The Commission also submits that since, according to the Federal Supervisory Boards own evidence,
procedure before that body is not inter partes, it cannot be regarded as a court or tribunal within the
meaning of art 177 of the Treaty.
31.
It must be reiterated that the requirement that the procedure before the hearing body concerned must
be inter partes is not an absolute criterion. Besides, under para 3(3) of the Verordnung ber das
Nachprfungsverfahren fr ffentliche Auftrge, the parties to the procedure before the procurement
review body must be heard before any determination is made by the chamber concerned.
32.
According to the Commission, the criterion relating to the application of rules of law is not met either,
because, under para 57c of the HGrG and para 3(1) of the Verordnung ber das Nachprfungsverfahren
fr ffentliche Auftrge, procedure before the Federal Supervisory Board is governed by rules of
procedure which it itself adopts, which do not take effect in relation to third parties and which are not
published.
33.
It is, however, undisputed that the Federal Supervisory Board is required to apply provisions governing
the award of public contracts which are laid down in Community directives and in domestic regulations
adopted to transpose them. Furthermore, general procedural requirements, such as the duty to hear the
parties, to make determinations by an absolute majority of votes and to give reasons for them are laid
down in para 3 of the Verordnung ber das Nachprfungsverfahren fr ffentliche Auftrge, which is
published in the Bundesgesetzblatt. Consequently, the Federal Supervisory Board applies rules of law.
34.
Finally, both Dorsch Consult and the Commission consider that the Federal Supervisory Board is not
independent. They point out that it is linked to the organisational structure of the Bundeskartellamt, which
is itself subject to supervision by the Ministry for Economic Affairs, that the term of office of the chairman
and the official assessors is not fixed and that the provisions for guaranteeing impartiality apply only to lay
members.
35.
It must be observed first of all that, according to para 57c(1) of the HGrG, the supervisory board
carries out its task independently and under its own responsibility. According to para 57c(2) of the HGrG,
the members of the chambers are independent and subject only to observance of the law.
36.
 Under para 57c(3) of the HGrG, the main provisions of the Richtergesetz concerning annulment or
withdrawal of their appointments and concerning their independence and removal from office apply by
analogy to official members of 285 the chambers. In general, the provisions of the Richtergesetz
concerning annulment and withdrawal of judges appointments apply also to lay members. Furthermore,
the impartiality of lay members is ensured by para 57c(2) of the HGrG, which provides that they must not
hear cases in which they themselves were involved through participation in the decision making process
regarding the award of a contract or in which they are, or were, tenderers or representatives of tenderers.
37.
It must also be pointed out that, in this particular instance, the Federal Supervisory Board exercises a
judicial function, for it can find that a determination made by a review body is unlawful and it can direct the
review body to make a fresh determination.
38.
It follows from all the foregoing that the Federal Supervisory Board, in the procedure which led to this
reference for a preliminary ruling, is to be regarded as a court or tribunal within the meaning of art 177 of
the Treaty, so that the question it has referred to the court is admissible.

Substance
39.
By its question, the Federal Supervisory Board is asking in effect whether it follows from art 41 of
Directive 92/50 that, if that directive has not been transposed by the end of the period laid down for that
purpose, the appeal bodies of the member states having competence in relation to procedures for the
award of public works contracts and public supply contracts may also hear appeals relating to procedures
for the award of public service contracts.
40.
It must be stated first of all that it is for the legal system of each member state to determine which
court or tribunal has jurisdiction to hear disputes involving individual rights derived from Community law.
However, it is the member states responsibility to ensure that those rights are effectively protected in
each case. Subject to that reservation, it is not for the court to involve itself in the resolution of questions
of jurisdiction to which the classification of certain legal situations based on Community law may give rise
in the national judicial system (see the judgment in SEIMSociedade de Exporto e Importo de
Materiais Ld v Subdirector-Geral das Alfndegas Case C-446/93 [1996] ECR I-73 (para 32)).
41.
Although art 41 of Directive 92/50 requires the member states to adopt the measures necessary to
ensure effective review in the field of public service contracts, it does not indicate which national bodies
are to be the competent bodies for this purpose or whether these bodies are to be the same as those
which the member states have designated in the field of public works contracts and public supply
contracts.
42.
It is, however, common ground that paras 57a to 57c of the HGrG were designed to transpose
Directive 89/665 and that para 57a was to be the basis for the transposition of Directive 92/50 which the
Federal government has still not undertaken.
43.
That being the case, it must be reiterated first of all that the member states obligation arising from a
directive to achieve the result envisaged by the directive and their duty under art 5 of the EC Treaty to
take all appropriate measures, whether general or particular, to ensure fulfilment of that obligation is
binding on all the authorities of member states, including, for matters within their jurisdiction, the courts. It
follows that, when applying national law, whether adopted before or after the directive, the national court
having to interpret that law must do so, as far as possible, in the light of the wording and the purpose of
286 the directive so as to achieve the result it has in view and thereby comply with the third para of art
189 of the EC Treaty (see the judgments in Marleasing SA v La Comercial Internacional de Alimentacin
SA Case C-106/89 [1990] ECR I-4135 (para 8), Wagner Miret v Fondo de Garanta Salarial Case C-
334/92 [1993] ECR I-6911 (para 20) and Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1,
[1994] ECR I-3325 (para 26)).
44.
Secondly, the question of the designation of a body competent to hear appeals in relation to public
service contracts is relevant even where Directive 92/50 has not been transposed. Where a member state
has failed to take the implementing measures required or has adopted measures which do not conform to
a directive, the court has recognised, subject to certain conditions, the right of individuals to rely in law on
a directive as against a defaulting member state. Although this minimum guarantee cannot justify a
member state absolving itself from taking in due time implementing measures sufficient to meet the
purpose of each directive (see esp the judgment in European Commission v Germany Case C-253/95
[1996] ECR I-2423 (para 13)), it may nevertheless have the effect of enabling individuals to rely, as
against a member state, on the substantive provisions of Directive 92/50.
45.
If the relevant domestic provisions cannot be interpreted in conformity with Directive 92/50, the
persons concerned, using the appropriate domestic law procedures, may claim compensation for the
damage incurred owing to the failure to transpose the directive within the time prescribed (see esp the
judgment in Dillenkofer v Germany Joined cases C-178179/94 and C-188190/94 [1996] All ER (EC)
917, [1996] ECR I-4845).
46.
 The answer to be given to the question referred to the court must accordingly be that it does not follow
from art 41 of Directive 92/50 that, where that directive has not been transposed by the end of the period
laid down for that purpose, the appeal bodies of the member states having competence in relation to
procedures for the award of public works contracts and public supply contracts may also hear appeals
relating to procedures for the award of public service contracts. However, in order to observe the
requirement that domestic law must be interpreted in conformity with Directive 92/50 and the requirement
that the rights of individuals must be protected effectively, the national court must determine whether the
relevant provisions of its domestic law allow recognition of a right for individuals to bring an appeal in
relation to awards of public service contracts. In circumstances such as those arising in the present case,
the national court must determine in particular whether such a right of appeal may be exercised before
the same bodies as those established to hear appeals concerning the award of public supply contracts
and public works contracts.

Costs
47.
The costs incurred by the German government and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the proceedings pending before the national tribunal, the
decision on costs is a matter for that body.
On those grounds, Court of Justice, in answer to the question referred to it by the
Vergabeberwachungsausschu des Bundes by order of 5 February 1996, hereby rules: it does not
follow from art 41 of Council Directive (EEC) 92/50 relating to the coordination of procedures for the
award of public service 287 contracts that, where that directive has not been transposed by the end of the
period laid down for that purpose, the appeal bodies of the member states having competence in relation
to procedures for the award of public works contracts and public supply contracts may also hear appeals
relating to procedures for the award of public service contracts. However, in order to observe the
requirement that domestic law must be interpreted in conformity with Directive 92/50 and the requirement
that the rights of individuals must be protected effectively, the national court must determine whether the
relevant provisions of its domestic law allow recognition of a right for individuals to bring an appeal in
relation to awards of public service contracts. In circumstances such as those arising in the present case,
the national court must determine in particular whether such a right of appeal may be exercised before
the same bodies as those established to hear appeals concerning the award of public supply contracts
and public works contracts.

288

[1998] All ER (EC) 289

van der Wal v European Commission

(Case T-83/96)

EUROPEAN COMMUNITY; Information

COURT OF FIRST INSTANCE OF THE EUROPEAN COMMUNITIES (FOURTH CHAMBER)
JUDGES LENAERTS (PRESIDENT), LINDH AND COOKE
25 SEPTEMBER 1997, 19 MARCH 1998
European Community Community institutions Access to information Public access to Commission
documents Applicant requesting access to documents provided by Commission to national courts
Access denied on grounds that disclosure detrimental to protection of public interest as documents
integral part of court proceedings Whether court proceedings exception to general principle of access
applying where Commission not party to proceedings for which documents produced Commission
Decision (ECSC, EC, Euratom) 94/90.

Between 1993 and 1995 three member state courts requested information from the European
Commission concerning the application and interpretation of various aspects of Community competition
rules. The requests were made within the framework of the Commissions Notice 93/C39/05 on co-
operation between national courts and the Commission in applying arts 85 and 86 of the EC Treaty and
were noted in the Commissions 24th Report on Competition Policy. In 1996 the applicant, a lawyer with a
firm which dealt with cases raising questions of competition law at Community level, requested copies of
the Commissions replies. The Commission refused the request on the grounds that the disclosure of the
requested letters would be detrimental to the protection of the public interest in so far as they formed an
integral part of court proceedings. The applicant subsequently brought an action before the Court of First
Instance of the European Communities for the annulment of the Commissions decision refusing access,
on the grounds, inter alia, that the Commission had infringed the Commission Decision (ECSC, EC,
Euratom) 94/90 on public access to Commission documents. The applicant contended, inter alia, that the
exception to the general principle of access in relation to court proceedings applied only to proceedings
to which the Commission was a party and could not therefore be relied on in the instant case.

Held The exception to the general principle of access to Commission documents set out in Decision
94/90, which was based on the protection of the public interest in relation to documents connected with
court proceedings, was designed to ensure respect for the fundamental right of every person to a fair
hearing. The scope of that exception was not therefore restricted solely to the protection of the interests of
the parties in the context of specific proceedings but encompassed the procedural autonomy of the
national and Community courts. The Commission was accordingly entitled to rely on that exception even
when it was not itself a party to the court proceedings which, in the particular case, justified the protection
of the public interest. Moreover, where documents were drafted by the Commission for the sole purposes
of a particular national court on the basis of the co-operation provided for by Notice 93/C39/05, the
decision as to whether to grant access to them was a matter for the appropriate national court on the
basis of its own national procedural law for as long as the court 289 proceedings giving rise to its
incorporation in a Commission document were pending. It followed that the Commission was required to
refuse access to the documents in the instant case (see p 298 c f g, post).

Cases cited
Delimitis v Henninger Bru AG Case C-234/89 [1991] ECR I-935.
European Commission v EU Council Case C-122/94 [1996] ECR I-881.
Johnston v Chief Constable of the Royal Ulster Constabulary Case C-222/84 [1986] 3 All ER 135, [1987]
QB 129, [1986] 3 WLR 1038, [1986] ECR 1651, ECJ.
Opinion 2/94 [1996] ECR I-1759.
Otto BV v Postbank NV Case C-60/92 [1993] ECR I-5683.
Stichting Certificatie Kraanverhuurbedrijf (SCK) v European Commission Joined cases T-213/95 and T-
18/96 (1997) Transcript, 22 October, CFI.
Socit franaise de Biscuits Delacre v EC Commission Case C-350/88 [1990] ECR I-395.
Weddel & Co AG v EC Commission Case C-54/90 [1992] ECR I-871.
WWF UK v European Commission Case T-105/95 [1997] All ER (EC) 300, [1997] ECR II-313, CFI.

Application
Gerard van der Wal, the applicant, applied to the Court of First Instance of the European Communities for
the annulment of the European Commissions decision of 29 March 1996 refusing him access to letters
sent by the Commissions Directorate General for Competition to national courts in the context of the
notice on cooperation between national courts and the Commission in applying arts 85 and 86 of the EC
Treaty. Mr van der Wal was represented initially by C P Bleeker and L Y J M Parret, of the Hague Bar and
the Brussels Bar respectively, and subsequently by L Y J M Parret, with an address for service in
Luxembourg at the Chambers of Aloyse May, 31 Grand-Rue. The Kingdom of the Netherlands intervened
in support of the applicant and was represented by M Fierstra and J S van den Oosterkamp, Legal
Advisers, acting as agents, with an address for service in Luxembourg at the Embassy of the
Netherlands, 5 Rue C M Spoo. The European Commission, the defendant, was represented by W Wils
and U Wlker, of its Legal Service, acting as agents, with an address for service in Luxembourg at the
office of Carlos Gmez de la Cruz, of its Legal Service, Wagner Centre, Kirchberg. The language of the
case was Dutch. The facts are set out in the judgment of the court.

19 March 1998.

The COURT OF FIRST INSTANCE (Fourth Chamber)

delivered the following judgment.

LEGAL BACKGROUND
1.
In the Final Act of the Treaty on European Union (the TEU) (Maastricht, 7 February 1992; TS12(1994);
Cmnd 2485) the member states incorporated a declaration (No 17) on the right of access to information in
these terms:
The Conference considers that transparency of the decision-making process strengthens the
democratic nature of the institutions and the publics confidence in the administration. The
Conference accordingly recommends that the Commission submit to the Council no later than 1993
a report on measures designed to improve public access to the information available to the
institutions.
290
2.
In response to that declaration, the European Commission published Communication 93/C 156/05
concerning public access to the institutions documents (OJ 1993 C156 p 5), which it sent to the EU
Council, the European Parliament and the European Economic and Social Committee on 5 May 1993. On
2 June 1993 it adopted Communication 93/C166/04 on openness in the Community (OJ 1993 C166 p 4).
3.
In the context of those preliminary steps towards implementation of the principle of transparency, on 6
December 1993 the Council and the Commission approved a code of conduct concerning public access
to Council and Commission documents (OJ 1993 L340 p 41), which sought to establish the principles
governing access to documents held by those institutions.
4.
Accordingly, in implementation of that agreement the Commission adopted, on 8 February 1994, on
the basis of art 162 of the EC Treaty, Commission Decision (ECSC, EC, Euratom) 94/90 on public access
to Commission documents (OJ 1994 L46 p 58), under art 1 of which the code of conduct was formally
adopted. The text of that code is set out in an annex to Decision 94/90.
5.
The Code of Conduct as thus adopted by the Commission sets out a general principle in these terms:
The public will have the widest possible access to documents held by the Commission and the Council.
6.
For those purposes the term document is defined in the code of conduct as meaning any written text,
whatever its medium, which contains existing data and is held by the Commission or the Council.
7.
After briefly setting out the rules governing the lodging and processing of requests for documents, the
code of conduct describes the procedure to be followed, where it is proposed to reject a request, in these
terms:
Where the relevant departments of the institution concerned intend to advise the institution to
reject an application, they will inform the applicant thereof and tell him that he has one month to
make a confirmatory application to the institution for that position to be reconsidered, failing which
he will be deemed to have withdrawn his original application. If a confirmatory application is
submitted, and if the institution concerned decides to refuse to release the document, that decision,
which must be made within a month of submission of the confirmatory application, will be notified in
writing to the applicant as soon as possible. The grounds of the decision must be given, and the
decision must indicate the means of redress that are available, i.e. judicial proceedings and
complaints to the ombudsman under the conditions specified in, respectively, Articles 173 and
138[e] of the Treaty establishing the European Community.
8.
The code of conduct describes the factors which may be invoked by an institution to ground the
rejection of a request for access to documents in these terms:
The institutions will refuse access to any document where disclosure could undermine: the
protection of the public interest (public security, international relations, monetary stability, court
proceedings, inspections and investigations), the protection of the individual and of privacy, the
protection of commercial and industrial secrecy, the protection of the Communitys financial
interests, the protection of confidentiality as requested by the natural or legal persons that
supplied the information or as 291 required by the legislation of the Member State that supplied the
information. They may also refuse access in order to protect the institutions interest in the
confidentiality of its proceedings.
9.
In 1993 the Commission adopted Notice 93/C39/05 on cooperation between national courts and the
Commission in applying arts 85 and 86 of the EC Treaty (OJ 1993 C39). In the notice, the Commission
stated that:
(37) national courts may, within the limits of their national procedural law, ask the
Commission and in particular its Directorate-General for Competition for the following information.
First, they may ask for information of a procedural nature to enable them to discover whether a
certain case is pending before the Commission, whether a case has been the subject of a
notification, whether the Commission has officially initiated a procedure or whether it has already
taken a position through an official decision or through a comfort letter sent by its services. If
necessary, national courts may also ask the Commission to give an opinion as to how much time is
likely to be required for granting or refusing individual exemption for notified agreements or
practices, so as to be able to determine the conditions for any decision to suspend proceedings or
whether interim measures need to be adopted. The Commission, for its part, will endeavour to give
priority to cases which are the subject of national proceedings suspended in this way, in particular
when the outcome of a civil dispute depends on them.
(38) Next, national courts may consult the Commission on points of law. Where the application
of Article 85(1) and Article 86 causes them particular difficulties, national courts may consult the
Commission on its customary practice in relation to the Community law at issue. As far as Articles
85 and 86 are concerned, these difficulties relate in particular to the conditions for applying these
arts as regards the effect on trade between member states and as regards the extent to which the
restriction of competition resulting from the practices specified in these provisions is appreciable. In
its replies, the Commission does not consider the merits of the case. In addition, where they have
doubts as to whether a contested agreement, decision or concerted practice is eligible for an
individual exemption, they may ask the Commission to provide them with an interim opinion. If the
Commission says that the case in question is unlikely to qualify for an exemption, national courts
will be able to waive a stay of proceedings and rule on the validity of the agreement, decision or
concerted practice.
(39) The answers given by the Commission are not binding on the courts which have requested
them. In its replies the Commission makes it clear that its view is not definitive and that the right for
the national court to refer to the Court of Justice, pursuant to Article 177, is not affected.
Nevertheless, the Commission considers that it gives them useful guidance for resolving disputes.
 (40) Lastly, national courts can obtain information from the Commission regarding factual data:
statistics, market studies and economic analyses. The Commission will endeavour to communicate
these data or will indicate the source from which they can be obtained.
292

FACTS
10.
The XXIVth Report on Competition Policy (1994) (the XXIVth Report) stated that the Commission had
received a number of questions from national courts pursuant to the procedure described in para 9,
above.
11.
By letter dated 23 January 1996 the applicant, in his capacity as a lawyer and member of a firm which
deals with cases raising questions of competition at Community level, requested copies of some of the
Commissions replies to those questions, namely: (1) the letter dated 2 August 1993 from the Director-
General of the Directorate-General for Competition (DG IV) to the Oberlandesgericht (the Higher Regional
Court), Dsseldorf, concerning the compatibility of a distribution agreement with Commission Regulation
(EEC) 1983/83 on the application of art 85(3) of the Treaty to categories of exclusive distribution
agreements (OJ 1983 L173 p 1); (2) the letter dated 13 September 1994 from Commissioner van Miert to
the Tribunal dInstance (the District Court), St Brieuc, concerning the interpretation of Council Regulation
(EEC) 26/62 applying certain rules of competition to production of and trade in agricultural products (OJ S
Edn 195962, p 129); and (3) The letter sent by the Commission in early 1995 to the Cour dAppel (the
Court of Appeal), Paris, which had asked it for an opinion on contractual provisions concerning sales
targets for motor vehicle agents in the light of art 85(1) of the Treaty and Commission Regulation (EEC)
123/85 on the application of art 85(3) of the Treaty to certain categories of motor vehicle distribution and
servicing agreements (OJ 1984 L15 p 16).
12.
By letter dated 23 February 1996 the Director-General of DG IV refused the applicants request on the
ground that disclosure of the requested letters would be detrimental to the protection of the public interest
(court proceedings). He explained that:
When the Commission replies to questions submitted to it by national courts before which an
action has been brought for the purposes of resolving a dispute, the Commission intervenes as an
amicus curiae. It is expected to show a certain reserve not only as regards acceptance of the
manner in which the questions are submitted to it but also as regards the use which it makes of the
replies to those questions. I consider that, once the replies have been sent, they form an integral
part of the proceedings and are in the hands of the court which raised the question. The points of
both law and fact contained in the replies must be regarded, in the context of the pending
proceedings, as part of the national courts file. The Commission has sent the replies to that
national court and the decision whether to publish that information and/or make it available to third
parties is a matter primarily for the national court to which the reply is sent.
13.
The Director-General also referred to the need to maintain a relationship of trust between the
Community executive and the national court authorities in the member states. Such considerations, which
are valid in all cases, must apply even more forcibly in cases such as the present, where no final
judgment has yet been given in respect of the matters dealt with in the questions submitted to the
Commission.
14.
By letter dated 29 February 1996 the applicant sent a confirmatory application to the Secretariat-
General of the Commission stating, inter alia, that he did not see how the conduct of the national
proceedings could be undermined if information of a non-confidential nature provided by the Commission
to the 293 national court in the context of application of Community competition law came to the attention
of third parties.
15.
By letter dated 29 March 1996 (the contested decision) the Secretary- General of the Commission
confirmed DG IVs decision on the ground that disclosure of the replies could undermine the protection of
the public interest and, more specifically, the sound administration of justice. He continued as follows:
there is a risk that disclosure of the replies requested, which comprise legal analyses, could
undermine the relationship and the necessary cooperation between the Commission and national
courts. A court which has submitted a question to the Commission would obviously not appreciate
the reply being disclosed, particularly where the question is relevant to a pending case.
16.
The Secretary-General added that the procedure in the present case differed considerably from that
under art 177 of the Treaty to which the applicant had referred in his confirmatory application.

PROCEDURE AND FORMS OF ORDER SOUGHT

17.
In those circumstances, by application lodged at the registry of the Court of First Instance of the
European Communities on 29 May 1996, the applicant brought this action.
18.
 By documents lodged at the registry of the Court of First Instance on 4 and 19 November 1996
respectively, the governments of the Kingdom of the Netherlands and the Kingdom of Sweden sought
leave to intervene in this case in support of the applicant.
19.
By order of the President of the Fourth Chamber of the Court of First Instance of 9 December 1996,
those two governments were granted leave to intervene in support of the applicant. Pursuant to its
request dated 14 March 1997, the Swedish governments application for leave to intervene was removed
from the register by order of the President of the Fourth Chamber of 12 May 1997 and it was ordered to
bear its own costs.
20.
The written procedure ended on 24 April 1997.
21.
The parties presented oral argument and replied to the courts questions at the hearing on 25
September 1997.
22.
The applicant, supported by the Kingdom of the Netherlands, claims that the court should: annul the
Commissions decision of 29 March 1996 and order the Commission to pay the costs.
23.
The Commission contends that the court should: dismiss the application and order the applicant to pay
the costs.

LAW
24.
The applicant raises two pleas in law in support of his action, alleging infringement of Decision 94/90
and infringement of art 190 of the Treaty, respectively.

The first plea, alleging infringement of Decision 94/90

Arguments of the parties
25.
The applicant submits, first, that the exception relating to court proceedings provided for in the code
of conduct applies only to proceedings to which the Commission is a party. The exception cannot
therefore be relied on in this case.
294
26.
If the court were to consider that the exception relating to court proceedings also applies to
proceedings to which the Commission is not a party, the applicant submits, in the alternative, that
disclosure of the documents in issue does not undermine the cooperation between the Commission and
national courts or adversely affect the public interest. There is no basis for the Commissions assertion
that disclosure of such documents would not be appreciated by the national court and, in any event, the
feelings of the national court cannot take precedence over the primary rule of public access.
27.
The Commission is wrong to claim that in certain circumstances it is required to refuse access to
documents. In any event it has a duty to state the extent to which its legitimate interests and the interest
in the proper conduct of legal proceedings require application of the exception to the rule of public access
(see Weddel & Co AG v EC Commission Case C-54/90 [1992] ECR I-871).
28.
The applicant considers that the information which might be provided by the Commission in the
context of cooperation with national courts is not in any way confidential and points out that in this case
the Commission has confirmed that none of the letters requested was of a confidential nature.
29.
It would be contrary to the tradition of public access, of review of acts of the administration and of the
conventional separation of powers if such communications from the administration to the judiciary were
not accessible.
30.
The Netherlands government considers that the Commission has given a broad interpretation to the
exceptions to the fundamental principle that the public will receive the widest possible access to
documents held by the Commission. The effect of its interpretation of Decision 94/90 is to exclude a
category of documents from public access with no consideration of whether the content of the documents
justifies recourse to the exceptions. Letters sent by the Commission to a national court fall within the
scope of the code of conduct and the Commission is incorrect in its view that it is for the national court
alone to decide whether and under what conditions such letters may be disclosed to third parties.
31.
The proper conduct of proceedings before the national court would not be impeded if information
provided by the Commission to the court in the context of an action between individuals came to the
attention of third parties. National courts would not be less disposed to request information from the
Commission if the reply might be disclosed, and even if that were in fact the case it would not constitute
sufficient grounds for holding disclosure to be incompatible with the public interest. The Netherlands
government acknowledges that disclosure of the documents might jeopardise the sound administration of
justice if, by having access to the information in those documents, individuals could avoid judicial
proceedings thus affecting the efficient and uniform application of Community law.
32.
The relationship between the Commission and national courts is governed by the principle of
cooperation in good faith, pursuant to art 5 of the Treaty. The Commission does not act as an expert in the
context of the notice.
33.
Finally, the Netherlands government submits that the Commission has not carried out an individual
assessment of each specific case.
34.
In response to the applicants first argument, the Commission points out that it starts from the
assumption that the contested decision falls within the scope of Decision 94/90. It rejects the
interpretation that the exception relating to court proceedings applies only to proceedings to which the
Commission is a 295 party. The rule set out in the code of conduct is broad enough to include
Commission letters drafted in the context of cooperation with national courts.
35.
As regards the second argument concerning the protection of the public interest, it is not necessary to
assess the interests at issue, since such an assessment is necessary only where the Commission refuses
access to a document in order to protect the confidentiality of its proceedings (see WWF UK v European
Commission Case T-105/95 [1997] All ER (EC) 300, [1997] ECR II-313 (para 59)). It is nevertheless
apparent from the two replies sent to the applicant that his request was examined on its own merits. In
this case it is sufficient that there is a possibility that disclosure might undermine the protection of the
public interest, in particular in the case of court proceedings, for the Commission to be obliged to refuse
access to the document at issue.
36.
Disclosure of the replies given by the Commission in the context of the notice could indeed undermine
the protection of the public interest (court proceedings), and not only in the situation evoked by the
Netherlands government. When a national court applies arts 85(1) and 86 of the Treaty, it does so on the
basis of its autonomous jurisdiction and in a manner determined in principle by national procedural rules
(see the judgments in Otto BV v Postbank NV Case C-60/92 [1993] ECR I-5683 (para 14) and Delimitis v
Henninger Bru AG Case C-234/89 [1991] ECR I-935 (para 53)). It follows from those principles that if a
national court submits a request to the Commission it is for that court alone to determine on the basis of
its national procedural law whether and, if so, when and under what conditions the Commissions reply
may be disclosed to third parties. That remains the case at least for as long as the litigation in question is
pending.
37.
The role of the Commission in the context of its cooperation with national courts differs fundamentally
from the role of the Court of Justice in proceedings under art 177 of the Treaty, to which the applicant
referred in his confirmatory application. When it answers a question referred for a preliminary ruling the
Court of Justice states the law and its judgment is binding on the national court. The Commission, by
contrast, plays a secondary role vis--vis the national court, which is entirely free to decide whether or not
to approach the Commission. The Commissions role can be compared to that of an expert commissioned
by a court to provide information or an opinion. The Commission sends its answer to the national court to
deal with as it sees fit.
38.
The Commission adds that its reason for refusing access to the documents is quite separate from the
question whether those documents contain business secrets or other confidential data which the
Commission is not authorised to disclose in the context of a procedure initiated pursuant to Council
Regulation (EEC) 17/62 implementing arts 85 and 86 of the Treaty (OJ S Edn 195962 p 87). It points
out, however, that it is required to observe the rules of confidentiality as regards competition procedures
pending before it. Within those limits, it seeks to be as open as possible.
39.
The Commission also rejects the Netherlands governments assertion that the principle that the public
should have the widest possible access to documents held by the European institutions is a fundamental
principle of Community law.
40.
As regards the transparency of relations between the executive and the judiciary, it considers that the
relationship between the Commission and national courts cannot simply be equated with the relationship
between the executive and the judiciary in a traditional state.
296

Findings of the court

41.
Decision 94/90 is a measure granting citizens a right of access to documents held by the Commission
(see WWF [1997] All ER (EC) 300, [1997] ECR II-313 (para 55)). It follows from the broad scheme of that
decision that it applies generally to requests for access to documents and that any person may request
access to any Commission document without needing to justify the request (see, in that respect,
Communication 93/C156/05). The exceptions to that right must be construed and applied strictly, in order
not to defeat the application of the general principle laid down in the decision (see [1997] All ER (EC) 300,
[1997] ECR II-313 (para 56)).
42.
Decision 94/90 established two categories of exception. According to the wording of the first category,
drafted in mandatory terms, the institutions will refuse access to any document where disclosure could
undermine [in particular] the protection of the public interest ( court proceedings). It follows that the
Commission is obliged to refuse access to documents falling under that exception once the relevant
circumstances are shown to exist (see [1997] All ER (EC) 300, [1997] ECR II-313 (para 58)).
43.
It follows from the use of the form could, in the present conditional, that in order to show that
disclosure of documents relating to court proceedings could undermine the protection of the public
interest, as required by the case law, before deciding on an application for access to such documents, the
Commission must consider in respect of each document requested whether, in the light of the information
available to it, disclosure is in fact likely to undermine one of the facets of public interest protected by the
first category of exception. Where that is the case, the Commission must refuse access to the documents
in issue (see para 42, above).
44.
It is thus necessary to consider whether, and if so, to what extent the Commission is entitled to rely on
the exception based on the protection of the public interest in order to refuse to grant access to
documents sent by it to a national court in response to a request from that court in the context of the
cooperation based on the notice, even though the Commission is not a party to the proceedings pending
before the national court which gave rise to the request.
45.
In that respect, it should be recalled that art 6 of the Convention for the protection of Human Rights
and Fundamental Freedoms (Rome, 4 November 1950; TS 71 (1953); Cmd 8969) assures the right of
everyone to a fair trial. In order to guarantee that right, the case must be heard, inter alia, by an
independent and impartial tribunal.
46.
It is settled case law that fundamental rights form an integral part of the general principles of law
whose observance the Community judicature ensures (see esp Opinion 2/94 [1996] ECR I-1759 (para 33)
and Stichting Certificatie Kraanverhuurbedrijf (SCK) v European Commission Joined cases T-213/95 and
T-18/96 (1997) CFI Transcript, 22 October (para 53)). For that purpose, the Court of Justice and the Court
of First Instance draw inspiration from the constitutional traditions common to the member states and from
the guidelines supplied by international treaties for the protection of human rights on which the member
states have agreed or to which they have acceded. In that regard the 1950 convention has special
significance (see esp the judgment in Johnston v Chief Constable of the Royal Ulster Constabulary Case
C-222/84 [1986] 3 All ER 135, [1986] ECR 1651 (para 18)). Furthermore, as provided for in art F(2) of the
TEU, which entered into force on 1 November 1993, the Union shall respect fundamental rights as
guaranteed by the [1950 convention], and as they result from the 297 constitutional traditions common to
the Member States, as general principles of Community law.
47.
The right of every person to a fair hearing by an independent tribunal means, inter alia, that both
national and Community courts must be free to apply their own rules of procedure concerning the powers
of the judge, the conduct of the proceedings in general and the confidentiality of the documents on the file
in particular.
48.
The exception to the general principle of access to Commission documents based on the protection of
the public interest when the documents at issue are connected with court proceedings, enshrined in
Decision 94/90, is designed to ensure respect for that fundamental right. The scope of that exception is
therefore not restricted solely to the protection of the interests of the parties in the context of specific court
proceedings, but encompasses the procedural autonomy of national and Community courts (see para 47,
above).
49.
Its scope therefore entitles the Commission to rely on that exception even when it is not itself party to
the court proceedings which, in the particular case, justify the protection of the public interest.
50.
In that respect, a distinction must be drawn between documents drafted by the Commission for the
sole purposes of a particular court case, such as the letters in the present case, and other documents
which exist independently of such proceedings. Application of the exception based on the protection of
the public interest can be justified only in respect of the first category of documents, because the decision
whether or not to grant access to such documents is a matter for the appropriate national court alone, in
accordance with the essential rationale of the exception based on the protection of the public interest in
the context of court proceedings (see para 48, above).
51.
When, in the context of proceedings pending before it, a national court requests certain information
from the Commission on the basis of the cooperation provided for by the notice, the Commissions reply is
expressly provided for the purposes of the court proceedings in question. In such circumstances, the
protection of the public interest must be regarded as requiring the Commission to refuse access to that
information, and therefore to the documents containing it, because the decision concerning access to
such information is a matter to be decided exclusively by the appropriate national court on the basis of its
own national procedural law for as long as the court proceedings giving rise to its incorporation in a
Commission document are pending.
52.
In this case, the applicant requested the production of three letters, all concerning pending court
proceedings. The applicant did not claim that those letters merely reproduced information which was
otherwise accessible on the basis of Decision 94/90. In that respect, furthermore, it should be noted that
the first letter related to the compatibility of a distribution agreement with Regulation 1983/83, the second
concerned the application of Regulation 26/62 and the third concerned the interpretation of Regulation
123/85 (see para 11, above). Those letters thus concerned points of law raised in the context of specific
pending proceedings.
53.
In that respect, as the Commission has already pointed out, it is irrelevant whether the three
documents in issue contained business secrets, since the Commissions refusal to disclose those replies
was justified on the grounds set out above (see paras 4552, above).
298
54.
The role played by the Commission in the context of the cooperation put in place by the notice also
differs from the role of the Court of Justice in proceedings under art 177 of the Treaty. The latter
establishes a special procedure between two court systems. The role of the Court of Justice in that
procedure is to rule on the questions raised by national courts. The national court formulates the
questions to be referred to the Court of Justice according to its own rules of procedure which will, where
necessary, ensure the confidentiality of sensitive information. Similarly, the instructions to the registrar of
the Court of Justice provide that, where appropriate, names or confidential data may be omitted in
publications concerning the case. In contrast, the cooperation referred to in the notice is not governed by
any such rules of procedure. There is therefore no reason to apply the rules concerning the publication of
judgments delivered in the context of proceedings under art 177 to replies given by the Commission in the
context of the notice.
55.
Finally, the applicant has not shown how the principle of the separation of powers and review of acts
of the administration would not be respected if the replies provided by the Commission to national courts
in the context of the notice were not made accessible to the public merely on request to the Commission.
That argument must therefore be dismissed as unfounded.
56.
For the reasons set out above, the first plea cannot be upheld.

The second plea, alleging infringement of art 190 of the EC Treaty

Arguments of the parties
57.
The applicant submits that the statement of reasons given by the Commission is insufficient.
58.
The Netherlands government submits that a statement of reasons must be commensurate to the
nature of the measure involved. It considers that the statement of reasons is incomprehensible because
different reasons were put forward by the Commission in the two letters. In the first the Commission
referred to court proceedings, whilst in the second it mentioned the proper administration of justice. The
recipient does not therefore have a clear view of the reasons which led the institution to decide in the way
it finally did.
59.
In the pleadings, the Commission again gives an essentially different justification for the contested
decision, by referring to the nature of the cooperation between the Commission and national courts, in the
context of which the Commission must be compared to an expert commissioned by the court to provide
information. Quite apart from the fact that the comparison is erroneous, that argument completely
obscures the grounds on which the Commission actually based its decision to refuse access to the
requested documents.
60.
The Netherlands government also considers that the two letters fail to show why or in what way the
supposed relationship of trust between the Commission and national courts could be jeopardised if the
applicant were to be granted access to the documents. The Commission provided no justification for its
view that the national court would not appreciate disclosure of the documents in issue. Furthermore, it is
not at all clear from the statement of reasons how the possible need to protect that relationship of trust
could have other consequences if the case at issue were no longer pending.
61.
The Commission considers that the contested decision is based on sufficient grounds, which are set
out not only in the contested decision itself but 299 also in the letter dated 23 February 1996 from the
Director-General of DG IV. Those two letters clearly indicate the reasons for which the application for
access was refused. Furthermore, the Secretary-General of the Commission also replied to some of the
arguments put forward by the applicant in his confirmatory application dated 29 February 1996.
62.
 In its observations on the Netherlands governments statement in intervention, the Commission
stresses that the grounds for the contested decision are to be found not only in the letter dated 29 March
1996 but also in the letter dated 23 February 1996. There is no contradiction or material difference
between the use of the expression court proceedings in one letter and proper administration of justice in
the other. The relationship of trust referred to by the Commission is clearly the relationship resulting from
the obligation of cooperation in good faith laid down in art 5 of the Treaty.
Findings of the court
63.
It should first be noted that the duty to give reasons for a decision has a twofold purpose, namely, on
the one hand, to permit interested parties to know the justification for the measure so as to enable them to
protect their rights and, on the other, to enable the Community judicature to exercise its power to review
the legality of the decision (see esp the judgments in Socit franaise de Biscuits Delacre v EC
Commission Case C-350/88 [1990] ECR I-395 (para 15) and WWF [1997] All ER (EC) 300, [1997] ECR II-
313 (para 66)). The question as to whether a statement of reasons satisfies those requirements must be
assessed with reference not only to its wording but also to its context and the whole body of legal rules
governing the matter in question (see the judgment in European Commission v EU Council Case C-
122/94 [1996] ECR I-881 (para 29)).
64.
The code of conduct provides that, where the initial application for access is rejected, the applicant is
entitled to ask the institution to reconsider that rejection without being obliged to put forward arguments
challenging the validity of the first decision. That procedure does not constitute an appeal against the
refusal but an opportunity to obtain a second assessment by the institution of the application for access.
65.
It follows that when a reply confirms the rejection of an application on the same grounds, it is
appropriate to consider the sufficiency of the reasons given in the light of all the exchanges between the
institution and the applicant, taking into account also the information available to the applicant concerning
the nature and content of the requested documents.
66.
In this case, it is clear from the applicants letter dated 23 January 1996 and the paras of the XXIVth
Report referred to therein that the applicant knew from the outset that the Commissions letters
constituted replies sent in the context of the notice to three national courts and that each concerned a
case pending before those courts. The subject matter of those letters was also described in general
terms.
67.
In his response dated 23 February 1996 the Director-General of DG IV relied on the exception based
on the protection of the public interest (court proceedings) and explained that the requested letters
contained points of both law and fact which should be regarded as forming part of the national courts
files, particularly since the cases at issue were still pending.
68.
The contested decision constitutes an express confirmation of that first refusal. Even though it refers to
the protection of the public interest and, specifically, the proper administration of justice, the applicant
could have been 300 in no doubt that the Secretary-Generals intention was to reject the application on
the basis of the same exception to the code. There is no conflict between the use of the expression court
proceedings in the first letter and proper administration of justice in the second, since the exception at
issue is intended to ensure respect for the proper administration of justice. It follows that the Commission
gave essentially the same explanation in both letters.
69.
Nor, contrary to what is claimed by the Netherlands government, does the fact that the Commission
referred to cooperation with national courts during the hearing constitute a new ground, as that
cooperation was already mentioned in the first letter, which speaks of a relationship of trust between the
Commission and the national court authorities in the member states, and was recalled in the second
letter, which refers to the necessary cooperation between the Commission and national courts and to the
risk that disclosure of the replies requested could undermine that cooperation.
70.
Nor can the Commission be criticised for having referred to proceedings under art 177 only in the
second letter, as its comments constitute a response to the comparison which the applicant sought to
draw in the confirmatory application between those proceedings and the procedure referred to in the
notice.
71.
It follows from the foregoing that the Commission clearly indicated the grounds on which it had applied
the exception based on the need to protect the public interest (court proceedings) in respect of the three
replies requested, whilst taking account of the nature of the information contained therein. The applicant
was thus in a position to know the reasons on which the contested decision was based and the court was
able to exercise its power to review the legality of that decision.
72.
It follows that the second plea cannot be accepted and the action must therefore be dismissed in its
entirety.

COSTS
73.
Under art 87(2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs, if
they have been asked for in the successful partys pleadings. Since the applicant has been unsuccessful
and the European Commission has asked for costs, he must be ordered to pay the costs. However, under
art 87(4) of the Rules of Procedure, the intervener must bear its own costs.
On those grounds, the Court of First Instance (Fourth Chamber) hereby: (1) dismisses the application;
(2) orders the applicant to pay the costs incurred by the defendant; (3) orders the Kingdom of the
Netherlands to bear its own costs.

301

[1998] All ER (EC) 302

R v Minister of Agriculture, Fisheries and Food, ex parte Compassion

in World Farming Ltd (CWIF)
(Case C-1/96)
EUROPEAN COMMUNITY; Free movement of goods
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, RAGNEMALM, WATHELET
(PRESIDENTS OF CHAMBERS), MANCINI (RAPPORTEUR), MOITINHO DE ALMEIDA, KAPTEYN,
MURRAY, EDWARD, PUISSOCHET, HIRSCH, JANN AND SEVN, ADVOCATE GENERAL LGER
27 MAY, 15 JULY 1997, 19 MARCH 1998
European Community Freedom of movement Goods Exports Quantitative restrictions Calves
exported from member state maintaining level of protection against intensive farming systems laid down
by international convention Calves exported to member states complying with Community directive but
not convention Validity of directive European Convention on the Protection of Animals Kept for
Farming Purposes Council Directive (EEC) 91/629 EC Treaty, art 36.

The applicant, an animal welfare body, requested the Minister for Agriculture, Fisheries and Food to
prohibit or restrict the export of calves intended for rearing in veal crates, a practice which was prohibited
in the United Kingdom. It contended that the UK government had power under art 36 1 of the EC Treaty to
restrict the export of veal calves to other member states, which had implemented Council Directive (EEC)
91/629 laying down minimum standards for the protection of calves, where the veal crate system was
likely to be used during the transitional periods under art 3 2. It relied on the stricter international standards
laid down by the European Convention on the Protection of Animals Kept for Farming Purposes 1976 and
the 1988 recommendation implementing the convention with regard to cattle, to which all the member
states and the Community had agreed to adhere and with which, they contended, the veal crate system
was inconsistent. The minister decided that the UK government had no such power and the applicants
applied for judicial review of that decision. The national court stayed the proceedings and referred to the
Court of Justice of the European Communities for a preliminary ruling the questions (a) whether Directive
91/629 was invalid in so far as it was inconsistent with the convention and the recommendation and (b)
whether a member state which had implemented the recommendation could rely on the derogations from
the principle of free movement of goods provided in art 36 of the Treaty on grounds of public morality,
policy or the protection of the health or life of animals, to restrict the export of live calves in order to
prevent those calves from being reared in veal crate systems in other member states which had
implemented the directive but not the recommendation.
1
Article 36, so far as material, provides: The provisions [on free movement of goods] shall not preclude prohibitions or
restrictions on exports justified on grounds of public morality, public policy the protection of health and life of
humans, animals or plants
2
Article 3, so far as material, is set out at p 325 d e, post

302

Held (1) Although the 1976 convention became an integral part of the Community order on its entry into
force, it was clear from its wording that its provisions were indicative only and limited to setting out
general principles to be elaborated in recommendations. Moreover, the 1988 recommendation expressly
provided that it was not directly applicable in the national law of the contracting parties and was to be
implemented according to the method which each party considered adequate; and it did not contain
legally binding obligations for the contracting parties or the Community. It followed that no factors had
been disclosed which affected the validity of the directive (see p 327 f to p 328 a and p 331 f, post).
(2) Recourse to art 36 of the Treaty was not possible where Community directives provided for the
harmonisation of measures necessary to achieve the specific objective which would be furthered by
reliance on that provision. It was also clear from the wording, context and objectives of Directive 91/629
that it laid down minimum common standards for the protection of calves that were confined for the
purposes of rearing and fattening: those minimum standards and any derogations therefrom were set
down exhaustively, and had to be complied with in accordance with a precise timetable. Since a ban on
exports would strike at the harmonisation achieved by the directive, it followed that a member state could
not rely on art 36 to restrict the export of calves to other member states for reasons relating to the
protection of the health of animals, which was the specific objective of that harmonisation. Nor could a
member state do so for reasons relating to the protection of public policy or public morality, which were
not the subject of the directive since it could not rely on the views or behaviour of a section of national
public opinion in order unilaterally to challenge a harmonising measure adopted by the Community
institutions (see p 329 b and p 330 a to c g to p 331 c g, post); R v Ministry of Agriculture, Fisheries and
Food, ex p Hedley Lomas (Ireland) Ltd Case C-5/94 [1996] All ER (EC) 493, [1996] ECR I-2553 applied.

Notes
For justifications under art 36 for restrictions on trade between member states, see 52 Halsburys Laws
(4th edn) paras 12981204.
For the EC Treaty, art 36, see 50 Halsburys Statutes (4th edn) 278.

Cases cited
Association comit conomique agricole rgional fruits et lgumes de Bretagne v Le Campion Case
218/85 [1986] ECR 3513.
Campus Oil Ltd v Minister for Industry and Energy Case 72/83 [1984] ECR 2727.
Conegate Ltd v Customs and Excise Comrs Case 121/85 [1986] 2 All ER 688, [1987] QB 254, [1987] 2
WLR 39, [1986] ECR 1007, ECJ.
Cullet v Centre Leclerc Toulouse Case 231/83 [1985] ECR 305.
Customs and Excise Comrs v Schindler Case C-275/92 [1994] 2 All ER 193, [1994] QB 610, [1994] 3
WLR 103, [1994] ECR I-1039, ECJ.
Danisco Sugar AB v Allmnna Ombudet Case C-27/96 (1997) Transcript, 27 November, ECJ.
Debus (Criminal proceedings against) Joined cases C-13/91 and C-113/91 [1992] ECR I-3617.
Denkavit (Firma) Futtermittel GmbH v Minister fr Ernhrung, Landwirtschaft und Forsten des Landes
Nordrhein-Westafalen Case 251/78 [1979] ECR 3369.
Deutsche Shell AG v Hauptzollamt Case C-188/91 [1993] ECR I-363.
Direction gnrale des impts v Forest Case 148/85 [1986] ECR 3449.
303
EC Commission v EC Council Case 22/70 [1971] ECR 263.
EC Commission v Germany Case 153/78 [1979] ECR 2555.
Germany v EU Council Case C-280/93 [1994] ECR I-4973.
Gourmetterie Van den Burg (Criminal proceedings against) Case C-169/89 [1990] ECR I-2143.
Greece v EC Commission Case 30/88 [1989] ECR 3711.
Holdijk Joined cases 141143/81 [1982] ECR 1299.
Hnig v Stadt Stockach Case C-128/94 [1995] ECR I-3389.
Kontogeorgas v Kartonpak AE Case C-104/95 [1996] ECR I-6643.
Kramer Joined cases 3/76, 4/76 and 6/76 [1976] ECR 1279.
Opinion 1/76 [1977] ECR 741.
Pigs Marketing Board v Redmond Case 83/78 [1978] ECR 2347.
Procureur de la Republique v Leclerc Case 34/84 [1985] ECR 2915.
Procureur du Roi v Dassonville Case 8/74 [1974] ECR 837.
R v Bouchereau Case 30/77 [1981] 2 All ER 924, [1978] QB 732, [1977] ECR 1999, ECJ.
R v Henn Case 34/79 [1980] 2 All ER 166, [1981] AC 850, [1980] 2 WLR 597, [1979] ECR 3795, ECJ.
R v Ministry of Agriculture, Fisheries and Food, ex p Hedley Lomas (Ireland) Ltd Case C-5/94 [1996] All
ER (EC) 493, [1996] ECR I-2553, ECJ.
Richardt (Criminal proceedings against) Case C-307/89 [1991] ECR I-4621.
Sevince v Staatssecretaris van Justitie Case C-192/89 [1990] ECR I-3461.
Simmenthal SpA v Italian Minister for Finance Case 35/76 [1976] ECR I-1871.
Socit Civile Agricole du Centre dInsmination de la Crespelle v Cooprative dlevage et
dInsemination Artificielle du Dpartment de la Mayenne Case C-323/93 [1994] ECR I-5077.
Spain v EU Council Case C-350/92 [1995] ECR I-1985.
Tedeschi v Denkavit Commerciale srl Case 5/77 [1977] ECR 1555.
Torfaen BC v B & Q plc Case C-145/88 [1990] 1 All ER 129, [1990] 2 QB 19, [1990] 2 WLR 1330, [1989]
ECR 3851, ECJ.
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER
(EC) 97, [1995] ECR I-4921, ECJ.
Unit Socio-Sanitaria Locale No 47 di Biella (USSL) v Instituto Nazionale per lAssicurazione contro gli
Infortum sul Lavoro (INAIL) Case C-134/95 [1997] ECR I-195.

Reference
By order of 12 December 1995, the Queens Bench Division of the High Court, referred to the Court of
Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty two
questions (set out at p 326 j to p 327 d, post) on the interpretation of arts 34 and 36 of the Treaty and on
the validity of Council Directive (EEC) 91/629 laying down minimum standards for the protection of calves.
Those questions arose in proceedings brought by the Royal Society for the Prevention of Cruelty to
Animals (the RSPCA) and Compassion in World Farming Ltd (CIWF) against the Minister of Agriculture,
Fisheries and Food challenging the ministers refusal to restrict, on the basis of art 36 of the Treaty, the
export of veal calves to other member states. On 8 May 1997, pursuant to an order of the High Court, the
RSPCA ceased to be a party. Written observations were submitted on behalf of: CIWF, by G Barling QC
and P Duffy, instructed by M Rose, Solicitor; the UK government, by J E Collins, Assistant Treasury
Solicitor, acting as agent, R Plender QC and S Masters, Barrister; the 304 French government, by C de
Salins, Assistant Director in the Legal Affairs Directorate of the Ministry of Foreign Affairs, and F Pascal,
Central Administrative Attach in the same directorate, acting as agents; the EU Council, by M Sims-
Robertson, Legal Adviser, acting as agent; and the European Commission, by R Wainwright, Principal
Legal Adviser, and H Stvlbk, of its Legal Service, acting as agents. Oral observations were made by:
CIWF, the UK government, the French government, the Council and the Commission. The language of
the case was English. The facts are set out in the opinion of the Advocate General.

15 July 1997.

The Advocate General (P Lger)

delivered the following opinion (translated from the French).
1.
The veal crate system, which is criticised for affecting the health and life of veal calves and for being
contrary to public morality and public policy, within the meaning of art 36 of the EC Treaty, has brought
about this reference by the Queens Bench Division of the English High Court.
2.
This court is asked to determine the extent of the right thus afforded to member states to oppose, on
those grounds, the export of animals to other member states and thereby derogate from the principle of
the free movement of goods.
3.
The High Court also refers to the existence of international .agreements and a Community directive
laying down different standards for the protection of veal calves. It therefore also asks this court to
determine the validity of Council Directive (EEC) 91/629 laying down minimum standards for the
protection of calves (OJ 1991 L340 p 28), with reference to the European Convention on the Protection of
Animals kept for Farming Purposes (17 March 1976; OJ 1978 L323 p 14) and the 1988 Recommendation
concerning cattle.

I LEGAL BACKGROUND TO THE DISPUTE

A. The provisions of international law

The European convention on the protection of animals kept for farming purposes
4.
The European Convention for the Protection of Animals kept for Farming Purposes was drawn up
within the Council of Europe with the aim of protecting animals kept for farming purposes, particularly in
modern intensive production systems3. Adopted in that context, the convention was approved in the
name of the EEC by Council Decision (EEC) 78/923 (OJ 1978 L323 p 12).
3
See the first recital in the preamble to Council Decision (EEC) 78/923 (OJ 1978 L323 p 12).

5.
According to art 1 of the convention, it applies to the keeping, care and housing of animals and in
particular to animals in modern intensive stock-farming systems.
6.
Chapter I lays down the general principles by which the convention is to ensure protection of animals
in stock-farming systems. Chapter II sets up a standing committee and lays down its rules of organisation
and functioning and defines its powers, whilst Ch III lays down how the convention is to enter into force.
7.
The provisions of the convention more particularly devoted to stock-farming systems, to housing space
for animals and to provision of water, food and care for them are set out in arts 3, 3a, 4 and 6.
305
8.
Under art 9(1), recommendations to the contracting parties are to be elaborated and adopted by the
standing committee for the implementation of the principles set out in the convention.
9.
A protocol of amendment to the convention was approved by art 1 of Council Decision (EEC) 92/583
on the conclusion of the Protocol of Amendment to the European Convention for the Protection of Animals
kept for Farming Purposes (OJ 1992 L395 p 21).
The recommendation concerning cattle
10.
The recommendation concerning cattle, dated 21 October 1988 and adopted by the standing
committee of the Council of Europe at its 17th meeting, in accordance with the rules of the convention,
sets out the general requirements relating to the rearing of cattle.
11.
Article 20 of the recommendation provides that it is not to be directly applicable within the national law
of the contracting parties and is to be implemented according to the method that each such party
considers appropriate, that is to say through legislation or through administrative practice.
Appendix C: special provisions for veal calves
12.
Appendix C to the recommendation, more particularly devoted to veal calves, was adopted on 8 June
1993.

B. Provisions of Community law

Regulation 805/68
13.
Council Regulation (EEC) 805/68 on the common organisation of the market in beef and veal (OJ S
Edn 1968 (II) p 187), provides for a pricing and trading system and, according to art 1 thereof, covers,
inter alia, live animals of the domestic bovine species.
14.
The second indent of art 22(1) prohibits any quantitative restriction or measure having equivalent
effect in the internal trade of the Community.
Directive 91/629
15.
Directive 91/629 lays down minimum standards for the protection of calves both in order to eliminate
differences which may distort conditions of competition within the organisation of the common market in
calves and in order to ensure the well-being of calves (see the fifth and sixth and seventh recitals).
16.
The first recital states that all the member states have ratified the 1976 convention and this has been
approved by Decision 78/923.
17.
The seventh recital in the preamble to the directive states that the Commission should, on the basis of
a report from the Scientific Veterinary Committee, pursue actively scientific research into the most efficient
stock-farming system or systems from the point of view of the well-being of calves and that provision
should accordingly be made for an interim period to enable the Commission to complete this task
successfully.
18.
Article 3(1) lays down a transitional period of four years during which all new holdings are to comply
with minimum requirements regarding the housing of calves in individual boxes or by tethering in stalls.
19.
Article 4(1) requires the member states to ensure that the conditions for rearing calves comply with the
general provisions laid down in the annex to the 306 directive. That annex lays down rules concerning,
inter alia, accommodation, feeding and watering of calves.
20.
Article 11(2) allows the member states, in compliance with the general rules of the Treaty, to maintain
or apply within the territories stricter provisions for the protection of calves than those laid down in the
directive.

II FACTS AND NATIONAL PROCEDURE

21.
The facts, as described in the order for reference, are as follows.
22.
In recent years, between 500,000 and 600,000 veal calves have been exported each year from the
United Kingdom to other member states of the Community, some of which allow a substantial proportion
of these animals to be reared in the veal crate system.
23.
A veal crate is a box-like structure used to house a single veal calf. According to the national court,
the term veal crate system refers to
a veal production system in which calves are reared in conditions which do not fulfil the
requirements as to the minimum width of veal crates and the composition of veal calves diet laid
down in the Convention and the Recommendation
24.
This method of rearing calves has been prohibited in the United Kingdom since 1 January 1990, the
date of entry into force of the Welfare of Calves Regulations 1987, SI 1987/2021.
25.
The Royal Society for the Prevention of Cruelty to Animals (the RSPCA) and Compassion in World
Farming Limited (CIWF), the applicants in the main proceedings, whose object is the protection of
animals, consider [the veal crate system] to be inconsistent with the health and well-being of the calves
and the cause of unnecessary suffering.
26.
Consequently, the RSPCA and CIWF asked the Minister of Agriculture, Fisheries and Food to have
resort to art 36 of the Treaty and adopt measures restricting exportation of veal calves to member states
in which the veal crate system was likely to be used, contrary to the standards enforced in the United
Kingdom and to the international standards of the convention, to which all the Member States and the EC
have agreed to adhere.
27.
In reply, the minister stated that the United Kingdom did not have the power to restrict exports of veal
calves and that, even in the event that he had such a power, he would not be minded, for policy reasons,
to impose a ban.
28.
The RSPCA and CIWF then brought proceedings before the High Court 4, which submitted the
following questions to this court.
4
The RSPCA ceased to be a party to the proceedings before the High Court pursuant to an order made by that court on 8
May 1997, after the reference had been made to this court.

III THE QUESTIONS REFERRED FOR A PRELIMINARY RULING

29.
Where: (a) all of the Member States have become parties to the European Convention for the
Protection of Animals kept for Farming Purposes 1976 (the Convention) and the Convention has
been approved by EC Decision 78/923/EEC of 19 June 1978 (OJ 1978 L323, p 12); (b) the 1988
Recommendation concerning cattle (the Recommendation) has been adopted by the Standing
Committee established pursuant to the Convention and has become effective under the terms of
the Convention; (c) the standards 307 laid down by and pursuant to the Convention contain
stipulations as to the minimum width of veal crates and the composition of veal calves diets; (d)
Council Directive 91/629/EEC lays down obligatory minimum standards for the protection of calves
which are lower than the standards laid down by and pursuant to the Convention in certain
respects, including the width of veal crates and the composition of calves diets; (e) the Directive
permits Member States to maintain or apply within their territories stricter provisions for the
protection of calves than those laid down in this Directive; (f) veal calves are exported from a
Member State (Member State A) to certain other Member States (Member State B) which have
implemented and/or complied with the directive but have not implemented and/or complied with the
standards indicated at paragraph (c) above although Member State A has implemented and
complied with those standards; (g) the export of calves to lace rearing contrary to the Convention is
considered to be cruel and immoral by animal welfare organisations and a considerable body of
public opinion, supported by authoritative scientific veterinary opinion, in the Member State from
which exports occur.
(1) In the circumstances set out above, may Member State A rely on Article 36 of the EC Treaty
and, in particular, the grounds of public morality and/or public policy and/or the protection of the
health or life of animals contained therein, to justify any restriction in relation to the export of live
calves from Member State A with a view to avoiding the rearing of those calves in the veal crate
systems in Member States B?
(2) If the effect of provisions of the Directive, if valid, would be to require the answer no to be
given to Question (1), are those provisions valid?

IV ADMISSIBILITY OF THE QUESTIONS REFERRED

30.
Before I go on to consider those questions, it is necessary to examine first the factors which, as
discussed during the proceedings, could prompt the court to decline jurisdiction.
31.
At the hearing, the French government, without raising any objection as to admissibility, stated
however that the first question was theoretical, even hypothetical in the absence of a measure restricting
intra-Community trade. It considers that the national court cannot judge the necessity or proportionality of
a decision which has not been adopted.
32.
According to the courts settled case law
it is for the national courts alone, before which the proceedings are pending and which must
assume responsibility for the judgment to be given, to determine, having regard to the particular
features of each case, both the need for a preliminary ruling to enable them to give judgment and
the relevance of the questions which they refer to the Court. (See the judgment in Kontogeorgas v
Kartonpak AE Case C-104/95 [1996] ECR I-6643 (para 11).)
The court later drew the conclusion that where the questions submitted by the national court concern the
interpretation of Community law, the Court of Justice is, in principle, bound to give a ruling (see the
judgment in Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93
[1996] All ER (EC) 97, [1995] ECR I-4921 (para 59)).
33.
It is clear that the questions submitted directly concern Community law since they relate to the
interpretation of art 36 of the Treaty and the assessment of the validity of Directive 91/629.
308
34.
However, the court has held that, in order to determine its jurisdiction, it has to examine the
circumstances in which the court was called upon by the national court to give a ruling. In this regard, it
considers that it cannot give a ruling on a preliminary question submitted by a national court
where it is quite obvious that the interpretation or assessment of validity of a Community rule
sought by the national court bears no relation to the actual facts of the main action or its purpose or
where the problem is hypothetical and the court does not have before it the factual or legal material
necessary to give a useful answer to the questions submitted to it (See [1996] All ER (EC) 97,
[1995] ECR I-4921 (paras 6061).)5

5
See, more recently, the judgment in Unit Socio-Sanitaria Locale No 47 di Biella (USSL) v Instituto Nazionale per
lAssicurazione contro gli Infortum sul Lavoro (INAIL) Case C-134/95 [1997] ECR I-195 (para 12)).

35.
My first observation is that the link between the questions submitted and the proceedings cannot be
seriously contested. The issue in the main proceedings, centred on the adoption of measures restricting
exports based on requirements of animal protection, is indissociable from the plea based on art 36 of the
Treaty, which is the subject matter of the first question. Similarly, the right of a member state to use art 36
depends, at least in part, as we will see, on the existence of a directive harmonising the field concerned,
so that the question of the validity of Directive 91/629, which lays down minimum standards for the
protection of veal calves, is determinative (see paras 51ff, below).
36.
Nor is the problem hypothetical. From the explanation of the factual and legal background of the case
given by the national court in its order for reference it is clear that the main action concerns the discretion
of the minister to refuse to adopt measures based on art 36. Examination of the national file reveals that
the action, which has been brought by bodies having the protection of animals as their object, seeks
precisely to have the ministers decision set aside6.
6
The affidavit of the representative of the RSPCA and CIWF, dated 5 July 1995, is entitled: In the matter of an application
for leave to apply for judicial review against the Minister of Agriculture, Fisheries and Food by the Royal Society for the
Protection of Cruelty to Animals and Compassion in World Farming Limited.

37.
It is true that the proceedings before the national court present the particularity of having arisen not
from any positive action taken by the United Kingdom but from a refusal to take any such action. It is
therefore the absence of any ministerial decision prohibiting or restricting exports of veal calves to the
other member states which is challenged before the High Court and which has led to the referral of the
questions.
38.
Consequently, the interpretation which the court will give of art 36 of the Treaty will not necessarily
enable the High Court to give an assessment of the lawfulness of the government decision.
39.
If that interpretation upholds the right to apply quantitative restrictions on exports, or measures having
equivalent effect, on the grounds indicated by the High Court, this courts ruling could not be of any use in
resolving the dispute since art 36 only lays down a mere right to adopt similar measures and in this case
the minister has stated that he does not consider it expedient to have recourse to them.
40.
Those objections do not suffice, however, to call the courts jurisdiction into question. If, on the other
hand, the court interprets art 36 as not allowing 309 restrictive measures to be taken in the circumstances
of the present case, that would support the lawfulness of the contested decision.
41.
Moreover, by the questions which it submits, the High Court does not expect this court to enable it to
determine the proportionality of a specific measure adopted by the UK minister. It is only a question of
ruling to what extent art 36 may apply in the circumstances of this ease.
42.
In those circumstances, and on any view, the questions submitted must be declared admissible.

V APPLICATION OF ARTICLE 36 OF THE TREATY

43.
By its first question the national court asks in effect whether art 36 of the Treaty allows a member state
to restrict or prevent the exportation of live animals to another member state on the ground that public
morality, public policy or the health or life of those animals are threatened by the rearing methods used in
the member state of exportation.
44.
It should be pointed out first of all that, as the applicant in the main proceedings accepts, a ban or
restriction on exports of live calves from the United Kingdom would constitute a quantitative restriction on
exports, or measure of equivalent effect within the meaning of Article 34 of the EC Treaty.
45.
However, in the the judgment in R v Ministry of Agriculture, Fisheries and Food, ex p Hedley Lomas
(Ireland) Ltd Case C-5/94 [1996] All ER (EC) 493, [1996] ECR I-2553 (para 18), the court reiterated that
art 36 of the Treaty allows the maintenance of restrictions on the free movement of goods, justified on
grounds of the protection of the health and life of animals. Nothing prevents that reasoning from being
transposed to the case where the reasons relied on justify, not the maintenance, but the introduction of
restrictions.
46.
Whilst the free movement of goods is indisputedly one of the pillars of the Community edifice, it is
significant that recently the court referred to the protection of life and health of animals as being a
fundamental requirement recognised by Community law (see [1996] All ER (EC) 493, [1996] ECR I-2553
(para 18))7.
7
In its judgment in Holdijk Joined cases 141143/81 [1982] ECR 1299 (para 13), the court had already referred to the
Communitys concern for the health and protection of animals, evinced, inler alia, by Article 36 of the Treaty and by
Council Decision No 78/923/EEC

47.
The grounds of public morality or public policy, also referred to in art 36 and to which the national court
refers, constitute other reasons capable of justifying a departure, allowed by the Treaty, from the principle
of free movement of goods.
48.
However, recourse to art 36 is no longer justified if Community rules provide for the measures
necessary to guarantee protection of the interests listed in that article (see eg the judgment in Campus
Oil Ltd v Minister for Industry and Energy Case 72/83 [1984] ECR 2727 (para 27)). In particular, the
member states are bound, where there is a regulation establishing a common organisation of the markets
in a particular field, to refrain from taking any measures which might undermine or create exceptions to it
(see the judgments in Pigs Marketing Board v Redmond Case 83/78 [1978] ECR 2347 (paras 5658), or,
more recently, Direction gnrale des impts v Forest Case 148/85 [1986] ECR 3449 (para 14)).
49.
In the present case, the beef and veal sector is the subject of a common organisation of the markets
but, as the court has held (Holdijk Joined cases 141143/81 [1982] ECR 1299 (para 12)):
310
the establishment of such an organization pursuant to Article 40 of the Treaty does not have
the effect of exempting agricultural producers from any national provisions intended to attain
objectives other than those covered by the common organization, even though such provisions
may, by affecting the conditions or production, have an impact on the volume or the cost of national
production and therefore on the operation of the common market in the sector concerned
50.
However, Regulation 805/68 does not pursue any aim of protecting animals.
51.
On the other hand, that concern is at the centre of Directive 91/629, adopted, for harmonisation
purposes, within the framework of the common agricultural policy. In this regard, the court has reiterated
that recourse to art 36 is no longer possible where Community directives provide for harmonisation of the
measures necessary to achieve the specific objective which would be furthered by reliance upon this
provision (see the judgment in Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (para
18))8.
8
See also the judgments in Tedeschi v Denkavit Commerciale srl Case 5/77 [1977] ECR 1555 (para 35) and Criminal
proceedings against Gourmetterie Van den Burg Case C-169/89 [1990] ECR I-2143 (para 8).

52.
It is therefore necessary to examine the terms of Directive 91/69 in order to determine the latitude left
to a member state which wishes to ensure observance of the aims referred to in art 36.

A. The degree of harmonisation carried out by Directive 91/629

53.
Harmonisation undertaken by a directive precludes application of art 36 by a member state if the
protective measures which it lays down specifically concern the field which the authorities having
recourse to it wish to protect.
54.
In the present case, as I have said, the directive is entirely devoted to the protection of veal calves, as
its title indicates. Above all, some of the rules which it lays down concern the actual system of veal
crates. Article 3 provides for a minimum standard for the accommodation of veal calves, characterised by
strict space measurements. Similarly, paras 11 to 14 of the annex to the directive lay down minimum rules
applicable to the feeding and watering of veal calves9.
9
Article 4 of the directive refers to the annex, thus rendering its content binding, to the same extent as the directive itself.

55.
However one assesses, with particular reference to the values laid down by the 1976 convention and
the 1988 recommendation, the degree of protection thus laid down, which is the issue raised by the
second question, it appears that the fields in which the applicant parties wish art 36 to be used are in fact
dealt with by the directive.
56.
The fact that the applicants claim that the level of protection is insufficient must not be taken as
evidence of incomplete harmonisation. The criterion based on harmonisation of a particular field,
designed to allow the competence left to the member states to be identified, is assessed with regard to
the scale of the area covered by Community rules, the main purpose of harmonisation being above is to
harmonise national laws in a given sector, even if this results in a level which is considered too weak.
57.
From this point of view, the directive fully harmonised the powers of the member states in the area
under consideration.
58.
The member states are certainly authorised to enact more protective measures for the life and health
of animals. This could point to the existence of 311 some leeway left by the directive to art 36 since, under
the first sentence of art 11(2), from the date set in paragraph 1 10, Member States may, in compliance with
the general rules of the Treaty, maintain or apply within their territories stricter provisions for the protection
of calves than those laid down in this Directive.
10
Article 11(1) provides: Member States shall bring into force the laws, regulations and administrative provisions,
including any sanctions, necessary to comply with this Directive not later than 1 January 1994. They shall forthwith
inform the Commission thereof.

59.
However, whilst the directive allows the member states some latitude in the field with which it deals,
where their competence in the matter of the protection of animals may still be exercised to a stricter
standard, the terms used show that the adoption by a member state of measures comparable to those
provided for by the directive is allowed only within the boundaries of its own territory and only in
accordance with the principles laid down by the Treaty.
60.
The measures authorised under art 11 are therefore confined, by the principle of the free movement of
goods, to strictly territorial boundaries; there can be no effect on intra-Community trade. The UK
government has in fact made use of this power when it prohibited the rearing of veal calves according to
the system in question.
61.
It would therefore be possible to conclude, like the United Kingdom, that owing to the exhaustiveness
of Directive 91/629 recourse to art 36 is impossible.
62.
However, it must be pointed out, as the Commission does, that under art 3(1) of the directive, some of
the requirements laid down by it for the accommodation of veal calves 11 apply from 1 January 1994 and
for a transitional period of four years, [to] all holdings newly built or rebuilt and/or brought into use for the
first time after that date.
11
Those requirements contained in art 3(1), are as follows: where calves are housed in groups, they must have
sufficient unobstructed floor space to be able to turn round and lie down without hindrance of at least 15m for each calf
of 150kg live weight, where calves are housed in individual boxes or by tethering in stalls, the boxes or stalls shall
have perforated walls and their width must he no less than 90cm plus or minus 10%, or 080 times the height at the
withers.

63.
Furthermore, art 3(4) states:
The duration of use of installations built: before 1 January 1994 which do not meet the
requirements of paragraph 1 shall under no circumstances extend beyond 31 December 2003;
during the transitional period, in accordance with paragraph 1, shall under no circumstances extend
beyond 31 December 2007, unless on that date they comply with the requirements of this
Directive.
64.
For reasons related to the progressive adaptation of agricultural holdings to the new standards, the
Community legislature has therefore accepted that the new standards are not to apply immediately to all
installations. Until the end of 2003, holdings built before 1 January 1994 can escape the standards
applicable to those built after that date and for a transitional period of four years. Similarly, holdings built
during the transitional period are to be brought into full conformity with the standards before 31 December
200712.
 12
The amendments made by Council Directive (EC) 97/2, amending Directive 91/629 (OJ 1997 L25 p 24), in particular art
3(3) and (4), do not change the argument nor the conclusions which I think must be drawn. Standards which are stricter
and more precise than those lad down by the original Directive 91/629 as regards accommodation and floor space for
veal calves are now to apply to new or rebuilt holdings and to those brought into operation after the end of the
transitional period. But, although the date for bringing installations built before 1 January 1994 into conformity with the
standards of the transitional period remains 31 December 2003, the date for bringing installations of the transitional
period into conformity with the new standards is put back to 31 December 2006. Only the level of harmonisation chosen
by the legislature and the schedule for achieving it have therefore been changed, but not the principle of achieving it in
stages. As long as the final stage, to be completed at the end of 2006, has still not been reached, harmonisation cannot
be regarded as having been fully and effectively achieved.

65.
Many categories of agricultural holdings, having different levels of protection for veal calves, may
therefore co-exist, so that harmonisation of the conditions for housing animals will not be truly achieved
until that last date. It is only then that all agricultural holdings covered by the directive will have to comply
with the same rules.
312
66.
Consequently, for a number of years yet, the member states may be obliged to allow the rearing,
export or import of veal calves whose method of rearing departs from the protective rules of the directive,
even as amended.
67.
In those circumstances, I think it is logical to apply to this ease the case law of the court which holds
that a directive does not have the effect of removing the competence which art 36 allows member states
when the period it gives them for adopting the provisions necessary to comply with it has not expired (see
the judgments in Simmenthal SpA v Italian Minister for Finance Case 35/76 [1976] ECR I-1871 (para 19)
and Firma Denkavit Futtermittel GmbH v Minister fr Ernhrung, Landwirtschaft und Forsten des Landes
Nordrhein-Westafalen Case 251/78 [1979] ECR 3369 (paras 1820)).
68.
In the present case, art 3 of the directive lays down no period for the transposition of the Community
standard into national law. It determines the immediate or deferred applicability to the operators
concerned of the obligations provided for by the directive according to the date on which the agricultural
holding was built or brought into service.
69.
However, the two cases are comparable in that actual implementation of the protective rules laid down
by the directive may be deferred in time.
70.
Until the periods for the implementation of the animal housing conditions laid down by the directive
have expired, as at the end of a transposition period, it remains within the power of the member states to
apply art 36 and this article may serve as a basis for action justified on grounds of the protection of the life
and health of animals.
71.
I would add that the directive, which does not have this purpose, is not designed to take into account
considerations of public morality and public policy, as referred to in the first preliminary question, as a
possible justification for recourse to art 36.
72.
Moreover, a directive might exhaustively harmonise a field concerning the protection of animals but
still not pacify pressure groups which, judging the level of protection adopted to be insufficient, would still
be likely to disturb public order.
73.
For those different reasons, Directive 91/629 does not suffice to prohibit the use of art 36 in the case
submitted to the court, so that it appears necessary for the court to rule on the scope of that article.

B. Scope of art 36
(1) The judgment in Ex p Hedley Lomas
74.
This case may be compared in many respects to the judgment in Ex p Hedley Lomas [1996] All ER
(EC) 493, [1996] ECR I-2553.
313
75.
The two cases specifically concern the fields covered by art 34 and art 36 in so far as the latter article
is relied on to advocate quantitative restrictions based on grounds of the protection of the life and health
of animals.
76.
Above all, the most characteristic point in common between them is that the restrictive measure
adopted, or sought, has its cause outside the territory of the member state which adopted it, or which is
asked to adopt such a measure, and within the territory of the Community.
77.
 This circumstance places the member state concerned in a situation delicate from the point of view of
Community principles and in particular from the point of view of the free movement of goods in which it is
has to judge the necessity or expediency of restricting exports of its goods to another member state for
reasons based or a practice which is largely unknown to it.
78.
However, in Ex p Hedley Lomas, the question was whether art 36 could be invoked by a member state
to justify a restriction on exports of animals to another member state on the ground that the latter state
was not complying with the requirements of a Community directive providing for harmonisation of the
measures necessary for achieving the aim in view. In the present case, on the other hand, farmers in
some member states rear veal calves using the veal crate method but without contravening the
provisions of the directive harmonising the protection standards.
79.
In spite of those differences, the aspects of the problem posed in each of the two cases could have
prompted the same approach.
80.
The directive in question in Ex p Hedley Lomas brought about full harmonisation of the measures
necessary to achieve the aim covered by recourse to art 36 but the minister contended that it was not
always respected in the member state of destination.
81.
In the case now before the court, the directive is applied but it does not bring about full harmonisation.
82.
In both cases, failure to implement measures enabling the aim stated by art 36 to be achieved can
make recourse to that provision legitimate.
83.
However, the factual context of Ex p Hedley Lomas, characterised by uncertainty as to the breach of
the Community legislation by the member state of importation, did not allow the court to rule on this point
nor on the fact that the acts in question were going on outside the territory of the state of exportation 13.

13
I stated in my opinion in Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (paras 2425) that breach
of the directive was not certain, or at any rate was not demonstrated, which was sufficient for the court to rule out, in
such circumstances, any recourse to art 36 by the member state of exportation.

84.
The court made a point of explaining more than once that the decision was given in that context 14. As a
result, the effect of the judgment is relative15. The answer given by the court is confined to the case where
non-observance of the provisions of a harmonising directive would be hypothetical 16. It is easy to see why,
in such a case, a member state cannot be entitled to take unilateral measures to restrict exports.
14
Ibid, paras 16, 2021.
15
The judgment in Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (para 16) is worded as follows: It
is against that factual background that the first question asked by the national court must be answered.
16
The court does not consider that a member state has a right to have recourse to art 36 in order to obviate any breach
by a member state of rules of Community law judgment in Ex p Hedley Lomas [1996] All ER (EC) 493, [1996]
ECR I-2553 (para 20); my emphasis).

314
85.
In view of those particular aspects, I consider that the present case lends itself better to the
interpretation of art 36 invoked to oppose extra-territorial practices.
(2) Use of art 36 to uphold certain of its aims against extra-territorial practices
86.
The national court inquires whether a member state has a right to justify, by reference to
considerations relating to public morality, public policy or the protection of the health and life of animals,
restrictions on the export of veal calves to member states in which the veal crate system is used.
87.
As I explained in my opinion in Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (para
31) I consider that a member state can rely on art 36 of the Treaty only in order to ensure protection,
within its own national territory, of an interest safeguarded by that article.
88.
In that same opinion, I cite Advocate General Trabucchi who, in his opinion in Procureur du Roi v
Dassonville Case 8/74 [1974] ECR 837 at 860, concluded:
states can derogate in the said manner [under art 36] only for the purpose of the protection
of their own interests and not for the protection of the interests of other states Article 36 allows
every state the right to protect exclusively its own national interests. Consequently, for the purpose
of protecting industrial and commercial property, each state can restrict the freedom of movement
of goods only with reference to the protection of individual rights and economic interests falling
under its own sphere of [responsibility].17

17
Cited in my opinion in Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (para 32).
89.
I simply refer to that part of my opinion explaining the reasons justifying, in my view, a strict limitation
of the scope of art 36 (see Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (paras 33
39)).
90.
It follows that export restrictions can be justified only if the acts or practices in question offend against
public morality or public policy in the state of exportation.
91.
At this stage of development of the European Community, it would not be acceptable for a member
state to be allowed to use that provision to oppose the export of goods on the ground that there might be
an infringement, according to its own criteria, to public policy or public morality of the member state of
importation.
92.
In the present case, it is common ground that the rearing of animals under the system criticised by
CIWF, although practised outside the United Kingdom, produces effects on British territory in which public
opinion, like specialised veterinary circles, opposes its maintenance in certain member states.
93.
The situation is not comparable as far as the aim of protecting animals is concerned. In this case, the
adverse effect on the health and life of animals resulting from the rearing system in question is produced
outside British borders and outside the sphere of responsibility of the United Kingdom.
94.
I will examine in turn each of the aims of art 36 of the Treaty to which the preliminary question makes
reference.
315
Public morality
95.
Whilst the court has held that it has competence to assess concepts such as public policy 18 or public
security19, the strictly national character of which is comparable to that of public morality, it considers
nevertheless that: In principle, it is for each member state to determine in accordance with its own scale
of values and in the form selected by it the requirements of public morality in its territory (see the
judgment in R v Henn Case 34/79 [1980] 2 All ER 166, [1979] ECR 3795 (para 15)) 20.
18
See esp the judgments in Cullet v Centre Leclerc Toulouse Case 231/83 [1985] ECR 305 (paras 3233) and Procureur
de la Republique v Leclerc Case 34/84 [1985] ECR 2915 (para 9)).
19
See esp the judgment in Campus Oil [1984] ECR 2727 (para 33), in which the court held: it must be decided whether
the concept of public security covers reasons such as those referred to in the question raised by the national court.
20
See also the judgment in Conegate [1986] 2 All ER 688, [1986] ECR 1007 (para 14).

96.
However, on the rare occasions when the court has had to rule on the protection of public morality,
within the meaning of art 36 of the Treaty, the cases concerned involved national rules in areas about
which it may be said that, generally, they are by tradition at the heart of the debate held on this subject in
most European societies21.

21
The judgments in Henn [1980] 2 All ER 166, [1979] ECR 3795 and Conegate [1986] 2 All ER 688, [1987] QB
254, [1986] ECR 1007, involved pornography and the judgment in Customs and Excise Comrs v Schindler Case C-
275/92 [1994] 2 All ER 193, [1994] ECR I-1039, relating, however, to legislation governing a services activity,
involved lotteries.

97.
Even in these cases, the court did not, however, fail to point out that the competence of the member
states was recognised in principle, which leaves the court with the power to derogate from the rule, if the
facts of the case justify this.
98.
The formulations used, which contain substantive assessments by the court of the rules in question,
seem to evince the courts intention to ensure that the meaning of the concept does not shift. In Henn, the
court stated that it could not be disputed that the statutory provisions in question came within the powers
reserved to the member states by art 36 (see Henn [1980] 2 All ER 166, [1979] ECR 3795 (para 15)). In
Conegate Ltd v Customs and Excise Comrs Case 121/85 [1986] 2 All ER 688, [1986] ECR 1007, in
holding that domestic rules banning the importation of certain indecent or obscene goods could not be
based on grounds of public morality, the court held that the fact that goods caused offence could not be
regarded as sufficiently serious when the legislation of the member state concerned did not lay down any
prohibition on the manufacture or marketing of those goods on its territory (see Henn [1980] 2 All ER 166,
[1979] ECR 3795 (para 15)).
99.
In the words of the courts settled case law
Article 36 is not designed to reserve certain matters to the exclusive jurisdiction of member
states but permits national laws to derogate from the principle of the freedom of goods to the extent
 to which such derogation is and continues to be justified for the attainment of the objectives
referred to in that article. (See the judgment in Spain v EU Council Case C-350/92 [1995] ECR I-
1985 (para 21).)22

22
See also the judgment in Simmenthal [1976] ECR I-1871 (para 24), Tedeschi [1977] ECR 1555 (para 34), EC
Commission v Germany Case 153/78 [1979] ECR 2555 (para 5), Denkavit [1979] ECR 3369 (para 14) and Campus Oil
[1984] ECR 2727 (para 32).

As a derogation from that principle, art 36 must be interpreted strictly (see the judgment in Criminal
proceedings against Richardt Case C-307/89 [1991] ECR I-4621 (para 20)).
316
100.
Despite its specific character, public morality cannot totally escape the rule. Reasons based on this
concept may not be invoked at every turn, or else there would be a risk of multiplying obstacles to intra-
Community trade.
101.
In this regard, I consider it necessary for the court to assume the right to carry out a minimal review of
the content of the concept of public morality, particularly in areas less generally condemned than those
with which it has had to deal in the past. The advantage of this approach will be to prevent the concept
from being given a broad meaning for the sole purpose of justifying unreasonably restrictive measures 23.

23
See, to this effect, the opinion of Advocate General Van Gerven in Torfaen BC v B & Q plc Case C-145/88 [1990] 1 All
ER 129, [1989] ECR 3851 (para 29). A broad interpretation is, incidentally, facilitated by the actual meaning of the
expression. Since morality is defined as Attitude, moral conduct, principles, reference must he made to the term
morals, which means: All the rules of conduct and values which operate as standards in a society, Le Petit Larousse.

102.
However, the fact remains that public morality varies from one member state to another, in time and
according to the member states socio-cultural particularities. Moreover, it is no part of the courts task to
judge values characterising the public morality of a member state and which are therefore quite specific to
it. Consequently, it would appear to be necessary to allow national authorities sufficient discretion to
determine the requirements which ensue from public morality, within the bounds imposed by the Treaty 24.

24
See the comparable reasoning followed by the court in the services field in Schindler [1994] 2 All ER 193, [1994]
ECR I-1039 (paras 6063).

103.
The complementary task for the court is to give an interpretation of art 36 which removes form its
scope domestic practices or domestic rules pursuing aims which clearly cannot be a matter of public
morality.
104.
As far as the present case is concerned, the fact that a member state should consider that harm
unjustifiably caused to the life or health of domestic animals, even for economic purposes, through the
use of a particular rearing method is a matter of public morality in that state does not appear to be
manifestly contrary to art 36.
105.
According to the national court
the export of calves to face rearing contrary to the Convention is considered to be cruel and
immoral by animal welfare organisations and a considerable body of public opinion, supported by
authoritative scientific veterinary opinion, in the member state from which exports occur.
106.
It must be added, however, that it must be proved by objective scientific evidence that the rearing
system in question causes unreasonable harm to the health or life of animals and that the measures
adopted must be proportionate to the aim in view25, which is for the national court to judge.
25
See eg the judgments in Campus Oil [1984] ECR 2727 (para 37) and Criminal proceedings against Debus Joined cases
C-13/91 and C-113/91 [1992] ECR I-3617 (para 16).

107.
The national court refers, secondly, to public policy.
Public policy
108.
In its judgment in R v Bouchereau Case 30/77 [1981] 2 All ER 924, [1977] ECR 1999 (para 35),
employing reasoning which can be transposed to the field of free movement of goods, the court held that

In so far as it may justify certain restrictions to the free movement of persons subject to
Community law, recourse by a national authority to the 317 concept of public policy presupposes, in
any event, the existence of a genuine and sufficiently serious threat to the requirements of public
policy affecting one of the fundamental interests of society.
109.
In the present case, the applicant in the main proceedings does not claim that public policy has been
contravened. The most that is asserted, in the applicants observations and, to a lesser extent, in the
order for reference, is: The inhumanity of the veal crate system has produced a strong public response in
the United Kingdom.
110.
The description thus given of the protest movement which has arisen in the United Kingdom against
the use of the rearing system in question does not reveal anything to establish a real threat to public
policy. In such a context, this ground provided for by art 36 cannot be relied on in support of an export
restriction.
111.
Moreover, it would seem to me to be dangerous for a Community principlein this case, freedom of
movement for goodsto be called into question on the ground that its application provokes a social
reaction, if there are no other reasons for limiting its application (see, on this point, the opinion of
Advocate General VerLoren van Themaat in Cullet v Centre Leclerc Toulouse Case 231/83 [1985] ECR
305 (para 5.3)).
112.
Consequently, I conclude that measures restricting exports of animals based on the risk that they may
provoke a reaction in national public opinion against the maintenance of a methods of rearing animals
which is considered to be cruel towards them cannot constitute measures for meeting a public policy
objective within the meaning of art 36.
Protection of the health and life of animals
113.
Article 36 does not allow a member state to restrict its exports on account of extra-territorial
circumstances which, even though they produce effects within its population, do not affect on its own
national territory the interest protected by this provision.
114.
The principle must, in my view, apply to the aim of protecting the health and life of animals.
115.
To allow a member state to prohibit or restrict the export of animals located on its territory in order to
protect them against practises affecting their health or life beyond its own borders would in practice mean
giving member states the right to monitor, and even to influence, the practices or rules applied by the
other member states.
116.
The Community legislature certainly did not intend art 36 to be interpreted so broadly. The function
conferred on the provision, which is meant to be an instrument at the service of the member states
against unreasonable effects which the free movement of goods has on their fundamental national
interests, precludes such an interpretation.
117.
In its judgment in Richardt [1991] ECR I-4621 (para 20), the court reaffirmed this principle of strict
interpretation of art 36 thus:
The Court has stated on several occasions (see the judgment in Campus Oil, cited above,
paragraph 37, concerning restriction on imports) that Article 36, as an exception to a fundamental
principle of the Treaty, must be interpreted in such a way that its scope is not extended any further
than is necessary for the protection of the interests which it is intended to secure. Measures
adopted on the basis of Article 36 can therefore be justified only if they are 318 such as to serve the
interest which that article protects and if they do not restrict intra-Community trade more than is
absolutely necessary. (My emphasis.)
118.
I am not arguing that the protection of the life and health of animals must give way to the economic
requirements of the free movement of goods but, more simply, that, given its function in the Treaty, which
largely determines its territorial scope of application, art 36 is not the appropriate instrument for resolving
a case such as that before the national court.
119.
Title XVI of the Treaty, on the environment, sets aims and lays down procedures enabling the
Community to adopt measures exceeding the strict territorial boundaries of the member states.
120.
The territorial limitation of the scope of art 36 does not in any way call in question its usefulness in the
field of the protection of the health and life of animals. It allows a member state to combine provisions
taken to preserve an animal species threatened on its national territory with measures restricting or
prohibiting export by individuals of the species concerned.
121.
If, as I propose, the court holds that recourse cannot be had to art 36 of the Treaty for reasons other
than the incompleteness of the harmonisation carried out by the directive, the second preliminary
question, on the validity of Directive 91/629, will not require any reply.
122.
 If, on the contrary, the court holds that the directive, being exhaustive, precludes application of art 36, I
must, in the alternative, examine its validity.

VI VALIDITY OF THE DIRECTIVE

123.
The applicant in the main proceedings proposes, likewise in the alternative, that the reply to the
second preliminary question should be that the directive is contrary to the 1976 convention and to the
1988 recommendation, the clear and unconditional provisions of which are, in its view, binding on the
Community.
124.
The UK government and the French government, as well as the Council and Commission, on the other
hand, consider that examination of the directive discloses no factor such as to affect its validity.
125.
The assessment of the validity of a rule of secondary Community law with reference to international
instruments concluded by the Community must follow the criteria which the court has laid down in
connection with procedures relating to the application of the General Agreement on Tariffs and Trade
(GATT). In its judgment in Germany v EU Council Case C-280/93 [1994] ECR I-4973 (para 105), the court
held that
In deciding whether the applicant can rely on certain provisions of GATT to challenge the
lawfulness of the Regulation, it should be noted that the court has held that the provisions of GATT
have the effect of binding the Community. However, it has also held that in assessing the scope of
GATT in the Community legal system, the spirit, the general scheme and the terms of GATT must
be considered.

A. Integration of the convention and the recommendation in Community law

126.
No party disputes the Communitys competence to conclude the convention nor the validity of Decision
78/923 which approved its terms26.
26
I would point out, for the sake of precision, that this competence flows from the power which the Community has, on the
internal level, in the matter of common agricultural policy and harmonisation of laws. In its judgment in Kramer Joined
cases 3/76, 4/76 and 6/76 [1976] ECR 1279 (paras 1520), the court held that, in the absence of specific provisions of
the Treaty authorising the Community to enter into international commitments in certain fields, reference had to be made
to the general system of Community law in the sphere of the external relations of the Community. The court stated that
art 210 of the Treaty, under which the Community has legal personality means that in its external relations the
Community enjoys the capacity to enter into international commitments over the whole field of objectives defined in Part
One of the Treaty. In those circumstances, the court concluded that such authority arises not only from an express
comment by the Treaty, but may equally flow implicitly from other provisions of the Treaty. Now, whilst the Treaty does
not contain any specific provisions authorising the Community to enter into international commitments in the field of
protection of animals during rearing, Decision 78/923 is based on arts 43 and 100 of the Treaty, relating to the
aforementioned fields of common agricultural policy and harmonisation of laws. Moreover, the fourth and fifth recitals in
its preamble refer to, respectively, disparities which may give rise to unequal conditions of competition and which
may consequently have an indirect effect on the proper functioning of the common market and to the common
agricultural policy. It would therefore appear that concluded the convention pursuant to powers conferred on it by those
instruments, in the corresponding fields and in order to attain the Community aims referred to in art 3(d), (f) and (h) of
the Treaty, as in force on the dale of adoption of Decision 78/923. See also the judgment of the court in EC Commission
v EC Council Case 22/70 [1971] ECR 263 (paras 1219) and Opinion 1/76 [1977] ECR 741 (paras 34).

127.
The Community is therefore bound by the convention, which since its entry into force has formed an
integral part of the Community legal order.
319
128.
According to the courts established case law, since they are directly linked to the agreement which
they implement, measures emanating from a body established by the agreement and entrusted with
responsibility for its implementation form part of the Community legal order 27. In the present case, the
recommendation and Appendix C thereto were adopted by the standing committee formed under art 8 of
the convention. This body is to be responsible for the elaboration and adoption of recommendations to
the Contracting Parties containing detailed provisions for the implementation of the principles set out in
Chapter I of this Convention (see Ch II, art 9(1) Detailed implementation).
27
See the judgments in Greece v EC Commission Case 30/88 [1989] ECR 3711 (para 13), Sevince v Staatssecretaris van
Justitie Case C-192/89 [1990] ECR I-3461 (para 9) and Deutsche Shell AG v Hauptzollamt Case C-188/91 [1993] ECR
I-363 (para 17).

B. The effect of the standards laid down by the convention and the recommendation
129.
Although the imperative form used in the provisions of the convention, with the word shall being
systematically used to lay down the principles contained in Ch I, those provisions are for the most part
characterised by imprecision (only Ch I lays down substantive rules applicable to the field dealt with by
the convention).
130.
Thus, the housing of animals, their feeding, and the care given to them must, in particular, according to
the convention, be appropriate to their physiological and ethological needs in accordance with
established practice and scientific knowledge (arts 3a, 4(2), 5). Similarly, the appropriateness of the
housing, feeding, watering and care is to be determined having regard to their species and to their
degree of development, adaptation and domestication (arts 3a, 4(1) and 5). The freedom of movement
for the animals and the food and liquid given to them must not cause the animals unnecessary suffering
or injury (arts 4(1) and 6, first paragraph). The same aim is given for the Community obligation to
thoroughly inspect the state of health of animals at intervals which are sufficient for this purpose. When
an animal is killed on the farm, this must be done competently and without causing unnecessary pain
or distress to the 320 animals (new art 7(2)). Finally, natural or artificial breeding or breeding procedures
which cause, or are likely to cause, suffering or injury are prohibited (new art 3 of the convention,
amended by the protocol of amendment).
131.
The only really precise provisions concern the minimum frequency of inspections of the condition and
state of health of animals and of technical equipment which must be inspected at least once a day in the
case of animals kept in intensive stock-farming systems (new art 7(1) and (3)).
132.
No precise rule can be found in the convention which could call in question the veal-crate system as
regards its main characteristics, namely the minimal width of the crates and the composition of the calves
food. Consequently, the concern expressed in the convention to make the contracting parties aware about
maintaining rearing conditions which respect the well-being of animals in vital areas is not reflected in the
definition of standards whose non-observance by the directive could affect its validity.
133.
According to art 20 of the recommendation, this text shall have no direct application within Parties and
shall be implemented according to the method that each party considers adequate, that is through
legislation or through administrative practice.
134.
Appendix C which, according to art 1(3) of the recommendation, constitutes an integral part of the
recommendation, is subject to the same conditions regarding its entry into force.
135.
The provisions of the recommendation cannot therefore enter into force without implementing
measures being taken by the signatories, each one for itself. Any binding force of the recommendation is
suspended upon adoption of such measures, so that the directive is not subordinate to it. The procedure
applicable in the event of failure to implement a recommendation is set out in art 9(3) and (4) of the
convention. Significantly, this provision allows the Community to bring the effects of the recommendation
to an end simply by giving notice.
136.
Irrespective of the conditions for its implementation, the actual content of the recommendation
precludes the calling in question of the validity of the directive, which, here again, was not bound to
observe its provisions. The rules laid down by the recommendation, being more precise than those laid
down by the convention, are, however, characterised by the use of the conditional tense, at least in
relation to the fields concerning the rearing system in question.
137.
For example, as regards the space available for animals and their feed the elements which
characterise the rearing system in question the first and third paragraphs of art 6(3) of the
recommendation provide that
The construction of accommodation for cattle whether tethered or in pens, should at all times
allow them sufficient freedom of movement to be able to groom themselves without difficulty and
sufficient room to lie down, to rest, to adopt sleeping postures and freely to stretch their limbs and
rise. The animals should be able to see and touch other cattle. (My emphasis.)
138.
Article 8 provides:
The space allowance for cattle housed in groups should be calculated in relation to the whole
environment, the age, sex, live weight and behavioural needs of the stock, taking account of the
presence or absence of horns and the size of the group. Lack of space or overstocking leading to
trampling, behavioural or other disorders shall he avoided. (My emphasis.)
321
139.
The point is that the conditional nature of those rules precludes them from being recognised as having
the slightest binding force and that where, on the other hand, a standard is laid down in mandatory terms,
its imprecision renders it unenforceable.
140.
Article 10, dealing with feeding, partly falls into this last category. It provides:
All animals shall have appropriate access to adequate, nutritious, hygienic and balanced feed
or wholesome liquid each day and to adequate supplies of water of suitable quality, so as to
maintain their full health and vigour and to meet their behavioural and physiological needs.
Sufficient roughage should be provided daily in accordance with the age and the physiological
needs of the animal. (My emphasis.)
141.
 Finally, a reading of Appendix C leads to the same conclusion.
142.
As regards the space allowance for calves, point 4, for example, states The dimensions of the
individual pen or stall shall be appropriate to the size of the animal and The width of the pen should be
not less than the height of the calf at the withers (my emphasis).
143.
Similarly, point 5 states: Where possible, the keeping of calves in groups should be advised.
144.
The second paragraph of point 8 states: Calves older than two weeks shall have access to a
palatable, digestible and nutritious diet containing a sufficient quantity of iron and roughage appropriate to
their age, weight and biological needs (my emphasis).
145.
Point 14 states:
Since some systems at present in use are not designed, constructed or operated in such a way
as to fulfil all the biological needs of calves, efforts must be made to develop and apply husbandry
systems which minimise the risk of injuries and disease and allow for all their biological needs to be
met, in particular by providing appropriate feeding regimes and by avoiding barren environments,
too restricted areas, and lack of social contact.
146.
It follows that neither the convention nor the recommendation create obligations which compel
adoption by the directive.
147.
However, the court does not go by the spirit, scheme or terms of the international agreement of which
the Community measure is alleged to be in breach. In its case law on GATT it has held that it must also
review the legality of the measure in question in the event that the Community intended to implement a
particular obligation entered into within the framework of GATT, or if the Community act expressly refers
to specific provisions of GATT.
148.
The first recital in the preamble to the directive explains that all the member states have ratified the
convention and that the Community has also approved it.
149.
Although that reference to the convention marks the Communitys intention to promote improvements
in rearing conditions for veal calves, which is already evinced by the title and content of the directive, the
wording of the recital, which merely mentions the stage which the member states and the Community
have reached in the procedure for the adoption of the convention, and the general nature of the
reference, do not support the conclusion that the Community wished to confer binding force on any
particular provision of the convention or of the 322 recommendation, or that it was its intention to make
the directive serve the purpose of implementing them.
150.
Consequently, I do not consider the validity of the directive to be affected by the provisions of the
convention or of the recommendation.

VII CONCLUSION
151.
I therefore propose that the court should rule as follows in reply to the questions submitted to it.
(1) Article 36 of the EC Treaty must be interpreted as not allowing a member state, even where
no directive provides for full harmonisation of the measures necessary to achieve the specific
objective which recourse to art 36 is meant to protect, to invoke grounds of public policy and/or the
protection of the health and life of animals in order to justify measures restricting the export of live
calves with a view to preventing those calves from being reared in the veal crate system used in
another member state. Article 36 of the Treaty must be interpreted as allowing a member state, in
the same circumstances, to justify such measures on the grounds of public morality where
protection of the health and life of animals is regarded in that member state as falling within that
field, the harm to the health or life of animals resulting from the rearing method in question is
established by objective scientific evidence and the measures adopted are proportionate to the
objective pursued.
(2) Consideration of Council Directive (EEC) 91/629 laying down minimum standards for the
protection of calves has disclosed no factor of such a kind as to affect its validity.

19 March 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 12 December 1995, received at the court on 2 January 1996, the Queens Bench Division
of the High Court, referred to the Court of Justice of the European Communities for a preliminary ruling
under art 177 of the EC Treaty two questions on the interpretation of arts 34 and 36 of the EC Treaty and
on the validity of Council Directive (EEC) 91/629 laying down minimum standards for the protection of
calves (OJ 1991 L340 p 28).
2.
Those questions have been raised in proceedings brought by the Royal Society for the Prevention of
Cruelty to Animals (the RSPCA) and Compassion in World Farming Limited (CIWF) against the Minister
of Agriculture, Fisheries and Food (the minister) challenging the ministers refusal to restrict, on the basis
of art 36 of the Treaty, the export of veal calves to other member states.

INTERNATIONAL LAW

European Convention on the Protection of Animals Kept for Farming Purposes

3.
The European Convention on the Protection of Animals kept for Farming Purposes (17 March 1976;
OJ 1978 L323 p 14) (the convention) was adopted within the framework of the Council of Europe. It was
approved on behalf of the European Economic Community by virtue of art 1 of Council Decision (EEC)
78/923 (OJ 1978 L323 p 12).
4.
Article 3 of the convention provides:
Animals shall be housed and provided with food, water and care in a manner whichhaving
regard to their species and to their degree of development, adaptation and domesticationis
appropriate to their 323 physiological and ethological needs in accordance with established
experience and scientific knowledge.
5.
Article 4(1) provides that the freedom of movement appropriate to an animal, having regard to its
species and in accordance with established experience and scientific knowledge, is not to be restricted in
such a manner as to cause it unnecessary suffering or injury. Under art 4(2), where an animal is
continuously or regularly tethered or confined, it is to be given the space appropriate to its physiological
and ethological needs in accordance with established experience and scientific knowledge.
6.
Pursuant to art 9(1), the standing committee is to be responsible for the elaboration and adoption of
recommendations to the contracting parties for the implementation of the principles set out in the
convention.

Recommendation concerning cattle

7.
The 1988 recommendation concerning cattle (the recommendation) was adopted by the standing
committee on 21 October 1988 and, by virtue of art 1(1) thereof, applies to all cattle kept for farming
purposes.
8.
The first sub-paragraph of art 6(3) of the recommendation states that the construction of
accommodation for tethered cattle and cattle in pens should at all times allow them sufficient freedom of
movement to be able to groom themselves without difficulty and sufficient room to lie down, to rest, to
adopt sleeping postures and freely to stretch their limbs and to rise.
9.
Article 10 provides that all animals are to have appropriate access to adequate, nutritious, hygienic
and balanced feed or wholesome liquid each day, and to adequate supplies of water of suitable quality, so
as to maintain their full health and vigour and to meet their behavioural and physiological needs. Sufficient
roughage should be provided daily.
10.
Article 20 provides that the recommendation is not to be directly applicable within the national law of
the contracting parties and is to be implemented according to the method that each party considers
adequate, that is to say through legislation or through administrative practice.

Appendix C to the recommendation

11.
Appendix C to the recommendation, which lays down special provisions for calves, was adopted by
the standing committee on 8 June 1993. Under para 4 thereof, the dimensions of an individual pen or stall
are to be appropriate to the size of the animal at the end of its stay in that pen or stall.
12.
Paragraph 8 of Appendix C provides that the stockkeeper should ensure that the newborn calf
receives sufficient colostrum from its dam or another suitable source. Calves older than two weeks are to
have access to a palatable, digestible and nutritious diet containing a sufficient quantity of iron and
roughage appropriate to their age, weight and biological needs in order to maintain good health and
vigour and allow for normal behaviour and normal development of the rumen. All calves are to receive
liquid food at least twice daily during the first four weeks and, in any case, until they are eating adequate
quantities of suitable solid food.
324
COMMUNITY LAW

Regulation 805/68
13.
Under art 1 of Council Regulation (EEC) 805/68 on the common organisation of the market in beef and
veal (OJ S Edn 1968 (I) p 187), that common market organisation is to comprise a price and trading
system and cover, inter alia, live animals of the domestic bovine species.
14.
The second indent of art 22(1) of the regulation prohibits any quantitative restriction or measure having
equivalent effect in the internal trade of the Community.

The directive
15.
The second indent of art 3(1) and art 3(4) of the directive provide as follows:
(1) Member States shall ensure that from 1 January 1994 and for a transitional period of four
years, all holdings newly built or rebuilt and/or brought into use for the first time after that date shall
comply with at least the following requirements where calves are housed in individual boxes or
by tethering in stalls, the boxes or stalls shall have perforated walls and their width must be no less
than 90cm plus or minus 10%, or 080 times the height at the withers (4) The duration of use of
installations built:before 1 January 1994 which do not meet the requirements of paragraph 1
shall under no circumstances extend beyond 31 December 2003;during the transitional period, in
accordance with paragraph 1, shall under no circumstances extend beyond 31 December 2007,
unless on that date they comply with the requirements of this Directive.
16.
Article 4(1) of the directive provides that the member states are to ensure that the conditions for
rearing calves comply with the general provisions laid down in the annex to the directive.
17.
Under art 11(2), from the date set in art 11(1), namely 1 January 1994, the member states may, in
compliance with the general rules of the Treaty, maintain or apply within their territories stricter provisions
for the protection of calves than those laid down in the directive.
18.
Paragraph 7 of the annex to the directive provides that accommodation for calves must be constructed
in such a way as to allow each calf to lie down, rest, stand up and groom itself without difficulty and to see
other calves.
19.
Under para 11 of the annex, all calves must be provided with an appropriate diet adapted to their age,
weight and behavioural and physiological needs, to promote a positive state of health and well-being. In
particular, in order to ensure a positive state of health and well-being as well as a healthy growth rate and
to meet their behavioural needs, their food must include sufficient iron and, as a rule, a minimum of dried
feed containing a digestible fibre.

NATIONAL LAW
20.
The veal crate system has been prohibited in the United Kingdom since 1 January 1990 under the
Welfare of Calves Regulations 1987, SI 1987/2021.
21.
The prohibition now in force is laid down in the Welfare of Livestock Regulations 1994, SI 1994/2126
and the Welfare of Livestock Regulations (Northern Ireland) 1995, SR 1995/172.
325

Background to the main proceedings

22.
According to the order for reference, in the years prior to 1995 between 500,000 and 600,000 veal
calves were exported annually from the United Kingdom to other member states. A substantial proportion
of those calves were then reared in a production system called the veal crate system. A veal crate is a
box-like structure used to house a single veal calf.
23.
The national court states that the rearing conditions under that system do not meet the requirements
relating to the minimum width of veal crates and the composition of veal calves diets set out in the
convention and the recommendation. When the calves are one or two weeks old, they are placed in
individual box-like structures, where they remain until they are removed for slaughter approximately five
months later.
24.
However, it is not disputed that the rearing conditions meet the requirements of the directive, having
regard to the temporary derogations authorised by it.
25.
 It is also apparent from the order for reference that the export of live calves to other member states
using the veal crate system is a topic of considerable public concern in the United Kingdom.
26.
The RSPCA and CIWF are animal welfare bodies with a particular interest in the prevention of cruelty
to farm animals. They asked the minister to prohibit or restrict the export of calves for rearing in veal
crates. They contended that the United Kingdom government had power under Community law to restrict
the export of veal calves to other member states where the system described above was likely to be
used, contrary to the standards in force in the United Kingdom and the international standards laid down
by the convention to which all the member states and the Community had agreed to adhere.
27.
On 22 May 1995 the minister replied to the RSPCA and CIWF that the United Kingdom had no power
to restrict the export of veal calves and that in any event, for policy reasons, he was not minded to impose
a ban even if he had the power to do so.
28.
The RSPCA and CIWF therefore applied to the High Court for judicial review. The RSPCA
subsequently ceased to be a party to those proceedings pursuant to an order made by the High Court on
8 May 1997 which was notified to the Court of Justice on 21 May 1997.

Questions referred for a preliminary ruling

29.
In those judicial review proceedings the High Court decided that, in order to resolve the dispute
between the parties, it was necessary to stay the proceedings and to refer the following two questions to
the Court of Justice for a preliminary ruling:
Where:(a) all of the Member States have become parties to the European Convention for the
Protection of Animals kept for Farming Purposes 1976 (the Convention) and the Convention has
been approved by EC Decision 78/923/EEC of 19 June 1978 (OJ 1978 L 323, p 12); (b) the 1988
Recommendation concerning Cattle (the Recommendation) has been adopted by the Standing
Committee established pursuant to the Convention and has become effective under the terms of
the Convention; (c) the standards laid down by and pursuant to the Convention contain stipulations
as to the minimum width of veal crates and the composition of veal calves diets; (d) Council
Directive 91/629/EEC lays down obligatory minimum standards for 326 the protection of calves
which are lower than the standards laid down by and pursuant to the Convention in certain
respects, including the width of veal crates and the composition of calves diets; (e) the directive
permits Member States to maintain or apply within their territories stricter provisions for the
protection of calves than those laid down in this Directive; (f) veal calves are exported from a
Member State (Member State A) to certain other Member States (Member States B) which have
implemented and/or complied with the directive but have not implemented and/or complied with the
standards indicated at paragraph (c) above although Member State A has implemented and
complied with those standards; (g) the export of calves to face rearing contrary to the Convention is
considered to be cruel and immoral by animal welfare organisations and a considerable body of
public opinion, supported by authoritative scientific veterinary opinion, in the Member State from
which exports occur.
(1) In the circumstances set out above, may Member State A rely on Article 36 of the EC Treaty
and, in particular, the grounds of public morality and/or public policy and/or the protection of the
health or life of animals contained therein, to justify any restriction in relation to the export of live
calves from Member State A with a view to avoiding the rearing of those calves in the veal crate
systems in Member States B?
(2) If the effect of provisions of the Directive, if valid, would be to require the answer no to be
given to Question (1), are those provisions valid?

The validity of the directive

30.
By its second question, which it is appropriate to consider first, the national court asks in effect
whether the directive is invalid in so far as it is inconsistent with the convention and the recommendation.
31.
As far as the convention is concerned, it should first be observed that it became an integral part of the
Community legal order upon its entry into force.
32.
However, it is clear from the actual wording of the provisions cited in paras 3 to 6 of this judgment that
the contracting parties have considerable discretion in the choice of the appropriate methods for
implementing the convention.
33.
As Advocate General Lger observes in para 132 of his opinion, above, the concern expressed in the
convention to make the contracting parties aware of the need to maintain rearing conditions which respect
the well-being of animals in vital areas is not followed up by the definition of standards non-observance of
which could affect the validity of the directive.
34.
 It is clear from the very wording of those provisions that they are indicative only and are limited to
providing for the elaboration of recommendations to the contracting parties with a view to application of
the principles which they set out.
35.
As far as the recommendation is concerned, art 20 thereof expressly provides that the
recommendation is not directly applicable in the national law of the contracting parties and that it is to be
implemented according to the method that each party considers adequate, that is to say through
legislation or through administrative practice.
36.
Secondly, even if the provisions of the recommendation and of its appendix relating to housing for
cattle and their diet may be more precise than those of the convention, a document of that kind
nevertheless does not contain legally binding obligations for the contracting parties and therefore for the
Community.
327
37.
The answer to the second question must therefore be that consideration of the directive has disclosed
no factor of such a kind as to affect its validity.

The possibility of relying on art 36 of the Treaty

38.
By its first question, the national court asks whether a member state which has implemented the
recommendation, drawn up to apply the principles of the convention, may rely on art 36 of the Treaty and,
in particular, on the grounds of public morality, public policy or the protection of the health or life of
animals referred to in that provision in order to justify restrictions on the export of live calves with a view to
preventing them from being reared in the veal crate systems used in other member states which have
implemented the directive but which do not apply the recommendation.
39.
First of all, a ban or restriction on the export of live calves from one member state to other member
states constitutes a quantitative restriction on exports contrary to art 34 of the Treaty.
40.
CIWF does not dispute this but maintains that such a restriction would be justified having regard to art
36 of the Treaty and thus compatible with Community law.
41.
It should be noted at the outset that, where there is a regulation on the common organisation of the
market in a given sector, the member states are under an obligation to refrain from taking any measures
which might undermine or create exceptions to it (see esp Direction gnrale des impts v Forest Case
148/85 [1986] ECR 3449 (para 14)). Rules which interfere with the proper functioning of a common
organisation of the market are also incompatible with such common organisation, even if the matter in
question has not been exhaustively regulated by it (see, to that effect, Association comit conomique
agricole rgional fruits et lgumes de Bretagne v Le Campion Case 218/85 [1986] ECR 3513 (para 13)
and Danisco Sugar AB v Allmnna Ombudet Case C-27/96 (1997) Transcript, 27 November (para 24)).
42.
Under art 1 of Regulation 805/68, live animals of the domestic bovine species are covered by a
common organisation of the market and, in accordance with the second indent of art 22(1) thereof, they
must be able to move freely between the member states since quantitative restrictions and measures
having equivalent effect are prohibited in the internal trade of the Community.
43.
Furthermore, the court has held that, in particular, any provisions or national practices which might
alter the pattern of imports or exports by preventing producers from buying and selling freely within the
state in which they are established, or in any other member state, on the conditions laid down by
Community rules are incompatible with the principles of a common organisation of the market (see Pigs
Marketing Board v Redmond Case 83/78 [1978] ECR 2347 (para 58)).
44.
In this case, a ban on the export of calves would, as the UK government has pointed out, affect the
structure of the market and, in particular, would have a considerable impact on the formation of market
prices, which would interfere with the proper functioning of the common organisation of the market.
45.
It is true that the court ruled in Holdijk Joined cases 141143/81 [1982] ECR 1299 that Community
law, as it then stood, did not prevent a member state from maintaining or introducing unilateral rules
concerning the standards which had to be observed in the installation of enclosures for fatting calves with
a view to protecting the animals and which applied without distinction to calves intended for the national
market and to calves intended for export.
328
46.
However, that judgment related to measures which a member state applied only within its own territory.
Furthermore, it was delivered before the Community legislature had adopted the directive and was
expressly founded on the absence, in the provisions governing the common organisation of the market, of
any provision for the protection of animals kept for farming purposes (see Holdijk [1982] ECR 1299 (para
13)).
47.
Next, while art 36 of the Treaty allows the maintenance of restrictions on the free movement of goods,
justified on grounds of public morality, public policy or the protection of the health and life of animals,
which constitute fundamental requirements recognised by Community law, recourse to art 36 is
nevertheless no longer possible where Community directives provide for harmonisation of the measures
necessary to achieve the specific objective which would be furthered by reliance upon this provision (see
esp R v Ministry of Agriculture, Fisheries and Food, ex p Hedley Lomas (Ireland) Ltd Case C-5/94 [1996]
All ER (EC) 493, [1996] ECR I-2553 (para 18)). In such a case, the appropriate checks must be carried
out and protective measures adopted within the framework outlined by the harmonising directive (see
Socit Civile Agricole du Centre dInsmination de la Crespelle v Cooprative dlevage et
dInsemination Artificielle du Dpartment de la Mayenne Case C-323/93 [1994] ECR I-5077 (para 31)). In
that regard, the member states must rely on mutual trust to carry out checks on their respective territories
(see, most recently, Ex p Hedley Lomas [1996] All ER (EC) 493, [1996] ECR I-2553 (para 19)).
48.
It must therefore be established whether the directive provides for the harmonisation of the measures
necessary for the protection of the health of calves, which would be the primary objective of reliance upon
art 36.
49.
As the court has held in previous cases, in interpreting provisions of Community law it is necessary to
consider not only their wording but also the context in which they occur and the objectives of the rules of
which they are part (see esp Hnig v Stadt Stockach Case C-128/94 [1995] ECR I-3389 (para 9)).
50.
As regards, first, the wording of the directive, art 3(1) lays down standards relating to the minimum
housing space for calves. In addition, under art 4, the member states are to ensure that the conditions for
rearing calves comply with the general provisions laid down in the annex to the directive, including the
minimum standards regarding their housing and diet laid down in paras 7 and 11.
51.
Next, as regards the context of the directive, it is apparent from the first two recitals in its preamble
that the provisions containing minimum requirements for the protection of calves were adopted on
account of a resolution of the European Parliament of 20 February 1987 on animal welfare policy (OJ
1987 C 76, p 185) and of Decision 78/923.
52.
Finally, as regards the objective of the directive, it is apparent from the fifth and sixth recitals in its
preamble that it is guided by the need, first, to eliminate the differences which, by distorting conditions of
competition, interfere with the smooth running of the organisation of the common market in calves and
calf products and, second, to establish common minimum standards for the protection of rearing calves
or calves for fattening in order to ensure rational development of production. In addition, the seventh
recital indicates that the purpose of the interim period is simply to enable the Commission actively to
pursue scientific research into the most efficient stock-farming system or systems from the perspective of
the well-being of calves.
53.
The Community legislature thus sought to reconcile the interests of animal protection and of the
smooth functioning of the organisation of the common market in calves and derived products.
329
54.
Thus, it follows from the wording of the directive, its context and the objectives which it pursues that it
lays down minimum common standards for the protection of calves that are confined for the purposes of
rearing and fattening.
55.
CIWF asserts, however, that the broad discretion accorded to the member states to grant derogations
for very long periods, in accordance with art 3(4) of the directive, shows that the directive is not a full
harmonisation measure excluding recourse to art 36.
56.
As to that, the court holds that in adopting the directive the Community legislature laid down
exhaustively common minimum standards as described above.
57.
Furthermore, the member states are required to implement those standards within their territory, in
accordance with a precise timetable, in order to ensure the well-being of veal calves. The temporary
derogations allowed are themselves laid down exhaustively in the directive.
58.
It cannot be argued that, under art 11(2) of the directive, the member states may, in compliance with
the general rules of the Treaty, maintain or apply within their territories stricter provisions for the protection
of calves than those laid down in the directive.
59.
It is indeed clear from the wording of art 11(2) of the directive, first, that the measures permitted on that
basis, which are limited to strictly territorial boundaries, may relate only to cattle-farms falling within the
jurisdiction of the member state in question and, second, that such measures may be adopted only in
compliance with the general rules of the Treaty.
60.
As the United Kingdom government has correctly observed, it follows from the express terms of that
provision that the member states are not entitled to adopt stricter measures for the protection of calves
other than provisions applying within their own territory.
61.
In adopting the Welfare of Livestock Regulations 1994 and the Welfare of Livestock Regulations
(Northern Ireland) 1995, the United Kingdom has, in accordance with art 11(2) of the directive, applied
within its territory stricter provisions than those laid down in the directive.
62.
However, a ban on exports imposed on account of conditions prevailing in other member states which
have in fact implemented the directive would fall outside the derogation allowed by art 11(2). A ban on
exports such as that called for by CIWF would strike at the harmonisation achieved by the directive.
63.
In those circumstances, the fact that the member states are authorised to adopt within their own
territory protective measures stricter than those laid down in a directive does not mean that the directive
has not exhaustively regulated the powers of the member states in the area of the protection of veal
calves (see, to that effect, Criminal proceedings against Gourmetterie Van den Burg Case C-169/89
[1990] ECR I-2143 (paras 9, 12)).
64.
It follows that a member state cannot rely on art 36 of the Treaty in order to restrict the export of calves
to other member states for reasons relating to the protection of the health of animals, which constitutes
the specific objective of the harmonisation undertaken by the directive.
65.
It remains to be examined whether a member state may rely on art 36 in order to restrict the export of
calves to other member states for reasons relating to the protection of public policy or public morality,
which are not the subject of the directive.
66.
CIWF supports recourse to those justifications simply by drawing attention to the views and reactions
of a section of national public opinion which believes 330 that the system put in place by the directive
does not adequately protect animal health. So, in reality, public policy and public morality are not being
invoked as a separate justification but are an aspect of the justification relating to the protection of animal
health, which is the subject of the harmonising directive.
67.
In any event, a member state cannot rely on the views or the behaviour of a section of national public
opinion, as CIWF maintains, in order unilaterally to challenge a harmonising measure adopted by the
Community institutions.
68.
Therefore, reliance on art 36 for the protection of public order or public morality in circumstances such
as those involved in the instant case is also ruled out.
69.
It follows that a member state which has implemented the recommendation drawn up to apply the
principles of the convention cannot rely on art 36 of the Treaty and, in particular, on the grounds of public
morality, public policy or the protection of the health or life of animals laid down in that article, in order to
justify restrictions on the export of live calves with a view to preventing those calves from being reared in
the veal crate systems used in other member states which have implemented the directive but which do
not apply that recommendation.

Costs
70.
The costs incurred by the French and UK governments, by the EU Council and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the action pending before the national
court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Queens Bench
Division of the High Court, by order of 12 December 1995, hereby rules: (1) Consideration of Council
Directive (EEC) 91/629 laying down minimum standards for the protection of calves has disclosed no
factor of such a kind as to affect its validity. (2) A member state which has implemented the 1988
Recommendation concerning cattle, drawn up to apply the principles of the European Convention on the
Protection of Animals kept for Farming Purposes, cannot rely on art 36 of the EC Treaty and, in particular,
on the grounds of public morality, public policy or the protection of the health or life of animals laid down in
that article, in order to justify restrictions on the export of live calves with a view to preventing those calves
from being reared in the veal crate systems used in other member states which have implemented
Directive 91/629 but which do not apply that recommendation.

331
 [1998] All ER (EC) 332

Bayerische Hypotheken- und Wechselbank AG v Dietzinger

(Case C-45/96)

EUROPEAN COMMUNITY; Consumer

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES WATHELET (PRESIDENT OF THE FIRST CHAMBER, ACTING FOR THE PRESIDENT OF
THE FIFTH CHAMBER), MOITINHO DE ALMEIDA, EDWARD, JANN AND SEVN (RAPPORTEUR),
ADVOCATE GENERAL JACOBS
22 JANUARY, 20 MARCH 1997, 17 MARCH 1998
European Community Consumer protection Contracts negotiated away from business premises
Defendant personally guaranteeing parents obligations to bank Agreement concluded at parents home
Defendant not notified of any right to cancel agreement Bank enforcing guarantee Whether
defendant able to rely on relevant Community rules to renounce guarantee Council Directive (EEC)
85/577, art 2.

The plaintiff bank granted Ds building firm an overdraft facility. At a meeting with a bank representative at
Ds home, the defendant, Ds son, gave a written guarantee for up to DM 100,000 of his parents
obligations to the bank; the defendant was not subsequently informed of any right of cancellation. In May
1993 the bank called in all of Ds debts and sued the defendant under the guarantee for DM 50,000. The
defendant sought to renounce the guarantee, on the ground that he had not been informed of his right of
cancellation in accordance with the requirements of the German law implementing Council Directive
(EEC) 85/577 to protect consumers in respect of contracts negotiated away from business premises. The
Oberlandesgericht found in favour of the defendant and the bank appealed to the Bundesgerichtshof,
which stayed the proceedings and referred to the Court of Justice of the European Communities for a
preliminary ruling the question whether a contract of guarantee concluded by a natural person who was
not acting in the course of a trade or profession was covered by the directive.

Held Directive 85/577 was designed to protect consumers by enabling them to withdraw from a contract
concluded on the initiative of the trader rather than the customer, where the customer might have been
unable to see all the implications of his act. Accordingly, a contract which benefited a third party standing
outside the contractual relationship, such as a guarantee securing the performance of a credit agreement,
could not be excluded from the scope of the directive on the sole ground that the goods or services
purchased were intended for the use of that third party. However, since the directive was designed to
protect only consumers, a contract of guarantee only came within the scope of the directive where, in
accordance with art 21, the guarantor had entered into a commitment for a purpose which could be
regarded as outside his trade or profession. It followed that a contract of guarantee concluded by a
natural person who was not acting in the course of his trade or profession did not come under the
protection of the directive where it guaranteed repayment of a debt contracted by another person who, for
his part, was acting within the course of his trade or profession (see p 344 f and p 345 a b e, post).
1
Article 2, so far as material, is set out at p 342 g, post

332

Cases cited
Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325, ECJ.
Von Colson v Land Nordrhein-Westfalen Case 14/83 [1984] ECR 1891.

Reference
By order of 11 January 1996, the Bundesgerichtshof (the Federal Court of Justice) referred to the Court of
Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty a question
(set out at p 433 d, post) concerning the interpretation of Council Directive (EEC) 85/577 to protect the
consumer in respect of contracts negotiated away from business premises. That question arose in
proceedings between Bayerische Hypotheken- und Wechselbank AG and Edgar Dietzinger concerning
the performance of a contract of guarantee concluded by Mr Dietzinger with the bank. Written
observations were submitted on behalf of: Mr Dietzinger, by E Bubb, Rechtsanwalt, Landshut; the
German government, by A Dittrich, Regierungsdirektor in the Federal Ministry of Justice, and B Kloke,
Oberregierungsrat in the Federal Ministry of Economic Affairs, acting as agents; the Belgian government,
by J Devadder, General Adviser in the Ministry of Foreign Affairs, Trade and Cooperation with Developing
Countries, acting as agent; the French government, by C de Salins, Head of Sub-directorate in the Legal
Affairs Directorate of the Ministry of Foreign Affairs, and R Loosli-Surrans, Charge de Mission in the
same directorate, acting as agents; the Finnish government, by T Pynn, Legal Adviser in the Ministry of
Foreign Affairs, acting as agent; and the European Commission, by C OReilly and U Wlker, of its Legal
Service, acting as agents. Oral observations were submitted by: Mr Dietzinger, the German government,
the French government, the Finnish government and the Commission. The language of the case was
German. The facts are set out in the opinion of the Advocate General.

20 March 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
The issue in this case, which comes by way of a reference from the Bundesgenchtshof (the Federal
Court of Justice), is essentially whether a guarantee given to a financial institution by an individual who is
not acting in the course of his trade or profession, in order to secure a loan by that institution to a third
party acting in the course of his trade or profession, falls within the scope of Council Directive (EEC)
85/577 to protect the consumer in respect of contracts negotiated away from business premises (OJ 1985
L372 p 31).

The facts
2.
The defendants father ran a building firm in respect of which the plaintiff bank granted a current
account overdraft facility. The defendants parents were visited by an employee of the bank; in the course
of that visit, the defendant gave a written guarantee of up to DM 100,000 of his parents obligations to the
bank. The defendant was not informed of any right to cancel the guarantee. The bank subsequently called
in loans granted to the defendants parents, totalling more than DM 16m, and claimed DM 50,000 from
the defendant under the guarantee.
3.
The defendant sought to renounce the guarantee in accordance with the German Gesetz ber den
Widerruf von Haustrgeschaften und hnlichen Geschften of 16 January 1986 (the Law on the
cancellation of doorstep transactions and analogous transactions). The issue was litigated and reached
the 333 Bundesgerichtshof, which referred the following question to the Court of Justice for a preliminary
ruling:
Where a contract of suretyship is concluded under German law between a financial institution
and a natural person who is not acting in that connection in the course of his trade or profession, in
order to secure a claim by the financial institution against a third party in respect of a loan, is it
covered by the words contracts under which a trader supplies goods or services to a consumer
(Article 1(1) of Council Directive 85/577/EEC of 20 December 1985 to protect the consumer in
respect of contracts negotiated away from business premises)?
4.
The domestic law issue appears to be whether a guarantee constitutes a contract for the supply of
goods or services concluded for valuable consideration within the meaning of the 1986 Law; the order for
reference sets out arguments either way, which appear to turn on the concept of consideration in national
law. Moreover there appears to be a divergence of view between the Ninth and the Eleventh Chambers of
the Bundesgerichtshof as to whether consideration is in fact necessary in order to bring a contract within
the scope of the 1986 Law. It appears in any event that, if the guarantee falls within the Law, the
defendant is entitled to renounce it.
5.
The 1986 Law appears to have been intended to implement the directive. If guarantees such as that at
issue in the main proceedings fall within the concept contracts under which a trader supplies goods or
services to a consumer within the meaning of art 1(1) of the directive, the 1986 Law should be construed
so that the guarantee falls within it (see the judgment in Von Colson v Land Nordrhein-Westfalen Case
14/83 [1984] ECR 1891).
6.
Written observations were submitted by the defendant, the Belgian, Finnish, French and German
governments and the European Commission, all of whom with the exception of the Belgian government
were also represented at the hearing.

The directive
7.
The thrust of the directive is to ensure that, in respect of transactions to which it applies, the consumer
has, and is notified of, a cooling-off period of at least seven days during which he can withdraw from the
contract (arts 45).
8.
The directive was adopted under art 100 of the EC Treaty. The preamble states as follows (first, third,
fourth and seventh recitals):
Whereas it is a common form of commercial practice in the Member States for the conclusion of
a contract or a unilateral engagement between a trader and consumer to be made away from the
business premises of the trader, and whereas such contracts and engagements are the subject of
legislation which differs from one member state to another Whereas the preliminary programme
of the European Economic Community for a consumer protection and information policy provides
inter alia, under paragraphs 24 and 25, that appropriate measures be taken to protect consumers
against unfair commercial practices in respect of doorstep selling; whereas the second programme
 of the European Economic Community for a consumer protection and information policy confirmed
that the action and priorities defined in the preliminary programme would be pursued; Whereas 334
the special feature of contracts concluded away from the business premises of the trader is that as
a rule it is the trader who initiates the contract negotiations, for which the consumer is unprepared
or which he does not expect; whereas the consumer is often unable to compare the quality and
price of the offer with other offers; whereas this surprise element generally exists not only in
contracts made at the doorstep but also in other forms of contract concluded by the trader away
from his business premises Whereas the freedom of Member States to maintain or introduce a
total or partial prohibition on the conclusion of contracts away from business premises, inasmuch
as they consider this to be in the interest of consumers, must not be affected
9.
Article 1 provides as follows:
(1) This Directive shall apply to contracts under which a trader supplies goods or services to a
consumer and which are concluded: during an excursion organized by the trader away from his
business premises, or during a visit by a trader (i) to the consumers home or to that of another
consumer; (ii) to the consumers place of work; where the visit does not take place at the express
request of the consumer. (2) This Directive shall also apply to contracts for the supply of goods or
services other than those concerning which the consumer requested the visit of the trader, provided
that when he requested the visit the consumer did not know, or could not reasonably have known,
that the supply of those other goods or services formed part of the traders commercial or
professional activities. (3) This Directive shall also apply to contracts in respect of which an offer
was made by the consumer under conditions similar to those described in paragraph 1 or
paragraph 2 although the consumer was not bound by that offer before its acceptance by the
trader. (4) This Directive shall also apply to offers made contractually by the consumer under
conditions similar to those described in paragraph 1 or paragraph 2 where the consumer is hound
by his offer.
10.
Consumer is defined as a natural person who, in transactions covered by this Directive, is acting for
purposes which can be regarded as outside his trade or profession. Trader is defined as a natural or
legal person who, for the transaction in question, acts in his commercial or professional capacity, and
anyone acting in the name or on behalf of a trader (art 2).
11.
Contract is nowhere defined.
12.
It may be noted at the outset that, on the assumption that the guarantee was not concluded at the
home of the defendant and on the basis that the defendants parents were not consumers in the context
of the transaction at issue, the guarantee would in any event not fall within art 1(1) of the directive, since it
was not concluded during a visit by a trader to a consumers home or the consumers place of work or
during an excursion organised by the trader away from his business premises (see the first, fourth and
seventh recitals in the preamble)2. The wording of art 1(1) appears surprisingly restrictive in the light of
the preamble to the directive, which suggests that the essential element 335 determining application of
the directive is that the contract in question was concluded away from the traders business premises. In
its earlier ruling on the directive in Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994]
ECR I-3325 (para 3) the court was not called upon to address that issue, although it was not obvious from
the order for reference in that case that the facts fell within art 1(1) 3; the court restricted itself to pointing
out to the national court the need to verify whether the contract was concluded in the circumstances
described by the directive4. In this case, the national court has phrased its question in such a way that it is
in any event both possible and appropriate for the court to provide an answer.
2
Article 1(1) of the Commissions original proposal for a Council Directive to protect the consumer in respect of contracts
which have been negotiated away from business premises (OJ 1977 C22 p 6) provided that the directive applied to
contracts negotiations for which have been initiated away from business premises.
3
See also Advocate General Lenzs opinion: [1995] All ER (EC) 1, [1994] ECR I-3325 (para 3).
4
See the judgment in Faccini Dori [1995] All ER (EC) 1, [1994] ECR I-3325 (para 14) and Advocate General Lenzs
opinion: [1995] All ER (EC) 1, [1994] ECR I-3325 (para 26).

13.
The transaction at issue in this case is a guarantee (or contract of suretyship; the terms are
interchangeable) under which the defendant, acting for purposes outside his trade or profession,
undertook to the plaintiff bank, acting in its commercial capacity, to stand surety for the loan provided by
that bank to the defendants parents, also acting in their commercial capacity. The referring court has
asked whether a guarantee such as that described above is a contract within the scope of the directive.
14.
The Belgian, Finnish, French and German governments all argue that the guarantee is not within the
directive. There are three distinct, albeit overlapping, lines of argument to that effect. First, it is argued
generally that the guarantee is not a contract for the purposes of the directive because the consumer
receives no consideration, or, in other words to the same effect, because the guarantee is not a
synallagmatic contractnamely a bilateral agreement involving mutual and reciprocal obligations or
dutiesbut a unilateral undertaking from the point of view of the guarantor. Secondly, it is argued in more
specific terms that the guarantee is not a contract as defined by art 1(1) because no goods or services
are supplied by the trader party to the consumer party. Thirdly, it is argued that the guarantor cannot be
correctly described as a consumer.
15.
The defendant and the Commission argue on the contrary that the guarantee is a contract, both on the
general basis that all contracts between a trader and a consumer (including unilateral undertakings given
by a consumer to is a trader) are covered and in more specific terms because contracts made between a
trader who supplies goods or services and a consumer are covered.
16.
In keeping with the courts approach to the interpretation of Community legislation, I will consider in
turn the terms, the scheme and the objectives of the directive with a view to determining whether a
guarantee of the type at issue falls within its scope.

The terms of the directive

17.
The English version of the directive suggests that there is no scope for it to apply to a transaction such
as the guarantee at issue in the main proceedings, since the transaction is not a contract under which a
trader supplies goods or services to a consumer within the meaning of art 1(1). Most of the other
language versions, however, are framed in wider terms, more or less to the same effect as the French:
contrats conclus entre un commerant fournissant des biens ou des services et un consommateur. On a
literal reading of those versions, it is 336 not necessary in order for the directive to apply that the goods or
services in question are supplied under the contract at issue: it suffices that one of the parties is a
supplier of goods or services. While it may be thought that the linguistic distinction is slight and that the
majority version may well have been intended to have the same meaning and effect as the minority, the
distinction may be crucial in this case: the guarantee at issue was concluded between a trader who
supplies goods or services, namely the plaintiff bank, and a consumer, namely the defendant, and on a
literal reading of the majority version it appears to be irrelevant whether the goods or services were
supplied to the other party to the contract at issue under that contract or to another party in the context of
a separate transaction. I will accordingly assume that it is not conclusive that the contract at issue does
not fall within the literal wording of the English version of art 1(1).
18.
In my view, notwithstanding the literal terms of the majority version of art 1(1) the directive applies
solely to contracts under which a trader supplies goods or services to a consumer.
19.
As the Belgian government submits, even if the wording (in the French version) of art 1(1) is broad, the
provision should be read in the light of the remainder of that article. It is apparent from art 1 as a whole in
particular in my view art 1(2)that the contracts covered are those under which a trader supplies goods
or services to a consumer.
20.
The Commission suggests that the reason for the reference to goods or services in art 1(1) is simply
to make clear that the directive is not restricted to traders supplying goods. At the hearing, the
Commission invoked the terms of art 1 of the original proposal, which referred simply to contracts
between a consumer and a trader, and unilateral engagements by a consumer towards a trader. It
argued that the addition of the reference to goods or services was intended to show that all types of
contract were covered by the directive. It would however be perverse for the legislature to seek to
broaden the meaning of an already general, unqualified noun (contracts) by adding a qualification
limiting the subject matter of the contract to goods and services. Far from broadening the scope of the
concept, in my view the ultimate addition to art 1(1) of the reference to goods or services is surely
designed to delineate unequivocally its scope.
21.
Moreover, the Commissions assertion that the original proposal for the directive was not restricted to
contracts for the supply of goods and services is not consistent with the scheme and tenor of that
proposal. While I do not wish to dwell on the details of a draft long since superseded by the final version, I
will, since the Commission has invoked the document in support of its view, refer to two provisions in the
proposal which to my mind demonstrate unequivocally that it was intended to apply only to synallagmatic
contracts for the supply of goods or services to a consumer.
22.
Article 3(2) provided that the contract (referred to as a doorstep contract) was to contain inter alia:
a description of the goods or service forming the subject matter of the contract, a time limit of delivery of
the goods or supply of services, the price,the terms of payment.
23.
Moreover, the proposal included as an annex a model form for the exercise by the consumer of his
right of cancellation. That form was as follows:
337

I hereby declare that I am cancelling the contract

for ..............................................................................................
(description of the goods or services)
amounting to .............................................................................
(price)
signed ........................................................................................
(date)
Name .........................................................................................
Address .....................................................................................
Date ..........................................................................................

24.
I accordingly cannot accept the Commissions contention that the proposal was intended to cover a
wider category of transactions than synallagmatic contracts for the supply of goods or services.
25.
At the hearing, the Commission put forward a further argument based on the wording of the directive:
it referred to art 3(2)(a), which provides that the directive does not apply to, inter alia, contracts for the
construction, sale and rental of immovable property or contracts concerning other rights relating to
immovable property. The Commission states that the last phrase refers to, for example, guarantees, and
concludes that guarantees must therefore be generally covered by the directive since otherwise there
would be no need for this specific exclusion. However, it is not obvious to me that other rights
necessarily encompasses guarantees: the examples given by the Commission in the explanatory
memorandum (COM(76) 544 final) to the original proposal of such other rights are applications for or
transfers of a mortgage, the granting of easements of right of way (see the commentary on art 2(d)).

The scheme of the directive

26.
As the French government points out, if a guarantee were covered by the directive so that the
guarantor had the option of withdrawing from it within a specified period, it would be necessary to provide
for the fate of the principal contract during that period: that there is no provision in the directive to that
effect supports the view that the directive was not intended to extend to such transactions. It is interesting
to note that the Commissions original proposal for the directive specifically prohibited the trader from
requiring the consumer to provide any surety for payment of the contract price before the expiration of the
cooling-off period (art 9): that provision was intended to prevent as far as possible the creation of a fait
accompli before the expiry of the cooling-off period (see the explanatory memorandum). It was not
included in the final version of the directive, which leaves the legal effects of renunciation to national law
(art 7).
27.
The Belgian government argues that a guarantee cannot be regarded as falling within the concept of
contract in art 1(1) of the directive: it is a non-synallagmatic contract where only one of the parties, the
guarantor, gives an undertaking to the other. In my view, that argument has some force; moreover I do not
accept the Commissions assertion that the reference to unilateral undertakings in the preamble to the
directive means that guarantees are covered. The Commission referred at the hearing to the first recital in
the preamble to the directive, which states:
Whereas it is a common form of commercial practice in the Member States for the conclusion of
a contract or a unilateral engagement between a 338 trader and consumer to be made away from
the business premises of the trader
The Commission asked what consequences could be drawn from that reference in the preamble other
than that the directive covers unilateral engagements such as the guarantee at issue.
28.
The answer to that question is in my view to be found in the legislative history of the directive and in
particular the Commissions own commentary on the original proposal for the directive. In that proposal,
art 1(1) stated that the directive applied to contracts between a consumer and a trader, and unilateral
engagements by a consumer towards a trader. The wording was changed between the amended
proposal in January 19785 and the final text.
5
Amendment to the proposal for a Council Directive to protect the consumer in respect of contracts which have been
negotiated away from business premises (OJ 1978 C127 p 6).

29.
As the German government points out, unilateral engagement in this context was probably intended
to refer to the making of an offer by the consumer in circumstances where the offer becomes binding
either on being made or on acceptance by the trader. That interpretation accords with the explanatory
memorandum to the original proposal6, in which the Commission stated:
6
Commentary on art 1.

The Directive also applies where the consumer engages himself unilaterally without any
corresponding obligations of the trader, for instance when ordering a[n] electrical appliance or when
binding himself unilaterally to acquire goods or to accept services. Even though a contract has not
yet been concluded in these cases, protection of the consumer is necessary, as the unilateral
engagement may already affect his interests.
30.
 Without express provision, a consumer who made an offer away from the traders business premises
would not benefit from the directive even though he would in effect have done what the directive seeks to
prevent, namely bound himself to a transaction away from the traders business premises with no
cooling-off period. The extension of the directive to such circumstances was originally achieved by the
express reference in art 1(1) to unilateral engagements. In the final text, the same result is achieved by
art 1(3) and 1(4), and it is presumably for that reason that the reference to unilateral engagements in art
1(1) was dropped. That interpretation is consistent with the retention of the reference to unilateral
engagements in the preamble.
31.
It may be concluded therefore that the directive is intended to apply to unilateral engagements of the
type mentioned, but not necessarily to all unilateral engagements.
32.
Finally, the Belgian, Finnish and German governments argue that the defendant was not a consumer
for the purposes of the directive, submitting that the notion of consumer encompasses the concept of
recipient of the goods or services supplied under the transaction in question. To my mind, there is some
force in that argument. The guarantee is clearly severable from the principal transaction (namely the
extension of credit), and in the context of the guarantee it is the guarantor who is the supplier of services
and the bank which is the recipient. The German government concludes that the directive is accordingly
inapplicable, since the consumer is the supplier. I prefer to turn the analysis 339 round and emphasise the
difficulty of identifying the guarantor as a consumer in any real sense.

The objectives of the directive

33.
Various parties refer to the objectives of the directive in support of their view as to whether the
directive applies to guarantees of the type at issue in the main proceedings.
34.
The defendant refers to the directives objective of consumer protection: in his view, the directive seeks
to establish as extensive consumer protection as possible 7.
7
Commentary on art 1.

35.
The Commission refers to the third and fourth recitals in the preamble, set out at para 8, above, and to
the explanatory memorandum to the original proposal for the directive in which it is indicated that the
proposed directive was intended to have as wide a scope as possible (commentary on art 1).
36.
The German government argues that the aim of the directive is not to give general protection to
individuals against any type of transaction which they undertake without forethought but rather to protect
them as consumers against particular and specified types of contract. In the case of a guarantee there is
no question of the satisfaction of the guarantors personal needs as envisaged by the directive: a
guarantor knows that he is not signing a consumer contract. That argument applies a fortiori where, as
here, the credit transaction in respect of which the guarantee is given is for both parties a commercial
transaction and the guarantee is as to the economic situation of one of the trading parties. In those
circumstances, the guarantor is not entitled to the rights granted to consumers acting in a purely private
capacity.
37.
The Belgian government submits that the directives objective of protection, achieved by conferring on
the consumer a right of withdrawal, is targeted at the dangers arising out of the particular form of selling
consisting in .direct marketing (restricting the consumers freedom of choice, lack of initiation by the
consumer, no possibility of comparing price or quality). The dangers threatening a guarantor, on the other
hand, arise not from the trader but rather from the principal debtor.
38.
In my view, the legislative history and context of the directive support the governments view that it was
not intended to extend to guarantees of the type at issue in the main proceedings.
39.
It is of course incontrovertible that the directive seeks to protect consumers. It does not however follow
from that proposition that all consumers are protected in all circumstances by the directive: like other
directives with consumer protection as their aim, the directive applies to certain transactions only 8.
8
See eg Council Directive (EEC) 84/450 relating to the approximation of the laws, regulations and administrative
provisions of the member states concerning misleading advertising (OJ 1984 L250 p 17), Council Directive (EEC)
87/102 for the approximation of the laws, regulations and administrative provisions of the member states concerning
consumer credit (OJ 1987 L42 p 48), Council Directive (EEC) 93/13 on unfair terms in consumer contracts (OJ 1993
L95 p 29) and most recently the proposed directive of the European Parliament and of the Council on the protection of
consumers in respect of contracts negotiated at a distance.

40.
The third recital in the preamble, set out above and invoked by the Commission in support of its
argument for a wide construction, refers to the 340 Communitys preliminary and secondary programmes
for a consumer protection and information policy. Those two programmes indicate that the intention was
(in the words of the preliminary programme) to protect purchasers of goods or services against the abuse
of power by the seller, in particular against high-pressure selling methods 9, and (in the words of the
second programme) to protect purchasers of goods or services against unfair sales practices and high-
pressure selling methods10.
9
See Council Resolution of 14 April 1975 on a preliminary programme of the European Economic Community for a
consumer protection and information policy (OJ 1975 C92 p 1, para 19(i) of the annex).
10
See Council Resolution of 19 May 1981 on a second programme of the European Economic Community for a consumer
protection and information policy (OJ 1981 C133 p 1, para 28(1) of the annex).

41.
Admittedly, it is stated in the explanatory memorandum to the original proposal for the directive that the
need for protecting the consumer is not limited to any specific type of contract for instance contracts for
the supply of goods, but extends to all contracts which are initiated by a trader away from business
premises (para 1.3). That statement however must be seen in its context: the explanatory memorandum
opens with the proposition Where contracts for goods or services are initiated away from business
premises of a trader the consumer generally needs special protection. Examples are given of certain
types of contracts [where] a particular consumer protection might be required: especially insurance
contracts, consumer credit contracts, contracts relating to the sale of shares, investment funds, etc. or
contracts on distance education (para 1.3). All those contracts are bipartite, synallagmatic contracts for
the provision of goods or trader to a consumer. The same may be said of all the other examples given in
the explanatory memorandum of transactions to which the proposed directive applies or would in the
absence of specific exemption apply, namely an order for an electrical appliance, a unilateral undertaking
to acquire goods or to accept services, contracts relating to the installation of a heating system, the
maintenance of a heating oil tank, the repair of a roof, and small doorstep sales such as milk and bread.
42.
I am accordingly not persuaded that the directive, which in my view aims to protect the consumer party
to contracts for the supply of goods or services negotiated away from business premises, extends to the
protection of a guarantor in the circumstances of the main proceedings.
43.
I would however emphasise that it does not follow from that conclusion that member states may not
protect a guarantor in such circumstances as a matter of national law. Article 8 of the directive permits
member states to adopt or maintain more favourable provisions to protect consumers in the field which it
covers It would not therefore be contrary to the directive for the national court in this case to find that as a
matter of German law the guarantee is within the 1986 Law. Moreover, if it is correctas is suggested in
certain of the observations made to the courtthat the defendant was wholly uninformed of the nature of
the commitment into which he was entering, and which he was even pressurised into signing, it may well
be that other national law remedies are available in the spheres of misrepresentation or undue influence.
341

Conclusion
44.
Accordingly the question referred by the Bundesgerichtshof should in my opinion be answered as
follows:
A guarantee given to a financial institution by an individual who is not acting in the course of his
trade or profession, in order to secure a loan by that institution to a third party, is not within the
scope of Council Directive (EEC) 85/577 to protect the consumer in respect of contracts negotiated
away from business premises.

17 March 1998.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By order of 11 January 1996, received at the Court of Justice of the European Communities on 15
February 1996, the Bundesgerichtshof (the Federal Court of Justice) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty a question concerning the interpretation of Council
Directive (EEC) 85/577 to protect the consumer in respect of contracts negotiated away from business
premises (OJ 1985 L372 p 31).
2.
That question has been raised in proceedings between Bayerische Hypotheken- und Wechselbank AG
(the bank) and Edgar Dietzinger concerning the performance of a contract of guarantee concluded by Mr
Dietzinger with the bank.
3.
Article 1(1) of Directive 85/577 provides as follows:
This Directive shall apply to contracts under which a trader supplies goods or services to a
consumer and which are concluded:during an excursion organised by the trader away from his
business premises, orduring a visit by a trader (i) to the consumers home or to that of another
consumer; (ii) to the consumers place of work; where the visit does not take place at the express
request of the consumer.
4.
Next, art 2 provides:
For the purposes of this Directive: consumer means a natural person who, in transactions
covered by this Directive, is acting for purposes which can be regarded as outside his trade or
profession; trader means a natural or legal person who, for the transaction in question, acts in his
commercial or professional capacity, and anyone acting in the name or on behalf of a trader.
5.
Under art 4 of Directive 85/577, traders are required to give consumers written notice of their right to
cancel the contract within a specified period. Article 5 provides that that period is to be not less than
seven days from receipt by the consumer of the notice of his right to renounce the effects of the contract.
6.
Mr Dietzingers father ran a building firm in respect of which the bank, inter alia, granted a current
account overdraft facility. On 11 September 1992, Mr Dietzinger gave a direct recourse written guarantee,
for a sum not to exceed DM 100,000, covering his parents obligations to the bank.
7.
The contract of guarantee was concluded at the house of Mr Dietzingers parents during a visit by an
employee of the bank to which Mr Dietzingers mother had agreed over the telephone. Mr Dietzinger was
not informed of his right of cancellation.
342
8.
In May 1993, the bank called in, with immediate effect, all the loans which it had granted to Mr
Dietzingers parents, which at that time totalled more than DM 16m. It also sued Mr Dietzinger for
payment of DM 50,000 under the guarantee. Mr Dietzinger sought to renounce the guarantee,
maintaining that he had not been informed of his right of cancellation, contrary to the Gesetz ber den
Widerruf von Haustrgeschften und hnlichen Geschften (the Law on the Cancellation of Doorstep
Transactions and Analogous Transactions, BGBl I, p 122) of 16 January 1986, which transposed Directive
85/577 into German law.
9.
The Landgericht (the Regional Court) found in favour of the bank. Mr Dietzinger then appealed to the
Oberlandesgericht (the Higher Regional Court), which quashed the decision given at first instance.
10.
The Bank then appealed on a point of law to the Bundesgerichtshof, which held that an interpretation
of Directive 85/577 was necessary in order to determine the dispute. It therefore referred the following
question to the Court of Justice for a preliminary ruling:
Where a contract of guarantee or suretyship is concluded under German law between a
financial institution and a natural person who is not acting in that connection in the course of his
trade or profession, in order to secure a claim by the financial institution against a third party in
respect of a loan, is it covered by the words contracts under which a trader supplies goods or
services to a consumer (Article 1(1) of Council Directive 85/577/EEC of 20 December 1985 to
protect the consumer in respect of contracts negotiated away from business premises, OJ 1985
L372, p 31)?
11.
By its question, the Bundesgerichtshof is asking in effect whether a contract of guarantee concluded
by a natural person who is not acting in the course of a trade or profession is covered by Directive 85/577.
12.
Mr Dietzinger and the Commission consider that Directive 85/577 applies to a contract of guarantee by
virtue of the directives aim, which is to protect those consumers who conclude a contract where, because
it involved doorstep selling, they were unable to prepare themselves for its negotiation. Like a purchaser,
a guarantor undertakes to perform obligations and is even more in need of protection since he receives
no consideration in exchange for his commitment.
13.
In the Commissions view, art 1 of Directive 85/577 is applicable to any contract concluded between a
natural person and a trader who, in the course of his business activities, supplies goods or services to
consumers in general, even if the contract in question does not involve such consideration. In referring to
contracts under which a trader supplies goods or services to a consumer, the directive is simply making
clear that its scope is not restricted to sellers of goods.
14.
The German, Belgian, French and Finnish governments, on the other hand, consider that guarantees
are not covered by Directive 85/577, essentially because a guarantee is not a contract under which a
trader supplies goods or services to a consumer within the meaning of art 1.
15.
According to those governments, the wording of the provision implies that goods or services are
supplied by a trader to a consumer who relies on the protection afforded by Directive 85/577, so that it is
not enough for the trader to be a supplier of goods or services in general. They point out that such an
interpretation is strongly suggested by the English version of the directive (contracts under which a trader
supplies goods or services to a consumer). In circumstances such as those of the instant case, the
guarantors commitment 343 gives rise to no consideration, in the sense that the guarantor receives no
goods or services from the trader to whom the commitment was given.
16.
Those governments argue further that Directive 85/577 does not cover guarantees; if it did, the
directive would have contained specific rules providing, in particular, for the fate of the contract whose
performance is guaranteed by the guarantor in the event of his exercising the right of cancellation.
Consequently, protection of guarantors is a matter for national law alone. In particular, the French
government argues that, since Directive 85/577 does not govern the effects, on the principal contract, of
possible invalidity of a contract of guarantee, such guarantees must, in view of their ancillary nature, be
excluded from the scope of the directive.
17.
The court observes that, according to art 1, Directive 85/577 applies to contracts under which a trader
supplies goods or services to a consumer which are concluded away from the traders business
premises, unless the trader was expressly requested by the consumer to visit him with a view to the
negotiation of the contract.
18.
In determining whether a contract of guarantee securing performance of a credit agreement by the
principal debtor can fall within the scope of Directive 85/577, it should be noted that, apart from the
exceptions listed in art 3(2), the scope of the directive is not limited according to the nature of the goods
or services to be supplied under a contract; the only requirement is that the goods or services must be
intended for private consumption. The grant of a credit facility is indeed the provision of a service, the
contract of guarantee being merely ancillary to the principal contract, of which in practice it is usually a
precondition.
19.
Furthermore, nothing in the wording of the directive requires that the person concluding the contract
under which goods or services are to be supplied be the person to whom they are supplied. Directive
85/577 is designed to protect consumers by enabling them to withdraw from a contract concluded on the
initiative of the trader rather than of the customer, where the customer may have been unable to see all
the implications of his act. Consequently, a contract benefiting a third party cannot be excluded from the
scope of the directive on the sole ground that the goods or services purchased were intended for the use
of the third party standing outside the contractual relationship in question.
20.
In view of the close link between a credit agreement and a guarantee securing its performance and the
fact that the person guaranteeing repayment of a debt may either assume joint and several liability for
payment of the debt or be the guarantor of its repayment, it cannot be excluded that the furnishing of a
guarantee falls within the scope of the directive.
21.
Moreover, the possible termination of a contract of guarantee concluded in the context of doorstep
selling within the meaning of Directive 85/577 is merely one particular situation where the question may
arise as to the effect of the possible invalidity of an ancillary contract upon the principal contract. In those
circumstances, the mere fact that the directive contains no provision governing the fate of the principal
contract where the guarantor exercises the right of renunciation conferred by art 5 cannot be taken to
mean that the directive does not apply to guarantees.
22.
However, it is apparent from the wording of art 1 of Directive 85/577 and from the ancillary nature of
guarantees that the directive covers only a guarantee ancillary to a contract whereby, in the context of
doorstep selling, a consumer assumes obligations towards the trader with a view to obtaining goods or
services from him. Furthermore, since the directive is designed to protect only 344 consumers, a
guarantee comes within the scope of the directive only where, in accordance with the first indent of art 2,
the guarantor has entered into a commitment for a purpose which can be regarded as unconnected with
his trade or profession.
23.
The answer to the question referred to the court must therefore be that, on a proper construction of the
first indent of art 2 of Directive 85/577, a contract of guarantee concluded by a natural person who is not
acting in the course of his trade or profession does not come within the scope of the directive where it
guarantees repayment of a debt contracted by another person who, for his part, is acting within the
course of his trade or profession.

Costs
24.
The costs incurred by the German, Belgian, French and Finnish governments and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the action pending before the national
court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the question referred to it by the
Bundesgerichtshof by order of 11 January 1996, hereby rules: on a proper construction of the first indent
of art 2 of Council Directive (EEC) 85/577 to protect the consumer in respect of contracts negotiated away
from business premises, a contract of guarantee concluded by a natural person who is not acting in the
course of his trade or profession does not come within the scope of the directive where it guarantees
repayment of a debt contracted by another person who, for his part, is acting within the course of his trade
or profession.

345

[1998] All ER (EC) 346

Jules Dethier quipement SA v Dassy and another

(Case C-319/94)
EUROPEAN COMMUNITY; Social policy, Law relating to undertakings: EMPLOYMENT; Transfer of
undertakings
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE CHAMBER), MANCINI (RAPPORTEUR), AND MURRAY,
ADVOCATE GENERAL LENZ
11 JULY 1996, 12 MARCH 1998
European Community Employment Continuity of employment Transfer of trade, undertaking or
business Transfer of undertaking being wound up by court Employee dismissed shortly before
transfer and not taken on by transferee Whether legal transfer for purposes of Community directive
Whether transferor and transferee having power to dismiss employees for economic, technical or
organisational reasons Council Directive (EEC) 77/187, arts 1, 4.

D was employed by a Belgian undertaking, Sovam. In 1991, Sovam fell into financial difficulties and, at
the request of the shareholders, the local commercial court put the company into liquidation. The court
appointed a liquidator who dismissed D and shortly thereafter transferred Sovams assets to another
company, Dethier. D brought an action before a local labour court seeking that both Sovam and Dethier
be held jointly and severally liable for sums due to him by way of compensation in lieu of notice, paid
holiday leave and bonuses. He contended that the transfer of Sovams assets constituted a contractual
transfer within the meaning of the Belgian legislation implementing Council Directive (EEC) 77/187 on the
approximation of the laws of the member states relating to the safeguarding of employees rights in the
event of transfers of undertakings, businesses or parts of businesses, and that Dethier was therefore
jointly and severally liable for those payments. The labour court ruled in Ds favour and Dethier appealed,
contending that the liquidation of Sovam was comparable to an insolvency, in which case the transfer fell
outside the scope of Directive 77/187, and further that any joint and several liability on the part of the
transferee could benefit only those employees taken on by the transferee and not those dismissed before
the transfer. After finding that there had been a transfer of an undertaking within the meaning of the
directive, the national court stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling the questions: (i) whether, on a proper construction of art 1(1) 1 of the
directive, the directive applied in the event of the transfer of an undertaking which was being wound up
voluntarily or by the court; and (ii) whether the right set out in art 4(1) 2 of the directive to dismiss
employees for economic, technical or organisational reasons in the context of a transfer was available
only to the transferee.
1
Article 1(1), so far as material, is set out at p 349 f, post
2
Article 4(1), so far as material, is set out at p 349 h, post

Held (1) When considering whether Directive 77/187 applied to the transfer of an undertaking subject to
an administrative or judicial procedure, the court should 346 take into account, primarily, the purpose of
the procedure in question and whether it entailed the continuation of the business with a view to its
retrieval: it should also take account of the form of the procedure, in particular in so far as it meant that
the undertaking continued or ceased trading. The criterion relating to the purpose of the procedure for
winding up by the court, in the instant case, was not conclusive, since liquidation proceedings could be
used to bring a companys activities to an end whatever the reason for that course. However it was
apparent from the form of the procedure, in which liquidator, although appointed by the court, was an
organ of the company who sold the assets under the supervision of the general meeting that the
undertaking continued to trade while it was being wound up by the court, and that the continuity of the
business was thus assured when the undertaking was transferred. There was accordingly no justification
for depriving the employees of the rights guaranteed to them under the directive. It followed that, on a
proper construction of art 1(1), the directive applied in the event of a transfer of an undertaking which was
being wound up by the court if the undertaking continued to trade (see p 363 j, p 364 b to g and p 365 h,
post); dUrso v Ercole Marelli Elettromeccanica Generale SpA Case C-362/89 [1991] ECR I-4105 and
Spano v Fiat Geotech SpA Case C-472/93 [1995] ECR I-4321 applied.
(2) On its proper construction, art 4(1) allowed both the transferor and the transferee to dismiss
employees for economic, technical or organisational reasons. Moreover, the contract of employment of a
person unlawfully dismissed before the transfer of an undertaking had to be regarded as still extant
against the transferee even if the dismissed employee was not taken on by him after the undertaking was
transferred. In those circumstances, the employee could claim, as against the transferee, that his
dismissal was unlawful (see p 365 b e f j, post); Bork (P) International A/S (in liq) v Foreningen af
Arbejasledere i Danmark Case 101/87 [1988] ECR 3057 and Foreningen af Arbejdsledere i Danmark v
Daddys Dance Hall A/S Case 324/86 [1988] ECR 739 applied.
Notes
For employees rights on transfer of undertakings, see 52 Halsburys Laws (4th edn) para 2120.

Cases cited
Abels v Administrative Board of the Bedrijfsvereniging voor de Metaalindustrie en de Electrotechnische
Industrie Case 135/83 [1985] ECR 469.
Bork (P) International A/S (in liq) v Foreningen af Arbejasledere i Danmark Case 101/87 [1988] ECR
3057.
Centro Servizi Spediporto Srl v Spedizioni Marittima del Golfo Srl Case C-96/94 [1995] ECR I-2883.
Corsica Ferries Italia Srl v Corpo dei Piloti del Porto di Genova Case C-18/93 [1994] ECR I-1783.
dUrso v Ercole Marelli Elettromeccanica Generale SpA Case C-362/89 [1991] ECR I-4105.
Foreningen af Arbejdsledere i Danmark v A/S Danmols Inventar (in liq) Case 105/84 [1985] ECR 2639.
Foreningen af Arbejdsledere i Danmark v Daddys Dance Hall A/S Case 324/86 [1988] ECR 739.
Landsorganisationen i Danmark for Tjenerforbundet i Danmark v Ny Mlle Kro Case 287/86 [1987] ECR
5465.
Loureno Dias v Director da Alfndega do Porto Case C-343/90 [1992] ECR I-4673.
347
Redmond (Dr Sophie) Stichting v Bartol Case C-29/91 [1992] I-ECR 3189.
Salonia v Poidomani Case 126/80 [1981] ECR 1563.
Spano v Fiat Geotech SpA Case C-472/93 [1995] ECR I-4321.
Spijkers v Gebr Benedik Abattoir CV Case 24/85 [1986] ECR 1119.

Reference
By judgment of 1 December 1994, the Cour du Travail (the Higher Labour Court), Lige, referred to the
Court of Justice of the Eureopan Communities for a preliminary ruling under art 177 of the EC Treaty a
number of questions (set out at p 362 d f, post) on the interpretation of Council Directive (EEC) 77/187 on
the approximation of the laws of the member states relating to the safeguarding of employees rights in
the event of transfers of undertakings, businesses or parts of businesses. Those questions arose in
proceedings between Jules Dethier quipement SA, on the one hand, and Jules Dassy and Sovam
SPRL, on the other, concerning payment of compensation in lieu of notice and of other sums of
compensation. Written observations were submitted on behalf of: the Belgian government, by J Devadder,
General Adviser in the Ministry of Foreign Affairs, External Trade and Cooperation with Developing
Countries, acting as agent and the European Commission, by M Wolfcarius, of its Legal Service, and H
Kreppel, a national civil servant on secondment to that service, acting as agents. The language of the
case was French. The facts are set out in the opinion of the Advocate General.

11 July 1996.

The Advocate General (C O Lenz)

delivered the follwing opinion (translated from the German).

A INTRODUCTION
1.
In this case, the Cour du Travail (the Higher Labour Court), Lige, has referred to the Court of Justice
of the European Communities questions on the application and interpretation of Council Directive (EEC)
77/187 on the approximation of the laws of the member states relating to the safeguarding of employees
rights in the event of transfers of undertakings, businesses or parts of businesses (OJ 1977 L61 p 26), in
particular the question whether the directive applies to the transfer of an undertaking which is undergoing
liquidation.
2.
The action before the national court is between Mr Dassy and Sovam SPRL, his former employer, on
the one hand, and Jules Dethier quipement SA, on the other.
3.
Mr Dassy had been employed by Sovam SPRL since the beginning of 1974 as head of its after-sales
service department. At the beginning of the 1990s, the company got into financial difficulties and suffered
a substantial reduction in turnover. In March 1991, an auditors report revealed that the value of the
companys net assets had fallen below the amount of its share capital. Since the shareholders could not
agree on the course to be followed, the company in the end applied to be wound up by the court. The
Tribunal de Commerce (the Commercial Court), Huy, accordingly made an order on 15 May 1991 putting
the company into liquidation under supervision of the court and appointed a liquidator.
4.
 According to the national court and the Commission, liquidation of a company under Belgian law must
be taken to mean all the steps which, following dissolution of a commercial company, are aimed at paying
creditors out of the corporate assets and distributing any balance to the members. That procedure 348
applies to all commercial companies with legal personality. The reason for the companys dissolution does
not matter. During the liquidation, the company acts through the liquidator. He is an organ of the company
and represents it vis--vis third parties. Under the law, liquidators are appointed either in the articles of
association or by general meeting. If the general meeting does not agree on a candidate by the requisite
majority, a liquidator is appointed by the court. In the latter event, the liquidation is one conducted under
court supervision, as in this case. The only difference from voluntary liquidation is that the procedure for
appointing a liquidator is different.
5.
On 5 June 1991, the liquidator appointed by the court in this case terminated Mr Dassys contract of
employment.
6.
By an agreement of 27 June 1991, the company in liquidation was transferred by the liquidator to
Dethier. On 10 July 1991, the Tribunal de Commerce confirmed that transfer of assets. The referring court
and the Commission both agree that this endorsement by the court was not necessary. In any event, it is
not required under the liquidation procedure.
7.
On 22 May 1992, Mr Dassy brought an action against Dethier, which on the basis of a contractual
transfer of the undertaking he considered to be jointly and severally liable for payment of the sums due
from Sovam SPRL (in liq). The Tribunal de Commerce fixed the sum payable by the company in
liquidation at BF 1,643,726 and held that Dethier was jointly and severally liable for payment of that sum.
8.
With the Cour du Travail, Lige, which has made this reference, Dethier has now lodged appeal. The
Belgian court takes the view that, in this case, an undertaking has been transferred for the purposes of
the directive. However, it is not sure whether the directive can apply because it is open to doubt whether
the transfer of a company in liquidation is a contractual transfer (cession conventionelle) of an
undertaking for the purposes of art 1(1) of the directive.
9.
Article 1(1) of the directive states: This Directive shall apply to the transfer of an undertaking, business
or part of a business to another employer as the result of a legal transfer 3 or merger.
3
Translators note: cession conventionelle in the French text.

10.
The safeguarding of employees rights in the event of transfer of an undertaking is dealt with in art
3(1): The transferors rights and obligations arising from a contract of employment or from an
employment relationship existing on the date of a transfer within the meaning of Article 1(1) shall, by
reason of such transfer, be transferred to the transferee.
11.
In order to prevent circumvention of that provision, art 4 prohibits dismissals effected solely on the
grounds of transfer of an undertaking. Article 4(1) states:
The transfer of an undertaking, business or part of a business shall not in itself constitute
grounds for dismissal by the transferor or the transferee. This provision shall not stand in the way of
dismissals that may take place for economic, technical or organizational reasons entailing changes
in the workforce.
12.
Finally, art 7 allows member states to apply or introduce laws, regulations or administrative provisions
which are more favourable to employees.
13.
According to the national court, the directive was transposed into Belgian law by Collective Agreement
No 32 bis, as amended by Collective Agreement No 34932 quater. That agreement safeguards
employees rights whenever there is a change of employer resulting from the contractual transfer of an
undertaking. It also guarantees certain rights for employees who are taken on in the case of a takeover of
assets following insolvency or of a court-approved arrangement with creditors. Collective Agreement No
32 did not apply to insolvency or similar proceedings. The progenitors of Collective Agreement No 32
considered, however, that employees of an undertaking adjudged insolvent or the subject of a court-
approved arrangement were in a similar position to employees of an undertaking that has been
transferred and that they, therefore, also deserved a minimum degree of protection.
14.
The collective agreement also provides that the transferor and transferee are to be jointly and
severally liable for the payment of debts existing at the date of the transfer and resulting from contracts of
employment existing at that date.
15.
Should the directive be applicable to this case, the national court raises the further question of how
termination of a contract of employment occurring immediately before the company is transferred is to be
viewed under the directive.
16.
It has therefore referred the following questions to the court for a preliminary ruling:
(1) Does Council Directive 77/187 apply where the transfer is effected by a company in
voluntary liquidation, a procedure whose aim, in the absence of continued trading, is liquidation by
realisation of the assets? Is the answer the same where the transferor is being wound up by the
court?
(2) Where the contracts of employment of all the employees have been terminated by the
liquidator and only some of those employees have been re-engaged for the purposes of the
liquidation, may the dismissals of the employees not subsequently taken on by the transferee be
regarded as having taken place for economic, technical or organizational reasons within the
meaning of Article 4(1) of the directive? Must the power to dismiss such employees for such
reasons be left, on the contrary, to the transferee alone?
May staff not taken on by the transferee claim, as against him, merely because an economic
entity was transferred shortly after their dismissal for economic, technical or organisational reasons,
that the measure taken in their regard by the transferor was unlawful if the transfer agreement does
not provide for their re-engagement?

B OPINION

The first question

17.
In the first question, the national court asks whether the directive is applicable if the transfer is effected
by a company in liquidation. It is apparent from the wording of the question and the rest of the preliminary
reference that the national court considers that a transfer of an undertaking has occurred for the purposes
of the directive. It explains that the liquidator of Sovam SPRL and Dethier entered into an agreement for
transfer of the business assets. That agreement provided, inter alia, that the liquidator was to transfer, for
a sum of BF 2m, the business assets (importation and distribution of domestic electrical appliances and
equipping and fitting out buildings), which included: the goodwill, commercial name and logo; the office
furniture, machines, tools and vehicles; all patents, franchises and licences vested in the transferor; the
lease.
350
In addition to other provisions, the agreement also required the transferee to take over, on the same
conditions, three members of staff designated by name and to inform the transferor of any other re-
engagements. According to the national court, the agreement therefore covers the entire business of
Sovam SPRL. It refers to the case law of the Court of Justice and concludes that in this case an
undertaking was transferred for the purposes of the directive.
18.
The Court of Justice has consistently held that it is for the national court to establish, on the basis of
the criteria laid down by the court, whether there is, in the case in point, a contractual transfer of an
undertaking for the purposes of the directive (see the judgments in Spijkers v Gebr Benedik Abattoir CV
Case 24/85 [1986] ECR 1119 (para 14) and Dr Sophie Redmond Stichting v Bartol Case C-29/91 [1992] I-
ECR 3189 (paras 2325)).
19.
However, since the transferred undertaking was in liquidation, the national court raises the question
whether the transfer in this case was also a contractual transfer.
20.
On that point reference should be made to the judgment of the Court of Justice in Abels v
Administrative Board of the Bedrijfsvereniging voor de Metaalindustrie en de Electrotechnische Industrie
Case 135/83 [1985] ECR 469. In that case, the court held that the scope of art 1(1) of the directive could
not be appraised solely on the basis of a textual interpretation: this was precluded by the various
language versions of the provision and the different meanings of the concept of contractual transfer in the
insolvency laws of individual member states. For that reason, the provisions meaning had to be clarified
in the light of the scheme of the directive, its place in the system of Community law in relation to the rules
on insolvency, and its purpose (see [1985] ECR 469 (paras 1113)).
21.
As regards the relationship between the directive and insolvency law, the Court of Justice referred to
the special procedures characterising insolvency law and intended to weigh up the various interests
involved, in particular those of the various classes of creditors. Those special features were encountered
in all the legal systems of the member states and were also confirmed in Community law. The fact that
insolvency law was the subject of specific rules in the legal systems of the member states and in the
Community legal order and that those rules were very different in the various member states led the court
to conclude that if the directive had been intended to apply also to transfers of undertakings occurring in
the context of such proceedings, an express provision would have been included for that purpose.
22.
According to the court, that interpretation also necessarily followed from consideration of the purpose
of the directive, namely to ensure that the restructuring of undertakings within the common market did not
adversely affect the employees in the undertakings concerned. Since there were major differences of
opinion with regard to the consequences, for the protection of employees, of applying the directive to
insolvency proceedings, a serious risk of general deterioration in working and living conditions of workers,
contrary to the social objectives of the Treaty, could not be ruled out. Extending the scope of the directive
to insolvency proceedings might dissuade a potential transferee from acquiring an undertaking on
conditions acceptable to its creditors, who would then have to sell the assets of the undertaking
separately. That would entail the loss of all the jobs in the undertaking, detracting from the effectiveness
of the directive.
23.
The court concluded that the member states were not obliged to extend the rules laid down in the
directive to transfers of undertakings, businesses or parts of 351 businesses occurring in the context of
insolvency proceedings. The member states were, however, at liberty independently to apply the
principles of the directive, wholly or in part, on the basis of their national law alone (see [1985] ECR 469
(paras 2324)).
24.
The court then had to rule on the question whether the directive applied to judicial leave to suspend
payment of debts. To do that, it compared that procedure with insolvency proceedings. It established first
that the procedure to obtain court leave to suspend payment of a debt was also of a judicial nature,
although the supervision exercised by the court was more limited than in the case of insolvency. That is to
say the special rules applicable to insolvency either did not apply or applied only to a limited extent.
25.
As a further criterion, it looked at the purpose of the procedure and found that it served primarily to
safeguard the assets of the undertaking and, where possible, enable the business to carry on, with a view
to reaching an arrangement ensuring that it would be able to continue trading in the future. The court
concluded that the reasons for not applying the directive in the context of insolvency proceedings did not
apply to procedures for judicial leave to suspend payment of debts. The fact that such procedures can
lead to the debtors being adjudged insolvent does not matter because it is clear that they take place at a
stage prior to insolvency (see [1985] ECR 469 (paras 2829)).
26.
In a further important case, dUrso v Ercole Marelli Elettromeccanica Generale SpA Case C-362/89
[1991] ECR I-4105, the court had to consider whether the directive applied to transfers of undertakings
effected by undertakings under special administration.
27.
The court first applied the criterion of the extent of court supervision, which it had also mentioned in
Abels. In view of the differences between the legal systems of the member states, to which reference had
already been made in Abels, the scope of the concept of contractual transfer could not be determined
according to the kind of supervision exercised by the administrative or judicial authority over transfers of
undertakings in the course of a specific creditors arrangement procedure. The court concluded: Given all
the considerations set out in the judgment in the Abels case, the decisive test is therefore the purpose of
the procedure in question (see [1991] ECR I-4105 (paras 2526)).
28.
The special administration procedure at issue was applied by a decree which had, or could have, two
kinds of effects. On the one hand, it had to be assimilated to the decree ordering compulsory
administrative liquidation, which had effects identical in substance to those of bankruptcy proceedings. On
the other hand, the decree could also authorise the undertaking to continue trading for a specified period
under the supervision of an auditor. The auditor had the power to draw up a programme which had to
contain a restructuring plan.
29.
According to the court, application of the directive was dependent on whether or not the undertaking
was authorised to continue trading. If no decision was taken was taken on authorising the undertaking to
continue trading or the period of validity of such authorisation had expired
the aim, consequences and risks of a procedure such as the compulsory administrative
liquidation procedure are comparable to those which led this Court to conclude, in its judgment in
the Abels case, that Article 1(1) of the directive did not apply to transfers of an undertaking,
business or part of a business in a situation in which the transferor had been adjudged insolvent.
Like insolvency proceedings, that procedure is designed to liquidate the 352 debtors assets in
order to satisfy the body of creditors, and transfers effected under this legal framework are
consequently excluded from the scope of the directive.
Without that exclusion, the risk of a deterioration in the living and ,working conditions of workers could not
be ruled out (see [1991] ECR I-4105 (para 31)).
30.
However, if the decree authorised the undertaking to continue trading under the supervision of an
auditor, the primary purpose of that procedure was to give the undertaking some stability allowing its
future activity to be safeguarded.
The social and economic objectives thus pursued cannot explain nor justify the circumstance
that, when all or part of the undertaking is transferred, its employees lose the rights which the
directive confers on them under the conditions which it lays down. (See [1991] ECR I-4105 (para
32).)
31.
Consequently, the essential criterion for the court is the purpose of the procedure in question. The
objective of the directive itself and the arrangement of the particular procedure are further criteria.
32.
 Those criteria must now be applied to the liquidation under examination in this case. The national court
explains that liquidation of a company must be taken to entail all the steps which, following dissolution of
a commercial company, are aimed at paying creditors out of the corporate assets and distributing any
balance to the members. According to legal writers, liquidation is to be preferred to commencement of
insolvency proceedings because it allows the optimal, or least disadvantageous, realisation of the assets,
and all or part of the economic activities which are still viable can be safeguarded.
33.
The Commission states that the tasks of a liquidator are to realise the assets, settle debts and
distribute any balance to the members. The objective, therefore, is liquidation by means of the realisation
of assets. Liquidation occurs for the benefit of the company, whereas in insolvency proceedings the
administrator acts exclusively for the benefit of the creditors. The assets must be realised at the best
possible price. According to the Commission, the Belgian legislature has here made use of a fiction.
Although a commercial companys legal personality is necessarily brought to an end by dissolution, it is
treated as continuing to exist for the purposes of liquidation. A company in liquidation has legal personality
solely for the purposes of realising the assets, settling debts and distributing any balance. Pursuing the
activity of the dissolved company per se is no longer permitted, save, exceptionally, in so far as it helps to
achieve the new object of the companyliquidationwhich has replaced the companys previous
economic activity. According to the Commission, a company in liquidation can, therefore, only conclude
business already started. It may not begin new business, even if it falls within the scope of its previous
corporate activity (except if it serves the purposes of the liquidation). It is often necessary for the company
to continue trading on such a temporary basis in order to prevent the value of the economic entity to he
transferred from depreciating. The entity must continue to trade in order to facilitate the subsequent
transfer. In that case the liquidator can also enter upon new transactions without special authorisation.
However, he must remain aware that continued trading is permitted only on a temporary basis and with a
view to subsequent realisation of the company assets under the best possible conditions. According to the
Commission, the liquidator is personally liable for losses if he continues to trade without restriction. There
is no need to prove negligence in his management of the company.
353
34.
The decision that the company is to continue trading for the time being is adopted by the requisite
majority at a general meeting. It does not require authorisation by the court.
35.
The Commission also states in its observations that voluntary liquidation is never an alternative to
insolvency. If the conditions for instituting insolvency proceedings are satisfied, liquidation is no longer
possible. The only exception is if the company is in temporary difficulties or unable for the time being to
determine whether the value of its assets exceed its liabilities. The guarantees enjoyed by creditors in
insolvency proceedings are much more extensive than in liquidation proceedings. Furthermore, in the
former the creditors are directly represented by the administrator. That could suggest that in liquidation
proceedings the objective of satisfying the creditors claims is not as central as in insolvency proceedings.
36.
The fact remains, however, that the purpose of liquidation is also to realise the assets, a purpose
almost identical to that of insolvency. Nevertheless, it appears to me that winding up the companys affairs
is more central to liquidation proceedings, whereas, in insolvency proceedings, only satisfaction of the
creditors claims is important. However, whether that in itself is sufficient for the directive to be applied to
this case, as the Belgian government suggests, seems to me to be questionable.
37.
The Commission also considers that the purposes of insolvency and liquidation are almost identical,
but in arriving at its proposed solution to this case it relies almost exclusively on the judgment in dUrso.
That is to say, it looks to see only whether or not a decision to continue trading was adopted. The
Commission considers it irrelevant that trading is continued not in order to permit the company to survive
but so as to be able to achieve more fully the purposes of the liquidation. In its view, the undertaking must
retain its identity and, while trading continues, must be able to continue the same type of business. It also
relies on the protective purpose of the directive, which would preclude the protection of employees from
being removed by a resolution in general meeting that the company go into liquidation.
38.
In the Commissions view, the directive does not apply, however, if the company in liquidation resolves
to cease business in order to sell the assets. Discontinuing trade definitively in that way stops the
directive from applying. The sole purpose of the liquidation is then to sell the assets and bring the
existence of the undertaking to an end. If the directive were to apply in such a case, the employees would
be exposed to an even greater risk. Furthermore, if trading is discontinued for too long, it can no longer be
assumed that the same entity continues to exist and that the business is being continued.
39.
In my view, the distinction drawn by the court in dUrso cannot simply be transposed to the case of
liquidation. In dUrso, the business was continued with a view to reconstruction, so as to ensure that it
could also trade in the future. The same was true of the proceedings for judicial leave to suspend
payment of debts in Abels. In this case, however, trading is being continued solely with the aim of
dissolving the company. Trading is not directed towards the future, but is being continued only until the
undertaking is sold.
40.
 In dUrso, the undertaking was in fact transferred in the end, but, having regard to the protective
purpose of the directive, that transfer could be treated differently from the transfer in this case. In dUrso,
an undertaking which had been, or was being, reconstructed was transferred, and a purchaser is perhaps
more easily found for such an undertaking than for one which is in liquidation. In 354 that respect,
applying the directive would have no disadvantages for the employees. That is not soas the court held
in Abelsin the case of insolvency proceedings: in this event, applying the directive might have an
adverse effect on the employees.
41.
In dUrso, however, the sole criterion applied by the court was whether the undertaking continued to
trade: the directive applied so long as trading was continued, since there was no reason why employees
should not enjoy the protection of the directive if the undertaking continued to trade. If it ceases to trade,
however, the same criteria met in insolvency proceedings apply and the directive ceases to be applicable
from that point. In other words, despite the applicability of criteria comparable to those met in the case of
insolvency, the directive applies solely because the company continues trading.
42.
In this case, therefore, the directive should apply, a fortiori if trading is continued. It is not a condition,
for an undertaking to be put into liquidation, that liabilities must exceed assets. Indeed, the conditions
signalling insolvency must not be satisfied, since otherwise liquidation would no longer be permitted.
Liquidation may be a stage preceding insolvency, but it can also occuras the Belgian government points
outif the members no longer wish to work together. In other words, the company does not necessarily
have to be in financial difficulties. The directive should accordingly apply in the case of liquidation if the
undertaking continues to trade.
43.
It does not matter that liquidation may be a stage preliminary to insolvency. In Abels [1985] ECR 469
(para 29), the court held that the directive applied in the case of judicial leave to suspend payment of
debts precisely because it was only such a preliminary stage. The court also held in Foreningen af
Arbejdsledere i Danmark v A/S Danmols Inventar (in liq) Case 105/84 [1985] ECR 2639 (para 10) that the
directive also applied to a transfer which is effected in the course of a procedure, or at a stage, prior to
the commencement of [insolvency] proceedings.
44.
Finally, I would like to refer to a further judgment, in which the court ruled on the applicability of the
directive to the transfer of undertakings declared to be in critical difficulties (see Spano v Fiat Geotech
SpA Case C-472/93 [1995] ECR I-4321). In that case, the court held ([1995] ECR I-4321 (para 28)):
Consequently, an undertaking found to be in critical difficulties is subject to a procedure which,
far from being aimed at the liquidation of the undertaking, is designed on the contrary to promote
the continuation of its business with a view to its subsequent recovery. (My emphasis.)
This case is no different, however. If an undertaking in liquidation continues to trade, this happensas
mentioned abovewith a view to the undertakings subsequent recovery. In this case, therefore, with the
undertaking in liquidation continuing to trade, one can likewise neither explain nor justify the
circumstance that, when all or part of the undertaking is transferred, its employees lose the rights which
the directive confers on them (see Spano [1995] ECR I-4321 (para 30) and dUrso [1991] ECR I-4105
(para 32)).
45.
So, the directive would be applicable to cases of liquidation if the business of the undertaking is
continued in the course of the liquidation.
46.
The same result is arrived at if liquidation procedure is compared with insolvency proceedings.
According to the national court, the special rules that result in the directive not applying in the case of
insolvency do not exist where liquidation is concerned. Thus, it is the general meeting which decides on
liquidation, appoints the liquidators and defines their powers. In insolvency 355 proceedings, on the other
hand, the company can itself file a declaration of insolvency, but it can also be declared insolvent on
application by a creditor or on the basis of the investigation committees work, whereupon the
administrator is appointed by the court and his powers are determined by law.
47.
On insolvency, there is a special procedure supervised by the court for establishing liabilities. That is
not so on liquidation. The liquidator can acknowledge the existence of a liability without having to refer the
matter to anybody else or have his decision confirmed by court judgment.
48.
In insolvency proceedings, a creditor can only have the amount of his claim determined; in liquidation
the position is somewhat different inasmuch as a creditor can obtain judgment against the company in
liquidation. A judgment can also be enforced against a company in liquidation. The liquidator can oppose
enforcement only if it would prejudice the rights of other creditors. In insolvency proceedings, on the other
hand, such enforcement actions are prohibited because the management and liquidation of the assets
destined to satisfy the creditors claims are matters regulated by law.
49.
A liquidator is an organ of the company, whereas an administrator in insolvency proceedings, who
represents the creditors, is a third party vis--vis the company.
50.
 An administrator in insolvency proceedings disposes of the assets under the supervision of the
insolvency judge, while a company liquidator performs that task under the supervision of the general
meeting so that transfer of the undertaking does not need the courts endorsement.
51.
There are therefore many differences compared with insolvency proceedings. In particular, it is clear
that the court has much less influence where liquidation is concerned and that there is no special
procedure such as that applicable on insolvency.
52.
The directive must accordingly be found to apply to the transfer of an undertaking where the
undertaking is in liquidation but the general meeting has resolved that trading is to continue. It does not
matter whether the undertaking is being wound up voluntarily or by the court. The only difference between
the two forms of liquidation is that, in the case of winding up by the court, the general meeting, acting by
the requisite majority, may not appoint the liquidators. In that case, they are appointed by the court.
53.
Whether the directive or its provisions are in any event applicable to liquidations by virtue of national
law, pursuant to Collective Agreement No 32, must be considered separately. As the court held in Abels
[1985] ECR 469 (para 24) the member states are at liberty to apply all or part of the directive
independently, on the basis of their national law alone. The national court must decide whether this is also
the position in the case of liquidation.

The second question

54.
In the first part of the second question, the national court seeks to ascertain whether dismissals by a
liquidator may be regarded as having taken place for economic, technical or organisational reasons within
the meaning of art 4(1) of the directive. In other words, the question is whether the power to dismiss
employees for such reasons belongs only to the transferee, or to the transferor as well.
55.
That uncertainty arises from the wording of art 4(1) of the directive. It prohibits dismissals on the
ground of the transfer alone. That prohibition expressly applies to both the transferor and the transferee.
The second sentence, which permits dismissals for economic, technical or organisational reasons 356
entailing changes in the workforce, does not indicate whether that right is conferred on the transferor, on
the transferee or on both.
56.
Both the Commission and the Belgian government considerin my view rightlythat that possibility
must be open to the transferor as well. The Belgian government relies here on the judgment in P Bork
International A/S (in liq) v Foreningen af Arbejasledere i Danmark Case 101/87 [1988] ECR 3057. In that
case, the court referred to its earlier judgment in Landsorganisationen i Danmark for Tjenerforbundet i
Danmark v Ny Mlle Kro Case 287/86 [1987] ECR 5465 in which it held that the directive could be relied
on solely by employees whose contract of employment or employment relationship was in existence at
the time of the transfer. Whether such a contract or relationship existed at that time had to be assessed
on the basis of national law, subject, however, to compliance with the mandatory provisions of the
directive concerning protection of employees from dismissal as a result of the transfer.
Accordingly, the employees of the undertaking whose contract of employment or employment
relationship was terminated with effect from a date prior to that of the transfer, contrary to Article
4(1) of the directive, must be regarded as still in the employ of the undertaking on the date of the
transfer, with the result, in particular, that the employers obligations towards them are automatically
transferred from the transferor to the transferee in accordance with Article 3(1) of the directive. In
order to determine whether the employees were dismissed solely as a result of the transfer,
contrary to Article 4(1), it is necessary to take into consideration the objective circumstances in
which the dismissal took place and, in particular, in a case such as this, the fact that it took effect
on a date close to that of the transfer and that the employees in question were taken on again by
the transferee. (See Bork [1988] ECR 3057 (para 18).)
57.
The judgment in Bork likewise does not indicate whether the transferor may dismiss employees for
economic, technical or organisational reasons. In my view, such a right follows, however, from this
consideration: according to the judgment in Bork, art 4(1) prohibits the dismissal of employees where the
sole reason is the transfer. It follows from the words solely as a result of the transfer that the transferor
can dismiss employees on other grounds. As a minimum, these should be those specified in the second
sentence of art 4(1) and be available to the transferor as well.
58.
If one considers the liquidation in this case, still othereconomicreasons support that interpretation.
It gives the liquidator the possibility of implementing rationalisation measures prior to the sale. Potential
purchasers will then be more willing to take over the company in liquidation. If the transferor is allowed to
dismiss employees for economic, technical or organisational reasons, the liquidation itself will also be
facilitated, jobs at the undertaking in liquidation will be made safe and worker protection will thereby be
enhanced.
59.
By the second part of the second question, the national court asks whether staff dismissed for
economic, technical or organisational reasons may claim, as against the transferee, that their dismissal
was unlawful for the simple reason that shortly after their dismissal an economic entity was transferred,
even if the transfer agreement does not provide for their re-engagement.
60.
As the Commission rightly points out, this question is wrongly put. If the dismissals were carried out for
economic, technical or organisational reasons, they cannotas already established abovebe unlawful,
so that neither the transferor 357 nor the transferee can be accused of acting unlawfully. The fact that an
economic entity was transferred shortly after their dismissal does not alter that conclusion. If the dismissal
was not carried out solely on account of the transfer, it is not rendered invalid by the transfer.
61.
This last question could, however, be understood and construed differently. It clearly refers to the
phrase in the judgment in Bork, as it contains both the criteria mentioned in the judgment: dismissals
being close in time to the transfer of an economic entity and re-engagement after the transfer of the
business. Those criteria were held in Bork to be evidence that the sole reason for the dismissals was the
transfer and that they were therefore contrary to art 4(1) of the directive. In this case, only one of those
criteria is satisfied: the dismissal effected by the transferor was close in time to the transfer of the
business. There was no re-engagement after the transfer. The question referred by the national court
could, then, be understood as seeking to ascertain whether fulfilment of just one of the criteria specified in
Bork is sufficient for the dismissal to be regarded as resulting solely from the transfer and, therefore, as
being unlawful.
62.
In order to answer that question, reference must again be made to precisely the judgment in Bork.
There the court states that, in order to determine whether a dismissal is unlawful, it is necessary first to
take into account the objective circumstances in which it took place. As examples for the case which
resulted in that reference, it gives the two criteria of closeness in time to the transfer of the undertaking
and re-engagement after the transfer. Since, in this case, the facts of the action before the national court
are different, that court must reach its decision on the basis of other factors. It is true that in this case the
dismissal took effect shortly before the undertaking was transferred. However, another important point is
that the undertaking was in liquidation at the time of the dismissal. That circumstance is indicative of a
dismissal for organisational reasons. The Commission also considers that this makes it easier for the
employer to prove that the dismissal was for economic, technical or organisational reasons.
63.
The answer to the last question must therefore be that in this case a dismissal cannot be alleged to be
unlawful merely because, shortly after it was carried out, the company was transferred: the national court
must take into consideration all objective circumstances relating to the dismissal, in this case, above all,
the fact that the company was in liquidation.
64.
I would add only for the sake of completeness that, should an employee he dismissed unlawfully by
the transferor, that unlawful action can be pleaded as against the transferee in so far as the contract of
employment is regarded as continuing to exist and is thereby transferred to the transferee as a result of
the transfer. Accordingly, the employee can assert his rights under the contract of employment against the
transferee.
65.
It does not matter whether the individual member states impose sanctions in this situation. The Belgian
government pointed out that in Belgian law there is no provision whereby a dismissal is rendered void, so
that the contract of employment cannot be maintained.
66.
The answer to the second question referred by the national court must therefore be that termination of
contracts of employment by the transferor can be regarded as dismissals for economic, technical or
organisational reasons within the meaning of art 4(1) of the directive and that this power does not belong
only to the transferee. Dismissals for economic, technical or organisational reasons are not unlawful
merely because, shortly after they are carried out, an economic entity is transferred. The national court
must take into consideration all the objective 358 circumstances relating to the dismissal, in this case
above all the fact that the compally was in liquidation.

C CONCLUSION
67.
In the light of the foregoing, I propose that the answer to the questions submitted by the national court
should be:
(1) Council Directive (EEC) 77/187 on the approximation of the laws of the member states
relating to the safeguarding of employees rights in the event of transfers of undertakings,
businesses or parts of businesses applies to a transfer effected by a company in voluntary
liquidation, but only on condition that the undertaking continues to trade in that context. This applies
irrespective of whether the liquidator is appointed by the court or by the general meeting.
(2) The liquidator, too, has the right to dismiss employees for economic, technical or
organisational reasons. Staff not taken on by the transferee may not claim, as against him, that
their dismissal was unlawful merely because shortly after their dismissal the business was
transferred as an economic entity. When assessing the legality of dismissals on the basis of the
criteria laid down by the Court of Justice, the national court must, however, take into consideration
 all the objective circumstances relating to the dismissals, in this case above all the fact that the
company was in liquidation.

12 March 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By judgment of 1 December 1994, received at the Court of Justice of the European Communities on 8
December 1994, the Cour du Travail (the Higher Labour Court), Lige, referred to the court for a
preliminary ruling under art 177 of the EC Treaty a number of questions on the interpretation of Council
Directive (EEC) 77/187 on the approximation of the laws of the member states relating to the
safeguarding of employees rights in the event of transfers of undertakings, businesses or parts of
businesses (OJ 1977 L61 p 26).
2.
Those questions have been raised in proceedings between Jules Dethier quipement SA (Dethier), on
the one hand, and Jules Dassy and Sovam SPRL (Sovam), on the other, concerning payment of
compensation in lieu of notice and of other sums of compensation.
3.
According to art 1(1) of the directive, the directive is to apply to the transfer of an undertaking,
business or part of a business to another employer as a result of a legal transfer or merger.
4.
Under the first sub-paragraph of art 3(1), the transferors rights and obligations arising from a contract
of employment or from an employment relationship existing on the date of a transfer are, by reason of
such transfer, to be transferred to the transferee.
5.
Article 4(1) of the directive provides that the transfer of an undertaking, business or part of a business
is not in itself to constitute grounds for dismissal by the transferor or the transferee. That provision is not,
however, to stand in the way of dismissals that may take place for economic, technical or organisational
reasons entailing changes in the workforce.
6.
The Directive was transposed into Belgian law in the second chapter of Collective Agreement No 32
bis of 7 June 1985 concerning the safeguarding of employees rights in the event of a change of employer
as a result of the 359 contractual transfer of an undertaking and regulating the rights of employees re-
engaged in the event of a takeover of assets following insolvency or judicial composition with transfer of
assets, made mandatory by Royal Decree of 25 July 1985 (Moniteur Belge, 9 August 1985, p 11527).
That agreement was amended by Collective Agreement No 32 quater of 19 December 1989, made
mandatory by Royal Decree of 6 March 1990.
7.
The procedure for winding up companies under Belgian law is dealt with in arts 178 to 188 of the Lois
Coordonnes sur les Socits Commerciales (the Consolidated Laws on Commercial Companies) (the
Consolidated Laws). It applies following the dissolution of a commercial company and is designed to allow
the company to conclude business transactions already entered into while precluding it, as a rule, from
engaging in new business. Article 178 of the Consolidated Laws provides that, following their dissolution,
commercial companies are deemed to exist for the purposes of their liquidation even if they have already
ceased all trading.
8.
Dissolution automatically brings to an end the functions of all the organs of the company, which are
replaced by one or more liquidators. The liquidator, who represents the company vis--vis third parties,
may already be named in the companys articles of association. In the absence of contrary agreement,
liquidators are appointed by the members in general meeting or, if the majority laid down by law is not
obtained, by the court: in the first case the liquidation is called voluntary winding up (liquidation
volontaire), while in the second it is referred to as winding up by the court (liquidation judiciaire).
9.
Although the national court states in its judgment making the reference that the objectives pursued in a
winding up, whether voluntary or by the court, are very close to those in an insolvency, that is to say
realisation of the companys assets, it draws attention to a number of points where the two procedures
differ, which may be summarised as follows: the decision to wind up the company, the appointment of
liquidators and the definition of their powers are matters for the general meeting of the company. It is only
where the requisite majority of the members is not achieved that the company must apply to the court for
a winding up order; the court appoints the liquidators in accordance with the articles of association of the
company, or in accordance with the resolution of the general meeting, except where it appears certain
that disagreement amongst the members will prevent the general meeting from passing a resolution, in
which case a liquidator is appointed by the court. As regards insolvency, the company may declare itself
insolvent, but it may also be declared insolvent following an action by a creditor or work by the
investigation commission, the court appointing the administrator whose powers are laid down by law; the
legal personality of the company survives for the purposes of the liquidation (art 178 of the Consolidated
Laws), which is not the case with a company in insolvency; the company retains its commercial character
throughout the winding up; thus, if it ceases subsequently to make payments or is unable to raise credit it
could be declared insolvent; in that case winding up is a procedure which precedes insolvency; while
there is a special insolvency procedure for establishing liabilities under the supervision of the court, that is
not so in the case of a winding up, whether voluntary or ordered by the court, where the liquidator may
recognise a debt without referring to anybody else; responsibility for that decision, which does not require
sanctioning by court order, rests with the liquidator; while, in an insolvency, creditors may only register
their debts as liabilities of the company, the position is different in a winding up where creditors may
obtain judgment against 360 it; in the case of a winding up, creditors may levy execution against the
company and the liquidator can oppose that step only if its effect is to prejudice the rights of the other
creditors, while, in an insolvency, such levying of execution is prohibited since the management and
liquidation of the assets to be used to pay off creditors are governed by law; the general meeting may
revoke its appointment of a liquidator, whereas only the court may revoke the appointment of a court-
appointed liquidator or of an administrator in an insolvency; from that point of view a distinction must thus
be drawn between voluntary winding up, on the one hand, and winding up by the court and insolvency
proceedings, on the other; a liquidator is the organ of the company during its winding up whereas an
administrator is a third party. An administrator represents the creditors as well as the company, while a
liquidator represents only the company even though he must ensure that the interests of creditors are
safeguarded; unlike an administrator in an insolvency, a liquidator cannot challenge certain payments in
the absence of a suspect period before the beginning of the winding up, nor can he bring an action to
make good a deficiency in assets or to establish liability on the part of the founders; an administrator sells
the assets under the supervision of the insolvency judge and with the authorisation of the court in certain
cases, while a liquidator performs that task under the supervision of the general meeting, so that the
transfer of the undertaking is not subject to court approval; in conclusion, insolvency proceedings afford
the creditors greater guarantees than liquidation proceedings in that the administrator is supervised by the
court and the creditors are more directly represented.
10.
Mr Dassy had been employed by Sovam since 1974. On 15 May 1991 the Tribunal de Commerce (the
Commercial Court), Huy, made an order putting that company into liquidation and appointing a liquidator.
On 5 June 1991 the liquidator dismissed Mr Dassy.
11.
On 27 June 1991 the liquidator transferred the assets of Sovam to Dethier under an agreement which
the Tribunal de Commerce approved on 10 July 1991.
12.
On 22 May 1992 Mr Dassy brought an action before the Tribunal du Travail (the Labour Court), Huy,
for Sovam and Dethier to be held jointly and severally liable to pay the sums due to him by way of
compensation in lieu of notice, paid holiday leave and end-of-year bonuses. He contended that the
transfer of the undertaking was contractual and that Dethier was therefore jointly and severally liable for
those payments.
13.
On 17 December 1993 the Tribunal du Travail, Huy, gave judgment against Sovam, in liquidation, and
Dethier, finding them jointly and severally liable to pay Mr Dassy the sum of BF 1,643,726. It held in
particular that the second chapter of Collective Agreement No 32 bis should be applied, since winding up
by the court could not be treated in the same way as insolvency where, as in this case, the transfer
resulted from a takeover plan which predated the winding up. It also considered that liability had to be
joint and several even where a contract was prematurely terminated before the date of the transfer
because, according to the case law of the Court of Justice, staff unlawfully dismissed before a transfer
had to be regarded as still employed by the undertaking on the date of the transfer.
14.
Dethier appealed against that judgment to the Cour du Travail, Lige, contending that the liquidation of
Sovam was comparable to an insolvency. Furthermore, any joint and several liability on the part of the
transferee could benefit only those employees taken on by the transferee and not those dismissed before
the transfer.
361
15.
After finding that in this case the entity formed by the assets of Sovam had retained its identity after its
being taken over, in accordance with the judgment in Spijkers v Gebr Benedik Abattoir CV Case 24/85
[1986] ECR 1119, and that there had therefore been a transfer of an undertaking within the meaning of
the directive, the national court raised the question whether that transfer could be regarded as a legal
transfer within the meaning thereof and, in particular, whether the winding up of a company amounted to a
procedure analogous to insolvency proceedings, which would then fall outside the field of application of
the directive.
16.
As regards the dismissal of Mr Dassy, the national court is unsure as to the interpretation of art 4 of the
directive in a case where an employee is dismissed by the liquidator shortly before the transfer and not
subsequently taken on by the transferee.
17.
The Cour du Travail, Lige, therefore decided to stay proceedings and to refer the following questions
to the Court of Justice:
(1) Does Council Directive 77/187 apply where the transfer is effected by a company in
voluntary liquidation, a procedure whose aim, in the absence of continued trading, is liquidation by
 realisation of the assets? Is the answer the same where the transferor is being wound up by the
court?
(2) Where the contracts of employment of all the employees have been terminated by the
liquidator and only some of those employees have been re-engaged for the purposes of the
liquidation, may the dismissals of the employees not subsequently taken on by the transferee be
regarded as having taken place for economic, technical or organizational reasons within the
meaning of Article 4(1) of the directive? Must the power to dismiss such employees for such
reasons be left, on the contrary, to the transferee alone?
May staff not taken on by the transferee claim, as against him, merely because an economic
entity was transferred shortly after their dismissal for economic, technical or organizational reasons,
that the measure taken in their regard by the transferor was unlawful if the transfer agreement does
not provide for their re-engagement?

The first question

18.
By its first question, the national court essentially seeks to ascertain whether, on a proper construction
of art 1(1) of the directive, the directive applies in the event of the transfer of an undertaking which is
being wound up voluntarily or by the court.
19.
It should be noted first of all that the court has consistently held that it cannot give preliminary rulings
on questions which bear no relation to the facts and the subject matter of the main action and are
therefore not strictly needed in order to decide the case (see esp Salonia v Poidomani Case 126/80
[1981] ECR 1563 (para 6), Loureno Dias v Director da Alfndega do Porto Case C-343/90 [1992] ECR I-
4673 (para 20), Corsica Ferries Italia Srl v Corpo dei Piloti del Porto di Genova Case C-18/93 [1994] ECR
I-1783 (para 14) and Centro Servizi Spediporto Srl v Spedizioni Marittima del Golfo Srl Case C-96/94
[1995] ECR I-2883 (para 45)).
20.
Since the main proceedings concern the transfer of an undertaking being wound up by the court, the
court is not required to interpret the directive with regard to the transfer of an undertaking being wound up
voluntarily, a situation which is unrelated to the subject matter of the main proceedings, whatever the
similarities between the procedure governing winding up by the court and that governing voluntary
winding up.
362
21.
Next, as the court ruled in Abels v Administrative Board of the Bedrijfsvereniging voor de
Metaalindustrie en de Electrotechnische Industrie Case 135/83 [1985] ECR 469, the directive does not
apply to the transfer of an undertaking, business or part of a business in the course of insolvency
proceedings.
22.
However, it is clear from the judgment in Abels [1985] ECR 469 (paras 2829) that the directive is
applicable to proceedings such as a sursance van betaling (judicial leave to suspend payment of
debts), even though they have certain features in common with insolvency proceedings. The court in fact
considered that the reasons for not having the directive apply in the case of insolvency proceedings were
not valid when the proceedings in question involved supervision by the court which was more limited than
in the case of insolvency and when they were intended primarily to safeguard the assets of the
undertaking and, if possible, to keep the undertaking in business by means of a collective suspension of
debt payment with a view to reaching a settlement allowing the undertaking to continue operating in the
future.
23.
Similarly, the court held in dUrso v Ercole Marelli Elettromeccanica Generale SpA Case C-362/89
[1991] ECR I-4105 (paras 28, 31, 34) that the directive did not apply to transfers of undertakings,
businesses or parts of a business made as part of a creditors arrangement procedure of the kind
provided for by the Italian legislation on compulsory administrative liquidation, whose effects were
comparable to those of bankruptcy proceedings. Conversely, it did apply when it was decided, under a
body of legislation such as that governing special administration for large undertakings in critical
difficulties, that the undertaking was to continue trading for as long as that decision remained in force. In
such cases, the primary purpose of the special administration procedure was to give the undertaking
some stability allowing its future activity to be safeguarded. The social and economic objectives thus
pursued could not explain or justify the circumstance that, when all or part of the undertaking was
transferred, its employees lost the rights which the directive conferred on them under the conditions which
it laid down (see [1991] ECR I-4105 (paras 29, 3234)).
24.
More recently, the court held in Spano v Fiat Geotech SpA Case C-472/93 [1995] ECR I-4321 that the
directive applied to the transfer of an undertaking declared to be in critical difficulties pursuant to Italian
Law No 675 of 12 August 1977. It pointed out in particular that the purpose of a declaration that an
undertaking was in critical difficulties was to enable the undertaking to retrieve its economic and financial
situation and above all to preserve jobs, that the procedure in question was designed to promote the
continuation of its business with a view to its subsequent recovery and that, by contrast with insolvency
proceedings, it did not involve any judicial supervision or any measure whereby the assets of the
undertaking were put under administration and did not provide for any suspension of payments (see
[1995] ECR I-4321 (paras 26, 2829)).
25.
It follows from that case law that, in deciding whether the directive applies to the transfer of an
undertaking subject to an administrative or judicial procedure, the determining factor to be taken into
consideration is the purpose of the procedure in question (see the judgments in dUrso [1991] ECR I-4105
(para 26) and Spano [1995] ECR I-4321 (para 24)). However, as the Advocate General has stated in his
opinion, account should also be taken of the form of the procedure in question, in particular in so far as it
means that the undertaking continues or ceases trading, and also of the directives objectives (see paras
31, 41, 45).
26.
In this case, it is apparent from the information provided by the Cour du Travail that the objective of the
Belgian procedure under which a company is 363 wound up by the court is liquidation by realising the
companys assets for the benefit of the company itself and, subsidiarily, of any creditors. It is not a
condition for putting a company into liquidation that its liabilities must exceed its assets. Although
liquidation may be a stage which precedes insolvency, it can also occur, as the Belgian government
states, when the members no longer wish to cooperate.
27.
It follows that, although the objectives of a winding up by the court may sometimes be similar to those
of insolvency proceedings, this is not necessarily the case, since liquidation proceedings may be used
whenever it is wished to bring a companys activities to an end and whatever the reasons for that course.
28.
Since the criterion relating to the purpose of the procedure for winding up by the court appears not to
be conclusive, it is necessary to consider that procedure in detail.
29.
According to the reference by the national court, in the case of a liquidation the liquidator, although
appointed by the court, is an organ of the company who sells the assets under the supervision of the
general meeting; there is no special procedure for establishing liabilities under the supervision of the
court; and a creditor may as a rule enforce his debt against the company and obtain judgment against it.
By contrast, in the case of an insolvency, the administrator, inasmuch as he represents the creditors, is a
third party vis--vis the company and realises the assets under the supervision of the court; the liabilities
of the company are established in accordance with a special procedure and individual enforcement
actions are prohibited.
30.
It is thus apparent that the situation of an undertaking being wound up by the court presents
considerable differences from that of an undertaking subject to insolvency proceedings and that the
reasons which have led the court to rule out application of the directive in the latter situation may be
absent in the case of an undertaking being wound up by the court.
31.
That is the case where, as in the main proceedings, the undertaking continues to trade while it is being
wound up by the court. In such circumstances continuity of the business is assured when the undertaking
is transferred. There is accordingly no justification for depriving the employees of the rights which the
directive guarantees them on the conditions it lays down.
32.
The answer to the first question submitted for a preliminary ruling must therefore be that, on a proper
construction of art 1(1) of the directive, the directive applies in the event of the transfer of an undertaking
which is being wound up by the court if the undertaking continues to trade.

The second question

33.
By the first part of the second question, the national court essentially asks whether, on a proper
construction of art 4(1) of the directive, only the transferee may dismiss employees for economic,
technical or organisational reasons or whether the transferor must also be accorded that power.
34.
Article 4(1) protects the rights of employees against a dismissal whose sole justification is the transfer,
both vis--vis the transferor as well as vis--vis the transferee.
35.
The court has held that employees whose contract of employment or employment relationship comes
to an end with effect from a date prior to that of the transfer, contrary to art 4(1), must be regarded as still
employed by the undertaking on the date of the transfer, with the result, in particular, that the employers
obligations towards them are automatically transferred from the 364 transferor to the transferee (see P
Bork International A/S (in liq) v Foreningen af Arbejasledere i Danmark Case 101/87 [1988] ECR 3057
(para 18)).
36.
Accordingly, inasmuch as art 4(1) precludes dismissals from taking place solely by reason of the
transfer, it does not restrict the power of the transferor any more than that of the transferee to effect
dismissals for the reasons which it allows.
37.
 The answer to the first part of the second question referred for a preliminary ruling must therefore be
that, on a proper construction of art 4(1) of the directive, both the transferor and the transferee may
dismiss employees for economic, technical or organisational reasons.
38.
By the second part of the second question, the national court essentially seeks to ascertain whether
employees unlawfully dismissed by the transferor shortly before the undertaking is transferred and not
taken on by the transferee may claim, as against the transferee, that their dismissal was unlawful.
39.
First, it follows from the judgment in Bork that employees dismissed before the undertaking was
transferred, contrary to art 4(1), must be regarded as still employed by the undertaking on the date of the
transfer.
40.
Secondly, it is settled case law that the rules of the directive, in particular those concerning the
protection of workers against dismissal by reason of the transfer, must be considered to be mandatory, so
that it is not possible to derogate from them in a manner unfavourable to employees (see esp Foreningen
af Arbejdsledere i Danmark v Daddys Dance Hall A/S Case 324/86 [1988] ECR 739 (para 14)).
41.
Therefore, the contract of employment of a person unlawfully dismissed shortly before the transfer
must be regarded as still extant as against the transferee even if the dismissed employee was not taken
on by him after the undertaking was transferred.
42.
For those reasons, the answer to the second part of the second question referred for a preliminary
ruling must be that employees unlawfully dismissed by the transferor shortly before the undertaking is
transferred and not taken on by the transferee may claim, as against the transferee, that their dismissal
was unlawful.

Costs
43.
The costs incurred by the Belgian government and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the action pending before the national court, the decision on
costs is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the questions referred to it by
the Cour du Travail, Lige, by judgment of 1 December 1994, hereby rules: (1) On a proper construction
of art 1(1) of Council Directive (EEC) 77/187 on the approximation of the laws of the member states
relating to the safeguarding of employees rights in the event of transfers of undertakings, businesses or
parts of businesses, that directive applies in the event of the transfer of an undertaking which is being
wound up by the court if the undertaking continues to trade. (2) On a proper construction of art 4(1) of
Directive 77/187, both the transferor and the transferee may dismiss employees for economic, technical
or organisational reasons. Employees unlawfully dismissed by the transferor shortly before the
undertaking is transferred and not taken on by the transferee may claim, as against the transferee, that
their dismissal was unlawful.

365

[1998] All ER (EC) 366

lurlaro v Istituto Nazionale della Previdenza Sociale (INPS)

(Case C-322/95)
EUROPEAN COMMUNITY; Free movement of workers: SOCIAL SECURITY
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES MOITINHO DE ALMEIDA (RAPPORTEUR AND PRESIDENT OF THE CHAMBER), SEVN,
EDWARD, JANN AND WATHELET, ADVOCATE GENERAL LGER
22 JANUARY, 13 MARCH, 17 SEPTEMBER 1997
European Community Workers Freedom of movement Social security Invalidity benefits
Entitlement to benefit Reference period during which contributions to be paid Extension of reference
period Effect of periods in another member state Whether periods of unemployment spent in one
member state extending reference period in another Council Regulation (EEC) 1408/71, art 9a
Council Regulation (EEC) 574/72, art 15 EC treaty, arts 4851.

The applicant, an Italian national, worked and paid social security contributions in Germany from 1959 to
1984. In 1984 he began receiving unemployment benefit and in 1989 he began to receive sickness
benefit. Finally, suffering from an infirmity which reduced his working capacity by more than two-thirds, he
applied for an invalidity allowance in Italy. His application was rejected on the ground that he did not
qualify under Italian legislation since, during the five-year period preceding his application for benefit, he
had drawn unemployment benefits but had not paid any social security contributions. The applicant
brought an action before the Pretura Circondariale di Roma in its capacity as an employment tribunal for a
declaration that the necessary condition for recognition of his entitlement to invalidity allowance had been
fulfilled. He contended, inter alia, that by virtue of art 9a 1 of Council Regulation (EEC) 1408/71 on the
application of social security schemes to employed persons, to self-employed persons and to members of
their families moving within the Community and art 15(1)(f) 2 of Council Regulation (EEC) 574/72 laying
down the procedure for implementing Regulation 1408/71, the tribunal was obliged to take into account
his earlier contributions, since the relevant German legislation, which also stipulated a five-year reference
period in which a minimum contribution was to be made, allowed for the extension of that period by any
amount of time spent within that period, receiving unemployment benefit. The Pretore di Roma stayed the
proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling the
question whether, in the light of arts 48 to 51 (free movement of workers) of the EC Treaty, art 9a of
Regulation 1408/71 and art 15(1)(f) of Regulation 574/72 required a member state whose legislation did
not allow for the prolongation of the reference period for the recognition of entitlement to invalidity
pension, to prolong that period where a worker had received unemployment benefit in another member
state in which such extension was provided for.
1
Article 9a, so far as material, is set out at p 380 j to p 381 a, post
2
Article 15(1)(f), so far as material, is set out at p 380 f g, post

366

Held Article 9a of Regulation 1408/71 was concerned solely with a situation where the legislation of a
member state which made entitlement to a benefit conditional upon completion of a minimum period of
insurance during a reference period itself allowed for the extension of that period by periods during which
certain benefits, and in particular unemployment benefits, were granted under the legislation of another
member state. Moreover, art 15(1)(f)(ii) of Regulation 574/72 only required a time limit to be extended for
the duration of insurance periods completed within that period under the legislation of another member
state where those periods gave rise, under the legislation of that state, only to the suspension of the time
limit. In the instant case, the Italian legislation did not provide for the extension of the reference period
and, under the German legislation, periods of unemployment did not give rise solely to the suspension of
the time limit but were taken into account for the purposes of calculating the amount of the invalidity
benefit. Accordingly, since art 51 of the Treaty and Regulation 1408/71 did not regulate the national
conditions under which insurance periods were constituted and the Italian legislation employed objective
criteria and applied without distinction to national workers and those of other member states, it followed
that arts 48 to 51 of the Treaty, art 9a of Regulation 1408/71 and art 15(1)(f)(ii) of Regulation 574/72 did
not require a member state to extend the reference period laid down by its legislation for determining the
minimum insurance requirement for the grant of an invalidity benefit, by a period equivalent to periods of
unemployment spent by the person concerned under the legislation of another member state which,
unlike that of the first member state, allowed such an extension where the periods of unemployment were
spent on national territory (see p 381 h, p 382 c to p 383 j and p 384 f g h, post).
Bestuur van de Nieuwe Algemene Bedrijfsvereniging v Drake Case C-12/93 [1994] ECR I-4337
applied.

Notes
For social security measures necessary for free movement in general and unemployment benefits in
particular, see 52 Halsburys Laws (4th edn) paras 2131, 2186-2189.
For the EC treaty, arts 4851, see 50 Halsburys Statutes (4th edn) 283, 284.

Cases cited
Baccini v Office National de lEmploi (ONEM) Case 232/82 [1983] ECR 583.
Bestuur van de Nieuwe Algemene Bedrijfsvereniging v V A Drake Case C-12/93 [1994] ECR I-4337.
Caisse auxiliaire dassurance maladie-invalidit (CAAMI) v Faux Case C-302/90 [1991] ECR I-4875.
Caisse de Compensation pour Allocations Familiales des Rgions de Charleroi et Namur v Laterza Case
733/79 [1980] ECR 1915.
de Jaeck v Staatssecretaris van Financin Case C-340/94 [1997] ECR I-461.
Gravina v Landesversicherungsanstalt Schwaben Case 807/79 [1980] ECR 2205.
Istituto Nazionale della Previdenza Sociale (INPS) v Baglieri Case C-297/92 [1993] ECR I-5211.
Labots (ne Hagenbeek) v Raad van Arbeid, Arnhem Case 100/63 [1966] ECR 425.
Lepore v Office National des Pensions (ONP) Joined cases C-4546/92 [1993] ECR I-6497.
Martnez Losada v Instituto Nacional de Empleo (INEM) Joined cases C-88/95, C-102103/95 [1997] All
ER (EC) 240, [1997] ECR I-869, ECJ.
367
Paraschi v Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501.
Rijksdienst voor Werknemerspensioenen v Celestre, Nationaal Pensioenenfonds voor Mijnwerkers Joined
cases 116117/80 and 119121/80 [1981] ECR 1737.
Rossi v Caisse de Compensation pour Allocations Familiales des Rgions de Charleroi et Namur Case
100/78 [1979] ECR 831.
Vella v Alliance nationale des mutualits chrtiennes Case C-324/88 [1990] ECR I-257.
Reference
By order of 3 October 1995, the Pretura Circondariale di Roma (the Rome District Magistrates Court)
referred to the court for a preliminary ruling under art 177 of the EC Treaty a question (set out at p 381 c,
post) on the interpretation of art 9a of Council Regulation (EEC) 1408/71 on the application of social
security schemes to employed persons, to self-employed persons and to members of their families
moving within the Community, as amended and updated by Council Regulation (EEC) 2001/83 and
subsequently amended by Council Regulation (EEC) 2332/89 and of art 15 of Council Regulation (EEC)
574/72 laying down the procedure for implementing Regulation 1408/71, as amended and updated. The
question arose in proceedings between Mr Iurlaro, an Italian national, and the Istituto Nazionale della
Previdenza Sociale (the INPS) concerning the latters refusal to take into account, for the purpose of the
acquisition of entitlement to an invalidity allowance under Italian law, periods of unemployment spent by
Mr Iurlaro in Germany. Written observations were submitted on behalf of: Mr Iurlaro, by R Maffei and P
Boer, both of the Rome Bar; the INPS, by M Pittelli, of the Rome Bar, and P Ciacci, of the Milan Bar, the
Italian government, by U Leanza, Head of Department of Contentious Diplomatic Affairs at the Ministry of
Foreign Affairs, acting as agent, assisted by D Del Gaizo, Avvocato dello Stato; and the European
Commission, by A Aresu and M Patakia, both of its Legal Service, acting as agents. Oral observations
were made by: Mr Iurlaro, represented by P Boer, the INPS, represented by M Pittelli and P Ciacci; and
the Commission, represented by M Patakia and P Stancanelli, of its Legal Service. The language of the
case was Italian. The facts are set out in the opinion of the Advocate General.

13 March 1997.

The Advocate General (Lger)

delivered the following opinion (translated from the French).
1.
The Pretura Cirondariale di Roma, in its capacity as an employment tribunal, has requested the Court
of Justice of the European Communities to interpret a number of Community provisions concerning the
application of social security schemes to migrant workers3, in the light of art 48 of the EC Treaty. It asks in
particular whether periods of unemployment completed in one member state (Germany) should be taken
into account when calculating the reference period for the purpose of acquisition of entitlement to
invalidity benefits in another member state (Italy).
3
Article 9a of Council Regulation (EEC) 1408/71 on the application of social security schemes to employed persons, to
self-employed persons and to members of their families moving within the Community (OJ S Edn 1971(11) p 416), as
introduced with retrospective effect by Council Regulation (EEC) 2332/89 (OJ 1989 L224 p 1); art 15(1)(f) of Council
Regulation (EEC) 574/72 laying down the procedure for implementing Regulation 1408/71 (OJ S Edn 1972 (I) p 160).

368
2.
I will suggest that the court should rule against such periods being taken into account in the case
submitted to it but, before doing so, I will first define the legal and factual context of the case.

RELEVANT NATIONAL PROVISIONS ITALIAN LEGISLATION

3.
Under Italian law the acquisition of entitlement to invalidity allowance presupposes, in addition to
acknowledgement of the invalidity, that the following cumulative insurance and contribution conditions are
satisfied4: at least five years must have elapsed since the person concerned joined the insurance
scheme; that person must have paid, or have had credited in his favour, at least 260 weekly contributions
(five years of pensionable service); finally, at least 156 weekly contributions (three years of pensionable
service) must have been paid or credited during a reference period of five years preceding his application
for benefit.
4
See art 4 of Law No 222 of 12 June 1984 concerning revision of the rules on invalidity pension (GURI No 165 of 16
June 1984), which refers to the conditions laid down by art 9(2) of Royal Decree-Law No 636 of 14 April 1939, which
became Law No 1272 of 6 July 1939 (which was replaced by art 2 of Law No 218 of 4 April 1952 (GURI No 89 of 15
April 1952, ordinary supplement), as amended by Law No 222, cited above.

4.
As regards the taking into account of periods of unemployment, art 4 of Law No 218, provides that
periods in respect of which the ordinary allowance in respect of compulsory unemployment
insurance is paid shall be regarded as contribution periods for the purposes of entitlement to
pension and of the amount of pension.

German legislation
5.
Under German legislation5 the person insured is entitled to an invalidity pension if he is unable to carry
on his occupation and: he has paid three years compulsory contributions during the five years last
preceding the onset of the invalidity, and he completed the general contribution period before the onset of
the invalidity.
 5
Sozialgesetzbuch (SGB), Book VI [BGBl III 860, as amended by the Gesetz zur Reform der gesetzlichen
Rentenversicherung (the Law on the Reform of Statutory Old Age Insurance) of 28 December 1989 (BGBl IS 2261)].

6.
The reference period of five years prior to the onset of the invalidity is extended by notional
contribution periods6, although the period during which the worker was registered on unemployment lists
in Germany and drew benefit from bodies governed by public law, whilst giving rise to a notional
contribution for calculating the pension7, merely allows extension of the reference period for the purposes
of the minimum contribution required for recognition of entitlement.
6
Paragraph 43, SGB.
7
Paragraph 58, SGB.

Distinction between the two schemes

7.
Thus, the way in which account is taken of periods during which the person concerned received
unemployment benefits differs widely between the one scheme and the other.
8.
Under the Italian scheme those periods are regarded as contribution periods for the purposes of
entitlement to social security benefits.
369
9.
Under the German scheme the periods in question constitute grounds for extending the reference
period for calculating the minimum insurance requirement.

FACTUAL AND PROCEDURAL CONTEXT

10.
Mr lurlaro, who is of Italian nationality, was initially insured in Italy from 1954 to 1956. He then lived
and worked without interruption in Germany, and, as from May 1983, went through a period of
unemployment in respect of which benefit was paid by the German social security system and which
continued until 31 December 19918.
8
Save for a period between 15 August 1984 and 1 October 1984, during which Mr lurlaro has said, in reply to a question
by the court, that he drew sickness benefits.

11.
On 18 October 1989, suffering from an infirmity which reduced his working capacity by more than two-
thirds, Mr lurlaro applied to the Istituto Nazionale della Previdenza Sociale (the INPS) for the Italian
invalidity pension9.
9
The INPS states that Mr lurlaro unsuccessfully made a similar claim before the competent institution in Germany, and
that the proceedings before the German courts in respect of that refusal are still pending (p 2, para 1 of the French
translation of its observations).

12.
That application was refused for failure to comply with one of the conditions laid down by the Italian
legislation concerning contributions. The INPS took the view that no contribution had been paid during the
five years preceding the application, that reference period extending from 18 October 1984 to 18 October
1989.
13.
The plaintiff disputes that assessment. He relies on the German provisions in support of his claim that
the period of unemployment in Germany in respect of which benefit was paid should have been taken into
account, thereby neutralising that period and causing the reference period used to determine the
existence of the minimum insurance condition to commence at an earlier date.
14.
Relying on arts 15(1)(f) of Regulation 574/72 and 9a of Regulation 1408/71, cited above, Mr lurlaro
brought an action before the Pretore di Roma for a declaration that the necessary condition for
recognition of his entitlement to invalidity allowance had been fulfilled.
15.
The national court considers that the scope of each of the Community provisions cited needs to be
clarified, since art 9a of Regulation 1408/71 seems to imply a condition that the legislation of both
countries must provide for the neutralisation of specified periods, excluding a situation such as that in the
present case, where that possibility is provided for in only one of the two member states.
16.
It therefore requests the Court of Justice to interpret
Article 15(1)(f) of Regulation 574/72 and Article 9a of Regulation 2332/89, in the light of Article
48 of the Treaty establishing the EEC, in order to ascertain whether Article 4 of Law No 222/1984
should be so applied as to extend the reference period for the recognition of entitlement to invalidity
pension where a worker has been in receipt of unemployment benefit in another member state (in
 this case Germany) in which such extension is provided for and, if so, whether such extension
should be subject to conditions.
17.
Before considering that question, which I will approach from two angles, I think it will be useful to give
a brief account of the Community provisions on social security for workers moving within the Community,
laid down by 370 Regulation 1408/71, and Regulation 574/72 which lays down the detailed rules for its
application.

Relevant Community legislation

18.
Within the Community, there are social security schemes in existence which not only differ sharply
from one member state to another, but which the Community rules that have been introduced make no
claim to supplant.
19.
Each member state retains its own competence in the field of social security, in accordance with arts
117ff of the Treaty, and in particular remains free to determine the conditions for affiliation to the various
social security schemes. The Court of Justice has, moreover, consistently reaffirmed this basic principle:
The Court has consistently held that the member states remain competent to define the conditions for
granting social security benefits (see Martnez Losada v Instituto Nacional de Empleo (INEM) Joined
cases C-88/95, C-102103/95 [1997] All ER (EC) 240, [1997] ECR I-869 (para 43)) 10.
10
See also Bestuur van de Nieuwe Algemene Bedrijfsvereniging v V A Drake Case C-12/93 [1994] ECR I-4337 (para 27).

20.
Nevertheless, divergences arising from the diversity of the schemes in force can constitute an obstacle
to the principle of freedom of movement for workers. Workers might hesitate to exercise that fundamental
right if they were not assured that it would not entail harmful consequences for them as regards their
social security cover. That is why art 51 of the EC Treaty provides for the adoption of such measures in
the field of social security as are necessary to provide freedom of movement for [migrant] workers (see
de Jaeck v Staatssecretaris van Financin Case C-340/94 [1997] ECR I-461 (para 18); my emphasis) 11.
11
See also the consistent case law of the court, for example in Istituto Nazionale della Previdenza Sociale (INPS) v
Baglieri Case C-297/92 [1993] ECR I-5211 (para 17) and Lepore v Office National des Pensions (ONP) Joined cases C-
4546/92 [1993] ECR I-6497 (para 34)).

21.
It was thus out of a concern to eliminate the obstacles and to encourage free movement that
Regulation 1408/71, and then the further legislation laying down the detailed rules for its application, were
adopted in accordance with the objectives set out in art 51 However, in view of the competence, referred
to above, retained by the member states, those provisions merely aim to provide for the coordination, not
the harmonization, of the legislations of the Member States.
22.
The Court of Justice has therefore incessantly pointed out that
the regulations did not set up a common scheme of social security, but allowed different
schemes to exist, creating different claims on different institutions against which the claimant
possesses direct rights by virtue either of national law alone or of national law supplemented,
where necessary, by Community law. (See Rossi v Caisse de Compensation pour Allocations
Familiales des Rgions de Charleroi et Namur Case 100/78 [1979] ECR 831 (para 13).)12

12
Confirmed eg by Caisse de Compensation pour Allocations Familiales des Rgions de Charleroi et Namur v Laterza
Case 733/79 [1980] ECR 1915 (para 8), Gravina v Landesversicherungsanstalt Schwaben Case 807/79 [1980] ECR
2205 (para 7) and Baccini v Office National de lEmploi (ONEM) Case 232/82 [1983] ECR 583 (para 17).

23.
That coordination, which is thus aimed at eliminating the negative effects which national legislation
might produce when a worker crosses a frontier, is centred on four basic principles. Those principles, as
expressed in general terms below, are also reaffirmed for nearly all the risks governed by Regulation
1408/71: the principle that the legislation of only one member state shall be applicable 13; the 371 principle
of equality of treatment between nationals and non-nationals 14; the principle of the conservation of
acquired rights15; the principle of the retention of rights in the course of acquisition, also referred to as the
principle of the aggregation of insurance periods16.
13
Article 13(1) of the regulation provides: persons to whom this Regulation applies shall be subject to the legislation of
a single member state only.
14
Article 3(1) of the regulation provides: persons resident in the territory of one of the Member States to whom this
Regulation applies shall be subject to the same obligations and enjoy the same benefits under the legislation of any
Member State as the nationals of that state.
15
Article 51(b) of the Treaty provides: The Council shall adopt such measures in the field of social security as are
necessary to secure for migrant workers payment of benefits to persons resident in the territories of the Member
States.
16
I shall return to this principle in the course of my further argument.

24.
 Those rules, and the principles underlying them, are applicable in this case.
25.
In the first place, the benefit applied for falls within the scope ratione materiae of Regulation 1408/71,
since, in the words of art 4(1) of that regulation, This Regulation shall apply to all legislation concerning
the following branches of social security (b) invalidity benefits
In that respect, amongst the provisions specific to that category of benefits, Ch 2 of Title III of
Regulation 1408/71 envisages two groups of cases for which different rules are laid down. The regulation
takes account of the fact that, where invalidity is concerned, two types of legislation exist in the member
states: those under which the amount of benefits is independent of the duration of periods of insurance
(arts 3739), and those under which the amount of the invalidity benefit depends on the duration of
insurance periods (art 40). Considering the nature of the Italian legislation on the matter, it can be seen at
the outset that only the provisions of art 40 are applicable to the situation of the applicant in the main
proceedings.
26.
His case also falls within the scope ratione personae of Regulation 1408/71, art 2(1) of which refers to
employed or self-employed persons who are or have been subject to the legislation of one or more
Member States and who are nationals of one of the Member States. Since employed person for the
purposes of the regulation is to be understood solely by reference to the social security scheme that is
applicable to him (art 1(a)), it follows that, in order to come within the scope of Regulation 1408/71 ratione
personae, it is sufficient for a national of a member state to be, or to have been, subject to a social
security scheme of one or more member states, which is the case here.
27.
After those introductory remarks, I turn now to the question referred by the national court. I shall begin
by considering whether the INPSs refusal of invalidity benefit to Mr lurlaro, on the ground that one of the
requirements of the Italian legislation was not met, was a proper one having regard to the need to comply
with the principle that insurance periods are to be aggregated. Only then will I examine the question as
formulated, concerning the exportation, for the purposes of determining the reference period in Italy, of
the neutralisation under German law of periods of unemployment for which benefit is paid.
372

ARGUMENT

The question whether the INPS properly refused to grant the applicant an invalidity
allowance
28.
I think it is necessary to dwell for a moment on the question whether in this case the principle that
insurance periods are to be aggregated was properly applied, since mere reading of the documents
produced during the written procedure could give rise to doubts on the matter.
29.
That principle, which, as I have already said, underlies the Community rules on this matter, appears in
art 51(a) of the Treaty, which provides for the adoption by the EU Council, in the field of social security, of
such measures as are necessary to provide freedom of movement for workers, particularly by making
arrangements to secure for migrant workers aggregation, for the purpose of acquiring and retaining the
right to benefit and of calculating the amount of benefit, of all periods taken into account under the laws of
the several countries.
30.
That principle is taken up by art 9(2) of Regulation 1408/71, which equates insurance periods
completed under the legislation of any member state with those completed under the legislation of the
state in which benefit is sought.
In the words of this court (Istituto Nazionale della Previdenza Sociale (INPS) v Baglieri Case C-297/92
[1993] ECR I-5211 (para 11)):
the object of that provision is to guarantee that periods of insurance completed in different
member states are treated as equivalent so that the persons concerned can satisfy the condition of
a minimum length of insurance periods where national legislation makes admission to a voluntary
or optional continued insurance scheme subject to such a condition.
31.
That principle is reaffirmed in respect of nearly all the risks governed by Regulation 1408/71 and,
specifically on the matter of invalidity, which is at issue here, in art 40(1) 17.
17
By reference to the application by analogy of art 45(1) of the regulation.

32.
The principle thus excludes the possibility that, whenever a worker begins a new insurance period in a
member state, he is regarded there as a newly insured person. The institution of the member state in
which an invalidity benefit is applied for is required to take into account, in so far as may be necessary,
periods completed by the migrant worker in any other member state as if they were periods completed
under the legislation which it administers, without discrimination as against other workers by reason of
the exercise of his right to freedom of movement (see Labots (ne Hagenbeek) v Raad van Arbeid,
Arnhem Case 100/63 [1966] ECR 425 at 430).
33.
In this case, Mr lurlaros application for recognition of his entitlement to an invalidity pension was
rejected on the ground that he did not fulfil the condition, laid down by Italian legislation, that the required
number of contributions (three years of pensionable service) have been paid on his account during the
reference period (five years preceding the application).
34.
During those five years Mr lurlaro benefited, during at least three of them, from the payment of
unemployment benefit by the competent German institution.
35.
For the purposes of calculating periods conferring entitlement to the benefit in question, Italian
legislation, as has been seen, equates periods of unemployment with contribution periods.
373
36.
In principle, therefore, under Italian legislation, Mr lurlaro may claim to have complied with the
condition the non-fulfilment of which had been raised in objection to his application. That objection
appears, in the final analysis, to be motivated solely by the consideration that the condition in question
was fulfilled under the legislation of a member state other than Italy.
37.
On the basis of those considerations alone, as for the reference, the refusal by the INPS would appear
above-mentioned principle that insurance periods are to set out in the order to be contrary to be
aggregated.
38.
The courts reasoning in its judgment in Lepore v Office National des Pensions (ONP) Joined cases C-
4546/92 [1993] ECR I-6497, concerning old-age pension, is in that respect entirely transposable to the
present case18.
18
For an illuminating commentary on that judgment, see Van Raepenbusch La scurit sociale des personnes qui se
dplacent lintrieur de la Communaut (mai 1992avril 1994) Journal des tribunaux. Droit europen n10 (1994) p
105.

In that judgment, the court held that the requirements of freedom of movement meant that, when
calculating old-age pension, periods of invalidity had to be treated as periods of active employment, even
where the applicable national legislation provided for such assimilation only where the incapacity for work
arose at a time when the worker was employed in the member state in question, and if, at the time the
incapacity for work arose, he was in reality employed in another member state.
It held that
it would be contrary to Articles 48 to 51 of the Treaty if, as a consequence of the exercise of
their right to freedom of movement, migrant workers were to lose the advantages in the field of
social security guaranteed to them by the laws of a single Member State; such a consequence
might discourage Community workers from exercising their right to freedom of movement and
would therefore constitute an obstacle to that freedom. (See Lepore [1993] ECR I-6497 (para 21).)
The court went on to observe (para 22):
the prospect of a workers losing, in one Member State, the right to have periods of invalidity
treated as periods of insurance, which would occur if he went to work in another Member State, is
likely, in certain circumstances, to discourage him from exercising his right to freedom of
movement.
39.
If regard is had solely to procedural documents lodged before the hearing, therefore, the question
referred might not appear to be directly relevant. A proper application of arts 48(2) and 51 of the Treaty
would suffice; the Italian legislation would encroach upon the freedom of movement for migrant workers if
it provided that only periods of insurance against unemployment completed on national territory could be
regarded as relevant contribution periods for the purposes of social advantages, to the exclusion of
similar periods completed on the territory of other member states.
40.
It would have to be concluded that Community law precluded the application of national legislation,
such as that administered by the INPS, which did not assimilate periods of unemployment insurance
completed in another member state to those completed on its territory, for the purposes of calculating the
minimum insurance requirement attached to the grant of the invalidity allowance.
374
41.
However, the representatives of the INPS put forward at the hearing a consideration which is
conclusive in this case. They explained that Italian legislation provides for the coverage of periods of
unemployment for only a limited period, which may not exceed six months.
42.
In that case, the line of reasoning set out above cannot apply.
43.
 Mr lurlaro could benefit from the rule, laid down by the Italian legislation, assimilating periods of
unemployment to contribution periods only in respect of a period not exceeding six months. The
requirement for entitlement to a pensionevidence of three years contributions during the reference
periodwould thus clearly not be fulfilled. The refusal of his claim by the INPS would therefore be
justified.
44.
Although, as Community law stands, member states are free to enact rules favouring certain migrant
workers19, Mr lurlaro cannot, in the absence of such national provisions, be granted advantages additional
to those which he would have been able to claim had he not exercised his right to freedom of movement,
merely because he exercised that right.
19
Thus the court held thatthere is no rule prohibiting a Member State from granting nationals who have worked in a
non-member country and then returned to their country of origin, where they no longer work, more favourable treatment
than nationals who have worked in another Member State and subsequently find themselves in the same situation (see
Baglieri [1993] ECR I-5211 (para 18)).

45.
As I have already mentioned, Community rules on social security envisage only the coordination of
existing national systems. They do not in any way introduce an independent system for migrant workers.
In that respect it would not be permissible for a migrant worker like Mr lurlaro to derive advantage from a
system that proved to be discriminatory to the detriment of non-migrant workers. The objective is not to
allow the accumulation of the most favourable conditions for acquisition of entitlement to benefit laid down
by the various national laws to which a migrant worker is successively subject, for the purpose of
acquiring entitlement to pension.
46.
That would, however, be the result if one were to disapply the rule on the maximum duration of
coverage of periods of unemployment which is laid down in the Italian legislation.
47.
It is in any case for the national court to satisfy itself of the existence and applicability of that rule from
the point of view of national law and, if appropriate, to draw the conclusion set out above.
48.
I shall now, as invited by the Pretore di Roma, examine the question whether Mr lurlaro might,
nevertheless, lay claim to an invalidity pension by requiring the INPS to apply the rule, known only to
German legislation, that the period of insurance against unemployment completed by the person
concerned in that other state is to be suspended for the purpose of extending the reference period.

The question of the exportation of the neutralisation which is provided for solely by the
German legislation
49.
On that point, the national court asks whether it is relevant to apply two provisions of Community
legislation, and requests an interpretation for that purpose. As will be seen, neither of those provisions
may be invoked in this case.
375
50.
The first of those provisions is art 9a of Regulation 1408/71, which provides, under Title I (General
Provisions) in relation to the prolongation of the reference period, that:
Where under the legislation of a Member State recognition of entitlement to a benefit is
conditional upon completion of a minimum period of insurance during a specific period preceding
the contingency insured against (reference period) and where the aforementioned legislation
provides that the period during which the benefits have been granted under the legislation of that
Member State or periods devoted to the upbringing of children in the territory of that Member State
shall give rise to prolongation of the reference period, periods during which invalidity pensions or
old-age pensions or sickness benefits, unemployment benefits or benefits for accidents at work
(except for pensions) have been awarded under the legislation of another Member State and
periods devoted to the upbringing of children in the territory of another Member State shall likewise
give rise to a prolongation of the aforesaid reference period. (My emphasis.)
51.
That provision thus imposes an obligation, on the member state in which prolongation of the reference
period is allowed, also to take account, within that period, of periods during which certain benefits were
awarded under the legislation of another member state, without, it seems, requiring that those benefits
should also have given rise to prolongation in that latter state.
That interpretation is confirmed by a reading of the third recital in the preamble to Regulation 2332/89,
cited above, which gives this reason for the insertion of art 9a in the basic regulation:
Whereas a provision should be introduced permitting a Member State whose legislation
provides for the prolongation, by certain facts or circumstances, of a reference period preceding the
occurrence of the risks insured against, during which a minimum period of insurance must be
fulfilled, for the recognition of the right to a benefit, to take account of similar facts or circumstances
occurring in another Member State for the aforementioned prolongation. (My emphasis.)
52.
 A mere reading of that provision is, it seems to me, sufficient to show that it does not apply to the case
with which we are concerned.
53.
What art 9a ultimately constitutes is the application to the particular case of states whose legislation
provides for the prolongation of the reference period of the principle, laid down in para 2 of the preceding
article, that insurance periods are to be aggregated 20. Since Italian legislation lays down no such rule in
respect of the benefits in question21, reliance on that provision serves no purpose in this case 22. The most
that can be said is that it would be different if the application for benefit had been made in Germany, since
that state lays down the principle of 376 prolongation of the reference period for the period in which the
person concerned was covered against unemployment23.
20
In his opinion in Paraschi v Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501 (para 15),
Advocate General Tesauro saw it as a declaratory provision expounding the obligation of non-discrimination laid down
by the Treaty.
21
The INPS has referred to certain periods, which alone are capable of being regarded as neutral, and thus capable of
being taken into account for the prolongation of the reference period, provided for in art 37 of Presidential Decree No
818 of 26 April 1957.
22
Nor does Mr lurlaro dispute that assessment: in so far as Italian legislation does not enact rules neutralising periods of
unemployment in respect of which benefit is paid, it does not appear that that provision can help to resolve the question
at issue in this case (see point IX, para 3 of his observations).
23
Thus the Court of Justice held, concerning the conditions for entitlement to an invalidity pension, that it was incompatible
with Community law for facts and circumstances allowing prolongation of the reference period not to be taken into
account by a member state which provides for that prolongation if they occur in another member state (see Paraschi
[1991] ECR I-4501 (paras 2425)).

54.
Turning now to art 15 of Regulation 574/72, that provision appears under Title IV, concerning
Implementation of the special provisions of the regulation [1408/71] relating to the various categories of
benefits and constitutes Ch 1, concerning General rules for the aggregation of periods. It provides:
1. In the cases referred to in Articles 18(1)24, 3825, 45(1) to (3)26, 6427 and 67(1) and (2)28 of the
Regulation, aggregation of periods shall be effected in accordance with the following rules (f)
Where under the legislation of one Member State certain periods of insurance or residence are
taken into account only if they have been completed within a specified time limit, the institution
which administers such legislation shall (ii) extend such time limit for the duration of periods of
insurance or residence completed wholly or partly within the said time limit under the legislation of
another Member State, where the periods of insurance or residence involved under the legislation
of the second Member State give rise only to the suspension of the time limit within which the
periods of insurance or residence must be completed.

24
The provision concerning Aggregation of periods of insurance, employment or residence in relation to sickness and
maternity benefits.
25
This provision concerns the Consideration of periods of insurance or of residence completed under the legislation to
which an employed person or a self-employed person was subject for the acquisition, retention or recovery of the right
to benefits for the purposes of granting invalidity benefits to workers subject to legislations under which the amount of
benefits is independent of the duration of periods of insurance.
26
Concerning the Consideration of periods of insurance or of residence completed under the legislation to which an
employed or self-employed person was subject, for the acquisition, retention or recovery of the right to benefits in
relation to old age and death (pensions).
27
Concerning the Aggregation of periods of insurance or residence in relation to death grants.
28
Concerning the Aggregation of periods of insurance or employment in relation to unemployment.

55.
Although that provision requires the competent institution of a member state to which application is
made for a pension to extend the reference period for the grant of that pension if the person concerned
has benefited from periods of insurance or residence in another member state where that prolongation is
provided for29, its scope is limited to those cases which it refers to expressly. Those cases do not include
that of Mr lurlaro, which concerns invalidity allowance for workers subject exclusively to legislation under
which the amount of the invalidity benefit depends on the duration of the insurance periods. That
provision makes no reference to arts 40ff of Regulation 1408/71, which, by contrast, applies inter alia to
the case of workers subject exclusively to legislation under which the amount of the benefits does not
depend on the duration of the insurance periods.
29
The court held in Rijksdienst voor Werknemerspensioenen v Celestre, Nationaal Pensioenenfonds voor Mijnwerkers
Joined cases 116117/80 and 119121/80 [1981] ECR 1737 (para 13), that: Regulation No 574/72 contains in
Articles 15 and 46 provisions governing the overlapping of periods of insurance completed under the legislation of two
or more Member States. Consequently, it is not permissible for the institution of a Member State to apply national rules
for the aggregation and apportionment of periods of insurance which are less favourable to the worker than those
contained in the regulation.

377
56.
Since neither of the two provisions referred to by the Pretore di Roma is applicable to this case, it is
not possible to deduce from either of them an obligation on the part of the INPS to prolong the reference
period, laid down by Italian legislation for calculating the minimum insurance requirement for acquisition of
entitlement to an invalidity benefit, by means of the export of the rule laid down by German legislation for
the suspension of the period during which Mr lurlaro drew unemployment benefit in that latter state, where
no such rule exists in Italian law.
57.
 On a more general note, I would mention once again the freedom enjoyed by the various member
states in laying down conditions for the acquisition of entitlements in social security matters. In the
absence of Community harmonisation, those states are in particular free not to provide for grounds for the
prolongation or neutralisation of the relevant periods, provided there is no discrimination under art 48(2) of
the Treaty.
58.
I must again stress the fact that Mr lurlaro cannot rely on an independent system for migrant workers
which would allow such persons to accumulate, for the purpose of acquiring entitlement, all the most
favourable conditions of the various member states to whose legislation they were successively subject.
59.
Nor, moreover, have Mr lurlaros legal rights been curtailed in any way by the exercise of his right to
free movement. He has not, it is true, acquired any new right to the grant of an invalidity benefit (although
the aim of the rules is not per se to benefit migrant workers). Nor, however, has he been deprived of a
right which would have been his if he had stayed in his country of origin. And that is the essential
objective of the Community rules on the matter.

CONCLUSION
60.
On the basis of the considerations set out above, I suggest that the reply to the question referred by
the Pretura Circondariale di Roma should be as follows:
Neither art 15(1)(f) of Council Regulation (EEC) 574/72 laying down the procedure for
implementing Council Regulation (EEC) 1408/71 on the application of social security schemes to
employed persons, to self-employed persons and to members of their families moving within the
Community, nor art 9a of Council Regulation (EEC) 1408/71, as introduced with retrospective effect
by Council Regulation (EEC) 2332/89, nor arts 48(2) and 51 of the EC Treaty preclude the refusal
by a member state (in this case the Italian Republic) to take into account the period during which a
worker who applies for the grant of an invalidity pension drew unemployment benefit in another
member state (in this case Germany) as a ground for prolonging the reference period for the grant
of the pension applied for, where that ground for prolongation exists under the legislation of the
second member state but not under its own legislation.

17 September 1997.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 3 October 1995, received at the Court of Justice of the European Communities on 16
October 1995, the Pretura Circondariale di Roma (the Rome District Magistrates Court) referred to the
Court of Justice for a preliminary ruling under art 177 of the EC Treaty a question on the interpretation of
art 9a of Council Regulation (EEC) 1408/71 on the application of social security schemes to employed
persons, to self-employed persons and to members of their families 378 moving within the Community
(OJ S Edn 1971 (II) p 416), as amended and updated by Council Regulation (EEC) 2001/83 (OJ 1983
L230 p 6) and subsequently amended by Council Regulation (EEC) 2332/89 (OJ 1989 L224, p 1), and of
art 15 of Council Regulation (EEC) 574/72 laying down the procedure for implementing Regulation
1408/71 (OJ S Edn 1972 (I) p 160), as amended and updated by Regulation 2001/83.
2.
The question was raised in proceedings between Mr Iurlaro, an Italian national, and the Istituto
Nazionale della Previdenza Sociale (the National Social Welfare Institution) (the INPS) concerning the
latters refusal to take into account, for the purpose of the acquisition of entitlement to an invalidity
allowance under Italian law, periods of unemployment spent by Mr Iurlaro in Germany.
3.
The documents before the court show that between January 1954 and March 1956 Mr Iurlaro was
insured in Italy and that, under that insurance, he paid 110 weekly social security contributions in that
country. He then moved to Germany, where he worked and paid social security contributions from 1959. It
is common ground that between 18 October 1984 and 18 October 1989 Mr Iurlaro received
unemployment benefit under German law, save for the period from 15 August 1989 to 1 October 1989,
during which sickness benefits were paid to him.
4.
On 18 October 1989, suffering from an infirmity which reduced his working capacity, in posts
corresponding to his abilities, by more than two-thirds, Mr Iurlaro applied to the INPS for an invalidity
allowance in Italy pursuant to art 1 of Law No 222 of 12 June 1984 concerning revision of the rules on
invalidity pension (GURI No 165 of 16 June 1984).
5.
Under art 4 of Law No 222 acquisition of entitlement to an invalidity allowance and an invalidity
pension is subject to fulfilment of the insurance and contribution conditions laid down in art 9(2) of
Decree-Law No 636 of 14 April 1939, which became Law No 1272 of 6 July 1939. That provision, which
was replaced by art 2 of Law No 218 of 4 April 1952 on the reform of pensions in respect of compulsory
invalidity, old-age and survivorship insurance (GURI No 89 of 15 April 1952, ordinary supplement), as
amended by Law No 222, requires for that purpose that at least five years are to have elapsed since the
person concerned joined the insurance scheme, and that at least 260 weekly contributions (five years of
pensionable service) are to have been paid or credited in his favour, of which at least 156 (three years of
pensionable service) are to have been paid during the five years preceding his application for benefit.
6.
Article 4 of Law No 218 provides that periods in respect of which the ordinary allowance in respect of
compulsory unemployment insurance is paid shall be regarded as contribution periods for the purposes of
entitlement to pension and of the amount of pension.
7.
In addition, the order for reference shows that, under para 43 of the Sozialgesetzbuch (the Social Law
Code) (BGBl III 860), as amended by the Gesetz zur Reform der gesetzlichen Rentenversicherung (the
Law on the Reform of Statutory Old-Age Insurance) of 28 December 1989 (BGBl IS 2261), insured
persons are entitled, until they reach the age of 65 years, to an invalidity pension if they are unable to
carry on their occupation, have three years of compulsory contributions during the five years last
preceding the onset of the invalidity, and have completed the general contribution periods before the
onset of the invalidity.
8.
Under the same provision the reference period of five years prior to the onset of the invalidity is
extended by periods which are assimilated to contribution periods (Anrechnungszeiten), such periods
including, pursuant to para 58 of the Sozialgesetzbuch, periods during which the worker was registered
on 379 unemployment lists in Germany and drew a benefit from bodies governed by public law.
Those periods, which are thus taken into account in determining the reference period but are not taken
into account as contribution periods for the purposes of acquiring entitlement to a pension in Germany,
give rise furthermore to the payment of notional contributions for the purposes of calculating the amount
of the pension.
9.
Mr Iurlaros application was rejected by the INPS on the ground that the minimum contribution
requirement laid down by Law No 222 was not fulfilled. In its view, whilst Mr Iurlaro had, since 1954, paid
110 weekly contributions in Italy and 854 weekly contributions in Germany, he had, during the reference
period referred to in art 4 of Law No 222, in this case running from 18 October 1984 until 18 October
1989, drawn unemployment benefits in Germany but not paid any social security contributions.
10.
His complaint against that decision having been rejected, Mr Iurlaro brought an action on 24 June
1994 before the Pretura Circondariale di Roma, arguing that the contribution requirement laid down by
Law No 222 was fulfilled on the ground, inter alia, that the effect of extension of the reference period
provided for by the German legislation as regards periods of unemployment was that the reference period
to be taken into account for the application of the Italian legislation began on 18 October 1978.
11.
Referring to art 48 of the EC Treaty, the Pretore di Roma is uncertain as to whether the rule in art 4 of
Law No 222 should be disapplied on the ground that it conflicts with art 15(1)(f) of Regulation 574/72,
which provides:
In the cases referred to in Articles 18(1), 38, 45(1) to (3), 64, and 67(1) and (2) of the
Regulation, aggregation of periods shall be effected in accordance with the following rules (f)
where, under the legislation of one Member State, certain periods of insurance or residence are
taken into account only if they have been completed within a specified time limit, the institution
which administers such legislation shall (ii) extend such time limit for the duration of periods of
insurance or residence completed wholly or partly within the said time limit under the legislation of
another Member State, where the periods of insurance or residence under the legislation of the
second Member State give rise only to the suspension of the time limit within which the periods of
insurance or residence must be completed.
12.
The national court finds that rule to be unclear, since art 9a of Regulation 1408/71 appears to make its
application conditional upon the legislation of both countries having made provision for the neutrality of
the specified periods, thereby excluding application of the rule in a situation where that possibility is
provided for in one of the two member states (in this case, Germany) but not in the other (in this case,
Italy).
13.
Under art 9a of Regulation 1408/71:
[W]here, under the legislation of a Member State, recognition of entitlement to a benefit is
conditional upon completion of a minimum period of insurance during a specific period preceding
the contingency insured against (reference period) and where the aforementioned legislation
provides that the periods during which the benefits have been granted under the legislation of that
Member State or periods devoted to the upbringing of children in the territory of that Member State
shall give rise to prolongation 380 of the reference period, periods during which invalidity pensions
or old-age pensions or sickness benefits, unemployment benefits or benefits for accidents at work
(except for pensions) have been awarded under the legislation of another Member State and
periods devoted to the upbringing of children in the territory of another Member State shall likewise
give rise to prolongation of the aforesaid reference period.
14.
 In view of those doubts the Pretore di Roma decided to stay proceedings and refer to the Court of
Justice
the question of the interpretation of Article 15(1)(f) of Regulation 574/72 and Article 9a of
Regulation 1408/71, in the light of Article 48 of the Treaty establishing the EEC, in order to
ascertain whether Article 4 of Law No 222/1984 should be so applied as to prolong the reference
period for the recognition of entitlement to invalidity pension where a worker has been in receipt of
unemployment benefit in another Member State (in this case Germany) in which such extension is
provided for and, if so, whether such prolongation should be subject to conditions.
15.
By that question the national court asks essentially whether arts 48 to 51 of the Treaty, art 9a of
Regulation 1408/71 and art 15(1)(f)(ii) of Regulation 574/72 must be interpreted as requiring a member
state to extend the reference period which is laid down by its legislation for determining the minimum
insurance requirement for the grant of an invalidity benefit by a period equivalent to the periods of
unemployment spent by the person concerned under the legislation of another member state which,
unlike that of the first member state, allows such an extension where the periods of unemployment are
spent on national territory.
16.
For the purposes of answering this question it should be borne in mind that, whilst the Italian
legislation applicable in this case takes account of periods of unemployment in respect of which benefit
has been paid both for the purpose of entitlement to a pension and for the purpose of calculating its
amount, under German law the period during which a worker draws unemployment benefits in Germany,
whilst giving rise to notional contributions for the purpose of determining the amount of the pension,
serves to extend the five-year reference period as regards the minimum contribution required for
recognition of entitlement to a pension.
17.
It must therefore be held at the outset that the requirements for applying art 9a of Regulation 1408/71,
to which the question put to the court refers, are not satisfied. As the Advocate General points out at para
51 of his opinion, above, the very wording of that provision shows that it is concerned solely with a
situation where the legislation of a member state which makes entitlement to a benefit conditional upon
completion of a minimum period of insurance during a reference period itself allows for the extension of
that period by periods during which certain benefits, and in particular unemployment benefits, were
granted under the legislation of another member state. That is not the situation in the case of legislation,
such as the Italian legislation in point in the main proceedings, under which periods of unemployment do
not extend the reference period.
18.
It should next be noted that Mr Iurlaro has been subject only to legislation on invalidity benefits of
Type B, that is to say, under which the amount of benefit is determined on the basis of the duration of the
insurance periods completed. In such a case art 40(1) of Regulation 1408/71 provides that the worker in
question is 381 to receive benefits under the provisions of Ch 3 (arts 44 to 51) concerning old-age and
death benefits, which are to apply by analogy.
19.
Under art 45(1) of that regulation the competent institution of a member state whose legislation makes
the acquisition of the right to old-age benefits conditional upon the completion of insurance periods is to
take account of periods completed under the legislation of any other member state as if they were periods
completed under the legislation which it administers.
20.
In that regard, art 15(1)(f)(ii) of Regulation 574/72, which applies in the cases referred to in Articles
18(1), 38, 45(1) to (3), 64, and 67(1) and (2) of the Regulation, provides that the institution of a member
state whose legislation takes certain insurance periods into account only if they have been completed
within a specified time limit must extend such time limit for the duration of insurance periods completed
within that time limit under the legislation of another member state, where those periods give rise, under
the legislation of that latter member state, only to the suspension of the time limit within which the
insurance periods are to be completed.
21.
In a case such as that in point in the main proceedings, however, the periods of unemployment which
were spent under German legislation during the reference period which is to be taken into account under
Italian legislation do not give rise, under the legislation of the first state, solely to the suspension of the
time limit within which the insurance periods are to be completed. Under the German legislation
applicable to the dispute in the main proceedings, as para 16 of this judgment makes clear, periods of
unemployment in respect of which benefit is paid are also taken into account for the purposes of
calculating the amount of the invalidity benefit.
22.
In those circumstances, the court finds that in this case art 15 of Regulation 574/72 cannot be relied
upon to any useful purpose, without its even being necessary to consider whether that provision covers
cases where, as here, art 45 of Regulation 1408/71 is applicable not directly but, under art 40 of the same
regulation, by analogy.
23.
Finally, the court refers to its case law, to the effect that art 51 of the EC Treaty and Regulation
1408/71 do not regulate the conditions under which insurance periods are constituted (see esp Bestuur
van de Nieuwe Algemene Bedrijfsvereniging v V A Drake Case C-12/93 [1994] ECR I-4337 (para 26)). It
is for each member state to determine the conditions governing the right or obligation to become a
member of a social security scheme, provided that there is no overt or covert discrimination in that regard
between nationals of the host member state and those of other member states.
24.
In this case the Italian legislation, in preventing the taking into consideration, in order to extend the
five-year reference period which it lays down for the purposes of calculating the minimum insurance
requirement for the grant of an invalidity benefit, of periods during which the person concerned spent
periods of unemployment under either Italian legislation or that of another member state, employs
objective criteria and applies without distinction to national workers and those of other member states.
25.
In those circumstances, the court concludes that arts 48 to 51 of the Treaty, art 9a of Regulation
1408/71 and art 15(1)(f)(ii) of Regulation 574/72 must be interpreted as not requiring a member state to
extend the reference period laid down by its legislation for determining the minimum insurance
requirement for the grant of an invalidity benefit, by a period equivalent to periods of unemployment spent
by the person concerned under the legislation of another 382 member state which, unlike that of the first
member state, allows such an extension where the periods of unemployment are spent on national
territory.
26.
Mr Iurlaro and the Commission also submit that the Italian institution, when applying the aggregation
rules laid down in art 45(1) of Regulation 1408/71, must take into account not only periods of
unemployment spent in Italy during the reference period in question, but also similar periods spent in
Germany, even if, under German legislation, periods of unemployment are not taken into account for the
purposes of the acquisition of entitlement to invalidity benefit. In this case Mr Iurlaro spent periods of
unemployment in Germany totalling more than three years during the five-year period prior to his
application for invalidity benefit in Italy. He therefore fulfilled the requirement under art 4 of Law No 222
that at least 156 weekly contributions (three years of pensionable service) must have been paid or
credited in favour of the person concerned during the five years preceding his application for benefit.
27.
It should be noted in that respect that the term periods of insurance is defined in art 1(r) of Regulation
1408/71 as
periods of contribution or periods of employment or self-employment as defined or recognised
as periods of insurance by the legislation under which they were completed or considered as
completed, and all periods treated as such, where they are regarded by the said legislation as
equivalent to periods of insurance.
28.
It follows that, for the purposes, inter alia, of applying art 45 of Regulation 1408/71, periods of
insurance means periods defined or recognised as such by the legislation under which they were
completed (see esp concerning periods treated as such, Vella v Alliance nationale des mutualits
chrtiennes Case C-324/88 [1990] ECR I-257), subject however to compliance with arts 48 to 51 of the
Treaty (see, on that point, Caisse auxiliaire dassurance maladie-invalidit (CAAMI) v Faux Case C-
302/90 [1991] ECR I-4875 (paras 2528)).
29.
National legislation which takes into account for the purposes of acquisition of entitlement to invalidity
allowance only such periods of insurance against unemployment that were completed on national
territory, to the exclusion of similar periods completed in the territory of other member states, infringes
those provisions of the Treaty.
30.
Such a requirement, for which no objective justification has been put forward and which establishes a
difference in treatment between workers who have not exercised their right to freedom of movement and
migrant workers, to the detriment of the latter, must be held to be discriminatory and thus in breach of the
fundamental rules of the Treaty which are designed to ensure freedom of movement for workers.
31.
It should, however, be pointed out that, at the hearing before the court, the INPS and the Italian
government stated, without being contradicted, that it is only for a period not exceeding six months that
the Italian scheme treats periods of unemployment as periods of contribution for the purposes of
calculating whether entitlement to invalidity benefit has been acquired.
32.
Having regard to those statements, the accuracy of which it is for the national court to verify, the court
finds that arts 48 to 51 of the Treaty do not preclude the legislation of a member state, such as the Italian
legislation, from limiting to a period of six months the taking into account, for the purposes of granting an
invalidity benefit, periods of unemployment spent in another member 383 state, where such limitation is
applicable also to cases where those periods are spent in the member state of the competent institution.
33.
The answer to the question referred to the court must therefore be that arts 48 to 51 of the Treaty, art
9a of Regulation 1408/71 and art 15(1)(f)(ii) of Regulation 574/72 must be interpreted as not requiring a
member state to extend the reference period laid down by its legislation for determining the minimum
insurance requirement for the grant of an invalidity benefit by a period equivalent to the periods of
unemployment spent by the person concerned under the legislation of another member state which,
unlike that of the first member state, allows such an extension where the periods of unemployment are
spent on national territory. Moreover, arts 48 to 51 of the Treaty do not preclude the legislation of a
member state from refusing to take into account, for the purposes of calculating whether the minimum
insurance requirement attaching to the grant of an invalidity benefit is satisfied, periods of insurance
against unemployment completed during a given period preceding the contingency insured against under
the legislation of another member state, over and above those which are taken into account by the
legislation of the first member state during that same period.

Costs
34.
The costs incurred by the Italian government and the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the proceedings pending before the national court, the decision on costs
is a matter for that court.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the question referred to it by the
Pretura Circondariale di Roma by order of 3 October 1995, hereby rules: arts 48 to 51 of the EC Treaty,
art 9a of Council Regulation (EEC) 1408/71 on the application of social security schemes to employed
persons, to self-employed persons and to members of their families moving within the Community, as
amended and updated by Council Regulation (EEC) 2001/83 and subsequently amended by Council
Regulation (EEC) 2332/89, and art 15(1)(f)(ii) of Council Regulation (EEC) 574/72 laying down the
procedure for implementing Regulation 1408/71, as amended and updated by Regulation 2001/83 must
be interpreted as not requiring a member state to extend the reference period laid down by its legislation
for determining the minimum insurance requirement for the grant of an invalidity benefit by a period
equivalent to the periods of unemployment spent by the person concerned under the legislation of
another member state which, unlike that of the first member state, allows such an extension where the
periods of unemployment are spent on national territory. Moreover, arts 48 to 51 of the Treaty do not
preclude the legislation of a member state from refusing to take into account, for the purposes of
calculating whether the minimum insurance requirement attaching to the grant of an invalidity benefit is
satisfied, periods of insurance against unemployment completed during a given period preceding the
contingency insured against under the legislation of another member state, over and above those which
are taken into account by the legislation of the first member state during that same period.

384

[1998] All ER (EC) 385

Caisse Nationale dAssurance Vieillesse des Travailleurs Salaris v

Thibault
(Case-136/95)

EUROPEAN COMMUNITY; Social policy: EMPLOYMENT; Discrimination

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT), SCHINTGEN, MANCINI, MURRAY (RAPPORTEUR) AND
HIRSCH. ADVOCATE GENERAL RUIZ-JARABO COLOMER
5 DECEMBER 1996, 9 JANUARY 1997, 30 APRIL 1998
European Community Equality of treatment of men and women Equal working conditions
Employers service regulations making right to performance assessment conditional on 6 out of 12
months presence at work Employee denied assessment due to seven months absence from work
Absence due mainly to maternity leave Whether principle of equality of treatment infringed Council
Directive (EEC) 76/207, arts 2(3), 5(1).

T had been employed by the national old age insurance fund for employees since 1973. In 1983 Ts
combined sickness and maternity leave meant that she was only at work for a period of about five
months. Relying on the standard service regulations, Ts employer refused to draw up a performance
assessment for her, since she did not meet the conditions laid down in the regulations, namely six
months presence at work. Since such assessments were a precondition to receiving a 2% discretionary
increase in pay, T brought the matter before the Labour Tribunal in Paris, claiming that the failure to
assess her performance, because of her absence on maternity leave, constituted discrimination on
grounds of sex and that she had lost an opportunity for promotion. Ts claim was upheld and her employer
was ordered to compensate her for the loss suffered. The employer appealed. Following a series of
appeals in which the courts considered the compatibility of the service regulations with the provisions of
the French Code du Travail implementing Council Directive (EEC) 76/207 on equal treatment for men and
women as regards access to employment, vocational training and promotion and working conditions, the
Cour de Cassation stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling the question whether, inter alia, art 5(1) 1 of that directive precluded
rules which deprived a woman of the right to an assessment of her performance because of absence
through maternity leave.
1
Article 5(1), so far as material, is set out at p 397 j, post
Held The right of any employee to have their performance assessed each year and, consequently, to
qualify for promotion, formed an integral part of the conditions of their contract of employment within the
meaning of art 5(1) of the directive. Moreover, art 2(3) 2 of the directive recognised the legitimacy, in terms
of the principle of equal treatment, of protecting women during pregnancy and after childbirth. Accordingly,
national rules concerning employment and working conditions had to be examined in the light of arts 5(1)
and 2(3) to determine whether they guaranteed men and women the same conditions 385 without
discrimination on grounds of sex. It was clear that a woman who was accorded unfavourable treatment
regarding her working conditions, in that she was deprived of the right to an annual assessment of her
performance and, therefore, of the opportunity to qualify for promotion as a result of absence through
maternity leave, was discriminated against on grounds of her pregnancy and her maternity leave. Such
conduct constituted discrimination based directly on grounds of sex within the meaning of the directive. It
followed that arts 2(3) and 5(1) precluded national rules which deprived a woman of the right to an
assessment of her performance because of absence through maternity leave (see p 400 f to j and p 401
e f j, post).
2
Article 2(3), so far as material, is set out at p 389 h, post

Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047, Habermann-Beltermann v

Arbeiterwohlfahrt, Bezirksverband Ndb./Opf-eV Case C-421/92 [1994] ECR I-1657 and Webb v EMO Air
Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115 applied.

Notes
For equal treatment relating to promotion and working conditions, see 52 Halsburys Laws (4th edn) para
2113.

Cases cited
EEC Commission v France Case 167/73 [1974] ECR 359.
Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990]
ECR I-3941.
Gillespie v Northern Health and Social Services Board Case C-342/93 [1996] All ER (EC) 284, [1996]
ECR I-475, ECJ.
Habermann-Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb./Opf-eV Case C-421/92 [1994] ECR I-
1657.
Handels- og Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiverforening Case C-179/88
[1990] ECR I-3979.
Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047.
Kempf v Staatssecretaris van Justitie Case 139/85 [1986] ECR 1741.
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Sinatra v Fonds national de retraite des ouvriers mineurs (FNROM) Case 296/84 [1986] ECR 1047.
Von Colson v Land Nordrhein-Westfalen Case 14/83 [1984] ECR 1891.
Wagner Miret v Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911.
Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115, [1994] QB 718, [1994] 1 WLR 941,
[1994] ECR I-3567, ECJ.

Reference
By order of 28 March 1995, the French Cour de Cassation referred to the Court of Justice of the
European Communities for a preliminary ruling under art 177 of the EC Treaty a question (set out at p 399
h, post) on the interpretation of Council Directive (EEC) 76/207 on the implementation of the principle of
equal treatment for men and women as regards access to employment, vocational training and
promotion, and working conditions. That question was raised in proceedings between the Caisse
Nationale dAssurance Vieillesse des Travailleurs Salaris (CNAVTS) and Mrs Thibault concerning the
refusal by the 386 CNAVTS to undertake an assessment of Mrs Thibaults work performance for 1983.
Written observations were submitted on behalf of: the French government, by C de Salins, Deputy
Director in the Legal Affairs Department of the Ministry of Foreign Affairs, and A de Bourgoing, Charg de
Mission in that department, acting as agents; the UK government, by J E Collins, Assistant Treasury
Solicitor, acting as agent; and the European Commission, by M Wolfcarius, of its Legal Service, acting as
agent. Oral observations were made by: the French government, represented by A de Bourgoing; the UK
government, represented by J E Collins and D Pannick QC; and the Commission, represented by M
Wolfcarius. The language of the case was French. The facts are set out in the opinion of the Advocate
General.

9 January 1997.
The Advocate General (Ruiz-Jarabo Colomer)
delivered the following opinion (translated from the Spanish).
1.
In this case, the court is called upon to answer a question submitted by the French Cour de Cassation
(the Court of Cassation) concerning the interpretation of certain provisions of Council Directive (EEC)
76/207 on the implementation of the principle of equal treatment for men and women as regards access
to employment, vocational training and promotion, and working conditions (OJ 1976 L39 p 40).
That question arose in relation to an application by the Caisse Nationale dAssurance Vieillesse des
Travailleurs Salaries (the National Old-Age Insurance Fund for Employees) (the CNAVTS), a social
security institution, for review of a judgment given in 1990 by the Conseil de Prudhommes (the Labour
Tribunal), Melun, in favour of Evelyne Thibault.
2.
As is evident from the order for reference, Mrs Thibault was engaged by the CNAVTS as an agent
technique in 1973 and promoted to the position of rdacteur juridique in 1983.
Throughout 1983, Mrs Thibaults attendance at work suffered from a certain lack of continuity: she was
absent on sick leave for a total of 52 days (4 to 13 February, 3 to 16 March and 16 May to 12 June); then
she took a 16-week period of maternity leave (13 June to 1 October), followed by a six-week period of
child-care leave (3 October to 16 November) under art 46 of the Convention Collective Nationale du
Travail du Personnel des Organismes de Scurit Sociale (the national collective labour agreement for
the staff of social security institutions). Her attendance at work for the whole year was recorded as
totalling 155 days.
3.
Relying on Ch XIII of the standard service regulations for the application of the collective agreement,
which provides that employees who are present at work for at least six months in any year must be the
subject of an assessment of performance by their immediate superiors, the CNAVTS refused to draw up
an assessment for Mrs Thibault in relation to 1983.
The fact that no assessment of performance was drawn up by her superiors meant that Mrs Thibault
could not be included on the list of staff, classified according to their assessment marks, which is used
annually for the purpose of awarding advancements on merit equivalent to 2% of salary.
4.
Mrs Thibault applied to the Conseil de Prudhommes, Paris, for an order that the CNAVTS draw up an
assessment for her in relation to 1983 and award her the appropriate advancement, claiming that the lack
of such an assessment, which was not drawn up for her because she had been absent on maternity
leave, 387constituted discrimination and that she had thereby lost an opportunity of obtaining the
advancement for the year in question.
By judgment of 17 December 1985, the defendant was ordered to pay the applicant arrears for the
period from 1 January to 31 December 1984 on the ground that she had been unfairly deprived of the
advancement, since maternity leave cannot be counted as absence for the purpose of calculating the
period of six months attendance at work which entitles employees to an assessment of performance.
5.
The CNAVTS applied for review of that judgment. The Cour de Cassation set it aside on 9 February
1989 and referred the case back to the Conseil de Prudhommes, Melun, for a new judgment.
In its judgment of 24 January 1990, the Conseil de Prudhommes allowed Mrs Thibaults application,
on the grounds that art L 123-1 of the Labour Code contains a clearly anti-discriminatory provision; that
her absence on maternity leave should have been counted as actual attendance for the purpose of
calculating the period necessary to create entitlement to an assessment of performance; and that,
accordingly, the applicant had been at work for more than six months, for which reason her immediate
superiors should have given her an assessment of performance. It therefore ordered the CNAVTS to pay
her arrears of F 3,334 for 1984.
6.
In the course of the proceedings brought by the CNAVTS for review of that judgment, the Cour de
Cassation, noting that the rule contained in art L 123-1(c) of the Labour Code was the result of the
transposition into national law of Directive 76/207, stayed the proceedings and referred the following
question to the Court of Justice for a preliminary ruling:
Must Articles 1(1), 2(1), 5(1) and, if relevant, 2(4) of Council Directive 76/207/EEC of 9
February 1976 be interpreted as meaning that a woman may not be deprived of the right to an
assessment of performance, and consequently to the possibility of an advancement in career, on
the ground that she was absent from work by reason of maternity leave?
7.
Under art 45 of the collective agreement, pregnant employees who prove a certain period of service
are entitled to sixteen weeks maternity leave on full pay, which may be extended to 28 weeks in some
cases. That time is not counted in the computation of sick leave and cannot give rise to any reduction in
annual leave. Moreover, under art 46, or completing her maternity leave, a woman is entitled to take
either three months leave on half-pay or one-and-a-half months leave on full pay, and to keep her job.
8.
Article L 122-26-2 of the Labour Code provides that the period of maternity leave is to be treated as a
period of actual work when determining a workers rights by virtue of length of service. Under art L 123-
1(c) of that code, subject to the special provisions laid down therein, unless belonging to a particular sex
is a decisive condition for holding a post or exercising a professional activity, no-one may adopt any
measure on the basis of sex, in particular in relation to pay, training, post, assessment, grading,
promotion or transfer.
9.
Articles 29 to 31 of the collective agreement provide for a system of advancements which, over a
period of time, can amount to a maximum of 40% of salary. First, on expiry of the second year after taking
up appointment, employees are entitled to receive, annually, 2% of salary by way of advancement linked
exclusively to length of service.
388
Secondly, once the third year after taking up appointment has elapsed and subject to a maximum
increase of 24%, the 2% may be supplemented, on a discretionary basis, by a further 2% which is
awarded on the basis of the employees being included on the list compiled annually by reference to the
mark which the quality of the employees work and performance have been assessed as meriting by his
or her immediate superiors. The proportion of recipients of this advancement on merit may not exceed
40% of the staff in each grade.
Once 24% of salary has been reached, and subject to a maximum of 40%, increases are granted only
on the basis of length of service, at a rate of 2%pa.
10.
Chapter XIII of the service regulations implements arts 29 to 31 of the collective agreement. In
particular, with regard to the discretionary award of the 2% increase, it provides that any employee who
has been present at work for at least six months must have an assessment of performance drawn up by
his or her immediate superiors.
11.
Moreover, the supplement of 13 November 1975, appended to the collective agreement, defines the
concept of professional experience as time actually spent at work. Article 3 thereof provides that, with
effect from 1 July 1973, in addition to actual attendance, the following absences must be counted as
periods of professional experience for the purpose of classifying posts: annual leave, special short-term
leave, time spent as an elected trade-union official, and absences on certain other specified grounds,
subject to a maximum of five working days per six-month period.
Article 3bis which was added by an annex of 15 December 1983 and entered into force on 16 January
1984, provides that maternity leave must be counted as a period of professional experience in the same
way as the absences listed in art 3.
12.
Article 1 of Directive 76/207 provides as follows:
(1) The purpose of this Directive is to put into effect in the Member States the principle of equal
treatment for men and women as regards access to employment, including promotion, and to
vocational training and as regards working conditions and, on the conditions referred to in
paragraph 2, social security. This principle is hereinafter referred to as the principle of equal
treatment.
13.
Article 2 provides:
(1) For the purposes of the following provisions, the principle of equal treatment shall mean that
there shall be no discrimination whatsoever on grounds of sex either directly or indirectly by
reference in particular to marital or family status (3) This Directive shall be without prejudice to
provisions concerning the protection of women, particularly as regards pregnancy and maternity.
(4) This Directive shall be without prejudice to measures to promote equal opportunity for men and
women, in particular by removing existing inequalities which affect womens opportunities in the
areas referred to in Article 1 (1).
14.
Finally, by virtue of art 5:
(1) Application of the principle of equal treatment with regard to working conditions, including
the conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex. (2) To this end, Member States shall take the
measures 389 necessary to ensure that: (a) any laws, regulations and administrative provisions
contrary to the principle of equal treatment shall be abolished; (b) any provisions contrary to the
principle of equal treatment which are included in collective agreements, individual contracts of
employment, internal rules of undertakings or in rules governing the independent occupations and
professions shall be, or may be declared, null and void or may be amended
15.
Written observations were submitted, within the time limit laid down for that purpose by art 20 of the
Protocol on the EC Statute of the Court of Justice, and oral observations were put forward at the hearing,
by the French government, the United Kingdom and the European Commission.
16.
The French government states that Directive 76/207 has been transposed into national law, and
specifically into art L 123-1 of the Labour Code, that infringement of its provisions entails criminal
sanctions, and that the prohibition of discrimination contained in the directive must be observed not only
in the legislation of the member states but also in the context of collective agreements concluded by
employers and workers.
In its opinion, the courts case law, according to which both refusal to enter into a contract of
employment with, and dismissal of, a pregnant woman constitute discrimination, must also apply to the
right to be given an assessment of performance and the right to advancement in career, since those
factors form part of working conditions for the purposes of art 5 of Directive 76/207 (see the judgments in
Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990]
ECR I-3941 and Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115, [1994] ECR I-3567).
In Mrs Thibaults case, the employer should have separated the period of maternity leave from the
period of sick leave; the latter is the only period which can be counted as absence from work for the
purpose of applying Ch XIII of the service regulations without infringing the principle of equal treatment,
since equating maternity leave with absence from work for any other reason is contrary to the courts case
law on equal treatment in so far as, in this case, it meant that an assessment of performance, the
existence of which is a necessary precondition for inclusion on the list of employees eligible to receive the
advancement on merit amounting to 2% of salary. was not drawn up for her in relation to 1983.
It adds that such discriminatory treatment arises not from the applicable provisions of laws and the
collective agreement but from the interpretation thereof by the CNAVTS, since under no circumstances is
an employer obliged to regard maternity leave as time absent from work, and those provisions must be
interpreted by the national court in such a way that they conform to the requirements of Community law.
The French government proposes that the court should reply in the affirmative to the question referred
to it by the Cour de Cassation for a preliminary ruling.
17.
The United Kingdom is of the opinion that Mrs Thibault was not the victim of discrimination on grounds
of sex, for four reasons: first, because the reason for which an assessment of performance was not drawn
up for her relates to her absence from work; second, because the situation of a woman who is absent
from work as a result of taking maternity leave cannot be compared to that of a man or woman who is
working; third, because the determination of all rights extended to women during maternity leave is a
matter for the member 390 states, without prejudice to the provisions of Council Directive (EEC) 92/85 on
the introduction of measures to encourage improvements in the safety and health at work of pregnant
workers and workers who have recently given birth or are breastfeeding (OJ 1992 L348, p 1); and, finally,
because Directive 76/207 does not confer any right to have periods of absence by reason of maternity
counted as periods of work which could give rise to entitlement to an assessment of performance.
It proposes that the court should reply to the national court that, where a workers right to have an
assessment of performance drawn up for him or her in relation to a particular period is subject to proof of
his or her attendance at work for a minimum length of time, Directive 76/207 does not require the
employer to make such an assessment in the case of a female worker who, having taken maternity leave,
was not at work for that minimum period.
18.
The Commission points out that pregnancy is a condition which can affect only women and that,
consequently, simply to treat maternity leave in the same way as sick leave even though they are two
distinct situations constitutes direct discrimination despite the fact that the criterion governing the
application of the rule, namely, absence from work for more than six months in the year in question, is
neutral. It adds that refusal to draw up an assessment of performance for a pregnant woman in the
circumstances described constitutes direct discrimination on grounds of sex, since that refusal is based
on a method of calculating periods of absence from work which adversely affects only pregnant women.
It takes the view that the question submitted for a preliminary ruling must be answered in the
affirmative.
19.
In view of the conflicting positions adopted, I will say at the outset that I disagree totally with the
opinions put forward by the United Kingdom and with the solution proposed in its written observations,
and that, conversely, I agree with most of the arguments put forward both by the French government and
by the Commission.
20.
Starting from the assumption that equality, as defined by the constitutional court of one of the member
states, is not a reality or an abstract mathematical concept, but rather unequal treatment of that which is
unequal or equal treatment of that which is similar or alike3 and taking into account the Court of Justices
consistent case law according to which discrimination can only arise through the application of different
rules to comparable situations or the application of the same rule to different situations 4, I shall now
consider whether, by simply treating maternity leave in the same way as sick leave, the contested rule in
the collective agreement, which applies to men and women alike, guarantees equal treatment because
both situations are similar, or whether, on the contrary, it introduces direct discrimination on grounds of
sex by applying the same rule to different situations since, for the purpose of calculating attendance at
work, it takes into consideration a category of leave which only women can take.
3
See Judgment No 29/1987 of the Spanish Tribunal Constitutional (Constitutional Court) of 6 March 1987; para 5 (b)
(BOE of 24 March 1987).
4
See the judgment in Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-
225 (para 30).

21.
 With regard to the protection of working women who are pregnant or have recently given birth, art 2(3)
of Directive 76/207 allows member states to 391 adopt provisions which introduce different treatment. In
its judgment in Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047 (para 25) the court held
that
by reserving to Member States the right to retain, or introduce provisions which are intended to
protect women in connection with pregnancy and maternity, the directive recognizes the
legitimacy, in terms of the principle of equal treatment, of protecting a womans needs in two
respects. First, it is legitimate to ensure the protection of a womans biological condition during
pregnancy and thereafter until such time as her physiological and mental functions have returned to
normal after childbirth; secondly, it is legitimate to protect the special relationship between a woman
and her child over the period which follows pregnancy and childbirth
Maternity leave, the rules for which are set out in arts 45 and 46 of the collective agreement governing
Mrs Thibaults employment relationship, which is normally split between the weeks immediately before
and after childbirth and is restricted to women, undoubtedly falls within the scope of that derogation.
22.
On this point, I agree with Advocate General Tesauro when he states in his opinion in Habermann-
Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb./Opf-eV Case C-421/92 [1994] ECR I-1657 (para
11):
On closer inspection the provisions adopted in implementation of Article 2 (3) of the directive
cannot properly be called derogations from the principle of equality, in that they seek rather to
ensure that that principle operates in substance, by permitting such inequalities as are necessary
in order to achieve equality. In short, different treatment is allowed or imposed, in favour of and to
protect female workers, in order to arrive at material and not formal equality, since that would
constitute a denial of equality.
23.
However, this case is concerned not with national provisions for the protection of women, adopted on
the basis of art 2(3) of Directive 76/207, but with the application of the principle of equal treatment as
regards access to employment and working conditions, laid down in art 5(1) of the directive.
24.
In interpreting Directive 76/207, the court has established clear and consistent case law setting out its
views on unfavourable treatment received by a woman in the labour market because of the fact that she
is pregnant.
Thus, in Dekker [1990] ECR I-3941 (para 14), it held that an employer is in direct contravention of the
principle of equal treatment embodied in arts 2(1) and 3(1) of Directive 76/207 if he refuses to enter into a
contract of employment with a female candidate whom he considers to be suitable for the job where such
refusal is based on the possible adverse consequences for him of employing a pregnant woman, owing to
rules on unfitness for work adopted by the public authorities which assimilate inability to work on account
of pregnancy and confinement to inability to work on account of illness.
In its judgment of the same date in Handels- og Kontorfunktionaerernes Foribund i Danmark v Dansk
Arbejdsgiverforening Case C-179/88 [1990] ECR I-3979 (para 13) the court stated:
It follows from the provisions of the Directive [arts 2(1)(3) and 5(1)] that the dismissal of a
female worker on account of pregnancy constitutes 392 direct discrimination on grounds of sex, as
is a refusal to appoint a pregnant woman.
25.
A few years later, in Habermann-Beltermann [1994] ECR I-1657 (para 26) the court confirmed that
case law and held that
Article 2 (1), read in conjunction with Articles 3(1) and 5(1) of Directive 76/207/EEC, precludes
an employment contract for an indefinite period for the performance of night-time work concluded
between an employer and a pregnant employee, both of whom were unaware of the pregnancy,
from being held to be void on account of the statutory prohibition on night-time work which applies,
by virtue of national law, during pregnancy and breastfeeding, or from being avoided by the
employer on account of a mistake on his part as to the essential personal characteristics of the
woman at the time when the contract was concluded.
26.
In its judgment in Webb [1994] 4 All ER 115, [1994] ECR I-3567 (paras 2426), the court refused to
comply with the national courts request that it consider whether the situation of a woman who finds
herself incapable, by reason of pregnancy discovered very shortly after the conclusion of an employment
contract, of performing the task for which she was recruited can be compared with that of a man similarly
incapable for medical or other reasons, since pregnancy is not in any way comparable with a pathological
condition, and even less so with unavailability for work on non-medical grounds. Further on, it added that

(26) dismissal of a pregnant woman recruited for an indefinite period cannot be justified on
grounds relating to her inability to fulfil a fundamental condition of her employment contract. The
availability of an employee is necessarily, for the employer, a precondition for the proper
performance of the employment contract. However, the protection afforded by Community law to a
woman during pregnancy and after childbirth cannot be dependent on whether her presence at
work during maternity is essential to the proper functioning of the undertaking in which she is
employed. Any contrary interpretation would render ineffective the provisions of Directive 76/207.
27.
It is clear from that case law that any decision by the employer which is directly related to the sex of a
female candidate, which affects the possibilities of access to a job or modifies the working conditions
offered to her as compared with those offered to a man, constitutes direct discrimination contrary to
Directive 76/207.
28.
I have no doubt whatsoever that the rule contained in Ch XIII of the service regulations annexed to the
collective agreement, under which any employee who has recorded a minimum of six months attendance
during the year must have an assessment of performance drawn up by his or her immediate superiors, as
a necessary, though not sufficient, precondition for receiving the advancement on merit of 2% of salary,
must be regarded as a working condition for the purposes of art 5(1) of the directive.
29.
Like the Commission, I note that that rule, which requires a minimum of six months attendance at
work in order for a right to an assessment of performance to arise, lays down a neutral criterion which
applies equally to men and women and is not bound, in principle, to affect either group adversely since
393 all are likely, for example, to be absent on account of illness. However, in reality, it is obvious that that
rule is liable to operate consistently to the disadvantage of women since it allows the employer to refuse
to assess a female employee whose absence during the year under consideration was due, in large
measure, to the fact that she took maternity leave.
30.
In fact, the way in which the CNAVTS applies that rule of the collective agreement means that any
female employee who takes the full period of maternity leave within the same year will find that, however
little she has been absent on account of illness, she will not have accumulated the six months attendance
necessary in order to establish the right to an assessment of performance, and it must also be borne in
mind that art 45 of the collective agreement now provides that, in certain circumstances, maternity leave,
which normally lasts sixteen weeks, may extend to 26 or 28 weeks. It goes without saying that, in the
latter event, merely because she had had a multiple birth, the woman would have no right to have the
quality of her work and her performance in the service of her employers during the rest of the year
assessed by her immediate superiors.
31.
I take the view that the right of all employees to have their work assessed annually is an integral part
of the conditions of their contracts of employment particularly since the existence or absence of such an
assessment has very practical consequences, such as, in the former case, the possibility of inclusion on
the list of recipients of an advancement on merit of 2% of salary if the marks awarded were sufficiently
high, or, in the latter case, denial of that possibility.
In those circumstances, it seems obvious to me that treating the period of maternity leave as a period
of sick leave has a direct adverse effect on the woman who has given birth, since the risk which she runs
on that account of receiving no annual assessment of performance is much greater than that run by a
man for whom only sick leave can be counted in relation to the same period.
32.
I conclude that, by being applied equally to different situations, the rule in the collective agreement
produces discriminatory effects. Consequently, in order to achieve the equal treatment sought, it will be
necessary to treat unequally that which is, in fact, different.
33.
For that reason, since the substantive equality between men and women as regards employment
precludes any consideration, either when they take up employment or during the employment
relationship, of a factor which by definition affects only women 5, it will not be possible, when calculating
the attendance at work of a female employee in order to establish a right to an assessment of
performance, to add the period of maternity leave either to periods of sick leave or to absences justified
for any other reason.
5
See the opinions of Advocate General Darmon in Dekker [1990] ECR I-3941 (para 26) and Advocate General Tesauro in
Webb [1994] 4 All ER 115, [1994] ECR I-3567 (para 8).

34.
I must admit that, where immediate superiors are required to comment on certain aspects of an
employees performance over a particular period of time, the requirement of a minimum time of
attendance at work is logical enough and cannot be regarded as disproportionate to the result pursued,
namely, that they should have a reasonable period of time in order to be able to assess the employees
performance. However, there is nothing to suggest that that period must necessarily be six months,
without exception, and I am not convinced that it is impossible to comment if the period is shorter.
394
In any event, given that this case involves direct discrimination on grounds of sex, it is not appropriate
even to consider the existence of any justification.
35.
At the hearing, the United Kingdom argued in favour of the applicability to this case of the precedent
established by the court in its judgment in Gillespie v Northern Health and Social Services Board Case C-
342/93 [1996] All ER (EC) 284, [1996] ECR I-475 (para 17) according to which
 women taking maternity leave provided for by national legislation are in a special position
which requires them to be afforded special protection, but which is not comparable either with that
of a man or with that of a woman actually at work.
It infers from that statement that Mrs Thibault, not having been at work for at least six months, is not
entitled to have an assessment of performance drawn up for her in relation to 1983 since otherwise she
would receive the same treatment as a man or woman who had been working.
36.
I disagree with that argument for two reasons. First, because the Gillespie case concerned application
of the principle that men and woman should receive equal pay for equal work, laid down by art 119 of the
EC Treaty and developed in Council Directive (EEC) 75/117 on the approximation of the laws of the
member states relating to the application of the principle of equal pay for men and women (OJ 1975 L45 p
19), and the court held that that principle does not require that women should continue to receive full pay
during maternity leave, a conclusion which appears logical in view of the fact that women on maternity
leave are not working.
However, the issue in this case is the application of the principle of equal treatment for men and
women as regards working conditions, and the court has stated, in the judgment in Gillespie [1996] All ER
(EC) 284, [1996] ECR I-475 (para 24), that Directive 76/207, as is clear from the second recital in its
preamble, does not apply to the principle of equal pay.
37.
Secondly, even if I took the view that the precedent set by Gillespie was applicable to this case, I
would still not agree with the conclusions which the United Kingdom draws from it. In fact, the court also
stated in that judgment that, since the benefit paid during maternity leave is calculated on the basis of the
average pay received by the woman while she was actually at work, the principle of non-discrimination
requires that she benefit from any pay rise, even if backdated, which is awarded between the period
covered by the reference pay and the end of maternity leave, adding that to deny such an increase to a
woman on maternity leave would discriminate against her purely in her capacity as a worker since, had
she not been pregnant, she would have received the pay rise.
In my opinion, if it were assumed that the courts judgment in Gillespie were applicable to this case, it
would serve to confirm that the act of equating maternity leave with sick leave, when calculating
attendance at work for the purpose of determining whether a woman is entitled to an assessment of
performance, is discriminatory. In fact, one could take up the courts wording and, adapting it to this case,
state that to deny a woman the right to an assessment of performance because she has been absent
from work on maternity leave would discriminate against her purely in her capacity as a worker since, had
she not been pregnant and given birth, she could not have been denied that right.
38.
For all the reasons which I have just set out, I propose that the court state in answer to the question
referred to it that art 5(1) of Directive 76/207 must be interpreted as precluding a neutrally worded national
provision contained in a 395 collective agreement, under which an employee who proves at least six
months attendance at work must be the subject of an assessment of performance by his or her
immediate superiors, but which, when applied in practice, produces direct discrimination on grounds of
sex in so far as it allows maternity leave to be counted as sick leave for the purpose of calculating the
time spent at work by a female employee.
39.
Before concluding, let me make one final point. The French government, in its observations, maintains
that the discrimination which Mrs Thibault has suffered arises neither from the relevant legislation (namely
art L 123-1 of the Labour Code which incorporates Directive 76/207 into national law) nor from the
applicable rules of the collective agreement and internal rules, but from the interpretation which the
CNAVTS has placed on those provisions.
40.
I am only partly in agreement with that view. Although those rules do not oblige the employer to count
maternity leave as absence from work for purposes of establishing the right to an assessment of
performance, they likewise do not prohibit him from so doing, thereby in fact allowing the interpretation
which the CNAVTS adopted in Mrs Thibaults case.
41.
Article 5(2)(b) of Directive 76/207 requires member states to ensure that any provisions contrary to the
principle of equal treatment which are included in, inter alia, collective agreements and internal rules of
undertakings shall be, or may be declared, null and void or may be amended. In my opinion, despite the
fact that art L 123-1 of the Labour Code has correctly transposed Directive 76/207 into national law, the
retention in the service regulations annexed to the collective agreement of the rule at issue creates for
individuals an ambiguous state of affairs by maintaining a state of uncertainty as to the possibilities
available to them of relying on Community law (see the judgment in EEC Commission v France Case
167/73 [1974] ECR 359 (para 41)).
That is clearly borne out by the fact that both Mrs Thibault, in the proceedings before the national
courts, and the French government and the Commission, both in their written observations and in the oral
observations submitted in these preliminary ruling proceedings, have had to rely, in order to show that
national law did not allow the interpretation adopted by the CNAVTS, on the one hand, on art L 122-26-2
of the Labour Code, which applies only for purposes of calculating length of service and, on the other, on
art 3bis of the agreement of 13 November 1975 appended to the collective agreement, which applies only
for the purpose of classifying posts. In any event, the latter provision did not enter into force until January
1984, that is, later than the year under consideration.
42.
In those circumstances, it is appropriate to refer to the courts case law according to which when it
interprets and applies national law, every national court must assume that the State had the intention of
fulfilling entirely the obligations arising from the directive concerned (see the judgment in Wagner Miret v
Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911 (para 20)). Moreover, the court has also
stated that
in applying national law, whether the provisions in question were adopted before or after the
directive, the national court called upon to interpret it is required to do so, as far as possible, in the
light of the wording and the purpose of the directive in order to achieve the result pursued by the
latter and thereby comply with the third paragraph of Article 189 of the Treaty. (See the judgment in
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135
(para 8).)
396

Conclusion
43.
In the light of the foregoing considerations, I propose that the question referred for a preliminary ruling
should be answered as follows:
Article 5(1) of Council Directive (EEC) 76/207 on the application of the principle of equal
treatment for men and women as regards access to employment, vocational training and
promotion, and working conditions must be interpreted as meaning that it precludes a neutrally
worded national provision contained in a collective agreement, under which an employee who
proves at least six months attendance at work must be the subject of an assessment of
performance by his or her immediate superiors, but which, when applied in practice, produces
direct discrimination on grounds of sex in so far as it allows maternity leave to be counted as sick
leave for the purpose of calculating the time spent at work by a female employee.

30 April 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By order of 28 March 1995, received at the Court of Justice of the European Communities on 28 April
1995, the French Cour de Cassation (the Court of Cassation) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty a question on the interpretation of Council Directive
(EEC) 76/207 on the implementation of the principle of equal treatment for men and women as regards
access to employment, vocational training and promotion, and working conditions (OJ 1976 L39 p 40).
2.
That question was raised in proceedings between the Caisse Nationale dAssurance Vieillesse des
Travailleurs Salaris (the CNAVTS) (the National Old-Age Insurance Fund for Employees) and Mrs
Thibault concerning the refusal by the CNAVTS to undertake an assessment of Mrs Thibaults
performance for 1983.
3.
Article 1(1) of the directive states that its purpose is to put into effect in the member states the principle
of equal treatment for men and women as regards access to employment, including promotion, and to
vocational training and as regards working conditions. That principle is known as the principle of equal
treatment. Article 2(1) of the directive defines that principle as meaning that there is to be no
discrimination whatsoever on grounds of sex either directly or indirectly by reference in particular to
marital or family status. Article 2(3) provides that the directive is to be without prejudice to provisions
concerning the protection of women, particularly as regards pregnancy and maternity.
4.
According to art 2(4), the directive is to be without prejudice to measures to promote equal opportunity
for men and women, in particular by removing existing inequalities which affect womens opportunities in
the areas referred to in art 1(1).
5.
Article 5(1) of Directive 76/207 provides that
Application of the principle of equal treatment with regard to working conditions, including the
conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex.
6.
Under art L 123-1(c) of the French Code du Travail (the Labour Code):
397
Subject to the special provisions of this code and save where the sex of the worker is an
essential condition for the performance of the duties attached to a post or an occupation: (c) no
measure may be adopted on grounds of sex, particularly in regard to remuneration, training,
assignment, qualification, classification, promotion or transfer.
7.
In France, under art 45 of the Convention Collective Nationale du Travail du Personnel des
Organismes de Scurit Sociale (the collective national labour agreement for the staff of social security
institutions), pregnant employees who have completed a minimum period of work with an institution are
entitled to 16 weeks maternity leave on full pay, and that period may be extended to 28 weeks. Under art
46 of the collective agreement, an employee may, on the expiry of her maternity leave, claim leave of
three months on half pay or leave of one-and-a-half months on full pay.
8.
Under art L 122-26-2 of the Code du Travail the period of maternity leave is to be treated as a period
of actual work for the purpose of determining a workers rights by virtue of length of service.
9.
Article 3 of the supplement of 13 November 1975 to the collective agreement provides that account
must be taken as professional experience for the purpose of classifying posts not only of actual
attendance at work but also of certain absences such as annual leave, movable holidays and special
leave, short-term leave, time spent as a trade union official and various other absences not exceeding five
working days in each six-month period. Article 3bis, added to the collective agreement by a supplement of
15 December 1983, provides that maternity leave must be taken into account as professional experience
on the same basis as the absences listed in art 3.
10.
Articles 29 to 31 of the collective agreement lay down the procedure for career advancement of
employees, which may amount to a maximum of 40% of their salary. Thus, under art 29 of the collective
agreement, upon the expiry of the second year after their entry into service employees are granted yearly,
by way of advancement based solely on length of service, a supplement of 2% of their salary. After the
third year and subject to a maximum of 24%, advancement under the collective agreement may rise from
2% to 4%, the additional 2% being based on the assessment made each year by the employees
superiors of their work and conduct. Between 24% and 40%, advancement under the collective
agreement is achieved at the rate of 2% per year.
11.
Chapter XIII of the CNAVTS standard service regulations amplifies arts 29 to 31 of the collective
agreement. As regards the discretionary advancement of 2%, it provides that any employee who has
been present at work for at least six months of the year must be the subject of an assessment of
performance by his superiors.
12.
Mrs Thibault was recruited by the CNAVTS in 1973 as an agent technique (skilled clerical worker) and
was promoted to rdacteur juridique (official responsible for legal drafting) in 1983. In that year, Mrs
Thibault was absent on account of sickness from 4 to 13 February, from 3 to 16 March and from 16 May
to 12 June. She then took maternity leave from 13 June to 1 October 1983, under art 45 of the collective
agreement, followed by maternity leave on half pay from 3 October to 16 November 1983 under art 46 of
the collective agreement.
13.
On the basis of Ch XIII of the standard service regulations, the CNAVTS refused to carry out an
assessment of performance for Mrs Thibault for 1983. In 398 its view, because of her absences, Mrs
Thibault did not meet the conditions laid down by that provision, namely six months presence at work.
14.
It is clear from the documents before the court that, in 1983, Mrs Thibault was at work for a period of
about five months. If she had not taken maternity leave between 13 June and 1 October 1983, she could
have relied on the six months attendance necessary for an assessment of performance under Ch XIII.
15.
Mrs Thibault then brought the matter before the Conseil de Prudhommes (the Labour Tribunal), Paris,
claiming that the failure to assess her performance, because of her absence on maternity leave,
constituted discrimination and that she had as a result lost an opportunity for promotion. By judgment of
17 December 1985, Mrs Thibaults claim was upheld and her employer was ordered to compensate her
for the loss she had suffered. The CNAVTS appealed against that decision.
16.
On 9 February 1989 the Cour de Cassation set that judgment aside on the ground that art 31 of the
collective agreement does not provide for inclusion as of right on the list of CNAVTS employees eligible
for advancement and referred the case to the Conseil de Prudhommes, Melun.
17.
By judgment of 24 January 1990, the Conseil de Prudhommes, Melun, held that the fact that Mrs
Thibaults performance had not been assessed deprived her of an opportunity for promotion. It considered
that her absence on account of maternity leave should have been treated as actual attendance at work
and that failure to take account of that absence constituted discrimination prohibited by art L 123-1(c) of
the Code du Travail. The Conseil de Prudhommes accordingly held that Mrs Thibault should have had
her performance assessed for 1983 and that she had missed an opportunity for promotion. The CNAVTS
was therefore ordered to award her back pay for 1984.
18.
The CNAVTS appealed against that judgment, contending that art 31 of the collective agreement does
not provide for automatic inclusion on the list for advancement on merit of employees who meet the
conditions laid down, that the period of professional experience prescribed by the collective agreement
should be severed from the period of actual attendance at work to be taken into account in order for an
employee to be eligible for an assessment of performance and that failure to assess Mrs Thibaults
performance was not based on grounds of sex, the principle of equality at work being applicable only to
rights potentially available to employees of both sexes, in accordance with art L 123-1(c) of the Code du
Travail.
19.
Since art L 123-1(c) of the Code du Travail transposes the directive into French law, the Cour de
Cassation decided to stay proceedings pending a preliminary ruling from the Court of Justice as to
whether
Articles 1(1), 2(1), 5(1) and, if relevant, 2(4) of Council Directive 76/207/EEC of 9 February
1976 must be interpreted as meaning that a woman may not be deprived of the right to an
assessment of performance, and consequently to the possibility of an advancement in career, on
the ground that she was absent from work by reason of maternity leave.
20.
According to the French government, the discrimination suffered by Mrs Thibault derives not from the
national legislation but from the interpretation thereof by the CNAVTS. It considers that art 3 of the
supplement to the collective agreement of 13 November 1975 and art 3bis of the supplement of 15
December 1983, although concerned with classification of the posts of staff of social security institutions
and not the arrangements for assessment of 399 performance referred to in art 31 of the collective
agreement, incontestably reflect the concern of the social partners not to discriminate against women on
maternity leave.
21.
On this point it should be recalled that, in accordance with the allocation of functions between the
Court of Justice and the national courts which underlies art 177 of the EC Treaty, it is for the national court
to establish the facts giving rise to the dispute, to interpret national legal provisions and to rule on their
application to the particular case (see, to that effect, Kempf v Staatssecretaris van Justitie Case 139/85
[1986] ECR 1741 (para 12) and Sinatra v Fonds national de retraite des ouvriers mineurs (FNROM) Case
296/84 [1986] ECR 1047 (para 11)).
22.
However, as the French government has stated, it is for the national court, within the limits of its
discretion under national law, to interpret and apply the law adopted to implement a directive in
accordance with the requirements of Community law (see the judgment in Von Colson v Land Nordrhein-
Westfalen Case 14/83 [1984] ECR 1891 (para 28)).
23.
Nevertheless, it must be emphasised that the national court has asked the Court of Justice to interpret
specific provisions of Community law in circumstances such as those of the case before it. Accordingly,
the observations of the French government regarding art 3 of the supplement of 13 November 1975 and
art 3bis of the supplement of 15 December 1983 are irrelevant.
24.
It must be borne in mind that the directive allows national provisions which guarantee women specific
rights on account of pregnancy and maternity, such as maternity leave (see Handels- og
Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiverforening Case C-179/88 [1990] ECR I-
3979 (para 15)).
25.
Furthermore, by reserving to member states the right to retain or introduce provisions which are
intended to protect women in connection with pregnancy and maternity, art 2(3) of the directive
recognises the legitimacy, in terms of the principle of equal treatment, first, of protecting a womans
biological condition during and after pregnancy and, second, of protecting the special relationship
between a woman and her child over the period which follows pregnancy and childbirth (see esp the
judgments in Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047 (para 25), Habermann-
Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb./Opf-eV Case C-421/92 [1994] ECR I-1657 (para
21) and Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115, [1994] ECR I-3567 (para
20)).
26.
The conferral of such rights, recognised by the directive, is intended to ensure implementation of the
principle of equal treatment for men and women regarding both access to employment (art 3(1)) and
working conditions (art 5(1)). Therefore, the exercise of the rights conferred on women under art 2(3)
cannot be the subject of unfavourable treatment regarding their access to employment or their working
conditions. In that light, the result pursued by the directive is substantive, not formal, equality.
27.
The right of any employee to have their performance assessed each year and, consequently, to qualify
for promotion, forms an integral part of the conditions of their contract of employment within the meaning
of art 5(1) of the directive.
28.
It is therefore in the light of art 5(1) of the directive, in conjunction with art 2(3), that rules such as
those at issue in this case must be examined to determine whether they guarantee men and women the
same conditions without discrimination on grounds of sex.
400
29.
The principle of non-discrimination requires that a woman who continues to be bound to her employer
by her contract of employment during maternity leave should not be deprived of the benefit of working
conditions which apply to both men and women and are the result of that employment relationship. In
circumstances such as those of this case, to deny a female employee the right to have her performance
assessed annually would discriminate against her merely in her capacity as a worker because, if she had
not been pregnant and had not taken the maternity leave to which she was entitled, she would have been
assessed for the year in question and could therefore have qualified for promotion.
30.
It is true, as the UK government was right to point out, that the court has recognised that the member
states have a discretion as to the social measures they adopt in order to guarantee, within the framework
laid down by the directive, protection of women in connection with pregnancy and maternity and as to the
nature of the protection measures and the detailed arrangements for their implementation (see, inter alia,
the judgment in Hofmann [1984] ECR 3047 (para 27)).
31.
Nevertheless, such discretion, which must be exercised within the bounds of the directive, cannot
serve as a basis for unfavourable treatment of a woman regarding her working conditions.
32.
It must therefore be held that a woman who is accorded unfavourable treatment regarding her working
conditions, in that she is deprived of the right to an annual assessment of her performance and, therefore,
of the opportunity of qualifying for promotion as a result of absence on account of maternity leave, is
discriminated against on grounds of her pregnancy and her maternity leave. Such conduct constitutes
discrimination based directly on grounds of sex within the meaning of the directive.
33.
The answer to the question must therefore be that arts 2(3) and 5(1) of the directive preclude national
rules which deprive a woman of the right to an assessment of her performance and, consequently, to the
possibility of qualifying for promotion because she was absent from the undertaking on account of
maternity leave.

Costs
34.
The costs incurred by the French and UK governments and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the proceedings pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the question referred to it by the
French Cour de Cassation by judgment of 28 March 1995, hereby rules: arts 2(3) and 5(1) of Council
Directive (EEC) 76/207 on the implementation of the principle of equal treatment for men and women as
regards access to employment, vocational training and promotion, and working conditions, preclude
national rules which deprive a woman of the right to an assessment of her performance and,
consequently, to the possibility of qualifying for promotion because she was absent from the undertaking
on account of maternity leave.

401

[1998] All ER (EC) 402

R v Customs and Excise Comrs, ex parte EMU Tabac SARL and others
(Imperial Tobacco intervening)
(Case C-296/95)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), RAGNEMALM , WATHELET, SCHINTGEN
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, MURRAY, PUISSOCHET
(RAPPORTEUR), HIRSCH, JANN AND SEVN ADVOCATE GENERAL RUIZ-JARABO COLOMER
4 MARCH, 17 APRIL 1997, 2 APRIL 1998
European Community Customs and excise Duties Goods already released for consumption
Company soliciting orders for tobacco and cigarettes from consumers in United Kingdom Company
purchasing goods in Luxembourg where lower duties on tobacco Purchases made for consumers
private use Company also arranging transport and delivery Goods detained by United Kingdom
customs authorities Member state in which duty payable Council Regulation (EEC) 92/12, arts 6, 7, 8,
10.

An English company, MBL, solicited and obtained orders for cigarettes and tobacco from private
individuals resident in the United Kingdom. MBL purchased the products from EMU, a company
established in Luxembourg, where excise rates on tobacco were generally lower, and arranged for their
importation into the United Kingdom by a private carrier in the name and on behalf of those individuals, in
return for a fee. During 1995 the Commissioners of Customs and Excise detained certain quantities of
those products when they were imported into Dover, as they were authorised to do under United Kingdom
legislation when excise duty was payable. MBL, EMU and one of MBLs directors applied for judicial
review of that detention; their application was refused and they subsequently appealed. The Court of
Appeal stayed the proceedings and referred to the Court of Justice of the European Communities for a
preliminary ruling questions concerning the interpretation of Council Directive (EEC) 92/12 on the general
arrangements for products subject to excise duty and on the holding, movement and monitoring of such
products and, in particular, art 81. Article 8 provided that where goods were held for personal use, excise
duty was payable in the member state in which they were purchased and no document was required
when they were transported to another member state, provided that the goods had been acquired by
private individuals for their own use and transported by them. The applicants contended: (i) that art 8
applied in a situation where purchase of goods chargeable to excise duty was effected through an agent
who also arranged for their transportation; and (ii) that if not, art 6 2 would have to be applied with the
result that excise duty would be payable in the country in which the goods were released for consumption,
ie Luxembourg in the instant case.
1
Article 8, so far as material, is set ou at p 419 j, post
2
Article 6, so far as material, is set ou at p 418 j to p 419 a, post

Held (1) On its true construction art 8 made no provision for the involvement of a third party and
therefore it was not applicable where the purchase and/or 402 transportation of goods subject to duty was
effected through an agent (see p 423 e f j to p 424 a, post).
(2) Although art 6 provided that duty would be chargeable at the time when goods were released for
consumption in a member state, that did not preclude excise duty from being subsequently levied in
another member state pursuant to arts 73, 94 or 105, whereupon duty paid in the first member state could
be reimbursed. The situation under consideration fell within the scope of arts 7 and 10, which provided
respectively that products already released for consumption in one member state became chargeable to
excise duty in another member state to which they were exported where they were held, in the member
state of destination, for the purposes of a trader carrying out an economic activity (art 7(2)) and where
they were dispatched by or on behalf of the vendor (art 10(2)). It followed that where goods from one
member state were carried to another member state on the instructions of a trader acting in return for
payment who had previously solicited customers in the latter state and had arranged for importation of the
goods, excise duty was chargeable in that latter state (see p 424 e to g and p 425 d to f, post).
3
Article 7, so far as material, is set ou at p 419 d to h, post
4
Article 9, so far as material, is set ou at p 420 a to e, post
5
Article 10, so far as material, is set ou at p 420 g to p 421 a, post

Cases cited
EC Commission v France Case C-30/89 [1990] ECR I-691.
Hydrotherm Gertebau GmbH v Compact del Dott Ing Mario Andreoli & C sas Case 170/93 [1984] ECR
2999.
Merck (Firma E) v Hauptzollamt Hamburg-Jonas Case 292/82 [1983] ECR 3781.
Nicolaus Corman & Fils SA v Hauptzollamt Gronau Case 64/81 [1982] ECR 13.
Wrsdorfer, ne Koschniske v Raad van Arbeid Case 9/79 [1979] ECR 2717.

Reference
By order of 31 July 1995, the Court of Appeal referred to the Court of Justice of the European
Communities for a preliminary ruling pursuant to art 177 of the EC Treaty two questions (set out at p 422
b to d, post) concerning the interpretation of Council Directive (EEC) 92/12 on the general arrangements
for products subject to excise duty and on the holding, movement and monitoring of such products, as
amended. The questions arose in proceedings between EMU Tobacco SARL, The Man in Black Ltd and
John Cunningham, a director of that company, and the Commissioners of Customs and Excise
concerning the charging of excise duty on cigarettes. Written observations were submitted on behalf of:
EMU Tabac SARL, The Man in Black Ltd and Mr Cunningham, by R Venables QC, T Lyons and A Hardy,
Barristers; Imperial Tobacco Ltd, by D Vaughan QC, and M Brealey, Barrister; the UK government, by L
Nicoll, of the Treasury Solicitors Department, acting as agent, and by S Richards and R Jay, Barristers;
the German government, by E Rder, Ministerialrat in the Federal Ministry of Economic Affairs, and B
Kloke, Oberregierungsrat in that ministry, acting as agents; the Greek government, by F Georgakopoulos,
Assistant Legal Adviser in the State Legal Service, and G Alexaki, Adviser in the Special Community
Legal Service of the Ministry of Foreign Affairs, acting as agents; the French government, by C de Salins,
Head of Sub-directorate in the Legal Affairs Directorate of the Ministry of Foreign Affairs, and F Pascal,
Charg de Mission in 403 the same Ministry acting as agents; the Irish government, by M A Buckley,
Chief State Solicitor, acting as agent, M Collins, SC, and J Payne, Barrister-at-Law; the Italian
government, by Professor U Leanza, Head of the Contentious Diplomatic Affairs Department of the
Ministry of Foreign Affairs, acting as agent, assisted by I Braguglia, Avvocato dello Stato, acting as agent;
the Netherlands government, by A Bos, Legal Adviser at the Ministry of Foreign Affairs, acting as agent;
the Austrian government, by F Cede, Ambassador, acting as agent; the Swedish government, by E
Brattgrd, Departementsrd in the Foreign Trade Department of the Ministry of Foreign Affairs, acting as
agent; and the European Commission, by E Traversa and P Oliver, of its Legal Service, acting as agents.
Oral observations were made by: EMU Tabac SARL, The Man in Black Limited and Mr Cunningham,
represented by R Venables, T Lyons and A Hardy; Imperial Tobacco Ltd, represented by D Vaughan and
M Brealey; the UK government, represented by L Nicoll, S Richards and R Jay; the Danish government,
represented by P Biering, Legal Adviser, Head of Department at the Ministry of Foreign Affairs, acting as
agent; the Greek government, represented by F Georgakopoulos; the French government, represented
by F Pascal; the Irish government, represented by A OCaoimh, SC, and N Hyland, Barrister-at-Law; the
Italian government, represented by I Braguglia; the Netherlands government, represented by M Fierstra,
Assistant Legal Adviser at the Ministry of Foreign Affairs, acting as agent; the Finnish government,
represented by T Pynn, Legal Adviser at the Ministry of Foreign Affairs, acting as agent; the Swedish
government, represented by E Brattgrd; and the Commission, represented by E Traversa and P Oliver.
The language of the case was English. The facts are set out in the opinion of the Advocate General.

17 April 1997.

The Advocate General (D Ruiz-Jarabo Colomer)

delivered the following opinion (translated from the Spanish).
1.
The Court of Appeal has referred to the Court of Justice of the European Communities for a
preliminary ruling two questions on the interpretation of Council Directive (EEC) 92/121 on the general
arrangements for products subject to excise duty and on the holding, movement and monitoring of such
products (OJ 1992 L76 p 1).
2.
The Court of Appeal wishes to know in which country the excise duty is chargeable when an individual
purchases, for his own use, goods subject to duty (specifically, tobacco products), the purchase takes
place in state B, from which the goods are transported to state A, where the individual concerned lives,
and both operations involve an agent acting for him.
3.
That is the position in this case, in which, as graphically stated by the national court, the appellants
have set up and participated in a scheme which enables a UK resident, without leaving the comfort of his
armchair, to obtain in this country tobacco which he has purchased in a shop in Luxembourg. In that way
he avoids payment of the excise duty applicable in the United Kingdom, which is heavier than that
payable in Luxembourg.

THE FACTS
4.
The appellants in the main proceedings are two companies and a natural person. The first company,
EMU Tabac Srl (the retailer) is a retailer of tobacco products in Luxembourg. The second, The Man in
Black Ltd (the agent), acts as agent for any buyers living in the United Kingdom. Mr Cunningham is a
director of the latter company.
404
5.
Both those companies are in turn subsidiaries of another company (The Enlightened Tobacco
Company). The relationship between the two subsidiaries is of a commercial nature and each has
different directors. Almost all EMU Tabac Srls sales are made in response to requests for specific
quantities of cigarettes or tobacco from private individuals living in the United Kingdom, although it also
sells to individuals living in Luxembourg.
6.
The system devised and implemented by the two companies is as follows: (a) The seller buys
cigarettes of a specified brand, under licence. (b) The seller and the agent have a commercial agreement
under which the first company has opened and maintains a credit account in the name of the second
covering all the purchases made by customers through the agent. The latter undertakes to transfer the
amount of the payments from customers directly into bank accounts in London or Luxembourg, having
furnished a guarantee to that effect. (c) The sale of the products to the customers follows a procedure
which both parties describe in the following terms:
(1) The offer to purchase products from the retailer is made by the customer. (2) The agent
shall receive and communicate that offer to the retailer in Luxembourg acting as agent for the
customer. (3) At its discretion, the retailer will accept that offer to purchase and will communicate its
acceptance to the agent (as agent of the customer) by delivering the products together with a
receipted copy of the agents offer summary document. (4) The contract for the sale and purchase
of the products will be made in Luxembourg and title in those products will pass in Luxembourg.
The contract will be subject to the law of England.
(d) The agent deals with transport of the goods for the customer, under contracts with a transport
undertaking: the contract of carriage is signed by the agent as agent for the customer. (e) Since the sales
are to individuals, no customer may buy more than 800 cigarettes per order. If more cigarettes are
requested in an order or a large number is ordered in a single month, the agent sends the customer a
printed form refusing to supply the goods, as they must be solely for personal use.
7.
 According to the order for reference, what happens in practice is as follows: (a) The individual
customer obtains an order form for cigarettes, fills it in and sends it to the agent, with a cheque for the
requisite amount. (b) Once the cheque has been paid into the agents customer account in London, that
fact is notified to the agents office in Luxembourg, together with details of the customers name and
address. (c) An employee of the agent goes to the retailers office (which is in the same building) and the
latter, having accepted the offer, hands over the goods. (d) The agent then packages the goods, which
are collected by the carrier and sent to the United Kingdom where they are delivered to the customer. (e)
Finally, the agent pays the retailer and the carrier out of a current account to which has previously been
transferred from the agents account in London a sum equivalent to the price of the goods and the cost of
transport. The remainder of what has been paid by the purchaser to the agent represents the agents
commission.

THE MAIN PROCEEDINGS AND THE PRELIMINARY QUESTIONS

8.
The UK tax administration (the Commissioners for Customs and Excise) ordered the seizure, in Dover,
of a quantity of tobacco imported under the 405 system described above and required the payment on
those goods of UK excise duty.
9.
The now appellants applied to the Queens Bench Division of the High Court for a declaration that the
UK excise duty was not payable, that the seizure of the tobacco was unlawful and that the UK tax
administration should refrain from seizing tobacco imported under the scheme.
10.
Those applications were dismissed by the High Court, whereupon the appellants brought the matter
before the Court of Appeal, which decided to stay the proceedings and seek a preliminary ruling from the
Court of Justice on the following questions:
(1) Does Directive 92/12/EEC and in particular Article 8 have the effect of precluding the
charging of excise duty on goods in a Member State A in circumstances where: (i) the goods were
acquired for the use of a private individual in Member State A; (ii) they were acquired in Member
State B by an agent acting on behalf of the private individual; (iii) transportation of the goods from
Member State B to Member State A was arranged by the agent; and (iv) the individual did not
himself travel with the goods from Member State B to A.
(2) Where a scheme has been commercially devised and marketed whereunder purchases
made in Member State B for the personal use of a private individual in Member State A are made
by an agent for that individual and those purchases are transported from Member State B to
Member State A as a result of arrangements made by such an agent, does Directive 92/12/EEC
have the effect of precluding the charging of excise duty on those purchases in Member State A?

The applicable Community provisions

11.
Excise duty is an indirect tax levied on certain goods such as tobacco products, alcoholic beverages
and hydrocarbons. The rates of duty applied by each member state are different, harmonisation thereof
not having so far been achieved.
12.
Products manufactured in, or imported into, the Community are subject to excise duty. Normally,
payment of the tax is deferred until the goods are released for consumption. Accordingly, the collection of
excise duty is in principle a matter for the member state where the goods are consumed.
13.
Before the creation of the single market, goods subject to excise duty could be held in a warehouse, or
move from one warehouse to another in various member states which were parties to the suspension
arrangement6, subject to border controls. Such controls are no longer possible since the advent of the
single market.
6
The suspension arrangement applies to the manufacture, processing, holding and movement of products subject to
excise duty, and under it, although the chargeable event has occurred, the duty is not yet chargeable.

14.
The directive lays down general conditions applicable to products subject to excise duty and makes
specific provision for the holding and movement of them and for the controls needed to ensure that tax
debts are discharged.
15.
Article 6 of the directive provides that, as a general rule, the duty becomes chargeable when the
goods are released for consumption7, the rates of duty in 406 force at the material time being applied in
the member state of release for consumption.
7
According to art 6, the duty also becomes chargeable when it is found that there are losses or differences which ought
to be subject to excise duty in accordance with art 14(3). In any event, release for consumption of goods subject to
excise duty is deemed to mean: (a) any departure, including irregular departure, from a suspension arrangement; (h)
any manufacture, including irregular manufacture, of the products concerned outside a suspension arrangement; (c) any
importation of those products, including irregular importation, where they have not been placed under a suspension
arrangement.
16.
Article 7 of the directive provides that, where products subject to excise duty already released for
consumption in a member state are held for commercial purposes in another member state, the excise
duty is to be levied in that other member state and the tax will be repayable by the member state from
which they came. The wording of the provision8 is as follows:
8
As amended by Council Directive (EEC) 92/108 amending Council Directive (EEC) 92/12 on the general arrangements
for products subject to excise duty and on the holding, movement and monitoring of products subject to excise duty (OJ
1992 L390 p 124).

(1) In the event of products subject to excise duty and already released for consumption in one
Member State being held for commercial purposes in another Member State, the excise duty shall
be levied in the Member State in which those products are held. (2) To that end, without prejudice
to Article 6, where products already released for consumption as defined in Article 6 in one Member
State are delivered, or intended for delivery in another Member State or used in another Member
State for the purposes of a trader carrying out an economic activity independently or for the
purposes of a body governed by public law, excise duty shall become chargeable in that other
Member State. (3) Depending on all the circumstances, the duty shall be due from the person
making the delivery or holding the products intended for delivery or from the person receiving the
products for use in a Member State other than the one where the products have already been
released for consumption, or from the relevant trader or body governed by public law. (4) The
products referred to in paragraph 1 shall move between the territories of the various Member
States under cover of an accompanying document listing the main data from the document referred
to in Article 18(1). The form and content of this document shall be established in accordance with
the procedure laid down in Article 24 of this Directive. (5) The person, trader or body referred to in
paragraph 3 must comply with the following requirements: (a) before the goods are dispatched,
make a declaration to the tax authorities of the Member State of destination and guarantee the
payment of the excise duty; (b) pay the excise duty of the Member State of destination in
accordance with the procedure paid down by that Member State; (c) consent to any check enabling
the administration of the Member State of destination to satisfy itself that the goods have actually
been received and that the excise duty to which they are liable has been paid. (6) The excise duty
paid in the first Member State referred to in paragraph 1 shall be reimbursed in accordance with
Article 22(3).
17.
Article 8 of the directive provides:
As regards products acquired by private individuals for their own use and transported by them,
the principle governing the internal market lays down 407 that excise duty shall be charged in the
Member State in which they are acquired.
18.
Article 9 of the directive lays down rules for products subject to excise duty where they are intended to
be used for commercial purposes:
(1) Without prejudice to Articles 6, 7 and 8, excise duty shall become chargeable where
products for consumption in a Member State are held for commercial purpose in another Member
State. In this case, the duty shall be due in the Member State in whose territory the products are
and shall become chargeable to the holder of the products. (2) To establish that the products
referred to in Article 8 are intended for commercial purposes Member States must take account,
inter alia, of the following: the commercial status of the holder of the products and his reasons for
holding them, the place where the products are located or, if appropriate, the mode of transport
used, any document relating to the products, the nature of the products, the quantity of the
products. For the purposes of applying the content of the fifth indent of the first subparagraph,
Member States may lay down guide levels, solely as a form of evidence. These guide levels may
not be lower than
19.
Article 10 of the directive contains specific provisions for particular kinds of transactions involving the
dispatch or transport of goods subject to excise duty:
(1) Products subject to excise duty purchased by persons who are not authorized
warehousekeepers or registered or non-registered traders and dispatched or transported directly or
indirectly by the vendor or on his behalf shall be liable to excise duty in the Member State of
destination. For the purposes of this Article, Member State of destination shall mean the Member
State of arrival of the dispatch or transport. (2) To that end, the delivery of products subject to
excise duty already released for consumption in a Member State and giving rise to the dispatch or
transport of those products to a person as referred to in paragraph 1, established in another
Member State, and which are dispatched or transported directly or indirectly by the vendor or on his
behalf shall cause excise duty to be chargeable on those products in the Member State of
destination. (3) The duty of the Member State of destination shall be chargeable to the vendor at
the time of delivery. However, Member States may adopt provisions stipulating that the excise duty
shall be payable by a tax representative, other than the consignee of the products. Such tax
representative must be established in the Member State of destination and approved by the tax
authorities of that Member State. The Member State in which the vendor is established must
ensure that he complies with the following requirements: guarantee payment of excise duty under
the conditions set by the Member State of destination prior to dispatch of the products and ensure
 that the excise duty is paid following arrival of the products, keep accounts of deliveries of products.
(4) In the case referred to in paragraph 2, the excise duty paid in the first Member State shall be
reimbursed in accordance with Article 22(4).

The first preliminary question

20.
The national court links its first question to art 8 of the directive, for the application of which three
cumulative conditions must be met in order for 408 products to be subject to the excise duty in force in the
country of purchase: (a) the products must be acquired by private individuals; (b) the products must be
acquired by private individuals for their own use; (c) the products must be transported by the private
individuals.
21.
In the present case there is no problem concerning the second condition, the legal aspects of which
are clear: both the parties in dispute and the national court recognise that private individuals resident in
the United Kingdom acquire tobacco products for their personal use, not for commercial purposes. No
doubts appear to arise as to the application of the criteria laid down in that connection by art 9 of the
directive.
22.
The same does not apply to the other two conditions. The dispute between the parties has focused
particularly on the requirement of transport, which raises considerable interpretative difficulties, but in my
opinion the first condition too must he analysed and examined in relation to the scheme described by the
UK court in its order for reference.
(i) The meaning of the words acquisition by private individuals used in art 8 of the directive
23.
I shall first mention a point which, judging from the observations of some of the parties, does not seem
to have been taken sufficiently into account: the directive of which interpretation is sought is a tax
measure, intended to deal with matters of taxation, and to that end it uses tax-law classifications. By this I
mean that the terms used in the directive, on the basis of the corresponding concepts in civil or mercantile
law, may have a rather independent meaning which will not coincide in every respect with that of their
counterparts in civil or mercantile law.
24.
In that light, it is possible that the private individual purchaser, as a subject of civil or mercantile law,
may not be precisely the same as the person with which art 8 of the directive is concerned. That applies
to a specific aspect of the involvement of the private individual in the purchase contract referred to in the
directive, where it envisages the possibility of his acting, in entering into the contract or giving effect to it,
through a third party.
25.
A similar possibility is generally available in civil law for all types of acts-in-the-law by virtue of the
principle that whoever does something for himself may also do it through someone else. However, from
the taxation point of view, it may be appropriate to distinguish between purchases made by someone who
acts for himself and those by someone who acts through an agent or attorney.
26.
Specifically, I consider that the concept of a private individual, with whose purchases art 8 of the
directive is concerned (I repeat, for the sole purpose of regulating the territorial chargeability of excise
duty), is of more limited scope, so that it excludes the instrumentality of third parties. There are two main
reasons for my taking that view.
27.
In the first place, the private individual in this case must carry out sequentially a number of personal
operations in order to benefit from the tax provisions in the member state of purchase: as we shall see
shortly, he must travel to that member state and from there he, personally, must transport the goods
acquired. It seems to me to follow logically from those requirements that the purchase, as such, should
also be deemed to be limited to a purchase made by the buyer himself, not through an agent or
middleman.
28.
Secondly, the presence of persons other than the purchaser, in the intra- Community movement of
goods subject to excise duty, is carefully regulated by the directive. In addition to action by authorised
warehousekeepers, registered 409 and non-registered traders, the directive provides for a tax
representative, other than the consignee of the products (art 10(3)). If it had been intended that the
purchaser could be replaced, purely and simply, by any natural or legal person to whom he had granted a
power of attorney in order to benefit from the rule in art 8 when the goods cross the frontier of a member
state, that article would have contained a specific provision to that effect.
29.
Therefore, art 8 of the directive provides only for action by the private individual on his own behalf,
actively participating in a number of successive operations (travel to another member state, purchase of
products subject to excise duty and transport to his own country) which he must carry out for himself.
Transactions which do not fulfil those requirementsas they do not in this casecannot benefit from the
rule contained in that article.
30.
 All this becomes clearer upon analysisan analysis which I shall now undertakeof the third
precondition for application of the rule laid down in art 8 of the directive.
(ii) The concept of products transported, according to a literal interpretation of art 8 of the
directive
31.
The first sentence of art 8 of the directive raises two problems of interpretation relevant here. First, the
various language versions differ from each other significantly. Second, the interpretation of the very term
transport used in that provision is not without difficulties.
32.
The expression used in art 8 of the directive (products acquired by private individuals and
transported by them) has been translated in various ways in the various languages of the member states
which made up the Community when it was adopted.
33.
Certain language versions coincide in using the expression transported by them without any
additional words as in the ease of the Spanish and English versions 9. Others, however, reinforce the
pronoun them with an additional term: that is so in the French 10, Portuguese11, Italian12, Dutch13 and
German14 versions. Finally, the Greekand Danish15versions lay even greater stress on the personal
aspect by using terms equivalent to transport in person.
9
Transportados por ellos.
10
Transports par eux-mmes.
11
Transportados pelos prprios.
12
Trasportati dai medesimi
13
door hen zelf vervoerde produkten.
14
Waren, die Privatpersonen sich selbst befrden.
15
Og som de selv medforer.

In my opinion, a literal interpretation of the provision implies that the person who does the transporting
must be, specifically, the private individual purchaser, and not any other person. Even in the absence of
the more radical language versions, the intrinsic meaning of the term transported by them or
transported by themselves used in the remainder of the language versions refers to transport carried out
and not merely arrangedby the person concerned.
35.
It is true that one person may entrust the transport of goods to another or avail himself of the services
of an agent to do so. But such operations go beyond the limits of art 8 of the directive since the private
individual is not then transporting, as required by that article, but is arranging or contracting for the
transport or entrusting it to another person.
410
36.
Article 8 does not say products transported on the directions, or on behalf, of private individuals. Nor
does it refer to the action of dispatching, which implies using the services of a third party. However, both
forms of words are found in other provisions of the directive. (a) Thus, art 10(1) of the directive refers to
products dispatched or transported directly or indirectly by the vendor or on his behalf (my emphasis).
The same terms are used in art 10(2). They thus use legal formulae that are wider and more flexible than
that used in art 8. (b) For its part, art 9(3) provides for the hypothesis of atypical modes of transport by
private individuals or on their behalf (my emphasis). By contrast, art 8 contains no reference to the latter
possibility.
37.
It is particularly significant that art 8, unlike art 10, dispensed with the possibilityprovided for and
regulated by the latterof having the goods in question sent, that is to say forwarded or dispatched by
private individuals to their own countries: on the contrary, it lays down the imperative requirement that
they be transported by the person who acquired them.
38.
The literal wording of the provision thus refers to transport operations carried out by the private
individual and no-one else. This implies, in my view, that the private individual who buys the products
subject to duty has travelled to another country and carries them with him.
39.
The status of traveller which the private individual must possess, if not apparent from art 8, must be
inferred as a matter of logic. When art 26 of the directive laid down exceptional rules for Denmark (later
extended to Sweden and Finland in their respective Acts of Accession) it created a direct link between art
8 and specified quantitative limits enforceable against travellers who go to that country or, if they live
there, leave it.
40.
Article 26 of the directive16, as a temporary exception to the conditions laid down by art 8, allows the
member states to which the directive is addressed to apply restrictive measures to travellers who bring
into their territories products subject to excise duty acquired by them in another member state 17. All this,
as a matter of sound logic, postulates the same interpretation of art 8 as the one which I advocate,
namely that the cross-frontier traveller must have personally acquired and transported those products with
him.
 16
The original wording of art 26 was as follows: (1) Without prejudice to Article 8, until 31 December 1996 and subject to a
review mechanism similar to that laid down in Article 28 (2) of Directive 77/388/EEC (OJ 1977 L 145, p.6), Denmark
shall be authorised to apply the specific arrangements laid down in paragraphs 2 and 3 for spirit drinks and
manufactured tobaccos within the general framework of the approximation of excise duty rates. (2) Denmark shall be
authorised to apply the following quantitative restrictions: private travellers entering Denmark shall benefit from the
allowance in force on 31 December 1992 for cigarettes, cigarillos or smoking tobacco and for spirit drinks, travellers
resident in Denmark and having been outside Denmark for less than the period in force on 31 December 1992 shall
benefit from the allowances in force in Denmark on 31 December 1992 for cigarettes and spirit drinks. (3) Denmark may
collect excise duties and carry out the necessary checks with respect to spirit drinks, cigarettes, cigarillos and smoking
tobacco. This article has recently been replaced, as from 1 January 1977, by the corresponding text of Council Directive
(EC) 96/99 amending Directive (EEC) 92/12 on the general arrangements for products subject to excise duty and on the
holding, movement and monitoring of such products (OJ 1997 L 8, p 12).
17
According to the new text of art 26 of the directive, in force as from 1 January 1997, Denmark and Finland are
authorised to restrict the grant of admission without payment of other duty to persons who have been absent from their
territory for more than 24 hours.

41.
It is true that art 8 likewise contains no express references to the travellers personal luggage, of
which the products purchased should form part. However, that omission does not in my view provide a
basis for arguing that transport may 411 be undertaken on behalf of the private individual. That
expression appears in art 28 of the directive only in relation to travellers who take a flight or a sea voyage
within the Community, who may be granted an exemption from duty 18. That fact has nothing to do with art
8, under which purchasers necessarily have to pay the duty applicable in the country of acquisition of the
products.
18
The member states may grant to such travellers, until 30 June 1999, exemption from excise duty on products which they
have bought at particular sales outlets in ports or airports or on board the aircraft or vessel carrying them. The
exemption applies only to products in quantities not exceeding, per person and journey, the limits laid down in the
Community provisions in force for travel between non-member countries and the Community.

42.
In any event, it must not be forgotten that art 8 of the directive, in referring to all products subject to
excise duty, allows the buyer to transport some of them for himself without necessarily including them in
his personal luggage: thus, for example, petrol which he has purchased in the country of destination and
is in his vehicles tank does not form part of his personal luggage in the strict sense, but the rule in art 8
nevertheless continues to apply. This explains why, therefore, that article did not include the requirement
that all the products referred to have to be included in travellers personal luggage.
43.
Accordingly, I agree with the position taken by all the member states which intervened in this case,
except France, and I differ from that taken by the appellants in the main proceedings, which on this point
coincides only partly with the view taken by the European Commission and the French government 19.
19
The Commission considers that art 8 may be applied even if the goods, acquired personally by the purchaser, are not
carried by him personally but by a carrier designated by him. The French government for its part considers that art 8
applies both if transport of the dutiable goods is effected by the buyer and if it is effected on his behalf (provided that the
seller does not assume responsibility for the transport operation).

(iii) The concept of products transported in art 8, in the legislative context of the directive
44.
Support for the same interpretation of the term transported by them can be drawn from other
considerations concerning the internal sense of art 8 within the legislative context of the directive.
45.
As is acknowledged in its preamble, the directive has to fulfil at the same time two imperative
requirements: on the one hand, movement from the territory of one member state to that of another may
not give rise to checks liable to impede free movement within the Community; on the other, each state
must be assured that it will collect the excise duty at the rates approved by it and in accordance with the
common rules on chargeability: accordingly the latter make it necessary to know of the movements of
products subject to excise duty.
46.
The system of excise duty in force at Community level since 1993 necessitates, in addition to
minimum rates of charge, harmonised taxation structures and common rules on the holding and
movement within the Community of the products affected. The latter may take three specific forms: (a)
Commercial movement of products under the suspension arrangement; following the abolition of frontier
controls, the principle of taxation at destination was maintained by providing for a system of detailed
information and checks in which traders, registered or otherwise, and accompanying documents play a
key role. (b) Commercial transactions involving products that have already been taxed, accounting for a
small proportion of intra-Community trade in this kind of product 20, in which the information mechanisms,
for which the economic agents 412 concerned are responsible, and control by the member states, are
also regulated in detail. (c) Cross-frontier purchases by private individuals, of the kind referred to in art 8
of the directive. By contrast with the other two systems, this one entails no obligation to provide
information since the intervention of economic agents as intermediaries is excluded.
20
This is stated in the Commissions report to the EU Council and the European Parliament of 13 September 1995
submitted in accordance with art 4 of Council Directive (EEC) 92/79 on the approximation of rates for other products
subject to excise duty (COM(95) 285 Final).
47.
Within that legislative context, the inclusion of art 8 must be interpreted in the light of the other
provisions of the directive so as not to negate the effectiveness of the remaining articles.
48.
First, art 9 of the directive, concerning products held for commercial purposes, would be rendered
largely ineffective if a system of purchase and transport of the kind at issue in the main proceedings
became general practice. Within such a system, after abolition of internal frontiers, the member states
would in practice find it hard to evaluate the combined effect of the factors mentioned in that provision to
determine whether the products were being transported on a commercial or a merely private basis. For
example, they would encounter difficulties in verifying whether or not a given private individual, using
several agents simultaneously or successively, had exceeded the levels which in principle imply that his
purchases are of a private nature.
49.
The fact that, under the system at issue in the main proceedings, the appellant undertakings refuse to
accept orders from private individuals which exceed those levels indicates their intention not to engage in
fraud, but that is not a sufficient argument to render acceptable the incitement to fraud which, objectively,
would result from generalisation of the system.
50.
Secondly, the rules on distance sales contained in art 10, observing the principle of taxation in the
country of destination and not that of purchase, would in all probability be systematically evaded in sales
between countries with widely differing tax rates, where logically there is an incentive to buy products in
the country which levies the least tax.
51.
In fact, the art 10 rules expressly mention only distance sales of products dispatched or transported
directly or indirectly by the vendor or on his behalf. For the appellants, the scope of that rule does not
extend to a mechanism such as that at issue in the main proceedings since there it is the buyers and not
the retailers who contract, through an agent, for the transport service.
52.
If that interpretation were allowed, there would be a lacuna in the directive regarding distance
purchases since, although the system used by EMU Tabac may not conform with art 10, interpreted
literally, it does not thereby cease to be, economically and legally, a distance purchase. The most
appropriate legal way of supplying that lacuna is recourse to analogy: application to the circumstances of
this case of the legal provision applicable to the most similar situation.
53.
As I see it, the legal concept closest to the distance purchasing carried out, in the comfort of their
armchairs, by UK residents is, precisely, the distance selling governed by art 10 of the directive, from
both the economic and the legal points of view.
54.
Therefore, if it is the case that the Community legislature wished to adhere to the principle of taxation
at destination for the distance sales provided for in art 10, the same principle must be deemed to apply to
the distance purchases at 413 issue in the main proceedings, with the result that distant purchasers must
pay the excise duty of the Sate in which they receive their tobacco products and not that of the state from
which they were dispatched.
55.
The principle of taxation at the point of origin (more properly, in the country of acquisition) in art 8 of
the directive does not therefore apply in such cases. Its scope is cut down to purchases made by private
individuals followed by personal transport of the products acquired, in all cases for the personal use of the
purchaser.
56.
The analogy is not invalidated by the fact that art 10 requires the transport to be carried out, directly or
indirectly, by the retailer or on his behalf. It would be sufficient, in order to side-step all the complex
obligations21 imposed by art 10 on sellers, for the latter to agree with their customers (or with their agents,
as occurs in this case) certain conditions concerning transport, perhaps even with guarantees or
assurances from the retailer.
21
The seller must guarantee payment of the excise duty under the conditions laid down by the member state of
destination before despatch of the products; it must satisfy itself that the duty is paid after the products arrive and keep
records of deliveries.

57.
Finally, I have already referred to the rules in the directive on the involvement of persons other than
the buyer in the case of dispatches from a distance of products subject to excise duty (authorised
warehousekeepers, registered or non-registered traders and tax representatives other than the
consignees of the products). The involvement of the latterwhose characteristic function coincides with
that of the agentdoes not exclude the rule of taxation at destination in the ease of distance sales. For
the reasons given earlier, that same conclusion must apply to the distance purchases at issue in the main
proceedings.
(iv) The arguments against the proposed interpretation of the term products transported
58.
 The appellants have put forward a number of arguments in support of a different view from the one I
have just set out. First, they rely on the general rule that can be inferred from art 6 of the directive and on
the allusion to the principle governing the internal market mentioned in art 8 of the directive. Secondly,
they state that to interpret that article in the manner that I have just advocated would give rise to double
taxation and would render art 10 of the directive superfluous.
59.
It is true that art 6 of the directive provides that the duty becomes chargeable at the time of release for
consumption of the products. But its legal force is neither greater nor lesser than that of the remaining
articles of the directive. Specifically, it does not support any inference that the rule of taxation in the
country of acquisition is to take precedence over the rule of taxation in the country of destination in the
case of intra-Community movements of such products. Rather, an analysis of the directive points to
precisely the opposite conclusion.
60.
Of the three systems of intra-Community movement of products subject to excise duty to which I
referred earlier (namely, commercial movement of products under the suspension arrangement,
commercial transactions involving products that have already been taxed and purchases by private
individuals for their own use) the first two are subject to the rule of taxation at destination.
61.
Both when products are being moved under the suspension arrangement and when they are moved
outside that arrangement, in the context of intra-Community 414 commercial transactionssuch as, for
example, distance salesthey are subject to duty at their destination.
62.
The rule of taxation at the point of origin is therefore of much more limited scope than the appellants
claim. It applies to the purchase of goods which are not subsequently going to move between the various
member states and, under art 8 of the directive, also to products acquired by private individuals and
transported by them, not by way of trade.
63.
The directive thus sought to strike a balance between the two rules (one requiring taxation at origin
and the other taxation at destination) which cannot simply be frustrated by declaring that the first rule is to
be of general application, extending to cases outside the sphere within which the Community legislature
sought to confine it.
64.
As regards the principle governing the internal market referred to in art 8 as the reason for making the
excise duty chargeable in the country of acquisition, its importance as a means of interpretation is much
less than the appellants suggest.
65.
The internal market is characterised by the abolition between the member states of harriers to the free
movement of goods, persons, services and capital. It involves an area without internal frontiers, in which
that free movement is guaranteed in accordance with the Treaty (art 7a of the EC Treaty).
66.
However, I consider that principle to be compatible with a taxation system which in specific eases
requires excise dutyor even other taxes such as VATto be paid at the destination of the goods, not at
the point of origin, where the transactions take place within the Community. Needless to say, the opposite
approach (taxation at origin) is also perfectly compatible with the internal market and also displays the
advantages of simplicity and certainty.
67.
In reality, the obstacle to the free movement of goods, and thus to attainment of the internal market,
would arise either from the existence of other discriminatory internal taxes and charges having an effect
equivalent to import customs duties (both of which, in this case are ruled out), or from double taxation of
the products subject to excise duty. Specifically in order to obviate the latter, the directive contains
detailed provisions as to the chargeability of the duty, providing in some cases for reimbursement of duty
already paid (art 22).
68.
In the case of art 8, double taxation is avoided and the principle governing the internal market is
complied with, facilitating intra-Community movement of goods without frontier controls simply by applying
the rule of taxation at origin, albeit subject to certain requirements which I have already mentioned. I do
not consider that the essence of that principle, as a criterion for interpretation of the rule, calls for the
limits of art 8 to be extended in the way contended for by the appellants.
69.
The appellants assert, however, that the interpretation of art 8 which I advocate would, in certain
circumstances, give rise to a clear case of double taxation, since they have to pay the excise duty on the
same products first in Luxembourg, when buying them, and then in the United Kingdom when receiving
them.
70.
I cannot accept that argument. The scheme of the directive is based on the principle of a single
payment of duty, to such an extent that a whole section (Title IV) is devoted specifically to reimbursement
of sums already paid, in cases of intra-Community trade, in order to avoid double payment of tax on the
same products.
71.
Article 22(1) of the directive provides:
415
In appropriate cases, products subject to excise duty which have been released for
consumption may, at the request of a trader in the course of his business, be eligible for
reimbursement of excise duty by the tax authorities of the Member State where they were released
for consumption when they are not intended for consumption in that Member State.
And paras 3 and 4 of that article deal more particularly with the conditions for reimbursement in the cases
referred to in arts 7 and 10.
72.
In my opinion, those legislative provisions are sufficient to avoid double taxation of the same products,
since the reimbursement procedure would apply to the appellants case, with the result that they would be
required to pay only the duty chargeable in the United Kingdom.
73.
Even if it is conceded that the directive contains a legal lacuna concerning distance purchases of the
kind at issue in the main proceedings, regarding which, in theory, the retailer takes neither direct nor
indirect responsibility for transport, recourse to analogy, as a means of fleshing out the legal order, would
enable the principle of reimbursement of duty already paid at origin to be applied to such circumstances,
since only the duty chargeable in the country of destination would be reimbursed.
74.
Finally, the parties argument that art 10 of the directive would become pointless if art 8 were
interpreted in the manner that I advocate, also seems to me to be unfounded.
75.
In my opinion, art 10 is useful not only in order to confirm the rule that distance sales should be
taxable at destination but also to determine the obligations of the retailer and thereby facilitate monitoring
of the exaction of duty in the relevant member state.
76.
The appellants are in part right to state that, if art 8 is interpreted as I propose, art 10 adds nothing of
substance to determination of the country which is to levy the duty. Indeed, whilst the rule in art 8
(taxation at origin) is limited to goods bought by private individuals and transported by them, distance
sales, a sensu contrario, do not follow the same rule. And that is precisely what art 10 provides.
77.
That conclusion does not mean, however, that art 10 is superfluous or redundant. The express
provision for distance sales, as an economic phenomenon particularly suited to some products subject to
excise duty, and the need to add detail to the applicable legal rules, are sufficient reasons for the
Community legislature to have chosen to include that article in the directive.
78.
In short, I consider that the answer to the first preliminary question must uphold the chargeability of
excise duty in the member state of destination in the circumstances outlined by the national court.

The second preliminary question

79.
Both the appellants and the intervener in support of the UK tax administration consider that no answer
will be needed to the second question if the interpretation of art 8 which I set out above is adopted.
80.
For my part, although essentially agreeing with that view, I see no problem in giving an express
answer to the national court, confirming that the circumstances set out in the second question represent
no obstaclequite the contraryto excise duty being levied in the state to which the products are sent.
81.
The terms of the second question are very similar to those of the first since the Court of Appeal raises
the same problem, in relation to the same factual 416 situation (purchase in state B for the personal use
of a private individual resident in state A, through an agent who acts for the private individual and
organises transport of the goods bought). The only difference is that in the second preliminary question it
is expressly stated that the scheme was commercially devised and marketed.
82.
In my view, the answer that I suggest to the first preliminary question a fortiori extends to the case
described in the second question. The existence of a scheme like the one described by the national court
cannot preclude the chargeability of excise duty in the state of destination when the products have neither
been bought by private individuals as such nor transported by them to that state but have been
dispatched using a complex commercial mechanism.
83.
The directive clearly distinguishes operations in which only private individuals take part from those
which, involving as they do economic agents, are of a commercial nature. The reference to private
individuals appears only in art 8 of the directive (and, derivatively, in art 9(3)), emphasising that they are
acting in a merely private capacity, that is to say neither commercially nor professionally.
84.
 In contrast, cases in which there is an economic agent acting as such, and in any event involving
himself in the intra-Community movement of goods subject to excise duty, fall outside the scope of art 8 of
the directive.
85.
On the basis of that distinction, a system for the distribution of tobacco products in which one or more
economic agents play a decisive role at the various stages, receiving orders, accepting or rejecting them,
giving effect to such orders, physically receiving the products and arranging for them to be transported to
the home address in another member state of the customer who has previously made payment is, in my
opinion, a clear example of a commercial operation which has nothing to do with the purchases by private
individual covered by art 8 of the directive.
86.
At the heart of that operation, moreover, the action of a commercial service undertaking, which acts as
agent for the purchaser, cannot be stripped of its inherent features: the undertaking plays a commercial
role in the operations of purchase, transport and delivery of the products subject to excise duty which
move from one member state to another. The fact that such operations are preceded by a contract signed
with the customer (the end consumer) does not deprive them of their commercial nature, on the basis of
which, for example, the company could be sued in the event of breach of contract.
87.
Finally, the peculiar characteristics of this case are such that any court would be entitled to reject the
claims of a plaintiff by construing the tax provisions according to the interpretative criteria least conducive
to tax evasion.
88.
From that standpoint, having regard to the connections between the retailer and the agent, which are
fellow subsidiaries of the same company, it is even possible that the national court, whose legal system
enables it to lift the veil of incorporation in order to obviate any abuse of the law, will choose to apply art
10 to the case before it, taking the view that it is the selling company itself which, by means of an
entrepreneurial subterfuge, in reality indirectly transports the goods to their destination. The expression
indirectly, when used in relation to transport, as it is in the said art 10 of the directive, is very flexible and
allows such d conclusion.
89.
In the same way, if it were necessary to do so as a last resort, the national court could decline to apply
the rule contended for by the appellants (taxation at origin) on the basis that to apply it to the present case
would clearly run counter 417 to the spirit and purpose of the directive and would be inimical to the
effectiveness of other provisions of it. By so doing it would merely be applying the general legal principle
prohibiting acts in fraud of the law.

CONCLUSION
90.
In view of the foregoing, I propose that the Court of Justice reply as follows to the questions referred to
it by the Court of Appeal:
(1) Council Directive (EEC) 92/12 on the general arrangements for products subject to excise
duty and on the holding, movement and monitoring of such products, and in particular art 5 thereof,
does not preclude the charging of such duty on goods in a member state A in circumstances where:
(a) a private individual resident in state A acquires those goods for his own use; (b) the purchase is
made in state B, from which the goods are transported to state A; both operations involve an agent
who acts in the name or on behalf of the private individual and arranges for transport of the
products; (d) the individual does not himself travel with the goods from member state B to member
state A.
(2) The same conclusion applies where purchases displaying the same characteristics take
place under a scheme commercially devised and adopted for commercial purposes by one or more
economic agents.

2 April 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 31 July 1995, received at the Court of Justice of the European Communities on 18
September 1995, the Court of Appeal referred to the Court of Justice for a preliminary ruling pursuant to
art 177 of the EC Treaty two questions concerning the interpretation of Council Directive (EEC) 92/12 on
the general arrangements for products subject to excise duty and on the holding, movement and
monitoring of such products (OJ 1992 L76, p 1), as amended by Council Directive (EEC) 92/108 (OJ 1992
L390, p 124).
2.
Those questions were raised in proceedings between EMU Tobacco SARL (EMU), The Man in Black
Ltd (MBL) and John Cunningham, one of the directors of that company, and the Commissioners of
Customs and Excise (the commissioners) concerning the charging of excise duty on cigarettes.
 The UK legislation
3.
Section 2(1) of the Tobacco Products Duty Act of 1979 provides that there shall be charged on
tobacco products imported into the UK a duty of excise. The Excise Goods (Holding, Movement,
Warehousing and REDS) Regulations 1992, SI 1992/3135 set out who is to be liable and when the
liability arises. The Excise Duties (Personal Reliefs) Order 1992, SI 1992/3155 contains certain
exemptions, but they appear to have no relevance for the present case.

The applicable Community legislation

4.
The directive was to be transposed by the member states by 31 December 1992.
5.
Article 6 provides:
(1) Excise duty shall become chargeable at the time of release for consumption or when
shortages are recorded which must be subject to excise duty in accordance with Article 14(3).
Release for consumption of products subject to excise duty shall mean: (a) any departure, including
418 irregular departure, from a suspension arrangement; (b) any manufacture, including irregular
manufacture, of those products outside a suspension arrangement; (c) any importation of those
products, including irregular importation, where those products have not been placed under a
suspension arrangement. (2) The chargeability conditions and rate of excise duty to be adopted
shall be those in force on the date on which duty becomes chargeable in the Member State where
release for consumption takes place or shortages are recorded. Excise duty shall be levied and
collected according to the procedure laid down by each Member State, it being understood that
Member States shall apply the same procedures for levying and collection to national products and
to those from other Member States.
6.
According to art 7:
(1) In the event of products subject to excise duty and already released for consumption in one
Member State being held for commercial purposes in another Member State, the excise duty shall
be levied in the member state in which those products are held. (2) To that end, without prejudice to
Article 6, where products already released for consumption as defined in Article 6 in one Member
State are delivered, intended for delivery in another Member State or used in another Member
State for the purposes of a trader carrying out an economic activity independently or for the
purposes of a body governed by public law, excise duty shall become chargeable in that other
Member State. (3) Depending on all the circumstances, the duty shall be due from the person
making the delivery or holding the products intended for delivery or from the person receiving the
products for use in a Member State other than the one where the products have already been
released for consumption, or from the relevant trader or body governed by public law. (4) The
products referred to in paragraph 1 shall move between the territories of the various Member
States under cover of an accompanying document listing the main data from the document referred
to in Article 18(1). The form and content of this document shall be established in accordance with
the procedure laid down in Article 24 of this Directive. (5) The person, trader or body referred to in
paragraph 3 must comply with the following requirements: (a) before the goods are dispatched,
make a declaration to the tax authorities of the Member State of destination and guarantee the
payment of the excise duty; (b) pay the excise duty of the Member State of destination in
accordance with the procedure laid down by that Member State; (c) consent to any check enabling
the administration of the Member State of destination to satisfy itself that the goods have actually
been received and that the excise duty to which they are liable has been paid. (6) The excise duty
paid in the first Member State referred to in paragraph 1 shall be reimbursed in accordance with
Article 22(3).
7.
Article 8 provides:
As regards products acquired by private individuals for their own use and transported by them,
the principle governing the internal market lays down that excise duty shall be charged in the
Member State in which they are acquired.
8.
Article 9(1) and (2) reads as follows:
419
(1) Without prejudice to Articles 6, 7 and 8, excise duty shall become chargeable where
products [released] for consumption in a Member State are held for commercial purposes in
another Member State. In this case, the duty shall be due in the Member State in whose territory
the products are and shall become chargeable to the holder of the products.
(2) To establish that the products referred to in Article 8 are intended for commercial purposes,
Member States must take account, inter alia, of the following:the commercial status of the holder
of the products and his reasons for holding them, the place where the products are located or, if
appropriate, the mode of transport used, any document relating to the products, the nature of
the products, the quantity of the products.
 For the purposes of applying the content of the fifth indent of the first subparagraph, Member
States may lay down guide levels, solely as a form of evidence. These guide levels may not be
lower than:
(a) Tobacco products
cigarettes, 800 items
cigarillos (cigars weighing not more than 3 g each), 400 items
cigars, 200 items
smoking tobacco, 10kg;
(b) Alcoholic beverages
spirit drinks, 10l
intermediate products, 20 l
wines (including a maximum of 60l of sparkling wines), 90l
beers, 110l.
Until 30 June 1997 Ireland shall be authorised to apply guide levels which may not be less than
45 litres for wine (including a maximum of 30 litres of sparkling wine) and 55 litres for beer.
9.
Article 10 provides:
(1) Products subject to excise duty purchased by persons who are not authorised
warehousekeepers or registered or non-registered traders and dispatched or transported directly or
indirectly by the vendor or on his behalf shall be liable to excise duty in the Member State of
destination. For the purposes of this Article, Member State of destination shall mean the Member
State of arrival of the dispatch or transport. (2) To that end, the delivery of products subject to
excise duty already released for consumption in a member state and giving rise to the dispatch or
transport of those products to a person as referred to in paragraph 1, established in another
Member State, and which are dispatched or transported directly or indirectly by the vendor or on his
behalf shall cause excise duty to be chargeable on those products in the Member State of
destination. (3) The duty of the Member State of destination shall be chargeable to the vendor at
the time of delivery. However, Member States may adopt provisions stipulating that the excise duty
shall be payable by a tax representative, other than the consignee of the products. Such tax
representative must be established in the Member State of destination and approved by the tax
authorities of that Member State. The Member State in which the vendor is established must
ensure that he complies with the following requirements:guarantee payment of excise duty under
the conditions set by the Member State of destination prior to dispatch of the products and ensure
that the excise duty is paid following arrival of the products, keep accounts of deliveries of 420
products. (4) In the case referred to in paragraph 2, the excise duty paid in the first Member State
shall be reimbursed in accordance with Article 22(4). (5) Subject to Community law, Member States
may lay down specific rules for applying this provision to products subject to excise duty which are
covered by special national distribution arrangements compatible with the Treaty.

The case in the main proceedings

10.
EMU is a company incorporated in Luxembourg, a subsidiary of The Enlightened Tobacco Company
(ETC), which specialises in selling tobacco products in the Grand Duchy of Luxembourg.
11.
MBL is a company incorporated in the United Kingdom and is also a subsidiary of ETC. Since
November 1994 it has been soliciting and obtaining from private individuals resident in the United
Kingdom orders for cigarettes and tobaccos to be purchased from EMU. Those orders are placed by
individuals, for the purposes of their own private requirements, on the basis of a price list drawn up in
Luxembourg francs. They are placed through MBL, which purchases the products and arranges for their
importation into the United Kingdom by a private carrier in the name and on behalf of those individuals, in
return for payment of a fee. Customers may not purchase more than 800 cigarettes at one time.
12.
The terms under which those transactions are carried out were set out in an agreement made between
EMU and MBL on 14 November 1994.
13.
Those terms included a provision that EMU was to open and maintain a credit account in the name of
MBL for all purchases made by customers through MBL.
14.
For its part, MBL undertook to pay all sums received from customers directly into bank accounts in
London or Luxembourg. It irrevocably guaranteed the payment by return of EMUs monthly credit account
statement in accordance with the latters usual payment terms. MBL further agreed that any variations in
the pound-sterling purchase price of the products caused by exchange rate fluctuations would be
absorbed by equivalent increases or reductions in its own fee to customers.
15.
 The contracts for the sale and purchase of the products would be made in the Grand Duchy of
Luxembourg and title in those products would pass in that country. The contract was to be subject to the
law of England.
16.
Excise duty in the United Kingdom is generally higher than in the Grand Duchy of Luxembourg.
17.
In the course of 1995, the commissioners detained certain quantities of tobacco products when they
were imported into Dover, as they are authorised to do by UK legislation when excise duty is payable.
18.
EMU, MBL and Mr Cunningham applied for judicial review of that detention. They sought declarations
that UK excise duty was not payable and that the detention of the tobacco was unlawful, and also an
injunction restraining the commissioners from detaining the products imported under their scheme. They
contended that excise duty was payable in the Grand Duchy of Luxembourg and that consequently the
products in question were exempt from that duty in the United Kingdom, with the result that their prices
were considerably reduced, by up to 40% in some cases.
19.
Their application was refused and on 30 May 1995 they appealed to the Court of Appeal.
421
20.
The Court of Appeal considered that the case turned on the interpretation of the directive, in particular
of art 8 thereof, and accordingly decided to stay proceedings and request the Court of Justice to give a
preliminary ruling the following questions:
(1) Does Directive 92/12/EEC and in particular Article 8 have the effect of precluding the
charging of excise duty on goods in Member State A in circumstances where(i) the goods were
acquired for the use of a private individual in Member State A; (ii) they were acquired in Member
State B by an agent acting on behalf of that private individual; (iii) transportation of the goods from
Member State B to Member State A was arranged by the agent; and (iv) the individual did not
himself travel with the goods from Member State B to Member State A?
(2) Where a scheme has been commercially devised and marketed whereunder purchases
made in Member State B for the personal use of a private individual in Member State A are made
by an agent for that individual and those purchases are transported from Member State B to
Member State A as a result of arrangements made by such an agent, does Directive 92/12/EEC
have the effect of precluding the charging of excise duty on those purchases in Member State A?
21.
It appears from the order for reference that by its two questions the national court seeks essentially to
ascertain whether the directive must be interpreted as precluding the levying of excise duty in member
state A on goods released for consumption in member state B, where the goods were acquired from a
company, X, for the use of private individuals in member state A, through a company, Y, acting in return
for payment as agent for those individuals, and where transportation of the goods from member state B to
member state A was also arranged by company Y on behalf of those individuals and effected by a
professional carrier charging for his services.
22.
It should first be observed that the clear purpose of the directive is to lay down a number of rules on
the holding, movement and monitoring of products subject to excise duty, in particular so as to ensure that
chargeability of excise duties is identical in all the member states.
23.
As is apparent from the third, fifth, sixth and eleventh recitals in its preamble, the directive draws a
distinction between, on the one hand, goods held for commercial purposes, in respect of which
accompanying documents are required for transportation purposes, and, on the other hand, goods held
for personal use.
24.
As regards the goods last mentioned, as art 8 provides that excise duty is payable in the member state
in which they were purchased, no document is required when they are transported to another member
state.
25.
For art 8 to apply, however, a number of conditions must be satisfied. The goods on which excise duty
is chargeable must have been acquired by private individuals for their own use and transported by them.
26.
Those conditions must make it possible to establish that goods on which duty is chargeable and which
are acquired in one state and then transported to another are held for strictly personal purposes.
27.
The applicants in the main proceedings submit, first, that that provision should be applied in a situation
such as this, where the purchase of the goods chargeable to excise duty was effected through an agent
who also arranged for their transportation.
422
28.
 In support of their submission, the applicants in the main proceedings argue that the maxim of Roman
law qui facit per alium facit per se, meaning that a person acting through an agent must be treated as if he
himself were so acting, constitutes a general principle in a number of legal systems, in particular in
English law, and must be applied in this case, a fortiori since neither the English version of the directive
nor the French, Italian, Spanish, German, Dutch or Portuguese versions exclude the possibility of using
an agent.
29.
That argument cannot be upheld.
30.
First, it is clear from the case law of the court that the Community legal order does not, in principle,
aim to define concepts on the basis of one or more national legal systems unless there is express
provision to that effect (see the judgment in Nicolaus Corman & Fils SA v Hauptzollamt Gronau Case
64/81 [1982] ECR 13 (para 8)). The text of art 8 of the directive does not contain any express reference to
national legal systems.
31.
Secondly, even if the above-mentioned principle were common to all the member states, it must be
noted that, as the Advocate General has observed, it is one which derives from civil law, and more
specifically from the law of obligations, and it does not necessarily fall to be applied in the sphere of fiscal
law, where the objectives are of a quite different nature.
32.
Lastly, it should be noted that where the Community legislature intended the directive to apply in the
event of the involvement of a third party it did so by means of an express provision formulated for that
purpose. That is the case in arts 9(3) and 10(1).
33.
As far as art 8 is concerned, it is evident that none of the language versions expressly provides for
such involvement and that, on the contrary, the Danish and Greek versions indicate particularly clearly
that, for excise duty to be payable in the country of purchase, transportation must be effected personally
by the purchaser of the products subject to duty.
34.
The applicants in the main proceedings accept that those two language versions preclude the
involvement of an agent. They submit, however, that if those versions are not consistent with the other
versions they are to be disregarded, on the ground that, at the time when the directive was adopted,
those two member states represented in total only 5% of the population of the 12 member states and their
languages are not easily understood by the nationals of the other member states.
35.
In that regard it must be observed that the contradiction between the Danish and Greek versions on
the one hand and the other language versions on the other only arises if the argument put forward by the
applicants in the main proceedings is accepted.
36.
Furthermore, to discount two language versions, as the applicants in the main proceedings suggest,
would run counter to the courts settled case law to the effect that the need for a uniform interpretation of
Community regulations makes it impossible for the text of a provision to be considered in isolation but
requires, on the contrary, that it should be interpreted and applied in the light of the versions existing in
the other official languages (see esp Wrsdorfer, ne Koschniske v Raad van Arbeid Case 9/79 [1979]
ECR 2717 (para 6)). Lastly, all the language versions must, in principle, be recognised as having the
same weight and thus cannot vary according to the size of the population of the member states using the
language in question.
37.
It follows from the foregoing that art 8 of the directive is not applicable where the purchase and/or
transportation of goods subject to duty is effected 423 through an agent. The conditions governing the
application of art 8 are not, therefore, satisfied in the situation described by the national court.
38.
The applicants in the main proceedings argue further that the principle of legal certainty, as set out in
particular in EC Commission v France Case C-30/89 [1990] ECR I-691, implies that the directive should
be so construed as to give the benefit of any ambiguity to private individuals in so far as it is liable to
entail financial consequences.
39.
In that connection, it is to be observed that the existence of possible ambiguity in a provision can only
be established by reference to the context of that provision (see, inter alia, Firma E Merck v Hauptzollamt
Hamburg-Jonas Case 292/82 [1983] ECR 3781 (para 12)).
40.
As already noted in this judgment, in particular at para 33, it is clear from the directive that at no point
did the Community legislature intend art 8 to apply in the event of the involvement of an agent.
Accordingly that provision is not ambiguous.
41.
The applicants in the main proceedings submit, secondly, that, should art 8 be interpreted as not
applying where an agent is involved, then art 6 would have to be applied, with the consequence that
excise duty would be payable, in this case, solely in the Grand Duchy of Luxembourg, the country in
which the goods were released for consumption.
42.
On that point, it should be observed that, although art 6 provides that duty is to become chargeable at
the time of release for consumption of goods in a member state, this does not, however, preclude excise
duty from being subsequently levied in another member state pursuant to arts 7, 9 or 10, whereupon duty
paid in the first state may be reimbursed pursuant to arts 7(6) or 10(4).
43.
A situation such as that described by the national court would appear to fall under both art 7 and art 10
of the directive.
44.
Article 10(2) provides that the delivery of products subject to excise duty already released for
consumption in a member state and giving rise to the dispatch or transport of those products to a person
established in another member state who is not an authorised warehousekeeper or registered or non-
registered trader, and which are dispatched or transported directly or indirectly by the vendor or on his
behalf is to cause excise duty to be chargeable on those products in the member state of destination.
45.
That provision was so drafted as to cover not only the case where the products are dispatched or
transported by the vendor himself, but also, more broadly, all cases of dispatch or transport on behalf of
the vendor.
46.
Moreover, the formula used in that provision clearly indicates that the Community legislature is more
concerned with the objective nature of the transaction than with its legal form.
47.
In the case in the main proceedings, EMU and MBL are subsidiaries of the same company and as
such they can be regarded as forming part of the same economic entity, despite the fact that they are
separate legal persons (see Hydrotherm Gertebau GmbH v Compact del Dott Ing Mario Andreoli & C
sas Case 170/93 [1984] ECR 2999 (para 11)).
48.
In addition, it appears from the order for reference that MBL does not act at the instigation of the
private individuals it represents, but solicits from them orders for tobacco and cigarettes which are then
placed exclusively with EMU, the vendor. Lastly, MBL and EMU set out in general terms their reciprocal
obligations in an agreement made in 1991 which makes clear that the risks 424 inherent in fluctuation of
exchange rates are to be borne by MBL and not by the purchasers themselves.
49.
Accordingly, although MBL acted as agent for the purchasers under English law, it must be concluded
that the goods in question in the main proceedings were dispatched or transported directly or indirectly by
the vendor or on his behalf in the sense contemplated by art 10 of the directive and that excise duty is
chargeable in the United Kingdom.
50.
As regards the double taxation that would result, according to the applicants in the main proceedings,
from the levying of excise duty in the United Kingdom, art 10(4) provides expressly that, if art 10(2)
applies, the excise duty paid in the member state where the goods were released for consumption is to be
reimbursed.
51.
With regard to art 7, para 1 thereof provides for the levying of excise duty in a member state in which
products already released for consumption in another member state are held for commercial purposes.
Article 7(2) provides, to that end, that where such products are delivered, intended for delivery in another
member state or used in another member state for the purposes of a trader carrying out an economic
activity independently or for the purposes of a body governed by public law, excise duty is to become
chargeable in that other member state.
52.
In a case such as that now before the national court, where goods from one member state are carried
to another member state on the instructions of a trader acting in return for payment who has previously
solicited customers in that latter state and has arranged for importation of the goods, it must be held that
excise duty is chargeable in that latter member state.
53.
It follows from the foregoing that the directive must be interpreted as not precluding the levying of
excise duty in member state A on goods released for consumption in member state B, where the goods
were acquired from a company, X, for the use of private individuals in member state A, through a
company, Y, acting in return for payment as agent for those individuals, and where transportation of the
goods from member state B to member state A was also arranged by company Y on behalf of those
individuals and effected by a professional carrier charging for his services.

Costs
54.
The costs incurred by the United Kingdom, Austrian, Danish, Finnish, French, German, Greek, Irish,
Italian, Netherlands and Swedish governments and by the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the proceedings pending before the national court, the decision on costs
is a matter for that court.
On those grounds, The Court of Justice in answer to the questions referred to it by the Court of Appeal
by order of 31 July 1995, hereby rules: Council Directive (EEC) 92/12 on the general arrangements for
products subject to excise duty and on the holding, movement and monitoring of such products, as
amended by Council Directive (EEC) 92/108, must be interpreted as not precluding the levying of excise
duty in member state A on goods released for consumption in member state B, where the goods were
acquired from a company, X, for the use of private individuals in member state A, through a company, Y,
acting in return for payment as agent for those by a professional carrier charging for his services.

425

[1998] All ER (EC) 426

Djabali v Caisse dAllocations Familiales de lEssonne

(Case C-314/96)
EUROPEAN COMMUNITY; Other European Community
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE CHAMBER), SCHINTGEN (RAPPORTEUR) AND
MANCINI ADVOCATE GENERAL JACOBS
7, 15 MAY 1997, 12 MARCH 1998
European Community Reference for a preliminary ruling Dispute in main proceedings resolved after
reference made Whether Court of Justice obliged to give ruling EC Treaty, art 177.

The justification for a preliminary reference is not that it enables advisory opinions on general or
hypothetical questions to be delivered but rather that it is necessary for the effective resolution of a
dispute (see p 432 h, post).
Foglia v Novello Case 244/80 [1981] ECR 3045 and Zabala Erasun v Instituto Nacional de Empleo
Joined cases C-422424/93 [1995] All ER (EC) 758 applied.

Notes
For the scope of the reference procedure, see 51 Halsburys Laws (4th edn) paras 21812184
For the EC Treaty, art 177 (as amended by art G.56 of the Treaty on European Union), see 50
Halsburys Statutes (4th edn) Current Service 98.

Cases cited
Amministrazione delle Finanze dello Stato v Simmenthal SpA Case 106/77 [1978] ECR 629.
Chanel SA v Cepeha Handelsmaatschappij NV Case 31/68 [1970] ECR 403.
Dzodzi v Belgium Joined cases C-297/88 and C-197/89 [1990] ECR I-3763.
Foglia v Novello Case 244/80 [1981] ECR 3045.
Gmurzynska-Bscher v Oberfinanzdirektion Kln Case C-231/89 [1990] ECR I-4003.
Krid v Caisse Nationale dAssurance Viellesse des Travailleurs Salaris CNAVTS Case C-103/94 [1995]
ECR I-719.
Office national de lemploi (Onem) v Kziber Case C-18/90 [1991] ECR I-199.
Pardini (Fratelli) SpA v Ministero del commercio con lestero Case 338/85 [1988] ECR 2041.
Politi SAS v Italian Ministry of Finance Case 43/71 [1971] ECR 1039.
Yousfi v Belgium Case C-58/93 [1994] ECR I-1353.
Zabala Erasun v Instituto Nacional de Empleo Joined cases C-422424/93 [1995] All ER (EC) 758, [1995]
ECR I-1567, ECJ.

Reference
By judgment of 28 May 1996, the Tribunal des Affaires de Scurit Sociale, Evry (France), referred to the
Court of Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty a
question (set out at p 432 h, post) on the interpretation of art 39(1) of the Cooperation Agreement
between the EEC and the Peoples Democratic Republic of Algeria, signed in Algiers on 26 426April 1976
and approved on behalf of the Community by Council Regulation (EEC) 2210/78. That question was
raised in proceedings between Mrs Djabali, an Algerian national, and the Caisse dAllocations Familiales
de lEssonne concerning the award of a disabled adults allowance. Written observations were submitted:
by Mrs Djabali; on behalf of the French government, by C de Salins, Deputy Director of the Legal Affairs
Directorate of the Ministry of Foreign Affairs, and C Chavance, Secretary of Foreign Affairs in the same
directorate, acting as agents; and on behalf of the European Commission, by J Forman, Legal Adviser,
and M Patakia, of its Legal Service, acting as agents. Oral observations were made by the French
government and the Commission. The language of the case was French. The facts are set out in the
opinion of the Advocate General.

15 May 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
This case, referred by the Tribunal des Affaires de Scurit Sociale dEvry (the Social Security Court,
Evry), concerns the entitlement of an Algerian national resident in France to a special allowance payable
in France to disabled adults. The case also raises the issue whether the court should rule where the
dispute giving rise to the reference appears to have been settled after the reference was lodged but the
national court does not withdraw the reference.

The relevant Community legislation

2.
The entitlement to social security benefits of Algerian workers and their families resident in the
Community is governed by the Cooperation Agreement between the EEC and the Peoples Democratic
Republic of Algeria, signed in Algiers on 26 April 1976 and approved on behalf of the Community by
Council Regulation (EEC) 2210/78 (OJ 1978 L263 p 1) (the agreement).
3.
The object of the agreement is to promote overall cooperation between the contracting parties with a
view to helping to strengthen relations between them and to contributing to the economic and social
development of Algeria (art 1).
4.
Article 39(1) provides that, subject to the following paras of art 39, none of which is relevant to the
present case, workers of Algerian nationality and any members of their families living with them shall
enjoy, in the field of social security, treatment free from any discrimination based on nationality in relation
to nationals of the member state in which they are employed.
5.
The court has ruled that, since the term social security in the identically worded provision of the
Cooperation Agreement between the EEC and the Kingdom of Morocco Signed at Rabat on 27 April 1976
and approved on behalf of the Community by Council Regulation (EEC) 2211/78 (OJ 1978 L 264 p 1)
cannot receive a definition different from that indicated in the context of Council Regulation (EEC)
1408/71 on the application of social security schemes to employed persons, to self-employed persons
and to members of their families moving within the Community 1, disability allowances come within the
purview of social security within the meaning of that provision (see the judgment in Yousfi v Belgium Case
C-58/93 [1994] ECR I-1353 (para 28)).
1
See for the consolidated version applicable at the material time Annex I to Council Regulation (EEC) 2001/83 (OJ 1983
L230, p 6). The latest consolidated version is published as Part I of Annex A to Council Regulation (EC) 118/97 (OJ 1997
L28 p 1). With regard to the inclusion of the French disabled adults allowance in the scope of Regulation 1408/71 (S
edn 1971(II) p 416), see in particular arts 4(2)(a) and 10a and Annex IIa.

427

The facts and the national legislation

6.
The allocation aux adultes handicap (disabled adults allowance) was introduced by Law No 75/534 of
30 June 1975. The grant of the allowance is governed by Title II of Ch VIII of the new French Social
Security Code. That code provides that the allowance is available to every French national or national of a
state that has concluded a reciprocal agreement, subject to certain conditions as to the extent of the
applicants handicap and the receipt of other benefits (art 821.1, first sub-paragraph).
7.
Mrs Djabali, a disabled Algerian national, applied to the Caisse dAllocations Familiales (the CAF),
Essonne, for a disabled adults allowance with effect from October 1993. It appears to be accepted that
she satisfies the above-mentioned conditions. The documents on the national courts file, submitted to the
court, suggest that that application was initially granted; however, it was presumably refused thereafter,
since Mrs Djabali appealed to the CAFs Commission de Recours Amiable (appeals board). On 13 July
1994 the appeals board dismissed her appeal, apparently on the ground that Mrs Djabali was neither a
French national nor a national of a state that had concluded a reciprocal social security agreement with
France. On 14 June 1995 Mrs Djabali brought an action before the Tribunal des Affaires de Scurit
Sociale (the tribunal), arguing that the refusal to award her the disabled adults allowance was in breach
of art 39 of the agreement.
8.
The CAF contended before the tribunal that the disabled adults allowance was to be seen as a social
security benefit only when the applicant was or had been a worker and had therefore contributed in
general to the social security system. It argued that Mrs Djabali was not entitled to the benefit on the basis
that she had never been employed in France and was hence not a worker or former migrant worker.
9.
 On 28 May 1996, the tribunal stayed the proceedings and referred the following question to the Court
of Justice for a preliminary ruling:
Does Article 39 of Council Regulation (EEC) No 2210/78 of 26 September 1978 concerning the
conclusion of the Cooperation Agreement between the EEC and the Peoples Democratic Republic
of Algeria apply to Mrs Djabali in regard to the award of a disabled adults allowance where she has
never been employed but will be entitled, possibly in December 1997, to a pension in her capacity
as a non-working mother (mre au foyer)?
10.
The reference to art 39 of Regulation 2210/78 should, of course, be to art 39 of the agreement.
11.
By letter dated 8 April 1997, the CAF informed the Court of Justice that the Ministre du Travail et des
Affaires Sociales had decided in November 1996 to grant the allowance to Mrs Djabali. Mrs Djabali had
accordingly received F 148,18845 arrears and since January 1997 had been in receipt of monthly
payments of F 3,982. The CAF enclosed with its letter to the court copies of letters (i) from the CAF to Mrs
Djabali dated 27 December 1996, informing her that the Ministre du Travail et des Affaires Sociales had
decided to grant her the allowance with effect from 1 October 1993, confirming that instructions to make
the necessary payment to her had been given and inviting her to withdraw her case from the tribunal and
(ii) from the CAF to the tribunal dated 6 December 1996, to the same effect.
12.
It appears that Mrs Djabali has not taken the necessary procedural steps formally to withdraw her
case.
428
13.
By letter dated 11 April 1997, the registry of the Court of Justice asked the referring court whether it
intended to maintain its request for a preliminary ruling.
14.
By letter dated 25 April 1997, the referring court informed the court that it had no power under national
procedural rules to withdraw a question duly referred to the court for a preliminary ruling. Accordingly, it
had no option but to maintain the request for a ruling.
15.
Written observations were submitted by the applicant, the French government and the European
Commission. The French government and the Commission were represented at the hearing.

Jurisdiction of the court

16.
Mrs Djabali has now, apparently, been granted the benefits to which she claimed entitlement. Although
that fact does not affect the admissibility of the reference, since all the conditions for making a reference
were fulfilled at the time the reference was made, it must now be doubtful whether a decision on the
question referred can be necessary to enable the national court to give judgment, as required by art 177
of the EC Treaty. In this case, the relevant social security body has, in accordance with its revised view
that Mrs Djabali is entitled to the allowance, apparently made full payment. If that is the case, an answer
to the question referred can no longer be regarded as necessary for the national court.
17.
The question accordingly arises whether the court can strike the case off even though the reference
has not been formally withdrawn by the national court.
18.
In Chanel SA v Cepeha Handelsmaatschappij NV Case 31/68 [1970] ECR 403 the court of its own
motion ordered that a case be removed from the register of the court where the reference had lost its
purpose following amendment on appeal of the judgment of the national court by which the reference had
been made.
19.
In cases other than those where the decision to refer is quashed on appeal, the principle appears to
be that the Court of Justice must continue the procedure until the reference is withdrawn 2. Such a course
of conduct appears incongruous where, as here, the reference has lost its purpose because the dispute
has been resolved but the national court does not withdraw the reference. Indeed there may be a stronger
case for not proceeding to a ruling in a case such as the present than in a case where the decision to
refer is quashed on appeal: in the latter case, the lower court may have to proceed with the main
proceedings without the benefit of a ruling by the Court of Justice, while in cases such as the present the
ruling would be given but ex hypothesi not applied.
2
See Politi SAS v Italian Ministry of Finance Case 43/71 [1971] ECR 1039, in particular the opinion of Advocate General
Dutheillet de Lamothe, at p 1054, and the judgment in Amministrazione delle Finanze dello Stato v Simmenthal SpA
Case 106/77 [1978] ECR 629 (para 10).

20.
Community law does not of course preclude the national court from withdrawing the reference (see the
judgment in Zabala Erasun v Instituto Nacional de Empleo Joined cases C-422424/93 [1995] All ER
(EC) 758, [1995] ECR I-1567). The question which arises is whether, if the national court does not do so
(for example, as is apparently the case here, because it is unable as a matter of national procedure to do
so), the court should none the less strike the case off its register on the ground that a decision can no
longer be necessary.
429
21.
It may, it is true, be dangerous for the court to strike out the case without further contact with the
national court. The Court of Justice will not necessarily be in a position to determine conclusively, on the
basis of the information provided by the parties, that there is no need for the national court to continue
with the case: that court may conceivably need to continue for some reason which is not apparent from
this courts file. Similarly, it would clearly not be appropriate for the Court of Justice to accept the assertion
of one party alone that a settlement has been reached. If, however, the national court and the parties
were to be given the opportunity specifically to address the point, the Court of Justice could then properly
strike the case off its register in the absence of any reply.
22.
Serious problems could arise for the Court of Justice if it continued with the reference. For example, if
the litigation giving rise to the reference were settled at an early stage, some or all parties might not
submit observations; it might accordingly be difficult for the Court of Justice to give a ruling. Furthermore,
if there were several questions, or if the issues raised were complex, it would surely be disproportionate
for the Court of Justice to be required to answer questions which were no longer material to the resolution
of the dispute which had given rise to them.
23.
A solution which might be applied in cases where this problem arises is for the registry not merely to
ask the national court whether it intends to maintain its request, but to ask the national court and the
parties whether there are any grounds for considering that a decision on a question is still necessary to
enable it to give judgment. If no such reasons were provided, then the case could be struck off on this
courts motion.
24.
That solution would be consistent with the principle developed by the court that the justification for a
preliminary reference, and hence for the jurisdiction of the court, is not that it enables advisory opinions
on general or hypothetical questions to be delivered, but rather that it is necessary for the effective
resolution of a dispute (see the judgments in Foglia v Novello Case 244/80 [1981] ECR 3045 (para 19)
and Zabala Erasun [1995] All ER (EC) 758, [1995] ECR I-1567 (para 29)).

The question referred

25.
In this case, the question referred can fortunately be answeredif that should prove necessary
relatively briefly.
26.
The French government argues that the reference is inadmissible since the order contains insufficient
information. While it is true that the order is somewhat reticent about the facts, the issue is in my view
sufficiently clear for the court to be able to answer the question.
27.
The court has recognised that art 39 of the agreement has direct effect so that persons to whom it
applies are entitled to rely on it in proceedings before national courts (see the judgment in Krid v Caisse
Nationale dAssurance Viellesse des Travailleurs Salaris CNAVTS Case C-103/94 [1995] ECR I-719
(para 24)).
28.
It is not clear from the file whether Mrs Djabalis husband is or was 3 employed in France, although it is
suggested by a document on the national courts file that he is a worker. On the assumption that he is or
was so employed, 430then art 39(1) clearly applies and Mrs Djabali, as a member of his family living with
him, is entitled to the allowance.
3
Worker in the identically worded provision of the EEC-Morocco Cooperation Agreement encompasses former worker:
see the judgment in Office national de lemploi (Onem) v Kziber Case C-18/90 [1991] ECR I-199 (para 27).

Conclusion
29.
If, following further contact with the national court and the parties and in the light of the responses
given to the court, it transpires that the dispute which gave rise to the reference has indeed been settled
and that there are no grounds for considering that a decision on the question referred is still necessary to
enable the national court to give judgment, I consider that the court should either rule that it has no
jurisdiction to give a preliminary ruling on the question referred or strike the case off the register of its own
motion.
30.
If a ruling proves still to be required, I consider that the question referred by the Tribunal des Affaires
de Scurit Sociale dEvry should be answered as follows:
Article 39(1) of the Cooperation Agreement between the EEC and the Peoples Democratic
Republic of Algeria, signed in Algiers on 26 April 1976 and approved on behalf of the Community by
Council Regulation (EEC) 2210/78, precludes a member state from refusing to grant a benefit such
as the disabled adults allowance, which is provided for under its legislation for its own nationals, to
 the wife of an Algerian who is or has been employed in the member state concerned and with
whom she resides in that member state, on the ground that she is of Algerian nationality.

12 March 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By judgment of 28 May 1996, received at the Court of Justice of the European Communities on 26
September 1996, the Tribunal des Affaires de Scurit Sociale (the Social Security Court), Evry (France),
referred to the court for a preliminary ruling under art 177 of the EC Treaty a question on the interpretation
of art 39(1) of the Cooperation Agreement between the EEC and the Peoples Democratic Republic of
Algeria, signed in Algiers on 26 April 1976 and approved on behalf of the Community by Council
Regulation (EEC) 2210/78 (OJ 1978 L263 p 1).
2.
That question was raised in proceedings between Mrs Djabali, an Algerian national, and the Caisse
dAllocations Familiales de lEssonne (the CAF) concerning the award of a disabled adults allowance.
3.
According to the documents before the court, Mrs Djabali is married to an Algerian national with whom
she lives at Longjumeau (France). She has never worked in France.
4.
After a surgical operation which she underwent in 1981, Mrs Djabali was left physically disabled; the
rate of disability being recognised as 80% as from October 1993, she applied for the disabled adults
allowance with effect from that date pursuant to the French legislation.
5.
The disabled adults allowance was introduced in France by Law No 75/534 of 30 June 1975 making
provision for disabled persons. It is governed by Title II of Book VIII of the new French Social Security
Code. The conditions for awarding it are laid down in arts L 821-1 to L 821-8.
6.
The first para of art L 821-1 of that code provides that any French national or national of a state which
has concluded a reciprocal agreement concerning the grant of allowances for disabled adults who resides
on French territory, who has 431 already reached the age at which entitlement to the special education
allowance provided for in art L 541-1 of the same code is acquired and whose permanent disability is at
least equal to a percentage fixed by decree is entitled to the disabled adults allowance, provided that the
person concerned is not entitled, under a social security or old-age pension scheme or some special
legislation, to claim an old-age or invalidity benefit or an industrial accident pension for an amount at least
equal to that allowance.
7.
On 16 November 1993 the Technical Committee for Occupational Rehabilitation and Guidance,
referred to in art L 821-4 of code, awarded Mrs Djabali the allowance she had claimed, subject to
administrative conditions.
8.
Nevertheless, on 13 July 1994 the CAF, which is responsible for paying the benefit in question,
rejected Mrs Djabalis application on the ground that she was neither French nor a national of a country
which had concluded with France a reciprocal agreement concerning the grant of allowances for disabled
adults.
9.
On 4 June 1995 Mrs Djabali brought an action before the Tribunal des Affaires de Scurit Sociale,
Evry, claiming that that decision was contrary to art 39(1) of the agreement.
10.
Article 39(1) provides as follows:
workers of Algerian nationality and any members of their family living with them shall enjoy,
in the field of social security, treatment free from any discrimination based on nationality in relation
to nationals of the Member States in which they are employed.
11.
According to Mrs Djabali, it follows that the agreement prohibits the authorities of a member state from
basing a refusal to award social security benefits on the ground that the claimant is of Algerian nationality.
12.
The CAF maintains before the tribunal, however, that the allowance at issue can be considered as a
social security benefit within the meaning of the agreement only where the applicant is a worker or former
worker already in receipt of contributory social security benefits as a result of previous employment. Mrs
Djabali does not satisfy that condition, however, and cannot therefore claim the payment of the disabled
adults allowance as a personal right.
13.
 Considering that settlement of the dispute depended on the interpretation to be given to art 39(1) of
the agreement, the Tribunal des Affaires de Scurit Sociale, Evry, decided to stay proceedings and refer
the following question to the Court of Justice for a preliminary ruling:
Does Article 39 of Council Regulation (EEC) No 2210/78 of 26 September 1978 concerning the
conclusion of the Cooperation Agreement between the EEC and the Peoples Democratic Republic
of Algeria apply to Mrs Djabali in regard to the award of a disabled adults allowance where she has
never been employed but will be entitled possibly in December 1997 to a pension in her capacity as
a non-working mother (mre au foyer)?
14.
By letter of 8 April 1997, the CAF informed the Court of Justice that the French Minister for
Employment and Social Affairs had decided to grant the disabled adults allowance to Mrs Djabali. An
enclosure in the letter shows that she has received F 148,18845 arrears for the period October 1993 to
December 1996 and that since 1 January 1997 she has been in receipt of monthly payments of F 3,982.
Consequently, Mrs Djabali has received full satisfaction and there is no longer any matter at issue
between her and the CAF. The CAF enclosed with its letter copies of two letters informing the Tribunal
des Affaires de Scurit 432 Sociale, Evry, and Mrs Djabali of the decision of the competent authorities to
grant her the disabled adults allowance with effect from 1 October 1993 and inviting her to withdraw her
case from the national court.
15.
First, it is not disputed that Mrs Djabali has not taken the necessary steps to discontinue her action
before the national court.
16.
Second, in reply to a letter from the Court registry asking the Tribunal des Affaires de Scurit Sociale,
Evry, whether in those circumstances it intended to maintain its request for a preliminary ruling, the
President of that court stated that under the national rules of procedure it had no power to withdraw a
question duly referred to the court for a preliminary ruling.
17.
According to settled case law, the procedure provided for in art 177 of the Treaty is an instrument of
cooperation between the Court of Justice and national courts by means of which the former provides the
latter with interpretation of such Community law as is necessary for them to give judgment in cases upon
which they are called to adjudicate (see, inter alia, the judgments in Dzodzi v Belgium Joined cases C-
297/88 and C-197/89 [1990] ECR I-3763 (para 33) and Gmurzynska-Bscher v Oberfinanzdirektion Kln
Case C-231/89 [1990] ECR I-4003 (para 18)).
18.
It is clear from both the wording and the scheme of art 177 of the Treaty that a national court or
tribunal is not empowered to bring a matter before the Court of Justice by way of a reference for a
preliminary ruling unless a case is pending before it, in which it is called upon to give a decision which is
capable of taking account of the preliminary ruling (see, to that effect, the judgments in Fratelli Pardini
SpA v Ministero del commercio con lestero Case 338/85 [1988] ECR 2041 (para 11) and Zabala Erasun
v Instituto Nacional de Empleo Joined cases C-422424/93 [1995] All ER (EC) 758, [1995] ECR I-1567
(para 28)).
19.
The justification for a preliminary reference is not that it enables advisory opinions on general or
hypothetical questions to be delivered but rather that it is necessary for the effective resolution of a
dispute (see the judgments in Foglia v Novello Case 244/80 [1981] ECR 3045 (para 18) and Zabala
Erasun [1995] All ER (EC) 758, [1995] ECR I-1567 (para 29)).
20.
In this instance, after the Tribunal des Affaires de Scurit Sociale, Evry, had referred the question to
the court, Mrs Djabali received the benefits she had claimed.
21.
It must therefore be stated that the claims of the applicant in the main proceedings have been satisfied
in full, with the result that the case pending before the national court now has no purpose.
22.
In those circumstances, for the court to reply to the question referred would be of no avail to the
Tribunal des Affaires de Scurit Sociale, Evry.
23.
Consequently, there is no need to reply to the question referred.

Costs
24.
The costs incurred by the French government and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the action pending before the national court, the decision on
costs is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the question referred to it by the
Tribunal des Affaires de Scurit Sociale dEvry by judgment of 28 May 1996, hereby rules: there is no
need to reply to the question referred.
 433

[1998] All ER (EC) 434

Clean Car Autoservice GesmbH v Landeshauptmann von Wien

(Case C-350/96)

EUROPEAN COMMUNITY; Free movement of workers

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE CHAMBER), SCHINTGEN (RAPPORTEUR), MANCINI,
MURRAY AND HIRSCH ADVOCATE GENERAL FENNELLY
23 OCTOBER, 4 DECEMBER 1997, 7 MAY 1998
European Community Freedom of movement Workers Discrimination on grounds of nationality
Employer prevented from trading in Austria on grounds that manager not resident there Whether
employer could rely on employees right to free movement Compatibility with Community law of
residence requirement EC Treaty, art 48.

The plaintiff company applied to the Vienna City Council to register as a motor vehicle service station
under the relevant domestic provisions. The city council found that the statutory prerequisites for the
exercise of that trade, which, inter alia, stipulated that the manager should be resident in Austria, had not
been fulfilled and refused the registration. Thereafter, the plaintiff brought proceedings before the
Verwaltungsgerichtshof, contending that its manager, although resident in Berlin, was entitled, as an
employee in its service and therefore as a worker, to enjoy the right to freedom of movement established
by art 481 of the EC Treaty. The court stayed the proceedings and referred to the Court of Justice of the
European Communities for a preliminary ruling the questions: (i) whether the rule of equal treatment in the
context of freedom of movement for workers set out in art 48 could be relied on by an employer in order to
employ, in the member state in which it was established, workers who were nationals of another member
state; and (ii) whether art 48 precluded a member state from providing that an undertaking exercising a
trade on its territory could not appoint as manager a person not resident there. The city council contended
that the residence requirement was justified on the grounds that it was intended to ensure that the
manager could be served with notice of fines which could be enforced against him and that he was in a
position to act effectively in the business.
1
Article 48, so far as material, provides: (2) freedom of movement shall entail the abolition of any discrimination based
on nationality between workers of the Member States as regards employment, remuneration and other conditions of
work and employment.

Held (1) The right of workers to be engaged and employed without discrimination necessarily entailed
as a corollary the employers entitlement to engage them in accordance with the rules governing freedom
of movement for workers. It followed that the rule of equal treatment in the context of free movement of
workers set out in art 48 could be relied on by an employer to employ, in the member state in which it was
established, workers who were nationals of other members states (see p 449 d j and p 451 j, post).
(2) A requirement that nationals of other member states reside in the state concerned in order to be
appointed managers of undertakings exercising a trade constituted indirect discrimination based on
nationality contrary to art 48(2). 434Accordingly, art 48 precluded a member state from providing that an
employer could not employ as a manager a person not resident in that state, unless the imposition of
such a requirement was based on objective considerations independent of the nationality of the
employees concerned and proportionate to a legitimate aim pursued by the national law. In the instant
case, although the national rules had a legitimate aim, that aim could be achieved with less restrictive
measures with the result that the residence requirement went beyond what was necessary for that
purpose. Furthermore, the residence requirement could not be justified on grounds of public policy within
the meaning of art 48(3) and was therefore precluded by art 48 (see p 450 d to h and p 541 c g j, post);
Schning- Kougebetopoulou v Freie und Hansestadt Hamburg Case C-15/96 [1998] All ER (EC) 97 and
R v Bouchereau Case 30/77 [1981] 2 All ER 924 applied.

Notes
For equal treatment in the context of the material scope of a workers right to free movement, see 52
Halsburys Laws (4th edn) para 1513.
For the EC Treaty, art 48, see 50 Halsburys Statutes (4th edn) 283.

Cases cited
Algemene Transport (NV)en Expeditie Ondememing van Gend & Loos v Nederlandse
Belastingadministratie Case 26/62 [1963] ECR 1.
Asscher v Staatssecretaris van Financin Case C-107/94 [1996] All ER (EC) 757, [1996] ECR I-3089,
ECJ.
Bachmann v Belgium Case C-204/90 [1994] STC 855, [1992] ECR I-249, ECJ.
Cowan v Trsor public Case 186/87 [1989] ECR 195.
EC Commission v Luxembourg Case C-351/90 [1992] ECR I-3945.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Futura Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471.
GB-INNO-BM v Confdration du Commerce Luxembourgeois Case C-362/88 [1990] ECR I-667.
Hfner v Macrotron GmbH Case C-41/90 [1991] ECR I-1979.
Lawrie-Blum v Land Baden-Wrttemberg Case 66/85 [1986] ECR 2121.
Levin v Staatssecretaris van Justitie Case 53/81 [1982] ECR 1035.
Luisi v Ministero del Tesoro Joined cases 286/82 and 26/83 [1984] ECR 377.
Merci Convenzionali Porto de Genova SpA v Siderurgica Gabrielli SpA Case C-179/90 [1991] ECR I-
5889.
Merino Garca v Bundesanstalt fr Arbeit Case C-266/95 [1997] ECR I-3279.
Peskelogou v Bundesanstalt fr Arbeit Case 77/82 [1983] ECR 1085.
R v Bouchereau Case 30/77 [1981] 2 All ER 924, [1978] QB 732, [1978] 2 WLR 250, [1977] ECR 1999,
ECJ.
R v Ministry of Agriculture, Fisheries and Food, ex p Agegate Ltd Case C-3/87 [1991] 1 All ER 6, [1990] 2
QB 151, [1990] 3 WLR 226, [1989] ECR I-4459, ECJ.
Ramrath v Ministre de la Justice Case C-106/91 [1992] ECR I-3351.
Rewe-Zentral AG v Bundesmonopolverwaltung fr Branntwein Case 120/78 [1979] ECR 649.
Rush Portuguesa Lda v Office national dimmigration Case C-113/89 [1990] ECR I-1417.
Scholz v Opera Universitaria di Cagliari Case C-419/92 [1994] ECR I-505.
Schning-Kougebetopoulou v Freie und Hansestadt Hamburg Case C-15/96 [1998] All ER (EC) 97, ECJ.
435
Stoeckl (Criminal proceedings against) Case C-345/89 [1991] ECR I-4047.
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER
(EC) 97, [1995] ECR I-4921, ECJ.
van Binsbergen v Bestuur van de Bedrijfsvereniging voor de Metaalnijverheid Case 33/74 [1974] ECR
1299.
Van Duyn v Home Office (No 2) Case 41/74 [1975] 3 All ER 190, [1974] ECR 1337, ECJ.
Walrave v Union Cycliste Internationale Case 36/74 [1974] ECR 1405.
Wielockx v Inspecteur der Directe Belastingen Case C-80/94 [1995] All ER (EC) 769, [1995] ECR I-2493,
ECJ.

Reference
By order of 8 October 1996, the Verwaltungsgerichtshof (the Administrative Court) referred to the Court of
Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty two
questions (set out at p 448 f, post) on the interpretation of art 48 of that treaty and arts 1 to 3 of Council
Regulation (EEC) 1612/68 on freedom of movement for workers within the Community. Those questions
were raised in proceedings brought by Fortress Immobilien Entwicklungs GesmbH, now Clean Car
Autoservice GesmbH, an Austrian company established in Vienna, against the Landeshauptmann von
Wien (the Prime Minister of the Land of Vienna), concerning the rejection of an application by Clean Car
to register for a trade on the ground that it had appointed as manager a person who did not reside in
Austria. Written observations were submitted on behalf of: Clean Car, by C Kerres, Rechtsanwalt, Vienna;
the Landeshauptmann von Wien, by E Hechtner, Senatsrat am Amt der Wiener Landesregierung; the
Austrian government, by F Cede, Ambassador, Federal Ministry of Foreign Affairs, acting as agent; and
the European Commission, by P Hillenkamp and P J Kuijper, Legal Advisers, acting as agents. Oral
observations were submitted by the Commission. The language of the case was German. The facts are
set out in the opinion of the Advocate General.

4 December 1997.

The Advocate General (N Fennelly)

delivered the following opinion.

I INTRODUCTION
1.
This case concerns Austrian rules requiring managers of certain businesses to reside in Austria. It
raises the preliminary question of the entitlement of employers to invoke in national proceedings the
Community-law rights of workers (inter alios, the manager), as well as an issue of indirect discrimination,
and of its possible justification by reference to the need to ensure the effective notification and execution
of administrative sanctions.

II LEGAL AND FACTUAL CONTEXT

2.
Trade activity in Austria is regulated by the Gewerbeordnung 1994 (the Trade Code) (the GewO).
Paragraph 5(1) of the GewO provides that trades be exercised on the basis of the application to register
the trade in question, pursuant to para 339. That paragraph provides for applications to be submitted to
the district administrative authority. Pursuant to para 340 of the GewO, this authority examines the
application to ensure that the statutory conditions for the exercise of the trade applied for are fulfilled by
the applicant at the location concerned, failing which exercise of the trade shall be prohibited by a
decision based on such a finding.
436
3.
Among the statutory conditions for exercise of a trade, para 9(1) of the GewO states that legal
persons, commercial-law partnerships and registered civil partnerships may exercise a trade, but must
have appointed a manager or tenant, and refers further to para 39 of the GewO.
4.
Paragraph 39 of the GewO provides as follows:
(1) The owner may for the exercise of his trade appoint a manager who is responsible to the
owner for the proper exercise of the trade and to the authorities (Paragraph 333) for compliance
with the provisions of the law on trades; he must appoint a manager if he is not resident in Austria.
(2) The manager must satisfy the personal requirements prescribed for the exercise of the trade,
be resident in Austria, and be in a position to act accordingly in the business. In the case of a trade
for which the production of proof of qualification is prescribed, the manager of a legal person, to be
appointed in accordance with Paragraph 9(1), must also: 1. belong to the statutory representative
organ of the legal person; or 2. be a worker employed in the business for at least half the normal
weekly working hours and subject to full compulsory insurance in accordance with the provisions of
social security law.
The manager of an owner who is not resident in Austria, to be appointed under subparagraph 1
for the exercise of a trade for which the production of proof of qualifications is prescribed, must be a
worker employed in the business for at least half the normal weekly working hours and subject to
full compulsory insurance in accordance with the provisions of social security law. The provisions of
Paragraph 39(2), in force until the coming into force of Federal law BGBl No 29/1993, shall
continue until 31 December 1998 to apply to persons who have been appointed as manager by 1
July 1993.
(3) In cases where a manager must be appointed, the owner must make use of a manager who
acts in the business accordingly.
5.
Pursuant to para 370(2) of the GewO, any fines regarding the conduct of a trade are to be imposed on
the manager, where a manager has been notified or approved.
6.
Fortress Immobilien Entwicklungs GesmbH, now Clean Car Autoservice GesmbH (Clean Car), a
company established in Austria, applied on 13 June 1995 to the Magistrat der Stadt Wien (the Vienna City
Council), District Office for Districts 13 and 14, to register the trade of maintenance and care of motor
vehicles (service station) excluding all artisanal activity. It notified the appointment of Mr Rudolf Henssen
as manager. It stated that Mr Henssen, a German national, was looking for accommodation in Austria, so
that evidence of Austrian residence would be produced later. It appears that Mr Henssen has since
moved to Vienna. The District Office for District 23 of Vienna decided on 20 July 1995 to prohibit the
exercise of the trade applied for, as the nominated manager resided in Berlin and, thus, did not comply
with the conditions in para 39(2) of the GewO that he reside in Austria and be in a position to act in the
business.
7.
On 10 August 1995, Clean Car commenced an administrative appeal before the Landeshauptmann
von Wien (the Prime Minister of Vienna). It submitted that Mr Henssen now resided in Vienna, and that, in
any event, residence anywhere in the European Union should be deemed to satisfy the residence
criterion. The Landeshauptmann rejected the appeal on 2 November 1995, on the basis that the material
time was the date of application, when the manager was not yet an Austrian resident.
437
8.
On 21 December 1995, Clean Car introduced a complaint (Beschwerde) against this decision before
the Verwaltungsgerichtshof, Wien (the Administrative Court, Vienna), on the basis that its arguments
based on Community law had been ignored. Clean Car referred in particular to arts 6 and 48 of the EC
Treaty and to the prohibition of covert discrimination, and asserted that Mr Henssen was an employee of
the company and, therefore, a worker.

III QUESTIONS
9.
In order to be able to give judgment in the case, the national court considered it necessary to refer the
following questions for a preliminary ruling pursuant to art 177 of the Treaty:
 (1) Are Article 48 of the EC Treaty and Articles 1 and 3 of Council Regulation (EEC) 1612/68 on
freedom of movement for workers within the Community (OJ S edn, First Series 1968 (II) p 475) to
be interpreted as meaning that employers in the host state also derive therefrom the right to employ
workers who are nationals of another Member State without being bound by conditions which
even if they do not depend on nationalityare typically linked with nationality?
(2) If employers of the host state have the right stated in Question 1: Are Article 48 of the EC
Treaty and Articles 1 and 3 of Regulation 1612/68 to be interpreted as meaning that a provision
such as Paragraph 39(2) of the Gewerbeordnung 1994, under which the owner of a trade may
appoint as a manager for trade law purposes only a person whose residence is in the host state
(Austria), is consistent therewith?
10.
The national court indicated that the first question essentially relates to the possibility of an employer
relying upon provisions which are couched in terms of workers rights. It also suggested that account be
taken, in answering the second question, of the fact that the manager is responsible to the authorities for
the observance of the provisions of Austrian trade law.

IV OBSERVATIONS
11 Written observations were submitted by Clean Car, the Landeshauptmann von Wien, the Republic
of Austria and the European Commission. Oral observations were submitted by the Commission.
12.
Clean Car argues that an interpretation of art 48 of the EC Treaty and of arts 1 and 3 of Regulation
1612/68, which did not grant employers the right to engage workers without their having to comply with
conditions which are typically connected with having the nationality of a particular state would undermine
the right of free movement. The potential derogations provided for in art 48(3) of the Treaty should be
interpreted restrictively and are not relevant to the present case (see the judgment in Van Duyn v Home
Office (No 2) Case 41/74 [1975] 3 All ER 190, [1974] ECR 1337). In particular, the public policy
justification would be applicable only if a worker from another member state were engaged in work which
was itself contrary to public policy. Clean Car adds that the requirement that a manager be in a position to
act effectively in a business could, in certain locations, be more easily satisfied by a frontier worker
resident in a contiguous part of Germany than by a person residing in a distant part of Austria.
13.
The Landeshauptmann von Wien accepts that employers may derive rights from art 48 of the Treaty
and from arts 1 and 3 of Regulation 1612/68, but maintains that the restriction at issue in the present case
is justified by reference to 438 considerations of general interest. He emphasises that a manager is
responsible to the Austrian authorities, on behalf of the owner of the business, for the observance of all
applicable legal provisions, and is liable to administrative sanctions for any breach thereof. He must
therefore reside where such sanctions can be notified to him and, if necessary, executed. The amended
version of para 39(2) of the GewO, applicable from 1 July 1996 (but not to the facts of this case), provides
that the manager must live in Austria, in so far as the notification and execution of any sanctions imposed
is not guaranteed pursuant to an international agreement. The Landeshauptmann von Wien compares
para 39(2) of the GewO with art 38(2) of the Rules of Procedure of the court, which requires that
applications to the court state an address for service in the place where the court has its seat and the
name of the person who is authorised and has expressed willingness to accept service. The function of
manager, moreover, is not limited to accepting service of administrative or other documents, but extends
to personal responsibility for the conduct of the business.
14.
The Austrian government submits, by reference to the case law of the court, that an employer does
not, as such, fall within the personal scope of application of art 48 of the Treaty (see the judgments in
Levin v Staatssecretaris van Justitie Case 53/81 [1982] ECR 1035 (para 9), Lawrie-Blum v Land Baden-
Wrttemberg Case 66/85 [1986] ECR 2121 (paras 16ff) and Asscher v Staatssecretaris van Financin
Case C-107/94 [1996] All ER (EC) 757, [1996] ECR I-3089 (para 25)). This entails a negative answer to
the first question, thereby removing any need to answer the second.
15.
In the alternative, the Austrian government argues that the material provisions of para 39(2) of the
GewO are justified by reference to considerations of the general interest (see the judgments in
Bachmann v Belgium Case C-204/90 [1994] STC 855, [1992] ECR I-249, Ramrath v Ministre de la
Justice Case C-106/91 [1992] ECR I-3351 (paras 29ff) and Finanzamt Kln-Altstadt v Schumacker Case
C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225). Service of notice of sanctions and their execution in
other member states of the European Union are possible only in very limited circumstances, other than
pursuant to bilateral accords. In these circumstances, consideration of arts 1 and 3 of Regulation 1612/68
is superfluous, as they merely implement art 48 of the Treaty.
16.
The Commission states that art 48 of the Treaty and Regulation 1612/68 grant rights to employees,
not employers. It, therefore, seeks to establish whether Mr Henssen is a worker within the meaning of
those provisions, as defined in the courts judgment in Lawrie-Blum [1986] ECR 2121. In that case, the
court stated that an employment relationship is defined in accordance with objective criteria by reference
to the rights and duties of the persons concerned, and that its essential feature is that for a certain period
of time a person performs services for and under the direction of another person in return for
remuneration (see [1986] ECR 2121 (para 17)). As the manager of a business is subject to the board of
the company and to the general meeting of the shareholders, who appoint him to perform certain
administrative tasks on their behalf and at their direction, and as he probably has a contract of
employment with the company, and as he cannot be deemed, in the case of a limited company, to be an
independent service provider within the meaning of art 52 unless he owns all the shares in the company,
the Commission concludes that a manager in circumstances such as those of the present proceedings is
a worker within the meaning of art 48 of the Treaty.
17.
The Commission adds that the principal effect of the residence rule in para 39(2) of the GewO is to
exclude non-Austrians. Furthermore, it prevents businesses from appointing managers responsible for
their activities in more than 439 one member state. The Commission submits, furthermore, that art 48(3)
of the Treaty is not applicable in the present case. While it is possible to justify national rules which are
not directly discriminatory by reference to considerations of the general interest, and there is a general
interest in securing the effective service and execution of administrative sanctions in the case of non-
compliance by a business with the applicable law, the Commission submits that the Austrian rules are
disproportionately restrictive of the freedom guaranteed by art 48 of the Treaty (see the judgments in
Bachmann [1994] STC 855, [1992] ECR I-249 (paras 21ff) and EC Commission v Luxembourg Case C-
351/90 [1992] ECR I-3945 (paras 19ff)). It would be sufficient to oblige the manager to have a
professional address in Austria, which could be that of the company itself where it is established in
Austria, or to require the company to furnish, by agreement with the authorities, a suitable guarantee
regarding possible future administrative sanctions.

V ANALYSIS

Question 1
18.
The national court has not raised a question regarding whether Mr Henssen is a worker within the
meaning of Community law. In fact, it refers to a manager as an employee in its order for reference, a
view which implicitly underlies both the first and the second questions. The court has indicated that
employees are to be treated as workers for the purposes of Community law (see the judgment in Lawrie-
Blum [1986] ECR 2121 (para 17)). It is also well established that art 48 of the Treaty gives rise to rights
which are directly effective before national courts (see eg the judgment in Walrave v Union Cycliste
Internationale Case 36/74 [1974] ECR 1405). The issue raised by the first question is whether an
employer, rather than an employee, can invoke before national courts rights deriving from art 48 of the
Treaty and arts 1 and 3 of the regulation. Since the latter provisions merely clarify and give effect to the
rights already conferred by Article 48, the answer to the first question must be sought in that article (see
the judgment in Scholz v Opera Universitaria di Cagliari Case C-419/92 [1994] ECR I-505 (para 6)).
19.
Austria has argued that employers do not fall within the personal scope of application of art 48 of the
Treaty. The cases which it cites2 establish the definition of a worker and of the employment relationship on
which that status depends, and state that the rights of freedom of movement are linked to that status
(see esp the judgment in Levin Case 53/81 [1982] ECR 1035 (para 9)). However, those decisions do not
address, nor do they exclude, either expressly or by implication, the extension of the benefit of
Community-law provisions on freedom of movement of workers to persons other than workers who, none
the less, have a material connection with a person who has that status. Nor can any such inference be
drawn from the texts of the relevant Treaty and legislative provisions. For example, art 49 of the Treaty is
the legal basis for the regulation, including its provisions on rights relating to the residence, housing,
employment and education of members of workers families, irrespective of their nationality.
2
See the judgments in Levin v Staatssecretaris van Justitie Case 53/81 [1982] ECR 1035, Lawrie-Blum [1986] ECR 2121
and Asscher [1996] All ER (EC) 757, [1996] ECR I-3089.

20.
Article 48(3) of the Treaty is couched in terms of rights which are, of their nature, attributed to workers:
to accept offers of employment, to move freely for this purpose, and to stay in a member state for the
purpose of employment. Article 1 of the regulation, similarly, speaks of the right of any member state
national, 440irrespective of his place of residence, to take up and pursue activity as an employed person
in another member state, with the same priority as its own nationals. Article 48(1) and (2), on the other
hand, does not expressly identify any particular beneficiary of freedom of movement of workers, which
shall be secured in the Community by the end of the transitional period at the latest, and which shall
entail the abolition of any discrimination based on nationality between workers of the Member States as
regards employment, remuneration and other conditions of work and employment. In the same vein, art 3
of the regulation simply states that national legal, regulatory or administrative provisions or administrative
practices shall not apply where they limit application for, offers of, or the taking up and pursuit of
employment, or subject these to conditions not applicable in respect of their own nationals or which,
though applicable irrespective of nationality, have as their exclusive or principal aim or effect to keep
nationals of other member states away from the employment offered.
21.
It would add greatly to the effectiveness of these rights and prohibitions if they could also be invoked
by economic actors other than workers, whose freedom of access to workers from member states other
than their own is restricted. For example, the right of workers under art 48(3) of the Treaty to accept offers
of employment actually made, could be set at nought if employers were not free to challenge national
restrictions on the making of such offers. It must also be borne in mind that, while freedom of movement
of workers may be conceived of, in part, in terms of workers personal rights, and is strengthened by their
efforts to secure such rights, inter alia before national courts, it ultimately serves an objective of general
interest, provided for in art 3(c) of the Treaty: the establishment of an internal market characterised by the
abolition, as between member states, of obstacles to the free movement of persons.
22.
The court has already directly addressed the entitlement of employers to invoke what are normally
characterised as workers rights in R v Ministry of Agriculture, Fisheries and Food, ex p Agegate Ltd Case
C-3/87 [1991] 1 All ER 6, [1989] ECR I-4459. That case concerned the interpretation of arts 55 and 56 of
the Act of Accession (1985) (EC 27 (1985); Cmnd 9634; OJ 985 L302 p 1) (Spain and Portugal), on
freedom of movement of workers. The court was asked if provisions of Community law precluded the
laying down in UK law of conditions regarding the nationality, residence and social security contributions
of crew members for the grant of fishing licences which excluded most Spanish nationals, and whether
such provisions could be relied upon in national courts by the owner of a fishing vessel with a partly
Spanish crew. The court stated that the concept of worker in art 55 of the Act of Accession was identical
to that in art 48 of the Treaty and that the crew members were not ineligible to be treated as workers by
reason of the manner in which they were paid (see [1991] 1 All ER 6, [1989] ECR I-4459 (para 34)). The
court ruled, furthermore, that the derogation in art 56(1) of the Act of Accession from the immediate
application of art 48 of the Treaty as between Spain and the existing member states must be interpreted
restrictively (see para 39)3. In particular, it could not be interpreted as permitting the introduction of new
restrictions, such as some of the conditions in the United Kingdom law at issue (see the judgment in R v
Ministry of Agriculture, Fisheries and Food, ex p Agegate Ltd Case C-3/87 [1991] 1 All ER 6, [1989] ECR
I-4459 (para 40)).
3
See also Peskelogou v Bundesanstalt fr Arbeit Case 77/82 [1983] ECR 1085.

In response to the question of the entitlement of the shipowner and employer to invoke these
provisions, the court responded simply that all of the provisions in 441 question had direct effect and that
they could, consequently, be relied upon by individuals before a national court (see [1991] 1 All ER 6,
[1989] ECR I-4459 (para 42)).
23.
In Merci Convenzionali Porto de Genova SpA v Siderurgica Gabrielli SpA Case C-179/90 [1991] ECR
I-5889 (para 23 and operative part), the court found that, even within the framework of art 90, the
provisions of art 48 have direct effect and give rise for interested parties to rights which the national courts
must protect. In that case, the interested party was an importer who complained of the fact that its own
ships crew was not permitted to unload a cargo at the port of Genoa because such dock work was
reserved to an undertaking whose workers were required to be of Italian nationality. Furthermore, in the
field of sexual equality in the workplace, which is probably also conceived of primarily in terms of workers
rights, the court implicitly accepted in Criminal proceedings against Stoeckl Case C-345/89 [1991] ECR I-
4047 the right of an employer to invoke, in his defence to a prosecution under national labour rules
prohibiting nightwork, the directly effective provisions of Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to employment,
vocational training and promotion, and working conditions (OJ 1976 L39 p 40).
24.
The extension to employers and other interested parties of the right to invoke Community-law
provisions on freedom of movement of workers is also consistent with the case law of the court regarding
the personal scope of other internal market freedoms. In Luisi v Ministero del Tesoro Joined cases 286/82
and 26/83 [1984] ECR 377 (paras 10, 16) and in Cowan v Trsor public Case 186/87 [1989] ECR 195
(para 15) the court found that the Treaty provisions on services, which speak only of the freedom to
provide services, can also be relied upon by the recipients of services, because this is a necessary
corollary thereof (see Luisi [1984] ECR 377 (para 10)); and had in fact been expressly envisaged from the
outset (see [1984] ECR 377 (paras 1214). Indeed, in Bachmann [1994] STC 855, [1992] ECR I-249
(para 31), the court held, in a case brought by a recipient of insurance services, that the impugned
provisions of Belgian tax law constituted a restriction on insurers freedom to provide services. It is worth
noting that art 59, like art 48(1) and (2), is couched, not in terms of the rights of any particular class of
persons, but in terms of the abolition of restrictions. A broad definition of the persons who derive rights
from the provisions of the Treaty on fundamental economic freedoms is not, therefore, inconsistent with
the relevant texts. It reflects the courts statement in NV Algemene Transporten Expeditie Ondememing
van Gend & Loos v Nederlandse Belastingadministratie Case 26/62 [1963] ECR 1 at 12 that the
functioning of the common market, which it is the objective of the Treaty to establish, is of direct concern
to interested parties in the Community. Thus, interested individuals could derive directly enforceable rights
even from a Treaty provision expressed in terms of a prohibition, and their vigilance could amount to an
effective supervision of the implementation of Community law in addition to that arising under arts 169
and 170 of the Treaty (see [1963] ECR 1 at 13).
25.
It would be strange if consumers had a directly effective Community-law right to travel to other
member states to avail of tourist or other services, or to shop for goods, but employers did not have an
equivalent right to travel to recruit workers in another member state (see the judgment in GB-INNO-BM v
Confdration du Commerce Luxembourgeois Case C-362/88 [1990] ECR I-667 (para 8)). It would also
be illogical if an employer could not complain of discriminatory limitations of his ability to hire workers
abroad, whereas a recruiting agency which he hired to do 442 so could complain before the national
courts about such restrictions on its freedom to provide services 4.
4
On the relationship of recruitment activities with the services provisions of the Treaty, see the judgment in Rush
Portuguesa Lda v Office national dimmigration Case C-113/89 [1990] ECR I-1417 (para 16); see also the judgment in
Hfner v Macrotron GmbH Case C-41/90 [1991] ECR I-1979 (paras 3540).
 Employers have a direct and real economic interest in the effectiveness of art 48. An employer is an
inescapable participant in the exercise by workers of the freedoms guaranteed to them. As interested
individuals, employers can play an effective role in the supervision of the achievement of the common
market, in addition to that of the Commission pursuant to art 169 of the Treaty.
26.
I conclude, therefore, in response to the first question, that employers in the host state derive from art
48 of the Treaty the directly effective right to employ workers who are nationals of another member state
without being bound by national rules which discriminate, directly or indirectly, on grounds of the
nationality of the workers in question.

Question 2
27.
It is contrary to art 48(2) of the Treaty, as well as to art 3(1) of the regulation, for member states to lay
down conditions for employment which are indirectly or overtly discriminatory on grounds of nationality.
The court has observed that national rules under which a distinction is drawn on the basis of residence
are liable to operate mainly to the detriment of nationals of other member states. Non-residents are in the
majority of cases foreigners (see eg the judgment in Schumacker [1995] All ER (EC) 319, [1995] ECR I-
225 (para 28)).
28.
It is evident, therefore, that the imposition of a condition that managers appointed by certain
companies in Austria be resident in that country may constitute indirect discrimination on grounds of
nationality. None the less, such apparent discrimination may be justified by reference to the requirements
of the general interest (see [1995] All ER (EC) 319, [1995] ECR I-225 (para 39)). Two such possible
justifications have been mentioned in the present case, both related to the fact that the manager is
responsible in Austrian law for the conduct of the business.
29.
Regard may be had to the courts judgment in van Binsbergen v Bestuur van de Bedrijfsvereniging
voor de Metaalnijverheid Case 33/74 [1974] ECR 1299. The courts analysis of the professional rules
imposed by member states on service-providers established in other member states can, for certain
purposes, be extended to the situation of workers who, though resident in one member state, occupy
positions of responsibility in another. The court stated ([1974] ECR 1299 (para 12)):
taking into account the particular nature of the services to be provided, specific requirements
imposed on the person providing the service cannot be considered incompatible with the Treaty
where they have as their purpose the application of professional rules justified by the general good
in particular rules relating to organization, qualifications, professional ethics, supervision and
liabilitywhich are binding upon any person established in the State in which the service is
provided, where the person providing the service would escape from the ambit of those rules being
established in another Member State.
30.
The first possible justification in the present case is that the residence requirement ensures
compliance with the condition that the manager be in a 443 position to act as such in the business, that is,
that he exercise a real rather than a merely formal role. In the light of the managers responsibility to the
authorities for the conduct of the business, this is a legitimate objective, which can be placed in the van
Binsbergen category of professional rules on organisation. However, the residence requirement may be,
according to the circumstances, either unnecessary for or, more fundamentally, unrelated to the
achievement of this aim. It is unnecessary if, as has been suggested, a frontier worker is able to perform
his managerial tasks without giving up his residence in a neighbouring member state. However, at the
material time, Mr Henssen appears to have been resident in Berlin. It is more germane to this case to
consider how the residence requirement is designed to secure the desired end. It is insufficient if; despite
residence in Austria, a manager still does not, or cannot, participate as required in the effective
management of the company. It would thus be less restrictive for the national authorities simply to impose
directly a condition of effective involvement, if necessary specifying, as is done in respect of certain firms
in the GewO, conditions regarding working hours, and to leave it to the manager to decide, in the light of
geographical and other circumstances, how to reconcile his residence with this condition.
31.
Secondly, it has been argued that it is necessary that the manager reside in Austria for the purpose of
the service of notice and execution of administrative sanctions in the case of breach of the rules
governing the conduct of the business. This is a legitimate objective. The member states have an obvious
interest in the public and regular conduct of traders registered pursuant to laws such as the GewO and
consequently in maintaining effective rules for their supervision. In my view, such supervision, just as
much as fiscal supervision and a number of other established mandatory requirements of general interest,
is capable of justifying a restriction on the exercise of fundamental freedoms guaranteed by the Treaty
(see the judgments in Rewe-Zentral AG v Bundesmonopolverwaltung fr Branntwein Case 120/78 [1979]
ECR 649 (para 8) and Futura Participations SA v Administration des Contributions Case C-250/95 [1997]
ECR I-2471 (para 31)). In the light of the apparent difficulty in enforcing administrative penalties outside
the states jurisdiction, conditions may be imposed to ensure that managers do not escape the ambit of
Austrias rules on professional responsibility by residing outside Austria.
32.
 It must also be considered, however, whether there are less restrictive ways in which this objective
could be achieved5. As regards service of notice of sanctions, the court recognised in it judgment in van
Binsbergen [1974] ECR 1299 (paras 14, 16), in the context of the need to ensure the observance of
professional rules of conduct connected with the administration of justice and with respect for professional
ethics, the adequacy of a lawyer established in another member state choosing an address for service in
the member state in question. In the context of the present case, this could conceivably be either at the
managers place of business or, as the Commission has suggested, the seat of the legal person which
employs him, if it is established in Austria.
5
It has not been suggested that managers must reside in Austria in order to permit the exercise of a jurisdiction to impose
administrative sanctions in the first place. Thus, the analysis which follows relates solely to the practical problems raised
by the notification and execution of such sanctions.

33.
However, van Binsbergen concerned a lawyer providing services in a member state other than his
own. The court was at pains to distinguish the position of persons permanently established. The analogy
with services provision is less compelling in the case of the enforcement of substantive professional rules
against 444 a worker engaged exclusively in one member state. Since I am not in doubt that Austria is
within its rights in requiring that there be an effective mechanism for the enforcement of administrative
sanctions, it needs to be seen whether there is a possible, less restrictive means to this end than the
residence requirement. The court recognised in Bachmann [1994] STC 855, [1992] ECR I-249 (para 24)
that the difficulties which a member states tax authorities might have in enforcing an undertaking by an
insurer established in another member state to pay tax on sums payable to a taxable person residing in
the former member state could justify indirectly discriminatory tax treatment of that persons insurance
contributions. However, while accepting that the expense involved would be prohibitive in the
circumstances of that case, the court also stated that It would certainly be possible in principle for such
an undertaking to be accompanied by the deposit by the insurer of a guarantee (see [1994] STC 855,
[1992] ECR I-249 (para 25)).
34.
By the same token, it would be less restrictive of the freedom of movement of workers if a foreign-
resident prospective manager had the option of furnishing a guarantee or security for compliance with
potential administrative sanctions to the Austrian authorities, whether by a deposit, the blocking of a
certain sum in a bank account, the nomination of a guarantor, or some other means, in lieu of changing
his residence. The material available does not enable a judgment to be made as to whether such financial
means would secure the Austrian objective. There may be features of the registered traders obligations
which can only be effectively enforced against a manager liable in person. It is appropriate, therefore, to
allow the national court to decide whether, in the light of all circumstances and the objectives of GewO, a
financial guarantee would meet the demands of the case. If so, I believe that the Austrian rule is more
restrictive than necessary. If not, I believe it would be justified, in the absence of some other means of
securing the objective of managerial responsibility.
35.
Of course, even the burdensome requirement that a manager provide a guarantee need only be
imposed where enforcement of administrative sanctions cannot otherwise be secured in the member
state where he resides. Similarly, an address for service need only be provided where service at his place
of residence cannot be ensured. Both of these restrictions could be avoided if notification and
enforcement of sanctions were secured through, for example, an international convention. This, indeed, is
recognised by the amended version of para 39 of the GewO, which, as from 1 July 1996, waives the
residence requirement for managers where Austria is a party to such a convention with their country of
residence. It appears that such a convention has been concluded between Austria and the Federal
Republic of Germany, the member state where Mr Henssen resided at the material time, and that it was
applicable at that time6. However, that is a matter for investigation by the national court. If such a
convention was in force, it follows from the courts decision in Wielockx v Inspecteur der Directe
Belastingen Case C-80/94 [1995] All ER (EC) 769, [1995] ECR I-2493 (para 25) that the relevant
mandatory requirement of general interestin that case fiscal cohesionwas sufficiently secured by the
possibility of resort to its terms. The fact that all member states may not have concluded such
conventions is not a bar to relying upon them to reduce as far as possible the restrictions imposed on
freedom of movement of workers, even if there results a difference in the conditions imposed by the host
state on the nationals of the various other member states. Therefore, assuming the 445 effectiveness of
the convention in question, the imposition of a requirement of Austrian residence on Mr Henssen
constituted a disproportionate measure, whose indirectly discriminatory character cannot be justified in
the light of the requirements of the general interest.
6
See the Vertrag zwischen der Republik sterreich und der Bundesrepublik Deutschland ber Amts- und Rechtshilfe in
Verwaltungssachen, Bundesgesetzblatt fr die Republik sterreich, 1990, No 526.

VI CONCLUSION
36.
In the light of the foregoing, I recommend that the Court of Justice answer the questions referred by
the national court as follows:
(1) Employers in the host state derive from art 48 of the EC Treaty the directly effective right to
employ workers who are nationals of another member state without being bound by national rules
which discriminate, directly or indirectly, on grounds of the nationality of the workers in question.
 (2) A national rule under which the owner of a business may only appoint as a manager for
trade-law purposes a person whose residence is in the host state constitutes indirect discrimination
on grounds of nationality.
(3) Such a national rule may be justified by the member states interest in securing compliance
with national rules or administrative decisions regarding the conduct of a registered trade unless
the same objective can be effectively secured by the alternative of a financial guarantee or the
terms of an applicable international convention.

7 May 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By order of 8 October 1996, received at the Court of Justice of the European Communities on 24
October 1996, the Verwaltungsgerichtshof (the Administrative Court) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty two questions on the interpretation of art 48 of that
Treaty and arts 1 to 3 of Council Regulation (EEC) 1612/68 on freedom of movement for workers within
the Community (OJ, S edn 1968 (II) p 475).
2.
Those questions were raised in proceedings brought by Fortress Immobilien Entwicklungs GesmbH,
now Clean Car Autoservice GesmbH (Clean Car), an Austrian company established in Vienna, against
the Landeshauptmann von Wien (the Prime Minister of the Land of Vienna), concerning the rejection of
an application by Clean Car to register for a trade on the ground that it had appointed as manager a
person who did not reside in Austria.

The Austrian legislation

3.
Under para 9(1) of the Gewerbeordnung 1994 (the Austrian Trade Code) (the GewO), legal persons,
commercial-law partnerships (whether general or limited) and registered civil partnerships (whether
general or limited) may exercise a trade, provided that they have appointed a manager or tenant in
accordance with paras 39 and 40 of the GewO.
4.
Paragraph 39 of the GewO provides as follows:
(1) The owner may for the exercise of his trade appoint a manager who is responsible to the
owner for the proper exercise of the trade and to the authorities (Paragraph 333) for compliance
with the provisions of the law on trades; he must appoint a manager if he is not resident in Austria.
(2) The manager must satisfy the personal requirements prescribed for the exercise of the trade,
be resident in Austria, and be in a position to act accordingly in the business. In the case of a trade
for which the production of 446 proof of qualification is prescribed, the manager of a legal person,
to be appointed in accordance with Paragraph 9(1), must also: 1. belong to the statutory
representative organ of the legal person; or 2. be a worker employed in the business for at least
half the normal weekly working hours and subject to full compulsory insurance in accordance with
the provisions of social security law.
The manager of an owner who is not resident in Austria, to be appointed under subparagraph 1
for the exercise of a trade for which the production of proof of qualifications is prescribed, must be a
worker employed in the business for at least half the normal weekly working hours and subject to
full compulsory insurance in accordance with the provisions of social security law. The provisions of
Paragraph 39(2), in force until the coming into force of Federal law BGBl No 29/1993, shall
continue until 31 December 1998 to apply to persons who have been appointed as manager by 1
July 1993.
(3) In cases where a manager must be appointed, the owner must make use of a manager who
acts in the business accordingly.
5.
Pursuant to para 370(2) of the GewO, where the appointment of a manager has been notified or
approved, any fines regarding the conduct of a trade are to be imposed on that manager.
6.
Paragraph 5(1) provides that trades may be exercised on the basis of the application to register the
trade in question, pursuant to para 339, when the general conditions and any specific conditions are
fulfilled, subject to certain exceptions which are not relevant in the present case.
7.
Under para 339(1), any person wishing to exercise a trade, other than a trade for which authorisation
is required and proof of competence other than a proficiency certificate must be produced, must apply for
registration to the administrative authority of the district in which the establishment is situated.
8.
Pursuant to para 340(1), the district administrative authority examines the application for registration
for a trade under para 339(1) to ensure that the statutory conditions for the exercise of the trade applied
for are satisfied by the applicant at the location concerned. If they are not, the district administrative
authority must, under para 340(7), make a finding to that effect by administrative decision and prohibit the
exercise of the trade in question.

The main proceedings

9.
On 13 June 1995, Clean Car applied to the Magistrat der Stadt Wien (the Vienna City Council) to
register for the trade of maintenance and care of motor vehicles (service station) excluding all artisanal
activity. When making that application, it stated that it had appointed Mr Rudolf Henssen, a German
national residing in Berlin, as manager in accordance with the GewO; it further indicated that Mr Henssen
was actively seeking to rent accommodation in Austria and that the declaration relating to his residence
there would be forwarded in due course.
10.
By decision of 20 July 1995, the Magistrat der Stadt Wien found that the statutory prerequisites for the
exercise of that trade were not satisfied and therefore prohibited it, on the ground that the manager must
satisfy the personal conditions laid down for the exercise of the trade in question, must have a residence
in Austria and must be in a position to act effectively as manager in the business, in accordance with para
39(2) of the GewO.
11.
On 10 August 1995, Clean Car lodged an administrative appeal against that decision with the
Landeshauptmann von Wien, submitting that the person 447 appointed as manager now had a residence
in Austria and that, in any event, since the accession of the Republic of Austria to the European Union,
residence anywhere in the European Union was sufficient to satisfy the statutory requirements.
12.
By decision of 2 November 1995, the Landeshauptmann von Wien dismissed the appeal, principally
on the ground that, because of the constitutive nature of the application to register a trade, the material
factual and legal situation was that pertaining at the time the application was lodged, when the person
appointed as manager did not yet have a residence in Austria.
13.
On 21 December 1995, Clean Car brought proceedings before the Verwaltungsgerichtshof, submitting
that the arguments based on Community law had been ignored in the decisions of both the Magistrat der
Stadt Wien and the Landeshauptmann von Wien. Clean Car referred in particular to arts 6 and 48 of the
EC Treaty and submitted that the person whom it had appointed as manager was entitled, as an
employee in its service and thus as a worker, to enjoy the right to freedom of movement established by art
48.
14.
Taking the view that, in order to reach a decision in the case, it must determine whether the Austrian
legislation prohibiting the owner of a trade undertaking from appointing as manager an employee not
resident in Austria is contrary to Community law as laid down in art 48 of the Treaty and arts 1 to 3 of
Regulation 1612/68, the Verwaltungsgerichtshof stayed proceedings and requested a preliminary ruling
from the Court of Justice on the following questions:
(1) Are Article 48 of the EC Treaty and Articles 1 to 3 of Regulation No 1612/68 to be interpreted
as meaning that employers in the host state also derive therefrom the right to employ workers who
are nationals of another Member State without being bound by conditions whicheven if they do
not depend on nationalityare typically linked with nationality?
(2) If employers of the host state have the right stated in Question 1: Are Article 48 of the EC
Treaty and Articles 1 to 3 of Regulation No 1612/68 to be interpreted as meaning that a provision
such as Paragraph 39(2) of the Gewerbeordnung 1994, under which the owner of a trade may
appoint as a manager for trade law purposes only a person whose residence is in the host state
(Austria), is consistent therewith?
15.
In its order for reference, the national court indicates that the first matter to be considered is whether
the provisions of Community law concerning freedom of movement for workers, which are addressed
primarily to workers as such, may also be relied upon by employers. If so, the question then arises
whether those provisions preclude a rule such as that in para 39(2) of the GewO, having regard in
particular to the limitations envisaged in art 48(3) of the Treaty and to the fact that, under para 370(2) of
the GewO, the manager is responsible, in the exercise of the trade, for compliance with the applicable
statutory provisions.

The first question

16.
By its first question, the national court seeks in substance to determine whether the rule of equal
treatment in the context of freedom of movement for workers, enshrined in art 48 of the Treaty and arts 1
to 3 of Regulation 1612/68, may also be relied upon by an employer in order to employ, in the member
state in which he is established, workers who are nationals of another member state.
448
17.
It should be borne in mind, first of all, that arts 1 to 3 of Regulation 1612/68 merely clarify and give
effect to the rights already conferred by art 48 of the Treaty (see, to that effect, Scholz v Opera
Universitaria di Cagliari Case C-419/92 [1994] ECR I-505 (para 6)).
18.
Next, it must be noted that art 48(1) states, in general terms, that freedom of movement for workers is
to be secured within the Community. Under art 48(2) and (3), such freedom of movement is to entail the
abolition of any discrimination based on nationality between workers of the member states as regards
employment, remuneration and other conditions of work and employment, and to entail the right, subject
to limitations justified on grounds of public policy, public security or public health, to accept offers of
employment actually made, to move freely within the territory of member states for that purpose, to stay in
a member state in order to be employed there under the same conditions as nationals of that state and to
remain there after such employment.
19.
Whilst those rights are undoubtedly enjoyed by those directly referred tonamely, workersthere is
nothing in the wording of art 48 to indicate that they may be relied upon by others, in particular employers.
20.
It must further be noted that, in order to be truly effective, the right of workers to be engaged and
employed without discrimination necessarily entails as a corollary the employers entitlement to engage
them in accordance with the rules governing freedom of movement for workers.
21.
Those rules could easily be rendered nugatory if member states could circumvent the prohibitions
which they contain merely by imposing on employers requirements to be met by any worker whom they
wish to employ which, if imposed directly on the worker, would constitute restrictions on the exercise of
the right to freedom of movement to which that worker is entitled under art 48 of the Treaty.
22.
Finally, the above interpretation is corroborated both by art 2 of Regulation 1612/68 and by the courts
case law.
23.
It is made explicitly clear in art 2 of Regulation 1612/68 that any employer pursuing an activity in the
territory of a member state and any national of a member state must be able to conclude and perform
contracts of employment in accordance with the provisions in force laid down by law, regulation or
administrative action, without any discrimination resulting therefrom.
24.
It is, furthermore, clear from, in particular, the judgment in Union Royale Belge des Socits de
Football Association ASBL v Bosman Case C-415/93 [1996] All ER (EC) 97, [1995] ECR I-4921 (paras
8486), that justifications on grounds of public policy, public security or public health, as envisaged in art
48(3) of the Treaty, may be relied upon not only by member states in order to justify limitations on
freedom of movement for workers under their laws, regulations or administrative provisions but also by
individuals in order to justify such limitations under agreements or other measures adopted by persons
governed by private law. Thus, if an employer may rely on a derogation under art 48(3), he must also be
able to rely on the same principles under, in particular, art 48(1) and (2).
25.
In the light of those considerations, the answer to the first question must be that the rule of equal
treatment in the context of freedom of movement for workers, enshrined in art 48 of the Treaty, may also
be relied upon by an employer in order to employ, in the member state in which he is established, workers
who are nationals of another member state.
449

The second question

26.
By its second question, the national court wishes to ascertain, in substance, whether art 48 of the
Treaty precludes a member state from providing that the owner of an undertaking exercising a trade on
the territory of that state may not appoint as manager a person not resident there.
27.
The court has consistently held that the rules of equal treatment prohibit not only overt discrimination
based on nationality but also all covert forms of discrimination which, by applying other distinguishing
criteria, achieve in practice the same result (see, inter alia, Merino Garca v Bundesanstalt fr Arbeit Case
C-266/95 [1997] ECR I-3279 (para 33)).
28.
It is true that a provision such as para 39(2) of the GewO applies without regard to the nationality of
the person to be appointed as manager.
29.
However, as the court has already held (see, inter alia, Finanzamt Kln-Altstadt v Schumacker Case
C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225 (para 28)), national rules under which a distinction is
drawn on the basis of residence are liable to operate mainly to the detriment of nationals of other member
states, as non-residents are in the majority of cases foreigners.
30.
A requirement that nationals of the other member states must reside in the state concerned in order to
be appointed managers of undertakings exercising a trade is therefore such as to constitute indirect
discrimination based on nationality, contrary to art 48(2) of the Treaty.
31.
 It would be otherwise only if the imposition of such a residence requirement were based on objective
considerations independent of the nationality of the employees concerned and proportionate to a
legitimate aim pursued by the national law (see, to that effect, Schning-Kougebetopoulou v Freie und
Hansestadt Hamburg Case C-15/96 [1998] All ER (EC) 97 (para 21)).
32.
In that context, as stated at para 15, above, the national court has expressly referred in its order for
reference to the fact that, under para 370(2) of the GewO, the person appointed as manager is
responsible for compliance with the applicable statutory provisions in the exercise of the trade concerned
and fines may be imposed upon him.
33.
In their written observations, the Landeshauptmann von Wien and the Austrian government have
explained that the residence requirement is intended to ensure that the manager can be served with
notice of the fines which may be imposed upon him and that they can be enforced against him. The
intention is also to ensure that the manager satisfies the other requirement imposed on him by para 39(2)
of the GewO, namely that he must be in a position to act effectively as such in the business.
34.
In that regard, the residence requirement must be held either to be inappropriate for ensuring that the
aim pursued is achieved or to go beyond what is necessary for that purpose.
35.
In the first place, the fact that the manager resides in the member state in which the undertaking is
established and exercises its trade does not itself necessarily ensure that he will be in a position to act
effectively as manager in the business. A manager residing in the state but at a considerable distance
from the place at which the undertaking exercises its trade should normally find it more difficult to act
effectively in the business than a person whose place of residence, even if in another member state, is at
no great distance from that at which the undertaking exercises its trade.
36.
Secondly, other less restrictive measures, such as serving notice of fines at the registered office of the
undertaking employing the manager and ensuring that 450 they will be paid by requiring a guarantee to
be provided beforehand, would make it possible to ensure that the manager can be served with notice of
any such fines imposed upon him and that they can be enforced against him.
37.
Finally, it must be added, even such measures as those just indicated are not justified by the aims in
question if the service of notice of fines imposed on a manager resident in another member state and
their enforcement against him are guaranteed by an international convention concluded between the
member state in which the undertaking exercises its trade and that in which the manager resides.
38.
It must be concluded, therefore, that the residence requirement in question constitutes indirect
discrimination.
39.
As regards the justifications based on art 48(3) of the Treaty, to which the national court has also
referred, it must be observed that a general rule of the kind in issue in the main proceedings cannot be
justified on any grounds of public security or public health.
40.
As regards the justification on grounds of public policy, also envisaged in art 48(3) of the Treaty, the
court has already held that in so far as it may justify certain restrictions on the free movement of persons
subject to Community law, recourse to the concept of public policy as used in that provision presupposes,
in any event, the existence, in addition to the perturbation of the social order which any infringement of
the law involves, of a genuine and sufficiently serious threat affecting one of the fundamental interests of
society (see the judgment R v Bouchereau Case 30/77 [1981] 2 All ER 924, [1977] ECR 1999).
41.
Here, however, it does not appear from the documents in the case that any such interest is liable to be
affected if the owner of an undertaking is free to appoint, for the purpose of exercising that undertakings
trade, a manager who does not reside in the member state concerned.
42.
It is thus also impossible for a national provision such as that in issue in the main proceedings, which
requires any worker appointed as manager for the exercise of a trade to reside in the state concerned, to
be justified on grounds of public policy within the meaning of art 48(3) of the Treaty.
43.
In view of the foregoing considerations, the answer to the second question must be that art 48 of the
Treaty precludes a member state from providing that the owner of an undertaking exercising a trade on
the territory of that state may not appoint as manager a person not resident there.

Costs
44.
The costs incurred by the Austrian government and by the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the action pending before the national court, the decision on
costs is a matter for that court.
 On those grounds, The Court of Justice (Sixth Chamber), in answer to the questions referred to it by
the Verwaltungsgerichtshof by order of 8 October 1996, hereby rules: (1) The rule of equal treatment in
the context of freedom of movement for workers, enshrined in art 48 of the EC Treaty, may also be relied
upon by an employer in order to employ, in the member state in which he is established, workers who are
nationals of another member state. (2) Article 48 of the Treaty precludes a member state from providing
that the owner of an undertaking exercising a trade on the territory of that state may not appoint as
manager a person not resident there.

451

[1998] All ER (EC) 452

Dafeki v Landesversicherungsanstalt Wrttemberg

(Case C-336/94)
EUROPEAN COMMUNITY; Free movement of workers: SOCIAL SECURITY
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RAGNEMALM (PRESIDENT OF THE FOURTH AND SIXTH CHAMBERS, ACTING FOR THE
PRESIDENT), MANCINI, MOITINHO DE ALMEIDA, MURRAY, EDWARD, PUISSOCHET, HIRSCH,
JANN (RAPPORTEUR) AND SEVN ADVOCATE GENERAL LA PERGOLA
22 OCTOBER, 3 DECEMBER 1996, 2 DECEMBER 1997
European Community Freedom of movement Workers Social security Discrimination on grounds
of nationality Greek national claiming early retirement benefit in Germany Benefits refused on
grounds that claimants age not proved by recently rectified Greek birth certificate Whether Community
law requiring national authorities to recognise certificates and analogous civil status documents issued by
other member states EC Treaty, art 48.

The applicant, a Greek national, had worked in Germany since May 1966. Initially her civil status
documents gave her date of birth as 3 December 1933, but, in 1986, that date was rectified by a Greek
district court and she was issued with a new birth certificate showing her date of birth as 20 February
1929. In 1988 she applied to the German pension fund for the early retirement benefit for women who had
reached the age of 60, producing for that purpose the new birth certificate and, at the request of the
pension fund, the judgment of the Greek court ordering rectification. Under German provisions on civil
status, certificates drawn up in another country did not benefit from the usual presumption of accuracy so
that a court seised of the matter had to evaluate the document before it and, in the event of inconsistency
between several documents of differing dates, there was a presumption that the one closest in time to the
event prevailed in the absence of other evidence. Although the applicant satisfied all the other conditions
of entitlement, the pension fund refused her application, basing its decision on the date of birth before
rectification. Thereafter the applicant brought an action before the Sozialgericht Hamburg, which stayed
the proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling
the question whether, in proceedings for determining entitlements to social security benefits of a migrant
worker who was a Community national, art 481 of the Treaty required that the competent social security
authorities and the courts of a member state recognise certificates and analogous documents concerning
civil status issued by the competent authorities of other member states.
1
Article 48, so far as material, provides: (2) freedom of movement shall entail the abolition of any discrimination based
on nationality between workers of the Member States as regards employment, remuneration and other conditions of
work and employment.

Held The possibility of successfully challenging the accuracy of a certificate of civil status depended not
only on the procedures followed but also on the conditions which had to be satisfied in order for such a
certificate to be altered, which could vary considerably from one member state to another. As a result,
452Community law did not require the administrative and judicial authorities of a member state to treat as
equivalent subsequent rectifications of certificates of civil status made by the competent authorities of
their own state and those made by the competent authorities of another member state. However, those
authorities had to accept certificates and analogous documents concerning personal status issued by the
competent authorities of other member states, unless their accuracy was seriously undermined by
concrete evidence regarding the individual in question; to rule otherwise would compromise the exercise
of rights arising from the freedom of movement for workers. In those circumstances, a national rule
establishing a presumption that, in the event of inconsistency between several documents of differing
dates, the document closest in time to the event to be proved prevailed in the absence of other sufficient
evidence could not justify refusal to take account of a rectification made by a court in another member
state (see p 465 a to j, post).

Notes
For equal treatment in the context of the material scope of a workers right to free movment, see 52
Halsburys Laws (4th edn) para 1513.
For the EC Treaty, art 48, see 50 Halsburys Statutes (4th edn) 283.
Cases cited
Amministrazione delle Finanze dello Stato v Ariete SpA Case 811/79 [1980] ECR 2545.
Amministrazione delle Finanze dello Stato v Denkavit Italiana Srl Case 61/79 [1980] ECR 1205.
Amministrazione delle Finanze dello Stato v San Giorgio Case 199/82 [1983] ECR 3595.
Amministrazione delle Finanze dello Stato v Sas Mediterranea Importazione Rappresentanze
Esportazione Commercio (MIRECO) Case 826/79 [1980] ECR 2559.
Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043.
Hans Just I/S v Danish Ministry for Fiscal Affairs Case 68/79 [1980] ECR 501.
Konstantinidis Case C-168/91 [1993] ECR I-1191.
Micheletti v Delegacin del Gobierno en Cantabria Case C-369/90 [1992] ECR I-4239.
Mund & Fester v Hatrex International Transport Case C-398/92 [1994] ECR I-467.
Paletta v Brennet AG Case C-45/90 [1992] ECR I-3423.
Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989.

Reference
By order of 12 September 1994, the Sozialgericht (the Social Court) Hamburg referred to the Court of
Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty a question
(set out at p 463 j, post) on the interpretation of arts 48 and 51 of that treaty in the light of German
provisions under which certificates of civil status were accorded different probative value, depending on
whether they were German or foreign. That question was raised in proceedings between Mrs Dafeki and
the Landesversicherungsanstalt Wrttemberg. Written observations were submitted on behalf of: Mrs
Dafeki, by J S Politis, of the Athens Bar; the German government, by E Rder, Ministerialrat in the
Federal Ministry of the Economy, and B Kloke, Regierungsrat in the same ministry, acting as agents; the
Greek government, by P Kamarineas, Legal Adviser at the State Legal Council, K Grigoriou, legal
representative to the state Legal Council, and I Galani-Maragkoudaki, Deputy Special Legal Adviser in the
Special Community Legal Affairs Department of the Ministry of Foreign Affairs, acting as 453 agents; and
the European Commission, by J Sack, Legal Adviser, acting as agent. Oral observations were made by:
Mrs Dafeki, represented by J S Politis; the Landesversicherungsanstalt Wrttemberg, represented by E
Graner, Regierungs- direktor, acting as agent; the German government, represented by S Maa,
Regierungsrtin zur Anstellung in the Federal Ministry of the Economy, acting as agent; the Greek
government, represented by F Georgakopoulos, Deputy Legal Adviser at the State Legal Council, acting
as agent, and I Galani-Maragkoudaki; and the Commission, represented by J Sack. The language of the
case was German. The facts are set out in the opinion of the Advocate General.

3 December 1996.

The Advocate General (A La Pergola)

delivered the following opinion (translated from the Italian).
1.
By order of 12 September 1994 the Sozialgericht (the Social Court), Hamburg referred the following
question to the Court of Justice for a preliminary ruling:
Whether, and to what extent, German social security authorities and courts are bound by the
rule of Community law to the effect that foreign certificates of civil status, and judgments of foreign
courts determining or rectifying data concerning matters of civil status, have binding force in
proceedings concerning entitlement to social security benefits.
2.
The case which gave rise to this question may be summarised as follows: Mrs Dafeki, a Greek
national, has been living in Germany since 1966 and was an employed person in that country until 1987.
Her identity papers showed her date of birth as 3 December 1933.
On 4 April 1986, at Mrs Dafekis request, the court in Trikala (Greece) rectified that date in accordance
with the special procedure provided for where national archives and registers of civil status were
destroyed during the war. Mrs Dafekis date of birth was consequently established as 20 February 1929.
The population register of the municipality of Glikomilea and the register of civil status of the municipality
of Chrissomilea were amended accordingly. Mrs Dafeki was then issued with a birth certificate attesting
her date of birth as rectified.
On 19 December 1988, Mrs Dafeki applied in Germany for the early retirement pension provided for
women who have reached the age of 60. Her application was, however, rejected. Basing its decision on
the documents drawn up before Mrs Dafekis date of birth had been rectified, the competent German
institution decided that Mrs Dafeki had not reached the age required for pension entitlement at the time
when her application was made. Mrs Dafeki contested that decision by proceedings before the
Sozialgericht Hamburg.
3.
The national court is asking the Court of Justice whether the German social security institution and any
courts seised of the matter are obliged to take into account the civil-status documents of another member
state and alterations made to such documents by the competent authorities of that state, including the
courts.
According to the national court, under German procedural and substantive law that question has to be
answered in the negative. German civil-status documents and rectifications made to them enjoy a
privileged authority because, provided that they comply with the prescribed formal requirements, they
benefit from a presumption of accuracy as to the matters to which they attest, subject however to any
interested party adducing evidence to the contrary. The same rules of evidence do not, however, apply to
foreign documents; they are subject to free evaluation by the courts which are, consequently, able to
disregard the matters 454 certified in them. Additionally and more particularly, according to the order for
reference, in the present case a rule of case law applies according to which if documents of differing
dates are inconsistent with one another, the document closest in time to the event to be proved takes
precedence. Consequently, the national court attaches greater probative value to Mrs Dafekis old identity
card than to the new documents issued after her date of birth had been rectified. The national court does,
however, raise the question whether or not the fact that German and foreign documents relating to civil
status are accorded different probative value constitutes discrimination based on nationality prohibited by
the Treaty and in particular by art 48 thereof. It has therefore submitted the above-mentioned question to
the court.
4.
The order for reference thus formulates the question for interpretation essentially in terms of the
prohibition of discrimination based on nationality laid down in art 48(2) of the Treaty.
On that point, both the representative of Mrs Dafeki and the Greek government submit, albeit for
different reasons, that the question referred should be answered in the affirmative: they consider this to
be a case of indirect discrimination contrary to arts 6, 48 and 51 of the Treaty and to art 3(1) of Council
Regulation (EEC) 1408/71 on the application of social security schemes to employed persons and their
families moving within the Community (OJ S Edn 1971 (II) p 416) which specifically applies to the field of
social security the general principle of non-discrimination. They submit that that principle requires the
competent authorities and the German courts to accord to Greek civil-status documents the same value
as the equivalent German documents.
Although the European Commission and the German government propose different replies as regards
the substance of the question referred, neither considers that there has been a breach of the principle of
non-discrimination in this case. That principle, they agree, requires that like situations should not be
treated differently. The point in this case, however, is, they maintain, that the situations are not the same.
And it is in fact true that the probative value of documents relating to civil status and of rectifications
thereof differs from one legal system to another according to the formalities and procedures prescribed for
them. Inasmuch as, it is argued, the differing treatment prescribed under the national rules in point
specifically concerns situations which are not the same, it does not constitute discrimination of the kind
alleged.
I subscribe to the last-mentioned view. Admittedly, the rules of evidence described by the national
court are, in theory, such as to affect the exercise of a freedom safeguarded by the Treaty. In other words,
they have a negative effectif only indirectlyon the legal position of a citizen exercising the right to
freedom of movement. Proof of status as regards age is a de facto prerequisite for the exercise of that
right, and, in Germany, it is rendered more difficult for nationals of other member states of the Community
to provide such proof than it is for German nationals. In that sense there is a difference in treatment, but
the sole factor relevant here is that this is not discrimination prohibited by Community law. This is because
each state makes itself responsible for the documentation of matters concerning the status of persons
subject to its authority by maintaining appropriate registers and issuing certificates. Both of those activities
are carried out in accordance with specific procedures and in compliance with the formalities laid down for
this. This explains the special probative value which attaches to the registers of civil status and to the
documents issued by the authorities vested with the power of certification in each legal system under
which these matters fall. The 455 fact is that, in the field with which we are concerned, form and probative
value are indissolubly linked, in the sense that the latter is dependent on the former.
Consequently, a distinction has clearly to be made in this case between German public documents on
the one hand and the equivalent documents in Greece or any other Community country on the other.
That, in my view, is the reason for which it is not possible to extend generally to documents from other
member states the effect which German law attaches to equivalent German documents.
5.
That, however, does not mean, in addition, that the German administrative authorities and courts are
free to disregard the content of foreign certificates or that the procedure prescribed for dealing with such
certificates, as described in the order for reference, must be deemed to be compatible with Community
law. Far from it, as I shall explain below. The plaintiff in the main proceedings has exercised a freedom
guaranteed by the Treaty, namely to work as an employed person in another member state, as a result of
which she is entitled to claim social security benefits in that state. The question of status in issue before
the national court concerning her date of birth involves a de facto prerequisite for the exercise of that
right. It is true that
in the absence of any relevant Community rules, it is for the national legal order of each
Member State to lay down the procedural rules for proceedings designed to ensure the
protection of the rights which individuals acquire through the direct effect of Community law
(See the judgment in Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case
33/76 [1976] ECR 1989 (para 5).)2

2
See also the judgments in Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043 (para 13), Hans
Just I/S v Danish Ministry for Fiscal Affairs Case 68/79 [1980] ECR 501, Amministrazione delle Finanze dello Stato v
 Denkavit Italiana Srl Case 61/79 [1980] ECR 1205, Amministrazione delle Finanze dello Stato v Ariete SpA Case 811/79
[1980] ECR 2545, Amministrazione delle Finanze dello Stato v Sas Mediterranea Importazione Rappresentanze
Esportazione Commercio (MIRECO) Case 826/79 [1980] ECR 2559 and Amministrazione delle Finanze dello Stato v
San Giorgio Case 199/82 [1983] ECR 3595 (para 12).

However, that principle of procedural autonomy is circumscribed by the requirement, emphasised by the
Court of Justice on several occasions, that the procedural rules laid down by the competent state must
not make it impossible in practice to exercise the right conferred by Community law 3.
3
Ibid.

And it is specifically in the light of that requirement that the present case falls to be examined. What
rules of evidence does the national court consider it has to apply here? It cites a rule of case law under
which, if two documents of differing dates are inconsistent with one another, it is the one which is closest
in time to the event to be proved which prevails. It is in application of that rule that the national court
considers that it has to attach greater probative value to Mrs Dafekis old identity card than to the new
birth certificate issued to her following a judicial decision rectifying her date of birth. Since the document
resulting from that rectification is obviously of later date than the document that was rectified, Mrs Dafeki
is precluded from proving the alteration in status, which, in turn, constitutes an indispensable prerequisite
for asserting the right to receive the pension guaranteed to her under Community law. Furthermore, and
quite apart the above-mentioned rule of case law, the German authorities refusal to take account of the
certificates submitted by Mrs Dafeki means that it is in practice impossible for her to be awarded the
pension. She has to prove that she has 456 reached the age of sixty, and I fail to see how she can do this
other than by relying on the documents issued by her country of birth. After all, it cannot be said that Mrs
Dafeki has not made every effort to provide proof of her age: she submitted to the court the new birth
certificate issued by the competent authorities of her country of origin together with the court order
constituting the authority for the rectification. Were any further condition to be imposed, it would, in my
view, be either impossible or at least extremely difficult to meet. It would therefore be impermissible under
the above-mentioned case law.
In essence, I consider that the rules of evidence that the national court intends applying in this case
make it impossible in practice to prove a fact which is a necessary prerequisite for the exercise of the right
in question. Those rules must therefore be disapplied by the national court.
6.
It remains to be clarified how, for the purposes of Community law, the German court is to evaluate, in a
case of the kind described in the order for reference, documents relating to civil status issued in another
member state. A preliminary point needs to be made here. There are circumstances in which an
individuals civil status is a condition for entitlement to a subjective legal position guaranteed under
Community law. Can it be accepted in such circumstances that the status of the person concerned should
be evaluated differently from one member state to another? The answer, in my view, has to be that it
cannot.
It cannot be accepted that an individuals status, in the sense of his legal position within the legal
system in questionin this instance the Community legal systemshould be evaluated differently
depending on the law of the state in which he is residing or working within the territory of the Community.
Were that to be the case, the way in which the competent national authorities treat the events which are
relevant for the purpose of defining the status of the person concerned would determine whether that right
of the individual was recognised or denied. That is incompatible with the basic concept underlying the
Treaty according to which subjective legal positions under Community law must enjoy equal recognition,
that is to say it must be possible to invoke them in the same way in every member state of the
Community. In other words, the immutability of statuswhenever, of course, it constitutes an element of
or prerequisite for a right of the individualderives from the necessity to guarantee in a uniform manner
the actual form of subjective legal positions under Community law and their protection. It would be
contrary to the very idea of integration were a right to exist and be enforceable in one member state but
not in another because the civil status of the person concernedand more specifically his age, upon
which that right dependsis assessed differently within the Community, though the Community itself is
conceived, also for the purposes under consideration here, as being a single area without internal
frontiers.
7.
The Court of Justice has already affirmed this principle of the uniformity of the status of Community
citizens in the judgment in Micheletti v Delegacin del Gobierno en Cantabria Case C-369/90 [1992] ECR
I-4239. The question raised in that case was whether Spain could refuse to accept, on the basis of its own
laws, the status of Mr Micheletti as an Italian citizen. The answer was that it could not: it is not
permissible for the legislation of a member state to restrict the effects of the grant of the nationality of
another Member State (see [1992] ECR I-4239 (para 10)). The reason for this was that the consequence
of allowing such a possibility would be that the class of persons to whom the Community rules on
freedom of establishment were applied might vary from one Member State to another (see [1992] ECR I-
4239 (para 12); my emphasis). In support of that interpretation, the court 457 cited Council Directive
(EEC) 73/148 on the abolition of restrictions on movements and residence within the Community for
nationals of member states with regard to establishment and the provision of services (OJ 1973 L172 p
14) which
(13) provides that Member States are to grant to the persons referred to in Article 1 the right
to enter their territory merely on production of a valid identity card or passport (Article 3) and are to
issue a residence card or permit to such persons, and to those mentioned in Article 4, upon
production, in particular, of the document with which they entered their territory (Article 6).
 (14) Thus, once the persons concerned have produced one of the documents mentioned in
Directive 73/148 in order to establish their status as nationals of a Member State, the other Member
States are not entitled to challenge that status (See [1992] ECR I-4239 (paras 1314).)
Two conclusions can be drawn from the judgment in Micheletti [1992] ECR I-4239 (paras 1314). First,
as Advocate General Tesauro rightly pointed out in his opinion
the provisions of Community law which require an individual to possess the nationality of a
Member State as a prerequisite for their application must be understood as referring to the national
law of the State whose nationality serves as the basis of the right relied upon. (See [1992] ECR I-
4239 (para 3).)
Second, once it has thus been established that a persons status is determined by the law of the state
to which he belongs and, consequently, that it is the authorities of that state which have the power to
certify that status, the determinations made by those authorities must be accepted uniformly in all the
member states and cannot be the subject of differing assessments 4.
4
Confirmation of the considerations so far set out is to be found in the judgment in Konstantinidis Case C-168/91 [1993]
ECR I-1191. The issue in that case was whether it was compatible with Community law for a member state to enter the
name of a Greek citizen in its own registers using a system of transliteration that had the effect of substantially altering
the name of the person concerned. The courts negative answer to that question was essentially based on the
consideration that that transliteration exposed the Greek citizen to the risk that potential clients may confuse him with
other persons (see [1993] ECR I-1191 (para 16)). That case clearly differed from the present case. The court did,
however, take account of the need for the citizen to be able to move in the Community with the name accorded him in
his country of origin. In that case too, therefore, the court confirmed the principle that an individual has to be treated as
the person he is with all of the personal attributes that distinguish him in the legal system of his country of origin.

8.
As I see it, although the Micheletti judgment concerns status civitatis, there are good reasons for
taking the view that the considerations set out by the court in that case may be applied by analogy to this
case. In providing that Member States shall grant to the persons referred to in Article 1 the right to enter
their territory merely on production of a valid identity card or passport art 3 of Directive 73/148 takes
account of a fundamental requirement for the effective realisation of the integration process 5: Community
citizens may move freely, 458taking with them all the elements needed to identify them as subjects, as
persons, and which are certified in the documents relating to their civil status. They must therefore be
able to rely on the documents issued to them by the competent authorities of their country of origin. That
is whyagain, of course, for the limited purposes of the exercise of the rights guaranteed under
Community lawrecognition must be accorded to the principle of the uniformity of a persons civil status;
he must be able to move within the Community precisely as he is individualised by that status, that is to
say with the relationships that are determined on the basis of age, sex, status civitatis, and so on. Only in
that way can citizens effectively exercise freedom of movement, now establishedby designas a right
of citizenship of the Union, throughout the entire territory of the Community.
5
We find that same approach in Council Directive (EEC) 68/360 on the abolition of restrictions on movement and
residence within the Community for workers of member states and their families (OJ S edn 1968, p 485). Article 3(1) of
the directive provides: Member States shall allow the persons referred to in Article 1 to enter their territory simply on
production of a valid identity card or passport. Article 4(1) then goes on to provide: Member States shall grant the right
of residence in their territory to the persons referred to in Article 1 who are able to produce the documents listed in
paragraph 3.

In this caseas the Commission has rightly pointed outthe proper functioning of the system
established by Regulation 1408/71 is necessarily based on implicit reference to the national legal systems
in regard to all those elements that are needed in order to determine the rights of workers and,
consequently, on reference to the documents concerning status which are issued to the person in
question by the authorities vested with the power of certification. In other words, those aspects of
personal status that affect the acquisition or exercise of a right under Community law are not governed
directly by the Community legal order but are implicitly and of necessity a matter for the legal system of
the particular member state that determines the status of the person concerned. A court or administrative
authority of a member state that is called upon to consider a case concerning an aspect of a persons
status that is governed by the law of another member state may not therefore disregard the facts as
determined by the authority vested, under that law, with the power to bring about, with respect to the
aspect of status in point, a special form of legal certainty and cause it to have effect in legal relations.
9.
At the hearing the Commission conceded that determining a persons age is a question of status but
ruled out the possibility of extending to this case the principle laid down in Micheletti in regard to
citizenship. To be more precise, the Commission acknowledges that it is a matter for the state of which
the person concerned is a national to determine and certify status. It adds that the authorities of the host
member state are normally required not to challenge that certification but considers that those authorities
may disregard the content of the document in cases of doubt. According to the Commission, that is
precisely the case of Mrs Dafeki who did not ask for rectification of her age until shortly before she applied
for early retirement benefit. Thataccording to the Commissionprovides grounds for more than mere
suspicion.
That is a view which, to be frank, I find baffling. In my opinion, if there are doubts concerning the scope
and effectiveness of the new birth certificate produced by Mrs Dafeki, the only avenue open to the
administrative and judicial authorities of the state concerned is to have recourse to the arrangements for
co-operation provided for by art 84 of Regulation 1408/71 and seek the requisite clarification from their
counterparts in the country in which the document in question was issued. If the latter confirm that the
certificate in question was indeed issued by the competent authority and produces legal certainty as
regards the age of the person concerned, that aspect may no longer be called in question.
In other words, the national court must confine itself to verifying that the certificate was issued by the
competent authorities and produces, within the legal system relied upon by the person concerned, the
requisite effects of legal certainty 459 as regards his age; it cannot hold the content of such certificates to
be unreliable, citing its freedom under its national law to evaluate evidence. Once the certificates have
been issued by the competent authorities of the state concerned, the principle of good faith and mutual
trust which has, pursuant to art 5 of the Treaty, to inform relations between the authorities of the different
member states, means that the accuracy and reliability of the data certified may not be challenged.
The opposite approach would be hard to reconcile with the need to ensure a relationship of trust and
scrupulous co-operation between authorities and institutions in the Member States; the importance of the
latter principle was rightly pointed out by Advocate General Gulmann in Paletta v Brennet AG Case C-
45/90 [1992] ECR I-3423 at 3453. The court held in the judgment in that case that
the competent institution is bound in fact and in law by the medical findings made by the
institution of the place of residence or temporary residence concerning the commencement and
duration of the incapacity for work, when it does not have the person concerned examined by a
doctor of its choice, as it may do under Article 18(5). (See [1992] ECR I-3423 (para 28).)
It is true that that case was covered by a specific provision of Council Regulation (EEC) 574/72 fixing the
procedure for implementing Council Regulation (EEC) 1408/71 on the application of social security
schemes to employed persons and their families moving within the Community (OJ S edn 1972 (I) p 159).
However, as the Advocate General rightly pointed out, that was a problem that involved a fundamental
principle
namely that there should be scrupulous cooperation between the institutions in the Member
States, built on mutual trust (see in this regard Article 84 of Regulation No 1408/71, in conjunction
with Article 5 of the EEC Treaty) and that the authorities in one Member State must recognize the
correctness of certificates issued by the authorities in other Member States (See [1992] ECR I-
3423 (para 12).)
10.
In my view, the solution to this case follows from the above considerations. Once it is accepted that the
question of Mrs Dafekis age is governed by Greek law and that the Greek authorities are vested with the
power of certification, it necessarily follows that those same authorities are also vested with the
consequential power to rectify the document they have issued. Reference to national law covers not only
the initial determinations but also any subsequent amendments to them. The power to rectify and the
power to certify are of the same legal nature. It is not possible to recognise the one and deny the other.
It therefore follows that both the German social security institution and the national court are bound by
the content of the administrative and judicial documents issued by the competent Greek authorities
relating to the status of Mrs Dafeki. This is because it is for those authorities alone to establish once and
for all and erga omnes Mrs Dafekis age, so far as concerns the exercise of rights arising out of
Community law.
11.
The rectification has therefore to be accorded the same value as is attached to it in the legal system of
the state of origin. The German government has contended that the rectification requested and obtained
by Mrs Dafeki is not valid in Greek law for the purpose of obtaining social security benefits 6; consequently,
were that possibility to be accepted in Germany, it would produce the absurd result of according a foreign
certificate greater value than is attached to it under 460 the law of its state of origin. I too agree that the
rectified document must be accorded an effect which is neither greater nor lesser than that which it
produces in the legal system under which it was drawn up. It is then for the national court, which has to
adjudicate on the substance of the case, to ascertain what that effect is and determine, in particular,
whether what is concerned is a document that creates legal certainty such that its contents must be held
to be binding. It is for the national court to inquire whether this is so. The Court of Justice can only
indicate to that court the principle to be applied. As I have said, that principle is to the effect that reference
must be made to the competent legal system, and exclusively, let it be said, as regards that aspect of
personal status upon which, in this case, the recognition by the German authorities of a right guaranteed
under Community law depends. The national court will have to ascertain whether the certificate submitted
by Mrs Dafeki in Germany constitutes proof under Greek law that she has reached the age of 60 years.
Whether the social security institution in the country of origin might, under the law of that country, require
other documents to be submitted or further formalities to be completed for pension purposes is a question
that does not fall within the ambit of the proceedings now before the court.
6
It need hardly be mentioned that the assumption on which the German government bases its argumentnamely that
the rectification obtained by Mrs Dafeki is of limited probative value in Greecewas discussed at the hearing but was in
no way substantiated and would in fact appear to have been denied by the Greek government, both in its written reply to
the question put to it on this subject by the Court of Justice and in its oral argument. That point is, in any event,
completely irrelevant for the purposes of a decision by the court, which is not required to rule on the substance of the
main proceedings. It will be for the national court to evaluate the effect that has to be accorded to the certificates in
question under the legal system under which they were issued.

12.
It must next be said that the case before the court does not in any way touch, as the Commission
appears to think, upon the thorny problem of the automatic recognition of judgments. The Commission
has stated that if the German social security institution were to accord to the rectification of Mrs Dafekis
age the same value as is accorded to it in Greece, this would in fact be tantamount to the mutual
recognition of judgments in matters of civil status. According to the Commission, the courts case law,
particularly Mund & Fester v Hatrex International Transport Case C-398/92 [1994] ECR I-467, shows that
such recognition may be accorded only on the basis of the appropriate international agreements; in this
case Germany and Greece are not bound by any instrument of public international law in the field in
question.
We have, however, to consider what are the circumstances in the present case. In the first place, Mrs
Dafeki submitted to the German institution a rectified birth certificate and not a judgment. According to the
order for reference, the decision of the Athens court merely constituted the authority for making that
rectification.
That aside, even if one were willing to contemplate the hypothesis that Mrs Dafeki had produced only
the judgment rectifying her birth certificate, I do not consider that this in itself would mean that there was
any question of automatic recognition of judgments in the absence of an appropriate agreement. Mrs
Dafekis request does not seek to have the Greek judgment enforced in Germany, that is to say to
establish in German law the legal situation brought about by that judgment, but to exercise a right for
which that legal situation (as actually produced) forms the prerequisite. In other words, the judgment is
not in this case to be considered in terms of its significance as a judicial decision; it is the findings to
which that judgment gives specific effect that have to be taken into account by 461 the German
authorities. The reason for this, as I have already explained, is that for the purposes in point here the
question of status is governed by the law of the state to which the person concerned belongs. That being
so, Mrs Dafekis age as established by judicial decision in her own country constitutes, for the German
authorities, a determination already made and given concrete form by the only authorities empowered to
do so. The situation would be different were Mrs Dafeki to seek to avail herself of the foreign judgment
from the point of view of its significance as a judicial decision, as would, for example, be the case if she
were primarily applying to have her age as rectified entered in the German registers of civil status. In that
case, in the absence of appropriate international agreements on the automatic recognition of such
judgments, there would have in fact to be an examination of the foreign judgment or other equivalent
procedure under German law.
13.
In conclusion, I do not find that this case discloses any prohibited discrimination of the kind envisaged
by the national court. German public documents are not the same as Greek public documents. It is
therefore not possible to extend to the latter the same probative value as that accorded to the former
under German procedural law. None the less, both the social security institution and the national court
have a duty to take into account documents relating to civil status issued to the person concerned by the
competent authorities of his country of origin. More specifically, they must accord to those documents the
same value as that attributed to them under the legal system under which they were drawn up. It is then
for the administrative authority and the court to evaluate the significance and effects of the document
under that legal system. Should the authorities of a member state have doubts concerning the legality or
effect of the document, they must consult their counterparts in the other member state and ask them to
provide the necessary clarification.

Conclusion
14.
The reply to the question referred to the Court of Justice by the Sozialgericht Hamburg should
therefore be:
In proceedings concerning entitlement to social security benefits, German social security
institutions and courts are not free to disregard the content of documents relating to civil status that
determine or rectify particulars of a Community citizens civil status and have been issued to the
person concerned by the competent authorities of their country of origin. Those institutions and
courts must accord to those certificates the same value as that attached to them under the legal
system of that country.

2 December 1997.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 12 September 1994, received at the Court of Justice of the European Communities on 28
December 1994, the Sozialgericht (the Social Court) Hamburg referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty a question on the interpretation of arts 48 and 51 of the
Treaty in the light of German provisions under which certificates of civil status are accorded different
probative value, depending on whether they are German or foreign.
2.
That question was raised in proceedings between Mrs Dafeki and the Landesversicherungsanstalt
Wrttemberg (a German retirement pension fund).
462
3.
Mrs Dafeki was born in Greece and has Greek nationality. She has worked in Germany since May
1966. Her civil-status documents gave her date of birth as 3 December 1933. By judgment of a single
judge of the District Court of Trikala of 4 April 1986, that date was rectified in accordance with the
procedure applicable where archives and registers have disappeared. Since then, the register of civil
status and Mrs Dafekis civil-status documents have shown that she was born on 20 February 1929. She
was accordingly issued with a new birth certificate.
4.
On 19 December 1988 Mrs Dafeki applied to the pension fund for the early retirement benefit for
women who have reached the age of 60. For that purpose, she produced first of all the new birth
certificate issued by the competent Greek authorities and then, at the request of the pension fund, the
judgment ordering rectification. Although she satisfied all the other conditions of entitlement, the pension
fund refused her application, basing its decision on the date of birth before rectification. Since her
subsequent objection was also rejected, Mrs Dafeki brought an action before the Sozialgericht Hamburg.
5.
In German law, para 66 of the Personenstandsgesetz (the Law on Civil Status) (the PStG) provides
that documents relative to civil status have the same probative value as registers of civil status; according
to para 60(1) of the PStG, provided that they are properly maintained, those registers constitute in
principle proof of marriages, births and the particulars entered in relation to those events. Evidence of
their inaccuracy may however be adduced. According to the case law of the Bundessozialgericht (the
Federal Social Court) and academic legal writing, para 66 of the PStG applies only to German and not to
foreign documents, including those relating to subsequent rectifications. It follows that where certificates
have been drawn up in another country, they do not benefit from the presumption of accuracy, so that the
court seised of the matter proceeds to an evaluation of the documents before it in accordance with the
rule of free assessment of evidence. In so doing, the court must take account in particular of a rule of
case law which establishes a presumption that, in the event of inconsistency between several documents
of differing dates, the document which prevails is generally, in the absence of other sufficient evidence,
the one closest in time to the event, and hence, in this case, the first extract from the register of births.
6.
The Sozialgericht Hamburg questions whether application of the rule of free assessment of evidence
to the probative value of certificates of civil status is compatible with Community law, in particular with arts
48 and 51 of the Treaty, as amounting to indirect discrimination on grounds of nationality. If Mrs Dafeki
had produced documents issuing from the German register of civil status, her rectified date of birth would
have been accepted without further enquiry.
7.
The Sozialgericht Hamburg accordingly decided to stay proceedings and refer the following question
to the Court of Justice for a preliminary ruling:
Are German social security authorities and courts bound, and if so to what extent, by the rule of
Community law to the effect that foreign certificates of civil status, and judgments of foreign courts
determining or rectifying data concerning matters of civil status, have binding force in proceedings
concerning entitlement to social security benefit?
8.
By its question, the national court is in substance asking whether, in proceedings for determining
entitlements to social security benefits of a migrant worker who is a Community national, art 48 of the
Treaty requires that the competent social security authorities and the courts of a member state recognise
463 certificates and analogous documents relative to civil status which have been issued by the
competent authorities of other member states.
9.
It should be borne in mind at the outset that, by virtue of art 48(2) of the Treaty, freedom of movement
for workers entails the abolition of any discrimination based on nationality between workers of the
member states as regards employment, remuneration and other conditions of work and employment.
10.
The situation of Mrs Dafeki, a national of a member state who has been employed in another member
state where she seeks the award of a retirement pension on the basis of that employment, falls within the
scope of that provision.
11.
In order to invoke the right to a social security benefit flowing from the exercise of the right of free
movement for workers guaranteed by the Treaty, workers must necessarily supply proof of certain
particulars entered in the registers of civil status.
12.
It is clear from the provisions of German law, as set out by the national court, that the probative value
accorded by those provisions to certificates of civil status issued by the competent authorities of another
member state is lower than that accorded to certificates drawn up by the German authorities.
13.
Thus, although they apply irrespective of the nationality of the worker, those rules operate in practice
to the detriment of workers who are nationals of other member states.
14.
The German government submits, however, that there are significant differences between the member
states as regards the provisions governing the maintenance and rectification of registers of civil status, in
view of the widely varying factual circumstances and legal considerations affecting legislative decisions.
In particular, the rules of authentication are not the same in the Hellenic Republic and the Federal
Republic of Germany. In the former state, amongst other things, alteration of a date of birth by judgment
of a single judge, for which the evidence of two witnesses suffices, is not uncommon. Not a few migrant
workers of Greek nationality have availed themselves of this possibility. The competent German
insurance institution has noted some hundreds of cases in which the date of birth declared on taking up
employment differed significantly from the date given on application for the award of a pension. As a
general rule, the alteration operates to the workers advantage.
15.
The Commission too points out that questions relating to civil status differ considerably from one
member state to another, since the respective systems have been strongly influenced by an extremely
wide variety of cultural phenomena and by various external events, such as wars and transfers of territory.
It is therefore difficult to start from the premiss that the factual and legal situations are identical or
equivalent. There are no common measures at Community level. Moreover, the Community has no
general competence to lay down rules concerning the law applicable to civil status or questions related to
the probative value of documents relative to civil status. In those circumstances, the Commission
considers that, as Community law now stands, it does not preclude the German practice.
16.
Account must be taken, first, of the considerable differences that exist between the national legal
orders as regards the conditions and procedures for rectification of a date of birth and, second, of the fact
that, for the time being, the member states have neither harmonised the matter nor established a system
of mutual recognition of such decisions, as has been done for judgments falling within the scope of the
Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968.
464
17.
The possibility of successfully challenging the accuracy of a certificate of civil status, such as that in
issue in the main proceedings, depends to a large extent on the procedure followed and on the conditions
which have to be satisfied in order for such a birth certificate to be altered. These may vary considerably
from one member state to another.
18.
Consequently, the administrative and judicial authorities of a member state are not required under
Community law to treat as equivalent subsequent rectifications of certificates of civil status made by the
competent authorities of their own state and those made by the competent authorities of another member
state.
19.
Nevertheless, exercise of the rights arising from freedom of movement for workers is not possible
without production of documents relative to personal status, which are generally issued by the workers
state of origin. It follows that the administrative and judicial authorities of a member state must accept
certificates and analogous documents relative to personal status issued by the competent authorities of
the other member states, unless their accuracy is seriously undermined by concrete evidence relating to
the individual case in question.
20.
In those circumstances, a rule of national law which establishes a general and abstract presumption
that, in the event of inconsistency between several documents of differing dates, it is the document
closest in time to the event to be proved which prevails in the absence of other sufficient evidence, cannot
justify refusal to take account of a rectification made by a court in another member state.
21.
The answer to be given to the question put to the court must therefore be that, in proceedings for
determining the entitlements to social security benefits of a migrant worker who is a Community national,
the competent social security institutions and the courts of a member state must accept certificates and
analogous documents relative to personal status issued by the competent authorities of the other member
states, unless their accuracy is seriously undermined by concrete evidence relating to the individual case
in question.

Costs
22.
The costs incurred by the German and Greek governments and by the European Commission, which
have submitted observations to the Court of Justice, are not recoverable. Since these proceedings are,
for the parties to the main proceedings, a step in the proceedings pending before the national court, the
decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the question referred to it by the Sozialgericht
Hamburg by order of 12 September 1994, hereby rules: in proceedings for determining the entitlements to
social security benefits of a migrant worker who is a Community national, the competent social security
institutions and the courts of a member state must accept certificates and analogous documents relative
to personal status issued by the competent authorities of the other member states, unless their accuracy
is seriously undermined by concrete evidence relating to the individual case in question.

465

[1998] All ER (EC) 466

Julius Fillibeck Shne Gmbh & Co KG v Finanzamt Neustadt

 (Case C-258/95)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES GULMANN (PRESIDENT OF CHAMBER), WATHELET, MOITINHO DE ALMEIDA, JANN AND
SEVN
ADVOCATE GENERAL LGER
5 DECEMBER 1996, 16 JANUARY, 16 OCTOBER 1997
European Community Value added tax Supply of goods or services Employer providing free
transport to workplace for employees Employer also commissioning one employee to provide additional
transport using own private vehicle Tax authorities charging employer value added tax on transport
Whether supply of services for consideration Whether transport services effected for purposes other
than those of business Council Directive (EEC) 77/388, arts 2(1), 6(2).

A building company, J & Co, provided free transport to work for employees who lived more than 6km from
their workplace. Employees were mainly transported in vehicles belonging to J & Co, although some were
transported by a fellow employee in his own private vehicle on the companys behalf. The employees
made no payment and no sum was deducted from their wages to pay for or contribute to the cost of the
transport services. The Finanzamt Neustadt considered that the provision of transport was taxable under
the German value added tax legislation. In proceedings in which J & Co disputed an assessment to value
added tax raised on that basis, the Bundesfinanzhof referred to the Court of Justice of the European
Communities for a preliminary ruling three questions to determine: (i) whether transport such as that
provided by J & Co constituted a service effected for consideration within the meaning of art 2(1) 1 of
Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover
taxescommon system of value added tax: uniform basis of assessment; (ii) whether the transport was
to be treated as a supply of services by virtue of art 6(2)2 in that it involved a use of goods forming part of
the assets of J & Cos business or a service carried out free of charge, even though it did not serve
purposes other than business purposes, as far as J & Co was concerned, but did serve the employees
private purposes; and (iii) if such transport was to be regarded as a supply of services under art 6(2),
whether the fact that an employee was commissioned to provide transport in his own private vehicle was
relevant to the application of art 6(2).
1
Article 2, so far as material, is set out at p 476 b, post
2
Article 6, so far as material, is set out at p 476 c, post

Held (1) A supply of services within the meaning of art 2(1) presupposed the existence of a direct link
betweeen the service provided and the consideration received. In the instant case, J & Cos employees
did not pay for the transportation provided and it was not possible to regard a proportion of the work
performed as consideration for that service, since the work to be performed and the wages received were
independent of whether or not the employees used 466 the transport provided to them by their employer.
In those circumstances it followed that there was no service effected for consideration for the purposes
of art 2(1) (see p 477 b e to j and p 480 c, post); Apple and Pear Development Council v Customs and
Excise Comrs Case 102/86 [1988] 2 All ER 922, applied.
(2) Under art 6(2) transport provided by an employer free of charge for employees between their
homes and the workplace in a company vehicle served, in principle, the employees private purposes and
thus served purposes other than those of the business. However, that provision did not apply when,
having regard to the difficulty of finding other suitable means of transport and changes in the place of
work, the requirements of the business made it necessary for the employer to provide transport for
employees, in which case the supply of those transport services was not effected for purposes other than
those of the business. That interpretation also applied when the employer did not convey the employees
in its own vehicles, but commissioned one of its employees to provide the transport using his own private
vehicle (see p 479 f g j to p 480 a d to f, post).

Notes
For supply of services within the context of the Sixth Directive, see 52 Halsburys Laws (4th edn) para
2017.

Cases cited
Apple and Pear Development Council v Customs and Excise Comrs Case 102/86 [1988] 2 All ER 922,
[1988] STC 221, [1988] ECR 1443, ECJ.
Argos Distributors Ltd v Customs and Excise Comrs Case C-288/94 [1996] STC 1359, [1996] I-5311.
Boots Co plc v Customs and Excise Comrs Case C-126/88 [1990] STC 387, [1990] ECR I-1235, ECJ.
Empire Stores Ltd v Customs and Excise Comrs Case C-33/93 [1994] 3 All ER 90, [1994] STC 623,
[1994] ECR I-2329, ECJ.
Enkler v Finanzamt Homburg Case C-230/94 [1996] STC 1316, [1996] ECR I-4517, ECJ.
Finanzamt Mnchen III v Mohsche Case C-193/91 [1997] STC 195, [1993] ECR I-2615, ECJ.
Glawe (H J) Spiel und Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt
Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994] STC 543, [1994] ECR I-1679, ECJ.
Khne v Finanzamt Mnchen III Case 50/88 [1990] STC 749, [1989] ECR 1925.
Naturally Yours Cosmetics Ltd v Customs and Excise Comrs Case 230/87 [1988] STC 879, [1988] ECR
6365, ECJ.
Staatssecretaris van Financin v Coperatieve Aardappelenbewaarplaats GA Case 154/80 [1981] ECR
445.
Staatssecretaris van Financin v Hong Kong Trade Development Council Case 89/81 [1982] ECR 1277.
Tolsma v Inspecteur der Omzetbelasting Case C-16/93 [1994] STC 509, [1994] ECR I-743, ECJ.

Reference
By order of 11 May 1995 the Bundesfinanzhof (the Federal Finance Court) referred to the Court of Justice
of the European Communities for a preliminary ruling three questions (set out at p 476 e to h, post)
concerning the interpretation of arts 2(1) and 6(2) of Council Directive (EEC) 77/388 on the harmonisation
of 467 the laws of the member states relating to turnover taxescommon system of value added tax:
uniform basis of assessment. Those questions were raised in proceedings between Julius Fillibeck Shne
GmbH & Co KG (Fillibeck) and the Finanzamt Neustadt (the Neustadt Tax Office), concerning the
imposition of value added tax on the free transport provided by Fillibeck for certain employees from their
homes to their place of work. Written observations were submitted on behalf of: Fillibeck, by K Heiniger,
Accountant and Tax Adviser, acting as agent; the Finanzamt Neustadt, by R Preuninger,
Oberregierungsrat, acting as agent; the German government, by E Rder, Ministerialrat at the Federal
Ministry of Economic Affairs, and B Kloke, Oberregierungsrat at the same ministry, acting as agents; the
UK government, by S Braviner, of the Treasury Solicitors Department, acting as agent, and N Paines,
Barrister; and the European Commission, by J Grunwald, Legal Adviser, acting as agent. Oral
observations were submitted on behalf of Fillibeck, represented by K Heininger; the Finanzamt Neustadt,
represented by W Widmann, Leitender Ministerialrat at the Ministry of Finance of Rheinland-Pfalz, acting
as agent; and the Commission, represented by J Grunwald. The language of the case was German. The
facts are set out in the opinion of the Advocate General.

16 January 1997.

The Advocate General (P Lger)

delivered the following opinion (translated from the French).
1.
The Bundesfinanzhof (the Federal Finance Court) has referred three questions to the Court of Justice
of the European Communities concerning the interpretation of certain provisions of Council Directive
(EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover taxescommon
system of value added tax: uniform basis of assessment (the Sixth Directive). The court is essentially
asked to consider the manner in which free collective transport provided by an employer to its employees,
from their home to the workplace, should be treated for value added tax (VAT) purposes.

Facts and procedure

2.
According to the order for reference, the facts in this case are as follows.
3.
Julius Fillibeck Shne GmbH & Co KG (the applicant) runs a building firm. In the years in question,
1980 to 1985, it provided transport for some of its employees free of charge from their homes to the
various building sites where they were required to work. Employees were mainly transported in vehicles
belonging to the employer, although some were transported by another employee in his own private
vehicle on the employers behalf. In neither case was any specific charge made to employees using the
transport, nor was any deduction made from their wages.
4.
Those transport services were provided pursuant to the Bundesrahmentarifvertrag fr das
Baugewerbe (the Federal Collective Framework Agreement for the Building Industry) where the
employees homes and the building sites were more than a minimum distance apart (six kilometres).
5.
In that context, the Finanzamt (the Tax Office) (the defendant) considered there to be taxable
transactions within the meaning of the German legislation on turnover tax (see para 1(1)(1), second
sentence, (b), of the Umsatzsteuergesetz of 1980) (the Law on turnover tax) (the UStG).
6.
In its appeal, the applicant challenged the view that those transport services were taxable. It claimed,
in that respect, that the judgment of the Bundesfinanzhof of 11 March 1988 3 did not apply to
circumstances such as those 468 in the present case and on that ground the transport at issue could
therefore not be regarded as a taxable supply for the purposes of the provisions of the UStG cited above 4.
3
See VR 30/84 and VR 114/83 (BFHE 153, 155,162, BStBl II 1988, 643, 651.
 4
The national court points out that, in that judgment, the Bundesfinanzhof abandoned the former case law. Whereas a
notional part of the work performed by the employees had previously been treated as consideration for the transport
supplied by the employer, it now considers that it is impossible to ascertain the value of that notional part of the work
constituting consideration.

Both the defendant and the Finanzgericht (the finance court), before which its action was
unsuccessful, it claimed, failed to recognise that the transport services at issue were not provided to
employees for their own private use, but were provided by the employer predominantly in the interests of
its business and pursuant to its obligation, in its capacity as a building firm, under the collective
agreement.
7.
The national court considers that this case raises questions relating to the taxation of so-called
benefits in kind provided by an employer to its employees where the employees do not have to provide
any specifically agreed, fixed return for the benefit in kind. It therefore referred the following questions to
the Court of Justice for a preliminary ruling:
(1) Does the transport provided by an employer constitute a service effected for consideration
within the meaning of Article 2(1) of the Sixth Directivethat is to say, effected for a proportion (to
be estimated) of the work performed by the employeeswhere, pursuant to a collective
agreement, the employer conveys employees (without specially agreed and calculated
consideration) from their homes to the workplace where they are more than a specified distance
apart, and the work performedwhich has no actual connection with such transport servicesis
already to be carried out in return for the agreed money wages as in the case of the other
employees?
(2) Does Article 6(2) of the Sixth Directive cover the use of goods forming part of the assets of
the business or a service carried out free of charge even whereas in the case of free transport for
employees from their homes to the workplace and back in a company vehicleit does not serve
purposes other than those of the business as far as the employer is concerned, but does serve the
employees private purposes and the employees are not charged turnover tax in this respect (on
account of their use free of charge of the transport service)?
(3) In the event that question 2 is answered in the affirmative: does Article 6(2) of the Sixth
Directive also cover a case where the employer does not convey the employees in its own vehicles,
but commissions a third party(in this case, one of its own employees) to effect the transport?

First question
8.
The Bundesfinanzhof is in doubt, first, as to whether, and if so the extent to which, the transport
services provided by the employer to its employees in this case constitute supplies effected for
consideration (in the form of a proportionto be estimated as appropriateof the work performed) within
the meaning of art 2(1) of the Sixth Directive.
9.
Article 2(1) provides:
469
The following shall be subject to value added tax: 1. the supply of goods or services effected for
consideration within the territory of the country by a taxable person acting as such
10.
It is not disputed that in this case the transport services constitute a supply of services by a taxable
person for the purposes of the Sixth Directive; the national courts only question is whether the supply is
effected for consideration.
11.
The question of whether a supply of services is effected for consideration has been addressed in
numerous cases and has now been settled.
12.
According to the court, in order for a supply of services to be regarded as effected for consideration,
and thus to be taxable, there must be a direct link between the service provided and the consideration
received.
13.
That principle was first laid down in Staatssecretaris van Financin v Coperatieve
Aardappelenbewaarplaats GA Case 154/80 [1981] ECR 445 (paras 1213), which concerned Council
Directive (EEC) 67/228 on the harmonisation of legislation of member states concerning turnover taxes
structure and procedures for application of the common system of value added tax (OJ S edn 1967 (I) p
16) (the Second VAT Directive).
14.
In Apple and Pear Development Council v Customs and Excise Comrs Case 102/86 [1988] 2 All ER
922, [1988] ECR 1443 the court confirmed that decision, this time in the context of the Sixth Directive, and
held that the body in question which, in pursuit of its task of defending the common interests of growers of
apples and pears, carried out collective advertising campaigns financed by a mandatory annual charge,
was not supplying services for consideration, in the absence of any direct link between the service
provided and the consideration received.
 The court held that if individual growers did receive benefits, then those benefits were only derived
indirectly from those accruing generally to the industry as a whole and, furthermore, that there was no link
between the level of benefit derived by individual growers from the services of the Apple and Pear
Development Council and the level of the mandatory charge which they had to pay.
15.
It is also useful to refer to Tolsma v Inspecteur der Omzetbelasting Case C-16/93 [1994] STC 509,
[1994] ECR I-743 (para 14), in which the court held that a supply of services is effected for consideration
within the meaning of art 2(1) of the Sixth Directive, only
if there is a legal relationship between the provider of the service and the recipient pursuant to
which there is reciprocal performance, the remuneration received by the provider of the service
constituting the value actually given in return for the service supplied to the recipient.
On that occasion, the court recalled ([1994] STC 509, [1994] ECR I-743 (para 12)) that
[it] has already held with reference to the concept of the provision of services against payment
in art 2(a) of the Second Directive, whose wording is similar to that of art 2(1) of the Sixth Directive,
that taxable transactions, within the framework of the VAT system, presuppose the existence of a
transaction between the parties in which a price or consideration is stipulated. The court concluded
that, where a persons activity consists exclusively in providing services for no direct consideration,
there is no basis of assessment and the services are therefore not subject to VAT (see the 470
judgment in Staatssecretaris van Financin v Hong Kong Trade Development Council (Case 89/81)
[1982] ECR 1277 at 1286, paras 9, 10).
And ([1994] STC 509, [1994] ECR I-743 (para 13)):
In its judgments in Staatssecretaris van Financin v Coperatieve Aardappelenbewaarplaats
GA (Case 154/80) [1981] ECR 445 at 454, para 12, and Naturally Yours Cosmetics Ltd v Customs
and Excise Comrs (Case 230/87) [1988] STC 879 at 886, [1988] ECR 6365 at 6389, para 11, the
court stated on this point that the basis of assessment for a provision of services is everything
which makes up the consideration for the service and that a provision of services is therefore
taxable only if there is a direct link between the service provided and the consideration received
(see also the judgment in Apple and Pear Development Council v Customs and Excise Comrs
(Case 102/86) [1988] STC 221 at 237, [1988] ECR 1443 at 1468, paras 11, 12).
16.
The three following criteria can be identified from those decisions, which makes it possible to define
the concept of the supply of services effected for consideration 5: there must be a direct link between the
service supplied and the consideration received6; the consideration must be capable of being expressed
in money7; the consideration must be the subjective value given to it by the parties 8.
5
See, to that effect, the opinion of Advocate General Lenz in Tolsma [1994] STC 509, [1994] ECR I-743 (para 14).
6
See the judgments in Coperatieve Aardappelenbewaarplaats [1981] ECR 445 (para 12), Apple and Pear Development
Council [1988] 2 All ER 922, [1988] ECR 1443 (para 11) and Naturally Yours Cosmetics [1988] STC 879, [1988]
ECR 6365 (para 11).
7
See the judgments in Coperatieve Aardappelenbewaarplaats [1981] ECR 445 (para 13) and Naturally Yours
Cosmetics [1988] STC 879, [1988] ECR 6365 (para 16).
8
See the judgments in Coperatieve Aardappelenbewaarplaats [1981] ECR 445 (paras 1314) and Naturally Yours
Cosmetics [1988] STC 879, [1988] ECR 6365 (para 16).

17.
The three conditions which the case law of this court requires to be satisfied in order for a supply of
services to be regarded as being effected for consideration and, as such, subject to VAT, are not satisfied
here.
18.
A mere proportion (to be estimated) of the work performed by the employee, to use the words of the
national court, cannot be considered real consideration for the transport services received by the
employee.
19.
In the present case, no direct link can be observed between the service supplied and the consideration
received.
20.
The service which, in this case, consists of the transport services is supplied independently of any
consideration received. Employees using the free collective transport made available to them by the
employer do not make any payment, nor is any sum deducted from their wages as consideration for that
service. Furthermore, the wages cannot be regarded as comprising consideration for that service, since
they are only paid as consideration for the work performed.
Just as in Apple and Pear Development Council in which all apple and pear growers paid the same
charge to the body in question and there was no link between the charge paid and the benefits derived,
the employees in this case are required to perform the same work and receive the same salary,
irrespective of whether they derive a benefit from the free transport; there is no consideration which has a
direct link with the transport provided by the employer.
The employer is required to make free transport available for all its employees under the relevant
collective agreement, but they are free to choose whether or not to make use of it.
471
21.
A distinct proportion of the work performed by the employee cannot therefore be regarded as the
(estimated) value of the consideration for the transport services.
22.
Furthermore, in the course of the proceedings, neither party proposed any estimation of the distinct
proportion of the work performed which represents the consideration for the transport services.
23.
As a result, in so far as no consideration having a direct link with the service supplied can be identified,
it is, a fortiori, impossible in this case to express the consideration in money, as required by the case law
cited above, and to recognise a subjective value given to it by the parties.
24.
Since transport services provided free of charge by an employer to its employees cannot be regarded
as effected for consideration within the meaning of art 6(2) of the Sixth Directive, those services do not
fall within the scope of that provision and on that basis cannot therefore be subject to VAT.
25.
The first question should therefore be answered in the negative.

Second question
26.
In the event that the first question is answered in the negative, the national court asks whether the
transport services at issue fall within the definition of taxable transactions for the purposes of art 6(2) of
the Sixth Directive.
27.
That provision reads as follows:
The following shall be treated as supplies of services for consideration: (a) the use of goods
forming part of the assets of a business for the private use of the taxable person or of his staff or
more generally for purposes other than those of his business where the value added tax on such
goods is wholly or partly deductible, (b) supplies of services carried out free of charge by the
taxable person for his own private use or that of his staff or more generally for purposes other than
those of his business.
Member States may derogate from the provisions of this paragraph provided that such
derogation does not lead to distortion of competition.
28.
According to art 11A(1)(c) of the Sixth Directive, the taxable amount of the transactions referred to in
art 6(2) is the full cost to the taxable person of providing the services.
29.
The wording of the national courts question seems to suggest that a distinction should be made
between the use of goods forming part of the assets of [the] business (art 6(2)(a)) and supplies of
services carried out free of charge (art 6(2)(b)).
However, that distinction does not seem to me to be necessary in the circumstances of this case. Only
the expressions private use of the taxable person or of his staff and purposes other than those of his
business, which are used in both provisions, are relevant.
30.
To summarise, either the transport at issue is regarded as being effected for the private use of the
staff, for purposes other than those of the business, or, alternatively, it is regarded as serving the
purposes of the business, in which case it is not subject to any tax under art 6(2) of the Sixth Directive.
31.
Arguments were submitted in favour of both approaches during the proceedings.
32.
First, the applicant claimed that the transport provided for employees directly and exclusively serves
the purposes of the business and therefore falls within the employment relationship and does not concern
the private domain 472 of employees. It submits, in that respect, that even if the transport did
theoretically concern the employees private sphere, it would have been brought within the employment
relationship by virtue of the collective agreement pursuant to which it is provided. It concluded therefore
that art 6(2) does not apply in this case.
33.
In contrast, the other parties considered that the free transport of employees serves their private
purposes and thus serves purposes other than those of the business; art 6(2) therefore applies.
34.
The UK government and the European Commission, however, consider that the special circumstances
of this case should be taken into account and justify non-taxation under art 6(2).
35.
I support that view.
36.
 First, it seems difficult to deny that the transport provided for employees from their home to the
workplace is a matter which in principle concerns their private life.
37.
The employee decides where his home will be with regard, where appropriate, to his place of work,
and that determines the distance between the two and the means of transport he is free to choose for that
journey. The employer is not involved in any way in those decisions, since the employees only obligation
towards him is to report to his place of work at the agreed time. The employee is responsible for taking
the necessary steps to arrive at the workplace and he alone is responsible for deciding how to do so. As a
result, under normal circumstances, transport provided for an employee from his home to the workplace is
for the private use of staff, and is effected for purposes other than those of the business. That is not
altered by the fact that the transport may be organised by the employer, who makes free collective
transport available to those employees wishing to use it.
38.
In general, therefore, the transport provided for employees by an employer free of charge using a
company vehicle must be considered to be subject to VAT, since it is treated as a supply of services
effected for consideration in accordance with art 6(2) of the Sixth Directive.
39.
That principle is consistent with the aim of that provision, which is designed to prevent the non-
taxation of business goods used for private purposes (see the judgments in Finanzamt Mnchen III v
Mohsche Case C-193/91 [1997] STC 195, [1993] ECR I-2615 (para 8) and Khne v Finanzamt Mnchen
III Case 50/88 [1990] STC 749, [1989] ECR 1925 (para 8)).
40.
However, I believe that, in exceptional cases and under special circumstances, which in this case
relate to the nature of the business, transport provided by an employer could be regarded as being
supplied for purposes which are not other than those of the business.
41.
The particular characteristic of a building firm is that its employees are required to work on building
sites in various locations according to the needs of the companys customers, rather than working on a
permanent basis at a place of work agreed in advance with the employer. The place in which they are
required to work can change from one day to the next, or even in the course of a single day. Employees
must be prepared to move as soon as the employer asks them to do so. That is a specific characteristic
which distinguishes employees in the building sector, in this case, from employees who are required to
work in a place determined in advance, for example on the premises of the company employing them.
Employees are consequently unable to exercise any degree of choice or control over the length of the
journey between their home and the workplace.
473
42.
The places where the employees concerned are required to work are generally difficult to reach by
traditional means, since they are building sites which may be situated, for example, on industrial estates
or in outlying areas.
43.
Furthermore, because those building sites may be some distance from the employees homes,
depending on where they are required to work, working hours in the building sector can mean that
employees have to get up particularly early in the morning in order to reach their place of work in time.
44.
In those circumstances, the risk involved for the employer in leaving employees to choose for
themselves how to travel from their homes to the workplace could be detrimental to the successful
operation of the company. It is possible that other more traditional means of transport, such as public
transport, might prove to be inadequate or even non-existent in those circumstances.
45.
Therefore, transport organised by an employer for its employees from their homes to their places of
work may, in certain special circumstances, such as those in the present case, be necessary to meet
fundamental requirements for the proper operation of the business. Although it cannot be denied that the
employees derive a benefit from the transport made available to them, that benefit seems to be of
secondary importance compared to the purposes of the business in respect of which the free collective
transport was organised.
46.
Although the fact that a collective agreement expressly provides for such transport is not conclusive, it
is none the less indicative of the fact that the building sector involves an activity with special
characteristics, which must be taken into account.
47.
I therefore consider that the answer to the second question should be that, although in general the
transport provided for employees from their homes to the workplace using a vehicle belonging to the
employer must be regarded as serving the private purposes of those employees and, on that basis,
subject to tax under art 6(2) of the Sixth Directive, that is not the case when, as a result of special
circumstances inherent, for example, in the nature of the business, those transport services serve
purposes which are not other than those of the business. Such a situation falls outside the scope of art
6(2) and the transaction is not taxable.
 Third question
48.
By this question, the Bundesfinanzhof asks the Court of Justice, in the event that the second question
is answered in the affirmative, whether the fact that an employee may be commissioned to provide the
transport at issue, using his own private vehicle, is relevant.
49.
In view of the suggested answer to the second question, there is no need to answer the third question
in this case.
50.
However, if the circumstances were not such as to justify regarding the transport as being provided for
purposes which are not other than those of the business, the answer to the third question would not differ
from that given to the previous question. An employee commissioned to carry out a task on behalf of the
taxable person, within the meaning of the Sixth Directive, is not acting independently within the meaning
of art 4(1) thereof: transport services supplied by an employee must be regarded as comparable to
transport services supplied by the taxable trader, since they are carried out on behalf of that trader.
474

Conclusion
51.
For those reasons, I propose that the questions referred by the Bundesfinanzhof should be answered
as follows:
(1) Article 2(1) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the
member states relating to turnover taxescommon system of value added tax: uniform basis of
assessment (the Sixth Directive) is to be interpreted as meaning that an employer who provides
transport for employees free of charge from their homes to the workplace, in the absence of any
real connection either with the work performed or the wages received, does not effect a supply of
services for consideration within the meaning of that provision.
(2) Article 6(2) of the Sixth Directive is to be interpreted as meaning that it applies where
employees are conveyed free of charge by the employer from their homes to the workplace and
back, in so far as the transport provided serves, in principle, the employees private purposes and
thus serves purposes other than those of the business. That provision does not, however, apply
where, as in this case, because of the specific nature of the business, the supply of those transport
services is not effected for purposes other than those of the business and therefore serves the
purposes of the business.
(3) The first sentence of the answer to question 2 also applies when the employer does not
convey the employees in its own vehicles, but commissions a third party (in this case, one of its
own employees) to do so.

16 October 1997.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By order of 11 May 1995, which was received at the Court of Justice of the European Communities on
31 July 1995, the Bundesfinanzhof (the Federal Finance Court) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty three questions concerning the interpretation of arts 2(1)
and 6(2) of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states
relating to turnover taxescommon system of value added tax: uniform basis of assessment (the Sixth
Directive).
2.
The questions were raised in proceedings between Julius Fillibeck Shne GmbH & Co KG (Fillibeck)
and the Finanzamt (the Tax Office) Neustadt, concerning the imposition of value added tax (VAT) on the
free transport provided by Fillibeck for its employees from their homes to their place of work.
3.
From 1980 to 1985, Fillibeck, which runs a building undertaking, conveyed some of its employees in
company vehicles free of charge from their homes to the various building sites where they were required
to work. During the same period it also required one of its employees to convey other employees from
their homes to their various places of work in his own private vehicle.
4.
Fillibeck provided that transport pursuant to the Bundesrahmentarifvertrag fr das Baugewerbe (the
Federal Collective Framework Agreement for the Building Industry) where the employees homes and
their places of work were more than a minimum distance apart.
5.
The Finanzamt Neustadt considered that that transport was subject to tax under the German VAT
legislation.
6.
 Fillibeck challenged the view that the transport was subject to VAT. Since its objection and the action it
brought were unsuccessful, it appealed to the 475 Bundesfinanzhof on a point of law; that court
considered that the dispute raised questions concerning the interpretation of arts 2(1) and 6(2) of the
Sixth Directive.
7.
Article 2(1) of the Sixth Directive provides:
The following shall be subject to value added tax: 1. The supply of goods or services effected
for consideration within the territory of the country by a taxable person acting as such
8.
Article 6(2) of the Sixth Directive provides:
The following transactions shall be treated as supplies of services for consideration: (a) the use
of goods forming part of the assets of a business for the private use of the taxable person or of his
staff or more generally for purposes other than those of his business where the value added tax on
such goods is wholly or partly deductible; (b) supplies of services carried out free of charge by the
taxable person for his own private use or that of his staff or more generally for purposes other than
those of his business.
9.
In those circumstances, the Bundesfinanzhof decided to stay the proceedings and to refer the
following questions to the Court of Justice for a preliminary ruling:
(1) Does transport provided by an employer constitute a service effected for consideration
within the meaning of Article 2(1) of the Sixth Directivethat is to say, effected for a proportion (to
be estimated) of the work performed by the employeeswhere, pursuant to a collective
agreement, the employer conveys employees (without specially agreed and calculated
consideration) from their homes to the workplace where they are more than a specified distance
apart, and the work performedwhich has no actual connection with such transport servicesis
already to be carried out in return for the agreed money wages as in the case of the other
employees?
(2) Does Article 6(2) of the Sixth Directive cover the use of goods forming part of the assets of
the business or a service carried out free of charge even whereas in the case of free transport for
employees from their homes to the workplace and back in a company vehicleit does not serve
purposes other than those of the business as far as the employer is concerned, but does serve the
employees private purposes and the employees are not charged turnover tax in this respect (on
account of their use free of charge of the transport service)?
(3) In the event that question 2 is answered in the affirmative: Does Article 6(2) of the Sixth
Directive also cover a case where the employer does not convey the employees in its own vehicles,
but commissions a third party (in this case, one of its own employees) to effect the transport?

The first question

10.
In the first question, the national court asks whether art 2(1) of the directive is to be interpreted as
meaning that transport provided by an employer for employees free of charge from their homes to the
workplace where they are more than a specified distance apart, in the absence of any real connection
with either the work performed or the wages received, constitutes a supply of services effected for
consideration within the meaning of that provision.
476
11.
Article 2(1) of the Sixth Directive provides that VAT is chargeable on the supply of services effected for
consideration within the territory of the country by a taxable person.
12.
It is apparent from the case law of the court that the concept of the supply of services effected for
consideration within the meaning of art 2(1) of the Sixth Directive presupposes the existence of a direct
link between the service provided and the consideration received (see, in particular, the judgment in
Apple and Pear Development Council v Customs and Excise Comrs Case 102/86 [1988] 2 All ER 922,
[1988] ECR 1443 (para 12)).
13.
It is also settled case law that the taxable amount for the supply of goods or services is represented by
the consideration actually received for them. That consideration is thus the subjective value, that is to say,
the value actually received, and not a value estimated according to objective criteria (see the judgments
in Staatssecretaris van Financin v Coperatieve Aardappelenbewaarplaats GA Case 154/80 [1981] ECR
445 (para 13), Naturally Yours Cosmetics Ltd v Customs and Excise Comrs Case 230/87 [1988] STC
879, [1988] ECR 6365 (para 16), Boots Co plc v Customs and Excise Comrs Case C-126/88 [1990] STC
387, [1990] ECR I-1235 (para 19), H J Glawe Spiel und Unterhaltungsgerte Aufstellungsgesellschaft
mbH & Co KG v Finanzamt Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994] STC 543, [1994] ECR I-
1679 (para 8), Empire Stores Ltd v Customs and Excise Comrs Case C-33/93 [1994] 3 All ER 90, [1994]
ECR I-2329 (para 18) and Argos Distributors Ltd v Customs and Excise Comrs Case C-288/94 [1996]
STC 1359, [1996] I-5311 (para 16)).
14.
 Furthermore, according to the same case law, that consideration must be capable of being expressed
in money (see the judgments in Coperatieve Aardappelenbewaarplaats [1981] ECR 445 (para 13),
Naturally Yours Cosmetics Ltd [1988] STC 879, [1988] ECR 6365 (para 16) and Argos Distributors Ltd
[1996] STC 1359, [1996] I-5311 (para 17)).
15.
It is clear from the order for reference that Fillibeck provides transport for its employees from their
homes to their workplace when they are more than a certain distance apart and that the employees do
not make any payment, nor is any sum deducted from their wages in respect of that service.
16.
Furthermore, since the work to be performed and the wages received are independent of the use or
otherwise by employees of the transport provided to them by their employer, it is not possible to regard a
proportion of the work performed as being consideration for the transport services.
17.
In those circumstances, there is no consideration which has a subjective value and a direct link with
the service provided. Consequently, the requirements relating to a supply of services effected for
consideration are not satisfied.
18.
The reply to the first question is therefore that art 2(1) of the Sixth Directive is to be interpreted as
meaning that an employer who provides transport for employees free of charge from their homes to the
workplace where they are more than a specified distance apart, in the absence of any real connection
either with the work performed or the wages received, does not effect a supply of services for
consideration within the meaning of that provision.

The second question

19.
In the second question, the national court is asking essentially whether transport provided for
employees free of charge between their homes and their workplace by the employer in a company vehicle
for purposes which are not other than those of the business but which, at the same time, serve the 477
employees private purposes, is to be treated as a supply of services for consideration within the meaning
of art 6(2) of the Sixth Directive.
20.
First of all, it should be noted that the national court refers to art 6(2)(a) and (b) without, however,
establishing which of those two provisions applies to the case before it. The parties to the main
proceedings consider that, in the present case, it is not necessary to determine whether sub-para (a) or
(b) applies. Furthermore, it is clear from the question that the requested interpretation of those provisions
concerns, more specifically, the concepts of private use of the taxable person or of his staff and
purposes other than those of his business, which appear in both provisions. The two provisions should,
consequently, be considered together.
21.
The national court pointed out that the transport provided by Fillibeck is for travel between the
employees homes and the companys various building sites where those employees are required to work.
22.
At the hearing, Fillibeck explained, in particular, that it has a number of building sites which are some
distance apart, that it is often not possible to reach those building sites by public transport and that
employees are moved between the different sites.
23.
It also pointed out that, since the carriage of the employees directly serves the purposes of the
business and consequently falls within the employment relationship, it does not concern the private
domain of employees. Furthermore, the transport is provided pursuant to a collective agreement.
24.
In contrast, the other parties who submitted observations claim that the transport provided by the
employer free of charge falls within art 6(2) of the Sixth Directive. The UK government and the
Commission pointed out, however, that certain circumstances may justify regarding the transport provided
for the employees as serving the purposes of the business.
25.
It should be recalled that the purpose of art 6(2) of the Sixth Directive is to ensure equal treatment as
between taxable persons and final consumers (see the judgment in Enkler v Finanzamt Homburg Case
C-230/94 [1996] STC 1316, [1996] ECR I-4517 (para 35)). It is designed to prevent the non-taxation of
business goods used for private purposes and of services provided free of charge by a taxable person for
private purposes (see, to that effect, the judgments in Khne v Finanzamt Mnchen III Case 50/88 [1990]
STC 749, [1989] ECR 1925 (para 8) and Finanzamt Mnchen III v Mohsche Case C-193/91 [1997] STC
195, [1993] ECR I-2615 (para 8)).
26.
In that respect, it should be noted that it is normally for the employee to decide where his home will be
with regard, where appropriate, to his place of work, and to determine the distance between the two and
the means of transport he intends to use. The employer is not involved in those decisions, since the
employees only obligation is to be present at his place of work at the agreed times. Consequently, under
normal circumstances, the transport services provided to employees are for the private use of the
employee within the meaning of art 6(2) of the Sixth Directive.
27.
The fact that an employee must travel between his home and the workplace in order to be present at
work and, consequently, to perform his duties, is not conclusive evidence that transport provided for an
employee from his home to his workplace is not to be considered as being for the employees private use
within the meaning of art 6(2). Indeed, it would be contrary to the purpose of that provision if such an
indirect link were sufficient, in itself, to prevent such travel being treated as a supply for consideration.
478
28.
Article 6(2) of the Sixth Directive must be interpreted thus in the usual situation where an employee
travels between his home and his fixed place of work, and has the possibility of using ordinary means of
transport.
29.
It should be acknowledged, however, that in certain circumstances the requirements of the business
may make it necessary for the employer to provide transport for employees between their homes and the
workplace. The fact that only the employer is able to provide suitable transport or that the workplace is not
always the same but is liable to change may mean that the employer is obliged to provide transport for its
employees.
30.
In such special circumstances, the transport is organised by the employer for purposes which are not
other than those of the business. The personal benefit derived by employees from such transport appears
to be of only secondary importance compared to the needs of the business.
31.
As regards the fact that the transport is provided pursuant to a collective agreement, even though such
an obligation is not in itself sufficient to determine the character of the supply of those services for the
purposes of art 6(2) of the Sixth Directive, it indicates that the transport is provided for purposes which are
not other than those of the business.
32.
The special characteristics of building firms, as described in particular by Fillibeck in the present case,
suggest that the transport may be organised for purposes which are not other than those of the business.
33.
It is for the national court to establish whether, in the light of the interpretations given by the court, the
particular characteristics of the case before it make it necessary, having regard to the requirements of the
business, for the employer to provide transport for employees between their homes and the workplace.
34.
The answer to the second question is therefore that art 6(2) of the Sixth Directive is to be interpreted
as meaning that transport provided for employees free of charge by the employer between their homes
and the workplace in a company vehicle serves, in principle, the employees private purposes and thus
serves purposes other than those of the business.
However, that provision does not apply when, having regard to certain circumstances, such as the
difficulty of finding other suitable means of transport and changes in the place of work, the requirements
of the business make it necessary for the employer to provide transport for employees, in which case the
supply of those transport services is not effected for purposes other than those of the business.

The third question

35.
In the third question, the national court asks essentially whether the answer to the second question
also applies when the employer does not convey the employees in its own vehicles, but commissions one
of its employees to provide the transport using his own private vehicle.
36.
In that respect, it is sufficient to note that the question whether transport services serve the private
purposes of employees of an undertaking or, more generally, serve purposes other than those of the
business, within the meaning of art 6(2) of the Sixth Directive, is not altered by the fact that, instead of
providing the transport in its own vehicles, the employer commissions one of its employees to provide that
transport using his own private vehicle.
37.
Consequently, the answer to the third question must be that the answer to the second question also
applies when the employer does not convey the 479 employees in its own vehicles, but commissions one
of its employees to provide the transport using his own private vehicle.

Costs
38.
The costs incurred by the German government, the UK government and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the proceedings pending before the
national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the questions referred to it by the
Bundesfinanzhof by order of 11 May 1995, hereby rules: (1) Article 2(1) of Council Directive (EEC) 77/388
on the harmonisation of the laws of the member states relating to turnover taxescommon system of
value added tax: uniform basis of assessment (the Sixth Directive) is to be interpreted as meaning that an
employer who provides transport for employees free of charge from their homes to the workplace where
they are more than a specified distance apart, in the absence of any real connection either with the work
performed or the wages received, does not effect a supply of services for consideration within the
meaning of that provision. (2) Article 6(2) of the Sixth Directive is to be interpreted as meaning that
transport provided for employees free of charge by the employer between their homes and the workplace
in a company vehicle serves, in principle, the employees private purposes and thus serves purposes
other than those of the business. However, that provision does not apply when, having regard to certain
circumstances, such as the difficulty of finding other suitable means of transport and changes in the place
of work, the requirements of the business make it necessary for the employer to provide transport for
employees, in which case the supply of those transport services is not effected for purposes other than
those of the business. (3) The answer to the second question also applies when the employer does not
convey the employees in its own vehicles, but commissions one of its employees to provide the transport
using his own private vehicle.

480

[1998] All ER (EC) 481

Note
John Deere Ltd v European Commission
(Case C-7/95 P)

EUROPEAN COMMUNITY; Other European Community

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES GULMANN (PRESIDENT OF THE CHAMBER), MOITINHO DE ALMEIDA, EDWARD, JANN
AND SEVN (RAPPORTEUR) ADVOCATE GENERAL RUIZ-JARABO COLOMER
3 JULY, 16 SEPTEMBER 1997, 28 MAY 1998
European Community Practice Scope of the courts review when hearing an appeal EC Statute of
the Court of Justice, art 51 Rules of procedure, art 112(1)(c) EC Treaty, art 168a.

Cases cited
Hilti v European Commission Case C-53/92 P [1994] ECR I-667.
San Marco v European Commission Case C-19/95 P [1996] ECR I-4435.

Notes
For the EC Statute of the Court of Justice, art 51, see 50 Halsburys Statutes (4th edn), Current Service, p
5.
For the EC Treaty, art 168a (as amended by art G.50 of the Treaty on European Union), see ibid p 96.

Appeal
By application lodged at the registry of the Court of Justice of the European Communities on 13 January
1995, John Deere Ltd, a company incorporated in the United Kingdom, brought an appeal pursuant to art
49 of the EC Statute of the Court of Justice against the judgment of the Court of First Instance of the
European Communities ([1994] ECR II-957), by which it dismissed John Deere Ltds application for
annulment of Commission Decision (EEC) 92/157 relating to a proceeding pursuant to art 85 of the EC
Treaty. In its subsequent judgment dismissing the appeal, the Court of Justice outlined the principles
applying to appeals, and, in particular, the extent of the courts jurisdiction when hearing an appeal. John
Deere Ltd were represented by H-J Niemeyer and R Bechtold, Rechtanwlte, Stuttgart, with an address
for service in Luxembourg at the Chambers of Loesch and Wolter, 11 Rue Goethe. The European
Commission were represented by J Currall, of its Legal Service, acting as agent, and N Forwood QC, with
an address for service in Luxembourg at the office of Carlos Gmez de la Cruz, of its Legal Service,
Wagner Centre, Kirchberg. The language of the case was English.

28 May 1998.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment, of which paras 17 to 22 are reported below.

SCOPE OF THE COURT OF JUSTICES REVIEW WHEN HEARING AN APPEAL

17.
Before the grounds of appeal put forward by the appellant are examined, certain principles applying to
appeals, and in particular to the extent of the Court of Justices jurisdiction, should be recalled.
481
18.
Article 168a of the EC Treaty and art 51 of the EC Statute of the Court of Justice state that an appeal
is to be limited to points of law and must be based on the grounds of lack of competence of the Court of
First Instance, breach of procedure before it which adversely affects the interests of the appellant or
infringement of Community law by the Court of First Instance. Article 112(1)(c) of the Court of Justices
Rules of Procedure provides that an appeal must contain the pleas in law and legal arguments relied on
(see OJ 1991 L176 p 7).
19.
It follows from those provisions that an appeal must indicate precisely the contested elements of the
judgment which the appellant seeks to have set aside, and also the legal arguments specifically advanced
in support of the appeal (see the order in San Marco v European Commission Case C-19/95 P [1996]
ECR I-4435 (para 37)).
20.
That requirement is not satisfied by an appeal confined to repeating or reproducing word for word the
pleas in law and arguments previously submitted to the Court of First Instance, including those based on
facts expressly rejected by that court; in so far as such an appeal does not contain any arguments
specifically contesting the judgment appealed against, it amounts in reality to no more than a request for
re-examination of the application submitted to the Court of First Instance, which under art 49 of the EC
Statute the Court of Justice does not have jurisdiction to undertake (see, to this effect, esp the order in
San Marco [1996] ECR I-4435 (para 38)).
21.
It also follows from the foregoing provisions that an appeal may be based only on grounds relating to
the infringement of rules of law, to the exclusion of any appraisal of the facts. The Court of First Instance
has exclusive jurisdiction, first, to establish the facts except where the substantive inaccuracy of its
findings is apparent from the documents submitted to it and, second, to assess those facts.
When the Court of First Instance has established or assessed the facts, the Court of Justice has
jurisdiction under art 168a of the Treaty to review the legal characterisation of those facts by the Court of
First Instance and the legal conclusions it has drawn from them (see esp the order in San Marco [1996]
ECR I-4435 (para 39)).
22.
The Court of Justice thus has no jurisdiction to establish the facts or, in principle, to examine the
evidence which the Court of First Instance accepted in support of those facts. Provided that the evidence
has been properly obtained and the general principles of law and the rules of procedure in relation to the
burden of proof and the taking of evidence have been observed, it is for the Court of First Instance alone
to assess the value which should be attached to the evidence produced to it (see esp the order in San
Marco [1996] ECR I-4435 (para 40)). The appraisal by the Court of First Instance of the evidence put
before it does not constitute, save where the evidence has been fundamentally misconstrued, a point of
law which is subject, as such, to review by the Court of Justice (see the judgment in Hilti v European
Commission Case C-53/92 P [1994] ECR I-667 (para 42)).

482

[1998] All ER (EC) 483

Drouot Assurances SA v Consolidated Metallurgical Industries (CMI

Industrial Sites) and others
(Case C-351/96)

EUROPEAN COMMUNITY; Other European Community: SHIPPING

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES GULMANN (PRESIDENT OF THE CHAMBER), WATHELET, MOITINHO DE ALMEIDA,
EDWARD (RAPPORTEUR) AND SEVN ADVOCATE GENERAL FENNELLY
13 NOVEMBER 1997, 15 JANUARY, 19 MAY 1998
European Community Jurisdiction Lis alibi pendens Definition of same parties Vessel sinking with
cargo Insurer of vessels hull refloating vessel Owner and insurer of cargo bringing action in Dutch
court for declaration of non-liability for contribution to general average Insurer of hull bringing
subsequent action in French court claiming contribution to general average from owner and insurer of
cargo Latter claiming lis alibi pendens Whether insurer of vessels hull same party as insured
Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968, art
21.

A South African company, CMI, arranged to have a cargo of ferrochromium transported from the
Netherlands to France on the Rhine. The vessel chartered for that purpose foundered with its cargo in
Dutch inland waters and Drouot, the insurer of the vessels hull, had the vessel refloated at its own
expense thereby enabling the cargo to be salvaged. In August 1990 CMI and Protea (the insurers of the
cargo) brought an action in the Rotterdam district court against the owner and charterer of the vessel for a
declaration, inter alia, that they were not obliged to contribute to the general average. In December 1990
Drouot brought proceedings before the commercial court in Paris against CMI, Protea and GIE (Proteas
agents at the time of the experts report on the salvaging costs) for payment of the sum set for their
contribution to the general average. CMI and Protea raised an objection of lis alibi pendens in the French
court on the grounds of their earlier action before the Dutch court. The court rejected that objection on the
grounds that neither the parties nor the subject matter of the two actions were the same. Following a
series of appeals, the Cour de Cassation stayed the proceedings and referred to the Court of Justice of
the European Communities for a preliminary ruling the question whether, in such circumstances, the
insurer of the hull of a vessel was deemed to be the same person as its insured for the purposes of the lis
pendens provisions of art 211 of the Convention on Jurisdiction and the Enforcement of Judgments in Civil
and Commercial Matters 1968.
1
Article 21, so far as material, is set out at p 496 g, post

Held Where, having regard to the subject matter of two disputes, there was such a degree of identity
between the interests of an insurer and those of its insured that a judgment delivered against one of them
would have the force of res judicata as against the other, the insurer and the insured had to be considered
to be the same party for the purposes of art 21 of the convention. However, that provision did not preclude
the insurer and its insured, where their interests 483 diverged, from asserting their respective interests
before the courts as against the other parties concerned and, in those circumstances, it was for the
national court to ascertain whether the interests of the two parties could be considered to be identical.
Moreover, the existence of a national procedural rule which precluded insurers from being brought into an
action was of no relevance to the solution of the dispute. Accordingly, art 21 would only apply in the
instant case if it could be established, with regard to the subject matter of the two disputes, that the
interests of Drouot were identical and indissoluble from those of the owner and charterer of the vessel
(see p 498 h j and p 499 c to e g to h, post).

Notes
For the Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters
1968, art 21 (as set out in Sch 1 to the Civil Jurisdiction and Judgments Act 1982), see 11 Halsburys
Statutes (4th edn, 1991 reissue) 1147.

Cases cited
Gubisch Maschinenfabrik KG v Palumbo Case 144/86 [1987] ECR 4861.
Maciej Rataj (The), Tatry (cargo owners) v Maciej Rataj (owners) Case C-406/92 [1995] All ER (EC) 229,
[1994] ECR I-5439, ECJ.
Zelger v Salinitri Case 129/83 [1984] ECR 2397.

Reference
By order of 8 October 1996, the French Cour de Cassation referred to the Court of Justice of the
European Communities for a preliminary ruling under the Protocol of 3 June 1971 on the interpretation by
the Court of Justice of the Convention on Jurisdiction and the Enforcement of Judgments in Civil and
Commercial Matters 1968, a question (set out at p 497 j to p 498 a, post) on the interpretation of art 21 of
that convention, as amended. That question was raised in proceedings brought by Drouot Assurances
SA, a company incorporated under French law (Drouot), against Consolidated Metallurgical Industries
(CMI) and Protea Assurance (Protea), companies incorporated under South African law, and Groupement
dIntrt conomique (GIE) Runion Europenne (GIE Runion), on the subject of the cost of refloating a
barge known as the Sequana which foundered in the inland waters of the Netherlands on 4 August 1989.
Written observations were submitted on behalf of: Drouot, by V Delaporte, of the Paris Bar; GIE Runion,
by D Le Prado, of the Paris Bar; the French government, by C de Salins, Head of Sub-directorate in the
Legal Directorate of the Ministry of Foreign Affairs, and J-M Belorgey, Charg de Mission in that
directorate, acting as agents; the German government, by J Pirrung, Ministerialrat in the Federal Ministry
of Justice, acting as agent; and the European Commission, by X Lewis, of its Legal Service, acting as
agent. Oral observations were made by: Drouot, represented by V Delaporte; CMI and Protea,
represented by J-C Balat, of the Paris Bar; of the French government, represented by J-M Belorgey; and
the Commission, represented by X Lewis. The language of the case was French. The facts are set out in
the opinion of the Advocate General.

15 January 1998.

The Advocate General (N Fennelly)

delivered the following opinion.

IINTRODUCTION
1.
The Court of Justice of the European Communities is asked, in this reference from the French Cour de
Cassation, to interpret the notion of the same 484 parties in art 21 of the Convention on Jurisdiction and
the Enforcement of Judgments in Civil and Commercial Matters 1968, as amended by the 1978
Accession Convention (Denmark, Ireland, United Kingdom) and the 1982 Accession Convention
(Greece)2. The issue is whether, in the context of a claim in general average made by an insurer in a
French court and an earlier claim for a negative declaration against the insured in a Dutch court that there
is no such liability, there is a lis alibi pendens for the purposes of art 21. Accordingly, the issue effectively
is whether the insured should be treated as the same party as its insurer.
2
Article 21 has been amended by art 8 of the 1989 Accession Convention (Spain, Portugual) (the San Sebastian
Convention), but the amended version entered into force between France and the Netherlandsthe two contracting
states concerned by the legal proceedings brought in the present caseonly on 1 February 1991. Although the facts of
this case arose in 1990, no changes made by the San Sebastian Convention are material to it.

IITHE LEGAL AND FACTUAL BACKGROUND TO THE REFERENCE

AThe relevant provisions of the convention

2.
Title II of the convention deals with jurisdiction. Whereas the general rules are set out in s 1, various
special rules qualifying those general rules are established by ss 2 to 9.
3.
Section 8 is entitled Lis Pendensrelated actions and consists of arts 21 to 23. Prior to its
amendment by the 1989 Accession Convention (Spain, Portugual) (the San Sebastian Convention), art 21
of the convention was worded as follows:
Where proceedings involving the same cause of action and between the same parties are
brought in the courts of different Contracting States, any court other than the court first seised shall
of its own motion decline jurisdiction in favour of that court.
A court which would be required to decline jurisdiction may stay its proceedings if the jurisdiction
of the other court is contested.
4.
In the case of related actions, the rule under art 22 of the convention is that the court other than that
first seised may stay its proceedings. Related actions are defined as those which are so closely
connected that it is expedient to hear and determine them together to avoid the risk of irreconcilable
judgments resulting from separate proceedings.
5.
Title III of the convention concerns Recognition and Enforcement. In accordance with the general
objective of the convention3, the general principle stated in art 26 is that a judgment given in a
Contracting State shall be recognized in the other Contracting States without any special procedure being
required. However, art 27 prescribes a limited number of grounds upon which such recognition may be
refused. For the purpose of the present case, only the third indent is potentially relevant: (3) if the
judgment is irreconcilable with a judgment given in a dispute between the same parties in the State in
which recognition is sought.
3
The first recital in the preamble to the convention refers to the desire of the contracting states to secure the
simplification of formalities governing the reciprocal recognition and enforcement of judgments of courts or tribunals.

485

BFactual background and procedure before the referring court

6.
Consolidated Metal Industries (CMI) charged Mr Velghe to transport on a barge, known as the
Sequana, a cargo of ferrochromium from Rotterdam in the Netherlands to Garlinghem-Aire-la-Lys in
France4. The vessel, having begun to take in water, foundered in Netherlands inland waters early on 4
August 1989, but not before the captain was apparently able to steer it from the main channel. Drouot
Assurances SA (Drouot), the insurer of the vessel (or hull), caused the ship to be refloated at its own
expense, thus salvaging CMIs cargo. Drouot brought proceedings on 11 and 13 December 1990 before
the Tribunal de Commerce (the Commercial Court), Paris5, against CMI, Protea Assurance (Protea), a
South African company, the insurer of the cargo, as well as Proteas European representative, the
Groupement dintrt conomique (GIE) Runion europenne (GIE) 6, for payment of the sum of f
99,48553, being the figure set by the average adjuster as the amount of the contribution of CMI and
Protea to the general average7. However, CMI and Protea pleaded, by way of a procedural objection to
the French action, a lis alibi pendens arising out of an action which they had earlier brought against Mr
Walbrecq and Mr Velghe8 before the Arrondissementsrechtbank (the District Court), Rotterdam, on 31
August 19909. According to the case file and the observations submitted to the court, CMI and Protea
sought, in the Netherlands action, a declaration to the effect that they were not obliged to contribute to the
general average. Such a negative declaration, which, apparently, could not have been sought if the action
had been initiated in France, was sought in the alternative to a claim that Mr Velghe be held responsible
for the occurrence of the accident on the ground that, as captain, he had permitted the barge to become
unseaworthy by overloading it at Rotterdam.
4
According to the order for reference, the vessel was owned by Mr Walbrecq and chartered by Mr Velghe. However, it
emerged at the oral hearing that Mr Walbrecq, having died in 1981, had actually been replaced as owner of the
Sequana by Mr Velghe sometime before the accident occurred. In their written observations Drouot and the Commission
submitted that Mr Velghe was also the captain of the barge at the time it sank, and that the barge had in fact been
chartered by another company not involved in either of the actions at issue in the main proceedings. This view, which
 was generally accepted at the hearing, is consistent with the judgment of the Cour dAppel, which is included in the case
file sent to the court.
5
For convenience, I shall hereinafter refer to this action as the French action.
6
It appears from Drouots written observations that the action against GIE was initiated separately on 11 February 1991.
7
The nature of the system of general average will be discussed below (see esp paras 1719). An average adjuster is a
professional who specialises in determining the amounts of the contributions which should be made, respectively, by
each of the participants in the venture to which the general average relates.
8
In view of the death in 1981 of Mr Walbrecq, the CMI and Protea action may now only concern Mr Velghe and the status
of Mr Walbrecq is not material to the reference before the court.
9
This action will for convenience hereinafter be referred to as the Netherlands action. When reference is hereinafter
made to both actions collectively, they will be described as the two actions.

7.
The lis alibi pendens plea was rejected on 11 March 1992 by the Tribunal de Commerce on the basis
that the parties to the two actions were not the same; viz Drouot was not a party to the Netherlands action
and Messrs Velghe and Walbrecq were not parties to the proceedings before it. Moreover, in the view of
the Tribunal de Commerce, the issues in the two actions were not the same. The defendants appealed to
the Cour dAppel (the Court of Appeal), Paris.
8.
According to the judgment of the Cour dAppel, CMI and Protea contended before that court that the
object of the two actions was the same and that Drouot was not a party to the Netherlands action only
because Dutch procedural rules 486 did not permit insurers to be named. In its judgment of 29 April 1994,
the Cour dAppel took the view that it was uncontested that Netherlands procedural rules precluded the
possibility for an insurance company to be present in a case involving its insured. Having recited the
broader scope of the Netherlands action (including, as it did, a claim regarding the owners liability for the
unseaworthy state of the vessel), it held that, in fact, it, none the less, encompassed the subject matter of
the French action. Furthermore, it held that Drouot could be regarded as a party to the Netherlands action
through the intermediary of the insured. Accordingly, the plea of lis alibi pendens was upheld.
9.
On appeal to the Cour de Cassation, Drouot contended, primarily, that the Cour dAppel should not
have upheld the lis alibi pendens plea, since neither the nature of the proceedings nor the identity of the
parties in the two actions was the same, and that the impugned judgment was incompatible with art 21 of
the convention10.
10
At the oral hearing, counsel for Drouot explained that, under the procedural rules of the Cour de Cassation, it had not
been possible for it to question the finding of the Cour dAppel regarding its presence through its insured in the
Netherlands action.

10.
The Cour de Cassation, being of the view that the appeal before it turned on the interpretation of the
concept of the same parties used in art 21 of the convention 11, decided to refer the following question to
the Court of Justice pursuant to arts 1 to 3 of the Protocol of 3 June 1971 on the interpretation of the
convention (see OJ 1978 L304 p 97):
11
The reference was received at the court on 25 October 1996. The report of the Cour de Cassations rapporteur, included
in the case file sent to this court, assists in appreciating the concerns that motivated the reference in the instant case.
The report notes that the courts case law regarding art 21 of the convention requires that it be interpreted autonomously
and that this would appear to preclude the application of the French-law principle under which there would be identity of
parties if the party opposing the plea of lis alibi pendens were represented by another party in allegedly related foreign
proceedings. While expressing some doubt as to the precise nature of the principle of Netherlands law relied upon in the
judgment of the Cour dAppel, the report suggests that the presumed presence at issue could satisfy the French-law
notion of effective representation.

whether, with regard in particular to the independent concept of same parties used in
Article 21 of the Brussels Convention, there is inter-state lis alibi pendens for the purposes of that
provision where a court of one Contracting State is seised by the insurer of a vessel that has been
shipwrecked with an action seeking from the owner and the insurer of the cargo on board partial
reimbursement, by way of contribution to the general average, of the refloating costs, when a court
of another Contracting State was seised previously by that owner and insurer with an action against
the owner and the charterer of the vessel for a declaration that they were not obliged to contribute
to the general average, and the court seised second declines jurisdiction, despite the parties in the
two cases not being strictly identical, on the ground that the procedural law applicable before the
court seised first restricts the opportunity for an insurer to be party to proceedings in which the
insured is involved so that the insurer of the vessel is in fact also involved, through the
intermediary of the insured, in the case brought first.

IIIOBSERVATIONS
11.
Written and oral observations have been submitted by Drouot, the French Republic and the European
Commission. On the other hand, GIE and the Federal Republic of Germany merely submitted written
observations while CMI 487 and Protea submitted joint oral observations. They may be summarised as
follows.
12.
At the hearing, counsel for Drouot pointed out that the general provisions of maritime law apply to
fluvial transport on the Rhine and Moselle rivers, under which, he submitted, Drouots contract of
insurance, in the absence of an express stipulation to the contrary, must be regarded as merely covering
the hull. Vessel insurance covers only liability for damage done by the vessel to other vessels or to port or
riverside installations. Drouot also submitted that the maritime rules of general average apply to
navigation on the Rhine and Moselle rivers (see further paras 17 to 19, below). Based on the case law
concerning art 21 of the convention12 Drouot submits that no lis alibi pendens may arise unless the same
parties figure in both proceedings. The criteria set out in art 21 must be construed autonomously from
related concepts in the respective laws of the contracting states. Both the judgment and Advocate
General Tesauros opinion in The Maciej Rataj, Tatry (cargo owners) v Maciej Rataj (owners) Case C-
406/92 [1995] All ER (EC) 229, [1994] ECR I-5439 support the view that there is lis alibi pendens only
when there is a strict identity of parties in the two procedures. At the hearing, Drouot submitted that, for
parties to be identical, they must have a common interest to defend or, at least, a common argument to
make, which, in its view, cannot be the case regarding its supposed representation by Mr Velghe in the
Netherlands action. It claims that, as insurer of the hull, it was not responsible for the general presumed
liability of the boat owner. In its written observations, Drouot submits that the status of a person as a party
before a court allegedly first seised must be determined by the law of the forum where the plea of lis alibi
pendens is raised, ie by French law in the present case 13. Finally, Drouot submitted that, apart from the
fact that it was neither voluntarily nor involuntarily a party to the Netherlands action, it had no interest in
that action, since, as insurer of the Sequana, it was liable to reimburse Mr Velghe in respect of his
contribution to the general average, notwithstanding the possible responsibility for the occurrence of the
accident.
12
It refers to Zelger v Salinitri Case 129/83 [1984] ECR 2397, Gubisch Maschinenfabrik KG v Palumbo Case 144/86
[1987] ECR 4861 and The Maciej Rataj, Tatry (cargo owners) v Maciej Rataj (owners) Case C-406/92 [1995] All ER
(EC) 229, [1994] ECR I-5439.
13
Drouot cites Zelger [1984] ECR 2397 (para 15), in favour of this contention and alleges that, in French law, it could not,
by virtue merely of being the insurer of the Sequana, be considered a party to the Netherlands action.

13.
In their oral observations, CMI and Protea, first, contended that Drouot was not merely, as claimed, the
insurer of the vessel but also of the personal liabilities of its captain and owner, namely Mr Velghe.
Secondly, counsel for CMI and Protea pointed out that his clients had initially brought an action in France
before the Tribunal de Commerce, Bthune (the place of Mr Velghes residence) against both Mr Velghe
and Drouot, in which, it was alleged, Drouot claimed that the action against it should have been brought in
Rotterdam. Accordingly, CMI and Protea instituted the Netherlands action but did not cite Drouot because
of the alleged Dutch procedural rules which prevent the insurer being a party to proceedings brought
against its insured. As regards the interpretation of art 21 of the convention, CMI and Protea contended
that the most important consideration was the avoidance of the adoption of incompatible decisions by
courts in different contracting states. They relied in particular on Gubisch Maschinenfabrik KG v Palumbo
Case 144/86 [1987] ECR 4861 as establishing the 488 identity of two causes of action concerning, on the
one hand, the annulment of a contract and, on the other, its enforcement.
14.
France submits that the autonomy of the conditions required for the occurrence of a lis alibi pendens
by art 21 of the convention would be undermined if the peculiarities of the procedural law of a contracting
state were to determine whether the same parties are present 14. It laid particular emphasis on the need
to respect the rights of defence, or fair hearing, of the insured. An insurer is not represented in national
proceedings by its insured. It does not have access to the court in legal proceedings involving its insured,
so as to permit it to present its own arguments or defend its position, and the legal rights and interests of
an insurer and its insured often do not coincide. Thus, even assuming identity between insurer and
insured, the rights of defence of insurance companies could only properly be guaranteed if art 21 of the
convention were interpreted as requiring, as a prerequisite to upholding a plea of lis alibi pendens, that
the parties supposedly present in both actions actually be involved as principal parties.

14
It refers esp to the opinion of Advocate General Tesauro in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-
5439 (para 19).

15.
Germany also stresses the importance of adopting an autonomous interpretation but argues for a
broad concept of same parties so as to avoid the occurrence of irreconcilable judgments within the
meaning of the third indent of art 27 of the convention. Given the relationship between the concepts of lis
alibi pendens and res judicata, Germany contends that parties to a second action, who are not formally
identical to those involved in a first action, should not be regarded as being the same, unless they would
be bound by the effects of the judgment given by the court first seised, so as to create a danger of
irreconcilable judgments if the plea were not upheld. For this purpose, the second court seised should
refer to either the substantive or procedural rules of the first court seised, as well as its own, in order to
determine whether any binding third party effects of the first courts judgment would be recognised.
Finally, Germany emphasises the need to ensure that art 21 be applied in a manner that respects the
requirements of effective access to court. Thus, where a plea of lis alibi pendens is upheld but the first
action is unsuccessful, it must remain possible for the party whose second action was affected by that
plea subsequently to resume its action before the second court.
16.
The Commission, though recognising that the question referred focuses only on the concept of the
same parties, submits that the two actions may be regarded as having the same cause of action and
object15. On the identity of the parties, the Commission observes that the question referred raises the
novel issue as to whether art 21 of the convention permits a court before which a plea of lis alibi pendens
is raised to go beyond the formal identification of the parties in the proceedings before the court first
seised. The concept of the same parties should, in its view, be interpreted strictly.
15
At the oral hearing the Commissions agent nevertheless expressed doubts as to whether this initial assessment was
correct.

Besides the need to preserve the autonomy of that concept from the laws of the contracting state
concerning matters such as subrogation, there are various other reasons which militate against
assimilating an insurer with its insured. The interests of an insured and an insurer are not necessarily
identical; the efficient administration of justice may not be well served by making the insurer await the
outcome of a first set of proceedings to which it is not a party whenever it wishes 489 to claim, in a
different action, that, for example, no subrogation actually arises. In the Commissions view the wording of
art 21, by referring to the same parties and not to other parties who may have rights or liabilities arising
from the rights or liabilities of the actual parties, supports this approach. At the hearing, the Commissions
agent questioned the practicality of the approach suggested by Germany: if the second court seised were
obliged systematically to assess the binding effects of a potential judgment to be given by the first court
seised before upholding a plea, the practical application of the notion of lis alibi pendens in art 21 would
become unduly complicated.

IVANALYSIS

AIntroduction
17.
The disputed claim in the French action concerns a contribution to general average. Although, as
France rightly points out, the court has not been asked to assess the similarity of that cause of action with
those involved in the Netherlands action, I think that it would still be helpful to take cognizance of the
nature of a claim for general average contribution. A short review of the particularities of the legal doctrine
of general average will, I think, assist the application in this particular case of the concept of identity of
parties in art 21 of the convention.

BThe notion of general average

18.
The notion of general average lies at the heart of both the proceedings instituted by Drouot in France
and the alternative claim by Protea and CMI in the Netherlands 16. It is a notion of maritime law of great
antiquity. Its origins can be traced to the Rhodian Sea Law, which also found later expression in Roman
Law, whereby the owner of cargo subjected to jettison (jactus factus levandae navis gratia) to save a
vessel could spread his loss, by claiming contribution, among the owner of the vessel and other cargo
owners (see eg Arnoulds Law of Marine Insurance and Average (1981, 16th edn) Vol 2, para 916; Ripert
Droit Maritime (1953, 4th edn) Tome III, paras 2213ff). It came, over time, to include claims based on
other types of damage and expense incurred to avoid it. In modern times, it generally takes the form, in
practice, of the incorporation into contracts of affreightment and policies of marine insurance of the
conventional York-Antwerp Rules. These rules, regularly revised since the first International General
Average Rules were adopted at York in 1864, appear to have been given effect in French law by Law No
67-545 of 7 July 1967 concerning incidents at sea (Loi n 67-545 du 7 juillet 1967 relative aux vnements
de mer). The nautical event giving rise to the claims at issue before the Dutch and French courts did not,
however, take place at sea but on a part of the Rhine and Moselle waterway system. Counsel for Drouot
explained at the hearing that, by virtue of a law of 1895, passed during the period of the attachment of
Alsace to Germany, maritime law was applied to navigation on the Rhine and Moselle. The rules
specifically applied in the charter-party in this case are those of Rhine, Antwerp and Rotterdam and, as
appears from the case file, general average is therein defined 490 as the sacrifices and expenses
reasonably made and incurred with the object of saving a boat and its cargo from common peril.
16
The term general average in the English language, is liable to lead to confusion, since it is quite distinct from the word
average in its normal sense. In fact, it shares an etymology with the French word avaries communes, the Dutch averij
grosse and the German groe Haverei (see Ledocte Legal Dictionary in Four Languages (1982); the word average in
this context thus means damage, which is the starting point of any general average claim.

19.
The gist, therefore, of the notion of general average is that participants in a common commercial
adventure should contribute equitably to the damage or loss of one of the participants who has, for the
benefit of all, suffered a sacrifice or, by extension, reasonably incurred expense to prevent loss. Insurers
of the hull and cargo are treated as participants and may claim or be liable for general average
contribution. A claim for general average contribution is, therefore, not necessarily made by an insurer.
20.
That conclusion seems to me to have a bearing on the question of the capacity of Mr Velghe (and Mr
Walbrecq) before the Dutch court. It is true that an insurer, having fully discharged its liability to its
insured, may in certain circumstances, by virtue of a right of subrogation, effectively step into the shoes of
the insured so as to pursue claims (including a claim for general average contribution) against third
parties. This right may, depending on the applicable law, be exercised either in the name of the insured or
by the insurer in its own name, but only so as to invoke the rights of the insured and after payment. That
clearly does not apply to the Netherlands action; nor does it apply to the French action where Drouots
claim is not based on subrogation to any rights of Mr Velghe but on its general average claim. The precise
nature of the action before the Dutch court is known only indirectly; but it is generally agreed that it
includes, as one element, a claim for a negative declaration that CMI and Protea be held not liable to
contribute to general average. Thus, in the particular context of a general average claim, it is difficult to
see how the prima facie lack of identity between Mr Velghe and Drouot should be merged into an identity
of interest.
21.
However, there are other problems of a practical nature to be addressed in considering whether a
party should be treated as representing his insurers for the purposes of art 21 of the Brussels Convention.
The court in the contracting state where the issue of lis alibi pendens in the sense of art 21 of the
convention is raised, and which is informed of the existence of a prior action claimed to be between the
same parties in another contracting state, will need to conclude that an insured, a named party in the
other contracting state, should be treated as being the same party as his insurer. To start with, that court
will need to investigate the relationship between insurer and insured. It is notorious that disputes between
insurer and insured take a great variety of forms. I will cite only a few. The very existence of the policy
may be in dispute; if its existence is established or admitted its validity may be contested on grounds of
fraud, misrepresentation or non-disclosure; its application to the particular loss may be in doubt, as may
be the quantum of the loss, the means of proving it or the time for payment. Even in the present case,
there is a dispute between CMI and Protea, on the one hand, and Drouot, on the other (though not, so far
as I am aware, between Drouot and Mr Velghe), regarding the scope of the cover under Drouots policy of
insurance. The second national court, in my hypothesis, might have to reach a conclusion on any of those
potentially complex issues by reference to the law applicable in the other, or even a third, contracting
state.
22.
The comparison of the capacity in which the claim for general average contribution is made by Drouot
in the French court with that of Mr Velghe in the Dutch court is the key to the application of art 21 of the
convention. The Drouot claim is made in its capacity as a party to the enterprise having made a sacrifice.
It is not made as representative of Mr Velghe. Indeed, any claim made by the owner of a vessel which
had sunk against the owner of the damaged cargo would 491 be met, to say the least, with surprise,
unless it were itself based on general average, ie on sacrifice. However, in the present case, Mr Velghe
would not appear to have participated in the salvage of the vessel and, thus, would scarcely have
incurred any such sacrifice.
23.
Consequently, viewing the matter exclusively within the context of the present claim for general
average contribution, there is no such identity of interest between Drouot and Mr Velghe as would justify,
even if that were permitted, overlooking their nominal and real differences of identity so as to treat them
as the same party.

CThe same parties

24.
In upholding the plea of lis alibi pendens in the present case, the Cour dAppel, in determining the
identity of the parties to the Netherlands action, has relied upon a principle which, in its opinion, forms
part of the procedural law of the Netherlands. It is right to point out that the provisions of the convention
were not cited to the Cour dAppel. Nevertheless, the approach which it has adopted is not, in my opinion,
compatible with the courts decisions in Gubisch and The Maciej Rataj [1995] All ER (EC) 229, [1994]
ECR I-5439, where the court declared unambiguously that the substantive conditions prescribed by art
21 of the convention for a successful plea of lis alibi pendens must be regarded as independent (see the
judgment in Gubisch [1987] ECR 4861 (para 11))17. These substantive conditions represent a considered
choice and the implicit rejection of the possibility of referring to the term lis alibi pendens as used in the
different national legal systems of the contracting states18. The application of the convention concept of
the same parties cannot, in my view, depend on the existence and scope of the Netherlands-law
principle relied upon by the Cour dAppel, since that represents a consideration of the law of the first court
seised.

17
See also the judgment in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439 (para 30).
18
See footnote 11, above, regarding the possible position in French law.

25.
Next, it is necessary to assess what guidance the Court of Justices case law furnishes in respect of
the autonomous convention concept of the same parties. Zelger v Salinitri Case 129/83 [1984] ECR
2397 is not relevant because it only concerned the procedural formalities regulating the temporal point at
which a court may be regarded as having been seised of an action (see [1984] ECR 2397 (paras 1316)).
In Gubisch [1987] ECR 4861 (para 13), the court was concerned with a situation whose salient features
were that
one of the parties [had] brought an action before a court of first instance for the enforcement of
an obligation stipulated in an international contract of sale [while] an action was subsequently
brought against him by the other party in another Contracting State for the rescission or discharge
of the same contract.
The court declared that the conditions laid down by art 21 are exhaustive; namely, the two actions must
be between the same parties and involve the same cause of action and the same subject-matter (see
[1987] ECR 4861 (para 14))19. It proceeded, in circumstances where the present issue of identity of
parties was not in doubt, to take the view that the concept of the same cause of action cannot be
restricted so as to mean two claims which are entirely identical (see [1987] ECR 4861 (para 17)). That
the court was clearly motivated by the need to avoid 492 conflicting judgments emerges clearly from its
statement that in circumstances such as those of the case before it ([1987] ECR 4861 (para 18)):
 19
The court, by oversight, stated ([1987] ECR 4861 (para 14)) that onlythe German version of Article 21 (which refers
to: Klagen wegen desselben Anspruchs zwischen denselben Parteien anhngig gemacht) does not expressly
distinguish between the terms subject-matter and cause of action, and that accordingly it must be construed in the
same manner as the other language versions, all of which make that distinction (my emphasis). In fact, at least the
English and Irish versions of art 21 of the convention also make no such distinction; thus, the English version refers to
proceedings involving the same cause of action and between the same parties while the Irish version refers to
imeachta leis an gcis channa chaingne agus idir na pirtithe canna (my emphases). The Danish text, for example,
would also appear merely to contain two conditions: derhar samme genstand og hviler pa samme grundlag (my
emphasis). However, in its judgment in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439 (para 38), the
court implicitly corrected the slip made in Gubisch by recognising that: the English version of art 21 does not expressly
distinguish between the concepts of object and cause of action. Reiterating the principle declared in Gubisch, the
court stated that that language version must however be construed in the same manner as the majority of the other
language versions in which that distinction is made.

There can be no doubt that a judgment given in a Contracting State requiring performance of
the contract would not be recognized in the state in which recognition is sought if a court in that
state had given a judgment rescinding or discharging the contract. Such a result, restricting the
effects of each judgment to the territory of the state concerned, would run counter to the objectives
of the convention, which is intended to strengthen legal protection throughout the territory of the
Community and to facilitate recognition in each Contracting State of judgments given in any other
Contracting State.
26.
CMI and Protea have argued from Gubisch for a broad and flexible interpretation of the concept of the
same cause of action and, by extension, of the same parties as those terms are used in art 21 of the
convention. It is true that, in Gubisch, the court held that the concept of the same subject-matter, which,
in effect, it interpreted into the English text by reference to the other language versions, could not be
restricted so as to mean two claims which are entirely identical ([1987] ECR 4861 (para 17)). In practice,
it applied that reasoning to the two actions one of which was brought to enforce and the other to rescind
or discharge the same contract. In doing so, it attached great importance to the purpose expressed, inter
alia, in art 27(3) of the convention of avoiding irreconcilable judgments between the same parties and how
such judgments could arise if the competing claims had to be entirely identical before a lis alibi pendens
plea could be upheld. That reasoning is not, however, equally applicable to the concept of the same
parties, since the judgment proceeds on the assumption that, whatever differences exist in the subject
matter, the parties are the same. Nothing in the judgment, in the text of art 21 or in the purpose of the
convention requires that a flexible approach be adopted in that instance. The contrary is rather the case.
Judgments are, in my view, truly irreconcilable only if they are contrary and given in actions between the
same parties.
27.
The judgment in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439 confirms this view and
has the merit of having addressed the issue of identity of parties, albeit not in the radical form suggested
in the instant case. The court had to examine whether the convention could be regarded as applicable in
the case of two sets of proceedings involving the same cause of action but where some but not all of the
parties were the same; ie where at least one of the plaintiffs 493 and one of the defendants to the
proceedings first commenced were among the plaintiffs and defendants in the second proceedings, or
vice versa (see [1995] All ER (EC) 229, [1994] ECR I-5439 (paras paras 2930)) 20. The court first agreed
with the recommendation of its Advocate General to the effect that the identity of the parties cannot
depend on the procedural position of each of them in the two actions, and that the plaintiff in the first
action may be the defendant in the second (see para 31) 21. Having regard to the wording and objective of
art 21, to prevent parallel proceedings, the court declared that art 21 must be understood as requiring,
as a condition of the obligation of the second court seised to decline jurisdiction, that the parties to the two
actions be identical (see paras 3233; my emphasis)22. It follows that this requirement must be strictly
construed.
20
The fifth question concerned the alleged identity of the causes of action in the two actions; namely, an action brought by
cargo owners in one contracting state (the Netherlands) in respect of damage caused during transit to their goods and
an action previously brought in another contracting state (the United Kingdom) by the shipowner whereby the latter
effectively sought a negative declaration regarding its possible liability for the damage to the cargo. The court took the
view that such causes of action were the same, since the positive and negative terms in which they were couched did
not render their object (to wit the determination of liability) different; see esp [1995] All ER (EC) 229, [1994] ECR I-
5439 (para 43).
21
See also the opinion of Advocate General Tesauro: [1995] All ER (EC) 229, [1994] ECR I-5439 (para 14).
22
In the English language, at least, the adjective identical refers to a thing that agrees in every detail with another thing
(see eg The Concise Oxford Dictionary (1990) p 585).

28.
Thus, in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439 (para 34), the court ruled that
where some of the parties are the same as the parties to an action which has already been
started, art 21 requires the second court seised to decline jurisdiction only to the extent to which the
parties to the proceedings pending before it are also parties to the action previously started before
the court of another contracting state. (My emphasis.)
I agree with the Commissions contention that the application of art 21 of the convention cannot depend
on the court before which the plea of lis alibi pendens is raised carrying out an inquiry into the true
capacity of parties before the court of a different contracting state.
29.
 My view, therefore, is that the concept of same parties is to be interpreted literally and strictly. The
court has used the word identical. This means that not only must the parties to the two actions be the
same in the literal sense of the same natural or legal person, but also that they must appear in the same
right. In particular, a person suing in his own right and for his own benefit is obviously not to be equated
with the same person suing or being sued in a purely representative capacity, for example, as the legal
personal representative of a deceased person or a person under a disability, or in any of the wide range
of cases where a person may, in law, be named to represent corporate bodies or their creditors in cases
of insolvency.
30.
This does not, to my mind, create too rigid a framework for the operation of art 21 of the convention.
On the contrary, it is in conformity with the objective of simplification of formalities governing the
reciprocal recognition and enforcement of judgments found in art 220 of the EC Treaty, which authorised
the member states, among other things, to enter into the convention. Simple and transparent rules,
capable of being applied on the basis of objective and readily accessible factors best serve this objective.
Cases where a broader discretion is 494 required to stay proceedings where related actions are brought
in the courts of different contracting states are catered for by art 22 of the convention. As Advocate
General Tesauro explained in his opinion in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439
(para 28), the idea of irreconcilable judgments in the third paragraph of art 22 does not bear the same
restrictive meaning as in art 27(3). According to him, art 22
is intended rather to improve coordination of the judicial function within the Community and to
avoid conflicting and contradictory decisions, even where the separate enforcement of each of
them is not precluded
that is to say
the rationale of the provision is therefore to encourage harmonious judicial decisions and
thereby obviate the danger of judgments which conflict with each other, albeit only as regards their
reasoning.
An unduly broad interpretation of the requirements of art 21 would run the risk of confusing related actions
with lis alibi pendens. No question has, however, been referred to the court in this case with regard to the
exercise of the discretionary power conferred by art 22.
31.
Moreover, I share the concerns expressed, particularly in the observations of France and the
Commission, that a more flexible approach to the application of the condition that the parties must be the
same in order for an obligation to decline jurisdiction to arise under art 21 of the convention could
seriously imperil the right to a fair hearing and, in some cases, even the efficient administration of justice.
In the instant case, it is apposite to recall that the principle of Netherlands procedural law which forms the
basis for the supposed implicit presence of Drouot in the Netherlands actionat least if the court were to
accept the uncontradicted description furnished by counsel for Drouot at the hearingwould appear to
have slipped onto the case file before the Cour dAppel23. As Advocate General Tesauro pointed out in
his opinion in The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439 (para 19), when
23
It should perhaps be added, although no criticism of the Cour dAppel is intended, that there is nothing on the case file
to indicate that its finding in respect of Dutch law was based either on expert evidence or other reliable sources of that
law.

reference to the domestic laws of contracting states [is] rendered necessary by the
incompleteness of the rules contained in the Brussels Convention [it] must be conducive to the
application of the provisions of the convention and may not in any circumstances lead to results
which conflict with its aims and rationale.
Accordingly, I do not think that the right of Drouot to a fair hearing would be served if art 21 of the
convention were to be interpreted as imposing an obligation, in circumstances such as those of the
present case, on the Cour dAppel to decline jurisdiction in respect of its claim in the French action when
its right to be heard in the Netherlands action would effectively depend on the attitude of Mr Velghe.
32.
The conclusion which must, to my mind, be drawn in circumstances such as those of the present case
is that no lis alibi pendens for the purposes of art 21 of the convention arises.
495

VCONCLUSION
33.
In the light of all the foregoing considerations, I recommend that the Court of Justice answer the
question referred by the French Cour de Cassation as follows:
No lis alibi pendens for the purposes of art 21 of the Convention on Jurisdiction and the
Enforcement of Judgments in Civil and Commercial Matters 1968, as amended, arises where a
court of one contracting state is seised by the insurer of a vessel that has been shipwrecked with
an action seeking from the owner and the insurer of the cargo on board partial reimbursement, by
way of contribution to the general average, of the refloating costs, when a court of another
contracting state was seised previously by that owner and insurer with an action against the owner
of the vessel for a declaration that they were not obliged to contribute to the general average.
19 May 1998.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By order of 8 October 1996, received at the Court of Justice of the European Communities on 25
October 1996, the French Cour de Cassation referred to the Court of Justice for a preliminary ruling under
the Protocol of 3 June 1971 on the interpretation by the Court of Justice of the Convention on Jurisdiction
and the Enforcement of Judgments in Civil and Commercial Matters 1968, a question on the interpretation
of art 21 of that convention, as amended (the convention).
2.
That question was raised in proceedings brought by Drouot Assurances SA, a company incorporated
under French law (Drouot), against Consolidated Metallurgical Industries (CMI Industrial Sites, CMI) and
Protea Assurance (Protea), companies incorporated under South African law, and Groupement dIntrt
conomique (GIE) Runion Europenne (GIE Runion), on the subject of the cost of refloating a barge
known as the Sequana which foundered in the inland waters of the Netherlands on 4 August 1989.
3.
Article 21 of the convention provides:
Where proceedings involving the same cause of action and between the same parties are
brought in the courts of different Contracting States, any court other than the court first seised shall
of its own motion decline jurisdiction in favour of that court.
A court which would be required to decline jurisdiction may stay its proceedings if the jurisdiction
of the other court is contested.
4.
In 1989 CMI arranged with a Mr Velghe to transport on the Sequana a cargo of ferrochromium
belonging to it from the Dutch port of Rotterdam to the French port of Garlinghem-Aire-la-Lys on the
Rhine. According to the referring court, at the time of these events, the Sequana belonged to a Mr
Walbrecq and was chartered by Mr Velghe.
5.
However, at the hearing the court was informed that, according to CMI and Protea, following the death
of Mr Walbrecq in 1981, Mr Velghe had become the owner of the Sequana and that, at the material time,
the vessel was chartered by another company. Drouot stated that it had no information on those matters,
but explained that, according to the custom followed on the Rhine, the master is the legal agent of the
owner and he is responsible for checking that the vessel is 496 fit to sail. Finally, according to CMI, Protea
and Drouot, Mr Velghe was the master of the Sequana at the time of the events in issue.
6.
On 4 August 1989 the Sequana foundered in the inland waters of the Netherlands. Drouot, the insurer
of the hull of the vessel, had it refloated at its own expense, thereby enabling CMIs cargo to be salvaged.
7.
On 11 and 13 December 1990, Drouot brought proceedings before the Tribunal de Commerce (the
Commercial Court), Paris, against CMI, Protea (the insurer of the cargo) and GIE Runion (Proteas
agent at the time of the experts report on the salvaging costs), for payment of the sum of f 99,48553, the
figure set by the average adjuster as the amount of the contribution of CMI and Protea to the general
average.
8.
However, in the French court, CMI and Protea raised an objection of lis alibi pendens on the basis of
an action they had brought on 31 August 1990 against Mr Walbrecq and Mr Velghe before the
Arrondissementsrechtbank (the District Court), Rotterdam, for a declaration, inter alia, that they were not
obliged to contribute to the general average.
9.
On 11 March 1992 the Tribunal de Commerce, Paris, rejected the plea of lis alibi pendens on the
ground that, as Drouot was not a party to the Netherlands action and Mr Walbrecq and Mr Velghe were
not parties to the French action, the parties to the two actions were not the same. Moreover, the Tribunal
took the view that the subject matter of the applications made in the two actions was not the same.
10.
CMI, Protea and GIE Runion then appealed to the Cour dAppel (the Court of Appeal), Paris. Before
that court, CMI and Protea argued that the only reason Drouot was not a party to the Netherlands action
was that procedural rules in the Netherlands did not permit insurers to be brought into the action. They
also argued that, since the subject matter of the dispute before the Netherlands court encompassed that
of the dispute before the French court, the subject matter of the two disputes was the same.
11.
In a judgment of 29 April 1994, the Cour dAppel, Paris, first noted that it was common ground that
Netherlands procedural rules restricted the opportunity for an insurer to be party to proceedings in which
the insured is involved. It then considered that Drouot was in fact present in the Netherlands action
through the intermediary of the insured. Finally, it held that, in the light of the arguments put forward by
CMI and Protea, the two disputes did have the same subject matter. Accordingly the plea of lis alibi
pendens was upheld.
12.
Drouot then appealed to the Cour de Cassation against that judgment, contending that the parties to
the two actions were not the same.
13.
The Cour de Cassation, taking the view that the appeal before it raised a difficulty concerning the
interpretation of art 21 of the convention, decided to stay proceedings and seek a ruling from the Court of
Justice on the question
whether with regard in particular to the independent concept of same parties used in Article
21 of the Convention of 27 September 1968 on Jurisdiction and the Enforcement of Judgments in
Civil and Commercial Matters, there is inter-state lis alibi pendens for the purposes of that provision
where a court of one Contracting State is seised by the insurer of a vessel that has been
shipwrecked with an action seeking from the owner and the insurer of the cargo on board partial
reimbursement, by way of contribution to the general average, of the refloating costs, when a court
of another contracting state was seised previously by that owner and insurer with an action against
497 the owner and the charterer of the vessel for a declaration that they were not obliged to
contribute to the general average, and the court seised second declines jurisdiction, despite the
parties in the two cases not being strictly identical, on the ground that the procedural law applicable
before the court seised first restricts the opportunity for an insurer to be party to proceedings in
which the insured is involved so that the insurer of the vessel is in fact also involved, through the
intermediary of the insured, in the case brought first.
14.
By that question, the national court is asking essentially whether art 21 of the convention is applicable
in the case of two actions for contribution to general average, one brought by the insurer of the hull of a
vessel which has foundered against the owner and the insurer of the cargo which the vessel was carrying
when it sank, the other brought by the latter two parties against the owner and the charterer of the vessel.
15.
As the parties to those two actions do not appear to be the same, it should be considered whether, in a
case such as that in the main proceedings, the insurer of the hull of a vessel must be deemed to be the
same person as its insured when applying the same parties criterion contained in art 21 of the
convention.
16.
According to consistent case law, the terms used in art 21 of the convention in order to determine
whether a situation of lis pendens arises must be regarded as independent (see the judgment in The
Maciej Rataj, Tatry (cargo owners) v Maciej Rataj (owners) Case C-406/92 [1995] All ER (EC) 229, [1994]
ECR I-5439 (para 30)).
17.
In the The Maciej Rataj [1995] All ER (EC) 229, [1994] ECR I-5439 (para 32), the court stressed that
art 21, together with art 22 on related actions, is contained in s 8 of Title II of the convention, a section
intended, in the interests of the proper administration of justice within the Community, to prevent parallel
proceedings before the courts of different contracting states and to avoid conflicts between decisions
which might result therefrom. Those rules are therefore designed to preclude, in so far as is possible, the
possibility of a situation arising such as that referred to in art 27(3), that is to say the non-recognition of a
judgment on account of its irreconcilability with a judgment given in a dispute between the same parties in
the state in which recognition is sought.
18.
In that same judgment the court held that, in the light of the wording of art 21 of the convention and the
objective set out above, that article must be understood as requiring, as a condition of the obligation of
the second court seised to decline jurisdiction, that the parties to the two actions be identical (see [1995]
All ER (EC) 229, [1994] ECR I-5439 (para 33)).
19.
It is certainly true that, as regards the subject matter of two disputes, there may be such a degree of
identity between the interests of an insurer and those of its insured that a judgment delivered against one
of them would have the force of res judicata as against the other. That would be the case, inter alia,
where an insurer, by virtue of its right of subrogation, brings or defends an action in the name of its
insured without the latter being in a position to influence the proceedings. In such a situation, insurer and
insured must be considered to be one and the same party for the purposes of the application of art 21 of
the convention.
20.
On the other hand, application of art 21 cannot have the effect of precluding the insurer and its
insured, where their interests diverge, from asserting their respective interests before the courts as
against the other parties concerned.
498
21.
In the present case, CMI and Protea made clear at the hearing that, in the Netherlands action, they
seek to have Mr Velghe declared exclusively liable for the foundering of the Sequana. As the insurer
merely of the hull of the vessel, however, Drouot takes the view that it cannot be held liable for the fault of
its insured and thus has no interest in the Netherlands action.
22.
 It appears, moreover, that, in the French action, Drouot has been acting not in its capacity as the
representative of its insured but in its capacity as a direct participant in the refloating of the Sequana.
23.
Thus, in this case, it does not appear that the interests of the insurer of the hull of the vessel can be
considered to be identical to and indissociable from those of its insured, the owner and the charterer of
that vessel. However, it is for the national court to ascertain whether this is in fact the case.
24.
In those circumstances, the existence or otherwise of a national procedural rule such as that
mentioned in the question referred for a preliminary ruling is of no relevance to the solution of the dispute.
25.
The answer to the question raised must thus be that art 21 of the convention is not applicable in the
case of two actions for contribution to general average, one brought by the insurer of the hull of a vessel
which has foundered against the owner and the insurer of the cargo which the vessel was carrying when
it sank, the other brought by the latter two parties against the owner and the charterer of the vessel,
unless it is established that, with regard to the subject matter of the two disputes, the interests of the
insurer of the hull of the vessel are identical to and indissociable from those of its insured, the owner and
the charterer of that vessel.

Costs
26.
The costs incurred by the French and German governments and by the European Commission, which
have submitted observations to the Court of Justice, are not recoverable. Since these proceedings are,
for the parties to the main proceedings, a step in the action pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the question referred to it by the
Cour de Cassation, France, by order of 8 October 1996, hereby rules: art 21 of the Convention on
Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters 1968, as amended, is not
applicable in the case of two actions for contribution to general average, one brought by the insurer of the
hull of a vessel which has foundered against the owner and the insurer of the cargo which the vessel was
carrying when it sank, the other brought by the latter two parties against the owner and the charterer of
the vessel, unless it is established that, with regard to the subject matter of the two disputes, the interests
of the insurer of the hull of the vessel are identical to and indissociable from those of its insured, the
owner and the charterer of that vessel.

499

[1998] All ER (EC) 500

Criminal proceedings against Goodwin and another

(Case C-3/97)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIRST CHAMBER)
JUDGES WATHELET (PRESIDENT OF THE CHAMBER), EDWARD AND JANN (RAPPORTEUR)
ADVOCATE GENERAL LGER
15 JANUARY, 12 MARCH, 28 MAY 1998
European Community Value added tax Supply of goods Defendants buying, selling and
manufacturing counterfeit perfumes without being registered for value added tax Defendants convicted
of fraudulently evading value added tax Whether supply of counterfeit perfume products a taxable
supply of goods Council Directive (EEC) 77/388, art 2.

The two defendants were accused respectively of purchasing and selling counterfeit perfume products
without being registered for value added tax, and of participating in the manufacture, production,
distribution and sale of counterfeit perfume products through a business organisation which was not
registered for value added tax. The Inner London Crown Court held that the supply of counterfeit
perfumes could be subject to value added tax under Council Directive (EEC) 77/388 on the harmonisation
of the laws of member states relating to turnover taxescommon system of value added tax: uniform
basis of assessment and convicted both defendants of fraudulently evading value added tax contrary to s
72(1)1 and (8) of the Value Added Tax Act 1994. The defendants appealed to the Court of Appeal,
contending inter alia that Directive 77/388 precluded the charging of value added tax on the supply of
counterfeit products. The Court of Appeal stayed the proceedings and referred to the Court of Justice of
the European Communities for a preliminary ruling the question whether, on a proper construction of art
22 of Directive 77/388, the supply of counterfeit perfumes constituted a supply goods which was subject
to value added tax.
1
Section 72, so far as material, provides: (1) If any person is knowingly concerned in the fraudulent evasion of VAT
he shall be liable(a) to a penalty of three times the amount of the VAT (8) Where a persons conduct must
have involved the commission by him of one or more offences under the preceding provisions he shall be liable
(a) to a penalty of three times the amount of any VAT that was or was intended to be evaded
 2
Article 2, so far as material, is set out at p 508 d, post

Held Directive 77/388 was based on the principle of fiscal neutrality which precluded a generalised
differentiation between lawful and unlawful transactions, except where, because of the special
characteristics of certain products, all competition between a lawful economic sector and an unlawful
sector was precluded. Although transactions involving counterfeit products infringed intellectual property
rights, any consequential prohibition of those goods was not linked to their nature, but to their detrimental
impact on the rights of third parties. Since such a prohibition was conditional and not absolute, as in the
case of narcotics or counterfeit currency, it was not sufficient to place trade in such products outside the
scope of the directive. Moreover, in so far as the trade in 500 counterfeit goods had a specific impact on
the lawful market in perfume products, the possibility of competition between counterfeit and lawfully
produced goods could not be ruled out and, accordingly, such goods could not, like narcotics or
counterfeit currency, be regarded as extra commercium. It followed that, on a proper construction of art 2
of the directive, value added tax was payable on counterfeit perfumes (see p 508 e j to p 509 a c to e h,
post).
Lange v Finanzamt Frstenfeldbruck Case C-111/92 [1993] ECR I-4677 and Witzemann v
Hauptzollamt Mnchen Case C-343/89 [1990] ECR I-4477 applied.

Notes
For the supply of goods in the context of Community VAT provisions, see 52 Halsburys Laws (4th edn)
para 2016.

Cases cited
Einberger v Hauptzollamt Freiburg Case 240/81 [1982] ECR 3699.
Einberger v Hauptzollamt Freiburg Case 294/82 [1984] ECR 1177.
Fischer v Finanzamt Donaueschingen Case C-283/95 (opinion) (1997) Transcript, 20 March, ECJ.
Horvath v Hauptzollamt Hamburg-Jonas Case 50/80 [1981] ECR 385.
Lange v Finanzamt Frstenfeldbruck Case C-111/92 [1993] ECR I-4677.
Mol v Inspecteur der Invoerrechten en Accijnzen Case 269/86 [1988] ECR 3627.
Vereiniging Happy Family Rustenburgerstraat v Inspecteur der Omzetbelasting Case 289/86 [1988] ECR
3655.
Witzemann v Hauptzollamt Mnchen Case C-343/89 [1990] ECR I-4477.
Wolf v Hauptzollamt Dsseldorf Case 221/81 [1982] ECR 3681.

Reference
By judgment of 24 December 1996, the English Court of Appeal referred to the Court of Justice of the
European Communities for a preliminary ruling under art 177 of the EC Treaty a question (set out at p 508
b, post) on the interpretation of Council Directive (EEC) 77/388 on the harmonisation of the laws of the
member states relating to turnover taxescommon system of value added tax: uniform basis of
assessment. That question arose in criminal proceedings against Mr Goodwin and Mr Unstead, who were
charged with the fraudulent evasion of VAT in respect of sales of counterfeit perfume products. Written
observations were submitted on behalf of: Mr Goodwin and Mr Unstead, by A Newman QC and P Guest,
Barrister, instructed by A Oxford, Solicitor, acting for Mr Unstead; the UK government, by J E Collins,
Assistant Treasury Solicitor, acting as agent, with S Richards and M Hoskins, Barristers; the Greek
government, by F Georgakopoulos, Legal Adviser to the State Legal Service, acting as agent, and A
Rokofyllou, Adviser to the Deputy Minister for Foreign Affairs, acting as agents; and the European
Commission, by H Michard and B Doherty, of its Legal Service, acting as agents. Oral observations were
made by: Mr Goodwin and Mr Unstead, represented by A Newman QC and P Guest; the UK government,
represented by J E Collins, with K Parker QC and M Hoskins; the Greek government, represented by F
Georgakopoulos and A Rokofyllou; and the Commission, represented by B Doherty. The language of the
case was English. The facts are set out in the opinion of the Advocate General.

501
12 March 1998.

The Advocate General (P Lger)

delivered the following opinion (translated from the French).
1.
Does the supply of counterfeit perfume products constitute a taxable transaction for value added tax
(VAT) purposes? By this question, the English Court of Appeal asks the Court of Justice of the European
Communities to clarify its case law under which certain unlawful transactions are held to fall outside the
scope of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating
to turnover taxescommon system of value added tax: uniform basis of assessment (OJ 1977 L145 p 1)
(the Sixth Directive).
 Facts
2.
Mr Goodwin and Mr Unstead were both convicted of the offence of being knowingly concerned in the
taking of steps with a view to the fraudulent evasion of VAT, contrary to s 72(1) of the Value Added Tax Act
1994. Mr Goodwin was accused of having purchased counterfeit perfume products and of having sold
them when he was not registered for VAT, ensuring thereby that he did not have to pay any VAT. Mr
Unstead was accused of having participated in the manufacture, production, distribution and sale of such
products through a business organisation which he ran together with other persons, and which was not
registered for VAT and therefore did not pay VAT.
3.
Mr Goodwin and Mr Unstead appealed against their convictions to the Court of Appeal. Without
denying that they had carried out the transactions as charged, they submitted that, on a proper
construction of the Sixth Directive, supplies of counterfeit perfume products are not subject to VAT.
4.
While agreeing with the Inner London Crown Court that, on the facts, VAT was payable, the Court of
Appeal considered that it would be helpful to obtain a preliminary ruling on this point. It therefore referred
the following question to the Court of Justice:
Does the supply of counterfeit perfume products fall within the scope of Council Directive
77/388/EEC of 17 May 1977 on the harmonization of the laws of the Member States relating to
turnover taxes (the Sixth Directive)?

Relevant legislation and case law

5.
The scope of the Sixth Directive is delimited by art 2 thereof, which provides:
The following shall be subject to value added tax: (1) The supply of goods or services effected
for consideration within the territory of the country by a taxable person acting as such; (2) The
importation of goods.
6.
In keeping with the general terms in which that provision is framed, the court considers the nature of
transactions to be irrelevant for the purposes of VAT liability, since the principle of fiscal neutrality
[precludes], as far as the levying of value added tax is concerned, a generalised differentiation between
lawful and unlawful transactions (see the judgments in Mol v Inspecteur der Invoerrechten en Accijnzen
Case 269/86 [1988] ECR 3627 (para 18), Vereiniging Happy Family Rustenburgerstraat v Inspecteur der
Omzetbelasting Case 289/86 [1988] ECR 3655 (para 20) and Lange v Finanzamt Frstenfeldbruck Case
C-111/92 [1993] ECR I-4677 (para 16)).
7.
However, the court has recognised certain exceptions to that principle.
502
8.
By transposing its case law on customs duties3 to cases concerning the levying of VAT on unlawful
imports and supplies, the court has on a number of occasions taken the view that certain products
3
In Horvath v Hauptzollamt Hamburg-Jonas Case 50/80 [1981] ECR 385, Wolf v Hauptzollamt Dsseldorf Case 221/81
[1982] ECR 3681 and Einberger v Hauptzollamt Freiburg Case 240/81 [1982] ECR 3699 (Einberger No 1), the court
took the view that no customs debt arises upon the importation of drugs such as morphine, heroin and cocaine, which
do not pass along the economic channels strictly controlled by the competent authorities for use for medical and
scientific purposes.

which have special characteristics inasmuch as, because of their very nature, they are subject
to a total prohibition on their being put into circulation in all the Member States, with the exception
of strictly controlled economic channels for use for medical and scientific purposes (See the
judgments in Mol [1988] ECR 3627 (para 18) and Happy Family [1988] ECR 3655 (para 20).)
or, inasmuch as they may not be marketed or incorporated into economic channels 4, are wholly alien to
the provisions of the Sixth Directive on the definition of the basis of assessment and, in consequence, to
the provisions on the origination of a turnover tax debt 5.
4
See the judgment in Lange [1993] ECR I-4677 (para 12)
5
See the judgments in Happy Family [1988] ECR 3655 (para 17), Mol [1988] ECR 3627 (para 15) and Witzemann v
Hauptzollamt Mnchen Case C-343/89 [1990] ECR I-4477 (para 19). See also the judgments in Einberger v
Hauptzollamt Freiburg Case 294/82 [1984] ECR 1177 (para 20) (Einberger No 2) and Lange [1993] ECR I-4677 (para
12).

9.
Two kinds of product have already been recognised as possessing such special characteristics. The
first is narcotics, such as amphetamines or hashish, whose supply or importation independently of an
economic channel strictly controlled by the competent authorities for use for medical and scientific
purposes can give rise only to penalties under the criminal law (see the judgments in Mol [1988] ECR
3627 (para 15) and Happy Family [1988] ECR 3655 (para 18)). Similarly, and a fortiori, VAT is not payable
on imports of counterfeit currency, there being no circumstances in which such imports could be
authorised (see the judgment in Witzemann v Hauptzollamt Mnchen Case C-343/89 [1990] ECR I-4477
(para 20)).
10.
In the case of unlawful goods of that kind, an exception to the principle of fiscal neutrality is justified,
since because of [their] special characteristics, all competition between a lawful economic sector and an
unlawful sector is precluded (see the judgment in Lange [1993] ECR I-4677 (para 16))6.
6
See also the judgments in Mol [1988] ECR 3627 (para 18) and Happy Family [1988] ECR 3655 (para 20).

11.
On the other hand, when that case law was relied on in respect of computer systems, where there is
no absolute prohibition based on the nature of the goods or their special characteristics, the court held
that the mere fact that export of those goods to certain destinations was prohibited, because of their
potential use for strategic purposes, [could] not be sufficient to remove those products from the scope of
the Sixth Directive (see [1993] ECR I-4677 (para 17)).
12.
Finally, I would refer to Fischer v Finanzamt Donaueschingen Case C-283/95, currently pending before
the court, in which Advocate General Jacobs proposes in his opinion that the court should conclude that
unlawful roulette transactions such as those in issue fall within the scope of VAT [and] are therefore
taxable 503 under Article 2(1) of the Directive (see (1997) ECJ Transcript, 20 March (para 20)).

The reply to the question

13.
Both the Court of Appeal and all parties to the proceedings take as their point of departure the case
law cited above. The various observations submitted diverge, however, on the question of its application.
Whereas the appellants infer from that case law that the Sixth Directive does not apply to their situation,
the European Commission and the Greek and UK governments maintain the contrary.
14.
I agree with the latter view, since to my mind the courts case law concerning narcotics and counterfeit
currency cannot be transposed to counterfeit products such as the perfumes at issue.
15.
First of all, I would emphasise the fundamental importance of the principle of fiscal neutrality which
underlies the Sixth Directive, by virtue of which VAT is payable without exception on all supplies of goods
and services effected for consideration within the territory of a member state by a taxable person, likewise
on all imports of goods. For VAT purposes, therefore, unlawful trade and lawful trade must be treated in
the same way, so that unlawful traders will not be given an advantage over their law-abiding competitors.
16.
In the present case, in conformity with the principle of fiscal neutrality, and thus notwithstanding the
reprehensible nature of the counterfeiting involved, the products in question ought to rank among those in
whose case transactions involving them give rise to the payment of VAT. The information before the court
does not disclose whether the counterfeit perfumes marketed by the appellants are manufactured within
the Community or imported from a non-member country. In either case, however, the transactions in
questionwhether involving a supply of goods or services effected for consideration within the territory of
the country by a taxable person acting as such or an importation of goodsrank among those specified
in art 2(1) or (2) of the Sixth Directive.
17.
As the Greek and UK governments emphasise, according to established case law, all exceptions to
that principle must be narrowly construed and confined exclusively to marginal cases where the nature of
the products concerned precludes any possibility of competition between a lawful sector and an unlawful
sector. According to the test adopted by the court, the goods in question must be goods which, by reason
of their special characteristics, may not be marketed or incorporated into economic channels.
18.
The argument put forward by the appellants, whereby they seek to have the transactions at issue
brought within that exceptional category, on the ground that they undermine the functioning of the
common market and can therefore only be alien to the Sixth Directive, first of all creates a sense of
unease which is hard to dispel. In flagrant disregard of the principle nemo auditur turpitudinem propriam
allegans, the appellants are seeking to rely on the unhealthy, and even dangerous, nature, from an
economic point of view, of their activities, in order to prove that they are not liable to pay VAT.
19.
More importantly, however, their argument fails if the tests already adopted in the case law are applied
to it in the circumstances of this case.
20.
In the first place, unlike narcotics or counterfeit currency, counterfeit products are not goods whose
marketing is prohibited because of their very nature, or their special characteristics.
504
21.
The transactions to which narcotics or counterfeit currency may give rise are intrinsically illegal,
because such products are systematically banned in all the member states by reason of their special
characteristics, namely their disastrous effects, in the case of narcotics, on the human body and, in the
case of counterfeit currency, on the economy of a country. Their prohibition is designed to protect what
may be termed the general interest. Moreover, the court has emphasised the fact that these two types of
product are universally regarded as of a reprehensible nature by referring to the international conventions
for their suppression7.
7
The relevant conventions are the Single Convention on Narcotic Drugs (New York, 30 March 1961; TS 23 (1979); Cmnd
7466), the Convention on Psychotropic Substances (Vienna, 21 February 1929; Misc 24 (1978); Cmnd 7330) and the
International Convention for the Suppression of Counterfeiting Currency (Geneva, 20 April 1929; TS 5 (1960); Cmd 932)
(see, respectively, the judgments in Happy Family [1988] ECR 3655 (para 25), Mol [1988] ECR 3627 (para 24) and
Witzemann [1990] ECR I-4477 (para 14)).

22.
Counterfeit products such as the perfumes at issue are not, as such, banned in the Community by
reason of their intrinsic nature. Obviously there is a lawful trade in perfumes, as indeed in any product
which can be counterfeited. While the marketing of counterfeit products may be banned, the object of this
is not the direct protection of a general interest; rather, the aim is either to prevent consumers from being
cheated as a result of purchasing goods other than those covered by the guarantees attaching to the
mark that is infringed, or, above all, to protect the individual interests of the holder of a right. Although
transactions involving such products infringe rules such as those governing intellectual property, the
ensuing prohibition is not linked to the nature or essential characteristics of the products.
23.
Counterfeiting is admittedly a plague which causes serious damage to lawful commercial activities 8,
and against which the Community has sought to guard itself, in the specific field of goods coming from or
destined for non-member countries, through the adoption of two successive regulations 9. Even today,
however, a finding that counterfeiting is illegal can only come about at Community level if the holder of the
flouted right takes action.
8
For comment, supported by statistics, see Gourdin-lamblin La lutte contre la contrefaon en droit communautaire
(1996) Revue du march commun et de lUnion europenne No 394.
9
Council Regulation (EEC) 3842/86 laying down measures to prohibit the release for free circulation of counterfeit goods
(OJ 1986 L357 p 1), replaced and amended by Council Regulation (EC) 3295/94 laying down measures to prohibit the
release for free circulation, export, re-export or entry for a suspensive procedure of counterfeit and pirated goods (OJ
1994 L341 p 8).

24.
The Community procedure which has been established opens as a rule with the lodging of an
application by the holder of the right which has been infringed (see art 3 of Council Regulation (EC)
3295/94 laying down measures to prohibit the release for free circulation, export, re-export or entry for a
suspensive procedure of counterfeit and pirated goods (OJ 1994 L341 p 8)). Regulation 3295/94 also
provides for the customs authorities to initiate a proceeding of their own motion by notifying the holder of
a right of a suspected risk of infringement and withholding the goods to enable him to lodge an application
for suspension of release (art 4). The fact remains that, even in such cases, continued suspension of
release depends on the holder of the right having applied to the competent authority for a decision on the
merits.
25.
Similarly, in cases involving not imported products, but goods manufactured in the Community (the
position in the present case cannot be 505 determined from the order for reference), the sale of those
goods must be presumed to be lawful unless and until the owner of the trade mark or a related right
shows that his exclusive rights have been infringed10.
10
In such cases the rules governing trade marks would apply, and in particular art 5 of Council Directive (EEC) 89/104 to
approximate the laws of the member states relating to trade marks (OJ 1989 L40 p 1), which provides that the owner of
a trade mark is entitled to prevent all third parties from using his trade mark or any sign where, because of its similarity
to that trade mark, there exists a likelihood of confusion.

26.
Thus, counterfeit goods, such as the perfumes at issue, are not unlawful at Community level by reason
of their inherent characteristics. They can only be prohibited if it is proved that they infringe a valid
property right. As the Commission has pointed out, the marketing of such products could at most be
subject to a conditional prohibition but not, as in the case of narcotics or counterfeit currency, wholly
banned. By contrast with those cases, the circumstances here are similar to those in Lange, where the
goods were not prohibited by reason of their very nature, and their exportation was banned only in
respect of certain countries where they could be used for strategic purposes. This is also the line of
reasoning which Advocate General Jacobs, in his opinion in Fischer, proposes that the court should follow
with regard to the unlawful organisation of roulette games.
27.
Since counterfeit products cannot, by reason of special characteristics, be ranked among goods which
may not be marketed or incorporated into economic channels, the possibility of competition with goods
which are traded lawfully can by no means be ruled out.
28.
As the UK government and the Commission rightly point out, unlike counterfeit currency or narcotics
(in respect of which either there is no lawful market or there is trading, but subject to such strict controls
that it can never compete with lawful trade), there is obviously a lawful market in perfume products which
forms an integral part of the Community economy. However, even though trade in counterfeit products is
prohibitedbecause of the harm it causes to the lawful business activities which are liable to be
supplanted by it, and in order to protect consumers 11and however much its existence may be deplored,
the fact that it still goes on cannot be denied. Refusal to take account of this (particularly from the tax
aspect, which is the one in point here) on the ground that, if such goods are banned, they are extra
commercium and cannot therefore be subject to VAT would confer on unlawful traders an unfair
competitive advantage over their legitimate rivals who, alone, would have to pay VAT.
11
See, to that effect, the second recital in the preamble to Regulation 3295/94.

29.
Moreover, the present case clearly illustrates the dangers of widening the courts case law concerning
exceptions to the principle of fiscal neutrality. It may very well be that the appellants calculated that,
notwithstanding the risk of prosecution entailed, trading in counterfeit perfume products would
nevertheless be worthwhile on the basis that VAT did not apply to those products. By the same token,
there can be no question of encouraging transactions of that type by guaranteeing them the benefit of an
unjustified tax exemption.
30.
Furthermore, as the UK government and the Commission emphasise 12, if levying of VAT were to
depend on whether or not a product is counterfeit, the whole integrity of the internal market would be
undermined. Given the lack of 506 harmonisation in this field, the catalogue of counterfeit goods varies
from one member state to another. Consequently, if the same tax treatment is to be accorded to all
transactions within the Community, the question whether VAT is payable on goods must be kept entirely
separate from the question whether or not those goods are counterfeit.
12
See also Advocate General Jacobs opinion in Fischer (1997) ECJ Transcript, 20 October (para 26).

31.
One last point. In all the cases cited above, the court has taken care to make it clear that the question
whether a transaction is subject to VAT is independent of the operation of other rules of domestic law,
particularly in the field of criminal law. In the present case, it is clear that this continues to hold true and
that the courts judgment will be wholly without prejudice to the powers of Member States to impose
appropriate penalties, including those with financial consequences, for contraventions of their legislation
(see, most recently, the judgment in Lange [1993] ECR I-4677 (para 24)).
32.
Accordingly, I conclude that transactions involving counterfeit products, such as the perfume products
at issue, fall within the scope of VAT. They are therefore subject to taxation under art 2 of the Sixth
Directive.

Conclusion
33.
In view of the foregoing considerations, I propose that the Court of Justice reply as follows to the
question referred by the Court of Appeal:
Article 2 of the Council Directive (EEC) 77/388 on the harmonisation of the laws of the member
states relating to turnover taxescommon system of value added tax: uniform basis of
assessment, must be interpreted as making the supply of counterfeit products, such as perfumes,
subject to VAT.

28 May 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By judgment of 24 December 1996, received at the Court of Justice of the European Communities on
9 January 1997, the English Court of Appeal referred to the Court of Justice for a preliminary ruling under
art 177 of the EC Treaty a question on the interpretation of Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment (OJ 1977 L145 p 1) (the Sixth Directive).
2.
That question was raised in criminal proceedings against Mr Goodwin and Mr Unstead, who were
charged with the fraudulent evasion of VAT in respect of sales of counterfeit perfume products.
3.
According to the documents before the court, Mr Goodwin was accused of having purchased
counterfeit perfume products and of having sold them without being registered for VAT. Mr Unstead was
accused of having participated in the manufacture, production, distribution and sale of counterfeit perfume
products through a business organisation which he ran together with other persons, and which was not
registered for VAT.
4.
At first instance, the case came before the Inner London Crown Court, which held that the Sixth
Directive did not preclude the charging of VAT on the manufacture, production, distribution and sale of
counterfeit perfume products, and found both defendants guilty of conduct contrary to s 72(1) and (8) of
the Value Added Tax Act 1994.
507
5.
Mr Goodwin and Mr Unstead have appealed against that decision to the Court of Appeal, arguing inter
alia that Community law precludes the levying of VAT in a situation such as theirs.
6.
The Court of Appeal believes that, as a matter of Community law, VAT is payable on supplies of
counterfeit perfume products for consideration. However, being somewhat doubtful on the point, it stayed
proceedings in order to refer the following question to the Court of Justice for a preliminary ruling:
Does the supply of counterfeit perfume products fall within the scope of Council Directive
77/388/EEC of 17 May 1977 on the harmonisation of the laws of the Member States relating to
turnover taxes (the Sixth Directive)?
7.
The crux of the question is whether, on a proper construction of art 2 of the Sixth Directive, VAT is
payable on the supply of counterfeit perfumes.
8.
Article 2 of the Sixth Directive provides:
The following shall be subject to value added tax: (1) The supply of goods or services effected
for consideration within the territory of the country by a taxable person acting as such; (2) The
importation of goods.
9.
It should be borne in mind at the outset that, according to established case law, the Sixth Directive,
whose purpose is to achieve widespread harmonisation in the area of VAT, is based on the principle of
fiscal neutrality. As regards the levying of VAT, that principle precludes a generalised differentiation
between lawful and unlawful transactions, except where, because of the special characteristics of certain
products, all competition between a lawful economic sector and an unlawful sector is precluded (see, inter
alia, Lange v Finanzamt Frstenfeldbruck Case C-111/92 [1993] ECR I-4677 (para 16)).
10.
Referring to a number of judgmentsEinberger v Hauptzollamt Freiburg Case 294/82 [1984] ECR
1177, Mol v Inspecteur der Invoerrechten en Accijnzen Case 269/86 [1988] ECR 3627, Vereiniging Happy
Family Rustenburgerstraat v Inspecteur der Omzetbelasting Case 289/86 [1988] ECR 3655 and
Witzemann v Hauptzollamt Mnchen Case C-343/89 [1990] ECR I-4477Mr Goodwin and Mr Unstead
submit that, since there is no lawful market in counterfeit perfumes, the present case falls within the scope
of that exception. In the United Kingdom, not only would a contract for the sale of counterfeit perfumes be
void for illegality, but such sales would also infringe a wide variety of intellectual property rights.
Furthermore, the placing of such products on the market seriously undermines the functioning of the
common market inasmuch as, unlike narcotics, trade in counterfeit perfumes is never permitted.
11.
In Einberger, Mol and Happy Family, the court ruled that no turnover tax arises upon the unlawful
importation into the Community of drugs or upon the unlawful supply of similar products effected for
consideration within a member state, in so far as the products in question are not confined within
economic channels strictly controlled by the competent authorities for use for medical and scientific
purposes. In the judgment in Witzemann [1990] ECR I-4477 (para 20), the court held that its reasoning in
relation to the illegal importation of drugs applies a fortiori to imports of counterfeit currency.
12.
In those four judgments, the court added that unlawful imports or supplies of goods such as those at
issue in those cases, release of which into the economic and commercial channels of the Community is
by definition absolutely precluded and which can give rise only to penalties under the criminal law, are
wholly alien to the provisions of the Sixth Directive (see the judgments in Einberger [1984] ECR 5081177
(paras 1920), Mol [1988] ECR 3627 (para 15), Happy Family [1988] ECR 3655 (para 17) and
Witzemann [1990] ECR I-4477 (para 19)). That line of case law thus concerns goods which, because of
their special characteristics, may not be placed on the market or incorporated into economic channels.
13.
However, that is not the position here. As the Greek government, the UK government and the
European Commission emphasised, the goods at issue in the main proceedings are not goods which
cannot be placed on the market because of their intrinsic nature or special characteristics.
14.
As Advocate General Lger pointed out in para 22 of his opinion, above, although transactions
involving counterfeit products infringe intellectual property rights, any consequential prohibition is not
linked to the nature or essential characteristics of such products, but to their detrimental impact on the
rights of third parties. Similarly, as the Commission noted in its observations, the prohibition on counterfeit
products, which stems from the fact that they infringe intellectual property rights, is conditional, not
absolute as in the case of narcotics or counterfeit currency. That prohibition is not sufficient, therefore, to
place trade in such products outside the scope of the Sixth Directive.
15.
 Furthermoreas the Commission also pointed outthe possibility of competition between counterfeit
products and goods which are lawfully traded cannot be ruled out in a case such as that before the
national court, in so far as there is a lawful market in perfume products on which counterfeit goods have a
specific impact. Accordingly, such goods cannot, like narcotics or counterfeit currency, be regarded as
extra commercium.
16.
It must therefore be stated in reply to the question referred that, on a proper construction of art 2 of the
Sixth Directive, VAT is payable on the supply of counterfeit perfumes.

Costs
17.
The costs incurred by the Greek and UK governments and the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main action, a step in the proceedings pending before the national court, the decision on
costs is a matter for that court.
On those grounds, the Court of Justice (First Chamber), in answer to the question referred to it by the
Court of Appeal by judgment of 24 December 1996, hereby rules: on a proper construction of art 2 of
Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover
taxescommon system of value added tax: uniform basis of assessment, VAT is payable on the supply of
counterfeit perfumes.

509

[1998] All ER (EC) 510

Jensen and another v Landbrugsministeriet, EF-Direktoratet

(Case C-132/95)
EUROPEAN COMMUNITY; Agriculture, Taxation: TAXATION; VAT and Customs and Excise
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, WATHELET (PRESIDENTS OF
CHAMBERS), MANCINI (RAPPORTEUR), MOITINHO DE ALMEIDA, MURRAY, PUISSOCHET, HIRSCH
AND SEVN ADVOCATE GENERAL FENNELLY
1 OCTOBER, 27 NOVEMBER 1997, 19 MAY 1998.
European Community Set-off Community aid Tax debt Member state setting off amounts payable
under common agricultural policy to farmers or their assignees against farmers value added tax debts
Whether such measures compatible with Community law Council Regulation (EEC) 1765/92.

The plaintiffs were a Danish farmer, who was entitled to a compensatory payment under Council
Regulation (EEC) 1756/92 establishing a support system for producers of certain arable crops, and KFK,
the assignee of a payment due to another Danish farmer under the same regulation. In the first case, the
farmer owed the state an amount in respect of value added tax which exceeded that of the payment to
which he was entitled under the regulation, with the result that the European Affairs Directorate of the
Danish Ministry of Agriculture set off the whole amount of the aid against that debt. In the second case the
farmer had assigned the annual amount of aid he could claim under the regulation to KFK. The document
recording the assignment had been notified to the directorate, which took note of it subject to the states
right of set off. Since the farmer had become indebted to the state before the date of the assignment, and
since his debt fell due before the aid could be paid, the directorate effected set off to cover the states
claim against the beneficiary of the aid and informed KFK that, as a result, it would not receive any
payment. The farmer in the first case and KFK both brought proceedings before the stre Landsret,
seeking payment of the amounts of aid to which they were entitled. The stre Landsret stayed the
proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling,
inter alia, the question whether Community law precluded a member state from setting off an amount due
to a beneficiary of aid payable under Community legislation against outstanding debts to that state.

Held Community law did not preclude a member state from setting off an amount due to a beneficiary of
aid payable under Community legislation and an outstanding debt to that member state. The position
would be different only if that practice interfered with the proper functioning of the common organisation of
the agricultural markets. In that respect, the capacity in which the member state granted aid under the
regulation, the fact that the rules of that member state on set-off required for set-off to be aviailable,
reciprocity of debts as between debtor and creditor, the practice generally followed by the member state
regarding set-off and the legal basis of the debt to the state involved in the set-off were of no importance,
provided that the national authorities ensured that the effectiveness of Community law was not in any way
undermined and that 510 economic operators enjoyed equal treatment; and it was for the national court to
determine whether that was the case (see p 541 j to p 542 b and p 544 a to c, post).

Cases cited
Albert Ruckdeschel & Co v Hauptzollamt Hamburg-St Annen, Diamalt AG v Hauptzollamt Itzehoe Joined
cases 117/76 and 16/77 [1977] ECR 1753.
Amministrazione delle Finanze dello Stato v San Giorgio Case 199/82 [1983] ECR 3595.
Association comit conomique agricole rgional fruits et lgumes de Bretagne (Cerafel) v Le Campion
Case 218/85 [1986] ECR 3513.
Balkon-Import-Export GmbH v Hauptzollamt Berling-Packhof Case 118/76 [1977] ECR 1177.
BayWa AG v Bundesanstalt fr Landwirtschaftliche Marktordnung Joined cases 146/81, 192193/81
[1982] ECR 1503.
Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043.
Continental Irish Meat Ltd v Minister for Agriculture Case 125/84 [1985] ECR 3441.
De Samvirkende Danske Landboforeninger v Ministry of Fiscal Affairs Case 297/82 [1983] ECR 3299.
DEKA Getreideprodukte GmbH & Co KG (in liq) v EEC Case 250/78 [1983] ECR 421.
Deutsche Milchkontor GmbH v Germany Joined cases 205-215/82 [1983] ECR 2633.
EC Commission v Germany Case 48/85 [1986] ECR 2549.
Eridania-Zuccherifici nazionali (SpA) v Minister of Agriculture and Forestry Case 230/78 [1979] ECR
2749.
Eurico Srl v EC Commission Case 109/83 [1984] ECR 3581.
Finsider v EC Commission Case 250/83 [1985] ECR 131.
France v EC Commission Case C-366/88 [1990] ECR I-3571.
Futura Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471.
Garage Molenheide BVBA v Belgium Joined cases C-286/94, C-340/95, C-401/95 and C-47/96 [1998] All
ER (EC) 61, ECJ.
Granaria BV v Hoofdproduktschap voor Akkerbouwprodukten Case 101/78 [1979] ECR 623.
Interagra (Compagnie) (SA) v EC Commission Case 217/81 [1982] ECR 2233.
Irish Creamery Milk Suppliers Association v Ireland Joined cases 36/80 and 71/80 [1981] ECR 735.
Italy v EC Commission Case C-34/89 [1990] ECR I-3603.
Peter v Hauptollamt Regensburg Case C-290/91 [1993] ECR I-2981.
Reichelt (Otto) GmbH v Hauptzollamt Berlin-Sd Case 113/81 [1982] ECR 1957.
Rewe-Zentral AG v Bundesmonopolverwaltung fr Branntwein Case 120/78 [1979] ECR 649.
Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989.
Russo v Azienda di Stato per gli Interventi Sud Mercato Agricolo (AIMA) Case 60/75 [1976] ECR 45.
Sucrimex SA v EC Commission Case 133/79 [1980] ECR 1299.
Unione Nazionali di Produktori di Olive (Unaprol) v Azienda di Stato per gli Interventi Sud Mercato
Agricolo (AIMA) Case C-186/93 [1994] ECR I-3615.
Wilhelm Fromme (Firma) v Bundesanstalt fr Landwirtschaftliche Marktordnung Case 54/81 [1982] ECR
1449.
511

Reference
By order of 10 April 1995, the stre Landsret (the Eastern Regional Court) referred to the Court of Justice
of the European Communities for a preliminary ruling under art 177 of the EC Treaty a number of
questions (set out at p 538 h to p 539 d, post) on the interpretation of Community law with regard to set-
off of amounts paid under Community law against debts payable to a member state and of arts 10(1) and
15(3) of Council Regulation (EEC) 1765/92 establishing a support system for producers of certain arable
crops. Those questions were raised in proceedings brought by Mr Jensen, a farmer entitled to a
compensatory payment under the regulation, and Korn- og Foderstofkompagniet A/S (KFK), the assignee
of another compensatory payment due under the same regulation, against Landbrugsministeriet, EF-
Direktoratet (the Directorate of European Affairs of the Ministry of Agriculture), concerning set-off by the
latter between the compensatory payments in question and outstanding debts payable to the state.
Written observations were submitted on behalf of: Mr Jensen, by A Philip of the Copenhagen Bar;
Landbrugsministeriet, EF-Direktoratet, by K Hagel- Srensen, of the Copenhagen Bar; the Danish
government, by P Biering, Head of Division in the Ministry of Foreign Affairs, acting as agent; the Irish
government, by M A Buckley, Chief State Solicitor, acting as agent, and E R Alkin, Barrister-at-law; the
Finnish government, by H Rotkirch, Ambassador, Head of the Legal Service in the Ministry of Foreign
Affairs, acting as agent; the Swedish government, by E Brattgrd, Departementsrd, acting as agent; the
UK government, by J E Collins, Assistant Treasury Solicitor, acting as agent, and K Parker QC; and the
European Commission, by H P Hartvig and T van Rijn, Legal Advisers, acting as agents. Oral
observations were made by Mr Jensen, represented by A Philip; Korn- og Foderstofkompagniet A/S,
represented by J Sberg, of the Silkeborg Bar; of Landbrugsministeriet; EF-Direktoratet, represented by K
Hagel-Srensen; the Danish government, represented by J Molde, Head of Division in the Ministry of
Foreign Affairs, acting as agent; the Greek government, represented by I Chalkias, Deputy Legal Adviser
to the State Legal Council, and E Mamouna, lawyer in the Special Department for Community
Proceedings of the Ministry of Foreign Affairs, acting as agents; the French government, represented by F
Pascal, Attach dAdministration Centrale in the Legal Directorate of the Ministry of Foreign Affairs, acting
as agent; the Irish government, represented by D Moloney, Barrister-at-law; the Finnish government,
represented by T Pynn, Legal Adviser in the Ministry of Foreign Affairs, acting as agent; and the
Commission, represented by H P Hartvig and T van Rijn. The language of the case was Danish. The facts
are set out in the opinion of the Advocate General.

27 November 1997.

The Advocate General (N Fennelly)

delivered the following opinion.

IINTRODUCTION
1.
The issue in this case is whether member states can apply national rules of set-off between fiscal
debts owed under national law and sums due to the same persons pursuant to Community law. The
applicable Community legislation, part of the McSharry agricultural reforms, requires that certain
agricultural aids be paid to their intended recipients in their entirety, by a specified date.
512

IILEGAL AND FACTUAL CONTEXT

(i) Community law

2.
The agricultural reforms in question replaced indirect subsidies by means of price supports with a
system approximating Community and world market prices, and with payment of compensating aids
directly to producers. They were initiated by Council Regulation (EEC) 3766/91 establishing a support
system for producers of soya beans, rape seed and colza seed and sunflower seed (OJ 1991 L356 p 17)
and Commission Regulation (EEC) 615/92 laying down detailed rules for a support system for producers
of soya beans, rape seed, colza seed and sunflower seed (OJ 1992 L67 p 11), and extended by Council
Regulation (EEC) 1765/92 establishing a support system for producers of certain arable crops (OJ 1992
L181 p 12) (the regulation). The second recital in the preamble to the regulation states that
in order to ensure better market balances a new support system has to be established to
compensate the loss of income caused by the reduction of the institutional prices by a
compensatory payment for producers who sow such products.
The recital adds that the area eligible should be restricted to the area down to arable crops or publicly
funded set aside in the past.
3.
Compensatory payments are based on the amount of land either under cultivation or set aside (arts
2(2) and 7). Applicants for aid are subject to an obligation to set aside a specified proportion of the land in
their holding, in return for compensation (arts 2(5) and 7(1)) 1. The amount of the compensatory payment
per hectare is calculated on a regional basis, by multiplying a specified basic monetary sum per tonne by
the average cereals yield from 1986/87 to 1990/91 determined for the region concerned (art 4(1) and (2)).
This amount, and the related compensation for set-aside obligations, may be changed in the light of
developments in production, productivity and the markets (art 15(1)). Where the sum of the individual
areas for which aid is claimed is greater than the regional base area in which arable crops were grown or
which were subject to publicly funded set-aside in the reference period 1989/91, the area eligible for aid is
proportionately reduced or, in subsequent years, proportionate increases in uncompensated set-aside
shall be imposed (art 2(6)). As an alternative to using a regional base area, member states have the
option of awarding aids on the basis of individual base areas, based on the average number of hectares
in arable cultivation or set-aside during the reference period in each holding in its territory, in which case
individual producers can apply for aid only in respect of the area attributed to them (art 2(3)).
1
Article 8(1) provides that small producers may apply under a simplified scheme which does not include a set-aside
obligation.

4.
Article 13 of the regulation states that the measures provided for shall be deemed to be interventions
intended to stabilise the agricultural markets within the meaning of art 3(1) of Council Regulation (EEC)
729/70 on the financing of the common agricultural policy (OJ S edn 1970 (I) p 218). Article 10(1)
provides that the compensatory payments shall be made between 16 October and 31 December following
the harvest. Article 15(3), which is central to the present case, states that [t]he payments referred to in
this Regulation are to be paid over to the beneficiaries in their entirety. Article 2(2) of Regulation 615/92
provided, similarly, that [t]he payments referred to in this Regulation shall be made to 513 producers
without any deductions except as otherwise provided in this Regulation.
5.
At the time of the events giving rise to the present action 2, art 4 of Regulation 729/70 provided for the
designation by member states of authorities and bodies, empowered to effect the expenditure provided
for, to whom the European Commission would make available the necessary credits. The second indent
of art 4(2) provided that [t]he member states shall ensure that those credits are used without delay and
solely for the purposes laid down. After the end of the 1987 financial year, the member states were
required to mobilise the financial resources to cover the necessary expenditure, in accordance with the
needs of their disbursing authorities, while the Commission made payments to cover expenditure already
effected by the member states3.
 2
Articles 4, 5 and 8 of Regulation 729/70 were amended by Council Regulation (EC) 1287/95, OJ 1995 L125 p 1, with
effect from 16 October 1995
3
Third sub-paragraph of arts 4(2) and 5(2)(a) of Regulation 729/70, as inserted and amended, respectively, by Council
Regulation (EEC) 3183/87 introducing special rules for the financing of the common agricultural policy, OJ 1987 L304 p
1, and further amended by Council Regulation (EEC) 2048/88, OJ 1988 L185 p 1. Article 5(2)(a) formerly provided for
the Commission to make advances to member states before expenditure was effected.

Article 8 of Regulation 729/70 provided at the relevant time:

(1) The Member States in accordance with national provisions laid down by law, regulation or
administrative action shall take the measures necessary to: satisfy themselves that transactions
financed by the [European Agricultural Guidance and Guarantee] Fund are actually carried out and
are executed correctly; prevent and deal with irregularities; recover sums lost as a result of
irregularities or negligence.
(2) In the absence of total recovery, the financial consequences of irregularities or negligence
shall be borne by the Community, with the exception of the consequences of irregularities or
negligence attributable to administrative authorities or other bodies of the Member States.
The sums recovered shall be paid to the paying authorities or bodies and deducted by them
from the expenditure financed by the Fund.
6.
Article 14 of Commission Regulation (EEC) 3887/92 laying down detailed rules for applying the
integrated administration and control system for certain Community aid schemes (OJ 1992 L391 p 36) 4 is
concerned with reimbursement of wrongly paid aids. Article 14(2) provides as follows:
4
Commission Regulation (EEC) 3387/92 is applicable from 1 February 1993 and was adopted on the basis of Council
Regulation (EEC) 3508/92 establishing an integrated administration and control system for certain Community aid
schemes, OJ 1992 L355 p 1, which applies to, inter alia, the aid scheme established by the regulation.

However, Member States may decide that the amount owed [in the case of wrongful payment to
a farmer] should not be paid over but deducted from the first advance or first payment due to the
farmer concerned after the date on which the repayment decision was taken. No interest shall run
after the beneficiary has been informed of the wrong payment.

(ii) Danish law

7.
Under general principles of Danish law, public authorities are entitled to recover fiscal debts owed to
them by the beneficiaries of public aid in three ways. 514First, a public authority, like any creditor, may
seek the attachment of aid moneys owed to its debtor, which, if resisted by the debtor, requires an order
of a competent court. The normal rules of priority apply. The aid will be paid directly to the grantee of such
an attachment order.
8.
Secondly, the debtor may voluntarily assign his interest in the moneys owed to him or expected to be
owed to him by a public authority to his creditors, including public authorities. In the case of an
assignment to a number of creditors, the rules of priority apply. In the case of an assignment to a private
creditor, the assignee will then be entitled to claim the aid in question directly from the responsible public
authority.
9.
Finally, the method at issue in this case is that of set-off in respect of debts in a liquidated sum. The
respective debts and credits must be mutual, that is, the debtor in respect of one is the creditor in respect
of the other. The different organs of the Danish state are considered to constitute a single entity for these
purposes5. Thus, for example, fiscal debts owed to the revenue authorities can normally be set off against
moneys owed to the same person by other ministries.
5
This statement in the order for reference is disputed by Mr Jensen, on the basis of a distinction between the state and
regional and local authorities. It is for the national court to resolve any subsisting dispute regarding the interpretation of
the national rules.

10.
However, set-off is, in certain cases, precluded by social and general considerations 6. It is not the
practice to set off debts to private parties arising under the law of property (eg for the supply of goods and
services or under contracts) against other debts, such as fiscal debts 7. In addition, it appears that the
states practice with regard to economic subsidies has varied. In certain fields, set-off has not been
operated in the past as regards subventions for specific projects, on the grounds that the project might not
otherwise be completed. Law No 284 of 27 April 1994 provides for set-off in certain specified cases of
amounts not to exceed 20% of such subventions, although it appears that this is without prejudice to the
states general entitlement to set off the full amount. Subventions within the province of the Ministry of
Agriculture which are subject, in practice, to this limited form of set-off include those for the development
of agricultural or fishery products, whether in respect of projects in primary agriculture or those concerning
the processing of such products. It appears that set-off does not take place at all in respect of subventions
outside the scope of Law No 284.
6
It appears that set-off is not possible in respect of payments necessary to satisfy basic needs, such as salaries,
pensions and social aids.
7
See Circular No 186 of the Minister for Justice, of 22 November 1983.
11.
The assignment of a debt will not normally affect the right of set-off. Where a creditor assigns a debt to
a third party, the debtor may still claim set-off between that debt and moneys owed to him by the original
creditor save where the latter debt arose after he knew or could have known of the assignment. If, at the
time the debtor knew or ought to have known of the assignment, the assigned debt was not yet due, the
debtor can still proceed to set it off against the original creditors debt to him where the latter debt
becomes due, at the latest, on the same day as the assigned debt (see art 28 of law No 669 of 23
September 1986).

(iii) The facts

12.
It appears that, in the past, the Danish state had extensive recourse to set-off of agricultural aids
against fiscal debts owed by farmers and others. The EC Directorate of the Danish Ministry of Agriculture,
the national authority 515 responsible for granting compensatory payments, sought advice from the
Commissions Directorate-General for Agriculture in July 1992 regarding this practice, in the light of
Regulation 615/92. The latter advised in a letter of 12 November 1992 that, where national law permitted
such set-off, there was no objection to it under Community law so long as the procedure was applied
neutrally, without discrimination as between debts arising under Community and national law, and the
Community payments regime was not thereby rendered impossible. Articles 2(2) of Regulation 615/92
and 15(3) of the regulation, it said, required only that no deduction for administrative expenses or other
reasons could be made from the amounts owed to producers. However, on 7 October 1994, the
Directorate-General, upon further consideration of the issue by the Commissions Legal Service, advised
the Danish Ministry of Agriculture of the reversal of its earlier provisional view. The Commissions Legal
Service had advised that the set-off of other debts against moneys owed under the regulation was not
permissible, due to the requirement in art 15(3) that aids be paid in their entirety. Furthermore, the system
of set-off would undermine the efficacy of the regime, as member states would be enabled to recover
fiscal debts from producers without following normal recovery procedures, in a way which was not
possible under the previous price-support regime 8. The Danish state thereupon desisted from the
practice, without accepting the correctness of the Commissions view.
8
See the Memorandum of 27 April 1994 from the Commission Legal Service to the Commission Director General for
Agriculture.

13.
The present litigation involves one case of simple set-off and one of set-off after an assignment, both
of which occurred in 1993, before the change of policy on the part of the Commission and the Danish
state. The aids owed to Mr Jensen at the end of 1993 (D Kr 33,563) were set off against VAT debts which
were due and which were greater than the amount of the aid. It appears that the aid would otherwise have
been used to make part payment of all of Mr Jensens debts, including that to the fiscal authorities, in the
framework of an agreement with his creditors, and that the set-off resulted in a significant reduction in his
payments to the other creditors.
14.
The Korn- og Foderstofkompagniet NS (KFK), a farm supplies company, had, for its part, received an
assignment in spring 1993 from a farmer, Mr Stenholt, of aid moneys which would be owed to him at the
end of the season (D Kr 45,574), in order to pay for certain supplies. However, the Danish state was able
to set off that amount against Mr Stenholts debts to the state under a debt-conversion loan dating from
1984, when his total debts to both public and private creditors were reduced in a procedure akin to
bankruptcy.
15.
Both Mr Jensen and KFK initiated proceedings before the stre Landsret (the Eastern Regional Court)
for the payment to them by the EC Directorate of the Ministry of Agriculture of the aids in question. The
national court suspended proceedings and referred the following questions for a preliminary ruling
pursuant to art 177 of the EC Treaty:
(1) Does Community law in general preclude a Member State from setting off an amount due to
the beneficiary of aid under a Community measure against outstanding debts to a Member State?
(2)(a) Is it of any significance for the answer to Question 1 whether the amount of aid under
Community law is paid in advance by the Member State 516 which has a claim to be reimbursed
for the aid paid out only if the rules of Community law on payment are satisfied, and which must
itself defray the expenditure involved in the administration of the support system? (b) Is it of any
significance for the answer to Question 1 that under the Member States rules on set-off it is a
condition for effecting set-off that there be reciprocity between the debtor under the principal claim
and the creditor under the counterclaim? (c) Is it of any significance for the answer to Question 1
that the Member States practice with regard to certain trade and environmental subsidies is
established in such a way as to permit set-off in an amount not exceeding 20% of the said state
subsidies? (d) Is it of any significance for the answer to Question 1 what legal basis exists for the
outstanding debt to the state against which set-off is to be effected? An answer is desired in
particular to the question whether the Member States have a greater scope to effect set-off if all or
part of the sum to be set off constitutes part of the Communitys own income?
(3) If Questions 1 and 2(a) to 2(d) are answered to the effect that set-off is in general possible,
or possible subject to certain conditions, is Article 15(3) of Council Regulation No 1765/92 to be
 interpreted as meaning that a Member State is precluded from requiring a national intervention
agency to effect set-off in the case of a beneficiary of compensatory payments with outstanding
debts to the state which could otherwise be involved in set-off?
(4) Is Article 10(1) of Council Regulation No 1765/92 to be interpreted as meaning that the
compensatory amounts in question are to be paid over immediately the intervention agency has
concluded the procedure with regard to the beneficiarys application, or is it permissible to delay the
payment for an investigation as to whether the state has outstanding claims against the beneficiary
in respect of which it wishes to effect a set-off, provided always that the payment is effected at the
latest by 31 December of the relevant support year?

IIIOBSERVATIONS
16.
Written observations were submitted by Mr Jensen, the EC Directorate of the Danish Ministry of
Agriculture, the Commission, the Kingdom of Denmark, the Republic of Finland, Ireland, the Kingdom of
Sweden and the United Kingdom of Great Britain and Northern Ireland. Oral observations were submitted
by Mr Jensen, KFK, the ministry, the Commission, the Kingdom of Denmark, the Republic of Finland, the
French Republic, the Hellenic Republic and Ireland.
17.
Generally speaking, all of the member states which have submitted observations, as well as the
ministry, support the possibility of set-off of compensatory payments under the regulation against fiscal
debts, for largely similar reasons, although they do not all make submissions regarding all of the
questions referred by the national court. I will not, therefore, seek to distinguish the arguments attributable
to each in respect of the first and third questions. I will refer to them, simply, as the member states. While
the Commission does not wish to exclude in all circumstances the possibility of set-off by member state
public authorities of Community aids against national fiscal debts, it submits, as do Mr Jensen and KFK,
that this is impermissible in respect of aids granted pursuant to the regulation. It is, therefore, convenient
to summarise the arguments put forward regarding each of the questions referred by the national 517
court in terms of these two broad schools of thought. As the observations raise related arguments
in respect of the more general issue of set-off in Question 1 and the more specific issue of legislative
interpretation in Question 3, I propose to treat those two questions together.

Questions 1 and 3
18.
Mr Jensen, KFK and the Commission submit that art 15(3) of the regulation contains a clear
prohibition of set-off. The set-off and, thus, satisfaction of a counter-debt cannot be deemed to be
equivalent to payment in the absence of receipt of the moneys in question. The Commission argues that
the farmer should receive the money in the same way as if he had sold his produce. The farmer should
have a choice regarding its expenditure, rather than having a particular use of the money imposed
unilaterally. Both Mr Jensen and the Commission rely on the objectives of the regulation, summarised
immediately below, to reinforce this interpretation of art 15(3). The Commission states that art 15(3) of the
regulation does not prejudice either the attachment procedure or the imposition and recovery of taxes,
provided that this is not discriminatory and does not undermine the common organisation of the markets.
In response to a question at the oral hearing, the Commission stated that art 15(3) of the regulation would
not preclude the set-off of aid against moneys owed by a farmer under Community law, as is confirmed by
the case law of the court.
19.
Mr Jensen submits that, irrespective of whether reciprocity between the debts in question is required
by Community law or merely by Danish law, this condition is not satisfied in the present case. The aid
scheme established by the regulation is regulated, paid for and supervised by the Community. He and
KFK state that the aid scheme is merely managed by the member states, and that aids must be passed
on faithfully to its intended beneficiaries (see Unione Nazionali di Produktori di Olive (Unaprol) v Azienda
di Stato per gli Interventi Sud Mercato Agricolo (AIMA) Case C-186/93 [1994] ECR I-3615). It is
immaterial, therefore, whether the member states capacity is described as that of an agent or
intermediary, as the aid is really owed by a third party, the Community. Mr Jensen and KFK argue that set-
off by the state must be prohibited in all cases, as a member state cannot be permitted unilaterally to
exploit its position as the distributor of aid, without any form of judicial control (see Italy v EC Commission
Case C-34/89 [1990] ECR I-3603). Member state authorities would otherwise be in a privileged position
relative to other creditors, a view echoed by the Commission at the oral hearing. Mr Jensen states that
the objective of the agricultural reforms was not to give member states an easy new tax-collection device.
Furthermore, KFK states that the distribution of aid by an organ of the state itself, rather than by an
independent designated body, was not imposed by Regulation 729/70, but was a matter of choice for the
member states.
20.
The Commission states that there is no general Community law obstacle to the imposition or recovery
of national taxes in the agricultural sector, in the light of its close links with the economy as a whole, so
long as the structure of the market or the functioning of market mechanisms established by the common
organisation of the markets is not thereby undermined (see art 39(2)(c) of the Treaty). There are no
Community rules regarding set-off of Community aids against national fiscal debts, though art 14 of
Regulation 3887/92 and the decisions of the court in DEKA Getreideprodukte GmbH & Co KG (in liq) v
EEC Case 518250/78 [1983] ECR 421 and Continental Irish Meat Ltd v Minister for Agriculture Case
125/84 [1985] ECR 3441 permit set-off in cases where the reciprocal debts concerned arise pursuant to
Community law. The Commission states that neutral national rules on set-off would, therefore, normally
be permitted to apply in the absence of Community rules. Although Mr Jensen suggests that the express
legislative provision for set-off in art 14(2) of Regulation 3887/92 excludes it in all other circumstances,
the Commission does not think that provision relevant to other situations.
21.
However, Mr Jensen, KFK and the Commission all argue that the application of national set-off rules in
the present case would prejudice the achievement of the objective of the regulation, which is to make a
direct contribution to farmers incomes and to compensate for losses from the scaling down of price
guarantees and export refunds. The aid is an integral part of the credit base of farming enterprises. This
would be threatened if their cash-flow were to be cut off, whereas the former intervention system, being
indirect, did not lend itself to the use of set-off by the state. KFK observes that a mortgage on the crop
alone does not provide sufficient security because of the reduction in prices entailed by the agricultural
reforms, and submits that farmers should be able to choose the creditors to whom they grant security. Mr
Jensen and the Commission also point out that attachment of debts would remain possible and claim that
this is a more transparent procedure which is open to all creditors.
22.
The member states argue that art 15(3) of the regulation is insufficiently precise to give rise to a
prohibition of set-off of fiscal debts against aid. There is nothing in the recitals in the preamble to the
regulation to suggest any such legislative intention. Variations on the formula used in art 15(3) are to be
found in other legislation, such as art 2(2) of Regulation 615/92 and art 7(2) of Commission Regulation
(EEC) 84/93 on the specific aid to be granted to producer groups in the raw tobacco sector (OJ 1993 L12
p 5). The Commissions linguistic interpretation of the article would lead to legal uncertainty, as it would
never be clear in any given case whether the operation of set-off would prejudice the wider objectives of
the aid scheme.
23.
The member states argue that set-off results in the full payment of aid to the affected farmer: in so far
as his indebtedness is reduced by that entire amount, he is thereby enriched. Payment can take place
other than by the issuing of a cheque or other banking transaction. Thus, set-off is neutral in its effects.
Article 15(3) of the regulation is designed to prohibit, in reality, the deduction by the member states from
aid payments of administrative charges, which are not financed by the Community, or of special taxes, as
distinct from those of general application. The Commission communication which accompanied the
proposal for the regulation is cited in support of this view (see COM(91) 379 final, 18 October 1991,
Reform of the common agricultural policy).
24.
The member states argue that, in the absence of Community rules, national rules can continue to
apply, provided they are non-discriminatory and do not undermine the common organisation of the market
(see Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das Saarland Case 33/76 [1976] ECR 1989,
Comet BV v Produktschap voor Siergewassen Case 45/76 [1976] ECR 2043 and Amministrazione delle
Finanze dello Stato v San Giorgio Case 199/82 [1983] ECR 3595). The courts case law recognises the
possibility of set-off, and its operation in the circumstances of the present case is no different, in principle,
from set-off in respect of wrongly paid Community aid pursuant to art 14(2) of Regulation 3887/92 (see
DEKA [1983] ECR 421). The provision for set-off in that provision 519 applies both to schemes which are
wholly financed by the Community and to those which are co-funded by the Community and the member
states. The universal right, in the member states legal systems, to set-off or to similar types of debt
recovery procedure should be considered to give rise to a general principle of Community law in its
favour. In the light of the principle of subsidiarity, any applicable national rules would have to be excluded
expressly.
25.
The consequences for debt collection of the agricultural reforms were not discussed by the Community
legislature, so that the objectives of those reforms cannot be invoked to prevent set-off. The reforms are
concerned with a move towards world market prices, rather than with guaranteeing a specific payment to
individual farmers, or their protection from creditors. Farmers do not escape the reach of the member
states general law9, and the close links between the agricultural sector and the rest of the economy can
vary the effect of the common agricultural policy in accordance with local economic conditions, including
levels of taxation (see art 39(2)(c) of the Treaty and Irish Creamery Milk Suppliers Association v Ireland
Joined cases 36/80 and 71/80 [1981] ECR 735).
9
See De Samvirkende Danske Landboforeninger v Ministry of Fiscal Affairs Case 297/82 [1983] ECR 3299.

26.
Prior to the reforms, there was nothing to prevent set-off of debts by operators, such as KFK, against
producers under the former system of indirect agricultural aids, nor does anything appear to preclude the
operation of set-off by private actors such as banks under the new regime. Furthermore, set-off was
possible as between public authorities and the operators under the price-support system. In any event, a
new system may give rise to new consequences for beneficiaries when managed within the framework of
a national legal system. The exclusion of the operation of set-off by member state tax authorities would
unjustifiably discriminate in favour of, and give a competitive advantage to, those farmers who do not pay
their taxes, to the detriment of the public treasury, of other categories of taxpayer, such as salaried
persons, and of farmers who punctually pay their taxes. Furthermore, private creditors such as KFK, who
are not even the intended beneficiaries of aid, would also benefit from this privilege. It would be especially
inappropriate since the state, unlike a private creditor, is not in a position to choose its debtors. The
recovery of taxes is a central state function, and the mandatory requirement of fiscal control is recognised
by Community law (see Rewe-Zentral AG v Bundesmonopolverwaltung fr Branntwein Case 120/78
[1979] ECR 649 (Cassis de Dijon)). There is a Community interest in the economies and fiscal health of
the member states. Moreover, it is unreasonable to force a solvent creditor to pay money to his insolvent
debtor.
27.
The member states emphasise the potential significance of this case for other types of debt recovery,
such as attachment. They take the view that the logic of the Commissions arguments would lead to the
condemnation of all types of seizure of aid, even by private parties, and would extend to set-off in favour
of the Community, which has already been expressly approved by the court. The exclusion of any type of
aid seizure for the purposes of debt recovery could make it more difficult for farmers to obtain credit. In
the alternative, continued acceptance of other debt recovery procedures such as attachment would
demonstrate the benefits of set-off, which is speedier and involves no additional cost for affected farmers.
The Commissions preference for judicial recovery procedures would interfere with the balance between
the rights of debtors and of creditors developed in the legal systems of the different member states.
520

Question 2
28.
Regarding Question 2(a), Mr Jensen submits that it is irrelevant that member states advance the
necessary moneys for aid payments from their own resources, as they have a right to claim a full
reimbursement from the Community. The change in payment procedure in 1988 (replacing the former
system of advance payment by the Community to the member states) cannot have altered the nature of
the agency relationship between the member states and the Community. The fact that member states can
only claim reimbursement if aid has been paid out in conformity with Community rules is in no way
inconsistent with such an agency relationship. The Commission took the view at the oral hearing that the
principle of reciprocity was not, in itself, a barrier to set-off because the member state acted on its own
account in distributing aid (see Eurico Srl v EC Commission Case 109/83 [1984] ECR 3581).
29.
The Ministry, Denmark and Finland argue that the debts in question are owed mutually by and to the
Danish state. The payment mechanism, the support of administrative costs by the member state, and the
fact that the member state is solely responsible to the Community for this expenditure, indicate that the
ministry is the farmers debtor in respect of the aid. The obligation to pay the aid is imposed directly upon
the member states, with the result that the beneficiary cannot claim aid directly from the Community (see
Unione Nazionali di Produktori di Olive (Unaprol) v Azienda di Stato per gli Interventi Sud Mercato
Agricolo (AIMA) Case C-186/93 [1994] ECR I-3615 and Eurico Srl [1984] ECR 3581).
30.
In response to Question 2(b), Mr Jensen argues that the general principle of equality requires that
reciprocity of debts be a condition of acceptance by Community law of the possibility of set-off. While the
court cannot interpret the Danish rules in this regard, it should make a finding as to the capacity in which
the ministry acts when managing Community aid schemes. Set-off would result in unacceptable inequality
as between member states, due to divergent practices.
31.
The Ministry agrees that the court should only rule on whether aid moneys are owed to the farmer by
the member state acting in its own capacity (on which it favours a positive response, however), and that
consequential questions of national law, regarding, for example, whether the state constitutes a single
juridical entity, should be left to be decided by the national court. The United Kingdom, on the other hand,
submits that reciprocity is a generally applicable condition for the operation of set-off in a Community law
context, but that it exists in the present context, as national authorities do not work for the account of the
Community, and that a wide margin of appreciation should be left to national law (see Eurico Srl [1984]
ECR 3581, Continental Irish Meat Ltd [1985] ECR 3441 and Unaprol [1994] ECR I-3615).
32.
Mr Jensen and the Commission argue, in respect of Question 2(c), that the operation of set-off in
respect of direct agricultural aids is discriminatory, and therefore unlawful, because equivalent national
subventions are either not subject to set-off at all or are subject to set-off limited to 20% of the total
amount (see BayWa AG v Bundesanstalt fr Landwirtschaftliche Marktordnung Joined cases 146/81,
192193/81 [1982] ECR 1503). The Commission adds that the highly discretionary manner in which set-
off is operated would in any event, create a permanent risk of discrimination. In the alternative, Mr Jensen
states that aid payments should be deemed to relate to the law of property, so that they cannot be set off
against fiscal debts.
521
33.
The Ministry argues that, although it is for the national court to undertake a comparison of Community
agricultural aids with national subventions, they are not equivalent, so that no unlawful discrimination
exists in Danish law.
34.
Mr Jensen suggests that the court should declare, in response to Question 2(d), that there is nothing
in Community law which indicates that member states should resort to otherwise impermissible recovery
techniques in respect of fiscal debts which contribute to the Communitys own resources. In any event,
only approximately 14% of VAT receipts form part of the Communitys own resources.
35.
 The Ministry submits that, if set-off of fiscal debts against Community agricultural aids is not generally
permissible, it should, at least, be possible in respect of taxes which contribute to the Communitys own
resources. It observes that Council Regulation (EEC, Euratom) 1552/89 implementing Council Decision
(EEC, Euratom) 88/376 on the system of the Communities own resources (OJ 1989 L155 p 1), requires
member states to do all in their power to collect the taxes in question.

Question 4
36.
Mr Jensen submits that, where examination of an aid application has been completed, it is contrary to
art 10(1) of the regulation for payment to be delayed, even before the final deadline set out in that
provision, in order to permit an additional check for possible fiscal debts. This delay serves member
states unilateral interests, and is contrary to the requirement in art 4(2) of Regulation 729/70 that member
states ensure that payments are made without delay and solely for the purposes laid down. Furthermore,
such delays result in inequality of treatment, because they will vary as between different member states
(see Deutsche Milchkontor GmbH v Germany Joined cases 205-215/82 [1983] ECR 2633). Mr Jensen
has sought interest from the ministry from the date on which the examination of his aid application ended.
The Commission, on the other hand, takes the view that art 10(1) of the regulation has no significance for
the present case independently of art 15(3).
37.
The Ministry remarks that the regulation contains no special rules on the processing of applications,
and that the member states have a wide margin of appreciation regarding the manner in which they
manage the payment of aids, provided the deadline is respected (see Unaprol [1994] ECR I-3615). It is
for the national court to determine whether a speedy examination of the application has taken place. Both
Ireland and the United Kingdom submit that payment should take place within a reasonable period after
the completion of the examination of the aid application, subject to observance of the final deadline
imposed by art 10(1).

IVANALYSIS
38.
The first three questions referred by the national court all concern the question whether there are any
obstacles to the application of the Danish rules on set-off in the circumstances of the present case, either,
firstly, on the basis of its inconsistency with Community law in general, taking into account factors such as
the capacity in which the ministry acts when distributing Community aids, or, secondly, on the particular
basis of the operation of the Danish rule of reciprocity, the allegedly discriminatory manner in which set-off
is operated, and the legal basis of the fiscal debt which is satisfied by way of set-off, or, thirdly, on the
basis of its inconsistency with art 15(3) of the regulation. I shall, therefore, treat these 522 three questions
together, although it will be convenient to examine first Question 3 regarding art 15(3), since that provision
is claimed to amount to an express negation of set-off, followed by the more general Question 1 and then
the more specific queries in Question 2. I will then examine separately Question 4, which relates to the
manner in which set-off is operated, and needs to be answered only if set-off is permissible.

Questions 1, 2 and 3
39.
The text of art 15(3) of the regulation, cited particularly by the Commission, does not expressly
mention set-off. The crediting of the aid to the beneficiary by way of set-off against and, pro tanto,
discharge of an existing fiscal debt to the state responsible for granting the aid is clearly distinguishable
from the deduction of special administrative fees or taxes from the aid, which are the only forms of
deduction mentioned in the Commission communication accompanying its proposal for the regulation.
However, the imposition of special charges to compensate for national administrative costs would secure
no benefit of equal value for the beneficiary, and would be clearly prohibited by art 15(3). In the case of
set-off, the beneficiary receives the entire monetary value of the aid, although he does not have full
control over its disposal. The same would be true in the case of attachment of the aid to discharge a debt,
or of any other process of execution applied against the beneficiary or his assets either in anticipation of
or after actual payment. These are all consequences of the fact that, as Advocate General Mancini put it
in DEKA [1983] ECR 421 at 439:
a person who is under an obligation to make a monetary payment offers his creditors a kind
of general guarantee in the form of his assets [T]he debtor is in a subordinate position in regard
to his own assets.
There would be no question of the aid not being considered to have been paid in full if it were transferred
to the beneficiary, formed (perhaps only momentarily) part of his assets and were then seized or attached,
against his will, for the benefit of any creditor. Execution before the actual transfer of money differs little,
from the point of view of the degree of liberty enjoyed by the beneficiary in respect of his assets, from any
form of post-payment execution10. Furthermore, there is no evidence that the Community legislature
intended, in adopting the rather laconic art 15(3), to limit the widely varying debt-recovery methods which
exist in national law. I conclude, therefore, that art 15(3) of the regulation does not prohibit the setting off,
under national rules, of aid under the regulation against prior debts to the state, where that is provided for
by national law and the beneficiary of the aid is credited with its full value.
10
Regarding its effects on creditors, see paras 47 to 54, below.

40.
 I now turn to the more fundamental inquiry required by Question 1, namely whether the application of
national rules on set-off in the context of the regulation is contrary to Community law.
41.
It is one of the fundamental characteristics of a common organisation of the market that the member
states are deprived of their powers unilaterally to regulate the sector concerned. The court said of
member states in EC Commission v Germany Case 48/85 [1986] ECR 2549 (para12), [t]heir legislative
competence can only be residual; it is limited to situations which are not governed by the Community
rules and to cases where those rules expressly give them power to act. The latter situation is exemplified
by art 8(1) of Regulation 729/70, which 523 provides for the supervision of the exercise of Community
policies and the prevention and remedying of irregularities to be governed by national law 11. In the former
situation, where Community law is silent, it is supplemented by national procedural or substantive rules in
order to give effect to Community law rights12. On the other hand, nothing in the common organisation of
markets deprives member states of their right generally to legislate in respect of independent aspects of
national policy. The court has, for example, ruled that the imposition of a temporary excise tax on
agricultural produce or a land tax on agricultural property, in pursuance of a national incomes policy
designed to share the burden of taxation, is not, in principle, incompatible with Community law. The
common agricultural policy is not intended to shield its beneficiaries from the effects of distinct national
policies of general application (see the judgments in Irish Creamery Milk Suppliers Association [1981]
ECR 735 (para 13) and De Samvirkende Danske Landboforeninger [1983] ECR 3299 (para 8)). The
courts recognition of the member states competence to pursue and protect their own fiscal interests 13
may, indeed, be flanked by recognition of a distinct Community interest in the fiscal efficacy of the
member states, not only because VAT and customs duties contribute to the Communitys own resources,
or because coordination of member state economic policies, including levels of public indebtedness, is a
matter of common concern in a developing economic and monetary union, but, ultimately, because the
proper functioning of the Community depends on the adequacy of the administrative, judicial and other
machinery which is operated and financed by the member states (see arts 3a(3) and 104c of the Treaty).
In my opinion, national rules governing the payment and collection of fiscal debts and all associated
matters, including processes of execution or attachment, priorities between creditors and, therefore, rights
of set-off come within the scope of that principle, provided they are, otherwise, compatible with
Community law.
11
See the judgments in Balkon-Import-Export GmbH v Hauptzollamt Berling-Packhof Case 118/76 [1977] ECR 1177 (para
5), Deutsche Milchkontor GmbH [1983] ECR 2633 (para 2), Italy v EC Commission Case C-34/89 [1990] ECR I-3603
(para 9) and BayWa AG [1982] ECR 1503 (para 29).
12
See the judgments in Rewe-Zentralfinanz [1976] ECR 1989, Comet BV [1976] ECR 2043 and San Giorgio [1983] ECR
3595 (para 12).
13
See also the judgment in Cassis de Dijon [1979] ECR 649 (para 8), Futura Participations SA v Administration des
Contributions Case C-250/95 [1997] ECR I-2471 (para 31) and, in a different context, my opinion in Garage Molenheide
BVBA v Belgium Joined cases C-286/94, C-340/95, C-401/95 and C-47/96 [1998] All ER (EC) 61 (para 43).

42.
In all such cases, the application of national rules is subject to conditions which, generally speaking,
are designed to ensure the effectiveness of Community law. For example, in Deutsche Milchkontor GmbH
[1983] ECR 2633 (paras 2223)14, the court outlined the conditions regarding national rules on recovery
of wrongly paid sums, expressly referred to in art 8(1) of Regulation 729/70:
14
See also the judgments in Balkon-Import-Export GmbH [1977] ECR 1177 (para 5), BayWa AG [1982] ECR 1503 (para
29) and Italy v EC Commission Case C-34/89 [1990] ECR I-3603 (para 9).

In the first place the application of national law must not affect the scope and effectiveness of
Community law. That would be the case in particular if the application of national law made it
impossible in practice to recover sums irregularly granted Secondly, national law must be
applied in a manner which is not discriminatory compared to procedures for deciding similar but
purely national disputes.
524
Where Community law relies implicitly on the supplementary framework of national substantive and
formal rules to secure the exercise of Community law rights, those rules are also subject to the conditions,
which I shall call the San Giorgio rules, that they may not be less favourable than those relating to similar
claims regarding national charges and they may not be so framed as to render virtually impossible the
exercise of rights conferred by Community law (see the judgment in San Giorgio [1983] ECR 3595 (para
12)).
43.
Independent national rules are subject to specific conditions, having regard to the fact that they are not
immediately or directly necessary for the implementation of Community law in the field in question. In the
field of the common organisation of agricultural markets, such national rules would be incompatible with
the Treaty if they interfered with the functioning of the machinery employed by those organizations in
order to achieve their ends (see the judgment in Irish Creamery Milk Suppliers Association [1981] ECR
735 (para 15))15. Such interference could, in the case of a national tax, be manifested by an effect on
price levels, on supplies on the market, or on the structure of agricultural production or of agricultural
holdings (see the judgments in Irish Creamery Milk Suppliers Association [1981] ECR 735 (para 22) and
De Samvirkende Danske Landboforeninger [1983] ECR 3299 (paras 1416)).
15
See also De Samvirkende Danske Landboforeninger [1983] ECR 3299 (para 1). The court has not found that the liability
to pay national taxes of farmers in receipt of Community support is contrary to the requirement in the applicable version
 of art 4(2) of Regulation 729/70 that member states use credits made available by the Commission solely for the
purposes laid down.

44.
The setting off, on the basis of national rules, of aids due to farmers under the regulation against fiscal
debts owed by those farmers to a member state belongs, in part, to both situations. Set-off is a procedure
provided for in Danish law which, while securing the payment in full of Community aids, does so in a
particular manner with a view to securing the independent national objective of recovering tax revenue. I
refer to the conditions regarding both types of rules in the analysis which follows. I do not think, however,
that in a hybrid case such as the present, the differently formulated conditions, recited in paras 42 and 43,
above, result in materially different requirements. The key issues are whether the implementation of set-
off renders actually or virtually impossible the enjoyment of rights under the regulation or otherwise affects
the functioning of the system established thereby, and whether set-off gives rise to discrimination relative
to the exercise of similar rights under national law.
45.
The court has already indicated in DEKA [1983] ECR 421 (paras 1314) that, in the case of an
insolvent trader, set-off by the Community of a reciprocal or related claim may be the only practicable way
to recover wrongly paid sums which are owed directly to the Community. Advocate General Mancini had
suggested in his opinion that, in order to ensure uniformity throughout the Community, such a right of set-
off should be deemed to be based on a general principle common to the legal systems of the member
states, and applied to claims having their basis in Community law ([1983] ECR 421 at 439). The same
reasoning clearly underlies the judgment ([1983] ECR 421 (paras 1315)) 16. The court noted that
Community rules on production and export refunds could give rise not only to debts which could be raised
by traders against the administering authorities, but also to claims against traders for the reimbursement
of wrongly paid sums, and that reciprocal and even related claims could result, as between 525
authorities and traders, which are an appropriate subject for set-off (see [1983] ECR 421 (para 13)).
16
Paragraph 15 refers expressly to a general principle against fraudulent assignment of claims.

46.
In Continental Irish Meat Ltd [1985] ECR 3441, the court accepted, without objection, the application of
national set-off rules in respect of monetary compensatory amounts owed, respectively, by and to the
intervention agency of the member state concerned. Two points may be made. First, the operation of set-
off is permitted for certain schemes by art 14(2) of Regulation 3887/92. Wrongly paid aids may be
deducted from aids still to be paid. Article 14(2) of Regulation 3887/92 does not, in my view, exclude, by a
contrario reasoning, the operation of set-off in other contexts. This much is clear, of course, from the
continuing applicability of the courts own case law in the matter, but also from the context of that
provision. Article 14 of Regulation 3887/92 is concerned generally with the recovery of wrongly paid sums,
and set-off is permitted as an alternative to other forms of recovery. It should not, therefore, be read as
prejudging the issue in respect of debts arising outside the particular context with which it is concerned.
Secondly, if Community law permits set-off, either directly in the Communitys own favour, or by the
member states in contexts where both debt and counter-debt arise under a Community scheme, then the
Communitys interest in the effectiveness of member state fiscal control dictates that set-off of aid under
the regulation against fiscal debts due to a member state should be permitted pursuant to national rules,
unless there is compelling evidence of a detrimental impact on the scope and effectiveness of Community
law, including that of the regulation. No distinction of principle should, in my view, be drawn, in the
absence of a clear contrary indication, between a debt owed to a member state authority charged with the
recovery of Community funds and one charged with the collection of taxes, such as would permit the first
but not the second to withhold payment by way of set-off in discharge of a debt admittedly due. To hold
otherwise would invert the San Giorgio rules by requiring national law to give preferential treatment to
Community law debts.
47.
The objectives of the regulation clearly relate to the general functioning of agricultural markets. This
emerges not only from the recitals quoted above, and from the statement in art 13 that expenditure under
the regulation is intended to stabilise agricultural markets, but also from the nature of the scheme
established by the regulation. Thus, while it is sought to compensate farmers for the drop in prices,
compensation is not calculated on the basis of particular farmers entitlements under the pre-existing
regime, but, rather, by reference to regional productivity, which may over or under compensate certain
farmers (art 4(1) and (2)). It is also subject to variation in the light of changes on the markets (art 15(1)),
while an excess of applications relative to the area forming the basis of the compensation calculation will
lead to proportionate aid reductions for all and, eventually, to increased, uncompensated set-aside (art
2(6)). Thus, contrary to the arguments of Mr Jensen and KFK, the regulation is not designed to protect
every individual farmer against any drop in disposable income. Furthermore, no such entitlement arises
from Community law generally (see SpA Eridania-Zuccherifici nazionali v Minister of Agriculture and
Forestry Case 230/78 [1979] ECR 2749). I accept the arguments advanced by the member states that
farmers were both directly and indirectly subject to the vagaries of set-off procedures under the old
regime (see para 25, above). Operators could set off the aid-subsidised prices owed to producers against
debts owed to them. Moreover, the intervention agency of each member state could operate set-off of
price 526 support aid against amounts owed by operators. In a situation of insolvency, this would
inevitably have resulted in a loss to the operators other creditors, possibly including farmers to whom
payment for produce was due17.
17
This could have occurred eg in the circumstances of Continental Irish Meat Ltd [1985] ECR 3441.
48.
In my opinion, the mere fact that direct set-off by the state against aids due to farmers was not
practicable in the past does not exclude its applicability now. The real question is whether the new system
established by the regulation would be so affected by set-off as to undermine its proper functioning.
49.
As I have already pointed out, set-off results in the farmer being credited with the full amount of the aid
in question, although his freedom to dispose of it as he pleases is restricted. In this, he is already at an
advantage relative to other forms of deduction which are, in principle, permitted by Community law, such
as the levying of taxation, which actually reduce the value of the aid to the farmer (see Irish Creamery
Milk Suppliers Association [1981] ECR 735 and De Samvirkende Danske Landboforeninger [1983] ECR
3299). The deduction of aid moneys at source to satisfy existing tax debts, in a fashion which ensures
that the full value of the aid none the less accrues to the farmer, having regard to his liabilities as well as
his disposable income, does not appear any more likely to affect the objectives of the common
organisation of the market than such direct taxation.
50.
Furthermore, not surprisingly, nobody has contested the right of creditors generally to avail of
attachment or other forms of execution or judicial attachment, against the debtor farmers will. The
distinguishing feature is not, however, the degree of protection of the farmer, whose aid may be diverted
in all cases, but that of other creditors, persons not directly addressed by the regulation. There is no
immediate Community interest in disturbing the balance drawn by national law between the rights of
different classes of creditors.
51.
The degree of protection of a farmers private creditors has been linked to the functioning of the
regulation, however, through the invocation of the need for liquidity. The interest of banks or supply
companies in offering credit to farmers is said to be reduced if either their expectation of being able to
attach aid, or their formal entitlement to the aid through the operation of an assignment, is endangered by
the intervention of set-off operated by the state, to the ultimate detriment of farmers. There is nothing to
suggest that the regulation is concerned with farmers liquidity as such. This depends on many structural
factors as well as the level of price or aid they receive factors like seasonality, levels of interest rates, land
values and so on. Putting the matter bluntly, the Community can hardly intend to contribute to farmers
liquidity by enabling them to delay or defeat legitimate tax claims against them.
52.
In any event, a prohibition of state set-off of national fiscal debts against aid on grounds which were
equally applicable to set-off by other parties and to attachment could ultimately have a much more
detrimental effect on farmers creditworthiness and liquidity than continued recourse by the state to one
among a number of forms of recovery provided for in national law.
53.
Set-off is of importance only in situations of actual or impending insolvency. Its application to mutual
running accounts between traders is routine and uncontroversial, because it suits the convenience of both
parties. It is inherently unlikely that the insolvent farmer, or one in financial difficulty, would personally
enjoy the free disposal of the aid moneys. The various forms of set-off and recovery which exist in the
legal systems of the member states are all 527 concerned with managing the risks posed to creditors by
insolvency18. The possibility of set-off grants an advantage to the creditor whose debtors debt to him is
matched, in whole or in part, by a debt he owes to his debtor, of which he can take advantage to alleviate
that risk. It has been asserted that the state should not be permitted to exploit its privileged access to aid
moneys by way of set-off. However, it is in the nature of set-offa practice which the court has not found
objectionable per se, in other contextsthat one creditor has the good fortune to find himself in such a
privileged position, relative to others. It is important not to view that privilege in isolation from others which
may exist. All forms of security are contingent on some kind of privileged access to the means of
recovering the debt in question. Other creditors may have organised their affairs in order to have equally
privileged access to other assets of the farmer. These might include ensuring that they owe him amounts
for produce which match the credit granted, retaining title in equipment supplied or securing a mortgage
on the farmers lands, crops or other property. The argument against such privileged access, by the state
or any other creditor, is essentially an argument based on fairness, relative to other creditors, rather than
one touching on the proper functioning of the regulation. I can only repeat that there is no apparent need
in Community law to disturb the balance drawn by the various national legal systems between the
entitlements of different types of creditor, secured or unsecured. The same is true of the relative status in
national law of the right to set-off and of rights under arrangements with creditors such as that which was
apparently concluded by Mr Jensen.
18
See, once more, the judgment in DEKA [1983] ECR 421 (para 14), as well as the opinion of Advocate General Mancini
([1983] ECR 421 at 439), his opinion in Continental Irish Meat [1985] ECR 3441 at 3447 and the reference to a
shortage of funds in my opinion in Garage Molenheide [1998] All ER (EC) 61 (para 41).

54.
It has also been argued that the farmers ability to secure private credit is reduced if set-off is availed
of by the state, to the detriment of his private creditors: the farmer cannot then use the benefit of the aid to
give security to the creditor of his choice. However, a farmer could never use a future aid payment to
secure credit from the fiscal authorities, who, as has been said, cannot avoid being his creditor, and, thus,
do not have to be enticed to grant credit. Just as public authorities are the only bodies likely to be obliged,
in the general interest, to be the creditors of any given economic actor (for taxes, customs duties, or other
dues), they are the bodies which are responsible, in the general interest, for the payment and distribution
of various types of aid and benefit (economic, social security, and so on) to wide sections of the
population. The fact that the state may have privileged access to various types of public aid, including
Community-funded aid, cannot, therefore, be examined in isolation from the fact that it has the burden of
collecting taxes and other dues which ultimately go to fund such aids, as well as the machinery for
administering them. While their functions are very different, no cogent argument has been advanced to
suggest why the benefits of risk management pursuant to ordinary national debt-recovery rules should be
denied to the state, whose role entails that it be both creditor and debtor, while they can continue to be
granted to creditors who freely accept the risk of advancing credit to freely contracting economic actors
such as farmers.
55.
It also follows from the states necessary role as fiscal creditor that, in the absence of set-off, farmers
in difficulties with their taxes would be able to avail of 528 this involuntary credit while simultaneously
borrowing on the security of their entitlement to aids under the regulation. Such borrowing would
inevitably result in a loss of revenue to the state. The court has already had occasion to condemn
assignments, where they amount to fraud on creditors (in that case, the Community itself) in DEKA [1983]
ECR 421 (paras 1518).
56.
Finally, my response to the argument that permitting set-off in this case will result in widespread
differences in treatment as between the different national legal systems, leading to a disruption in the
functioning of the common organisation of the markets is that, if the disparities prove to be such as to
compromise the equal treatment of producers in different member states or to distort or impair the
functioning of the common market, it is not for the court to exclude the application of all such national
rules on debt recovery, but for the Community legislature to adopt the provisions needed to remedy such
disparities (see the judgment in Deutsche Milchkontor GmbH [1983] ECR 2633 (para 24)).
57.
I shall now turn to Question 2(a) to (d), where the national court draws attention to certain specific
aspects of the national rules.

Question 2(a) and (b)

58.
Parts (a) and (b) of Question 2 both relate, in effect, to the question of the reciprocity of debts
subjected to set-off. Question 2(a) asks, essentially, if the capacity in which the state acts when granting
aid under the regulation affects, as a matter of Community law, its national law right to set off against it
fiscal debts owed to it. Question 2(b) asks if it is relevant that the national law on set-off imposes a
reciprocity requirement. The court has not been asked to define the capacity in which the designated
member state authority acts when distributing Community funds in the form of aid.
59.
The court indicated in DEKA [1983] ECR 421 (para 13) that reciprocal or even related claims are an
appropriate subject of set-off in Community law. As disclosed in the case law, reciprocity requirements
exist in the various national legal systems which permit set-off, although they vary greatly in detail. In
some member states, the state and its agencies do not constitute a single entity for that purpose. In
Ireland, for example, each government minister is a distinct corporation sole, so that one minister cannot
set off an amount due to another against a payment due by him. It is an obvious pre-condition for the use
of set-off in the present case that Danish law treat the ministry and the tax-collecting authority as a single
entity for that purpose, which is a matter for the national court. Community law does not determine the
meaning or range of application of reciprocity or mutuality demanded for the application of national rules
of set-off19. The test in Community law remains the more general one of whether the national rule affects
the scope and effectiveness of Community law, including the regulation.
19
The situation is different, of course, where set-off is operated directly by the Community itself, as in DEKA, in which
case, in the absence of specific rules, a Community standard must be devised based on the general principles common
to the laws of the member states.

60.
In its judgment in Unaprol [1994] ECR I-3615, for example, the court noted that Community law did not
regulate the right to retain interest earned on sums held by a national intervention agency pending
disbursement to beneficiaries in the form of aid. The existence of such a right was thus a matter to be
settled by national law, provided that the rules in question did not jeopardise the uniform application or
effectiveness of Community law (see [1994] ECR I-3615 (paras 24 52925, 28)). The court did not need,
therefore, to classify the national agency either as an intermediary or as owner of the sums in question
(see para 12).
61.
I have already stated my opinion that the interests of aid beneficiaries are not prejudiced by set-off in
any way which disturbs the proper functioning or the achievement of the objectives of the regulation, or
which contravenes the provisions of Regulation 729/70. For the same reasons, it cannot be said that the
Community suffers any prejudice thereby.
62.
It may be helpful, none the less, to state my view on the capacity in which the authorities designated
by member states act, as a matter of Community law, when they pay out moneys under the regulation, in
case the court should decide that the capacity in which these agencies act when granting aid moneys is
independently significant, or in case it might be useful to the national court in applying the national rules.
In my opinion, the relationship of the member states and their agencies with the Community 20 in this
context differs little from that which arises when they take steps to comply with any obligation imposed by
Community law, be it the collection of VAT, the enforcement of directly effective Community rules, or the
implementation of directives. They act on their own account, and exercise their own powers, as subjects
both of national and Community law. They do not act as the Communitys agent or intermediary, or on its
behalf, in anything other than the colloquial sense. They apply Community law on their own responsibility
(see the judgments in SA Compagnie Interagra v EC Commission Case 217/81 [1982] ECR 2233 (para 7)
and Sucrimex SA v EC Commission Case 133/79 [1980] ECR 1299). Thus, in a series of cases, the court
has held, with respect to various losses suffered by private individuals as a result of implementation of the
common agricultural policy, that, regardless of the role played by the Commission, it is the intervention
agencies which are responsible, due to their independence from the Community authorities 21. This
conclusion is not in any way undermined by the fact that the Community funds the necessary expenditure
under the regulation (subject, of course, to the financially significant exception of administrative costs).
The fact that the Community provides the necessary funds before or after the corresponding payment of
aid is not, therefore, of any relevance.
20
As I have already pointed out, the relationship of the various branches of the state inter se is determined by national
rules on reciprocity.
21
See the judgments in Eurico Srl [1984] ECR 3581 (para 18), Russo v Azienda di Stato per gli Interventi Sud Mercato
Agricolo (AIMA) Case 60/75 [1976] ECR 45, Granaria BV v Hoofdproduktschap voor Akkerbouwprodukten Case 101/78
[1979] ECR 623, Sucrimex SA [1980] ECR 1299 and Interagra [1982] ECR 2233.

63.
The court has often stated that the common agricultural policy operates on the basis of a division of
powers between the Community and the member states (see the judgments in Unaprol [1994] ECR I-
3615 (para 27) and Balkon-Import-Export GmbH [1977] ECR 1177 (para 5))22. This division is expressed
in the agricultural sector by, inter alia, arts 4 and 8 of Regulation 729/70. Thus, in Italy v EC Commission
Case C-34/89 [1990] ECR I-3603 (para 10), the court distinguished between the relationship, under art 8
of Regulation 729/70, of intervention bodies and economic operators, which is governed by national law,
and that between member states and the Commission, which
22
See also the discussion by Advocate General van Gerven in his opinion in Unaprol [1994] ECR I-3615 (paras 1415).

530
is concerned not with the granting of aid or the recovery of excessive advance payments as
such, but with whether the Member State concerned or the Community has to bear the relevant
financial burden.23

23
This division of powers is also evidenced by the supervisory regime established by art 9 of Regulation 729/70. See the
judgment in France v EC Commission Case C-366/88 [1990] ECR I-3571 (paras 2023).

Question 2(c)
64.
Here the national court raises the question of discrimination and, in particular, that of the significance
of a limitation in practice of set-off of certain national environmental and development subsidies against
their beneficiaries fiscal debts to the state. Any question of discrimination can only be resolved through a
comparison by the national court of the rules applicable to different types of debts, and of the material
characteristics of those debts, though the court is competent to provide guidelines for that process of
comparison. The overriding consideration is that national rules must not allow set-off to be used more
freely against Community law than against national-law debts.
65.
The first possibly discriminatory aspect of the application of the Danish rules relates to the existence of
what is referred to in the order for reference as an exception for general and social considerations. While
the precise nature of this exception is unclear, unlawful discrimination would exist if it were not also
available, in so far as it is applicable, to the process of set-off of fiscal debts against Community
agricultural aids. If, for example, the exception permits, or requires, the personal circumstances of the
debtor to be taken into account, it would be contrary to Community law for a farmer to be deprived of the
full sum of aid, by virtue of his fiscal debts, while another person, in a similar position, or even the same
person, could claim a derogation in respect of all or part of a debt owed by the state under national law. If,
on the other hand, entire categories of payments by the state, such as, for example, salaries, pensions or
social security entitlements, or specified minimum amounts of such payments, are excluded from the
scope of national rules on set-off on social or general grounds, the national court must examine whether
such debts are comparable to aids under the regulation. To that end, the national court should take into
account that, while the regulation is structural in nature, it is still, as the second recital in the preamble
indicates, concerned with farmers incomes, and that art 39(1)(b) of the Treaty lays down that one of the
objectives of the common agricultural policy shall be to ensure a fair standard of living for the agricultural
community. Save in the case of compensatory payments in respect of set-aside, aids paid under the
regulation merely supplement the price received by the farmer for his crop, but this would only be material
if the national law restrictions on set-off of salaries and so on are lifted in the light of additional income
sources.
66.
Secondly, the exceptional treatment of certain national economic subventions must be examined. At
the time when the set-offs at issue in the present case were effected, it appears that at least certain of
such subventions were not, in practice, subject to set-off at all, for fear of giving rise to a disincentive to
implementing the projects in question. Certain of the exempted subventions related to projects in the
agricultural sector, regarding both primary production and secondary processing. It is not necessary, in
my view, for the exempted national aids to be comparable in all respects and in detail with aids granted
under the regulation, in order for a finding of discrimination to be made. As Advocate 531 General Mancini
observed in his opinion in San Giorgio [1983] ECR 3595 at 3634 in the context of a national tax system, to
divide national tax measures (or, as in the present case, national subventions) into a number of sectors,
and then to confine the applicability of the principle of non-discrimination within each of those sectors
would not just amount to a weakening of that principle but would come close to destroying it 24. Thus, the
court spoke in Finsider v EC Commission Case 250/83 [1985] ECR 131 (para 8), of differences in
treatment being justifiable where they were based on objective and substantial differences (my
emphasis). Any aid directed at the support of economic activity which might not otherwise be possible,
given the structure of the market, should be seen as comparable. It is not even necessary for such a
finding that the exemption apply to aid in the agricultural sector, as we are not necessarily concerned with
discrimination as between different types of aid received by the same beneficiary. If compensatory
payments under the regulation are found to be comparable to any of the various types of national
subventions, unlawful discrimination will exist, even if national subventions are, in theory, subject to full
set-off, if; in practice, they are not.
24
See also the judgment in Albert Ruckdeschel & Co v Hauptzollamt Hamburg-St Annen, Diamalt AG v Hauptzollamt
Itzehoe Joined cases 117/76 and 16/77 [1977] ECR 1753 (para 7).

67.
The existence of discretion on the part of the Danish authorities regarding the application of set-off
rules is not necessarily unlawful, in my view, by reason of what the Commission describes as the
permanent risk of discrimination. The decision to avail of set-off is probably always discretionary, in the
sense that it is an option of the creditor, whose resort to it will turn on his perception of the solvency of the
debtor. Derogations in the light of the personal circumstances of tax debtors or of the likelihood of their
being able to remain in business are also inherently discretionary in character. On the one hand, the court
has ruled out the application of national rules which entail the exercise of a discretion as to the
expediency of waiving recovery of sums unduly or irregularly paid from Community funds, particularly in
accordance with economic criteria (see the judgments in BayWa AG [1982] ECR 1503 (para 30),
Deutsche Milchkontor GmbH [1983] ECR 2633 (para 22) and Balkon-Import-Export GmbH [1977] ECR
1177 (para 5)). On the other hand, the court has permitted the waiver of the recovery of such sums on
grounds of personal equity, pursuant to non-discriminatory national rules which do not impair the
objectives of the common organisation of the market in question (see the judgment in Peter v Hauptollamt
Regensburg Case C-290/91 [1993] ECR I-2981 (para 11)) 25. The court has also approved the non-
discriminatory application of national rules which permit, on equitable grounds, the remission after the
normal national time limit of over-paid customs duties, where this does not alter the effect of the relevant
Community-law rules (see the judgment in Otto Reichelt GmbH v Hauptzollamt Berlin-Sd Case 113/81
[1982] ECR 1957 (para 16)). The discretionary grant to a farmer of compensatory payments under the
regulation, on social or equitable grounds, instead of their retention by way of set-off, clearly does not
impair the functioning of the regulation. It is not, in my view, comparable to the waiver of recovery of funds
improperly paid from Community funds. It is much closer to the very different circumstances of the Otto
Reichelt GmbH case, where the exercise of a discretion in accordance with national equitable rules was
permitted to counteract a national rule on limitation periods which restricted an importers ability to 532
recover overpaid duties to which he was, in principle, entitled as a matter of Community law. For that
reason, I also think that the exercise of a discretion regarding the exercise of set-off on broader economic
grounds would also be permissible. However, this remains subject to the rule against discrimination. if
some comparable national subventions are, in practice, never subject to set-off, the mere possibility of a
discretionary waiver of set-off in respect of payments under the regulation will not satisfy the requirement
of equal treatment of Community-law and national-law payments.
25
Advocate General Jacobs distinguishes an equitable discretion from a discretion as to the expediency of waiver of
recovery: see the opinion in Peter [1993] ECR I-2981 (para 22).

Question 2(d)
68.
In the light of the foregoing discussion, I do not think that it is necessary for the application of non-
discriminatory national set-off rules to aids under the regulation that the counter-debt have as its legal
basis Community law, or that part of the counter-debt is required to be passed on to the Community as
part of its own resources. However, in the event of the court disagreeing with my analysis, the fact that
member states are required by Community law to implement a system of value added tax, the tax which
Mr Jensen had failed to pay, would be significant. The member states perform essentially the same
function when collecting VAT as when they distribute Community agricultural aid, that is, as independent
subjects of Community law complying with their duties thereunder26. The facts that, in the former case,
they distribute money provided entirely by the Community, while underwriting administrative costs
themselves, and, in the latter case, they keep the lions share of tax receipts, passing on only a small
fraction to the Community, are immaterial. In that case, the debts could be said to be reciprocal in the
same way as those in DEKA [1983] ECR 421, where, although both of the debts at issue arose under the
common agricultural policy, they did not relate to the same schemeone related to production refunds,
the other to wrongly paid export refunds. If Community law permits set-off by the Community between
debts arising under two different legislative schemes, it should also permit the member states to set off,
pursuant to non-discriminatory national rules, tax debts which they are obliged by Community law to
collect against agricultural aids which they are obliged by Community law to pay out to the tax debtor.
26
For the basic obligation to introduce a system of value added tax, see art 1 of Council Directive (EEC) 67/228 on the
harmonisation of legislation of member states concerning turnover taxesStructure and procedures for application of
the common system of value added tax, (OJ S edn 1970(I), p 16).

Question 4
69.
Article 10(1) of the regulation lays down an eleven-week time period within which aid is to be paid out,
but does not expressly require that payment be made as soon as the designated member state agency
has verified that the applicant farmer is qualified to receive such a compensatory payment. However, it
would, in my view, interfere with the proper functioning of the regulation for payment to be delayed once
an applicants entitlement is established, even if the prescribed time limit is respected. While it is not, in
principle, contrary to Community law for member state authorities to use compensatory payments for the
purposes of set-off, this does not mean that their efforts to check whether a farmer is also a tax debtor
can be permitted to disturb the ordinary functioning of the regulation. Such delays would, presumably,
affect all farmers, and not just 533 those who are found, upon enquiry, to be indebted to the fiscal
authorities. Any delay in payment could be costly to farmers, especially to those who wish to use the aids
to repay debts to third parties on which interest is payable. Article 13 of the regulation refers to Regulation
729/70, the applicable version of art 4(2) of which states that member states shall ensure that credits
supplied by the Commission are used without delay. Although this provision has been deprived of much of
its effect by the fact that the member states have, since 1988, received advances only in respect of
expenditure already incurred out of national resources, it still testifies to the Communitys interest in the
prompt payment of, inter alia, sums intended to stabilise agricultural markets. Furthermore, the delay in
question is probably objectively unnecessary.

VCONCLUSION
70.
In the light of the foregoing, I propose that the Court of Justice answer the questions referred by the
stre Landsret as follows:
(1) It is not contrary to art 15(3) of Council Regulation (EEC) 1765/92 establishing a support
system for producers of certain arable crops for a member state, pursuant to national rules, to set
off, without additional cost, compensatory payments to which a farmer is entitled under that
Regulation against fiscal debts owed by that farmer to the member state;
(2) Community law does not preclude a member state from setting off, pursuant to national
rules, an amount due to the beneficiary of aid under a Community measure against outstanding
debts to a member state, provided that those national rules do not render impossible the enjoyment
of the Community law right in question or hinder the functioning of the system established by the
Community measure, and are not applied in a discriminatory fashion;
(3) The setting off by a member state, pursuant to national rules, of compensatory payments
under Regulation 1765/92 against outstanding fiscal debts to the member state does not render
impossible the enjoyment of rights under that regulation or hinder the functioning of the system
established by it;
(4) The answer given at point (3) is not affected by the fact that the amount of aid is paid in
advance by the member state, which has only a claim to be reimbursed for the aid paid out,
provided that the rules of Community law on payment are satisfied, and which must itself defray the
expenditure involved in the administration of the system established by Regulation 1765/92;
(5) The answer given at point (3) is not affected by the existence of national rules which require,
as a condition for effecting set-off, that there be reciprocity between the debtor under the principal
claim and the creditor under the counterclaim;
(6) The answer given at point (3) is not affected by the legal basis for the outstanding debt to the
state against which set-off is to be effected nor, in particular, by the fact that part of the sum to be
set off constitutes part of the Communitys own resources;
(7) It is for the national court to determine whether national subventions, which are subjected, in
practice, to set-off in an amount not exceeding 20% of the sum involved, are comparable to
compensatory payments under Regulation 1765/92, for the purpose of establishing whether the
national rules on set-off are applied in a discriminatory fashion. It is, equally, for the 534 national
court to determine, for the same purpose, whether compensatory payments under that Regulation
are comparable to certain other types of payment in respect of which set-off is restricted on social
grounds;
(8) It is contrary to Community law for the disbursement of compensatory payments under
Regulation 1765/92 to be delayed in order to permit an investigation of whether the state has
outstanding claims against the beneficiary in respect of which it wishes to effect a set-off.

19 May 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 10 April 1995, received at the Court of Justice of the European Communities on 24 April
1995, the stre Landsret (the Eastern Regional Court) referred to the Court of Justice for a preliminary
ruling under art 177 of the EC Treaty a number of questions on the interpretation of Community law with
regard to set-off of amounts paid under Community law against debts payable to a member state and of
arts 10(1) and 15(3) of Council Regulation (EEC) 1765/92 establishing a support system for producers of
certain arable crops (OJ 1992 L181 p 12).
2.
Those questions were raised in proceedings brought by Mr Jensen, a farmer entitled to a
compensatory payment under the regulation, and Korn- og Foderstofkompagniet A/S (KFK), the assignee
of another compensatory payment due under the same regulation, against Landbrugsministeriet, EF-
Direktoratet (the Directorate of European Affairs of the Ministry of Agriculture), concerning set-off by the
latter between the compensatory payments in question and outstanding debts payable to the state.

THE COMMUNITY LEGISLATION

Regulation 729/70
3.
Article 4(1) of Council Regulation (EEC) 729/70 on the financing of the common agricultural policy (OJ
S edn 1970(I) p 218) provides that member states are to designate the authorities and bodies which they
will empower to effect the expenditure referred to in arts 2 and 3. Under art 4(2), the European
Commission is to make available to member states the necessary credits so that the designated
authorities and bodies may, in accordance with Community rules and national legislation, make the
payments referred to in art 4(1). The member states are to ensure that those credits are used without
delay and solely for the purposes laid down.
The regulation
4.
According to the second recital in the preamble to the regulation, the best way of ensuring better
market balances is to approximate the Community prices of certain arable crops to the prices of the world
market and to compensate the loss of income caused by the reduction of the institutional prices by a
compensatory payment for producers who sow such products. According to the 18th recital, it is
necessary to determine certain conditions for applying for compensatory payments and to specify when
producers are to be paid.
5.
Article 2(1) of the regulation provides that Community producers of arable crops may apply for a
compensatory payment under the conditions set out in its first title. In particular, under the second sub-
paragraph of art 2(2), the compensatory payment is to be granted for the area which is down to arable
535 crops or subject to set-aside in accordance with art 7 of the regulation and which does not exceed a
regional base area.
6.
Article 10(1) provides that compensatory payments for cereals and protein crops and the
compensation for the set-aside obligation are to be paid between 16 October and 31 December next
following the harvest.
7.
Under art 15(3), the payments referred to in the regulation are to be paid over to the beneficiaries in
their entirety.
The national provisions
8.
It appears from the order for reference that, under general rules of Danish law, public authorities may
recover fiscal debts owed by the beneficiaries of public aid in three ways.
9.
First, like any creditor, they may apply for attachment of the aid due to their debtor; if the debtor
objects, a decision of the competent court is necessary. The state, like any other creditor, may obtain an
attachment order in respect of entitlement to the aid itself; the aid is then paid directly to it, as it would be
paid to any private creditor who resorted to attachment proceedings. In the event of the aid entitlement
becoming subject to several attachment orders, the general rules concerning priority apply.
10.
Second, public authorities may recover their debts by accepting an assignment by the beneficiary of
aid of his entitlement to it. Where assignments are made to more than one creditor, the rules concerning
priority apply.
11.
Lastly, public authorities may recover their debts by effecting set-off between aid granted by a national
administration and the debt owed to that administration by the person entitled to the aid.
12.
For set-off of that kind between unconnected debts to be available, whether for the benefit of public or
of private creditors, several conditions must be fulfilled. First, there must be reciprocity between the two
debts in that the creditor in one case must be the debtor in the other. Next, the debt payable to the person
seeking set-off must have fallen due. Finally, there must be two debts of sums of money or two
obligations relating to fungibles of the same kind.
13.
Under Justitsministeriets circulreskrivelse No 186 (Circular No 186 of the Ministry of Justice) of 22
November 1983, which was sent to all departments of the central administration, it is not permissible to
effect set-off between claims not arising under the law of property and obligations such as tax debts,
VAT liabilities and fines and debts payable by the state which do fall within that law (in general, debts
originating in contract).
14.
Finally, Lov No 284 om aendring af forsekellige lovbestemmelser om inddrivelse af statskrav (the Law
amending certain provisions of legislation on the recovery of state debts) of 27 April 1994, which entered
into force on 1 July 1994, made set-off available, up to a maximum of 20%, in respect of a number of
specific trade or environmental subsidies granted by the Danish state within the areas of responsibility of
the Ministries of Energy, Industry, Agriculture and the Environment.

The disputes in the main proceedings

Danish practice regarding set-off
15.
According to the order for reference, the Directorate has been effecting set-offs since 1978 against, in
particular, debts to the state for VAT or other fiscal 536 liabilities when paying out aid due in relation to
common organisations of markets in agricultural products.
16.
By letter of 28 July 1992 the Directorate asked the Commission whether that practice could be
maintained.
17.
By letter of 12 November 1992 the Commission replied that, in so far as national legislation allowed
such set-off, it had no objection to the paying authoritys resorting to set-off against debts to the state
when paying out aid in respect of the 1992/93 marketing year under Commission Regulation (EEC)
615/92 laying down detailed rules for a support system for producers of soya beans, rape seed, colza
seed and sunflower seed (OJ 1992 L67 p 11), provided that the national legislation drew no distinction
between the payment of such aid and that of national aid and that payment thereof was not rendered
impossible by the national legislation concerning set-off.
18.
However, the Commission emphasised in particular that art 15(3) of Regulation 1765/92 and art 2(2) of
Regulation 615/92 required that the producer receive the full amount of the aid without reduction of any
kind; therefore, in its view, the member states could not impose, directly or indirectly, any charges for
processing applications for the compensatory payments provided for by those regulations.
19.
Following that letter, the Directorate, adhering to its previous practice, effected set-off in respect of the
amounts of aid paid in 1993 under both Regulation 615/92 and Regulation 1765/92.
20.
By letter of 7 October 1994 the Commission then drew the attention of the Danish Ministry of
Agriculture to the most recent opinion of its Legal Service, to the effect that the regulation precluded
national authorities from effecting set-off between Community aid and amounts due to them under
national schemes or provisions. A memorandum of 27 April 1994 from the Commissions Legal Service to
the Director General for Agriculture concerning the problem of set-off between national fiscal debts and
Community aid emphasised in particular that the regulation, which took precedence over national law,
contained a specific provision to the effect that the various payments were to be made to producers in
their entirety. Moreover, the set-off mechanism had undermined the aim of the system of direct income
support, since a member state could recover tax debts from farmers without having to follow the
procedures normally prescribed for that purpose.
21.
Following that letter, although disputing the Commissions position, the Danish Minister of Agriculture
decided to cease effecting set-off between aid payable to farmers for 1994 under the regulation and debts
to the state relating, in particular, to VAT and other levies.
Mr Jensens application for aid
22.
On 9 May 1993 Mr Jensen applied to the Directorate for area aid for the 1993 harvest. It is undisputed
that he fulfilled the conditions laid down by the regulation.
23.
Since, however, in December 1993 he owed the state an amount in respect of VAT which exceeded
that of the aid, he was informed on 20 December 1993before the Commission changed its view
concerning the Danish practice regarding set-offthat the entire amount of the aid, namely D Kr 33,563,
would go towards discharge of his VAT debt.
537
24.
 The Directorate dismissed the complaint lodged by Mr Jensen against that decision on the ground that
the set-off was justified and that all the relevant conditions laid down by Danish law were satisfied.
25.
Mr Jensen then instituted proceedings against the Directorate before the stre Landsret to secure
payment of the D Kr 33,563 in respect of area aid under the regulation.
26.
According to the case file, Mr Jensen, having experienced financial problems in 1993, sought to come
to an arrangement with his creditors under which they would receive only a percentage of their claims. As
it was regarded as income for the purposes of the arrangement, the area aid was to form part of the
overall amount to be shared by the creditors, one of whom was the tax administration.
Whilst the set-off allowed that administration to secure full payment of its claim, it also nearly
compromised implementation of the scheme of arrangement; the arrangement was still put into effect, but
payments to other creditors were significantly reduced.
The assignment to KFK
27.
According to the order for reference, in the spring of 1993 another farmer, Mr Stenholt, had assigned
to KFK the annual amount of aid he could claim under the regulation, namely D Kr 45,574. The document
recording the assignment had been notified to the Directorate, which took note of it, subject to the states
right of set-off.
28.
Since Mr Stenholt had become indebted to the state before the date of the assignment to KFK and
since his debt to the state fell due before the aid could be paid, the Directorate effected set-off to cover
the states claim against the beneficiary of the aid and informed KFK that, as a result, it would not receive
any payment.
29.
KFK therefore also instituted proceedings against the Directorate before the stre Landsret, seeking
payment of the aid assigned to it by Mr Stenholt.
The questions on which a ruling is sought
30.
Considering that an interpretation of Community law was required from the Court of Justice, the stre
Landsret stayed proceedings and submitted the following questions for a preliminary ruling:
(1) Does Community law in general preclude a Member State from setting off an amount due to
the beneficiary of aid under a Community measure against outstanding debts to a Member State?
(2)(a) Is it of any significance for the answer to Question 1 whether the amount of aid under
Community law is paid in advance by the Member State which has a claim to be reimbursed for the
aid paid out only if the rules of Community law on payment are satisfied and which must itself
defray the expenditure involved in the administration of the support system? (b) Is it of any
significance for the answer to Question 1 that under the Member States rules on set-off it is a
condition for effecting set-off that there be reciprocity between the debtor under the principal claim
and the creditor under the counterclaim? (c) Is it of any significance for the answer to Question 1
that the Member States practice with regard to certain trade and environmental subsidies is
established in such a way as to permit set-off in an amount not exceeding 20% of the said state
subsidies? (d) Is it of any 538 significance for the answer to Question 1 what legal basis exists for
the outstanding debt to the state against which set-off is to be effected? An answer is desired in
particular to the question whether the Member States have greater scope to effect set-off if all or
part of the sum to be set off constitutes part of the Communitys own income.
(3) If Questions 1 and 2(a) to 2(d) are answered to the effect that set-off is in general possible,
or possible subject to certain conditions, is Article 15(3) of Council Regulation No 1765/92 to be
interpreted as meaning that a Member State is precluded from requiring a national intervention
agency to effect set-off in the case of a beneficiary of compensatory payments with outstanding
debts to the state which could otherwise be involved in set-off?
(4) Is Article 10(1) of Council Regulation No 1765/92 to be interpreted as meaning that the
compensatory payments in question are to be paid over immediately the intervention agency has
concluded the procedure with regard to the beneficiarys application, or is it permissible to delay the
payment for an investigation as to whether the state has outstanding claims against the beneficiary
in respect of which it wishes to effect a set-off, provided always that the payment is effected at the
latest by 31 December of the relevant support year?

The first two questions

31.
By its first two questions the national court seeks essentially to ascertain whether Community law
precludes a member state from effecting set-off between an amount due to a beneficiary of aid payable
under Community legislation and an outstanding debt to that member state. It also seeks to ascertain
whether the answer to be given to that question is affected by (i) the capacity in which the member state
grants the aid provided for by the regulation and the requirement, imposed by national law, that claims
must be reciprocal between debtor and creditor, (ii) the practice generally followed by the member state
regarding set-off and (iii) the legal basis of the debt to the state against which the amount payable is set
off.
32.
The plaintiffs in the main proceedings maintain that it is incompatible with the general principles of
Community law to effect set-off between aid paid in the context of common market organisations and
debts to a member state. The objective of those common market organisations would be distorted by
such a procedure. Rather than enforced set-off of the kind in issue in the main proceedings, the Danish
authorities could have used, for example, attachment proceedings.
33.
Conversely, the Directorate and the Danish government contend that, in the absence of Community
rules on the subject, the national rules on set-off may continue to be applied provided that they are not
discriminatory and do not undermine the common market organisation involved.
34.
It must first be observed that, under the scheme for financing the common agricultural policy, the
Community grants aid in the context of a division of powers with the member states. The sums
corresponding to the aid are placed at the disposal of the member states, which must ensure that they are
properly managed (see Unione Nazionali di Produktori di Olive (Unaprol) v Azienda di Stato per gli
Interventi Sud Mercato Agricolo (AIMA) Case C-186/93 [1994] ECR I-3615 (para 27)).
539
35.
As it now stands, Community law contains no general rules on the right of national authorities to effect
set-off between amounts paid under Community legislation and outstanding debts to a member state.
36.
On the other hand, the court has held, in the case of an insolvent trader to whom funds had been
wrongly paid, that set-off may in fact constitute the only practicable way open to the authorities to recover
such sums (see the judgment in DEKA Getreideprodukte GmbH & Co KG (in liq) v EEC Case 250/78
[1983] ECR 421 (para 14)).
37.
However, national rules would be incompatible with the Treaty and with the rules on the common
organisation of markets if they authorised practices liable to interfere with the functioning of the machinery
employed by those organisations in order to achieve their ends (see, to that effect, Irish Creamery Milk
Suppliers Association v Ireland Joined cases 36/80 and 71/80 [1981] ECR 735 (para 15) and Association
comit conomique agricole rgional fruits et lgumes de Bretagne (Cerafel) v Le Campion Case 218/85
[1986] ECR 3513 (para 13)).
38.
In this case, it must be observed that the aim of the regulation isbecause of the suppression,
following reform of the common agricultural policy, of income support for farmers under the price policy
to guarantee direct income support in the form of compensatory payments. For the reasons set out in
paras 47 to 55 of Advocate General Fennellys opinion, above, national rules of the kind at issue in the
present proceedings, which are designed to ensure more effective recovery of debts payable to public
authorities, do not appear, a priori, to be such as to undermine the effectiveness of Community law.
39.
In any event, as Advocate General Fennelly has observed in para 56 of his opinion, above, if it were
found that permitting set-off resulted in widespread differences in treatment as between the different
national legal systems such as to compromise the equal treatment of producers in different member
states, it would be for the Community legislature to adopt the provisions needed to remedy such
disparities.
40.
It is nevertheless necessary to verify whether the circumstances mentioned by the national court in its
second question are such as to affect the conformity with Community law of the national practice
regarding set-off.
41.
First, with respect to the capacity in which the member state grants aid under the regulation and the
requirement imposed by national law that claims must be reciprocal between debtor and creditor, it must
be borne in mind that set-off is not expressly regulated by Community law.
42.
In those circumstances it is, in principle, for each member state to define the conditions under which
its national authorities may resort to set-off and to deal with all incidental issues.
43.
Second, in view of the practice generally followed by their member state in regard to set-off, the
plaintiffs in the main proceedings maintain that the manner in which direct agricultural aid granted under
Community law is set off against debts to the state is discriminatory because equivalent national aid is
either excluded from set-off or is available for set-off only up to 20% of its amount, whereas in a case like
that with which the main proceedings are concerned the Directorate could effect total set-off.
44.
Whilst emphasising that it is for the national court to compare the legal rules on Community
agricultural aid and those governing national subsidies, the Danish government considers that those two
types of aid are not equivalent and that no discrimination can therefore arise.
540
45.
It must be noted here, first, that, in the context of art 177 of the Treaty, the court has no jurisdiction to
rule on the compatibility of a national provision with Community law.
46.
It is therefore for the national court to decide whether the national rules applicable to set-off are
discriminatory, in that they are applied differently to amounts payable under Community law and to
amounts paid under national law alone.
47.
However, the Court of Justice may provide the national court with certain information relating to the
interpretation of Community law, for the purposes of the assessment which it is to undertake.
48.
It is true that, in the specific context of the common agricultural policy, the court has already made it
clear that, in so far as Community law, including its general principles, does not include common rules on
the point at issue, the national authorities, when implementing Community regulations, act in accordance
with the procedural and substantive rules of their own national law (see the judgment in Deutsche
Milchkontor GmbH v Germany Joined cases 205-215/82 [1983] ECR 2633 (para 17)).
49.
Nevertheless, in the first place, that rule must be reconciled with the need to apply Community law
uniformly so as to avoid unequal treatment of producers and traders (see Deutsche Milchkontor [1983]
ECR 2633 (para 17)).
50.
Second, the national rules on set-off must not make set-off between an amount due to the beneficiary
of aid under a Community measure and an outstanding debt to the member state subject to less
favourable conditions or procedures than those applicable to set-off between claims of purely domestic
origin (see, mutatis mutandis, the judgments in Rewe-Zentralfinanz eG v Landwirtschaftskammer fr das
Saarland Case 33/76 [1976] ECR 1989 (para 5), Comet BV v Produktschap voor Siergewassen Case
45/76 [1976] ECR 2043 (para 13) and Firma Wilhelm Fromme v Bundesanstalt fr Landwirtschaftliche
Marktordnung Case 54/81 [1982] ECR 1449 (para 6)).
51.
Indeed, the principle of non-discrimination implies that the obligations imposed by national legislation
on beneficiaries of Community aid must not be more stringent than those imposed on the beneficiaries of
similar advantages or aid based on national lawalways provided that the two groups of recipients are in
comparable situations and therefore that that different treatment is not objectively justifiable (see Firma
Wilhelm Fromme [1982] ECR 1449 (para 7)).
52.
Finally, with regard to the legal basis of the debt to the state involved in the set-off, Mr Jensen claims
that the amount of area aid paid by the Danish state to farmers, and reimbursed to it by the Community,
derives in reality from the Communitys own resources. It is therefore entirely logical for the Community to
prohibit a member state from using those resources to recover its own debts rather than to attain the
objectives pursued by the Community area aid.
53.
However, it must be noted here that neither the legal basis of the debt to the state nor the fact that the
amount set off against it may derive from the Communitys own resources in any way affects the member
states right to effect set-off between fiscal debts payable to it and compensatory payments due under
Community law.
54.
The answer to the first two questions must therefore be that Community law does not preclude a
member state from effecting set-off between an amount due to a beneficiary of aid payable under
Community legislation and an outstanding debt to that member state. The position would be different only
if that practice were to interfere with the proper functioning of the common 541 organisation of the
agricultural markets. In that regard, the capacity in which the member state grants aid under the
regulation, the fact that the rules of that member state on set-off require, for set-off to be available,
reciprocity of debts as between debtor and creditor, the practice generally followed by the member state
regarding set-off and the legal basis of the debt to the state involved in the set-off are of no importance,
provided that the national authorities ensure that the effectiveness of Community law is not in any way
undermined and that economic operators enjoy equal treatment. It is for the national court to determine
whether that is the case.

The third question

55.
By its third question, the national court seeks essentially to ascertain whether, on a proper construction
of art 15(3) of the regulation, member states are precluded from requiring a national intervention agency
to effect set-off with debts payable to the state in the case of a beneficiary of compensatory payments.
56.
The plaintiffs in the main proceedings and the Commission submit, on the basis of the objective of the
regulation, which is to contribute directly to maintaining the income of farmers and to offset losses
resulting from the progressive reduction of price guarantees and export refunds, that set-off is clearly
prohibited by that provision.
57.
 The Danish government and the Directorate, for their part, consider that the wording of art 15(3) of the
regulation indicates merely that the aid must not be subject to deduction of taxes or similar levies which
would reduce its overall amount.
58.
First, it must be observed here that, for the purposes of the main proceedings, it is unnecessary to
determine whether art 15(3) of the regulation precludes the national authorities from claiming payments
from beneficiaries of Community aid to cover administrative costs relating to their applications. The type
of case with which the national court is concerned involves set-off between debts to the state for which
such set-off would normally be available and amounts paid under Community law.
59.
Next, it must be noted that it is expressly stated in the second recital in the preamble to the regulation
that the object of the compensatory payments is to compensate the loss of income caused by the
reduction of the institutional prices as part of a new support system for the producers of certain arable
crops as a result of reform of the common agricultural policy.
60.
It is true that, under art 15(3) of the regulation, the payments referred to therein are to be paid over to
the beneficiaries in their entirety. However, as the Advocate General points out in para 39 of his opinion,
the wording of that provision contains no evidence that the Community legislature intended to limit the
widely varying debt-recovery methods which exist in national law.
61.
Set-off between compensatory payments made under the regulation and outstanding debts payable to
a member state does not have the effect of reducing the amount of the aid.
62.
Consequently, subject to the provisos mentioned in the answer to the first two questions, set-off is not
contrary to art 15(3) of the regulation.
63.
The answer to the third question must therefore be that, on a proper construction of art 15(3) of the
regulation, member states are not precluded from requiring a national intervention agency to effect set-off
with debts payable to the state in the case of a beneficiary of compensatory payments.
542

The fourth question

64.
By its fourth question, the national court seeks essentially to ascertain whether, on a proper
construction of art 10(1) of the regulation, the compensatory payments to which it refers must be paid as
soon as the intervention agency has completed its examination of the application from the beneficiary or
whether payment may be deferred until a check has been made as to whether debts are payable by him
to the state against which set-off may be available, provided that the payment is made no later than 31
December of the year concerned.
65.
Mr Jensen claims that, although the national authorities are entitled to lay down appropriate
administrative procedures for processing applications for Community area aid, such procedures must
conform to the fundamental principles of Community law and enable area aid to be paid in the period
between 16 October and 31 December in accordance with art 10(1) of the regulation. Any delay in paying
Community aid might result in discrimination between farmers in the same member state and between
farmers in different member states.
66.
The Directorate, on the other hand, considers that, in the absence of Community rules concerning the
right to check whether the beneficiary of the aid is in any way indebted to the state or to the Community, it
is the responsibility of each member state to lay down its own rules on the matter, observing the general
principles of Community law.
67.
First of all, as the Directorate was right to observe, although the member states remain entitled to
apply their national rules on set-off, in doing so they must nevertheless observe Community law without
undermining its effectiveness or the proper functioning of the common organisation of the markets.
68.
Next, it is expressly stated in art 10(1) of the regulation that the compensatory payments to producers
are to be made between 16 October and 31 December next following the harvest. Therefore, although the
national authorities should make those payments as soon as possible, they are not legally required to do
so before 31 December.
69.
Finally, Community law does not preclude a member state from verifying expeditiously whether it has
claims against the beneficiary of aid which might be available for set-off, provided that set-off is effected in
conformity with the general principles of Community law laid down in the answer to the first two questions.
70.
Accordingly, the answer to the fourth question must be that, on a proper construction of art 10(1) of the
regulation, payment of the compensatory payments to which it refers may be deferred until it has been
verified whether the state has any claim against the beneficiary thereof for which set-off may be available,
provided that the payment is made no later than 31 December in the year in question.

Costs
71.
The costs incurred by the Danish, Greek, French, Irish, Finnish, Swedish and UK governments and by
the European Commission, which have submitted observations to the Court of Justice, are not
recoverable. Since these proceedings are, for the parties to the main proceedings, a step in the
proceedings pending before the national court, the decision on costs is a matter for that court.
543
On those grounds, the Court of Justice, in answer to the questions referred to it by the stre Landsret
by order of 10 April 1995, hereby rules: (1) Community law does not preclude a member state from
effecting set-off between an amount due to a beneficiary of aid payable under Community legislation and
an outstanding debt to that member state. The position would be different only if that practice were to
interfere with the proper functioning of the common organisation of the agricultural markets. In that
regard, the capacity in which the member state grants aid under Council Regulation (EEC) 1765/92
establishing a support system for producers of certain arable crops, the fact that the rules of that member
state on set-off require, for set-off to be available, reciprocity of debts as between debtor and creditor, the
practice generally followed by the member state regarding set-off and the legal basis of the debt to the
state involved in the set-off are of no importance, provided that the national authorities ensure that the
effectiveness of Community law is not in any way undermined and that economic operators enjoy equal
treatment. It is for the national court to determine whether that is the case. (2) On a proper construction of
art 15(3) of Regulation 1765/92, member states are not precluded from requiring a national intervention
agency to effect set-off with debts payable to the state in the case of a beneficiary of compensatory
payments. (3) On a proper construction of art 10(1) of Regulation 1765/92, payment of the compensatory
payments to which it refers may be deferred until it has been verified whether the state has any claim
against the beneficiary thereof for which set-off may be available, provided that the payment is made no
later than 31 December in the year in question.

544

[1998] All ER (EC) 545

Svenska Journalistfrbundet (supported by Sweden and others) v EU

Council (supported by France and another)
(Case T-174/95)

EUROPEAN COMMUNITY; Other European Community: ADMINISTRATION OF JUSTICE; Courts

COURT OF FIRST INSTANCE OF THE EUROPEAN COMMUNITIES (FOURTH CHAMBER,
EXTENDED COMPOSITION)
JUDGES LENAERTS (PRESIDENT), LINDH, AZIZI, COOKE AND JAEGER
17 SEPTEMBER 1997, 17 JUNE 1998
European Community Action for annulment Admissibility Applicant requesting access to EU Council
documents concerning European Police Office Application partially refused Applicant seeking
annulment of refusal Whether rules on access to documents applying to Justice and Home Affairs
matters Whether court having jurisdiction to review questions of access to such documents Council
Decision (EC) 93/731 Treaty on European Union, Title VI EC Treaty, art 173.
European Community Action for annulment Abuse of procedure Partial refusal of application for
access to Council documents concerning Justice and Home affairs matters Applicant seeking
annulment of refusal Applicant publishing Councils defence on internet Whether publication
prejudicial to proper course of procedure.

In order to test the way in which the Swedish authorities applied their citizens right of access to
information of documents relating to European Union activities, the applicant, a journalists union,
contacted a number of national authorities seeking access to 20 documents relating to the setting up of
the European Police Office. Although access was granted to 18 of the 20 documents, certain passages
had been deleted and the applicant therefore applied to the EU Council for access to the same 20
documents. The Council allowed access to only four of the documents refusing access to the other 16
documents pursuant to art 4 of Council Decision (EC) 93/731 on public access to Council documents, on
the grounds that their release could be harmful to the public interest (public security) and that they were
subject to the principle of confidentiality. The applicant subsequently brought proceedings against the EU
Council seeking an annulment of the decision refusing access, contending principally that the Council had
failed to give sufficient reasons for its decision. Shortly after proceedings were commenced, the applicant
published certain material documents, including an edited version Councils defence and the names and
contact details of the Councils agents in the case, on the internet. Before the Court of First Instance of
the European Communities, which ultimately annulled the contested decision, issues were raised as to: (i)
whether the applicant had locus standi to bring the action; (ii) whether the Court of First Instance had
jurisdiction to entertain the action in view of the fact that the contested decision concerned the
arrangements for access to documents adopted on the basis of Title VI of the Treaty on European Union
(Justice and Home affairs) which fell outside the scope of the EC Treaty and were matters for inter-
governmental co-operation; and (iii) whether the applicants conduct in publishing the Councils defence
on the internet was prejudicial to the proper course of the procedure.

545

Held (1) The addressee of a contested decision and was not obliged to prove that the decision was of
direct and individual concern to it; he need only prove that he had an interest in the annulment of the
decision. Moreover the objective of Decision 93/731 was to give effect to the principle of the largest
possible access for citizens to information with a view to strengthening the democratic character of the
institutions and the trust of the public in the administration and, accordingly, it did not require members of
the public to give reasons for seeking access to requested documents. It followed that a person who was
refused access to a document or to part of a document had, by virtue of that very fact, established an
interest in the annulment of the decision. In the instant case, the contested decision denied access to 16
of the 20 documents requested and, as such, the applicant had proved an interest in the annulment of
that decision (see p 555 j, p 556 b to c, post); Interporc Im- und Export GmbH v European Commission
Case T-124/96 (1998) CFI Transcript, 6 February applied.
(2) In the absence of any provision to the contrary in Decision 93/731 itself, its provisions applied to all
Council documents, irrespective of their content. The fact that the court had, by virtue of art L of the Treaty
on European Union, no jurisdiction to review the legality of measures adopted under Title VI did not curtail
its jurisdiction in the matter of public access to those measures. The assessment of the legality of the
contested decision was based upon its jurisdiction to review the legality of Council decisions taken under
Decision 93/731, on the basis of art 173 of the EC Treaty, and did not in any way bear on the inter-
governmental co-operation in the spheres of Justice and Home Affairs as such. Moreover, the fact that the
documents related to Title VI was only relevant in so far as their contents might possibly come within the
scope of one or more of the exceptions provided for in Decision 93/731 and was thus relevant only to the
examination of the substantive lawfulness of the decision taken by the Council and not to the admissibility
of the application as such (see p 557 h to p 558 c, post).
(3) Under the rules which governed procedure before the Court of First Instance, parties were entitled
to protection against the misuse of pleadings and evidence. Thus, in accordance with the third sub-
paragraph of art 15(3) of the Instructions to the Registrar, no third party, private or public, could have
access to the case file or to the procedural documents without the express authorisation of the President,
after the parties had been heard; and, in accordance with art 116(2) of the Rules of Procedure, the
President could exclude secret or confidential documents from those furnished to an intervener. Those
provisions reflected a general principle in the due administration of justice according to which parties had
the right to defend their interests free from all external influences and particularly from the influence of
members of the public. It followed that a party who was granted access to the procedural documents of
other parties was only entitled to use those documents for the purpose of pursuing his own case and for
no other purpose, including that of inciting criticism on the part of the public in relation to arguments
raised by other parties. In the instant case, it was clear that the publication on the internet by the applicant
of the defence together with an invitation to the public to submit comments to the Council, had as its
purpose to bring pressure to bear upon the Council and to provoke criticism of its agents in the
performance of their duties. Accordingly, the applicants actions involved an abuse of procedure which
would be taken into account in awarding costs (see p 565 h to p 566 c, post).

546

Notes
For the EC Treaty art 173 (as amended by art G.53 of the Treaty on European Union), see 50 Halsburys
Statutes (4th edn), Current Service, 97.

Cases cited
Campus Oil Ltd v Minister for Industry and Energy Case 72/83 [1984] ECR 2727.
Carvel v EU Council Case T-194/94 [1996] All ER (EC) 53, [1995] ECR II-2765, CFI.
Comit International de la Rayonne et des Fibres Synthtiques (CIRFS) v EC Commission Case C-
313/90 [1993] ECR I-1125.
Association des Acieries Europennes Independantes (EISA) v European Commission Case T-239/94
[1997] ECR II-1839.
Eugnio Branco Lda v European Commission Case T-85/94 [1995] ECR II-45.
France v EC Commission Case C-59/91 [1992] ECR I-525.
Fritz Werner Industrie-Ausrstungen GmbH v Germany Case C-70/94 [1995] ECR I-3189.
Interporc Im- und Export GmbH v European Commission Case T-124/96 (1998) Transcript, 6 February,
CFI.
Netherlands v EU Council Case C-58/94 [1996] ECR I-2169.
Socit franaise des Biscuits Delacre v EC Commission Case C-350/88 [1990] ECR I-395.
van der Wal v European Commission Case T-83/96 [1998] All ER (EC) 289, CFI.
WWF UK v European Commission Case T-105/95 [1997] All ER (EC) 300, [1997] ECR II-313, CFI.
Application
By application lodged on 22 September 1995 at the registry of the Court of First Instance of the European
Communities, Svenska Journalistfrbundet, an association governed by Swedish law, applied for the
annulment of the EU Councils decision of 6 July 1995 refusing access to certain documents concerning
the European Police Office, requested under Council Decision (EC) 93/731 on public access to Council
documents (OJ 1993 L340 p 43). By order of 23 April 1996, the Kingdom of Denmark, the Kingdom of the
Netherlands and the Kingdom of Sweden were granted leave to intervene in support of the applicant, and
the French Republic and the United Kingdom were granted leave to intervene in support of the defendant.
Svenska Journalistfrbundet was represented by O W Brouwer, of the Amsterdam Bar, and F P Louis, of
the Brussels Bar, assisted by D Curtin, Professor at the University of Utrecht, with an address for service
in Luxembourg at the Chambers of Loesch and Wolter, 11 Rue Goethe. The Kingdom of Sweden was
represented by L Nordling, Director-General of the Legal Service of the Ministry of Foreign Affairs, acting
as agent; the Kingdom of Denmark was represented by P Biering, Head of Department in the Ministry of
Foreign Affairs, and L Miklsen, Ambassador, acting as agents, with an address for service in
Luxembourg at the Danish Embassy, 4 Boulevard Royal; and the Kingdom of the Netherlands was
represented by M Fierstra and J Steven van den Oosterkamp, Legal Advisers, acting as agents, with an
address for service in Luxembourg at the Embassy of the Netherlands, 5 Rue C M Spoo. The EU Council
was represented by G Maganza and D Canga Fano, Legal Advisers, acting as agents, with an address for
service in Luxembourg at the office of A Morbilli, Manager of the Legal Affairs Directorate of the European
Investment Bank, 100 Boulevard Konrad Adenauer. The French Republic was represented by C de
Salins, Assistant Director in the 547 Legal Department of the Ministry of Foreign Affairs, and D Wibaux,
Secretary for Foreign Affairs in the same ministry, acting as agents, with an address for service in
Luxembourg at the French Embassy, 8B Boulevard Joseph II; and the United Kingdom was represented
by J Collins, of the Treasury Solicitors Department, acting as agent, with an address for service in
Luxembourg at the British Embassy, 14 Boulevard Roosevelt. The language of the case was English. The
facts are set out in the judgment of the court.

17 June 1998.

The COURT OF FIRST INSTANCE (Fourth Chamber, Extended Composition)

delivered the following judgment.
1.
In the Final Act of the Treaty on European Union (the TEU) (Maastricht, 7 February 1992; TS 12
(1994); Cmnd 2485), the member states incorporated a Declaration (No 17) on the right of access to
information, in the following terms:
The Conference considers that transparency of the decision-making process strengthens the
democratic nature of the institutions and the publics confidence in the administration. The
Conference accordingly recommends that the Commission submit to the Council no later than 1993
a report on measures designed to improve public access to the information available to the
institutions.
2.
On 8 June 1993 the European Commission published Communication 93/C156/05 on public access to
the institutions documents (OJ 1993 C156 p 5), which had been submitted to the EU Council, the
Parliament and the Economic and Social Committee on 5 May 1993. On 17 June 1993 it published
Communication 93/C166/04 on openness in the Community (OJ 1993 C166 p 4), which had also been
submitted to the Council, the Parliament and the Economic and Social Committee on 2 June 1993.
3.
On 6 December 1993 the Council and the Commission approved a code of conduct concerning public
access to Council and Commission documents (OJ 1993 L340 p 41), and each undertook to take steps to
implement the principles thereby laid down before 1 January 1994.
4.
In order to put that undertaking into effect, the Council adopted on 20 December 1993 Council
Decision (EC) 93/731 on public access to Council documents (OJ 1993 L340 p 43), the aim of which was
to implement the principles established by the code of conduct. It adopted that decision on the basis of art
151(3) of the EC Treaty, which states that The Council shall adopt its Rules of Procedure.
5.
Article 1 of Decision 93/731 provides:
(1) The public shall have access to Council documents under the conditions laid down in this
Decision. (2) Council document means any written text, whatever its medium, containing existing
data and held by the Council, subject to Article 2(2).
6.
Article 2(2) provides that applications for documents the author of which is not the Council must be
sent directly to the author.
7.
Article 4(1) of Decision 93/731 provides:
Access to a Council document shall not be granted where its disclosure could undermine: the
protection of the public interest (public security, international relations, monetary stability, court
 proceedings, inspections and investigations), the protection of the individual and of privacy, the
548 protection of commercial and industrial secrecy, the protection of the Communitys financial
interests, the protection of confidentiality as requested by the natural or legal person who
supplied any of the information contained in the document or as required by the legislation of the
member State which supplied any of that information.
8.
Article 4(2) adds that Access to a Council document may be refused in order to protect the
confidentiality of the Councils proceedings.
9.
Articles 2(1), 3, 5 and 6 of Decision 93/731 set out in particular the procedure for submitting
applications for access to documents and the procedure to be followed by the Council when replying to
such applications.
10.
Article 7 provides:
(1) The applicant shall be informed in writing within a month by the relevant departments of the
General Secretariat either that his application has been approved or that the intention is to reject it.
In the latter case, the applicant shall also be informed of the reasons for this intention and that he
has one month to make a confirmatory application for that position to be reconsidered, failing which
he will be deemed to have withdrawn his original application. (2) Failure to reply to an application
within a month of submission shall be equivalent to a refusal, except where the applicant makes a
confirmatory application, as referred to above, within the following month. (3) Any decision to reject
a confirmatory application, which shall be taken within a month of submission of such application,
shall state the grounds on which it is based. The applicant shall be notified of the decision in writing
as soon as possible and at the same time informed of the content of Articles 138e and 173 of the
Treaty establishing the European Community, relating respectively to the conditions for referral to
the Ombudsman by natural persons and review by the Court of Justice of the legality of Council
acts. (4) Failure to reply within a month of submission of the confirmatory application shall be
equivalent to a refusal.

THE FACTS
11.
Following Swedens accession to the European Union on 1 January 1995, the applicant decided to test
the way in which the Swedish authorities applied Swedish citizens right of access to information in
respect of documents relating to EU activities. For that purpose it contacted 46 Swedish authorities,
among whom were the Swedish Ministry of Justice and the National Police Authority (the
Rikspolisstyrelsen), seeking access to a number of Council documents relating to the setting up of the
European Police Office (Europol), including eight documents held by the National Police Authority and 12
held by the Ministry of Justice. In response to its requests the applicant was granted access to 18 of the
20 documents requested. It was refused access by the Ministry of Justice to two documents on the
ground that they concerned the negotiating positions of the Netherlands and German governments.
Furthermore, certain passages in the documents to which access was granted had been deleted. In some
documents it was difficult to ascertain whether passages had been deleted or not.
12.
On 2 May 1995 the applicant also applied to the Council for access to the same 20 documents.
13.
By letter dated 1 June 1995, the Councils General Secretariat allowed access to two documents only,
those being documents which contained communications by the future French Presidency of its priorities
in the field of 549 asylum and immigration and in the field of justice. Access to the other 18 documents
was refused on the ground that documents 1 to 15 and 18 to 20 are subject to the principle of
confidentiality as laid down in Article 4(1) of Decision 93/731.
14.
On 8 June 1995 the applicant submitted a confirmatory application to the Council in order to obtain re-
examination of the decision refusing access.
15.
The competent department of the Councils General Secretariat, together with the Councils Legal
Service, then prepared a note for the attention of the Information Working Party of the Permanent
Representatives Committee (Coreper) and the Council. A draft reply, together with the exchange of
correspondence that had taken place previously between the applicant and the General Secretariat, was
distributed with a note dated 15 May 1995 prepared by Mr Elsen, Director-General of the Councils
Justice and Home Affairs Directorate (DG H), when the first application was being examined. That note
provided a brief summary of the contents of the documents and a preliminary assessment as to whether
they could be released. It was communicated to the applicant for the first time in the course of the present
proceedings as an annex to the Councils defence. On 3 July 1995 the Information Working Party decided
to release two other documents but to refuse access to the remaining 16. At a meeting on 5 July 1995
Coreper approved the terms of the draft reply proposed by the Working Party.
16.
The Council points out that all the documents concerned were at the disposal of the members of the
Council and that copies of the documents were also available for examination at the Information Working
Party meeting of 3 July.
17.
After the Coreper meeting, the Council replied to the confirmatory application by a letter dated 6 July
1995 (the contested decision), in which it agreed to grant access to two other documents but rejected the
application for the remaining 16 documents.
18.
It explained that:
[i]n the Councils opinion access to those documents cannot be granted because their release
could be harmful to the public interest (public security) and because they relate to the Councils
proceedings, including the positions taken by the members of the Council, and are therefore
covered by the duty of confidentiality. Lastly, I would like to draw your attention to the provisions of
Articles 138e and 173 of the EC Treaty concerning, respectively, the conditions governing the
lodging of a complaint with the Ombudsman and the institution of proceedings before the Court of
Justice by a natural person against acts of the Council.

PROCEDURE
19.
By application lodged at the registry of the Court of First Instance of the European Communities on 22
September 1995 the applicant instituted this action.
20.
By a letter lodged on 9 February 1996, the European Parliament sought leave to intervene in the case
in support of the applicant. It subsequently withdrew its intervention.
21.
By order of the President of the Fourth Chamber of the Court of First Instance of 23 April 1996, the
Kingdom of Denmark, the Kingdom of the Netherlands and the Kingdom of Sweden were granted leave to
intervene in 550 support of the applicant, and the French Republic and the United Kingdom were granted
leave to intervene in support of the defendant.
22.
By letter received on 3 April 1996 the Council drew the attention of the Court of First Instance to the
fact that certain material documents, including the Councils defence, had been published on the Internet.
The Council considered that the applicants conduct was prejudicial to the proper course of the procedure.
It requested the court to take appropriate measures in order to avoid further such action on the part of the
applicant.
23.
The court decided to treat this incident as a preliminary issue within the meaning of art 114(1) of the
Rules of Procedure, and accordingly invited the parties to submit observations on the matter. The written
procedure was suspended in the meantime. Observations were received from the applicant and from the
Danish, French, Netherlands, Swedish and UK governments.
24.
In the light of those observations the court decided that the proceedings would be resumed, without
prejudice to the consequences it would attach to that preliminary issue (see paras 135 to 139, below).
25.
By decision of 4 June 1996, the court referred the case to the Fourth Chamber, Extended
Composition. It did not accede to a request by the Council of 20 June 1996 that the case be referred to
the court sitting in plenary session.
26.
The written procedure was concluded on 7 April 1997.

FORMS OF ORDER SOUGHT BY THE PARTIES

27.
The applicant, supported by the Kingdom of Denmark and the Kingdom of the Netherlands, requests
the court to: annul the contested decision; order the Council to pay the costs.
28.
The Kingdom of Sweden requests the court to annul the contested decision.
29.
The Council requests the court to: declare the application inadmissible in its entirety; alternatively,
declare the application inadmissible in so far as it relates to documents which have already been received
by the applicant and do not contain deleted passages; in the further alternative, reject it as unfounded;
order the applicant to pay the costs.
30.
The French Republic requests the court to: dismiss the application; order the applicant to pay the
costs.
31.
The United Kingdom requests the court to dismiss the application as inadmissible or, in the alternative,
as unfounded.

ADMISSIBILITY
32.
The Council claims that the application is inadmissible on several grounds, relating to the identity of
the applicant, non-compliance with the time limit for bringing an action, the applicants lack of interest in
bringing the action and the courts lack of jurisdiction. Each of those grounds will be examined in turn.

The identity of the applicant

33.
Svenska Journalistfrbundet is the Swedish Journalists Union. It owns and publishes a newspaper
entitled Tidningen Journalisten. The application is headed Svenska Journalistfrbundets tidning and
Tidningen Journalisten. The application states that the applicant is the magazine of the Swedish
Journalists Union, but the link between the two entities is not clearly explained. During the written
procedure Tidningen Journalisten was therefore designated as the applicant.
551
Arguments of the parties
34.
In reply to a written question from the court, the applicants lawyers indicated by fax message of 4
August 1997 that the application should be regarded as having been lodged by the Swedish Journalists
Union as the proprietor of the magazine, since it alone of the two entities had capacity to sue under
Swedish law.
35.
At the hearing they added that any distinction between the Swedish Journalists Union and Tidningen
Journalisten was artificial. The application and confirmatory application sent to the Council had been
presented on headed paper of Svenska Journalistfrbundet and Tidningen Journalisten and the Council
replied to Svenska Journalistfrbundets Tidning. Svenska Journalistfrbundet was thus a party to the
case from the outset.
36.
The Netherlands government considers that it would be too formalistic to consider that an action
instituted by an independent division of a legal person could not be attributed to that legal person, given
that it is now clear that adequate proof of authority was produced when the action was instituted and the
interests of the parties to the proceedings have not been injured in any way.
37.
In a letter of 9 September 1997, the Council contends that in the light of the replies of the applicants
lawyers Tidningen Journalisten, which it had regarded as the applicant in the case, had no capacity to sue
under Swedish law.
38.
It further contends that even if the Swedish Journalists Union could be substituted for Tidningen
Journalisten, the former could not be regarded as the addressee of the Councils reply of 6 July 1995, nor
as directly and individually concerned by that decision.
39.
It therefore asks the court to dismiss the application as inadmissible.
Findings of the court
40.
The first page of the application refers to both Tidningen Journalisten and Svenska
Journalistfrbundets tidning.
41.
The proof of authority granted to the applicants lawyers as required by art 44(5)(b) of the Rules of
Procedure was signed on behalf of the Swedish Journalists Union by Lennart Lund, Editor in Chief of the
magazine Tidningen Journalisten. In that regard, the applicant has lodged, as an annex to its fax
message of 4 August 1997 (see para 34, above), a certificate confirming that the Swedish Journalists
Union had instructed Lennart Lund to bring the present application before the court.
42.
In those circumstances it is clear that the application has, in reality, been brought by the Swedish
Journalists Union as proprietor of Tidningen Journalisten.
43.
The Swedish Journalists Union being a legal person entitled to sue under Swedish law, the Council
cannot object to the admissibility of the application on this basis.
44.
Moreover, given that the Council had addressed the two negative replies of 1 June 1995 and 6 July
1995 to Mr Christoph Andersson, Svenska Journalistfrbundets tidning, it cannot at this stage argue that
the Swedish Journalists Union was not the addressee of the contested decision.
552

The time limit for bringing the action

Arguments of the parties
45.
 The Council questions whether the action was brought within the prescribed time limit. It maintains that
the applicant received the contested decision on 10 July 1995. It then had two months from that date to
bring an action for its annulment.
46.
The Council points out that art 1 of Annex II to the courts Rules of Procedure, in the version then
applicable, provided that procedural time limits were to be extended for parties not habitually resident in
the Grand Duchy of Luxembourg by the following: for the Kingdom of Belgium: two days, for the
Federal Republic of Germany, the European territory of the French Republic and the European territory of
the Kingdom of the Netherlands: six days, for the European territory of the Kingdom of Denmark, for the
Hellenic Republic, for Ireland, for the Italian Republic, for the Kingdom of Spain, for the Portuguese
Republic (with the exception of the Azores and Madeira) and for the United Kingdom: ten days, for
other European countries and territories: two weeks.
47.
The Council, supported by the French government, doubts that the rule applicable to non-member
states should also apply to member states of the European Union and considers that the applicant should
have brought its action in compliance with a time limit extended on account of distance by ten days, in
order to avoid any discrimination between applicants from countries that are further away from
Luxembourg than Sweden, which are entitled only to a ten-day extension.
48.
The applicant relies on the actual terms of art 1 of Annex II in the version reproduced above, and
considers that they do not support the Councils contention. There is no reference to member states or
non-member states. In the absence of any specific extension for Sweden, that country was entitled to the
extension of two weeks applicable to all the European states not specifically mentioned. The Councils
argument concerning discrimination does not carry conviction, since numerous places in Belgium are
further away from Luxembourg than certain places in the Netherlands, yet all inhabitants of Belgium are
entitled to a two-day extension while all inhabitants of the Netherlands are entitled to a six-day extension.
Only the applicants interpretation satisfies the requirement of legal certainty.
49.
The Swedish and Netherlands governments support that interpretation. At the hearing the Swedish
governments agent pointed out that it was formerly entitled to an extension of two weeks.
Findings of the court
50.
It is settled law that the Community rules governing procedural time limits must be strictly observed
both in the interest of legal certainty and in order to avoid any discrimination or arbitrary treatment in the
administration of justice (see France v EC Commission Case C-59/91 [1992] ECR I-525 (para 8)).
51.
The wording of art 1 of Annex II to the Rules of Procedure, in the version in force when the application
was brought, does not support the submission that the extension for distance applicable in the case of
Sweden was ten days and not two weeks. In fact, the ten-day extension applied only to certain
designated countries, of which Sweden was not one. The extension of two weeks thus applied to all
European countries and territories for which a shorter period was not laid down, including Sweden.
553
52.
It follows that the action was commenced within time.

The applicants interest in seeking annulment

Arguments of the parties
53.
The Council also doubts that the application is admissible inasmuch as it concerns documents that the
applicant had already received from the Swedish authorities, at least to the extent that those documents
do not contain deleted passages. The Council was not informed that the purpose of the applicants
request was to identify any passages in those documents which had been deleted. The applicants
interest is general and political in nature, its intention being to ensure that the Council gives proper effect
to its own code of conduct and Decision 93/731.
54.
In the circumstances, although the Council is conscious of the fact that the applicant is the addressee
of the contested decision, it questions whether the applicant is really affected by that decision within the
meaning of art 173 of the EC Treaty. That article does not allow individual actions in the public interest,
but only permits individuals to challenge acts which concern them in a way in which they do not concern
other individuals.
55.
In this case the applicant cannot derive any benefit from obtaining access to documents which are
already in its possession. Its insufficient interest in the outcome of the proceedings constitutes an abuse
of procedure.
56.
Supported by the French government, the Council further contends that the release of the documents
in question by the Swedish authorities to the applicant constitutes a breach of Community law, since no
decision had been taken to authorise such a disclosure. It is contrary to the system of legal remedies
provided for by Community law to take advantage of a breach of Community law and then to ask the court
to annul a decision whose effects have been circumvented as a consequence of such a breach. The fact
that the documents in question were brought into the public domain following an act contrary to
Community law should therefore preclude the applicant from bringing an action in this case.
57.
The applicant replies that the Council is confusing the rules on the admissibility of actions for the
annulment of decisions brought by their addressees with the rules on the admissibility of actions for the
annulment of regulations brought by certain individuals. Addressees must show that they have an interest
in bringing their action but do not have to prove that they are individually concerned.
58.
In this case the applicant considers that it has a sufficient interest in bringing the action and that that
interest is neither political nor general in nature. It points out that Tidningen Journalisten publishes articles
on specific subjects of general interest and on the functioning of public authorities and other matters
concerning the way in which Swedish journalists can go about their job. It therefore has a direct interest in
gaining access to Council documents and, if it is refused access for reasons which demonstrate that the
Council is misapplying the relevant rules, in obtaining the annulment of the decision concerned so as to
ensure that the Council rectifies its approach in the future. The fact that it has received documents from
another source does not therefore mean that it has no interest in bringing the action.
59.
In so far as the Council considers that the documents obtained from the Swedish authorities without its
prior authorisation were obtained unlawfully, the 554 applicant has a further ground for the application to
be held admissible even as regards documents obtained in full from the Swedish authorities. Any use
which the applicant may make of those documents will otherwise be thrown into doubt.
60.
The applicant also rejects the Councils argument that the insufficient interest the applicant has in the
present proceedings makes the application an abuse of procedure. It points out that at the time when it
requested access to the Councils documents it had asked for and obtained from the National Police
Authority only 8 of the 20 documents in question. The other 12 documents were requested from the
Swedish Ministry of Justice on the same day as it sent its request for the 20 documents to the Council.
Furthermore, many of the documents obtained appeared to have deleted passages and the applicant
could not, therefore, be sure that it had received all the documents in full. The Council itself has not
indicated to the court which documents contain deleted passages, although it has asked the court to
declare the application inadmissible to the extent that it concerns documents which the applicant has
obtained and which do not contain deleted passages. The applicant is therefore not in a position to know
which documents do not contain any such passages.
61.
The Swedish government supports the applicants arguments as to admissibility. It does not share the
Councils view that the release of the documents in Sweden constituted a breach of Community law.
There is no implied Community rule based on a common legal tradition whereby only the author of a
document may decide whether a document is to be released or not.
62.
The Netherlands government rejects the Councils argument as regards the applicants lack of interest
in bringing proceedings. It states that it was precisely in the public interest that Decision 93/731 was
adopted. The applicant is not required therefore to show a particular interest in order to be able to rely on
it. The application seeks to preserve the applicants rights as the addressee of the contested decision and
is not an action in the general interest. The applicant has an interest in seeking to prevent the Council
from applying a restrictive policy in regard to requests by the applicant for access to documents in the
future. Moreover, the Councils allegation that the applicant is in possession of documents in breach of
Community law is sufficient to show that the latter does have a legitimate interest.
It goes without saying that the interest recognised by Decision 93/731 relates to legally obtained
access to a document.
63.
The UK government contends that the application is inadmissible because the applicant has no
sufficient interest in the outcome of the proceedings. The application is therefore an abuse of procedure.
None of the reasons given by the applicant is sufficient to give rise to an interest in bringing proceedings
under art 173 of the EC Treaty.
Findings of the court
64.
The applicant is the addressee of the contested decision and, as such, is not obliged to prove that the
decision is of direct and individual concern to it. It need only prove that it has an interest in the annulment
of the decision.
65.
In the case of Commission Decision (ECSC, EC, Euratom) 94/90 on public access to Commission
documents (OJ 1994 L46 p 58), the court has already held that from its overall scheme, it is clear that
Decision 94/90 is intended to apply generally to requests for access to documents, and that, by virtue of
that decision, any person may request access to any unpublished Commission document, and is not
required to give a reason for the request (see Interporc Im- und Export GmbH 555 v European
Commission Case T-124/96 (1998) CFI Transcript, 6 February (para 48)).
66.
 The objective of Decision 93/731 (public access to Council documents) is to give effect to the principle
of the largest possible access for citizens to information with a view to strengthening the democratic
character of the institutions and the trust of the public in the administration. Decision 93/731, like Decision
94/90, does not require that members of the public must put forward reasons for seeking access to
requested documents.
67.
It follows that a person who is refused access to a document or to part of a document has, by virtue of
that very fact, established an interest in the annulment of the decision.
68.
In this case the contested decision denied access to 16 of the 20 documents requested. The applicant
has therefore proved an interest in the annulment of that decision.
69.
The fact that the requested documents were already in the public domain is irrelevant in this
connection.

The jurisdiction of the court

Arguments of the parties
70.
The French government states that the contested decision concerns the arrangements for access to
documents adopted on the basis of Title VI of the TEU. No provision of Title VI governs the conditions of
access to documents adopted on the basis of its provisions. In the absence of an express provision,
Decision 93/731, which was adopted on the basis of art 151(3) of the EC Treaty, is not applicable to acts
adopted on the basis of Title VI of the TEU.
71.
The UK government contends that the jurisdiction of the Court of First Instance does not extend to the
matters covered by Title VI of the TEU, and therefore to the question of access to the documents
concerning those matters. Justice and Home Affairs fall outside the scope of the EC Treaty and are
matters for inter-governmental co-operation. It is clear from art E of the TEU that in relation to Justice and
Home Affairs the institutions in question are to exercise their powers under the conditions and for the
purposes provided for by Title VI of the TEU. In exercising those powers they are acting within the scope
of Title VI, not of the EC Treaty. It follows from art L of the TEU that the provisions of the EC Treaty
concerning the powers of the court do not apply to Title VI of the TEU. Accordingly the jurisdiction of the
court is excluded as much in procedural matters as in matters of substance. In any event, it is frequently
impossible to draw a clear-cut distinction between the two.
72.
The UK government accepts that Decision 93/731 applies to Title VI documents, but considers that it
does not follow that the court may exercise jurisdiction over a refusal to allow access to such documents.
In particular, the court does not acquire jurisdiction simply because Decision 93/731 was adopted
pursuant to art 151 of the EC Treaty. Article 7(3) of Decision 93/731 is irrelevant in that connection, since
reference to the possibility of an action under art 173 of the EC Treaty cannot enlarge the jurisdiction of
the court.
73.
According to the applicant, Decision 93/731 itself expressly confirms that the court has jurisdiction in
cases concerning application of that decision, since it specifies that its provisions are applicable to any
document held by the Council. The criterion for application of Decision 93/731 is therefore the fact that the
document is held by the Council, irrespective of its subject matter, with the 556 exception of documents
drawn up outside the Council. In Carvel v EU Council Case T-194/94 [1996] All ER (EC) 53, [1995] ECR
II-2765, the Court of First Instance annulled a decision whereby the Council had refused the applicants
access to the decisions adopted by the Justice and Home Affairs Council; the Council did not contest the
jurisdiction of the court to adjudicate on access to documents falling under Title VI of the TEU in that case.
74.
That argument is supported by the Swedish, Danish and Netherlands governments. Although the court
has no jurisdiction to review the legality of Title VI documents, it does have jurisdiction over matters
concerning public access to those documents.
75.
The Netherlands government adds that the contested decision was not adopted on the basis of Title
VI, nor does that title constitute the legal basis of Decision 93/731. The court will not therefore be required
to adjudicate on co-operation in Justice and Home Affairs as such.
Findings of the court
76.
Before considering the objection raised by the French and UK governments, it is appropriate to
consider its admissibility in the light of the Rules of Procedure.
77.
This objection was not raised by the Council in the written pleadings. Furthermore, an application to
intervene is to be limited to supporting the form of order sought by one of the parties (see the final
paragraph of art 37 of the EC Statute of the Court of Justice, applicable to the Court of First Instance by
virtue of art 46 of that statute).
78.
 It follows that the French and UK governments are not entitled to raise an objection to admissibility
and that the court is not therefore obliged to consider the submissions they have made in that regard (see
the judgments in Comit International de la Rayonne et des Fibres Synthtiques (CIRFS) v EC
Commission Case C-313/90 [1993] ECR I-1125 (para 22)).
79.
However, under art 113 of the Rules of Procedure, the court may at any time of its own motion
consider whether there exists any absolute bar to proceeding with a case, including any raised by
interveners (see Association des Acieries Europennes Independantes (EISA) v European Commission
Case T-239/94 [1997] ECR II-1839 (para 26)).
80.
In this case the issue as to admissibility raised by the French and UK governments does involve an
absolute bar to proceeding in that it turns upon the jurisdiction of the court to entertain the application. It
can accordingly be examined by the court of its own motion.
81.
In this regard, Decision 93/731, in arts 1(2) and 2(2), expressly provides that it is to apply to all Council
documents. Decision 93/731 therefore applies irrespective of the contents of the documents requested.
82.
Moreover, pursuant to art K.8(1) of the TEU, measures adopted pursuant to art 151(3) of the EC
Treaty, which is the legal basis for Decision 93/731, are applicable to measures within the scope of Title
VI of the TEU.
83.
Thus, Council Decision (EC) 93/662 adopting the Councils Rules of Procedure (OJ 1993 L304 p 1),
which was adopted on the basis of inter alia art 151(3) of the EC Treaty, also applies to meetings of the
Council relating to Title VI of the TEU.
84.
It follows that, in the absence of any provision to the contrary in Decision 93/731 itself, its provisions
apply to documents relating to Title VI of the TEU.
557
85.
The fact that the court has, by virtue of art L of the TEU, no jurisdiction to review the legality of
measures adopted under Title VI does not curtail its jurisdiction in the matter of public access to those
measures. The assessment of the legality of the contested decision is based upon its jurisdiction to
review the legality of decisions of the Council taken under Decision 93/731, on the basis of art 173 of the
EC Treaty, and does not in any way bear upon the intergovernmental co-operation in the spheres of
Justice and Home Affairs as such. In any event, in the contested decision the Council itself drew the
applicants attention to its entitlement to appeal under art 173 of the EC Treaty (see para 18, above).
86.
The fact that the documents relate to Title VI only is relevant in so far as the contents of the
documents might possibly come within the scope of one or more of the exceptions provided for in
Decision 93/731. That fact is thus relevant only to the examination of the substantive lawfulness of the
decision taken by the Council and not to the admissibility of the application as such.
87.
It follows from the foregoing that the application is admissible.

SUBSTANCE
88.
The applicant puts forward five pleas in law in support of its application for the annulment of the
contested decision, namely: breach of the fundamental principle of Community law that citizens of the
European Union should be granted the widest and fullest possible access to Community institutions
documents; breach of the principle of protection of legitimate expectations; infringement of art 4(1) of
Decision 93/731; infringement of art 4(2) of Decision 93/731; and infringement of art 190 of the EC Treaty.
89.
The court will first examine the third and fifth pleas together.

Third and fifth pleas in law: infringement of art 4(1) of Decision 93/731 and infringement of
art 190 of the EC Treaty
Arguments of the parties
(i) Infringement of art 4(1) of Decision 93/731
90.
The applicant claims that the Council did not make a real assessment of the likely impact that granting
access to the documents requested might have on public security in the European Union. On the
contrary, the fact that a confirmatory application was necessary before the Council agreed to release one
of the documents which had already been handed over to the European Parliament and was thus fully in
the public domain is particularly disturbing in that respect.
91.
In the absence of a definition of public security in Decision 93/731, the applicant suggests the following
definition:
 documents or passages of documents whose access by the public would expose Community
citizens, Community institutions or Member States authorities to terrorism, crime, espionage,
insurrection, destabilisation and revolution, or would directly hinder the authorities in their efforts to
prevent such activities, shall not be accessible by virtue of the public security exception.
558
92.
The applicant then gives a precise description of the contents of all the documents requested that are
in its possession, in support of its argument that the public security exception was applied in an unlawful
manner by the Council.
93.
It rejects the Councils assertion that it would not be in the interest of public security to allow those
involved in illicit activities to obtain detailed knowledge of the structures and means available to police co-
operation in the EU. That assertion simply bears no relation to the actual content of the documents in
question. The applicant points out that the two documents to which the Swedish authorities refused
access concerned not public security but the negotiating positions of the Kingdom of the Netherlands and
the Federal Republic of Germany.
94.
The Council denies that it considered all the documents relating to Europol to be covered by the public
security exception. The fact that four documents were disclosed shows that a real assessment was
carried out, the outcome of which was that some of the requested documents could be released, whilst
others could not.
95.
The Council, supported by the French and UK governments, contends that there is in any case no
need to adopt a restrictive definition of public security for the purposes of the application of Decision
93/731. Public security must be defined in a flexible way in order to meet changing circumstances. In any
event, an assessment as to whether the release of a specific document could undermine the protection of
the public interest (public security) can only be made by the Council itself.
96.
That applies particularly as regards documents dealing exclusively with issues which fall under Titles V
and VI of the TEU. The Council trusts that, should the court consider that it has jurisdiction in matters
concerning access to documents dealing exclusively with matters falling under Title VI, it would
nevertheless refrain from substituting its assessment for that of the Council in this regard.
97.
The Council considers that the applicants summary of the documents in question is neither objective
nor precise.
98.
The Swedish government takes issue with the description given by the Council of the way in which the
Information Working Party and Coreper dealt with the request for access to the documents in question.
99.
In particular the documents requested were not made available to the Swedish representative in the
Information Working Party before its meeting. The matter could not be dealt with satisfactorily in the short
time available.
100.
As far as Coreper was concerned, the only matter addressed by it was whether a decision concerning
the request for disclosure could be taken by written procedure. When Coreper voted on 5 July 1995, the
Swedish government and four other member States abstained. The Swedish government made a
statement expressing its dissatisfaction at the way the case had been handled.
101.
The Danish government shares to a large extent the Swedish governments criticism of the way the
case was handled. It considers that the Councils assessment of the various documents was purely
formalistic. In the Council Secretariat the possibilities of derogation in art 4(1) of Decision 93/731 were
first examined and it was thought that considerations of public security could justify withholding of
documents relating to Europol in general. When the confirmatory application was being examined, doubts
arose as to whether public security considerations could really be applied generally as a ground for 559
withholding Europol documents. Accordingly, it was then decided to retreat to a statement of reasons
based on the very general considerations of art 4(2) of Decision 93/731. The discussion in the Council
Secretariat did not focus on whether publication would entail a risk of real adverse consequences either
for public security or the requirement of confidentiality.
102.
The Netherlands government, having examined the documents in question, considers that the refusal
to grant access to the documents cannot under any circumstances be justified by the requirements of
public security. However, it reserves its opinion as far as a document which is not in its possession is
concerned. In its view, in order to establish whether the Council was justified in refusing access to the
documents in question on the ground of public security, it is necessary to examine, document by
document, whether access to them would undermine the fundamental interests of the Community or of
the member states to the extent that their existence would be jeopardised. It points out that the Council
later agreed to make available at least four of those documents to a journalist, Mr T, and that the refusal
to grant the applicant access to those documents therefore constitutes arbitrary discrimination.
103.
The Council insists that the content of the documents was in fact examined. It considers that there is
no evidence that the other members of the Council who abstained did so for the same reasons as the
Swedish government. No member state voted against the confirmatory decision or associated itself with
the Swedish governments statement.
(ii) Infringement of art 190 of the EC Treaty
104.
The applicant claims that the refusal, expressed in a single sentence, to grant access to 16 of the 20
documents does not satisfy the requirements of art 190 of the EC Treaty or art 7(3) of Decision 93/731. It
was impossible for it to assess whether the refusal should be challenged before the court, and equally
impossible for the court to assess whether the Council had made proper use of the exceptions referred to
above. It was only because the applicant had in its possession most of the documents concerned, in full
or in part, that it was able to show that the Council had applied those exceptions unlawfully in the present
case. It asks the court to examine the documents concerned in order to assess whether the Council was
justified in availing itself of the exceptions cited.
105.
The Council, supported by the French and UK governments, contends that the statement of reasons
for the contested decision discloses the essential objective pursued by the Council and its decision is
therefore duly reasoned. It would be excessive to require a specific statement of reasons for each of the
technical choices made by the institution. If it were necessary to provide a very detailed statement of
reasons in the case of negative responses to requests for access, the underlying objectives of art 4(1)
would be compromised. Decision 93/731 lays down very tight time limits for replying to applications.
Consequently, when applications cover many documents involving large numbers of pages, the statement
of reasons which can be provided will inevitably be rather briefer than the statement of reasons given in
reply to applications of a more limited scope. Furthermore, the requested documents clearly had an
essentially common subject matter.
106.
The Swedish government maintains that the balancing of the Councils interest in maintaining the
confidentiality of its proceedings and the publics interest in having access to documents should be
undertaken in relation to each separate document and that the decision does not state sufficient reasons.
It 560 claims that the Council does not indicate whether both the reasons given for maintaining
confidentiality are applicable to all the documents or, if that is not the case, which reason or reasons for
maintaining confidentiality are applicable to each particular document. The public is entitled to know, from
the particular circumstances surrounding each separate action or matter, why a specific document is to be
kept confidential.
107.
The Danish government states that it is not sufficient to refer in general to the possibilities of
derogation and to reproduce the terms of Decision 93/731. Refusal under art 4(1) of that decision cannot
lawfully be explained by indicating that a particular interest which is included therein can be regarded
generally as affected, just as the option of derogation with regard to the duty of confidentiality in art 4(2)
cannot form the basis of a refusal in general terms. The principle of assessment on the facts is applicable
and in certain cases the Council might be required to produce a document with any information requiring
protection under art 4 deleted.
108.
The Netherlands government also states that the Councils reason for refusing access to the various
documents is obscure. The contested decision confines itself to repeating the criteria in art 4 of Decision
93/731 and does not reveal which documents were withheld on the basis of art 4(1) and which withheld
on the basis of art 4(2). As regards the documents to which access was refused on the ground of
confidentiality of the Councils proceedings, it does not appear, moreover, from the contested decision,
that the requisite balancing of interests took place.
Findings of the court
109.
Decision 93/731 is a measure which confers on citizens rights of access to documents held by the
Council. It is clear from the scheme of the decision that it applies generally to requests for access to
documents and that any person is entitled to ask for access to any Council document without being
obliged to put forward reasons for the request (see para 65, above).
110.
There are two categories of exception to the principle of general access for citizens to Council
documents set out in art 4 of Decision 93/731. These exceptions must be construed and applied
restrictively so as not to defeat the general principle enshrined in the decision (see, in relation to the
analogous provisions of Decision 94/90, WWF UK v European Commission Case T-105/95 [1997] All ER
(EC) 300, [1997] ECR II-313 (para 56)).
111.
The wording of the first category of exceptions, drafted in mandatory terms, provides that access to a
Council document cannot be granted if its disclosure could undermine the protection of the public interest
(public security, international relations, monetary stability, court proceedings, inspections and
investigations) (see para 7, above). Accordingly, the Council is obliged to refuse access to documents
which come within any one of the exceptions in this category once the relevant circumstances are shown
to exist (see Carvel [1996] All ER (EC) 53, [1995] ECR II-2765 (para 64)).
112.
Nevertheless, it follows from the use of the verb could, in the present conditional, that in order to
demonstrate that the disclosure of particular documents could undermine the protection of the public
interest, the Council is obliged to consider in respect of each requested document whether, in the light of
the information available to it, disclosure is in fact likely to undermine one of the facets of public interest
protected by this first category of exceptions. If that is the case, the Council is obliged to refuse access to
the documents in question 561(see Interporc (1998) CFI Transcript, 6 February (para 52) and van der
Wal v European Commission Case T-83/96 [1998] All ER (EC) 289 (para 43)).
113.
By way of contrast, the wording of the second category, drafted in enabling terms, provides that the
Council may also refuse access in order to protect the confidentiality of its proceedings (see above, para
8). It follows that the Council enjoys a margin of discretion which enables it, if need be, to refuse access
to documents which touch upon its deliberations. It must, nevertheless, exercise this discretion by striking
a genuine balance between on the one hand, the interest of the citizen in obtaining access to the
documents and, on the other, any interest of its own in maintaining the confidentiality of its deliberations
(see Carvel [1996] All ER (EC) 53, [1995] ECR II-2765 (paras 6465)).
114.
The Council is also entitled to rely jointly on an exception derived from the first category and one
relating to the second category in order to refuse to grant access to documents which it holds, there being
no provision in Decision 93/731 which precludes it from so doing. The possibility cannot be ruled out that
the disclosure of particular documents by the Council could cause damage both to the interest protected
by the first category of exception and to the Councils interest in maintaining the confidentiality of its
deliberations (see WWF UK [1997] All ER (EC) 300, [1997] ECR II-313 (para 61)).
115.
In the light of these considerations, it is necessary to consider whether the contested decision satisfies
the criteria laid down by art 190 of the Treaty regarding the statement of reasons.
116.
The duty to state reasons in individual decisions has the double purpose of permitting, on the one
hand, interested parties to know the reasons for the adoption of the measure so that they can protect their
own interests and, on the other hand, enabling the Community court to exercise its jurisdiction to review
the validity of the decision (see esp Socit franaise des Biscuits Delacre v EC Commission Case C-
350/88 [1990] ECR I-395 (para 15) and Eugnio Branco Lda v European Commission Case T-85/94
[1995] ECR II-45 (para 32)).
117.
The statement of reasons for a decision refusing access to a document must therefore containat
least for each category of documents concernedthe particular reasons for which the Commission
considers that disclosure of the requested documents comes within the scope of one of the exceptions
provided for in Decision 93/731 (see WWF UK [1997] All ER (EC) 300, [1997] ECR II-313 (paras 64, 74)
and Interporc (1998) CFI Transcript, 6 February (para 54)).
118.
In the contested decision the Council indicated only that the disclosure of the 16 documents in
question would prejudice the protection of the public interest (public security) and that the documents
related to the proceedings of the Council, particularly the views expressed by members of the Council,
and for that reason fell within the scope of the duty of confidentiality (see para 18, above).
119.
Although the Council was at once invoking both the mandatory exception based upon the protection of
the public interest (public security) and also the discretionary exception based upon protection of the
confidentiality of its proceedings, it did not specify whether it was invoking both exceptions in respect of all
of the documents refused or whether it considered that some documents were covered by the first
exception while others were covered by the second.
120.
In that respect, the court notes that although the initial refusal contained in the letter of 1 June 1995
was based only upon the principle of confidentiality as set out in Article 4(1) of Decision 93/731 the
Council was nevertheless able to grant access to two further documents in the course of its consideration
of the confirmatory request, namely a report on the activities of the Europol Drugs Unit 562(document No
4533/95) and a provisional agenda for a meeting of Committee K.4 (document No 4135/95), documents
clearly relating to the activities of the Council within the scope of Title VI of the TEU. If the fact that such
documents related to Title VI meant that they were automatically covered by the exception based upon
the protection of the public interest (public security), the Council had no entitlement to grant access to the
documents. Moreover, given that the Council considered that it was entitled to grant access to these two
documents, having first balanced the interests involved, it follows that the Council must necessarily have
considered that all of the documents relating to Title VI did not automatically fall within the scope of the
first exception based upon the protection of the public interest (public security). Furthermore, the Council
itself admitted that it had not considered that all of the documents connected with Europol were covered
by the exception relating to public security.
121.
The case law of the Court of Justice shows that the concept of public security does not have a single
and specific meaning. Thus, the concept covers both the internal security of a member state and its
external security (see the judgment in Fritz Werner Industrie-Ausrstungen GmbH v Germany Case C-
70/94 [1995] ECR I-3189 (para 25)), as well as the interruption of supplies of essential commodities such
as petroleum products which may threaten the very existence of a country (see the judgment in Campus
Oil Ltd v Minister for Industry and Energy Case 72/83 [1984] ECR 2727 (para 34)). The concept could
equally well encompass situations in which public access to particular documents could obstruct the
attempts of authorities to prevent criminal activities, as the applicant has argued.
122.
Mr Elsens note demonstrates that most of the documents to which access was refused were
concerned only with negotiations on the adoption of the Europol Convention, in particular the proposals of
the Presidency and of other delegations with regard to those negotiations, and not with operational
matters of Europol itself (see para 15, above). Thus, in the absence of any explanation on the part of the
Council as to why the disclosure of these documents would in fact be liable to prejudice a particular
aspect of public security, it was not possible for the applicant to know the reasons for the adoption of the
measures and therefore to defend its interests. It follows that it is also impossible for the court to assess
why the documents to which access was refused fall within the exception based upon the protection of
the public interest (public security) and not within the exception based upon the protection of the
confidentiality of the Councils proceedings.
123.
Nor can the Council claim that, in this instance, it was unable to explain why the exception applied
without undermining the essential purpose of the exception, given the very nature of the interest to be
protected and the mandatory character of the exception. In fact, Mr Elsens note clearly shows that it was
possible to give an indication of the reasons why certain documents could not be disclosed to the
applicant without at the same time disclosing their contents.
124.
Finally, so far as concerns the exception in favour of the protection of the confidentiality of its
proceedings, the Council did not specifically indicate in the contested decision that all of the documents
included in the applicants request were covered by the exception based upon the protection of the public
interest (see para 119, above). The applicant could not therefore rule out the possibility that access to
some of the documents in question was being refused because they 563 were covered only by the
exception based upon the protection of the confidentiality of its proceedings.
125.
The terms of the contested decision do not, however, permit the applicant and, therefore, the court to
check whether the Council has complied with its duty to carry out a genuine balancing of the interests
concerned as the application of art 4(2) of Decision 93/731 requires. In fact, the contested decision
mentions only the fact that the requested documents related to proceedings of the Council, including the
views expressed by members of the Council, without saying whether it had made any comparative
analysis which sought to balance, on the one hand, the interest of the citizens seeking the information
and, on the other hand, the criteria for confidentiality of the proceedings of the Council (see Carvel [1996]
All ER (EC) 53, [1995] ECR II-2765 (para 74)).
126.
Moreover, the first reply from the Councilsent to the applicant in French although the applicant had
written the initial request in Germanconfined itself to citing the provisions of art 4(1) of Decision 93/731,
in support of its view that the documents were subject to the principle of confidentiality. It did not
therefore permit the applicant or the court to confirm that the Council had genuinely balanced the interests
involved at the stage of its consideration of the applicants first request.
127.
It follows from all of the foregoing that the contested decision does not comply with the requirements
for reasoning as laid down in art 190 of the EC Treaty and must therefore be annulled without there being
any need to consider the other grounds raised by the applicant or to look at the contents of the
documents themselves.

The request of the Netherlands government that the Court of First Instance invite the Court
of Justice to produce a note drafted by its services
128.
The Netherlands government requests that the Court of First Instance invite the Court of Justice to
produce a note drafted by the Research and Documentation service of the court for the purposes of that
courts judgment in Netherlands v EU Council Case C-58/94 [1996] ECR I-2169.
129.
As the present judgment is not based upon that note, there is no need to rule on this request.

Publication of the defence on the Internet

Arguments of the parties
130.
As indicated in para 22, above, by letter received on 3 April 1996 the Council drew the attention of the
court to the fact that certain pertinent documents, including the Councils defence, had been published on
the Internet. It considers that the applicants conduct was prejudicial to the proper course of the
procedure. The Council laid particular stress on the fact that the text of the defence had been edited by
the applicant before it was placed on the Internet. Furthermore, the names and contact details of the
Councils agents in the case were given and the public encouraged to send their comments on the case
to those agents. The Council requested the court to take any measures which might be appropriate in
order to avoid further such action on the part of the applicant.
131.
 By letter received on 3 May 1996, the applicants lawyers explained that they had played no role in the
placing of the defence and other documents concerning the case on the Internet. They had no knowledge
of those facts before receiving the letter from the registry of the Court of First Instance. They 564 had
immediately asked the applicant to remove all the documents from the Internet, and informed it that they
would no longer be able to represent it if that was not done.
132.
In its observations received on 24 May 1996, the applicant confirmed that it had placed the documents
on the Internet without informing its lawyers. It explained that the editing of the defence had been carried
out for purely practical reasons and that its intention was not to alter its contents or weaken the Councils
case. It simply wanted to shorten the defence by not reproducing certain passages in view of the time
required to put the defence on the Internet. It had no intention of putting pressure on the Council and
added that the names and contact details of the Councils agents were included simply because they
knew about the case, not to encourage the public to contact them directly as individuals.
133.
The applicant undertook to refrain from placing on the Internet or in any other way making available to
the public any further documents exchanged between the parties in the case. It would thenceforth restrict
itself to normal media reports on the case. The applicant further indicated that it had taken the decision to
have the defence withdrawn from the Internet. However, the document had been placed on the Internet
by an independent organisation, Grvande Journalister (an association of Swedish investigative reporters
and editors), which refused to withdraw it. Under Swedish law the applicant had no legal means of forcing
that association to withdraw the document and the latter was therefore responsible for keeping the
defence on the Internet.
134.
By letter received on 28 May 1996, the Swedish government explained that the Legal Director at the
Ministry of Justice had received a copy of the defence from the applicant and the Legal Director had
subsequently released a copy to a journalist without any objection on the applicants part. In doing so, the
Legal Director had taken into account the fact that the applicant had already published a detailed report
on the main elements of the defence and had given the names of the representatives of the Council
concerned. Another factor in that decision was that the document had not been transmitted to the
Swedish government by a Community institution, but by a private individual who had the right to dispose
of the document and had already demonstrated his willingness to disseminate it. The ministry was in no
way involved in the publication of the defence on the Internet and the newspapers action in that respect
was regarded as a provocation.
Findings of the court
135.
Under the rules which govern procedure in cases before the Court of First Instance, parties are
entitled to protection against the misuse of pleadings and evidence. Thus, in accordance with the third
subparagraph of art 5(3) of the Instructions to the Registrar of 3 March 1994 (OJ 1994 L78 p 32), no third
party, private or public, may have access to the case-file or to the procedural documents without the
express authorisation of the President, after the parties have been heard. Moreover, in accordance with
art 116(2) of the Rules of Procedure, the President may exclude secret or confidential documents from
those furnished to an intervener in a case.
136.
These provisions reflect a general principle in the due administration of justice according to which
parties have the right to defend their interests free from all external influences and particularly from
influences on the part of members of the public.
565
137.
It follows that a party who is granted access to the procedural documents of other parties is entitled to
use those documents only for the purpose of pursuing his own case and for no other purpose, including
that of inciting criticism on the part of the public in relation to arguments raised by other parties in the
case.
138.
In the present case, it is clear that the actions of the applicant in publishing an edited version of the
defence on the Internet in conjunction with an invitation to the public to send their comments to the agents
of the Council and in providing the telephone and telefax numbers of those agents, had as their purpose
to bring pressure to bear upon the Council and to provoke public criticism of the agents of the institution in
the performance of their duties.
139.
These actions on the part of the applicant involved an abuse of procedure which will be taken into
account in awarding costs, having regard, in particular, to the fact that this incident led to a suspension of
the proceedings and made it necessary for the parties in the case to lodge additional submissions in this
respect.

COSTS
140.
Under art 87(2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if
they have been applied for in the successful partys pleadings. In this case the applicant asked that the
Council be ordered to pay the costs. However, under art 87(3) of the rules, the court may, where the
circumstances are exceptional, order that the costs be shared or that each party bear its own costs. In
view of the abuse of procedure found to have been committed by the applicant, the Council will be
ordered to pay only two-thirds of the applicants costs.
141.
f)Pursuant to art 87(4) of the Rules of Procedure, the interveners will be ordered to pay their own
costs.
On those grounds, the Court of First Instance (Fourth Chamber, Extended Composition) hereby: (1)
annuls the Councils decision of 6 July 1995 refusing the applicant access to certain documents relating to
the European Police Office (Europol); (2) orders the Council to pay two-thirds of the applicants costs as
well as its own costs; (3) orders the Kingdom of Denmark, the French Republic, the Kingdom of the
Netherlands, the Kingdom of Sweden and the United Kingdom to bear their own costs.

566

[1998] All ER (EC) 567

Fischer v Finanzamt Donaueschingen

(Case C-283/95)

EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE CHAMBER), MANCINI, KAPTEYN, MURRAY AND
HIRSCH ADVOCATE GENERAL JACOBS
30 JANUARY, 20 MARCH 1997, 11 JUNE 1998
European Community Value added tax Supply of goods and services Supply Unlawful gaming
Whether supply of services within scope of Community law Council Directive (EEC) 77/388, art 2(1).
European Community Value added tax Exemptions Betting, gaming and lotteries Unlawful gaming
Whether exempt supplies Council Directive (EEC) 77/388, art 13B(f).

Under a German law transposing art 13B(f)1 of Council Directive (EEC) 77/388 on the harmonisation of
the laws of the member states relating to turnover taxescommon system of value added tax: uniform
basis of assessment, the turnover of licensed public casinos was exempted from value added tax. F,
without having the relevant permit, organised games of chance which resembled the game of roulette as
played in licensed casinos, and was assessed to value added tax on the basis of the estimated takings. In
proceedings in which F contested the assessment, the Finanzgericht Baden-Wrttemberg, Freiburg,
referred to the Court of Justice of the European Communities for a preliminary ruling, inter alia, the
questions whether (i) the unlawful operation of games of chance constituted supplies of services within
the scope of art 2(1)2 of the Sixth Directive, and (ii) such an activity could be charged to value added tax
when a corresponding activity carried on lawfully was exempted.
1
Article 13B, so far as material, is set out at p 581 b, post
2
Article 2(1), so far as material, is set out at p 580 j, post

Held (1) The only unlawful activities which did not give rise to any liability to value added tax were ones
relating to products such as illegally imported or supplied narcotic drugs or counterfeit currency which had
special characteristics such that their release into the economic and commercial channels of the
Community was by definition precluded and could give rise only to penalties under the criminal law. It
followed that betting and gambling transcations, which were lawful in a number of member states, fell
within the scope of the Sixth Directive (see p 583 b to f and p 584 g h, post); Witzemann v Hauptzollamt
Mnchen-Mitte Case 343/89 [1993] STC 108, [1990] ECR I-4477 and Lange v Finanzamt
Frstenfeldbruck Case C-111/92 [1997] STC 564, [1993] ECR I-4677 applied.
(2) Outside those cases where all competition between a lawful and unlawful economic sector was
ruled out, the principle of fiscal neutrality precluded a generalised distinction from being drawn in the
levying of value added tax between lawful and unlawful transactions. Moreover, games of chance, and in
particular roulette, were lawfully played in a number of member states, and since 567 unlawful games
were in competition with lawful ones, the principle of fiscal neutrality precluded their being treated
differently for value added tax purposes. Furthermore, any exemption under art 13B(f) had to be applied
in accordance with that principle, even where member states had exercised their power under the
directive to lay down conditions and limitations to the exemption. It followed that member states could not
reserve the exemption solely to lawful games of chance (see p 583 j to p 584 b g h, post); EC
Commission v Italy Case C-45/95 [1997] STC 1062, [1997] ECR I-3605 applied.

Notes
For supply of goods and services and exemptions in relation to Community value added tax rules, see 52
Halsburys Laws (4th edn) paras 2017, 2028.
Cases cited
Boots Co plc v Customs and Excise Comrs Case C-126/88 [1990] STC 387, [1990] ECR I-1235, ECJ.
EC Commission v Italy Case C-45/95 [1997] STC 1062, [1997] ECR I-3605, ECJ.
Einberger v Hauptzollamt Freiburg Case 240/81 [1982] ECR 3699.
Einberger v Hauptzollamt Freiburg Case 294/82 [1984] ECR 1177.
Glawe (H J) Spiel-und Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt
Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994] STC 543, [1994] ECR I-1679, ECJ.
Horvath v Hauptzollamt Hamburg-Jonas Case 50/80 [1981] ECR 385.
Lange v Finanzamt Frstenfeldbruck Case C-111/92 [1997] STC 564, [1993] ECR I-4677, ECJ.
Mol v Inspecteur der Invoerrechten en Accijnzen Case 269/86 [1988] ECR 3627.
Naturally Yours Cosmetics Ltd v Customs and Excise Comrs Case 230/87 [1988] STC 879, [1988] ECR
6365, ECJ.
Vereniging Happy Family Rustenburgerstraat v Inspecteur der Omzetbelasting Case 289/86 [1988] ECR
3655.
Witzemann v Hauptzollamt Mnchen-Mitte Case 343/89 [1993] STC 108, [1990] ECR I-4477, ECJ.
Wolf v Hauptzollamt Dsseldorf Case 221/81 [1982] ECR 3681.

Reference
By order of 21 August 1995 the Finanzgericht Baden-Wrttemberg, Freiburg, referred to the Court of
Justice of the European Communities for a preliminary ruling three questions (set out at p 582 g, post) on
the interpretation of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member
states relating to turnover taxescommon system of value added tax: uniform basis of assessment.
Those questions were raised in proceedings between Karlheinz Fischer and the Finanzamt
Donaueschingen concerning the payment of turnover tax on unlawful and punishable games of chance.
Written observations were submitted on behalf of: the German government, by E Rder, Ministerialrat in
the Federal Ministry of Economic Affairs, and B Kloke, Oberregierungsrat in that ministry, acting as
agents; the United Kingdom government, by S Braviner, of the Treasury Solicitors Department, acting as
agent, and P Mantle, Barrister; and the European Commission, by J Grunwald, Legal Adviser, acting as
agent. Oral observations were submitted on behalf of: the German government, represented by E Rder;
the UK government, represented by J E Collins, Assistant Treasury Solicitor, acting as agent, and K
Parker QC; and the Commission, represented by 568 J Grunwald. The language of the case was
German. The facts are set out in the opinion of the Advocate General.

20 March 1997.

The Advocate General (F Jacobs)

delivered the following opinion.
1.
The issues raised by the present case, referred to the Court of Justice of the European Communities
by the Finanzgericht (the finance court) Baden- Wrttemberg, are as follows: whether transactions
consisting in the unlawful provision of roulette games fall within the scope of value added tax (VAT); if so,
whether a member state must exempt them from VAT when it exempts the provision of such games by
licensed public casinos; if not, what constitutes the taxable amount for VAT purposes.

Relevant Community and national provisions

2.
Article 2 of Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states
relating to turnover taxescommon system of value added tax: uniform basis of assessment (the Sixth
Directive) (OJ 1977 L145 p 1) provides:
The following shall be subject to value added tax: (1) the supply of goods or services effected
for consideration within the territory of the country by a taxable person acting as such
3.
Article 11A(1)(a) of the Sixth Directive provides that the taxable amount for domestic transactions is
everything which constitutes the consideration which has been or is to be obtained by the supplier from
the purchaser, the customer or a third party for such supplies including subsidies directly linked to the
price of such supplies.
4.
Article 13B(f) exempts from tax betting, lotteries and other forms of gambling, subject to conditions
and limitations laid down by each Member State.
5.
Article 33 of the Sixth Directive provides:
Without prejudice to other Community provisions, [the provisions of] this Directive shall not
prevent a Member State from maintaining or introducing taxes on insurance contracts, taxes on
 betting and gambling, excise duties, stamp duties and, more generally, any taxes, duties or charges
which cannot be characterised as turnover taxes
Paragraph 1(1)(1) of the Umsatzteuergesetz (the UstG) (the Law on Turnover Tax) implements art 2(1)
of the Sixth Directive by subjecting to VAT domestic transactions effected for consideration by a trader
within the course of his business. Pursuant to art 13B(f) of the Sixth Directive, para 4(9)(b) of the UstG
exempts, inter alia, the turnover of licensed public casinos arising from the operation of such casinos.

The facts and the national courts questions

7.
Under German law the game of roulette may be organised commercially only by licensed public
casinos. From 1987 to 1989 Mr Fischer operated games of roulette at a number of locations in the
Federal Republic of Germany. Mr Fischer was not licensed to run a casino and therefore to operate
games of roulette but had a permit to operate a game of skill using a machine called a Roulette Opta II.
However, he departed from the terms of the permit in such a way that the game amounted to a game of
roulette.
569
The equipment used by Mr Fischer consisted of a wheel bearing the numbers 1 to 24 and also the
numbers 0 and X. Numbers 1 to 12 were black and numbers 13 to 24 red. The object of the game was to
predict, by placing chips on the appropriate squares of the gaming table, where the ball thrown by a
croupier would come to rest. The players purchased the chips for DM 5 each. In each game they could
place one or more chips on the squares for one of the numbers 1 to 24, 0 and X (en plein), on the line
between the two numbers ( cheval) and/or on the red and/or black numbers squares. A player received
24 times the value of his stake where he placed his stake on the square for the number on which the ball
came to rest, 12 times his stake where the ball came to rest on one of the two numbers between whose
squares he had placed his stake and twice his stake where he correctly predicted that the ball would
come to rest on red or black. Winnings were paid in chips after each game, and the chips collected by
croupiers were available for payment of winnings and for purchase. Players who wished to discontinue
the game could exchange their remaining chips for cash.
9.
Mr Fischer recorded as taxable turnover his net takings, ie the surplus amounts which he had taken at
the end of each day. The Finanzamt (the tax office), however, took the view that the taxable amount was
the sums of money (in the form of chips) staked by players in each game or series of games as reduced
by their winnings (also in the form of chips). Although it is not entirely clear, it appears that the Finanzamt
sought to take account only of sessions in which the house made a net gain, ignoring its net losses in
other games. Since Mr Fischer had not recorded that figure (which would have involved recording the
stakes and winnings for each player), the Finanzamt estimated it by multiplying his net takings by a factor
of six, based on the probability of each player winning or losing.
10.
The Finanzgericht raises first of all the question whether the provision of the roulette games is taxable.
Mr Fischer did not operate a licensed public casino within the meaning of para 4(9)(b) of the UstG and
indeed was convicted under para 284 of the German Criminal Code for operating a prohibited game of
chance. Since Mr Fischers turnover therefore does not qualify for exemption under para 4(9)(b), the
Finanzgericht concludes that it is taxable under German law by virtue of the general charging provision in
para 1(1)(1), first sentence, of the UstG. However, referring to the courts judgments in Vereniging Happy
Family Rustenburgerstraat v Inspecteur der Omzetbelasting Case 289/86 [1988] ECR 3655 and Mol v
Inspecteur der Invoerrechten en Accijnzen Case 269/86 [1988] ECR 3627 where it was held that unlawful
supplies of narcotic drugs fell outside the scope of VAT, the Finanzgericht wonders whether as a matter of
Community law the fact that Mr Fischers transactions were unlawful under German law precluded them
from being taxed.
11.
Secondly, on the assumption that Mr Fischers transactions are taxable, the Finanzgericht seeks
guidance on the taxable amount. The Finanzgericht is inclined to accept the view of the Finanzamt but
considers that it is precluded from giving judgment in its favour by the courts judgment in H J Glawe
Spiel-und Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt Hamburg-Barmbek-
Uhlenhorst Case C-38/93 [1994] STC 543, [1994] ECR I-1679, where it was held that the taxable amount
for services consisting in the operation of slot machines was the net takings emptied from the machine
after payment of the winnings rather than the total stakes inserted into the machine by players. The
Finanzgericht notes the similarities between the two forms of gambling. It adds that, although in its
judgment the court noted that the slot machines were 570 by law set in such a way as to pay out winnings
of at least 60% of the stakes inserted, that point cannot be considered crucial to the courts ruling.
12.
The Finanzgericht has therefore put the following questions to the Court of Justice:
(1) Is Article 2(1) of the Sixth Directive to be interpreted as meaning that services which the
organizer of unlawful and punishable games of chance provides to the players are not taxable?
(2) If question 1 is to be answered in the negative: is Article 11A(1)(a) of the Sixth Directive to be
interpreted as meaning that, in the case of unlawful gaming in the form of roulette, the basis of
assessment for the operators services to the players is the amount retained by the operator during
a tax period?
(3) If question 2 is to be answered in the negative: how is the basis of assessment to be
determined in cases described under questions 1 and 2?
 Question 1
13.
The courts case law concerning VAT on illegal transactions has its origins in the case law on customs
duties. In its judgments in Horvath v Hauptzollamt Hamburg-Jonas Case 50/80 [1981] ECR 385, Wolf v
Hauptzollamt Dsseldorf Case 221/81 [1982] ECR 3681 and Einberger v Hauptzollamt Freiburg Case
240/81 [1982] ECR 3699 (Einberger No 1), the court held that no customs debt arose upon the
importation of drugs otherwise than through economic channels strictly controlled by the competent
authorities for use for medical and scientific purposes. The court reasoned that the importation and
marketing of narcotic drugs otherwise than through such channels were prohibited in the member states
by virtue of an international agreement to which all the member states were signatories, namely the
Single Convention on Narcotic Drugs 1961 (New York, 30 March 1961; TS 23 (1979); Cmnd 7466). A
customs debt could not arise upon the importation of drugs which could not be marketed and integrated
into the economy of the Community. It added that unlawful imports of drugs fell wholly outside the scope
of the objectives assigned to the Community in art 2 of the treaty and the guidelines laid down in art 29 for
the operation of the customs union.
14.
In Einberger v Hauptzollamt Freiburg Case 294/82 [1984] ECR 1177 (Einberger No 2) the court
extended its case law on customs duties to VAT on unlawful imports of narcotic drugs. Subsequently in
Happy Family and Mol the court held that unlawful supplies of narcotic drugs made within the territory of
the country also fell outside the scope of VAT. It reasoned that unlawful transactions in drugs were alien to
the objectives of the Sixth Directive, namely to assist effective liberalisation of the movement of persons,
goods, services and capital, the integration of national economies and the achievement of a common
market permitting fair competition and resembling a real internal market (see the judgments in Mol [1988]
ECR 3627 (paras 14, 16) and Happy Family [1988] ECR 3655 (paras 16, 18).
15.
In two further cases the court was asked to consider the same issue in relation to two further
categories of goods. In Witzemann v Hauptzollamt Mnchen-Mitte Case 343/89 [1993] STC 108, [1990]
ECR I-4477, it extended the case law on customs duties and VAT to imports of counterfeit currency,
taking the view that its reasoning in that case law applied with even greater force to counterfeit currency
which, unlike narcotic drugs, was subject to a total 571 prohibition on importation and marketing. In Lange
v Finanzamt Frstenfeldbruck Case C-111/92 [1997] STC 564, [1993] ECR I-4677, on the other hand, the
court held that exports of goods to eastern Europe made in contravention of an export ban adopted within
the framework of the Co-ordinating Committee for Multilateral Export Controls (COCOM) fell within the
scope of the tax. In contrast to previous cases, the goods were not such that all transactions were
prohibited by reason of their nature or particular characteristics; all that was prohibited was their export to
certain destinations owing to possible use for strategic purposes.
16.
In the present case the court has received observations from the German and UK governments and
the European Commission. All three take the view that, in contrast to unlawful transactions in narcotic
drugs and counterfeit currency, the unlawful provision of roulette games falls within the scope of the Sixth
Directive. I share that view.
17.
It is clear that the courts case law on narcotic drugs and counterfeit currency constitutes an exception
to the normal rule that lawful and unlawful transactions should be accorded the same tax treatment. In
Mol [1988] ECR 3627 (para 18) and Happy Family [1988] ECR 3655 (para 20) the court held:
It must be acknowledged that the principle of fiscal neutrality does in fact preclude, as far as the
levying of value-added tax is concerned, a generalised differentiation between lawful and unlawful
transactions. However, that is not true in the case of the supply of products, such as narcotic drugs,
which have special characteristics inasmuch as, because of their very nature, they are subject to a
total prohibition on their being put into circulation in all the Member States, with the exception of
strictly controlled economic channels for use for medical and scientific purposes. In a specific
situation of that kind where all competition between a lawful economic sector and an unlawful
sector is precluded, the fact that no liability to value-added tax arises cannot affect the principle of
fiscal neutrality.
18.
Thus the court, responding to the concern expressed by the French, German and Netherlands
governments in those cases that more favourable tax treatment of unlawful trade would undermine the
principle of fiscal neutrality, made it clear that, where there existed the possibility of competition between
lawful and unlawful trade, the fiscal treatment should be the same (see Mol [1988] ECR 3627 (para 17)
and Happy Family [1988] ECR 3655 (para 19)).
19.
That is the case here. The transactions in issue in the present case are plainly distinguishable from
transactions in narcotic drugs or counterfeit currency. The provision of gambling services such as roulette
forms the subject of a substantial lawful trade which is an integral part of the economy of the Community.
It is not, to use the courts words in Happy Family and Mol, alien to the objectives of the Sixth Directive.
The transactions in issue in the present case are unlawful solely because a lawful game for which a
permit was held was converted into a game which may lawfully be provided only in licensed public
casinos. The case is more akin to the situation in Lange, in which the goods were not prohibited by their
nature and were the subject of an export ban solely because they were destined for certain countries for
use for strategic purposes. In the present case there exists the possibility of competition between lawful
and unlawful trade, and the principle of fiscal neutrality demands that the VAT system should not favour
unlawful gambling by placing it at an advantage over lawful gambling.
572
20.
I conclude therefore that unlawful roulette transactions such as those in issue fall within the scope of
VAT. They are therefore taxable under art 2(1) of the Sixth Directive unless specifically exempted
pursuant to art 13B(f).
21.
In that connection there is a further issue, not raised by the national court, which is relevant to the
outcome of the dispute, namely whether Germany is entitled under art 13B(f) to limit exemption from tax
to roulette games provided by licensed public casinos. Article 13B(f) permits member states to impose
conditions and limitations on the exemption provided for in that provision. However, the question remains
whether the member states discretion in limiting the scope of the exemption is curtailed by the principle of
fiscal neutrality.
22.
As the United Kingdom pointed out in its written observations, the ruling in Lange shows that the
application of the principle of fiscal neutrality is not confined to circumstances in which, but for its
application, lawful trade would be placed at a disadvantage in relation to unlawful trade. In that case the
principle of fiscal neutrality was applied in circumstances in which unlawful trade would otherwise have
been placed at a disadvantage. The court held that an exporting member state was not entitled to
withhold the export exemption provided for by art 15 of the Sixth Directive on the ground that the exports
in question were unlawful. The court observed that, in accordance with the principle of fiscal neutrality on
which the Sixth Directive was based, art 15 made no distinction for the purpose of exemptions between
lawful and unlawful exports. Moreover, noting that member states were required under art 17(3) of the
Sixth Directive to grant deduction or refund of input tax on goods qualifying for an export exemption under
art 15, the court added that the purpose of such exemptions was to ensure that consumers in non-
member countries were not subject to Community VAT. Consequently, the refusal by a member state to
apply an exemption provided for by the Sixth Directive to an export transaction in order to punish an
infringement of a national provision requiring authorisation would pursue an objective alien to the Sixth
Directive (see [1997] STC 564, [1993] ECR I-4677 (paras 1920)).
23.
In its written observations the United Kingdom took the view that, in the light of that ruling, the principle
of fiscal neutrality precluded Germany from restricting the exemption from tax in art 13B(f) to lawful
transactions. At the hearing, however, it withdrew that contention, stating that it was based on a
misunderstanding of the German rule. The United Kingdom observed that, while a member state was not
entitled to make a general distinction for the purpose of granting exemptions between lawful and unlawful
transactions, it was permitted to lay down conditions and limitations to the exemption, including a
condition that the transaction in question be carried out in licensed public casinos. Thus Germany was
entitled to exclude from the scope of the exemption transactions not performed in licensed public casinos
but could not, for example, refuse exemption for unlawful transactions performed by such a casino or
transactions rendered unlawful because the casino was late in renewing its licence.
24.
At the hearing the German government sought to justify the limitation on different grounds. It
contended that the principle of fiscal neutrality did not require the exemption to be extended to unlawful
transactions because the two categories of transaction were subject to different conditions of competition.
Under German law the counterpart to the VAT exemption for lawful roulette 573 games is that they are
subject to a special gaming or casino tax; transactions which are not exempt from VAT are not subject to
that tax.
25.
It appears from the order for reference that the games in issue in the main proceedings are
substantially identical to the roulette games provided by licensed casinos. It is therefore difficult to see
how it would be consistent with the aim of fiscal neutrality underlying the Sixth Directive for a member
state to refuse exemption to the former on the ground that they are provided unlawfully.
26.
As the United Kingdom correctly pointed out in its written observations, the ruling in Lange shows that
the principle of fiscal neutrality precludes the use of the VAT system to penalise unlawful transactions.
The incidence of VAT should not be made dependent upon the particularities of a member states criminal
legislation. To recognise an exception in the present case would undermine the consistency which should
characterise the application of fiscal legislation.
27.
I do not think it possible in this case to contend, as the United Kingdom sought to do at the hearing,
that Germany is merely laying down the conditions under which the exemption applies rather than making
a general distinction which the United Kingdom accepts is not permissiblebetween lawful and unlawful
trade. The national court presents the German rules as drawing a distinction between lawful and unlawful
roulette; its first question is put on that basis. There is nothing in the documents before the court to cast
doubt on that analysis. On the contrary, it appears that only lawful roulette is exempt from tax because
roulette games can be provided commercially only by licensed casinos. That the distinction between
lawful and unlawful roulette results from a condition under which the exemption is expressed to apply is of
little consequence in that respect.
28.
 It is therefore unnecessary in the present proceedings to consider the more general issue of the extent
to which, leaving aside distinctions between lawful and unlawful trade, the principle of fiscal neutrality may
curtail the discretion, accorded to member states by the introductory words of art 13B or by individual
exemptions to lay down conditions, limitations or exclusions in respect of certain exemptions. Given that
the transactions covered by art 13B of the Sixth Directive take place on competitive markets, the principle
of fiscal neutrality, if taken too far, would virtually remove any discretion accorded by the directive to
member states to define the terms of the relevant exemptions. It nevertheless seems to me that the
directive may impose limits on the power of member states, in defining the scope of exemptions, to
distinguish between competing taxable persons performing substantially identical transactions.
29.
The German governments argument that the principle of fiscal neutrality does not preclude the
German rules because different conditions apply to lawful and unlawful trade is in my view also
untenable. It seems to me that, for the purpose of applying the principle of fiscal neutrality in cases such
as the present, VAT must be considered in isolation. It would be unworkable to allow member states to
take account of the imposition on lawful transactions of other unharmonised taxes, whether direct or
indirect, the amount of which may or may not correspond to the VAT which is sought to be levied on
unlawful transactions.
30.
It is clear from art 33 of the Sixth Directive that excise duties and other special taxes may be imposed
in addition to VAT where applicable; indeed art 11A(2)(a) provides that such duties and taxes are to be
included in the taxable amount for VAT purposes. A member state is therefore free to impose special 574
taxes on gambling or gambling establishments and to determine the scope of such taxes. It may, if it
wishes, define the scope of any national taxes in such a way that a transaction is not subject both to VAT
and to the national tax; but it is not obliged to do so. What a member state is not entitled to do is to define
the contours of the harmonised Community VAT system by reference to unharmonised national taxes.
31.
I conclude therefore that a member state is not entitled to limit the scope of the exemption provided for
in art 13B(f) of the Sixth Directive to the lawful provision of roulette games.

Question 2
32.
By its second question the national court asks whether, in the case of the transactions in issue in the
main proceedings, the taxable amount is the amount retained by the operator during a tax period. This
question arises only if the view is taken that Germany is entitled to tax the transactions although it
exempts equivalent transactions effected by licensed public casinos.
33.
The purpose of the national courts question is to determine whether the courts ruling in Glawe
concerning slot machine transactions applies to the roulette transactions in issue here. The slot machines
in question in that case were located in bars and restaurants. Players activated a machine by inserting
coins which either entered its winnings reserve in order to be subsequently paid out as winnings or, if the
reserve was full, entered the machines cash box, which was periodically emptied by the operator. The
machines were required by law to pay out at least 60% of the stakes inserted into the machines as
winnings. The German government argued that the taxable amount consisted of the total stakes inserted
into the machines by players, including those stakes which were subsequently paid out as winnings.
34.
The court, following my opinion, held that the taxable amount for the purposes of art 11A(1)(a) of the
Sixth Directive did not include the statutorily prescribed proportion of the total stakes inserted
corresponding to the winnings paid out. Referring to its judgments in Boots Co plc v Customs and Excise
Comrs Case C-126/88 [1990] STC 387, [1990] ECR I-1235 and Naturally Yours Cosmetics Ltd v Customs
and Excise Comrs Case 230/87 [1988] STC 879, [1988] ECR 6365, the court reasoned that the taxable
amount for VAT purposes was the consideration actually received by a taxable person. The consideration
actually received by an operator of slot machines was limited to the proportion of the stakes which he
could actually take for himself, ie the coins which entered the machines cash box.
35.
In the present proceedings the German government asks the court to reconsider its ruling in Glawe. In
order to explain its criticisms of the ruling it may be helpful to set out a short passage from my opinion in
the case (see [1994] STC 543, [1994] ECR I-1679 (paras 1921)):
(19) In my view the consideration which the operator obtains for his services for the purposes of
art 11A(1)(a) is limited to the amounts which he empties from the machine. That is apparent from
an analysis of the transactions in issue and of other forms of gambling.
(20) Whilst gambling for money entails expenditure by gamblers, it does not in its simplest form
give rise to consumption of goods or services. Suppose, for example, that A enters into a private
bet with B, both placing their respective bets on the table. A wins the bet and collects the money on
575 the table. In such a case it would be absurd to suggest that A and B provide services to each
other for a consideration equal to the amount of their respective bets. The placing of the bets and
collection of the winnings is simply part of the gambling transaction. The placing of the bets,
although it involves the outlay of money, does not constitute the consumption of goods or services
which is the taxable event under the VAT system.
(21) Commercial gambling is different in so far as the person organising the gambling arranges
matters in such a way that on average his winnings are sufficient to meet his costs in organising the
 gambling and to provide him with a reasonable profit. For example, a bookmaker will set the odds
for bets on horse races at a level intended to ensure that he makes an overall profit on bets placed.
To that extent the person organising the gambling may perhaps be regarded as not only taking part
in the gambling himself but also providing a service to the other gamblers consisting in organising
the gambling. On that view his reward for that service would not, however, be the total amount of
the bets placed by gamblers. As already stated, the placing of bets and payment of winnings form
the nucleus of the gambling activity. The service provided by the organiser consists in providing the
framework within which that activity can take place, his reward for that service being the surplus of
winnings which he arranges for himself, together with any specific commission which he may
charge.
36.
In the present proceedings the German government rejects the premise that the nucleus of the
gambling activity does not involve consumption of goods or services. It considers that the gambling
activity itself involves an exchange of services. An organiser of commercial gambling does not merely
provide the framework for the gambling activity but also participates in the game himself. The player does
not, according to the German government, furnish consideration to the organiser with the mere aim of
participating in the game; otherwise it would suffice for the organiser to charge an entrance fee. What is
important to the player is the provision of the chance of winning.
37.
The German government points out that the fact that art 13B(f) provides for the possibility of
exemption necessarily implies that games of chance may be subject to the tax. According to the ruling in
Glawe, the taxable amount is nil, effectively removing gambling from the scope of the tax: unless the
organiser makes a net gain.
38.
The German government adds that its analysis is consistent with the principle underlying the Sixth
Directive that VAT is imposed on individual transactions. Under the ruling in Glawe, according to which the
taxable amount consists of the total of net receipts less winnings paid out over a period, the reference to
specific transactions is lost. Moreover, under the ruling in Glawe the consideration for the provision of the
framework for a game is the operators gross margin, which is inconsistent with the principle that VAT is
payable on a taxable persons turnover.
39.
In the alternative the German government contends that the transactions in Glawe are distinguishable
from those in the present case. It points out that, in contrast to the situation in Glawe, players winnings do
not correspond to a fixed percentage laid down by law. Nor do the stakes to be retained by the organiser
enter a separate cash box. The chips placed on the table cannot therefore be divided into two separate
categories, namely winnings and the organisers turnover.
576
40.
While not asking the court to reconsider its ruling in Glawe, the Commission does not think the same
analysis can be applied to roulette. It contends that the taxable amount consists of the chips purchased
by players. By purchasing chips a player gains access to the casino, its infrastructure, its particular
atmosphere and the tables. The range of possibilities offered by a casino is infinitely greater and more
varied than that offered by a gaming machine fixed on the wall of a restaurant.
41.
The United Kingdom, on the other hand, takes the view that the courts ruling in Glawe was correct
and should be extended to roulette. It emphasises however that gambling is a special case and that the
ruling is not applicable to other categories of transaction.
42.
In considering this issue it is important above all to have regard to the basic principles underlying the
common VAT system, as set out in particular in art 2 of Council Directive (EEC) 67/227 on the
harmonisation of legislation of member states concerning turnover taxes (see OJ S edn 1967 p 4) (the
First Directive). That provision states that the principle of the common VAT system
involves the application to goods and services of a general tax on consumption exactly
proportional to the price of goods and services, whatever the number of transactions which take
place in the production and distribution process before the stage at which tax is charged.
43.
The normal functioning of the VAT system in conformity with that provision may be illustrated by the
following example. Suppose that during a particular period a manufacturer sells goods for DM 3m, plus
DM 420,000 VAT (calculated at the rate of 14% applicable in Germany at the material time). The
manufacturers VAT-inclusive takings are therefore DM 3,420,000, of which he is obliged to pay DM
420,000 (less any VAT already paid on his purchases) to the tax authorities. The remaining DM 3m is
available to cover his profit margin, the cost components of his supplies and any other taxes for which he
may be liable (eg profits tax). The proportion of the price of his goods, ie of his total takings, represented
by VAT corresponds to the 14% VAT rate applicable in Germany (DM 420,000/DM 3m). The tax is
therefore exactly proportional to the price of [his] goods as required by art 2 of the First Directive 3.
3
For the purpose of this example I shall ignore the small amounts of hidden VAT which may have been passed on to him
in exempt purchases of goods or services (eg insurances).
44.
The application of VAT to gambling transactions is admittedly less straightforward. Indeed the reason
for the exemption in art 13B(f) is that such transactions are better suited to other forms of taxation.
Nevertheless, despite the inherent difficulties, the ruling in Glawe provides in my view the basis for
applying VAT to such transactions in a manner which is consistent with the basic principles of the VAT
system as set out and illustrated above. It seems to me that the ruling is equally applicable to roulette
transactions.
45.
If the ruling in Glawe is applied to the present case, the effect will be that VAT is payable on the
organisers net takings (after payment of all winnings) over a given period. That analysis produces results
most closely resembling those applicable in the case of more typical transactions as in the example given
above. It is ultimately only the amount retained by the organiser after payment of winnings which is
available to cover his profit margin, the costs of running the gambling establishment, the VAT and other
taxes which may be payable on his activities. It is that amount which may therefore be equated with the
577 manufacturers (VAT-inclusive) turnover from the sale of his goods. By calculating the tax by
reference to that amount the tax remains exactly proportional to the organisers turnover.
46.
The German governments argument that such an analysis removes gambling from the scope of the
tax ignores the fact that organisers of commercial gambling of the type in issue arrange the odds in such
a way as to ensure that they make a profit in the long run. Nor do I share the German governments view
that the ruling in Glawe conflicts with the principle that VAT is imposed on individual transactions. In my
opinion in Glawe [1994] STC 543, [1994] ECR I-1679 (para 29) I noted:
each stake must be regarded as consisting of two components. One component is the price
paid for the services provided by the operator (including the VAT payable on that amount). The
remainder of the stake may be regarded as an amount contributed to the common pool available to
be paid out as winnings. Over a given period, those components will correspond to the amounts
collected respectively by the cash box and the reserve of the machine.
47.
The same applies to the present case. As a matter of legal analysis each chip placed on the table
comprises two components: (a) the wager; and (b) the consideration for the organisers service, ie the
price paid by players for the right to participate in the game and obtain the chance of winning. That price,
consisting in the advantage which the house reserves to itself by virtue of the odds being in its favour, can
be calculated precisely and is a standard percentage varying according to the version of roulette played. It
is paid by each player each time he places a chip on the table. It would be perfectly possible for an
organiser to separate the two components by eliminating the advantage for the house and replacing it
with a separate charge to cover his costs and provide him with a profit.
48.
For similar reasons the German governments argument that the ruling in Glawe entails taxation of a
taxable persons gross margin rather than his turnover contrary to the basic principles of the Sixth
Directive is also misconceived. The organisers turnover is limited to that proportion of each chip
representing the price of the organisers service.
49.
In practice individual calculations based on each chip placed on the table are unnecessary. The total of
the amounts received by way of consideration for individual transactions corresponds to the organisers
net takings (after payment of winnings) during a given period. Over a period the organisers net takings
necessarily correspond to the advantage which he reserves to himself. The fact that there is in practice an
easier method of determining the taxable amount does not however mean that tax is not levied on
individual transactions. I therefore do not accept the Commissions view that such an analysis is
theoretically unsound. On the contrary it seems to me that it is an illustration of how a theoretically sound
solution is often easier to apply in practice.
50.
It is, I think, instructive to examine by way of contrast the consequences of the other interpretations
which have been suggested in these proceedings. It appears from the order for reference that the
Finanzamts analysis is based on a judgment of the Finanzgericht Dsseldorf of 29 January 1986. The
Finanzgericht reasoned that, if a player lost DM 50 during a session, the price which he paid for the
chance of winning was DM 50. If, on the other hand, he won DM 100 during a session, he paid nothing for
the chance of winning. Since the operator had no 578 record of players net winnings or losses at each
session, it was necessary for the Finanzgericht to lay down a method for estimating his taxable turnover.
After undertaking a calculation of probabilities the Finanzgericht concluded that the estimate was to be
arrived at by multiplying the operators net takings by six.
51.
At the hearing the German government explained that it did not share the Finanzamts analysis of the
transaction. In keeping with its analysis in Glawe, the German government considers that the taxable
amount consists of the total chips placed on the table by a player in each game, no account being taken
of his winnings.
52.
The Commission on the other hand takes the view that the taxable amount consists of the chips
purchased, again no account being taken of winnings.
53.
 The basic flaw in all those analyses is that they produce a tax burden which is not proportional to the
organisers real turnover. The essential reason for that is that a large part of the amount staked is repaid
to players as winnings yet is treated as the organisers turnover for VAT purposes. The result is an
artificial inflation of his turnover. Such analyses ignore the fact that the nucleus of a gambling activity,
although involving the use of money, does not involve the consumption of goods or services.
54.
The effect of the Finanzamts method may be seen from the figures given in the order for reference; for
example, on the basis of Mr Fischers actual takings of DM 344,880 for 1987 the Finanzamt calculated a
taxable amount of DM 2,069,280, on which it assessed VAT of DM 289,69920 at the rate of 14%. The
result of that method, however, is to impose an effective rate of tax on Mr Fischers actual takings of no
less than 84% (DM 289,69920/DM 344,880).
55.
The German governments analysis would undoubtedly lead to an even higher tax burden. Suppose
that a player purchases ten chips for DM 5 each. He stakes two chips in each of the first five games. In
the first four games he loses. In the fifth game he wins, restoring his total chips to ten. He then plays
another five games, again staking two chips in each game. He loses the sixth to ninth games but then
wins the tenth game, restoring the number of his chips to eight. On the German governments view tax
would be payable on the total chips staked, ie DM 100 (20 chips of DM 5 each). Yet the organisers actual
takings from the series of games would be only DM 10 (two chips of DM 5 each). That amount alone is
the (VAT-inclusive) turnover available to cover his costs, his profit margin, the VAT and any other taxes
which may be payable. Yet on the German governments analysis his VAT liability alone would amount to
DM 1228 (DM 100 x 14/114). His actual takings would therefore be insufficient even to meet his VAT
liability.
56.
Moreover, on both the German governments and the Finanzamts view the taxable amount can only
be estimated since it would plainly be impracticable for operators to keep a record of every chip staked or
the results of each session at the table. The correct multiplication factor to be applied has evidently been
the subject of some debate in Germany. From the taxable persons viewpoint the factor of six adopted by
the Finanzgericht Dsseldorf compares very favourably with the factor of 25 which had been suggested
by the tax authorities in the proceedings before that court. At the hearing in the present proceedings the
German government was unable to state the method which it would use to arrive at an estimate of Mr
Fischers turnover.
57.
The Commissions analysis would probably result in a taxable amount somewhat lower than that
resulting from the German governments analysis 579 since the taxable amount would not include chips
won by players and placed as stakes in further games. Moreover, taxable persons could perhaps
reasonably be required to keep a record of chips sold, making the use of estimates unnecessary.
However, there is no reason to suppose that the method proposed by the Commission, which takes no
account of winnings paid to players, would result in the imposition of a tax burden which was proportional
to the organisers actual takings from his activities.
58.
I consider therefore that, on the assumption that the national courts second question calls for a reply,
the court should uphold its ruling in Glawe and rule that it also applies to roulette transactions such as
those in issue in the main proceedings. I do not think the particular workings of the slot machines in
question in Glawe or the fact that the minimum winnings paid out were fixed by law were critical to the
courts ruling. The courts essential concern was to provide an interpretation which would ensure that a
taxable persons VAT liability was proportional to his actual turnover.
59.
I should, however, finally emphasise that, for reasons already mentioned, gambling transactions of the
kind in issue which involve betting with money are something of a special case, and it must not be thought
that the foregoing analysis can necessarily be extended to other transactions.
60.
In view of the answer which I propose to the second question, the national courts third question does
not arise.

Conclusion
61.
Accordingly, I am of the opinion that the Court of Justice should reply as follows to the questions put by
the Finanzgericht Baden-Wrttemberg:
Article 2(1) of the Sixth Directive is to be interpreted as subjecting to VAT the unlawful provision
of roulette games. A member state is not entitled to limit the scope of the exemption provided for in
art 13B(f) of the Sixth Directive to the lawful provision of such games.

11 June 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
 By order of 21 August 1995, received at the Court of Justice of the European Communities on 25
August 1995, the Finanzgericht (the finance court) Baden-Wrttemberg, Freiburg, referred to the Court of
Justice for a preliminary ruling under art 177 of the EC Treaty three questions on the interpretation of
Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover
taxescommon system of value added tax: uniform basis of assessment (the Sixth Directive) (OJ 1977
L145 p 1).
2.
Those questions were raised in proceedings between Mr Fischer and the Finanzamt (the tax office),
Donaueschingen, concerning the payment of turnover tax on unlawful and punishable games of chance.

Relevant provisions
3.
Article 2 of the Sixth Directive states:
The following shall be subject to value added tax: (1) the supply of goods or services effected
for consideration within the territory of the country by a taxable person acting as such
580
4.
Under art 11A(1)(a), the taxable amount in respect of supplies of goods and services is generally
everything which constitutes the consideration which has been or is to be obtained by the supplier from
the purchaser, the customer or a third party for such supplies including subsidies directly linked to the
price of such supplies.
5.
Article 13B provides:
Without prejudice to other Community provisions, Member States shall exempt the following
under conditions which they shall lay down for the purpose of ensuring the correct and
straightforward application of the exemptions and of preventing any possible evasion, avoidance or
abuse (f) betting, lotteries and other forms of gambling, subject to conditions and limitations laid
down by each Member State
6.
Article 33, in the version in force at the material time, stated that, without prejudice to other Community
provisions, the provisions of the Sixth Directive were not to prevent a member state from maintaining or
introducing taxes on insurance contracts, taxes on betting and gambling, excise duties, stamp duties and,
more generally, any taxes, duties or charges which could not be characterised as turnover taxes.
7.
In German law, under para 1(1) of the Umsatzsteuergesetz (the UStG) (the Law on Turnover Tax,
Bundesgesetzblatt) turnover tax is payable on supplies of goods and services effected for consideration in
the relevant tax area by a trader in the course of his business.
8.
Paragraph 4(9)(b) of the UStG exempts from the tax turnover covered by the Law on Betting and
Lotteries and gaming turnover of licensed public casinos.
9.
Under para 284 of the German Criminal Code, any person who organises games of chance open to
the public without official authorisation is to be liable to a fine or a term of imprisonment not exceeding two
years. Games of chance organised in clubs or private societies where games of chance are routinely
organised are also to be treated as open to the public.
10.
In accordance with para 40 of the Abgabenordnung (the 1977 tax code), it is irrelevant for taxation
purposes whether an act which wholly or partly meets the criteria for liability to tax under tax legislation is
contrary to a statutory requirement or prohibition or to public morality.

Facts of the main proceedings

11.
Between 1987 and 1989 Mr Fischer organised games of chance at several locations in Germany. He
had the official permit needed to operate a game of skill calling for use of a machine called Roulette Opta
II. However, he departed from the terms of that permit to such an extent that in the end the game
resembled the game of roulette as played in duly licensed public casinos.
12.
The equipment used by Mr Fischer consisted of a disk with spaces numbered from 1 to 24 and
coloured red or black together with two spaces marked 0 and X. The object of the game was to predict,
by placing chips on the appropriate squares of the gaming table, where the ball thrown by a croupier
would come to rest. The players purchased chips with a value of DM 5 each and could stake one or more
of them on each game. If the ball came to rest on the number, the line or the colour on which the stake
had been placed, the croupier paid out 24 times, 12 times and twice the stake respectively.
581
13.
Winnings were paid in the form of chips immediately after each game and the other chips staked were
collected by the croupier. Players who wished to stop playing could cash in their remaining chips.
14.
After carrying out an investigation, the Finanzamt took the view that Mr Fischer supplied services
subject to turnover tax, consisting of giving players opportunities of winning. According to the Finanzamt,
the basis of assessment should have been the amount of the stakes placed by players, less the winnings
paid out. However, as Mr Fischer had not recorded those transactions, it assessed the taxable amount on
the basis of a probability calculus, multiplying the takings by a factor of six.

The questions referred for a preliminary ruling

15.
The Finanzgericht Baden-Wrttemberg, before which Mr Fischer brought proceedings, was, first,
uncertain whether, in the light of Mol v Inspecteur der Invoerrechten en Accijnzen Case 269/86 [1988]
ECR 3627 and Vereniging Happy Family Rustenburgerstraat v Inspecteur der Omzetbelasting Case
289/86 [1988] ECR 3655, the transactions at issue in the main proceedings actually constituted supplies
of services within the meaning of art 2(1) of the Sixth Directive or whether, as unlawful activities, they fell
outside the scope of that directive. Secondly, if the services at issue were taxable under the Sixth
Directive, it raised the question whether the method of calculating the taxable amount established by the
court in H J Glawe Spiel-und Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt
Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994] STC 543, [1994] ECR I-1679, which involved the
taking into account of solely the net takings after payment of the players winnings, applied in the case
before it. Thirdly, if that was not the case, another method of calculating the taxable amount needed to be
established.
16.
The national court therefore decided to stay proceedings and refer the following questions to the Court
of Justice for a preliminary ruling:
(1) Is Article 2(1) of the Sixth Directive to be interpreted as meaning that services which the
organiser of unlawful and punishable games of chance provides to the players are not taxable?
(2) If question 1 is to be answered in the negative: is Article 11A(1)(a) of the Sixth Directive to be
interpreted as meaning that, in the case of unlawful gaming in the form of roulette, the basis of
assessment for the operators services to the players is the amount retained by the operator during
a tax period?
(3) If question 2 is to be answered in the negative: how is the basis of assessment to be
determined in cases described under questions 1 and 2?

Question 1
17.
For the purpose of providing the national court with a helpful answer, the first question must be
understood as seeking to determine whether the unlawful operation of a game of chance, in this case
roulette, falls within the scope of the Sixth Directive and whether a member state may impose value
added tax (VAT) on that activity when the corresponding activity carried on by a licensed public casino is
exempted.
18.
It should first be noted that the Sixth Directive expressly refers to forms of gambling in art 13B(f),
where it provides for their exemption, and in art 33, where it states that its provisions are not to prevent
the member states from maintaining 582 or introducing taxes on betting and gambling. It is thus clear that
such transactions do not, as such, fall outside the Sixth Directive.
19.
It is appropriate, next, to consider the doubts expressed by the national court concerning the possibility
of imposing VAT on unlawful activities. In Einberger v Hauptzollamt Freiburg Case 294/82 [1984] ECR
1177 (paras 1920), Mol [1988] ECR 3627 (para 15), Happy Family [1988] ECR 3655 (para 17) and
Witzemann v Hauptzollamt Mnchen-Mitte Case 343/89 [1993] STC 108, [1990] ECR I-4477 (para 19),
the court held that illegal imports or supplies of narcotic drugs or counterfeit currency, whose release into
the economic and commercial channels of the Community was by definition precluded and which could
give rise only to penalties under the criminal law, were wholly alien to the provisions of the Sixth Directive
and did not give rise to any VAT debt.
20.
As the court pointed out in Lange v Finanzamt Frstenfeldbruck Case C-111/92 [1997] STC 564,
[1993] ECR I-4677 (para 12), that principle relates only to products which, because of their special
characteristics, may not be marketed or incorporated into economic channels.
21.
By contrast, outside those cases where all competition between a lawful economic sector and an
unlawful sector is ruled out, the principle of fiscal neutrality precludes a generalised distinction from being
drawn in the levying of VAT between unlawful and lawful transactions. The court has accordingly held that
a ban on the export of certain products to specific destinations because they might be used for strategic
purposes cannot in itself be sufficient to remove such exports from the scope of the Sixth Directive (see
Lange [1997] STC 564, [1993] ECR I-4677 (paras 1617)).
22.
The foregoing considerations relating to the import or supply of goods apply equally to the supply of
services such as the organisation of games of chance. Those games, and roulette in particular, are
lawfully played in a number of member states. Since the unlawful transactions at issue in the main
proceedings are in competition with lawful activities, the principle of fiscal neutrality precludes their being
treated differently as regards VAT.
23.
The unlawful operation of games of chance therefore falls within the scope of the Sixth Directive.
24.
It must, finally, be ascertained whether, as the UK government asserted in its written observations, art
13B(f) of the Sixth Directive prohibits the member states from imposing VAT on unlawful games of chance
when the corresponding activity lawfully carried on in duly licensed public casinos is exempted.
25.
It is clear from the very wording of that provision that gambling is in principle to be exempted from VAT.
The member states retain the power to lay down the conditions and limitations of that exemption.
26.
The Commission has maintained, however, that art 13B(f) does not involve an absolute prohibition on
the taxation of games of chance. At the hearing, the UK government departed from the view which it had
taken in its written observations and maintained that when the member states lay down the conditions
and limitations of the exemption they are entitled to require that the transactions in question take place in
duly licensed public casinos.
27.
In that regard, it must be pointed out that the exemptions provided for by art 13B are to be applied in
accordance with the principle of fiscal neutrality inherent in the common system of VAT (see, to that effect,
EC Commission v Italy Case C-45/95 [1997] STC 1062, [1997] ECR I-3605 (para 15)). That requirement
also applies when the member states exercise their power under art 13B(f) to lay down the conditions and
limitations of the exemption. In according that power 583 to the member states, the Community legislature
did not authorise them to undermine the principle of fiscal neutrality which underlies the Sixth Directive.
28.
As pointed out in para 21 of this judgment, it is clear from the judgment in Lange that the principle of
fiscal neutrality precludes a generalised distinction from being drawn in the levying of VAT between
unlawful and lawful transactions. It follows that member states cannot reserve the exemption solely to
lawful games of chance.
29.
It cannot be contended, as the German government has done, that the conditions in which lawful
games take place are not comparable to those pertaining in the case of unlawful games on the ground
that licensed casinos are subject to a levy calculated on the basis of their profits.
30.
First, the common system of VAT would be distorted if member states could adjust its application on
the basis that there are other, unharmonised, taxes. Secondly, as the German government itself
acknowledged at the hearing, there is nothing to prevent levies of the same kind as those payable by
licensed casinos from also being imposed on organisers of unlawful games of chance.
31.
The answer to the first question must therefore be that the unlawful operation of a game of chance, in
the event roulette, falls within the scope of the Sixth Directive. Article 13B(f) of that directive must be
interpreted as meaning that a member state may not impose VAT on that activity when the corresponding
activity carried on by a licensed public casino is exempted.

Questions 2 and 3
32.
In view of the answer given to the first question, there is no need to reply to the second and third
questions asked by the national court.

Costs
33.
The costs incurred by the German and UK governments and by the European Commission, which
have submitted observations to the Court of Justice, are not recoverable. Since these proceedings are,
for the parties to the main proceedings, a step in the action pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the questions referred to it by
the Finanzgericht Baden-Wrttemberg, Freiburg, hereby rules: the unlawful operation of a game of
chance, in the event roulette, falls within the scope of the Council Directive (EEC) 77/388 on the
harmonisation of the laws of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment. Article 13B(f) of that directive must be interpreted as meaning
that a member state may not impose VAT on that activity when the corresponding activity carried on by a
licensed public casino is exempted.

584

[1998] All ER (EC) 585

 Imperial Chemical Industries plc v Colmer (Inspector of Taxes)
(Case C-264/96)

EUROPEAN COMMUNITY; TAXATION: TAXATION; Corporation Tax

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), RAGNEMALM, WATHELET (RAPPORTEUR) AND
SCHINTGEN (PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, MURRAY,
EDWARD, JANN, SEVN AND IOANNOU ADVOCATE GENERAL TESAURO
14 OCTOBER, 16 DECEMBER 1997, 16 JULY 1998
European Community Freedom of establishment Discrimination on grounds of nationality
Corporation tax Consortium relief Surrender to one company by another in same group of tax relief of
trading losses Relief subject to residence requirement Whether compatible with Community law EC
Treaty, arts 52, 56, 58.
European Community National legislation Construction Construction in light of Community
obligations Scope of duty of co-operation where offending national legislation applicable in Community
and non-Community context EC Treaty, art 5.

ICI was one of a consortium of companies which owned a holding company, H Ltd. H Ltd held shares in
23 trading companies, four of which were established in the United Kingdom, six in other member states
and thirteen in non-member states. ICI sought to set a proportion of the losses of one of the UK trading
companies against its own chargeable profits for the corresponding periods by way of tax relief, pursuant
to ss 258 to 2641 of the Income and Corporation Taxes Act 1970. The Inland Revenue refused ICIs
application for tax relief on the ground that H Ltd did not constitute a holding company within the
meaning of s 258(5)(b) of the 1970 Act, since the majority of its subsidiaries (19 out of 23) were not
bodies corporate resident in the United Kingdom as required by s 258(7). ICI contested the refusal,
contending, principally, that the residence requirement amounted to a restriction, in the form of a
discriminatory tax regime, on the freedom of establishment for companies and firms and therefore
infringed arts 52 and 582 of the EC Treaty. After a series of appeals, the House of Lords stayed the
proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling the
questions: (i) whether art 52 precluded member states from making a particular form of tax relief subject
to a residence requirement; and (ii) whether, in the event that such legislation was incompatible with
Community law, the national court was obliged, under art 53, to disapply that legislation or to construe it in
a way conforming with Community law, where the holding 585 company controlled mainly subsidiaries
having their seat in non-member countries.
1
Sections 258 to 264, so far as material are set out at p 597 j to p 598 f, post
2
Articles 52 and 58, so far as material, provide: (52) restrictions on freedom of establishment of nationals of a
Member State in the territory of another Member State shall be abolished Such abolition shall also apply to
restrictions on the setting-up of agencies, branches or subsidiaries (58) Companies or firms shall, for the
purposes of this Chapter, be treated in the same way as natural persons who are nationals of Member States.
3
Article 5, so far as material, provided: Member States shall take all appropriate measures to ensure fulfilment of the
obligations arising out of action taken by the institutions of the Community.

Held (1) Article 52 of the Treaty precluded legislation of a member state which, as in the instant case,
made a particular form of tax relief subject to the requirement that the holding companys business
consisted wholly or mainly in the holding of shares in subsidiaries that were established in the member
states concerned. Moreover, the need to maintain the cohesion of tax systems did not, in the instant case,
provide sufficient justification for maintaining rules restricting fundamental freedoms, since there was no
direct link between the consortium relief granted for losses incurred by a resident subsidiary and the
taxation of profits made by non-resident subsidiaries (see p 602 f to h and p 603 h, post); Bachmann v
Belgium Case C-204/90 [1992] ECR I-249 distinguished.
(2) The difference of treatment applied according to whether or not the business of the holding
company consisted wholly or mainly in holding shares in subsidiaries having their seat in non-member
countries lay outside the scope of Community law. Accordingly, art 5 of the Treaty did not require the
national court to interpret its legislation in conformity with Community law, or to disapply the legislation in
a situation falling outside the scope of Community law. However, where a particular provision had to be
disapplied in a situation covered by Community law, but that same provision could remain applicable to a
situation not so covered, it was for the competent body of the state concerned to remove that legal
uncertainty in so far as it might affect rights deriving from Community rules (see p 603 c to e j, post).

Notes
For the EC Treaty, arts 5, 52 and 58, see 50 Halsburys Statutes (4th edn) 267, 285 and 287.

Cases cited
Asscher v Staatssecretaris van Financin Case C-107/94 [1996] All ER (EC) 757, [1996] ECR I-3089,
ECJ.
Bachmann v Belgium Case C-204/90 [1992] ECR I-249.
Bond van Adverteerders v Netherlands Case 352/85 [1988] ECR 2085.
BP Supergas Anonimos Etaira Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece Case C-62/93
[1995] All ER (EC) 684, [1995] ECR I-1883, ECJ.
Decker v Caisse de Maladie des Employs Privs Case C-120/95 (1997) Transcript (opinion), 16
September, ECJ.
EC Commission v Belgium Case C-300/90 [1992] ECR I-305.
EC Commission v France Case 270/83 [1986] ECR 273.
EC Commission v UK Case C-246/89 [1991] ECR I-4585.
Enderby v Frenchay Health Authority Case C-127/92 [1994] 1 All ER 495, [1994] ICR 112, [1993] ECR I-
5535, ECJ.
Eurico Italia Srl v Ente Nazionale Risi Joined cases C-332333/92 and C-335/92 [1994] ECR I-711.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Furlanis Costruzioni General SpA v Azienda Nazionale Autonoma Strade (ANAS) C-143/94 [1995] ECR I-
3633.
Futura Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471.
586
Gebhard v Consiglio dellOrdine degli Avvocati e Procuratori di Milano Case C-55/94 [1996] All ER (EC)
189, [1995] ECR I-4165, ECJ.
Grado Case C-291/97 (1997) Transcript, 9 October, ECJ.
Kohl v Union des Caisses de Maladie Case C-158/95 (1997) Transcript (opinion), 16 September, ECJ.
MacLachlan v Caisse Nationale dAssurance Vieillesse des Travailleurs Salaris de la Rgion dIle-de-
France (CNAVTS) Case C-146/93 [1994] ECR I-3229.
Monin AutomobilesMaison du Deux-Roues Case C-428/93 [1994] ECR I-1707.
Pigs Marketing Board v Redmond Case 83/78 [1978] ECR 2347.
R v HM Treasury, ex p Daily Mail Case 81/87 [1989] 1 All ER 328, [1989] 1 QB 446. [1989] 2 WLR 908,
[1988] ECR 5483, ECJ.
R v IRC, ex p Commerzbank AG Case C-330/91 [1993] ECR I-4017.
Safir v Skattemyndigheten i Darlarnas Ln (formerly Skattemyndigheten i Kopparbergs Ln) Case C-
118/96 (1997) Transcript (opinion), 23 September, ECJ.
Stichting Collectieve Antennevoorziening Gouda v Commissariaat voor de Media Case C-288/89 [1991]
ECR I-4007.
Svensson v Ministre du Logement et de lUrbanisme Case C-484/93 [1995] ECR I-3955.
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER
(EC) 97, [1995] ECR I-4921, ECJ.
Unit Socio-Sanitarua Locale No 47 di Biella (USSL) v Istituto Nazionale per LAssicurazione Contro gli
Infortuni sul Lavaro (INAIL) Case C-134/95 [1997] ECR I-195.
Wielockx v Inspecteur der Directe Belastingen Case C-80/94 [1995] All ER (EC) 769, [1995] ECR I-2493,
ECJ.

Reference
By order of 24 July 1996, the House of Lords referred to the Court of Justice of the European
Communities for a preliminary ruling under art 177 of the EC Treaty two questions (set out at p 599 f to j,
post) on the interpretation of arts 5 and 52 of the Treaty. Those questions were raised in proceedings
between Imperial Chemical Industries plc (ICI) and the Inland Revenue concerning the latters refusal to
grant to ICI tax relief in respect of trading losses incurred by a subsidiary of the holding company
beneficially owned by ICI through a consortium. Written observations were submitted on behalf of: ICI, by
P Whiteman QC and C Vajda, Barrister, instructed by Hammond Suddards, Solicitors; the United
Kingdom government, by J E Collins, Assistant Treasury Solicitor, acting as agent, with D Wyatt QC and
R Singh, Barrister; and the European Commission, by P Oliver and H Michard, of its Legal Service, acting
as agents. Oral observations were made by ICI, the UK government and the Commission. The language
of the case was English. The facts are set out in the opinion of the Advocate General.

16 December 1997.

The Advocate General (G Tesauro)

delivered the following opinion (translated from the Italian).
1.
The points at issue in these proceedings are, first, the compatibility with art 52 of the EC Treaty of
domestic legislation which makes a particular form of tax relief available to companies belonging to a
consortium subject to the condition that, where the consortium controls a holding company, most of the
subsidiaries thereof are resident in the national territory; and, secondly, in the event that such legislation
is incompatible, the importance and extent of the national courts obligation under art 5 of the Treaty to
adopt an interpretation 587 which is consistent with Community law. The reference has been made by the
House of Lords, and the relevant legislation is that of the United Kingdom.
THE NATIONAL LEGISLATION
2.
The legislation applicable in the present case is to be found in ss 258 to 264 of the Income and
Corporation Taxes Act 1970, which have since been replaced by similar provisions in the Income and
Corporation Taxes Act 1988.
Those provisions govern, inter alia, consortium relief 4. This essentially enables a company which is a
member of a consortium to use losses incurred by subsidiaries controlled through a holding company to
offset tax on its profits. Thus, pursuant to the legislation in question, the company belonging to the
consortium may set losses incurred by a subsidiary against its chargeable profitsin proportion to the
size of its shareholdingfor the purposes of computing tax liability. The reasons why the legislation
makes this option available have been explained in the course of the proceedings. However, they need
not be considered here, save in order to assess whether the domestic legislation gives rise to a restriction
on freedom of establishment, contrary to the prohibition laid down in art 52 of the Treaty, and, if so,
whether that restriction is capable of being justified.
4
The expression consortium is used here to refer to an agreement between undertakings to form a joint venture to
operate at international level.

3.
Specifically, pursuant to s 258(1) of the 1970 Act, relief to which companies are entitled for trading
losses may be surrendered by a company which is a member of a group of companies (the surrendering
company) to another company in the same group (the claimant company). Under s 258(2), group relief is
also available in situations involving consortia. For instance, it is available where one of the companies
involved is a member of a consortium and the other is a company controlled by a holding company which
is in turn owned by a consortium5. In accordance with s 259(1) and (8)(a) of the Act, in cases where the
claimant company is a member of a consortium, only a fraction of the losses incurred by the surrendering
company may be set off, that fraction being equal to the claimant companys share in the consortium.
5
Pursuant to s 258(8), a company is owned by a consortium if three-quarters or more of the ordinary share capital of the
company is beneficially owned between them by companies of which none beneficially owns less than one-twentieth of
that capital, and those companies are called the members of the consortium.

The availability of consortium relief is also conditional on the company owned by the consortium
being a holding company as defined in s 258(5)(b) of the Act, namely a company the business of which
consists wholly or mainly in the holding of shares or securities of companies which are its 90 per cent
subsidiaries, and which are trading companies.
4.
Lastly, s 258(7) provides that references in this and the following sections of this Chapter to a
company apply only to bodies corporate resident in the United Kingdom. This is the provision whose
interpretation and application have given rise to these proceedings.

THE FACTS AND THE QUESTIONS REFERRED

5.
Coopers Animal Health (Holdings) Ltd (Holdings) was set up on 17 May 1984, its shares being
beneficially owned by a consortium formed by Wellcome Foundation Ltd and Imperial Chemical Industries
plc (ICI) which, respectively, have a 51% and a 49% interest in Holdings. The latter carries on no
business save that of holding shares in subsidiaries. Of its 23 subsidiaries, only four are resident 588 in
the United Kingdom, six being resident in other member states and the remaining thirteen in non-member
countries.
6.
One of the companies controlled by Holdings and resident in the United Kingdom is Coopers Animal
Health Ltd (CAH), which incurred considerable losses, particularly in the accounting periods ending,
respectively, in 1985, 1986 and 1987. ICI accordingly applied to the Inland Revenue under s 258 of the
Act for relief in respect of 49% of CAHs losses (the fraction corresponding to ICIs shareholding in
Holdings).
The Inland Revenue refused to grant the relief sought, on the ground that, although all the companies
involved (ICI, Holdings and CAH) were resident in the United Kingdom, most of the companies controlled
by Holdings were resident abroad. In the light of s 258(7) of the Actaccording to the Inland Revenue
that fact precluded Holdings from meeting the requirements for recognition as a holding company and,
accordingly, for securing the related tax relief.
7.
ICI brought an action challenging that interpretation. Both the High Court and the Court of Appeal
upheld its claim owing to their adoption of a different interpretation of the relevant legislation and, in
particular, of s 258(7), from that proposed by the Inland Revenue. In brief, both courts took the view that
access to tax relief cannot be denied in cases such as this, where both the surrendering company and the
claimant company are resident in the United Kingdom. It was not intended that, whenever the term
company is used in the text of s 258 (including, that is to say, references to the holding company or the
subsidiaries), it must be read in conjunction with the reference to company in the opening words of s
258(7), which merely defines the companies which may take advantage of the relief provided for in that
section. Thus, according to that construction, companies resident in the United Kingdom cannot be
denied relief in respect of losses incurred by subsidiaries which are also resident there.
8.
On appeal by the Inland Revenue, however, the House of Lords in its capacity as court of last instance
upheld the tax authorities interpretation, thereby findingsolely on the basis of domestic lawthat ICI
was not entitled to the tax relief sought.
Before the House of Lords, however, ICI introduced a fresh argumentbased on Community lawto
challenge the denial of relief. In short, ICI claimed that the legislation at issueor at least the Inland
Revenues interpretation thereofwas incompatible with arts 52 and 58 of the EC Treaty in so far as the
requirement that most of the companies controlled by Holdings had to be resident in the United Kingdom
constituted a restriction (albeit an indirect one) on ICIs freedom of establishment and in particular of its
right to own shares through a holding company in subsidiary companies resident in another member
state. In any event, according to ICI, in view of the fact that the relevant legislation was open to two
possible interpretationsthat adopted by the courts at first and second instance, and that favoured by the
Inland Revenueart 5 of the Treaty placed the national court under a duty to choose the first, if it enabled
any conflict, actual or potential, with Community law to be avoided.
9.
Taking the view that an interpretation of the aforesaid provisions of Community law was necessary in
order to enable it to give judgment in the dispute before it, the House of Lords referred the following two
questions to the Court of Justice of the European Communities for a preliminary ruling:
(1) In a situation where: (i) a company (Company A) is resident in a Member State of the
European Union (ii) Company A is part of a consortium 589 with another company (Company B)
also resident in that Member State (iii) Company A and B jointly own a holding company (Company
C) also resident in the Member State (iv) Company C has a number of trading subsidiaries, which
are resident either in that Member State, other Member States of the European Union or elsewhere
in the world, and (v) Company A is precluded from being entitled to claim against its corporation tax
liability relief in respect of trading losses incurred by a trading subsidiary (also resident in that
Member State) of Company C because the national legislation, construed as a matter of national
law, required that the business of Company C should consist wholly or mainly in the holding of
shares in subsidiaries which are resident in that Member State: Does the requirement identified
at (v) constitute a restriction on the freedom of establishment under Article 52 of the EC Treaty? If
so, is such treatment nevertheless justified under Community law?
(2) If the requirement under (v) is an unjustified restriction under Community law, does Article 5
of the EC Treaty require a national court to interpret the relevant national legislation, so far as is
possible, so as to comply with Community law, even though neither Company A, Company B nor
Company C is itself seeking to exercise any rights under Community law, and even if an
interpretation of national legislation which would comply with Community law would have the effect
of giving relief where the business of Company C consisted mainly in the holding of shares in
subsidiaries established outside the EC/EEA? Or does Article 5 have the consequence only that
the national legislation, despite its interpretation, takes effect subject to the requirements of
Community law in a case where these requirements are in point?

QUESTION 1
10.
By its first question, the House of Lords asks the Court of Justice whether art 52 of the Treaty
precludes application of legislation such as that described above. In particular, on the assumption that the
interpretation advocated by the Inland Revenue is correct, the House of Lords asks whether the pre-
condition for tax reliefthat most of the subsidiaries controlled by the holding company must be resident
in the United Kingdomentails an unjustified restriction on the freedom of establishment guaranteed by
art 52.

Relevance
11.
First of all, I should point out that doubts have been expressed in the course of the proceedings as to
whether this question has any bearing on adjudication of the dispute in the main proceedings.
Specifically, the UK government maintained that even if the legislation at issue were found to entail a
restriction on freedom of establishment, incompatible with a proper interpretation of art 52, that would
have no relevance for the purposes of resolving the dispute in the main proceedings. ICI would in any
event be denied the tax relief provided for by the Act, since the majority of the companies controlled by
Holdings (as many as 13 out of 23) are resident, not in other member states of the Community, but
elsewhere.
12.
The European Commission has taken a different view. Given that the Court of Justice declines only in
exceptional circumstances to give a ruling on questions referred under art 177 of the Treaty, the
Commission has pointed out that, in the light of s 258(5), the House of Lords itself acknowledged that the
quantitative criterion is not the only test which can be applied in order to 590 evaluate the business of a
holding company; other yardsticks may be used, such as the turnover of the companies controlled.
According to the Commission, the reference in s 258(5)(b), read in conjunction with s 258(7), to business
consisting wholly or mainly in the holding of shares or securities of trading companies resident in the
United Kingdom is not open to only one interpretation. In any event, it is for the national court to decide
which test to apply, while the court must provide any guidance which would be of assistance in resolving
the dispute.
13.
 The first point I would make in that connection is that, according to established case law, it is for the
national court to assess the relevance of and the need for a preliminary ruling. Given its direct knowledge
of the facts of the case and the relevant points of law, that court is in the best position to gauge the
relevance of any questions concerning Community law raised in the dispute (see the judgments in Pigs
Marketing Board v Redmond Case 83/78 [1978] ECR 2347 (para 25) and MacLachlan v Caisse Nationale
dAssurance Vieillesse des Travailleurs Salaris de la Rgion dIle-de-France (CNAVTS) Case C-146/93
[1994] ECR I-3229 (para 20)). In principle, therefore, the Court of Justice considers itself bound to
answer, except in cases where the questions referred are purely hypothetical or where it is quite obvious
that the requested interpretation or ruling on the validity of a provision of Community law has no bearing
on the facts or purpose of the main action6.
6
See the order in Monin AutomobilesMaison du Deux-Roues Case C-428/93 [1994] ECR I-1707 and the judgments in
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER (EC) 97,
[1995] ECR I-4921 (para 61), Unit Socio-Sanitarua Locale No 47 di Biella (USSL) v Istituto Nazionale per
LAssicurazione Contro gli Infortuni sul Lavaro (INAIL) Case C-134/95 [1997] ECR I-195 (para 12) and Grado Case C-
291/97 (1997) ECJ Transcript, 9 October (para 12).

14.
However, although I am somewhat sceptical as to whether an interpretation of art 52 is really
necessary in order to resolve the dispute before the House of Lords, it must be said that the present case
does not fall within one of the admittedly exceptional situations described above. In particular, this case
does not to my mind exhibit the characteristics which have hitherto led the Court of Justice to regard a
reference as manifestly irrelevant to a decision on the dispute in the main proceedings. It is apparent from
the order for reference that the proper construction of s 258(5) of the Act remains an open question.
Indeed, it is only if the availability of tax relief is based on a quantitative criterion related to the residence
of subsidiaries that it could appear fruitless to seek an interpretation of art 52 since the majority of the
companies in question are established outside the Community. The position would be different if, as
contemplated in the order for reference itself, the national court were to use turnover as a criterion or
apply some other test. In that case, appraisal of the compatibility of the legislation in question with art 52
of the Treaty could well have a bearing on the decision as to whether or not ICI is entitled to the relief
sought, if it transpired, for example, on the basis of the information available, that the turnover of the
companies controlled were essentially attributable to those resident in the Community.
Accordingly, in so far as, for the purpose of evaluating the business of a holding company, factors
other than the quantitative criterion may be taken into account when interpreting the domestic legislation, I
consider it useful to provide the House of Lords with an answer to the first question.
591

Substance
15.
That said, I would first of all observe that, as the Court of Justice itself has stated on several
occasions:
although, as Community law stands at present, direct taxation does not as such fall within
the purview of the Community, the powers retained by the Member States must nevertheless be
exercised consistently with Community law.7

7
See also EC Commission v UK Case C-246/89 [1991] ECR I-4585 (para 12), Finanzamt Kln-Altstadt v Schumacker
Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225 (para 21), Asscher v Staatssecretaris van Financin
Case C-107/94 [1996] All ER (EC) 757, [1996] ECR I-3089 (para 36) and Futura Participations SA v Administration
des Contributions Case C-250/95 [1997] ECR I-2471 (para 19). See also, however, Council Directive (EEC) 90/435 on
the common system of taxation applicable in the case of parent companies and subsidiaries in different member states
(OJ 1990 L225 p 6).

In the field of direct taxation, therefore, member states may not adopt measures which would have the
effect of unjustifiably impeding freedom of movement for natural or legal persons carrying on an activity in
a self-employed capacity (see the opinion of Advocate General Lger in Asscher v Staatssecretaris van
Financin Case C-107/94 [1996] All ER (EC) 757, [1996] ECR I-3089 (para 55)). It scarcely needs to be
mentioned that taxation which is discriminatory or which somehow impedes or limits the exercise of the
right of establishment is undoubtedly caught by art 52 (see the judgment in R v IRC, ex p Commerzbank
AG Case C-330/91 [1993] ECR I-4017 (para 20)).
It is therefore necessary to determine, in relation to the present case, whether art 52 of the Treaty
precludes the legislation at issue from making consortium relief conditional on the holding companys
business consisting, wholly or mainly, in the holding of shares of subsidiaries resident in the United
Kingdom.
16.
The requirement that most of the subsidiaries must be resident in the United Kingdom appears prima
facie to be a restriction on freedom of establishment, prohibited by the first paragraph of art 52. Relief is
thereby precluded in all cases where the holding companys business consists, wholly or mainly, in the
holding of shares of companies resident outside the United Kingdom, and thus even where such
companies are established in other member states. It is the latter aspect which is of significance for
Community law, since in those circumstances the legislation at issue limits, or at least discourages, the
exercise by British companies of the right to create corporate structures in other member states.
17.
 To my mind there can be no doubt that such legislation is restrictive. On that point, suffice it to recall
the judgment in R v HM Treasury, ex p Daily Mail Case 81/87 [1989] 1 All ER 328, [1988] ECR 5483 (para
15) in which the court reaffirmed that freedom of establishment constitutes one of the fundamental
principles of the Community and that the provisions of the Treaty guaranteeing that freedom have been
directly applicable since the end of the transitional period, before going on to explain that
even though those provisions are directed mainly to ensuring that foreign nationals and
companies are treated in the host member state in the same way as nationals of that state, they
also prohibit the member state of origin from hindering the establishment in another member state
of one of its nationals or of a company incorporated under its legislation which comes within the
592 definition contained in art 58. (See [1989] 1 All ER 328, [1988] ECR 5483 (para 16).)
18.
This is typical of restrictions on exits. Tax disincentives undoubtedly make the creation of cross-
border corporate structures a less attractive prospect for companies established in the United Kingdom. In
so far as such a restriction applies to subsidiaries resident in other member states, the UK legislation
entailsI repeatan obstacle to the freedom of establishment guaranteed by art 52 of the Treaty.
Furthermore, the legislation at issue appears particularly unfavourable to companies which belong to a
consortium as opposed to a group, since in the latter case the setting-off of losses against profits would
still be possible (a point made by the Commission and not disputed).
Nor is it a valid objection to argueas does the UK governmentthat a distinction based on the
residence of a companys subsidiaries does not amount to discrimination since the situations involved are
not comparable. The legislation at issue concerns companies which are liable to tax in the United
Kingdom and makes tax relief conditional on the manner in which the right of establishment is exercised
in other member states of the Community as well.
19.
In those circumstances, it only remains to determine whether the restriction in question may be
justified in the light of Community law.
In that connection, both ICI and the Commission have ruled out that possibility. According to the United
Kingdom, on the other hand, it is a measure justified in terms of its objective, which is to prevent the
creation of foreign subsidiaries from being used as an easy means of depriving the United Kingdom
Treasury of tax revenue.
20.
The first difficulty which arises in this connection is whether or not to class the restriction at issue as
giving rise to discrimination based on the place of establishment. The implications in respect of a possible
justification will vary according to the solution adopted. The court has consistently held that a
discriminatory measure is compatible with Community law only if it falls within the scope of one of the
derogations expressly provided by the Treaty8. Where, however, the measure in question applies without
distinction to all persons including foreigners, the measures restricting freedom of establishment are
compatible if they are in furtherance of imperative requirements in the general interest, if they are suitable
for securing the attainment of the objective pursued and if they do not go beyond what is necessary to
attain it9.
8
See the judgment in Bond van Adverteerders v Netherlands Case 352/85 [1988] ECR 2085 (para 32), in which the court
stated that [discriminatory] national rules are compatible with Community law only if they can be brought within the
scope of an express derogation.
9
See, most recently, the judgment in Gebhard v Consiglio dellOrdine degli Avvocati e Procuratori di Milano Case C-55/94
[1996] All ER (EC) 189, [1995] ECR I-4165 (para 37), in which the court referred without distinction to all the
fundamental freedoms guaranteed by Community law.

21.
Once again, the answer to the question referred depends on where emphasis is placed. It is apparent,
for example, that the legislation at issue discriminates between companies resident in the United
Kingdom, according to whether or not they have exercised their freedom of establishment in other
member states, through a holding company, for instance. In other words, the distinction affects companies
whose registered office is in the same member state and is linked to their decision whether or not to avail
themselves of the possibility, guaranteed by art 52 of the Treaty, of setting up branches or subsidiaries in
other countries, even if they are member states of the Community.
593
22.
Admittedly, even if art 52 ensured that all subsidiaries resident in the Community were placed on an
equal footing with those resident in the United Kingdom, a further level of discrimination evidently cannot
be ruled out. That is to say, there would still be discrimination between the companies which exercise the
right of establishment, depending on the precise form this takes: tax relief would be granted where the
holding companys business consisted, wholly or mainly, in holding shares of companies established in
the territory of member states, but denied where only a minority of the companies were resident in the
territory concerned.
That detail is especially significant in the present case, where some of the companies controlled by
Holdings are resident in member states other than the United Kingdom. However, that form of
discrimination clearly cannot be challenged on the basis of art 52 of the Treaty, since there is no
restriction on freedom of establishment in the Community. Although discrimination based on the place of
establishment might have been eliminated in compliance with art 52 in respect of the United Kingdom or
other member states of the Community, the UK legislation discourages, if anything, the creation of
subsidiaries in countries outside the Community. That is why, as we shall have occasion to verify when
examining the second question, ICI seeks to rely on art 5 of the Treaty with a view to securing in any
event the tax relief provided by the Act.
23.
The domestic legislation, in so far as it gives rise to discrimination, may clearly be justified only in the
exceptional circumstances envisaged by the Treaty. This is the approach taken by the Commission, which
has made a short study of the problem of justificatory grounds, from which it concludes that none of the
derogations provided for in art 56 (public policy, public security or public health) applies in the present
case. Considerations of a purely economic nature, such as loss of tax revenue, cannot justify restrictions
of a discriminatory character which fall within the scope of art 52 of the Treaty 10.
10
See the judgments in Bond van Adverteerders [1988] ECR 2085 and Stichting Collectieve Antennevoorziening Gouda v
Commissariaat voor de Media Case C-288/89 [1991] ECR I-4007 (para 11). In the judgment in Svensson v Ministre du
Logement et de lUrbanisme Case C-484/93 [1995] ECR I-3955 (para 15), given that the Luxembourgish legislation on
interest rate subsidies in respect of loans for the construction of housing entailed discrimination based on the place of
establishment, the court added that such discrimination can only be justified on the general interest grounds referred to
in Article 56(1) of the Treaty which do not include economic aims. It should be noted, however, that on the same
occasion the court also considered whether the legislation at issue, albeit classed as discriminatory, was necessary in
order to safeguard the cohesion of the tax system. In so doing, however, the court also determined whether the
measure in question could be justified in terms of requirements which may be taken into account only in the case of
measures which apply without distinction. In my joint opinion in Decker v Caisse de Maladie des Employs Privs Case
C-120/95 and Kohl v Union des Caisses de Maladie Case C-158/95, I have already explained the difficulties in regard to
consistency, raised by the courts recent case law (see esp (1997) ECJ Transcript (opinion), 16 September (paras 49
50)).

24.
However, even if the measure at issue were to be regarded as applying without distinction, in view of
the fact that the requirement is imposed on companies which are in any event liable to taxation in the
United Kingdom, it would still be incompatible with the rules regarding freedom of establishment. I have
no hesitation in stating that the arguments put forward in this case to justify the legislation at issue are
devoid of substance.
25.
Admittedly, on a number of occasions the court has acknowledged that the need for cohesion in the
application of tax systems can constitute sufficient justification, linked to mandatory requirements in the
general interest, for 594 imposing a restriction on freedom of establishment11. It is also true, however, that
the problem in question has in general arisen in respect of domestic legislation which distinguished
between legal or natural persons on grounds of their being resident or having their registered office in the
territory of another member state.

11
See the judgments in Bachmann v Belgium Case C-204/90 [1992] ECR I-249 (para 21), Schumacker [1995] All ER
(EC) 319, [1995] ECR I-225 (para 47), Wielockx v Inspecteur der Directe Belastingen Case C-80/94 [1995] All ER
(EC) 769, [1995] ECR I-2493 (para 25) and Asscher [1996] All ER (EC) 757, [1996] ECR I-3089 (para 59). See
also my opinion in Safir v Skattemyndigheten i Darlarnas Ln (formerly Skattemyndigheten i Kopparbergs Ln) Case C-
118/96 (1997) Transcript (opinion), 23 September, ECJ (paras 20ff).

In Bachmann v Belgium Case C-204/90 [1992] ECR I-249, which concerned the application to
residents of domestic legislation making the deduction of certain contributions from taxable income
conditional on those contributions having been paid in that member state, the court stated that the aim of
the Belgian legislation was to enable the loss of tax revenue resulting from the deduction of life assurance
contributions to be offset by the taxation of pensions, annuities or capital sums payable by the insurers.
The cohesion of the tax system would thus have been undermined if the Belgian state had been
compelled to offer the same tax advantages to persons insured with companies established abroad, in
view of the difficulty of collecting tax on earnings paid abroad (see [1992] ECR I-249 (paras 2223)).
Given that the domestic legislation was expressly stated to be non-discriminatory, the court therefore
concluded that it could not be regarded as incompatible with art 59 since it was justified by requirements
in the general interest.
26.
Returning to the instant case, it therefore remains to be determined whether the objective of
preventing the creation of subsidiaries outside the United Kingdom, and thus in other member states as
well, depriving the United Kingdom Treasury of tax revenue is capable of justifying the restriction on
freedom of establishment resulting from the legislation on consortium relief.
27.
According to the United Kingdom, that question should be answered in the affirmative. Obviously,
there is no UK tax charge on a non-resident subsidiary. Accordingly, relief on losses incurred by a
subsidiary resident in the United Kingdom would not be compensated by taxation of the profits made by
other subsidiaries, resident in other states. In the United Kingdoms view, that is incompatible with the
rationale underlying consortium relief, which is to extend the same tax treatment to a company when it is
a member of a consortium as it would receive if it participated directly in the business undertaken by the
joint venture.
28.
I have serious reservations regarding that argument. The objective is not so much that of preserving
the cohesion of the tax system as, quite simply, of preventing a fall in tax revenue. If that is indeed the
position, I do not believe that it can justify a derogation from a fundamental principle guaranteed by the
Treaty.
 That is not all, however. Even if the objective pursued were deemed to be valid under Community law,
it would still have to pass the proportionality test. Here, too, I have misgivings. It is highly doubtful whether
the restrictive measure in question is suited to attaining the objective pursued. Indeed, in circumstances
where tax relief is denied solely on account of Holdings exercise of freedom of establishment in other
member states, I do not believe it can seriously be maintained that the legislation at issue is an effective
means of ensuring the cohesion of the tax system.
29.
I find it difficult to reconcile the need to prevent tax evasion in order to preserve the cohesion of the tax
system with the fact that consortium relief is 595 granted whenever only a minority of companies is
resident outside the United Kingdom, and denied whenever such companies are in the majority. To my
mind the risk of evasion, if indeed it exists, is also present in the former set of circumstances, albeit
according to the proportion of non-resident companiesto a lesser degree.
30.
Furthermore, it remains to be demonstrated that no other measures, equally effective but less
restrictive of freedom of establishment, are available. On that point, I would suggest that neither the
Inland Revenue nor the UK government in its observations has established that the measures at issue
are the only ones available and that the objective could not be effectively pursued by other means.
31.
It seems to me that all the foregoing observations adequately support the conclusion that domestic
legislation which makes consortium relief available to companies only if the business of the holding
company controlled by the company seeking relief consists, wholly or mainly, in holding shares of
subsidiaries resident in the member state concerned constitutes a restriction on freedom of
establishment, which is prohibited by the Treaty and cannot otherwise be justified.

QUESTION 2
32.
Once again I would refer to the particular features of the present case and its implications for
Community law. Article 52 of the Treaty is relevant in so far as a requirement imposed by domestic
legislation in respect of tax relief also affects companies availing themselves of the right of establishment
in other member states of the Community. What this means in practice is that, in the present case, the
domestic legislation is contrary to art 52 in so far as it restricts freedom of establishment in other member
states of the Community.
As regards the further difficulty, namely discrimination against companies which choose to set up
subsidiaries mostly in non-member countries, art 52 is of no avail, since the matter falls outside the scope
of Community law.
If that is indeed the position, as I believe it undoubtedly is, not even the interpretation of art 5 of the
Treaty sought by the House of Lords can be of any assistance. In the first place, in so far as one aspect of
the present case is covered by art 52, which has direct effect, the national courts duty to interpret
domestic legislation consistently with Community law is irrelevant. The result sought by harmonisation of
national and Community law is already achieved by virtue of the fact that individuals may rely on
Community law in proceedings before the national courts.
Secondly, nor can the duty of consistent interpretation laid down by art 5 of the Treaty be relied on in
relation to the aspect of the present case which is not covered by art 52. The discrimination against
companies which choose to hold shares in subsidiaries, the majority of which are resident in non-member
countries, by comparison with those whose subsidiaries are all resident in the United Kingdom (or in the
Community) or which have only a minority of subsidiaries resident outside the United Kingdom (or the
Community), is not relevant for the purposes of Community law. It follows that neither art 52 nor art 5
applies. Accordingly, the national court is under no obligation pursuant to art 5 of the Treaty to adopt an
interpretation consistent with Community law in respect of a situation, or, as in the present case, aspects
of a situation to which Community law does not apply.
596
33.
In the light of the foregoing, I therefore propose that the Court of Justice should reply as follows to the
questions referred by the House of Lords:
(1) Article 52 of the Treaty is to be interpreted as precluding the application of legislation of a
member state which prevents a company established in the territory of that state from obtaining tax
relief in respect of losses incurred by another company, established in the same state and
controlled by the first company through a holding company, in cases where the holding companys
business consists, wholly or mainly, in holding shares of subsidiaries resident outside that state, in
so far as such legislation constitutes a restriction on the exercise of the right of establishment in
other member states of the European Union.
(2) Article 5 of the Treaty does not require the national courts to interpret domestic legislation
consistently with Community law in respect of a situation, or aspects of a situation, falling outside
the scope of Community law.

16 July 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 24 July 1996, received at the Court of Justice of the European Communities on 29 July
1996, the House of Lords referred to the Court of Justice for a preliminary ruling under art 177 of the EC
Treaty two questions on the interpretation of arts 5 and 52 of the Treaty.
2.
Those questions were raised in proceedings between Imperial Chemical Industries plc (ICI) and the
United Kingdom tax authorities (the Inland Revenue) concerning the latters refusal to grant to ICI tax
relief in respect of trading losses incurred by a subsidiary of the holding company beneficially owned by
ICI through a consortium.
3.
ICI and Wellcome Foundation Ltd, both of which are companies resident in the United Kingdom,
together form a consortium through which they beneficially own 49% and 51%, respectively, of Coopers
Animal Health (Holdings) Ltd (Holdings).
4.
The sole business of Holdings is to hold shares in some 23 trading companies which are its
subsidiaries and which operate in many countries. Of those 23 subsidiaries, fourincluding Coopers
Animal Health Ltd (CAH)are resident in the United Kingdom, six in other member states and thirteen in
non-member countries.
5.
CAH incurred losses on its UK trade in the accounting periods ending in 1985, 1986 and 1987. ICI
sought, pursuant to ss 258 to 264 of the Income and Corporation Taxes Act 1970, to set 49% of CAHs
losses for those periods (the proportion corresponding to its shareholding in Holdings) against its
chargeable profits for the corresponding periods by way of tax relief.
6.
As regards the conditions for and the detailed rules governing tax relief as claimed by ICI, the 1970 Act
provides as follows:
258(1) Relief for trading losses and other amounts eligible for relief from corporation tax may
in accordance with the following provisions of this Chapter be surrendered by a company (called
the surrendering company) which is a member of a group of companies and, on the making of a
claim by another company (called the claimant company) which is a member of the same group,
may be allowed to the claimant company by way of relief from corporation tax called group relief.
(2) Group relief shall also be available 597 in accordance with the said provisions in the case of a
surrendering company and a claimant company where either of them is a member of a consortium
and the other is(a) a trading company which is owned by the consortium and which is not a 75
per cent subsidiary of any company; or (b) a trading company(i) which is a 90 per cent subsidiary
of a holding company which is owned by the consortium; and (ii) which is not a 75 per cent
subsidiary of a company other than the holding company; or (c) a holding company which is owned
by the consortium and which is not a 75 per cent subsidiary of any company (5) For the purpose
of this s and the following ss of this Chapter(a) two companies shall be deemed to be members
of a group of companies if one is the 75 per cent subsidiary of the other or both are 75 per cent
subsidiaries of a third company, (b) holding company means a company the business of which
consists wholly or mainly in the holding of shares or securities of companies which are its 90 per
cent subsidiaries, and which are trading companies, (c) trading company means a company
whose business consists wholly or mainly of the carrying on of a trade or trades (7) References
in this and the following sections of this Chapter to a company apply only to bodies corporate
resident in the United Kingdom; and in determining for the purposes of this and the following ss of
this Chapter whether one company is a 75 per cent subsidiary of another, the other company shall
be treated as not being the owner(a) of any share capital which it owns directly in a body
corporate if a profit on a sale of the shares would be treated as a trading receipt of its trade, or (b)
of any share capital which it owns indirectly, and which is owned directly by a body corporate for
which a profit on the sale of the shares would be a trading receipt, or (c) of any share capital which
it owns directly or indirectly in a body corporate not resident in the United Kingdom. (8) For the
purposes of this and the following sections of this Chapter, a company is owned by a consortium if
three-quarters or more of the ordinary share capital of the company is beneficially owned between
them by companies of which none beneficially owns less than one-twentieth of that capital, and
those companies are called the members of the consortium.
259(1) If in any accounting period the surrendering company has incurred a loss, computed
as for the purposes of subsection (2) of section 177 of this Act, in carrying on a trade, the amount of
the loss may be set off for the purposes of corporation tax against the total profits of the claimant
company for its corresponding accounting period.
7.
The Inland Revenue refused ICIs application for tax relief on the ground that Holdings does not
constitute a holding company within the meaning of s 258(5)(b) read together with s 258(7). Even though
Holdings sole business is to hold shares or securities of companies which are its 90% subsidiaries, and
which are trading companies, the majority of its subsidiaries (19 out of 23) are not bodies corporate
resident in the United Kingdom as required by the opening words of s 258(7) and therefore Holdings
main business cannot be recognised as that of a holding company within the meaning of sub-s 5(b).
8.
 Contesting that interpretation of the domestic legislation, ICI brought an action against the decision
rejecting its claim. The High Court found in ICIs favour and its decision was subsequently affirmed by the
Court of Appeal.
598
9.
On appeal, the House of Lords concluded that the Inland Revenues refusal was justified in terms of
the Act, but felt it necessary to consider the arguments, based on Community law, advanced by ICI to
contest the refusal.
10.
In ICIs submission, the requirement that a holding companys business consist wholly or mainly in the
holding of shares or securities of companies resident in the United Kingdom amounts to a restriction, in
the form of a discriminatory tax regime, on freedom of establishment for companies and firms, and
therefore infringes arts 52 and 58 of the Treaty.
11.
It claims that the discrimination arises from the fact that tax relief for losses incurred by a resident
company which is a subsidiary of a resident holding company is granted to a member of a consortium
where all, or most of, the subsidiaries controlled by the holding company are resident, whereas, other
things being equal, it will be refused where the holding companybecause it has exercised its right to
freedom of establishment conferred by the EC Treatycontrols mainly subsidiaries resident in other
member states.
12.
ICI maintains that, faced with such discrimination, it is the national courts duty, even in a case such as
that before the House of Lords, where the holding company controls 23 subsidiaries, of which only ten are
resident in the United Kingdom or another member state, to set aside the residence requirement laid
down by the Act as being contrary to Community law.
13.
The House of Lords considered an interpretation of Community law to be necessary as regards both
the compatibility of the residence requirement laid down by the Act for the grant of tax relief as claimed by
ICI with the rules of the Treaty and, should the Act prove to be contrary to Community law, the approach
to be taken by national courts in such a situation. It therefore decided to stay proceedings and refer the
following questions to the Court of Justice for a preliminary ruling:
(1) In a situation where: (i) a company (Company A) is resident in a Member State of the
European Union (ii) Company A is part of a consortium with another company (Company B) also
resident in that Member State (iii) Company A and B jointly own a holding company (Company C)
also resident in the Member State (iv) Company C has a number of trading subsidiaries, which are
resident either in that Member State, other Member States of the European Union or elsewhere in
the world, and (v) Company A is precluded from being entitled to claim against its corporation tax
liability relief in respect of trading losses incurred by a trading subsidiary (also resident in that
Member State) of Company C because the national legislation, construed as a matter of national
law, required that the business of Company C should consist wholly or mainly in the holding of
shares in subsidiaries which are resident in that Member State: Does the requirement identified
at (v) constitute a restriction on the freedom of establishment under Article 52 of the EC Treaty? If
so, is such treatment nevertheless justified under Community law?
(2) If the requirement under (v) is an unjustified restriction under Community law, does Article 5
of the EC Treaty require a national court to interpret the relevant national legislation, so far as is
possible, so as to comply with Community law, even though neither Company A, Company B nor
Company C is itself seeking to exercise any rights under Community law, and even if an
interpretation of national legislation which would comply with Community law would have the effect
of giving relief where the business of Company C consisted mainly in the holding of shares in
subsidiaries 599 established outside the EC/EEA? Or does Article 5 have the consequence only
that the national legislation, despite its interpretation, takes effect subject to the requirements of
Community law in a case where these requirements are in point?

ADMISSIBILITY
14.
The UK government has expressed doubts as to the relevance of the first question in determining the
issue in the main proceedings. It argues that, even if the Act were found to entail a restriction on freedom
of establishment, incompatible with art 52 of the Treaty, this would have no bearing on the determination
of the proceedings. ICI would in any event be denied the tax relief provided for under the Act, since the
majority of the companies controlled by Holdings (13 out of 23) are resident, not in other member states,
but in non-member countries.
15.
According to established case law, it is solely for the national courts before which proceedings are
pending, and which must assume responsibility for the judgment to be given, to determine in the light of
the particular circumstances of each case both the need for a preliminary ruling to enable them to give
judgment and the relevance of the questions which they submit to the Court of Justice (see, inter alia, the
judgments in Enderby v Frenchay Health Authority Case C-127/92 [1994] 1 All ER 495, [1994] ICR 112,
[1993] ECR I-5535 (para 10), Eurico Italia Srl v Ente Nazionale Risi Joined cases C-332333/92 and C-
335/92 [1994] ECR I-711 (para 17) and MacLachlan v Caisse Nationale dAssurance Vieillesse des
Travailleurs Salaris de la Rgion dIle-de-France (CNAVTS) Case C-146/93 [1994] ECR I-3229 (para
20)). A request for a preliminary ruling from a national court may be rejected only if it is manifest that the
interpretation of Community law or the examination of the validity of a rule of Community law sought by
that court bears no relation to the true facts or the subject matter of the main proceedings (see the
judgments in BP Supergas Anonimos Etaira Geniki Emporiki-Viomichaniki kai Antiprossopeion v Greece
Case C-62/93 [1995] All ER (EC) 684, [1995] ECR I-1883 (para 10) and Furlanis Costruzioni General SpA
v Azienda Nazionale Autonoma Strade (ANAS) C-143/94 [1995] ECR I-3633 (para 12)).
16.
However, that is not the situation in the present case. The House of Lords observes that opinion differs
as to the proper construction of s 258(5), in terms of which, in order to qualify as a holding company
within the meaning of the Act, it is necessary to hold shares wholly or mainly in companies which are
resident in the United Kingdom, and, more specifically, as to the notion of control of a majority of
subsidiaries resident in the United Kingdom, one interpretation of which makes it necessary to determine
whether the Act is compatible with art 52 of the Treaty.
17.
In those circumstances, it is necessary to consider the questions referred by the House of Lords.

SUBSTANCE

The first question

18.
By its first question, the House of Lords asks essentially whether art 52 of the Treaty precludes
legislation of a member state which, in the case of companies established in that state belonging to a
consortium through which they control a holding company, makes a particular form of tax relief subject to
the requirement 600 that the holding companys business consist wholly or mainly in the holding of shares
in subsidiaries that are established in the member state concerned.
19.
Although direct taxation is a matter for the member states, they must nevertheless exercise their direct
taxation powers consistently with Community law (see the judgments in Finanzamt Kln-Altstadt v
Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225 (para 21), Wielockx v Inspecteur
der Directe Belastingen Case C-80/94 [1995] All ER (EC) 769, [1995] ECR I-2493 (para 16), Asscher v
Staatssecretaris van Financin Case C-107/94 [1996] All ER (EC) 757, [1996] ECR I-3089 (para 36) and
Futura Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471 (para 19)).
20.
According to established case law, the freedom of establishment which art 52 grants to nationals of the
member states and which entails the right for them to take up and pursue activities as self-employed
persons under the conditions laid down for its own nationals by the law of the member state where such
establishment is effected, includes, pursuant to art 58 of the Treaty, the right of companies or firms formed
in accordance with the law of a member state and having their registered office, central administration or
principal place of business within the Community, to pursue their activities in the member state concerned
through a branch or agency. With regard to companies, it should be noted in this context that it is their
corporate seat in the above sense that serves as the connecting factor with the legal system of a
particular state, like nationality in the case of natural persons (see the judgments in EC Commission v
France Case 270/83 [1986] ECR 273 (para 18) and R v IRC, ex p Commerzbank AG Case C-330/91
[1993] ECR I-4017 (para 13)).
21.
It should also be pointed out that, even though, according to their wording, the provisions concerning
freedom of establishment are directed mainly to ensuring that foreign nationals and companies are
treated in the host member state in the same way as nationals of that state, they also prohibit the member
state of origin from hindering the establishment in another member state of one of its nationals or of a
company incorporated under its legislation which comes within the definition contained in art 58 (see the
judgment in R v HM Treasury, ex p Daily Mail Case 81/87 [1989] 1 All ER 328, [1988] ECR 5483 (para
16)).
22.
It should be noted here that, under the legislation at issue in the main proceedings, companies
belonging to a resident consortium which have, through a holding company, exercised their right to
freedom of establishment in order to set up subsidiaries in other member states are denied tax relief on
losses incurred by a resident subsidiary where the majority of the subsidiaries controlled by the holding
company have their seat outside the United Kingdom.
23.
Such legislation, therefore, applies the test of the subsidiaries seat to establish differential tax
treatment of consortium companies established in the United Kingdom. Consortium relief is available only
to companies controlling, wholly or mainly, subsidiaries whose seats are in the national territory.
24.
It is therefore necessary to determine whether there is any justification for such inequality of treatment
under the Treatys provisions on freedom of establishment.
25.
The UK government maintains that, for the purposes of direct taxation, the respective situations of
resident and non-resident companies are not, as a general rule, comparable. It puts forward two types of
justification. First, the legislation at issue is designed to reduce the risk of tax avoidance arising, in the
present case, from the possibility for members of a consortium to channel the charges of non-resident
subsidiaries to a subsidiary resident in the United Kingdom and to 601 have profits accrue to non-resident
subsidiaries. The purpose of the legislation at issue is therefore to prevent the creation of foreign
subsidiaries from being used as a means of depriving the United Kingdom Treasury of taxable revenues.
A further objective is to prevent a reduction in revenue caused by the mere existence of non-resident
subsidiaries, since the Inland Revenue cannot tax profits made by subsidiaries located outside the United
Kingdom in order to offset the revenue lost through the granting of relief on losses incurred by resident
subsidiaries.
26.
As regards the justification based on the risk of tax avoidance, suffice it to note that the legislation at
issue in the main proceedings does not have the specific purpose of preventing wholly artificial
arrangements, set up to circumvent UK tax legislation, from attracting tax benefits, but applies generally
to all situations in which the majority of a groups subsidiaries are established, for whatever reason,
outside the United Kingdom. However, the establishment of a company outside the United Kingdom does
not, of itself, necessarily entail tax avoidance, since that company will in any event be subject to the tax
legislation of the state of establishment.
27.
Furthermore, the risk of charges being transferred, which the legislation at issue is designed to
prevent, is entirely independent of whether or not the majority of subsidiaries are resident in the United
Kingdom. The existence of only one non-resident subsidiary is enough to create the risk invoked by the
UK government.
28.
In answer to the argument that revenue lost through the granting of tax relief on losses incurred by
resident subsidiaries cannot be offset by taxing the profits of non-resident subsidiaries, it must be pointed
out that diminution of tax revenue occurring in this way is not one of the grounds listed in art 56 of the
Treaty and cannot be regarded as a matter of overriding general interest which may be relied upon in
order to justify unequal treatment that is, in principle, incompatible with art 52 of the Treaty.
29.
It is true that in the past the court has accepted that the need to maintain the cohesion of tax systems
could, in certain circumstances, provide sufficient justification for maintaining rules restricting fundamental
freedoms (see, to this effect, Bachmann v Belgium Case C-204/90 [1992] ECR I-249 and EC
Commission v Belgium Case C-300/90 [1992] ECR I-305). Nevertheless, in the cases cited, there was a
direct link between the deductibility of contributions from taxable income and the taxation of sums payable
by insurers under old-age and life assurance policies, and that link had to be maintained in order to
preserve the cohesion of the tax system in question. In the present case, there is no such direct link
between the consortium relief granted for losses incurred by a resident subsidiary and the taxation of
profits made by non-resident subsidiaries.
30.
Consequently, the answer to be given to the first question must be that art 52 of the Treaty precludes
legislation of a member state which, in the case of companies established in that state belonging to a
consortium through which they control a holding company, by means of which they exercise their right to
freedom of establishment in order to set up subsidiaries in other member states, makes a particular form
of tax relief subject to the requirement that the holding companys business consist wholly or mainly in the
holding of shares in subsidiaries that are established in the member state concerned.

The second question

31.
By its second question the House of Lords essentially asks the Court of Justice to explain the scope of
the duty to cooperate in good faith, laid down by 602 art 5 of the Treaty. More specifically, if it were to
follow from the reply to the first question that the legislation at issue in the main proceedings is
incompatible with Community law in not granting tax relief where the holding company owned by the
consortium controls mainly subsidiaries having their seat in the Community, in a case where this condition
is not fulfilled by subsidiaries resident in the United Kingdom, the House of Lords asks whether it must
likewise disapply that legislation, or construe it in a way conforming with Community law, where the
holding company controls mainly subsidiaries having their seat in non-member countries.
32.
It must be emphasised that the difference of treatment applied according to whether or not the
business of the holding company belonging to the consortium consists wholly or mainly in holding shares
in subsidiaries having their seat in non-member countries lies outside the scope of Community law.
33.
Consequently, arts 52 and 58 of the Treaty do not preclude domestic legislation under which tax relief
is not granted to a resident consortium member where the business of the holding company owned by
that consortium consists wholly or mainly in holding shares in subsidiaries which have their seat in non-
member countries. Nor does art 5 of the Treaty apply.
34.
Accordingly, when deciding an issue concerning a situation which lies outside the scope of Community
law, the national court is not required, under Community law, either to interpret its legislation in a way
conforming with Community law or to disapply that legislation. Where a particular provision must be
disapplied in a situation covered by Community law, but that same provision could remain applicable to a
situation not so covered, it is for the competent body of the state concerned to remove that legal
uncertainty in so far as it might affect rights deriving from Community rules.
35.
Consequently, in circumstances such as those in point in the main proceedings, art 5 of the Treaty
does not require the national court to interpret its legislation in conformity with Community law or to
disapply the legislation in a situation falling outside the scope of Community law.

COSTS
36.
The costs incurred by the UK government and by the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the proceedings pending before the national court, the decision on costs
is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the House of Lords
by order of 24 July 1996, hereby rules: (1) Article 52 of the EC Treaty precludes legislation of a member
state which, in the case of companies established in that state belonging to a consortium through which
they control a holding company, by means of which they exercise their right to freedom of establishment
in order to set up subsidiaries in other member states, makes a particular form of tax relief subject to the
requirement that the holding companys business consist wholly or mainly in the holding of shares in
subsidiaries that are established in the member state concerned. (2) In circumstances such as those in
point in the main proceedings, art 5 of the Treaty does not require the national court to interpret its
legislation in conformity with Community law or to disapply the legislation in a situation falling outside the
scope of Community law.

603

[1998] All ER (EC) 604

Criminal proceedings against Lemmens

(Case C-226/97)

EUROPEAN COMMUNITY; Free movement of goods: CRIMINAL; Criminal Procedure

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES GULMANN (RAPPORTEUR AND PRESIDENT OF THE THIRD AND FIFTH CHAMBERS,
ACTING FOR THE PRESIDENT), RAGNEMALM, WATHELET, SCHINTGEN (PRESIDENTS OF
CHAMBERS), MANCINI, KAPTEYN, MURRAY, EDWARD, PUISSOCHET, JANN AND SEVN
ADVOCATE GENERAL FENNELLY
16 DECEMBER 1997, 12 FEBRUARY, 16 JUNE 1998
European Community Freedom of movement Goods Technical standards Netherlands failing to
notify technical regulations applying to breathalyser equipment to Commission Whether Community
provisions applicable to technical regulations ancillary to criminal law Whether failure to notify rendering
evidence from breathalysers unusable in criminal proceedings Council Directive (EEC) 83/189, arts 8,
9.

L was charged in the Netherlands with driving a vehicle while under the influence of alcohol. By virtue of
domestic provisions on the conduct of breathalyser tests, the breath-analysis apparatus used had to
comply with certain standards of quality, performance and test methods. Before the
Arrondissementsrechtbank, Maastricht, L drew attention to the fact that the technical regulation in which
those requirements were set out had not been properly notified to the European Commission in
accordance with Council Directive (EEC) 83/189 laying down a procedure for the provision of information
in the field of technical standards and regulations. Articles 8 and 9 of the directive required member states
to communicate to the Commission any draft technical regulation falling within the scope of the directive
and to postpone the adoption of such draft regulations for three months. Since the Court of Justice of the
European Communities had already ruled that a breach of the obligation to notify rendered technical
regulations inapplicable and therefore unenforceable against individuals, the national court stayed the
proceedings and referred to the Court of Justice for a preliminary ruling the principal question whether an
infringement of the obligation under art 8 to notify a technical regulation on breath-analysis apparatus
made it impossible for evidence obtained by means of such apparatus, authorised in accordance with that
regulation, to be relied on against an individual charged with driving while under the influence of alcohol.
Before the Court of Justice the Netherlands and French governments contended respectively (i) that since
the technical regulation at issue applied in the field of criminal law, it fell outside the sphere of Community
law and (ii) that the directive did not apply to products which were intended to be used in connection with
the exercise of a public authority and a fortiori in criminal proceedings instituted by member states.

Held Although in principle criminal legislation and the rules of criminal procedure were matters for which
the member states were responsible, it did not follow that that branch of the law could not be affected by
Community law. There was nothing in Directive 83/189 to exclude technical regulations from the
notification requirements because they fell within the scope of criminal law, or to limit the scope of the
directive to regulations which were intended to be used otherwise than in connection with the exercise of
public authority. The regulation 604 at issue therefore constituted a technical regulation which should,
prior to its adoption, have been notified to the Commission in accordance with art 8 of the directive.
Moreover, although failure to notify technical regulations rendered them inapplicable inasmuch as they
hindered the use or marketing of a product which did not conform with its provisions, it did not have the
effect of rendering unlawful any use of a product which conformed with unnotified regulations. It followed
that breach of the obligation imposed by art 8 to notify a technical regulation on breath-analysis apparatus
did not make it impossible for evidence obtained by means of such apparatus, authorised in accordance
with that regulation, to be relied on against an individual charged with driving while under the influence of
alcohol (see p 617 e f, p 618 b c j and p 619 b d e, post)
Criminal proceedings against Casati Case 203/80 [1981] ECR 2595, Cowan v Trsor public Case
186/87 [1989] ECR 195 and Bic Benelux SA v Belgium Case C-13/96 [1997] ECR I-1753 applied; CIA
Security International SA v Signalson SA Case C-194/94 [1996] All ER (EC) 557, [1996] ECR I-2201
distinguished.

Cases cited
Bic Benelux SA v Belgium Case C-13/96 [1997] ECR I-1753.
Casati (Criminal proceedings against) Case 203/80 [1981] ECR 2595.
Clean Car Autoservice GesmbH v Landeshauptman von Wien Case C-350/96 [1998] All ER (EC) 434,
ECJ.
CIA Security International SA v Signalson SA Case C-194/94 [1996] All ER (EC) 557, [1996] ECR I-2201,
ECJ.
Cowan v Trsor public Case 186/87 [1989] ECR 195.
European Commission v Netherlands Case C-273/94 [1996] ECR I-31.
Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325, ECJ.
Gallotti (Criminal proceedings against) Joined cases 58/95, C-75/95, C-112/95, C-119/95, C-123/95, C-
135/95, C-140/95, C-141/95, C-154/95 and C-157/95 [1996] ECR I-4345.
Holdijk Joined cases 141143/81 [1982] ECR 1299.
Leur-Bloem v Inspecteur der Belastingdienst/Ondernemingen Amsterdam 2 Case C-28/95 [1997] All ER
(EC) 738, [1997] ECR I-4161, ECJ.
Ministre Public v Stoeckel Case C-345/89 [1991] ECR I-4047.
Sunino (Criminal proceedings against) Case C-2/96 [1996] ECR I-1543.
Telemarsicabruzzo SpA v Circostel Joined cases C-320322/90 [1993] ECR I-393.
Van de Haar, Kaveka de Meern BV (Criminal proceedings against) Joined cases 177178/82 [1984] ECR
1797.
Vaneetveld v Le Foyer SA Case C-316/93 [1994] ECR I-763.
Verholen v Sociale Verzekeringsbank Amsterdam Joined cases C-8789/90 [1991] ECR I-3757.

Reference
By order of 13 June 1997, the Arrondissementsrechtbank, Maastricht, referred to the Court of Justice of
the European Communities for a preliminary ruling under art 177 of the EC Treaty two questions (set out
at p 616 e f, post) concerning the interpretation of Council Directive (EEC) 83/189 laying down a
procedure for the provision of information in the field of technical standards and regulations. Those
questions were raised in criminal proceedings instituted against Mr Lemmens, who was charged with
driving a vehicle while under the influence of alcohol. Written observations were submitted on behalf of:
the Netherlands government, by J G Lammers, Deputy Legal Adviser at the Ministry of Foreign Affairs,
acting as 605 agent; the United Kingdom government, by J E Collins, Assistant Treasury Solicitor, acting
as agent, and N Green, Barrister; and the European Commission, by B J Drijber, of its Legal Service, and
M Schotter, a national civil servant seconded to the Legal Service, acting as agents. Oral observations
were made by: the Netherlands government, represented by M Fierstra, Deputy Legal Adviser at the
Ministry of Foreign Affairs, acting as agent; the French government, represented by R Loosli-Surrans,
Charg de Mission in the Directorate for Legal Affairs, Ministry of Foreign Affairs, acting as agent; and the
Commission, represented by B J Drijber. The language of the case was Dutch. The facts are set out in the
opinion of the Advocate General.

12 February 1998.

The Advocate General (N Fennelly)

delivered the following opinion.

IINTRODUCTION
1.
What happens if the breathalyser is not notified to Brussels? The present proceedings raise the
question of whether a defendant in criminal proceedings before a national court can rely upon the fact that
national provisions governing the use of breath-testing apparatus, which prima facie constitute technical
regulations for the purposes of Council Directive (EEC) 83/189 laying down a procedure for the provision
of information in the field of technical standards and regulations (OJ 1983 L109 p 8) 1, have not been
notified to the European Commission.
 1
The directive has subsequently been amended by Council Directive (EEC) 88/182 (OJ 1988 L81 p 75 (the 1988
directive), and by Council Directive (EC) 94/10 (OJ 1994 L100 p 30) (the 1994 directive); the question of which version
of the directive applies in principle to the present proceedings is examined in paras 11 and 12, below.

IIFACTUAL AND LEGAL BACKGROUND

(a) The facts and the order for reference

2.
The judgment of the Court of Justice of the European Communities in CIA Security International SA v
Signalson SA Case C-194/94 [1996] All ER (EC) 557, [1996] ECR I-2201 seems to have caused
something of a commotion in the Netherlands. The judgment prompted the government to draw up a list
of some 400 measures which might be considered technical regulations within the meaning of the
directive, which had not been notified and against which, therefore, the direct effect of art 8 of the
directive could in principle be invoked before the national courts. The matter was taken up by the press in
June 1997 and, following proceedings under the Netherlands law on openness in government, the list
was published on 21 July 1997. By then the Netherlands had notified to the Commission, in accordance
with the urgent procedure provided under the directive, the 1997 Regulation on breath analysis, whose
provisions are identical to the national measure which is at issue in the present proceedings.
3.
The facts as they appear from the order for reference and from information supplied by the
Netherlands in its observations are that the defendant in the main proceedings was accused of driving a
motor vehicle at a time when the level of alcohol in his breath exceeded the legal limit. Before the
Arrondissements- rechtbank (the district court), Maastricht, on 13 June 1997, the defendant made the
following statement:
606
I understand from the press that there are difficulties regarding the breath-sampling apparatus. I
refer to the fact that this apparatus has not been notified to Brussels and wonder what the
consequences, if any, of this could be for my case.
On the ground that a decision on certain points was necessary to enable it to give judgment, the
Arrondissementsrechtbank Maastricht referred the following questions to the Court of Justice:
(1) Can a defendant in criminal proceedings instituted for an offence contrary to Article 8(2)(a)
of the Wegenverkeerswet (Road Traffic Law) 1994 successfully rely on the application of the
Regeling Ademanalyse (Regulation on breath analysis) (Ned Stcrt 1987, 187), as amended, laying
down further rules concerning the requirements for breath-analysis apparatus and the tests which
such apparatus must undergo, being set aside where, for the purposes of the investigation
referred to in Article 8(2)(a) of the Wegenverkeerswet 1994, that Regulation is based, pursuant to
Article 65 of the Invoeringswet Wegenverkeerswet (Law implementing the Road Traffic Law) 1994,
on Article 163 of the Wegenverkeerswet viewed in conjunction with Article 5 of the Besluit
Alcoholonderzoeken (Decree on the conduct of tests to determine alcohol level) (Stb 1987, 432), as
amended, in view of the failure to notify the Regulation to the European Commission, as required
by Article 8 of Directive 83/189/EEC?
(2) Should a court in criminal proceedings of the kind referred to above disapply that Regulation
of its own motion on account of the failure to notify it as required?
4.
The Netherlands, the United Kingdom and the Commission submitted written observations. The
Netherlands, France and the Commission took part in the oral hearing.

(b) The national provisions

5.
Article 8(2)(a) of the Wegenverkeerswet 1994 reads in relevant part:
[it] shall be an offence for any person to drive or ride a vehicle after consuming intoxicating
liquor so that, upon testing, the alcohol content of his breath is shown to be higher than 220
micrograms of alcohol per litre of expelled breath.
Article 163 of the same Law establishes the procedure for carrying out breathalyser tests, while art
163(10) provides for the adoption of provisions for the implementation of this article and of art 160(5). The
Minister for Justice is charged with adopting measures to give effect to these implementing provisions.
6.
Article 3 of the Decree on alcohol testing of 24 September 1987, as amended to take account of the
1994 road traffic law, provides that breath analysis may only be carried out using breath-analysis
apparatus of a type designated by the Minister for Justice. Article 5 of the Decree charges the Minister for
Justice with laying down detailed requirements for breath-analysis apparatus and for testing such
apparatus.
7.
Articles 2 and 3 of the regulation on breath analysis of 25 September 1987, as amended (the 1987
regulation), provide for the testing and type approval of breath-testing apparatus by a control body in
accordance with paras 4.3 to 4.5 of Annex 1 thereto, concerning, respectively, the type approval test, the
first individual test and the repeat individual test.
 607

IIIOPINION

(a) Admissibility
8.
The factual and legal background set out in the order for reference is rather short on detail. None of
the member states which have submitted observations on the request has suggested that this should be
declared inadmissible on the ground that the information supplied is so defective as to prevent the court
interpreting the relevant provisions of Community law in the light of the situation at issue, as in
Telemarsicabruzzo SpA v Circostel Joined cases C-320322/90 [1993] ECR I-393, nor has the lack of
detail prevented their submitting pertinent observations on the main questions of Community law which
arise (see the judgments in Holdijk Joined cases 141143/81 [1982] ECR 1299 (para 6) and Criminal
proceedings against Sunino Case C-2/96 [1996] ECR I-1543 (para 5)). The Commission has suggested
that, while the court could declare the present request inadmissible, it should not do so in the present
case, as there is a Community interest in clarifying the scope of the judgment in CIA Security.
9.
It seems to me that, although the order for reference is extremely terse, the court has at its disposal
sufficient information on the factual and legal situation in the main proceedings to provide a useful
answer. The order records the fact that the defendant is charged with an offence of driving with excess
alcohol in his breath and specifies, in the questions, the national provisions whose non-notification is at
issue. To that extent this case bears comparison with Criminal proceedings against Gallotti Joined cases
58/95, C-75/95, C-112/95, C-119/95, C-123/95, C-135/95, C-140/95, C-141/95, C-154/95 and C-157/95
[1996] ECR I-4345. In that case the several orders for reference merely stated that the various
defendants were charged with offences under Italian legislation on waste and described the issue of
Community law whose interpretation was sought. The court held that
having regard to the very general nature of the questions submitted and the detailed
interpretation of Directive 91/156 set out by the national court in the grounds of the orders, the
Court has sufficient information to provide a helpful answer to those questions. (See [1996] ECR I-
4345 (para 9).)
Similarly, in Vaneetveld v Le Foyer SA Case C-316/93 [1994] ECR I-763 (para 13), the court held that the
information requirement was
less pressing where the questions relate to specific technical points and enable the court to
give a useful reply even where the national court has not given an exhaustive description of the
legal and factual situation.
10.
In my view, the sufficiency of the information supplied by the national court must be judged in the light
of the questions of Community law posed. Since the court in the present proceedings has raised an issue
of interpretation of Community law and adequately, if briefly, recited the relevant facts, the court is bound
to provide a ruling (see eg Leur-Bloem v Inspecteur der Belastingdienst/Ondernemingen Amsterdam 2
Case C-28/95 [1997] All ER (EC) 738, [1997] ECR I-4161 (paras 2527) and the cases cited).

(b) The temporal application of the directive

11.
A second preliminary matter which has been raised is the need to identify the version of the directive
which was applicable to the factual situation which gave rise to the present proceedings. The Commission
maintains that this is the 608 original version prior to the 1988 and 1994 amendments, as it is this version
which was in force at the time of the adoption of the 1987 regulation. The Netherlands and the United
Kingdom, on the other hand, appear to take the view that it is the version of the directive as amended by
both subsequent directives which applies, while France has not expressed any view on the matter.
12.
The successive modifications of the directive have in particular sought to extend and clarify its
provisions; while the concept of technical regulation has been widened on each occasion, these
amendments do not appear germane to the evaluation of the national provisions at issue in the present
case. It is therefore permissible to use the definition of this term provided in the original version of the
directive, as suggested by the Commission, as the modifications, even if they were found to apply on the
facts, would not affect the classification of the 1987 regulation.

(c) The substantive questions referred

13.
The first substantive question which arises in the present case is whether the 1987 regulation is a
technical regulation within the meaning of the directive. In essence, a technical regulation is a set of
technical specifications the observance of which is compulsory, de jure or de facto, in the case of
marketing or use [of a product] in a member state or a major part thereof, except those laid down by local
authorities (art 1(5)). A technical specification is in turn defined by the directive as any specification
which lays down the characteristics required of a product (art 1(1)). Subject to certain exceptions which
are not here relevant, art 8 obliges member states to communicate to the Commission all draft technical
regulations, and, if necessary, the basic legislative or regulatory provisions. In CIA Security [1996] All ER
(EC) 557, [1996] ECR I-2201 (para 54), the court held that Directive 83/189 must be interpreted as
meaning that breach of the obligation to notify renders the technical regulations concerned inapplicable,
so that they are unenforceable against individuals. It subsequently held in Bic Benelux SA v Belgium
Case C-13/96 [1997] ECR I-1753 (para 19) that the technical regulations caught by the directive were any
national measures which are capable of hindering, directly or indirectly, actually or potentially, intra-
Community trade in goods.
14.
The Netherlands has argued that the 1987 regulation does not seek to implement a provision of
Community law and, being a provision of criminal law, governs a matter outside the scope of Community
law. It therefore argues that, for these reasons, the regulation may not be considered a technical
regulation.
15.
I do not agree. It is clear from the established case law of the court that [although] in principle criminal
legislation and the rules of criminal procedure are matters for which the member states are responsible
Community law sets certain limits to their power (see the judgments in Cowan v Trsor public Case
186/87 [1989] ECR 195 (para 19) and Criminal proceedings against Casati Case 203/80 [1981] ECR
2595 (para 27)). The directive does not allow any derogation in favour of provisions of criminal law which
could constitute technical regulations and, as a matter of principle, the obligation to notify could therefore
arise in respect of such provisions.
16.
In the light of the judgments in CIA Security and Bic Benelux, the following criteria apply for
determining whether a national measure is a technical regulation:does it lay down technical
specifications?2 is its observance 609 compulsory, de jure or de facto, in the case of marketing or use in
a member state or a major part thereof? and is it capable of hindering, directly or indirectly, actually or
potentially, intra-Community trade in goods?
2
Where the directive as amended by the 1994 Directive applies to a particular set of facts, the words or other
requirements should be added; this term is defined in art 1(3) of the amended Directive.

17.
There is little doubt that the 1987 regulation contains technical specifications; Annex 1 lays down in
detail the characteristics required of breath-testing apparatus, particularly as regards quality,
performance, testing and test methods and conformity-assessment procedures. There is also little doubt
that observance of the regulation is compulsory de jure, in the sense that the police must use approved
apparatus in order to prove the essence of the offence. There remains the question of whether it could be
said to be capable of hindering trade within the Bic Benelux test.
18.
It may be true, as the Netherlands has observed, that the 1987 regulation sets out to ensure the
reliability of breath-analysis apparatus, rather than to establish rules regarding the marketing or use of
such apparatus. It is in my view clear from the courts case law, however, that it is the potential effects on
trade of a national regulation which give rise to the obligation to notify, rather than the objective these
provisions seek to attain. In European Commission v Netherlands Case C-273/94 [1996] ECR I-31 (paras
1415), the court held that an order defining the conditions under which substitute products could be
marketed as margarine was a technical regulation:
(14) [that] assessment cannot be undermined by the Netherlands Governments argument that
the effect of the disputed order is to encourage the marketing of margarine
(15) [the] obligation [to notify] cannot be subject to the unilateral assessment by the Member
State which drafted the regulation of the effects which it may have on trade between Member
States.
More generally, in Bic Benelux [1997] ECR I-1753 (paras 1920), the court held that hindrances to trade
may arise from the adoption of national technical regulations irrespective of the grounds on which they
were adopted, and continued:
(20) the fact that a national measure was adopted in order to protect the environment or that
it does not implement a technical standard which may itself constitute a barrier to free movement
does not mean that the measure in question cannot be a technical regulation within the meaning of
Directive 83/189.
19.
The Netherlands has also pointed out that the 1987 regulation does not apply globally to the market in
breath-testing apparatus, but only to a single purchaser, the police, and that apparatus which is not in
conformity with the regulation may be fully marketed and used. It cites the example of retailers of
intoxicating beverages who may wish to make such apparatus available to their customers.
20.
It may be the case that there exists, or could exist, a secondary market in breath-testing apparatus;
this does not, in my opinion, suffice to exclude the possibility that the 1987 regulation is capable of
hindering trade between the member states in such goods. As the court held in Criminal proceedings
against Van de Haar Joined cases 177178/82 [1984] ECR 1797 (para 13):
Article 30 of the Treaty does not distinguish between measures having an effect equivalent to
quantitative restrictions according to the degree to which trade between Member States is affected.
If a national measure is capable of 610 hindering imports it must be regarded as a measure having
an effect equivalent to a quantitative restriction, even though the hindrance is slight and even
though it is possible for imported products to be marketed in other ways.
 As the directive embraces all potential obstacles to trade, the obligation to notify must, in my view,
include measures such as the 1987 regulation, even if it were to be shown that other breath-testing
apparatus could be marketed in the Netherlands. I am therefore of the opinion that the 1987 regulation
constitutes a technical regulation for the purposes of the directive.
21.
The next and more difficult issue which arises is whether an individual in a legal situation such as that
of the defendant in national criminal proceedings can, as a matter of Community law, rely on the
Netherlands failure to notify the 1987 regulation. The question raises the delicate problem of ascertaining
who may rely on Community law before a national court and the extent of the direct effect of provisions of
directives (see the opinion of Advocate General Darmon in Verholen v Sociale Verzekeringsbank
Amsterdam Joined cases C-8789/90 [1991] ECR I-3757 (para 32)).
22.
In this context, it is perhaps useful to recall that, in accordance with the third para of art 189 of the
Treaty, [a] directive shall be binding, as to the result to be achieved, upon each member state to which it
is addressed. It therefore follows that results which could arise as a consequence of the application of a
directive but which are not those it seeks to achieve are in principle not binding, as a matter of Community
law, on the member state.
23.
The results the directive at issue in the present proceedings seeks to achieve were identified by the
court in CIA Security [1996] All ER (EC) 557, [1996] ECR I-2201 (para 40) in the following terms:
Directive 83/189 is designed to protect, by means of preventive control, freedom of movement
for goods, which is one of the foundations of the Community. This control serves a useful purpose
in that technical regulations covered by the directive may constitute obstacles to trade in goods
between member states, such obstacles being permissible only if they are necessary to satisfy
compelling public interest requirements.
24.
After holding that arts 8 and 9 of the directive were unconditional and sufficiently precise to be relied
upon against conflicting national provisions, the court went on to establish ([1996] All ER (EC) 557, [1996]
ECR I-2201 (para 45)):
the legal consequences to be drawn from a breach by the member states of their obligation to
notify and, more precisely, whether Directive 83/189 is to be interpreted as meaning that a breach
of the obligation to notify, constituting a procedural defect in the adoption of the technical
regulations concerned, renders such technical regulations inapplicable so that they may not be
enforced against individuals.
This same formulation is repeated in the courts conclusion on this point (see [1996] All ER (EC) 557,
[1996] ECR I-2201 (para 54)).
25.
For the purposes of the directive, the consequence of the non-notification of a technical regulation is
that a member state authority may not enforce this against individuals. It does not follow that non-notified
technical regulations are inapplicable for all purposes, and hence in effect null and void; such a
consequence would only arise if the Community had a power to annul provisions of national 611 law.
Such a power has never been claimed by the court. The qualification by the court of the obligation to
notify as procedural does not imply that it is in principle any different from the general obligation properly
to transpose, implement and enforce substantive provisions of directives. Nor does it imply that the failure
to respect such a procedural obligation has legal effects any different from those which arise from the
failure to respect substantive obligations.
26.
The inapplicability of national provisions which are on their face incompatible with Community
provisions is not universal in its effects, but depends on the result the directive seeks to achieve. The
conflicting national provisions are, therefore, only inapplicable in so far as this is required for the purposes
of [facilitating] the free movement of goods by preventing the creation of new barriers to trade, as the
objective of the directive is described in the second recital in the preamble to the 1988 Directive. In the
circumstances of the present case, I cannot see how the non-application of the 1987 regulation in the
criminal proceedings against the defendant would contribute to facilitating the free movement of goods.
Such a consequence is therefore not encompassed within the result to be achieved by the directive.
27.
The direct effect of provisions of directives may be limited ratione personae in accordance with art 189
of the Treaty. In Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325 (para
22), for example, the court noted that
the case law on the possibility of relying on directives against state entities is based on the fact
that under art 189 a directive is binding only in relation to each Member State to which it is
addressed. That case law seeks to prevent the state from taking advantage of its own failure to
comply with Community law.
In such cases, a provision of national law may be ruled inapplicable for some purposes, but remain fully
applicable to situations outside the reach of Community law.
28.
The provisions of a directive which have direct effect may be relied upon by persons other than those
for whose primary benefit it was adopted, where this is justified to achieve the result intended. Thus, in
Ministre Public v Stoeckel Case C-345/89 [1991] ECR I-40473, an employer was entitled to rely upon the
injunction that member states ensure that any laws, regulations and administrative provisions contrary
to the principle of equal treatment shall be abolished laid down in art 5(2)(a) of Directive (EEC) 76/207 on
the implementation of the principle of equal treatment for men and women as regards access to
employment, vocational training and promotions, and working conditions (OJ 1976 L39 p 40). In this case,
the objective of the directive was defined in art 5(1) primarily in terms of workers rights: [application] of
the principle of equal treatment means that men and women shall be guaranteed the same conditions
without discrimination on grounds of sex. Admittedly, this decision may be explained on the basis that the
useful effect of a prohibition of discrimination against workers inescapably involves their employers, at
least. That is undoubtedly so, in so far as it imposes obligations. But Stoeckel is an example of an
employer claiming the benefit of it.

3
See also my opinion in Clean Car Autoservice GesmbH v Landeshauptman von Wien Case C-350/96 [1998] All ER
(EC) 434.

612
29.
In Verholen [1991] ECR I-3757 (para 23), on the other hand, a case which has been cited by a number
of the parties to these proceedings, the court held that
the right to rely on the provisions of Directive 79/7 is not confined to individuals coming within
the scope ratione personae of the directive, in so far as the possibility cannot be ruled out that other
persons may have a direct interest in ensuring that the principle of non-discrimination is respected
as regards persons who are protected.
This judgment has been interpreted in some quarters as meaning that not only those within the personal
scope of a directive, but also those having a direct interest in the application of a directives provision
must be able to rely on it (see Prechal Directives in European Community Law (1995) Proefschrift
Universiteit Amsterdam, SLSN, p 167).
30.
It can, of course, be argued that the defendant in the main proceedings has an interest in a literal
sense in the application of the directive. He would benefit from a declaration by the national court that the
1987 regulation is inapplicable in his case because this might lead to the exclusion of the proof on which
the criminal charges laid against him are based. It seems to me, however, that the expression direct
interest used by the court in Verholen does not cover such merely consequential effects but refers to an
interest which arises by virtue of Community law. Article 4(1) of Directive 79/7 defines its scope ratione
materiae as covering the calculation of benefits including increases due in respect of a spouse and for
dependants and the conditions governing the duration and retention of entitlement of benefits. In the
main proceedings in that case, the amount of the pension to which the applicant before the national court,
Mr Heiderijk, was entitled depended directly on the periods for which his wife was an insured person;
under the relevant national provisions, a pension representing rights acquired by and on behalf of each of
the two spouses was paid to the husband alone (as noted by Advocate General Darmon in his opinion in
Verholen [1991] ECR I-3757 (para 35)). The court held that an individual who bears the effects of a
discriminatory national provision may be allowed to rely on Directive 79/7 only if his wife, who is the victim
of the discrimination, herself comes within the scope of that directive (see the judgment in Verholen
[1991] ECR I-3757 (para 25)).
31.
Although the calculation of the increase of benefits for spouses was within the material scope of
Directive 79/7, Mr Heiderijks interest in its application was held to depend on his spouses coming within
its terms; if Mrs Heiderijk were found not to come within the personal scope of the directive, then her
husbands reliance on art 4(1) would not contribute to the implementation of the principle of equal
treatment, and he therefore had no Community-law interest in ensuring the application of this provision. It
appears to me that the defendant in the present proceedings is in a position analogous to that of Mr
Heiderijk, assuming that the latters wife was outside the personal scope of Directive 79/7; though each
may have an interest in the application of the relevant provisions of the directive, this interest arises
exclusively as a matter of national law, and they may not therefore rely upon the directive in that context.
Contrariwise, both the employer in Stoeckel, and Mr Heiderijk if his wife was within the personal scope of
Directive 79/7, had an interest arising from Community law in the proper application of the relevant
provision of the directive.
32.
In the case of Directive 83/189, it seems to me that those who would be able to rely on the Treaty rules
guaranteeing the free movement of goods have an interest as against the member state in the application
of the directive, and may 613 therefore rely upon it in the national court. Such, for example, was clearly
the case of the applicant in CIA Security, whose alarm system comprised goods manufactured in two
other member states as well as Belgium (see the opinion of Advocate General Elmer: [1996] All ER (EC)
557, [1996] ECR I-2201 (para 29)). Once an interest which was cognisable in Community law had been
established, the fact that the applicants interest arose in proceedings under national law against
commercial competitors was essentially irrelevant. Equally clearly, in my view, the defendant in the
present case has no such interest; no element of his legal situation could be said to concern the free
movement of goods within the Community.
33.
 In the present case, it has not been suggested that the national provision which the Netherlands is
enforcing against the defendant, to wit, art 8(2)(a) of the Wegenverkeerswet, is a technical regulation for
the purposes of the directive.
Contrariwise, while the 1987 regulation may be a technical regulation, its enforcement against the
defendant in circumstances such as the present is not a matter of which he is entitled to complain as a
matter of Community law. He is not affected by provisions defining the technical characteristics of such
apparatus by reason of any personal or professional characteristic or activity which comes within the
ambit of Community law. If the defendant is not so affected by the 1987 regulation considered alone, the
fact that he may be concerned by the application of this regulation by virtue of national provisions of
criminal law defining the methods of proving a road traffic offence does not change his position under
Community law.
34.
It follows from the above that the first question should be answered in the negative, in which case no
answer to the second question is necessary.

IV CONCLUSION
35.
In the light of the foregoing, I recommend to the Court of Justice that it answer the questions referred
by the Arrondissementsrechtbank Maastricht on 13 June 1997 as follows:
Council Directive (EEC) 83/189 laying down a procedure for the provision of information in the
field of technical standards and regulations should be interpreted as not requiring a member state
to abstain from applying national provisions such as the Regeling Ademanalyse, which have not
been notified to the Commission in accordance with art 8 of the directive, in criminal proceedings
instituted for an offence contrary to art 8(2)(a) of the Wegenverkeerswet 1994.

16 June 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 13 June 1997, received at the Court of Justice of the European Communities on 18 June
1997, the Arrondissementsrechtbank, Maastricht, referred to the Court of Justice for a preliminary ruling
under art 177 of the EC Treaty two questions concerning the interpretation of Council Directive (EEC)
83/189 laying down a procedure for the provision of information in the field of technical standards and
regulations (OJ 1983 L109 p 8).
2.
Those questions were raised in criminal proceedings instituted against Mr Lemmens, who was
charged with driving a vehicle while under the influence of alcohol.
3.
Article 8(2)(a) of the Wegenverkeerswet 1994 (the Road Traffic Law, Stb 1995, 475) makes it an
offence for any person to drive or ride a vehicle after 614 consuming intoxicating liquor so that, upon
testing, the alcohol content of his breath is shown to be higher than 220 micrograms of alcohol per litre of
expelled breath.
4.
Article 163(1) and (2) of that Law is worded as follows:
(1) If he suspects the driver of a vehicle of infringing Article 8, the police officer may order him to
undergo a test within the meaning of Article 8(2)(a). (2) The driver to whom the order referred to in
paragraph 1 is addressed is required to exhale his breath in an apparatus used for the test and to
comply with all the instructions given to him by the police officer in order to ensure that the test is
properly carried out.
5.
Article 163(10) provides that the Minister for Justice is to lay down, by an administrative regulation,
further rules for the application of those provisions.
6.
The Besluit Alcoholonderzoeken (the Decree on the conduct of tests to determine the alcohol level) of
24 September 1987 (Stb 432), as frequently amended (the 1987 Decree), provides in art 3 that in order to
carry out a breathalyser test, breath-analysis apparatus must be used of a type designated by the Minister
for Justice, which presupposes that the apparatus has been approved following a test carried out by the
control body designated by the Minister for Justice. Article 5 of the 1987 Decree further provides that the
Minister for Justice is to lay down the requirements for breath-analysis apparatus and the rules for testing
such apparatus.
7.
The rules referred to in art 5 of the 1987 Decree were laid down by the Regeling Ademanalyse (the
regulation on breath analysis) of 25 September 1987 (Stcrt 1987), as frequently amended (the 1987
regulation). That regulation provides, in art 2(1), that in order to be capable of being used by the police for
measuring alcohol content, breath-analysis apparatus must be of a type approved in a test carried out by
the control body on the basis of para 4.3 of Annex 1 thereto and, in art 3(1), that the breath-analysis
apparatus must also have been approved in a test carried out by the control body on the basis of para 4.4
or 4.5 of that annex.
8.
Annex 1 to the 1987 regulation sets out the characteristics which breath- analysis apparatus must
exhibit, in particular so far as concerns quality, performance, testing and test methods as well as the
procedures for assessing conformity.
9.
Articles 8 and 9 of the directive, in the version initially in force at the time when the 1987 Decree and
the 1987 regulation were adopted, require member states to communicate to the European Commission
any draft technical regulation falling within the scope of the directive, and to postpone the adoption of
such draft regulations for three months, except in specific cases where urgent reasons are involved,
which are described in art 9(3). In certain circumstances, described in art 9(1) and (2), that period is to be
extended by three or by nine months.
10.
In its judgment in CIA Security International SA v Signalson SA Case C-194/94 [1996] All ER (EC) 557,
[1996] ECR I-2201 (para 54), the court interpreted the directive as meaning that breach of the obligation
to notify, imposed by arts 8 and 9 thereof, renders the technical regulations concerned inapplicable, so
that they are unenforceable against individuals. The court therefore held that individuals may rely on the
aforesaid provisions in proceedings before the national court which must decline to apply a national
technical regulation which has not been notified in accordance with the directive.
615
11.
Following that judgment, the Netherlands government drew up a list of national measures which
should possibly have been notified to the Commission in accordance with the directive and which
included both the 1987 Decree and the 1987 regulation. In an attempt to make up for lost time, the
Netherlands government notified the Commission of the 1997 Decree on the conduct of tests to
determine the alcohol level and the 1997 regulation on breath analysis, which are identical to their
predecessors from 1987.
12.
It is apparent from the order for reference that, in the course of the criminal proceedings instituted
against him, Mr Lemmens said:
I understand from the press that there are difficulties regarding the breath-analysis apparatus. I
maintain that this apparatus has not been notified to Brussels and wonder what the consequences
of this could be for my case.
13.
In those circumstances, the national court decided to stay proceedings and to refer the following
questions to the Court of Justice for a preliminary ruling:
(1) Can a defendant in criminal proceedings instituted for an offence contrary to Article 8(2)(a)
of the Wegenverkeerswet (Road Traffic Law) 1994 successfully rely on the application of the
Regeling Ademanalyse (Regulation on breath analysis) as amended, laying down further rules
concerning the requirements for breath-analysis apparatus and the tests which such apparatus
must undergo, being set aside where, for the purposes of the investigation referred to in Article
8(2)(a) of the Wegenverkeerswet 1994, that Regulation is based, pursuant to Article 65 of the
Invoeringswet Wegenverkeerswet (Law implementing the Road Traffic Law) 1994, on Article 163 of
the Wegenverkeerswet viewed in conjunction with Article 5 of the Besluit Alcoholonderzoeken
(Decree on the conduct of tests to determine alcohol level) as amended, in view of the failure
to notify the Regulation to the European Commission, as required by Article 8 of Directive
83/189/EEC?
(2) Should a court in criminal proceedings of the kind referred to above disapply that Regulation
of its own motion on account of the failure to notify it as required?
14.
As the Netherlands and French governments contend that the 1987 regulation is not a technical
regulation within the meaning of the directive, it is appropriate to examine this point at the outset.
15.
In that regard, it should be noted that the term technical regulation is defined in art 1(5) of the
directive as
technical specifications, including the relevant administrative provisions, the observance which
is compulsory, de jure or de facto, in the case of marketing or use in a Member State or a major
part thereof, except those laid down by local authorities.
According to art 1(1), the term technical specification is to be interpreted as meaning
a specification contained in a document which lays down the characteristics required of a
product such as levels of quality, performance, safety or dimensions, including the requirements
applicable to the product as regards terminology, symbols, testing and test methods, packaging,
marking or labelling.
616
16.
 The Netherlands and French governments submit that, although the 1987 regulation sets out the
characteristics for breath-analysis apparatus and the police is required to use approved apparatus in
order to establish that an infringement has actually been committed, the regulation does not fall within the
scope of the directive for two reasons.
17.
To begin with, the Netherlands government submits that the 1987 regulation applies in the field of
criminal law, which falls outside the sphere of Community law.
18.
The French government, for its part, contends that the directive does not apply to products which, like
those in this case, are intended to be used in connection with the exercise of public authority and a fortiori
in criminal proceedings instituted by the member states.
19.
Those arguments cannot be accepted. Although in principle criminal legislation and the rules of
criminal procedure are matters for which the member states are responsible, it does not follow that this
branch of the law cannot be affected by Community law (see, to that effect, Cowan v Trsor public Case
186/87 [1989] ECR 195 (para 19) and Criminal proceedings against Casati Case 203/80 [1981] ECR
2595 (para 27)).
20.
In this case, there is nothing in the directive to suggest that technical regulations within the meaning of
art 1 thereof are excluded from the notification requirement because they fall within the scope of criminal
law, or that the scope of the directive is limited to products intended to be used otherwise than in
connection with the exercise of public authority. As the court has already stated in Bic Benelux SA v
Belgium Case C-13/96 [1997] ECR I-1753 (para 19), a directive applies to technical regulations
irrespective of the grounds on which they were adopted.
21.
Secondly, the Netherlands government submits that the regulations in question apply not to persons
manufacturing or marketing breath-analysis apparatus, but only to a certain type of purchaser, namely the
police. Breath-analysis apparatus not in conformity with the 1987 regulation can be marketed and used
without any restrictions on the part of the market in such apparatus which is not concerned.
22.
Furthermore, according to the Netherlands government, the instructions contained in the 1987
regulation are addressed to police officers and are designed to ensure that breath-analysis apparatus is
reliable inasmuch as it provides evidence that a driver is under the influence of alcohol, but they do not
specify the conditions for marketing such apparatus.
23.
The French government states, along the same lines, that, in order to be classified as technical
regulations within the meaning of the directive, the regulations concerned must apply to products intended
for everyday use.
24.
Those arguments cannot be accepted. Admittedly, certain measures may impose, in respect of a
product intended for a particular group of users, technical specifications whose content depends on the
specific objective pursued by that group and which are too remote in terms of their relationship with the
production and marketing of the product to be classified as technical regulations within the meaning of the
directive. However, that is not the case here.
25.
Even on the assumption that in the Netherlands there is a market for breath-analysis apparatus that is
not in conformity with the 1987 regulation, the fact remains that the rules laid down by that regulation
must be complied with by those who sell such apparatus to the police, which is a major user of that
product on the Netherlands market.
617
26.
The 1987 regulation therefore constitutes a technical regulation which should, prior to its adoption,
have been notified to the Commission in accordance with art 8 of the directive.
27.
In its first question, the national court essentially asks whether the directive is to be interpreted as
meaning that, if the obligation imposed by art 8 thereof to notify a technical regulation on breath-analysis
apparatus has been infringed, the effect is that evidence obtained by means of such apparatus,
authorised in accordance with that regulation, cannot be relied upon against an individual charged with
driving while under the influence of alcohol.
28.
The Commission and the governments which have submitted observations to the court consider that
this question calls for a negative answer.
29.
The Netherlands government submits, in particular, that, in this case, the accused has no interest
worthy of protection in pleading that the technical regulation is inapplicable. Since the aim of the directive
is to guarantee the free movement of goods, only individuals who manufacture or import goods have a
direct interest in the technical rules contained in the 1987 regulation having been verified on the basis of
the directive.
30.
According to the UK government, it follows from the judgment in CIA Security that it is the technical
regulation itself, which has not been notified, that cannot be enforced against an individual. The directive,
which is aimed merely at removing obstacles to trade, does not have as its purpose to render unlawful the
use of a product marketed in accordance with a technical regulation which has not been notified.
31.
The Commission and the French government submit that, while the conditions for art 8 of the directive
to have direct effect and the penalty for breach of that provision, namely that a regulation which has not
been notified may not be enforced against individuals, are laid down in mandatory terms by Community
law, it is for national law to determine the content and specific consequences of that penalty, on the
understanding, first, that the conditions in which an individual may rely on the breach of Community law
cannot be less favourable to him than those which he must comply with in the event of a comparable
breach of national law and, secondly, that the effectiveness of Community law must remain intact. The
Commission therefore considers that Community law does not preclude the application of a principle of
national law to the effect that breach of the obligation to notify technical regulations does not render
inapplicable the legislation on driving while under the influence of alcohol, provided the accused is also
precluded from relying on breach of a similar obligation under national law.
32.
In that regard, it should be noted that in its judgment in CIA Security [1996] All ER (EC) 557, [1996]
ECR I-2201 (para 40), the court emphasised that the directive is designed to protect, by means of
preventive control, freedom of movement for goods, which is one of the foundations of the Community.
This control serves a useful purpose in that technical regulations covered by the directive may constitute
obstacles to trade in goods between member states, such obstacles being permissible only if they are
necessary to satisfy compelling requirements relating to the public interest.
33.
In that judgment, the court pointed out that the obligation to notify is essential for achieving such
Community control and went on to state that the effectiveness of such control will be that much greater if
the directive is interpreted as meaning that breach of the obligation to notify constitutes a substantial
procedural defect such as to render the technical regulations in question 618 inapplicable, and thus
unenforceable against individuals (see [1996] All ER (EC) 557, [1996] ECR I-2201 (paras 48, 54)).
34.
In criminal proceedings such as those in the main action, the regulations applied to the accused are
those which, on the one hand, prohibit and penalise driving while under the influence of alcohol and, on
the other, require a driver to exhale his breath into an apparatus designed to measure the alcohol content,
the result of that test constituting evidence in criminal proceedings. Such regulations differ from those
which, not having been notified to the Commission in accordance with the directive, are unenforceable
against individuals.
35.
While failure to notify technical regulations, which constitutes a procedural defect in their adoption,
renders such regulations inapplicable inasmuch as they hinder the use or marketing of a product which is
not in conformity therewith, it does not have the effect of rendering unlawful any use of a product which is
in conformity with regulations which have not been notified.
36.
The use of the product by the public authorities, in a case such as this, is not liable to create an
obstacle to trade which could have been avoided if the notification procedure had been followed.
37.
The answer to the first question must therefore be that the directive is to be interpreted as meaning
that breach of the obligation imposed by art 8 thereof to notify a technical regulation on breath-analysis
apparatus does not have the effect of making it impossible for evidence obtained by means of such
apparatus, authorised in accordance with regulations which have not been notified, to be relied upon
against an individual charged with driving while under the influence of alcohol.
38.
In the light of the answer to the first question, there is no need to answer the second question.

Costs
39.
The costs incurred by the Netherlands, French and UK governments, and by the European
Commission, which have submitted observations to the Court of Justice, are not recoverable. Since these
proceedings are, for the parties to the main proceedings, a step in the proceedings pending before the
national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the
Arrondissementsrechtbank, Maastricht, by order of 13 June 1997, hereby rules: Council Directive (EEC)
83/189 laying down a procedure for the provision of information in the field of technical standards and
regulations is to be interpreted as meaning that breach of the obligation imposed by art 8 thereof to notify
a technical regulation on breath-analysis apparatus does not have the effect of making it impossible for
evidence obtained by means of such apparatus, authorised in accordance with regulations which have
not been notified, to be relied upon against an individual charged with driving while under the influence of
alcohol.

619
 [1998] All ER (EC) 620

Stichting Greenpeace Council (Greenpeace International) and others v

European Commission
(Case C-321/95 P)
EUROPEAN COMMUNITY; Environment, Other European Community
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, RAGNEMALM, WATHELET
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA (RAPPORTEUR), KAPTEYN,
MURRAY, EDWARD, PUISSOCHET, HIRSCH, JANN AND SEVN ADVOCATE GENERAL COSMAS
17 JUNE, 23 SEPTEMBER 1997, 2 APRIL 1998
European Community Action for annulment Locus standi Commission authorising expenditure on
project to build power stations Authorities and builders failing to carry out environmental impact
assessments Nature conservancy foundation bringing action for annulment of Commissions
authorisation of further finance Whether foundation having locus standi EC Treaty, art 173.

The European Commission granted financial assistance to Spain for the building of two power stations by
a company, Unelco, on the Canary Islands. Several private individuals subsequently informed the
Commission that Unelco had started work before the planning and environmental authorities had issued a
declaration of environmental impact required under national legislation and had itself failed to undertake
an environmental impact assessment in accordance with Council Directive (EEC) 85/337 on the
assessment of the effects of certain public and private projects on the environment. When those
authorities did publish the required declarations of environmental impact, they were challenged in the
administrative courts by a local environmental protection organisation, and the validity of authorisations
issued to Unelco by the regional Ministry of Industry Commerce and Consumption were challenged by the
Spanish branch of the international nature conservancy foundation, Greenpeace. Following a meeting
between Greenpeace and the Commission at which the Commission confirmed the total amount of
money released to the power station project so far, Greenpeace and others brought an action before the
Court of First Instance of the European Communities seeking annulment of a decision allegedly taken by
the Commission to release a further tranche of aid to Spain in relation to the project. The Commission
raised an objection of inadmissibility concerning, inter alia, Greenpeaces locus standi, which was upheld
by the Court of First Instance on the basis that Greenpeace was not directly or individually concerned by
the measure for the purposes of art 1731 of the EC Treaty. Greenpeace appealed to the Court of Justice.
1
Article 173, so far as material, is set out at p 629 c, post

Held For the purposes of art 173 of the Treaty, a natural person was not directly or individually
concerned by a contested act where his specific situation was not taken into consideration in the adoption
of that act. The same applied to associations which claimed to have locus standi on the basis of the fact
that the persons they represented were individually concerned by the contested decision. Moreover,
since, in the instant case, it was the building of the two power stations which was 620 liable to affect the
environmental rights ensuing under Directive 85/337, the contested decision could affect those rights only
indirectly. Further, although the subject matter of the proceedings Greenpeace had brought before the
national courts challenging the administrative authorisations issued to Unelco concerning the construction
of the power stations differed from the subject matter of the action brought before the Court of First
Instance, both actions were based on rights afforded to individuals by Directive 85/337. Accordingly, the
rights derived from Directive 85/337 were fully protected by the national courts which could, if necessary,
refer a question to the Court of Justice for a preliminary ruling under art 177. It followed that Greenpeace
did not have locus standi and that the Court of First Instance had not erred in law. The appeal would
therefore be dismissed (see p 672 a to j, post).
Associazione Tecnico Economica del Cemento v European Commission Joined cases T-447449/93
[1995] ECR II-1971, Kwekerij Gebroeders Van der Kooy BV v EC Commission Joined cases 67/85, 68/85
and 70/85 [1988] ECR 219 and Comit International de la Rayonne et des Fibres Synthtiques (CIRFS) v
EC Commission Case C-313/90 [1993] ECR I-1125 applied.

Notes
For locus standi in relation to actions for annulment, see 51 Halsburys Statutes (4th edn) paras 239
242.
For the EC Treaty art 173 (as amended by art G.53 of the Treaty on European Union), see 50
Halsburys Statutes (4th edn), Current Service, 97.

Cases cited
Aannemersbedrijf PK Kraaijeveld BV v Gedeputeerde Staten van Zuid-Holland Case C-72/95 [1995] All
ER (EC) 134, [1996] ECR I-5431, ECJ.
Abertal SAT Ltda v EC Council Case C-264/91 [1993] ECR I-3265.
Air France (Socit Anonyme Participation Ouvrire Cie) v European Commission Case T-2/93 [1994]
ECR II-323.
Air France (Socit Anonyme Participation Ouvrire Cie) v European Commission Case T-3/93 [1994]
ECR II-121.
Alusuisse Italia SpA v EC Council Case 307/81 [1982] ECR 3463.
An TaisceThe National Trust for Ireland v European Commission Case C-325/94 P [1996] ECR I-3727.
Antillean Rice Mills NV v EC Commission Joined cases T-480/93 and T-483/93 [1995] ECR II-2305.
Arcaro (Criminal proceedings against) Case C-168/95 [1997] All ER 82, [1996] ECR I-4705, ECJ.
Asociacin Profesional de Empresarios de Pesca Comunitarios (Apesco) v EC Commission Case 207/86
[1988] ECR 2151.
Association pour la Protection des Animaux Sauvages v Prfet de Maine et Loire Case C-435/92 [1994]
ECR I-67.
Associazione Agricoltori della Provincia di Rovigo v European Commission Case T-117/94 [1995] ECR II-
455.
Associazione Italiana per il World Wildlife Fund v Regione Veneto Case C-118/94 [1996] ECR I-1223.
Associazione Tecnico Economica del Cemento v European Commission Joined cases T-447449/93
[1995] ECR II-1971.
Benzine en Petroleum Handelsmaatschappij BV v EC Commission Case 77/77 [1978] ECR 1513.
Bock v EC Commission Case 62/70 [1971] ECR 897.
621
Buralux SA v EU Council Case C-209/94 P [1996] ECR I-615.
CMC Cooperativa muratori e cementisti v EC Commission Case 118/83 [1985] ECR 2325.
Codorniu SA v EU Council Case C-309/89 [1994] ECR I-1853.
Comitato di Coordinamento per la Difesa della Cava v Regione Lombardia Case C-236/92 [1994] I-483.
Comit de dveloppement et de promotion du textile et de lhabillement (DEFI) v EC Commission Case
282/85 [1986] ECR 2469.
Comit International de la Rayonne et des Fibres Synthtiques (CIRFS) v EC Commission Case C-
313/90 [1993] ECR I-1125.
Compagnie franaise de lazote (Cofaz) SA v EC Commission Case 169/84 [1986] ECR 391.
Confederazione italiana dirigenti di azienda (CIDA) v EC Council Case 297/86 [1988] ECR 3531.
Consorzio Gruppo di Azione Locale Murgia Messapica v EC Commission Case T-465/93 [1994] ECR II-
361.
Deutz und Geldermann, Sektkellerei Breisach/Baden GmbH v EC Council Case 26/86 [1987] ECR 941.
Diego Cal & Figli Srl v Servici Ecologici Porto di Genova Spa (SEPG) Case C-343/95 [1997] ECR I-1547.
EC Commission v Belgium Case 247/85 [1987] ECR 3029.
EC Commission v Belgium Case C-2/90 [1992] ECR I-4431.
EC Commission v Denmark Case 302/86 [1988] ECR 4607.
EC Commission v EC Council Case C-155/91 [1993] ECR I-939.
EC Commission v France Case 252/85 [1988] ECR 2243.
EC Commission v Germany Case C-288/88 [1990] ECR I-2721.
EC Commission v Germany Case C-131/88 [1991] ECR I-825.
EC Commission v Germany Case C-361/88 [1991] ECR I-2567.
EC Commission v Germany Case C-195/90 [1992] ECR I-3141.
EC Commission v Germany Case C-431/92 [1995] ECR I-2189.
EC Commission v Netherlands Case 236/85 [1987] ECR 3989.
EC Commission v Spain Case C-355/90 [1993] ECR I-4221.
EEC Seed Crushers and Oil Producers Federation (Fediol) v EC Commission Case 191/82 [1983] ECR
2913.
European Commission v Greece Joined cases C-232233/95 (1997) Transcript, 11 June, ECJ.
Extramet Industrie SA v EC Council Case C-358/89 [1991] ECR I-2501.
Fdration nationale de la boucherie en gros et du commerce en gros des viandes v EEC Council Joined
cases 1922/62 [1962] ECR 491.
Fdration Nationale des Producteurs de Vins de Table et Vins de Pays v EC Commission Case 60/79
[1979] ECR 2429.
Fdration Rgionale des Syndicats dExploitants Agricoles (FRSEA) v EC Council Case T-476/93 [1993]
ECR II-1187.
Geotronics v European Commission Case C-395/95 P [1997] ECR I-2271.
International Business Machines Corp v EC Commission Case 60/81 [1981] ECR 2639.
Italsolar SpA v EC Commission Case C-257/90 [1993] ECR I-9.
Johnston v Chief Constable of the Royal Ulster Constabulary Case C-222/84 [1986] 3 All ER 135, [1987]
QB 129, [1986] 3 WLR 1038, [1986] ECR 1651, ECJ.
Kwekerij Gebroeders Van der Kooy BV v EC Commission Joined cases 67/85, 68/85 and 70/85 [1988]
ECR 219.
Lezzi Pietro & C Srl v EC Commission Case C-123/92 [1993] ECR I-809.
Matra SA v EC Commission Case C-225/91 [1993] ECR I-3203.
Metro SB Gromrkte GmbH & Co KG v EC Commission Case 26/76 [1977] ECR 1875.
622
Ministre public v Traen Joined cases 372374/85 [1987] ECR 2141.
Moksel Import-Export GmbH & Co Handels KG v EC Commission Case 45/81 [1982] ECR 1129.
Neotype Techmaschexport GmbH v EC Commission Joined cases C-305/86 and C-160/87 [1990] ECR I-
2945.
Nuova Ceam Srl v EC Commission Case 205/87 [1987] ECR 4427.
Oleificio Borelli SpA v EC Commission Case C-97/91 [1992] ECR I-6313.
Parti cologiste Les Verts v European Parliament Case 294/83 [1986] ECR 1339.
Peralta (Criminal proceedings against) Case C-379/92 [1994] ECR I-3453.
Piraiki-Patraiki v EC Commission Case 11/82 [1985] ECR 207.
Plaumann & Co v EEC Commission Case 25/62 [1963] ECR 95.
Procureur de la Rpublique v Association de dfense des brleurs dhuiles usages Case 240/83 [1985]
ECR 531.
R v Secretary of State for the Environment, ex p Royal Society for the Protection of Birds Case C-44/95
[1996] ECR I-3805.
R v Secretary of State for Trade and Industry, ex p Duddridge [1995] ELR 151, CA.
Sermes (SA) v EC Commission Case 279/86 [1987] ECR 3109.
Sierra Club v Morton (1972) 405 US 727, 31 L Ed 2d 636, US SC.
Societ Eridania Zuccherifici Nazionali v EEC Commission Joined cases 10/68 and 18/68 [1969] ECR
459.
Sofrimport SARL v EC Commission Case C-152/88 [1990] ECR I-2477.
Spijker Kwasten BV v EC Commission Case 231/82 [1983] ECR 2559.
Star Fruit Company SA v EC Commission Case 247/87 [1989] ECR 291.
STS Consorzio per Sistemi di Telecommunicazione via Satellite SpA v EC Commission Case 126/83
[1984] ECR 2769.
Timex Corp v EC Council Case 264/82 [1985] ECR 849.
Toepfer v EEC Commission Joined cases 106107/63 [1965] ECR 405.
Tombesi (Criminal proceedings against) Joined cases C-304/94, C-330/94, C-342/94 and C-224/95
[1997] All ER (EC) 639, [1997] ECR I-3561, ECJ.
Union de Federaciones Agrarias de Espaa (UFADE) v EC Council Case 117/86 [1986] ECR 3255.
Union Deutsche Lebensmittelwerke GmbH v EC Commission Case 97/85 [1987] ECR 2265.
Union nationale des entraneurs et Cadres techniques professionnels du football (Unectef) v Heylens
Case 222/86 [1987] ECR 4097.
Union syndicale-Service public europen v EC Council Case 72/74 [1975] ECR 401.
Van Parijs v EC Council Case C-257/93 [1993] ECR I-3335.
von Bonkewitz-Lindner v European Parliament Case 13/86 [1987] ECR 1417.
William Cook plc v EC Commission Case C-198/91 [1993] ECR I-2487.
William v EC Commission Case 246/81 [1982] ECR 2277.
WWF UK v European Commission Case T-105/95 [1997] All ER (EC) 300, [1997] ECR II-313, CFI.
Zuckerfabrik Watenstedt GmbH v EC Council Case 6/68 [1968] ECR 409.

Appeal
Stichting Greenpeace Council (Greenpeace International) and others appealed against the order of the
Court of First Instance of the European Communities ([1995] ECR II-2205) dismissing the action brought
by them for annulment of the Commission Decision to pay to the Spanish government, in addition to
amounts initially granted, ECU 12m to cover expenditure incurred in the construction of two electricity
power stations in the Canary Islands. Greenpeace and others were represented by P Sands and M
Hoskins, Barristers, instructed by Leigh, Day & Co, 623Solicitors, with an address for service in
Luxembourg at the Chambers of Jean-Paul Noesen, 18 Rue des Glacis. The European Commission was
represented by P Oliver, of its Legal Service, acting as agent, with an address for service in Luxembourg
at the office of Carlos Gmez de la Cruz, of its Legal Service, Wagner Centre, Kirchberg. The Kingdom of
Spain was represented by A J Navarro Gonzlez, Director-General for Legal and Institutional
Coordination in Community Matters, and G Calvo Daz, Abogado del Estado, of the State Legal Service
for matters before the Court of Justice of the European Communities, acting as agents, with an address
for service in Luxembourg at the Spanish Embassy, 46 Boulevard Emmanuel Servais. The language of
the case was English. The facts are set out in the opinion of the Advocate General.
23 September 1997.

The Advocate General (G Cosmas)

delivered the following opinion (translated from the Greek).

IINTRODUCTORY OBSERVATIONS
1.
In the present case the Court of Justice of the European Communities is called upon to adjudicate on
the appeal brought under art 49 of the EC Statute of the Court of Justice against the order of 9 August
1995 of the Court of First Instance of the European Communities (Case T-585/93 [1995] ECR II-2205).
The order appealed against dismissed the action brought by certain environmental associations and
individuals, pursuant to the fourth paragraph of art 173 of the EC Treaty, for annulment of the
Commissions decision to pay to the Spanish government, in addition to amounts initially granted, ECU
12m to cover expenditure incurred in the construction of two electricity power stations in the Canary
Islands. That decision, according to the applicants, was taken between 7 March 1991, the date of
adoption of Commission Decision C(91)440 granting financial assistance for the works in question, and
29 October 1993, when the Commission confirmed that the above mentioned amounts had already been
disbursed to the Spanish government.
2.
At the outset, it is worth noting that the court will be required in this case to form a view on the manner
and terms of application of the fourth paragraph of art 173 of the EC Treaty in circumstances in which the
applicants found the legal interest on which they rely on the consequences for the environment alleged to
flow from the act at issue adopted by the Community institution (sic) 2.
2
Paragraph 2 of this opinion is subject to revision by the Court of Justices translation service.

IIFACTS AND PROCEDURE

3.
The facts of the case under examination, as described in the order appealed against, may be stated
as follows:
By Decision C(91)440 adopted on 7 March 1991 pursuant to Council Regulation (EEC) 1787/84 on the
European Regional Development Fund (OJ 1984 L169 p 1), as amended by Council Regulation (EEC)
3641/85 (OJ 1985 L350 p 40), the European Commission agreed to grant the Kingdom of Spain financial
assistance from the European Regional Development Fund (the ERDF) up to a maximum of ECU
108,578,419 for the construction of two electricity power stations in the Canary Islands, on Gran Canaria
and Tenerife, by Union Electrica de Canarias SA (UNELCO). That financing was to be spread over four
years, the first instalment being payable on adoption of aforementioned Decision C(91)440. Subsequent
instalments of aid could be reduced or suspended if examination of progress of the works were to reveal
irregularities and, in particular, any significant alteration affecting the implementation of the decision for
which the Commissions approval 624 had not been sought. That condition was laid down in art 5 of the
decision, and also in paras A.20, A.21 and C.2 of Annex III thereto.
4.
As early as 23 December 1991, two of the appellants had informed the Commission by letter that, as
regards the works commenced on Gran Canaria, UNELCO had failed to undertake an environmental
impact assessment study in accordance with the applicable Community provisions (see Council Directive
(EEC) 85/337 on the assessment of the effects of certain public and private projects on the environment
(OJ 1985 L175 p 40)). Furthermore, another of the appellants reported to the Commission, by letter dated
23 November 1992, that UNELCO had started construction work on Gran Canaria and Tenerife without
the Comisin de Urbanismo y Medjo Ambiente de Canarias (CUMAC) (the Canary Islands Commission
for Planning and the Environment) having issued its environmental impact statement in accordance with
the applicable national legislation.
5.
In fact, on 3 December 1992, CUMAC issued two environmental impact statements relating to the
construction of the power stations on Gran Canaria and Tenerife. Administrative appeals against those
statements were lodged, on 26 March 1993 and 2 April 1993 respectively, by local environmental
protection associations which are amongst the appellants in the present proceedings 3. On 18 December
1993 Greenpeace Spain, an environmental protection association responsible in Spain for the attainment
in that country of the objectives of Stichting Greenpeace Council (Greenpeace), a nature conservancy
foundation established in the Netherlands and the first appellant, brought proceedings before the national
courts challenging the validity of the administrative authorisations issued to UNELCO by the competent
Spanish administrative authorities.
3
Those associations are Tagoror Ecologista Alternativo (TEA) and Comisin Canaria contra la Contamincion (CIC).

6.
At the same time, Greenpeace initiated a series of contacts with the Commission. In particular, by
letter dated 17 March 1993, Greenpeace asked the Commission to confirm whether Community structural
funds had been disbursed for the construction of the two power stations in the Canary Islands and to
inform it of the date of release of those funds. In its reply of 13 April 1993, the Director General of
Directorate-General XVI of the Commission recommended Greenpeace to read Decision C(91) 440
which gave details of the conditions of the grant of Community support and of the financing plan. By a
further letter of 17 May 1993, Greenpeace asked the Commission for full disclosure of measures which it
had taken in regard to the construction of the two power stations in the Canary Islands, in accordance
with art 7 of Council Regulation (EEC) 2052/88 on the tasks of the Structural Funds and their
effectiveness and on coordination of their activities between themselves and with the operations of the
European Investment Bank and the other existing financial instruments (OJ 1988 L185 p 9), which
provides:
Measures financed by the Funds shall be in keeping with the provisions of the Treaties, with
the instruments adopted pursuant thereto and with Community policies, including those concerning
environmental protection.
By letter dated 23 June 1993, the Director General of DG XVI replied to Greenpeace that it was unable to
provide the information requested since it concerned the Commissions internal decision-making
procedures, but gave the assurance that its decision was taken only after full consultation between the
various services concerned. On 29 October 1993, a meeting took place in Brussels between Greenpeace
and the competent officials of Directorate-General XVI of the 625 Commission, specifically concerning the
financing by the ERDF of the construction of the power stations on Gran Canaria and Tenerife.
7.
On 21 December 1993, the appellants brought an action before the Court of First Instance for
annulment of the above mentioned Commission decision to continue financing of the works. The Kingdom
of Spain intervened in the proceedings before the Court of First Instance. In its above mentioned order of
9 August 1995 the Court of First Instance upheld the objection of inadmissibility raised by the Commission
and on that ground dismissed the action.
8.
The unsuccessful parties at first instance appealed against that order, requesting the court to set aside
the order, to declare admissible the action brought before the Court of First Instance and to order the
Commission to pay the costs. The Commission requests the court to dismiss the appeal or, in the
alternative, to dismiss the original application as inadmissible on one of the other grounds advanced by
the Commission at first instance and, finally, to order the appellants to bear the costs. For its part, the
Kingdom of Spain seeks the dismissal of the appeal, affirmation of the correctness of the order at first
instance, and an order for costs against the appellants.

IIITHE ORDER APPEALED AGAINST

9.
In the order appealed against the Court of First Instance examined whether the applicants had
standing to bring the action pending before it. In so doing it distinguished between the applicant
individuals and applicant environmental associations.
10.
(a) As regards the former, it based itself on the settled case law of both the Court of Justice and the
Court of First Instance, according to which
persons other than the addressees may claim that a decision is of direct concern to them only if
that decision affects them by reason of certain attributes which are peculiar to them, or by reason of
factual circumstances which differentiate them from all other persons and thereby distinguish them
individually in the same way as the person addressed.4

4
See the judgments in Plaumann & Co v EEC Commission Case 25/62 [1963] ECR 95, Spijker Kwasten BV v EC
Commission Case 231/82 [1983] ECR 2559, Union Deutsche Lebensmittelwerke GmbH v EC Commission Case 97/85
[1987] ECR 2265, William Cook plc v EC Commission Case C-198/91 [1993] ECR I-2487, Matra SA v EC Commission
Case C-225/91 [1993] ECR I-3203, Socit Anonyme Participation Ouvrire Cie Air France v European Commission
Case T-2/93 [1994] ECR II-323 and Consorzio Gruppo di Azione Locale Murgia Messapica v EC Commission Case T-
465/93 [1994] ECR II-361.

In so doing, it did not follow the interpretative approach to the fourth paragraph of art 173, contended for
on behalf of the applicants to the effect that the hitherto restrictive case law cannot be transposed to
situations in which the legal interests affected by the contested decision are not economic in nature but
arise from the negative environmental impact of the unlawful conduct by the Community institutions. The
Court of First Instance held in that connection that the essential criterion applied by the Court of Justice in
its settled case law, namely
in substance, a combination of circumstances sufficient for the third-party applicant to be able to
claim that he is affected by the contested decision in a manner which differentiates him from all
other personsremains applicable whatever the nature, economic or otherwise, of those of the
applicants interests which are affected. (See [1995] ECR II-2205 (para 50).)
626
Reliance solely on the harm likely to be suffered, generally and in the abstract, by individuals who
cannot be determined in advance in a way which distinguishes them individually in the same way as the
addressee of a decision cannot suffice to confer locus standi (see [1995] ECR II-2205 (para 51)). The
Court of First Instance further considers that that case law constitutes the only correct interpretation of the
fourth paragraph of art 173 of the Treaty. In its order it expressly states that the contested decision must
directly and individually affect the applicant. That view of the matter cannot be affected either by the
finding that there has been a tendency in recent years by the competent national courts to make locus
standi solely dependent on the existence of a sufficient interest on the part of the applicants 5. For the
above reasons, the Court of First Instance held that it had to be examined whether, in the case pending
before it, the contested decision was of individual concern to the applicants. The court held that it was not,
in light of two matters of relevance in its view, enlarged upon in the order appealed against, which read as
follows ([1995] ECR II-2205 (paras 5456):
5
That argument based on comparative law was developed at some length by the applicants.

11.
(54) The applicants are 16 private individuals who rely either on their objective status as local
resident, fisherman or farmer or on their position as persons concerned by the consequences
which the building of two power stations might have on local tourism, on the health of Canary Island
residents and on the environment. They do not, therefore, rely on any attribute substantially distinct
from those of all the people who live or pursue an activity in the areas concerned and so for them
the contested decision, in so far as it grants financial assistance for the construction of two power
stations on Gran Canaria and Tenerife, is a measure whose effects are likely to impinge on,
objectively, generally and in the abstract, various categories of person and in fact any person
residing or staying temporarily in the areas concerned.
12.
(55) The applicants thus cannot be affected by the contested decision other than in the same
manner as any other local resident, fisherman, farmer or tourist who is, or might be in the future, in
the same situation [see the judgment in Spijker Kwasten BV v EC Commission Case 231/82 [1983]
ECR 2559 (para 9) and the order in Associazione Agricoltori della Provincia di Rovigo v European
Commission Case T-117/94 [1995] ECR II-455 (para 25)].
13.
(56) Nor can the fact that the second, fifth and sixth applicants have submitted a complaint to
the Commission constitute a special circumstance distinguishing them individually from all other
persons and thereby giving them locus standi to bring an action under Article 173 of the Treaty. No
specific procedures are provided for whereby individuals may be associated with the adoption,
implementation and monitoring of decisions taken in the field of financial assistance granted by the
ERDF. Merely submitting a complaint and subsequently exchanging correspondence with the
Commission cannot therefore give a complainant locus standi to bring an action under Article 173.
As the Court of Justice has held, although a person who asks an institution, not to take a decision
in respect of him, but to open an inquiry with regard to third parties, may be considered to have an
indirect interest, he is nevertheless not in 627 the precise legal position of the actual or potential
addressee of a measure which may be annulled under Article 173 of the Treaty [see the judgment
in William v EC Commission Case 246/81 [1982] ECR 2277].
14.
(b) As regards locus standi of the applicant associations, the Court of First Instance followed settled
case law according to which an association formed to further the collective interests of a category of
persons does not have locus standi to challenge a measure adopted by a Community institution affecting
the general interests of that category where the members of that association do not have standing to do
so individually6, unless that association had played a role in a procedure which led to the adoption of an
act within the meaning of art 173 of the Treaty, subject to the proviso, certainly, that the particular role is of
such a nature as to justify admissibility of the action brought by the association, even though its members
are not directly and individually affected by the act in question (see the judgments in Kwekerij Gebroeders
Van der Kooy BV v EC Commission Joined cases 67/85, 68/85 and 70/85 [1988] ECR 219 and Comit
International de la Rayonne et des Fibres Synthtiques (CIRFS) v EC Commission Case C-313/90 [1993]
ECR I-1125).
6
The Court of First Instance relied (a) on: the judgments of the Court of Justice in Fdration nationale de la boucherie
en gros et du commerce en gros des viandes v EEC Council Joined cases 1922/62 [1962] ECR 491 and Union
syndicale-Service public europen v EC Council Case 72/74 [1975] ECR 401; the order of the Court of Justice in
Fdration Nationale des Producteurs de Vins de Table et Vins de Pays v EC Commission Case 60/79 [1979] ECR
2429; the judgment in Comit de dveloppement et de promotion du textile et de lhabillement (DEFI) v EC Commission
Case 282/85 [1986] ECR 2469; the order in Union de Federaciones Agrarias de Espaa (UFADE) v EC Council Case
117/86 [1986] ECR 3255 (para 12) and (b) on the judgment of the Court of First Instance in Associazione Tecnico
Economica del Cemento v European Commission Joined cases T-447449/93 [1995] ECR II-1971 (paras 5859).

15.
In that context it is stated in the order appealed against that the applicant associations did not adduce
any special circumstances to demonstrate the individual interest of their members as opposed to that of
any other person residing in the areas of Gran Canaria and Tenerife. The possible effect on the legal
position of the members of the applicant associations cannot, therefore, be any different from that alleged
here by the applicants who are private individuals (see [1995] ECR II-2205 (para 60)). Consequently, in
so far as the applicants in the present case could not be considered to be individually concerned by the
contested decision, nor in logic, according to the court, could the members of the applicant associations.
16.
The Court of First Instance went on to examine whether the existence of contacts between one of
those associations, Greenpeace, and the Commission was sufficient to confer locus standi on
Greenpeace. The reply given to that question was in the negative since it was considered, in the first
place, that
 (62) the Commission did not, prior to the adoption of the contested decision, initiate any
procedure in which Greenpeace participated; nor was Greenpeace in any way the interlocutor of
the Commission with regard to the adoption of the basic Decision C(91)440 and/or of the contested
decision
and, secondly, that
(63) the correspondence which took place between Greenpeace and the Commission and its
subsequent meeting with members of the Commissions staff were for purposes of information only,
since the Commission was under no duty either to consult or to hear the applicants in the context of
the implementation of Decision C(91)440. (See [1995] ECR II-2205 (paras 6263).)
628
17.
On those grounds the Court of First Instance, in its above mentioned order, dismissed the action in its
entirety as inadmissible, on the ground that the applicants were not individually concerned by the
contested decision, without further examination of the remaining pleas of inadmissibility raised by the
Commission.

IVLEGAL FRAMEWORK
18.
As mentioned above, the legal discussion turns on the interpretation of the fourth paragraph of art 173
of the EC Treaty, which reads as follows:
Any natural or legal person may, under the same conditions, institute proceedings against a
decision addressed to that person or against a decision which, although in the form of a regulation
or a decision addressed to another person, is of direct and individual concern to the former.

VARGUMENTS OF THE PARTIES

A The appellants arguments

19.
(a) In their pleadings the appellants begin by setting out the relevant legal context for the solution of
the present case which, in their view, was incorrectly appraised by the Court of First Instance. Thus, they
consider that the lower court, in applying the fourth paragraph of art 173 of the Treaty, ought to have
taken account of the following matters: under Decision C(91)440, which expressly refers in its preamble to
the necessity of carrying out the works which are the subject of financing in compliance with the
requirements of Community law on environmental protection and, in particular, with Directive 85/337,
owing to the likely impact of the works on the environment, the Commission was required, if it observed
that during the construction works Community policy was not being complied with, to withhold payment
of Community funds, to notify the competent national authorities and to determine the measures to be
taken (see the seventh recital in the preamble to Decision C(91)440). Furthermore, Regulation 2052/88,
pursuant to which Decision C(91)440 was adopted, expressly provides that works subject to financing are
to comply with Community provisions on environmental protection.
20.
From the foregoing the appellants infer that, in implementing Decision C(91)440, the Commission
ought, on the one hand, to have systematically monitored whether the construction projects benefiting
from financing were being carried out in compliance with Community law and, in particular, with Directive
85/337 and, on the other, to have refused to disburse further funds if it observed non-compliance with
Community provisions on environmental law. They further consider that, inasmuch as the construction
projects were commenced before an environmental impact assessment had been obtained and the
Commission was aware of that irregularity, the decision by the latter to disburse to Spain further funds
amounting to ECU 12m is in direct breach of its above mentioned obligations.
21.
In that regard the appellants see fit to recall that, in accordance with their interpretation of Directive
85/337, the latter instrument confers a series of rights on people who may be concerned by public and
private projects likely to have significant effects on the environment, including a right of access to
information relating to the project concerned and the right to express an opinion before the project is
initiated (art 6(2)). There are, then, certain rights which are conferred on individuals in connection with
environmental protection on which the persons entitled may directly rely, inasmuch as the relevant
provisions of Directive 85/337 629have been adjudged by the Court of Justice to be of a clear and
unconditional nature, thus producing direct effects in the domestic legal order 7.
7
The appellants cite the judgment in EC Commission v Germany Case C-431/92 [1995] ECR I-2189 (paras 3740).

22.
In summary form, the reasoning on which the appellants claim for judicial protection is based is as
follows. The Commission was under a specific obligation not to disburse payments and to withhold
financing of the relevant works if it found that they were being carried out in contravention of Community
law. In fact, the works were commenced without a prior environmental impact assessment in
contravention of the provisions of Directive 85/337 which, in its turn, confers rights on a certain category
of Community nationals. Consequently, by not withholding the additional financing, even though it had
been informed in due time of the above mentioned infringement, the Commission violated its obligations
under Community law which consist, at least in part, in the defence of the appellants right to protection of
the environment arising in general out of Community law and, in particular, out of the above mentioned
provisions of Directive 85/337.
23.
(b) The appellants contend that the Court of First Instance erred in its interpretation and application of
art 173 of the EC Treaty because it did not apply the proper criteria for assessing whether the applicants
before it were individually affected, within the meaning of art 173, by the contested decision adopted by
the Commission. In particular, the Court of First Instance followed the case law of the Court of Justice
developed solely in the context of economic issues to the effect that an individual must belong to a closed
class in order to be individually concerned by a Community act. For that reason, the appellants assert,
the Court of First Instance did not take account of the specific nature and character of the environmental
legal interest, since indeed the Court of Justice has not hitherto adjudicated on the question of locus
standi of individuals, where they allege that the legal interest affected is that of environmental protection.
24.
The appellants consider that the interpretative approach followed by the Court of First Instance creates
a vacuum in the judicial protection afforded by the Community legal order, whenever it is a question of
reviewing compliance by Community institutions with Community environmental legislation. Since the
interest in environmental protection is common to and is shared by all Community citizens, there cannot,
in the event of environmental harm, be a closed class of persons affected by harm to the environment, at
least not in the sense attributed to the notion of a closed class by the Court of First Instance. It may be
clearly inferred from the order appealed against, the appellants assert, that the application, in cases
concerning the environment, of the traditional case law to the effect that an act must be of individual
concern to the applicant will have the practical effect of never allowing individuals to challenge
Community measures which affect their environmental interests. The legal vacuum becomes more
apparent from the finding that, in cases such as the present dispute, privileged applicants under art 173,
that is to say member states, Council and Commission) are not interested in availing themselves of the
possibility afforded them by the second paragraph of art 173.
25.
Nor, moreover, can it be maintained, in the appellants view, that that vacuum may be filled by the
possibility afforded to individuals not having locus standi of bringing proceedings before a national court in
order to vindicate their rights, having recourse to the legal remedies available under national law. In the
present case, the proceedings brought before the Spanish courts can only concern the legality of national
acts and non-compliance by the Spanish authorities with their 630 obligations under Directive 85/337.
Conversely, the legality of the Commission decision challenged before the court cannot be called in
question in the context of national proceedings, since the Spanish courts do not have jurisdiction to
review the legality under Community law of payment by the Commission of certain funds for the financing
of works8.
8
The appellants cite the judgment of the Court of First Instance in Air France [1994] ECR II-121 (para 69), in which it is
expressly stated that the existence of a remedy before the national courts cannot preclude the possibility of contesting
the legality of a decision adopted by a Community institution directly before the Community judicature under art 173 of
the Treaty.

26.
The appellants go on to stress that the solution adopted by the Court of First Instance in the order
appealed against directly contrasts with that of courts of the member states, and with recent
developments in international law. In that connection the appellants cite a series of comparative studies
from which it may be inferred that the legal systems of the member states have all developed in the
direction of a widening of the procedural right of citizens to seek judicial protection in cases in which their
legal interests in environmental protection are infringed. Indeed, the appellants observe that, if they had
been required to bring proceedings before a court of a member state, actions brought by some or all of
them would have been adjudged admissible.
27.
Moreover, it is contended that the restrictive interpretation adopted by the Court of First Instance in the
order appealed against runs counter to developments in Community and international law in relation to
environmental protection. Reliance is placed on case law of the Court of Justice in which environmental
protection has been proclaimed as one of the Communitys essential objectives 9 and in which EC
environmental legislation has been held to be capable of creating rights and obligations for individuals 10.
Reference is also made to the Fifth Environmental Action Programme approved by the Council and by the
representatives of the member states11, to Principle 10 of the Rio Declaration on Environment and
Development (1992 United Nations Conference on Environment and Development), to Agenda 21, to the
Council of Europe Convention on Civil Liability for Damage Resulting from Activities Dangerous to the
Environment (Lugano, 21 June 1993; ETS No 150), to recent judgments of the European Court of Human
Rights in Strasbourg, and to the system of administrative review introduced by the World Bank as from
1993 in the case of activities which pose a threat to the environment.
9
See the judgments in Procureur de la Rpublique v Association de dfense des brleurs dhuiles usages Case 240/83
[1985] ECR 531 (para 13) and EC Commission v Denmark Case 302/86 [1988] ECR 4607 (para 8).
10
See the judgments in EC Commission v Germany Case C-131/88 [1991] ECR I-825 (para 7) and EC Commission v
Germany Case C-361/88 [1991] ECR I-2567 (paras 1516).
11
Resolution of the Council and the Representatives of the governments of the member states, meeting within the Council
of 1 February 1993, on a Community programme of policy and action in relation to the environment and sustainable
development (OJ 1993 C138 p 1).

28.
 In light of the foregoing, the appellants propose a different interpretation of the fourth paragraph of art
173 of the Treaty. In order to ascertain whether an individual is affected individually by a Community act
when that person pleads damage to the environment caused by an infringement by the Community
institutions of their obligations under Community law, the Community judicature should, according to the
appellants reasoning, require the applicant to demonstrate that:
29.
(a) he/she has personally suffered (or is likely to suffer) some actual or threatened detriment as a
result of the allegedly illegal conduct of the Community 631 institution concerned, such as a violation of
his or her environmental rights or interference with his or her environmental interests; (b) the detriment
can be traced to the act challenged; and (c) the detriment is capable of being redressed by a favourable
judgment.
30.
It is contended on behalf of the appellants that they satisfy these three criteria. As to the first criterion,
they reiterate the evidence relied on before the Court of First Instance in which they described the nature
and extent of the detriment suffered as a result of the Commissions acts. As to the second criterion, they
point out that the above mentioned harm can be traced to the Commissions acts because the national
(Spanish) authorities have no discretion as to the use to which the funds advanced pursuant to Decision
C(91)440 are to be put. In other words, in disbursing the necessary funds to Spain to continue the
construction of projects being carried out in a manner contrary to provisions of Community environmental
law and, in particular, to Directive 85/337, the Commission directly contributed to the detriment suffered
by the appellants as a result of the harm to the environment. As to the third criterion, it is submitted that
the above mentioned detriment would be capable of being redressed if a court judgment were to annul
the Commission decision to continue financing the works. For, if the necessary funding for the
continuation of the projects were not available, it is to be expected, the appellants maintain, that work on
the power stations would be suspended.
31.
Specifically in relation to the appellant environmental associations, it is stated in the notice of appeal
that those associations should be recognised as having standing because one or more of their members
belong to the category of persons who, under the fourth paragraph of art 173, are entitled to bring an
action before the Court of First Instance, and also because they can rely in their own right, in order to
have their standing recognised, on the fact that their primordial objective is the protection of the
environment and that specifically included in their objects, moreover, is protection of the environment in
the geographical area in which the works in question are being carried out. Specifically concerning the
role of the representative organisations or associations and the need to recognise their procedural right of
action before the Community judicature, whenever a Community act affects the interests and objectives
for whose protection and furtherance they were formed, the appellant organisations and associations
refer to the judgment of the Court of First Instance in Associazione Tecnico Economica del Cemento v
European Commission Joined cases T-447449/93 [1995] ECR II-1971 (paras 5362) and to the opinion
of Advocate General Lenz in Confederazione italiana dirigenti di azienda (CIDA) v EC Council Case
297/86 [1988] ECR 353112.
12
See [1988] ECR 3531 (para 15): Nor should it be overlooked that, unlike Cases 282/85 and 117/86, this case is not
concerned with a measure which primarily affects the interests of the members of an association and not the
associations own interests. The question under discussion is the appropriate degree of participation of economic and
social categories in the Economic and Social Committee, which the Council must assess under Article 195 of the EEC
Treaty. In such a situation, nothing could be more natural than to leave the defence of the interests of the various
categories referred to in Article 195 to organized groups, that is to say associations, especially since the individual
members of the group, since they are not individually affected, are generally not entitled to bring an action before the
Court.

32.
In conclusion, the appellants consider that the solution proposed by them is the most apposite, and
refer to the courts judgment in Plaumann & Co v EEC Commission Case 25/62 [1963] ECR 95 from
which they infer that the procedural criteria of art 173 are not to be narrowly construed.
632

B The Commissions arguments

33.
The Commission observes at the outset that, even if the interpretative approach to art 173 contended
for on behalf of the appellants were accepted, the latter do not satisfy the criteria on which they rely. But,
in any event, it considers that the solution proposed by the appellants cannot be adopted because it
would be tantamount to abandoning the express requirement of the fourth paragraph of art 173 that an
applicant must be individually concerned by the contested measure 13. It is further pointed out by the
Commission that the court has never departed from its settled case law to the effect that persons other
than addressees of a decision may claim that a decision is of individual concern to them only if it affects
them by reason of certain attributes peculiar to them or by reason of circumstances differentiating them
from all other persons and, by virtue of that fact, distinguishes them individually in the same way as the
addressee. That definition is also to be found in the grounds of the judgments in Piraiki-Patraiki v EC
Commission Case 11/82 [1985] ECR 207, Extramet Industrie SA v EC Council Case C-358/89 [1991]
ECR I-2501 and Antillean Rice Mills NV v EC Commission Joined cases T-480/93 and T-483/93 [1995]
ECR II-2305, relied on by the appellants in support of the contra legem interpretation which they propose.
In that way, the Commission says, the appellants are substituting for the letter of the law their own
formulation in accordance with which it is sufficient to demonstrate that a Community act personally and
not necessarily individually concerns the applicant for the latter to acquire locus standi before the
Community judicature. In particular, the Commission points out that, if one were ultimately to abandon the
interpretative approach which requires a person, under the fourth paragraph of art 173, to be
differentiated in relation to the contested measure, then that measure would no longer be required to
concern the applicant individually: it would be sufficient for him to be personally concerned. That contra
legem interpretation would result, in the Commissions view, in every inhabitant of the islands of Tenerife
and Gran Canaria being recognised as having a legal interest in challenging the measure at issue
adopted by the Commission14.
13
For that reason, according to the Commissions arguments, the elements of comparative law relied on by the appellants
are not relevant. On the one hand, the international texts on which they base their view have no binding effect. On the
other hand, judicial developments in the legal systems of the member states cannot be transposed, the Commission
maintains, to Community procedural law. At national level there is no procedural requirement such as that imposed by
the fourth paragraph of art 173 of the Treaty that the contested measure must individually concern the applicant.
14
That is to say, the potential applicants would amount to tens, if not hundreds, nay, thousands, of persons. In the
Commissions view, that argument is not affected by the vague position adopted by the appellants in their pleadings to
the effect that it is clear that, in the present case, most of the inhabitants of Tenerife and Gran Canaria would not satisfy
the conditions actually proposed by the appellants.

34.
The Commission goes on to point out that, in its view, the appellants cannot validly argue that the
solution adopted by the Court of First Instance in the order appealed against creates a legal vacuum in
judicial protection and review of conformity by the Community institutions with Community environmental
legislation. In particular, it stresses that the appellants are entitled, in so far as the environmental
dimension of the works subject to financing is concerned, to bring proceedings before the Spanish courts,
and that certain of them had availed themselves of that right. Nevertheless, neither the financing, as such,
of the works nor the role of the Commission in monitoring whether Spain is observing the relevant
Community provisions concerns the appellants. Consequently, they do not have a legal interest in
obtaining a judicial determination in relation to those 633 questions. The absence of a legal interest
cannot be remedied by the elements of comparative law relied on by the appellants in their pleadings.
35.
As to the rights which the appellants maintain may be inferred from Directive 85/337, the Commission
makes the following observations: even though that directive, as regards certain at least of its provisions,
may produce direct effects, reliance on it cannot avail the appellants in the context of the present
proceedings. Even if it is acknowledged that the appellants, or certain of them, derive certain individual
rights at the level of national law from the directive, those rights do not go so far as to afford them the
possibility of contesting the legality of an act, such as Decision C(91)440, before the Court of First
Instance or to differentiate them, for the purposes of the fourth paragraph of art 173, in relation to the
contested act.
36.
The Commission also rebuts the submission contained in the appeal to the effect that the appellant
environmental associations have locus standi because, on the one hand, one or more of their members
are individually affected by the contested measure and, on the other, as representative organisations
whose object is the protection of the environment in the Canary Islands, they automatically have a legal
interest in matters concerning the environment15.
15
The Commission points out that that criterion is unsupported by authority, save for the opinion of Advocate General Lenz
in CIDA [1988] ECR 3531, which was not followed by the court.

37.
Furthermore, the Commission observes that the notice of appeal does not repeat the submission put
forward at first instance to the effect that two at least of the appellants have standing by virtue of the fact
that they had lodged complaints with the Commission concerning the conditions under which the relevant
construction works were being carried out. It infers from this that the appellants have tacitly abandoned
that submission. Finally, the Commission requests the court, should it decide to set aside the order of the
Court of First Instance, to adjudicate on the two other objections of inadmissibility raised at first instance.
For it has come to the conclusion that the legal redress sought by the appellants before the Court of First
Instance could be granted only in the event of an amendment of the fourth paragraph of art 173 of the
Treaty.
38.
At the hearing, the Commission further pointed out that construction of one of the electricity power
generating plants had already been started in 1990, that is to say before the issue of Community
financing of the works arose. The appellants ought therefore to have taken steps, as at that moment in
time, to oppose the carrying out of the works or, at least, the financing of them, by challenging the original
Commission Decision C(91)440. However, they failed to show due diligence and, for that reason, brought
proceedings at a later date against the decision to continue the financing. The Commissions
representative also stated that the decision to continue financing of the works was taken after there had
been found to be no infringement of the provisions of Directive 85/337. Thus, the Commission carried out
the review which it had bound itself to do under the original Decision C(91)440.
39.
It further pointed out that the Community legal order also affords legal protection to persons which
goes beyond the strict confines of the fourth paragraph of art 173 of the Treaty. It recalled in that
connection recent legislative developments concerning access by citizens to the archives of the
Community institutions, the right of petition to the European Parliament and the possibility of bringing a
matter before the Ombudsman. Conversely, no easing of the procedural restrictions laid down in the
fourth paragraph of art 173 can be expected, nor has there been any such proposal in the context of the
Intergovernmental Conference.
634

C The Spanish governments arguments

40.
For its part, Spain proposes that the appeal be dismissed on the basis of the arguments set out below.
It begins by rehearsing the case law of the court on the issue of locus standi of individuals, from which it
infers that, in order for individuals to bring an admissible claim before the Community judicature, they
must be individually concerned by the act which they are contesting, in the sense that it affects those
persons by reason of certain attributes peculiar to them or by reason of circumstances in which they are
differentiated from all other persons and, by virtue of those factors, distinguishes them individually in the
same way as the addressee of the decision. Consequently, Spain maintains, the only Community acts
open to challenge are those concerning a closed class of persons which is ascertained or ascertainable at
the time of adoption of the act and whose rights the author of the act has sought to regulate. For that
purpose, it is not enough that it may be possible to ascertain with a greater or less degree of accuracy the
number or even the identity of the persons to whom a measure applies, if the first of these three elements
is not also present, that is to say if there are no circumstances which differentiate and distinguish them in
the same way as the person addressed16. Thus, the court confers the right to challenge a Community act
only on the addressees thereof and on persons who may be equated with them.
16
Reference is made to the order of the court in Van Parijs v EC Council Case C-257/93 [1993] ECR I-3335 and to the
order of the Court of First Instance in Fdration Rgionale des Syndicats dExploitants Agricoles (FRSEA) v EC
Council Case T-476/93 [1993] ECR II-1187, and the judgment of the court in Buralux SA v EU Council Case C-209/94 P
[1996] ECR I-615.

41.
As a logical consequence, the appellants, who were not the addressees of the act which they are
contesting and cannot be equated with the addressee of that act, namely Spain, do not have locus standi
and the Court of First Instance was right to dismiss their action. For that reason, the appellant
environmental associations do not have locus standi. Also, it would be contrary to the principle of legality
for the actions of the environmental associations to be adjudged to be admissible and at the same time
for the actions brought by natural persons to be dismissed, that is to say for a wider legal interest to be
conferred on representative associations than on individuals. As to the latter, the Kingdom of Spain
observes that they also lack locus standi to challenge the decision of the Commission to continue the
financing. For one thing, they do not invoke circumstances or attributes peculiar to them which distinguish
them individually in the same manner as the addressee of the contested decision 17.
17
In accordance with the reasoning adopted by the Kingdom of Spain, if the interpretative approach advocated by the
appellants were accepted, then a legal interest would have to be recognised in the case of any person invoking the
impact of an interference with the environment in connection with fishing, farming, public health, tourism, quality of life,
childrens education, damage to the interests of taxi-drivers, local workers, tourists, windsurfers, and those concerned
with the protection of bird-life. Consequently, the appellants are affected by the contested act just as much as any other
individual carrying on any kind of activity in the area.

42.
Consequently, they do not belong to any closed class of persons, ascertained or ascertainable at the
time of adoption of the act, nor are they in a position analogous to that of the addressee thereof. Again, as
regards those appellants who had submitted complaints to the Commission, that fact does not suffice to
distinguish them, because the Commission is not under any obligation to initiate proceedings under art
169 of the Treaty, whenever it receives a complaint in that connection (see the judgment in Star Fruit
Company SA v EC Commission Case 247/87 [1989] ECR 291). For all the above mentioned reasons, the
Kingdom of Spain agrees 635 with the solution adopted by the Court of First Instance in the order
appealed against.
43.
The Kingdom of Spain goes on to counter the appellants argument concerning the creation of a legal
vacuum in judicial protection caused by an inability to review infringements by the Commission in matters
of environmental protection. It stresses that, in keeping with the system established by the Treaties,
individuals enjoy the legal protection ensured by national courts which are the ordinary courts under the
Community judicial system. Accordingly, only the
Member states and Community institutions have a legal interest in seeking the annulment of measures
of a general nature, as is clearly indicated by art 173 of the Treaty and laid down by the relevant case law
of the court. Moreover, in light of the characteristics of the present case, the Kingdom of Spain considers
that, in actual fact, the appellants are attempting to fabricate a dispute before the Community judicature.
Under the guise of a challenge to the legality of a Community act, they are in fact seeking to prompt the
Community judicature to adjudicate on the legality of an act of a national authority and, indirectly, to annul
it. Specifically, the appellants rely on no financial interest in connection with the mode of operation of the
ERDF but, in a general and undefined manner, on their interest in environmental protection which can be
affected only by the act of a national authority.
44.
Thus, even if it were accepted that the competent Spanish authorities infringed the provisions of
Community law on environmental protection which, in so far as they have been transposed into the
national legal order, constitute provisions of national law, there would then be an infringement of a
national rule of law by a national public authority, in which case the dispute arising would fall within the
exclusive competence of the national courts which, in a proper case, would be able to request the Court
of Justice for a preliminary ruling on the interpretation of the relevant provisions of Community law 18.
Finally, Spain submits, since the Commission decision at issue merely gives effect to Decision C(91)440,
it neither affects nor can affect the environment or, in any event, the rights which the appellants might
derive from the protection thereof. For that reason, Spain submits, there can be no question of an
infringement of the rights of the citizen deriving from the protection of the environment.
18
Spain refers to the judgment in Oleificio Borelli SpA v EC Commission Case C-97/91 [1992] ECR I-6313 in which it was
held that the court is not competent to decide on the legality of an act adopted by a national authority even though that
act was adopted in the context of a procedure for the adoption of a Community decision.

45.
Spain goes on to stress that, in its view, apart from having no individual interest, the appellants also
have no direct interest in challenging the Commission decision to continue the financing. It points out in
that connection that Decision C(91)440, and acts adopted in implementation thereof, are of a financial
nature and relate only indirectly to the environment, in the same way as any area of human activity.
46.
Finally, the representative of Spain maintained at the hearing before the court that the change in case
law advocated by the appellants could not be regarded as expedient since it would result in the
enshrinement of an actio popularis for environmental questions. Protection of the environment, moreover,
was but one legal concern among so many others within a legal system and there was no justification
whatever for according it any different procedural treatment.
636

VIMY VIEWS ON THE APPEAL

47.
I consider it important to stress, at the outset, the novelty of the present case. For the first time, the
court is being asked to form a view on a series of legal issues which were not dealt with in the context of
the An TaisceThe National Trust for Ireland v European Commission Case C-325/94 P [1996] ECR I-
3727 and Rovigo [1995] ECR II-455 cases19, whose subject matter concerned environmental protection.
The issue before the court in this case concerns the interpretation of the fourth paragraph of art 173 of the
Treaty. The legal debate turns on the precondition laid down in the above mentioned provision to the
effect that, in order for a natural or legal person to be able to challenge decisions of Community
institutions addressed to another person, those decisions must be of direct and individual concern to the
former. That wording has imposed a significant procedural limitation on the right of action before the
Community judicature and is appraised, moreover, of the courts own motion 20. I consider that, in order
that the dispute pending before the court may be fully ventilated, the court will need to give replies to the
questions set out below:
19
In the first of these two cases the court gave judgment on an appeal brought by two environmental associations against
the judgment by the Court of First Instance dismissing as inadmissible their action for annulment of a Commission
decision. In that decision the Commission refused to suspend or revoke the grant of appropriations from the Structural
Fund for the construction of a nature observatory in Ireland. The Court of First Instance based its order in that case on
the ground that it had not been proven that the Commission had decided not to make use of its power under the
relevant Community provisions in relation to the suspension or reduction of Community funding for the construction of
the above-mentioned centre, a power which, according to the Court of First Instance, is always available. That view of
the matter was upheld by the Court of Justice, as is apparent from paras 30 and 31 of its judgment. In the second of the
two cases the appellants (local occupational associations and individuals) had appealed against an order of the Court of
First Instance dismissing as inadmissible their action for annulment of a Commission decision approving funding for the
protection of the residents and natural environment of the Padua delta zone in Italy under Council Regulation (EEC)
1973/92. The Court of Justice upheld the view taken by the Court of First Instance that the applicants were not affected
individually by the contested act because, contrary to their assertions, Community law did not confer on them a right to
participate in the procedure for structuring the programme of funding actions. It may be observed, then, that in those two
cases the Court of First Instance in whole or in part founded its view as to the inadmissibility of the actions on grounds
other than those adopted in the present case.
20
Pursuant to art 92(2) of the Rules of Procedure of the Court. See the order in von Bonkewitz-Lindner v European
Parliament Case 13/86 [1987] ECR 1417, and judgment in Neotype Techmaschexport GmbH v EC Commission Joined
cases C-305/86 and C-160/87 [1990] ECR I-2945 (para 18).

48.
First, can the contested Commission decision concern the appellants in the sense that they are able
to derive a judicially protected right or legal interest from the interest, within the Community legal order, in
conservation of the environment? Secondly, if the first question is answered in the affirmative, was the
Court of First Instance right to hold that that decision did not individually concern the persons challenging
it? Thirdly, if the second question is answered in the negative, was there any other obstacle to
admissibility of the originating application?

A As to the possibility of a Community national deriving a judicially protected right or legal

interest from the interest, within the Community legal order, in conservation of the
environment
49.
The fourth paragraph of art 173 of the Treaty lays down the condition that a natural or legal person
bringing an action against a decision of a Community institution must be the holder of some right or legal
interest which is affected by the contested act. That condition, which does not apply to the privileged
parties mentioned in the second paragraph of that article, generally consists in the express requirement
that the applicant be directly and individually affected by the contested 637 act. Where the contested act
exhibits those characteristics, it is rare that that condition of a legal interest will not be satisfied 21. In any
event, the existence of a legal interest is one of the elements of admissibility of the action examined by
the Community judicature22.
21
Nevertheless, it is not improbable: see judgment in International Business Machines Corp v EC Commission Case 60/81
[1981] ECR 2639.
22
As examples of cases in which the court has accepted the existence of a legal interest, see judgments in Benzine en
Petroleum Handelsmaatschappij BV v EC Commission Case 77/77 [1978] ECR 1513; Asociacin Profesional de
Empresarios de Pesca Comunitarios (Apesco) v EC Commission Case 207/86 [1988] ECR 2151; and order of the court
in Lezzi Pietro & C Srl v EC Commission Case C-123/92 [1993] ECR I-809.

50.
I believe that that issue must be examined at the outset, precisely on account of the particular features
of the present case. Moreover, the Community judicature has not hitherto adjudicated upon the existence
of an environmental legal interest. For that reason, it is useful to clarify the position occupied by the issue
of environmental protection in the Community legal order and to analyse the courts case law on that
question.
(a) Public dimension of environmental protection
51.
In the first place, there can be no doubt that ensuring the quality of the environment is directly linked to
what might be termed the Community public interest. The Community judicature has proclaimed in its
judgments that environmental protection is one of the fundamental objectives of the Community 23.
Support for that view is, moreover, to be found in the letter of the Treaty. Following the revision of 7
February 1992, art 2 of the Treaty states that the Community shall have as its task to promote
growth respecting the environment. For those purposes the activity of the Community is to comprise,
pursuant to art 3(k) of the Treaty, a policy in the sphere of the environment. The general outlines of that
policy are elaborated in Title XVI of the Treaty. That policy is to contribute, under art 130r of the Treaty,
inter alia, to preserving, protecting and improving the quality of the environment, protecting human
health, and prudent and rational utilisation of natural resources. Paragraph 2 of art 130r further provides
that environmental protection requirements must be integrated into the definition and implementation of
other Community policies24. Secondary Community legislation has been enacted in order to give effect to
the objectives of art 130r of the Treaty25; it is the task of the Community judicature to ensure the correct
application of that legislation.
23
See eg judgments in Procureur de la Rpublique v Association de dfense des brleurs dhuiles usages Case 240/83
[1985] ECR 531, EC Commission v Denmark Case 302/86 [1988] ECR 4607 and EC Commission v Germany Case C-
195/90 [1992] ECR I-3141 (para 29) and my opinion in Diego Cal & Figli Srl v Servici Ecologici Porto di Genova Spa
(SEPG) Case C-343/95 [1997] ECR I-1547 (paras 5564).
24
Eg in EC Commission v Belgium Case C-2/90 [1992] ECR I-4431 (para 22ff), the court seeks to reconcile the
fundamental Community freedom of movement of goods enshrined in art 30 of the Treaty with the requirement of
environmental protection, as described in art 130r of the Treaty. Thus, although I consider that waste products may be
classified as goods within the meaning of art 30 of the Treaty, in the final analysis, the imposition of restrictions on their
circulation is accepted on the following grounds ([1992] ECR I-4431 (para 34)): The principle that environmental
damage should as a matter of priority be remedied at source, laid down by Article 130r(2) of the Treaty as a basis for
action by the Community relating to the environment, entails that it is for each region, municipality or other local
authority to take appropriate steps to ensure that its own waste is collected, treated and disposed of; it must accordingly
be disposed of as close as possible to the place where it is produced, in order to limit as far as possible the transport of
waste. See also the judgment in EC Commission v EC Council Case C-155/91 [1993] ECR I-939.
25
By means of secondary Community legislation a number of significant obligations have been imposed on the member
states requiring them to ensure protection of the environment. The number of directives concerning the safeguarding of
the environment is particularly considerable. An act or omission on the part of national authorities affecting the
environment may cumulatively infringe not one but several directives. For example, following complaints by individuals
against Greece concerning water quality in Lake Veyoritidas, the River Soulos and Payasitikos Bay, the Commission
brought an action against Greece under art 169 of the Treaty for the alleged infringement of Council Directive (EEC)
76/464 on pollution caused by certain dangerous substances discharged into the aquatic environment of the Community
(OJ 1976 L129 p 23). Nevertheless, in its allegations in the reasoned opinion it also referred to Council Directive (EEC)
76/160 concerning the quality of bathing water (OJ 1975 L31 p 1), as regards Payasitikos Bay, and to the following
directives, as regards the particular situation of Lake Veyoritidas and the River Soulos: Council Directive (EEC) 75/440
concerning the quality required of surface water intended for the abstraction of drinking water in the member states (OJ
1975 L194 p 26), Council Directive (EEC) 79/869 concerning the methods of measurement and frequencies of sampling
and analysis of surface water intended for the abstraction of drinking water in the member states (OJ 1979 L271 p 44),
Council Directive (EEC) 80/778 relating to the quality of water intended for human consumption (OJ 1980 L229 p 1),
Council Directive (EEC) 78/659 on the quality of fresh water needing protection or improvement in order to support fish
life (OJ 1978 L222 p 1), Council Directive (EEC) 75/442 on waste (OJ 1975 L194 p 39) and Council Directive (EEC)
78/319 on toxic and dangerous waste (OJ 1978 L84 p 43). On that case, see the opinion of Advocate General Tesauro
in European Commission v Greece Joined cases C-232233/95 (1997) Transcript, 11 June, ECJ.

638
52.
From the foregoing illustration it may be inferred that the environment has a public dimension which
constitutes a general Community interest whose protection is the responsibility of the member states and
the Community institutions, in accordance with the equilibrium established by the Treaty 26. Thus,
protection of the environment, as enshrined in primary Community law and as a matter of principle
affirmed by the case law of the court, constitutes a fundamental obligation on the part of the public
authoritie27if Community institutions and national authorities may be subsumed under one and the same
term. And it is in the implementation of that obligation that obligations on the part of individuals may
arise.28
26
An interesting example of the manner in which environmental protection may be secured as a result of a procedure
between Community institutions and member states may be found in the judgment in EC Commission v Germany Case
C-431/92 [1995] ECR I-2189. In that case the Commission had brought an action under art 169 of the Treaty for a
declaration that Germany, by granting authorisation for the construction of a new section of the Grosskrotzenburg power
 station without a prior environmental impact assessment, had failed to fulfil its obligations under arts 5 and 189 of the
Treaty in conjunction with Directive 85/337. It is worth noting that the Commission did not proceed against Germany on
the ground of an incorrect transposition of the directive in general but confined itself to the issue of the construction of
the thermal power station. The court acknowledged that the Commission may bring an action against a member state
under art 169 on the sole basis of an actual interference with the environment contrary to secondary Community law.
Indeed it considers it to be part of the Commissions task in the general interest of the Community, to ensure that the
member states give effect to the Treaty and the provisions adopted by the institutions thereunder and to obtain a
declaration of any failure to fulfil the obligations deriving therefrom. The court also clearly distinguished between the
need for environmental protection in the public interest and the existence or otherwise of rights conferred by Community
law on individuals in connection with such protection. It stated ([1995] ECR I-2189 (para 26)): In its application, the
Commission complains that Germany has not observed, in a specific case, the obligation flowing directly from the
directive to assess the environmental impact of the project concerned. The question which arises is thus whether the
directive is to be construed as imposing that obligation. That question is quite separate from the question whether
individuals may rely as against the state on provisions of an unimplemented directive which are unconditional and
sufficiently clear and precise, a right which has been recognized by the Court of Justice.
27
On this point Community law is analogous to the national constitutional systems of Germany, Greece and the
Netherlands. In those countries protection of the environment is expressly recognised as a fundamental obligation of the
state, not necessarily corresponding to a general individual right of the citizen to require observance of that obligation.
28
That does not, of course, mean that horizontal direct effect is conferred on directives with an environmental content (see
footnote 37, below). See eg the judgment in Criminal proceedings against Peralta Case C-379/92 [1994] ECR I-3453
(para 59) on prohibition of dumping into the sea of noxious chemical substances and the judgment in Criminal
proceedings against Tombesi Joined cases C-304/94, C-330/94, C-342/94 and C-224/95 [1997] All ER (EC) 639,
[1997] ECR I-3561 concerning the application of Directives 41/156 and 91/689 on waste.

53.
However, the fact that legality must be observed per se within the Community, including the obligation
to protect the environment, does not automatically confer on a natural or legal person a right or legal
interest enforceable by an action under the fourth paragraph of art 173 of the Treaty. The Community
legal order does not recognise an actio popularis29 in environmental matters either. It is not, therefore,
possible to rely, as the sole ground for claiming locus standi, on the legal vacuum which would be likely to
be created by the fact that certain infringements by the Commission cannot with certainty be remedied if
the task of submitting them for judicial review is entrusted exclusively to the member states and the
Community institutions, which do not in practice have an interest in that regard.
29
Contrary to the situation prevailing, at least to some extent, in Spain and Portugal. The reason why the procedural
criteria are so generous in those countries must be sought in the fact that the Constitutions in those states expressly
enshrine a general constitutional right to the environment on the part of citizens (art 45 of the Spanish Constitution and
art 46 of the Portuguese Constitution).

(b) Recognition by secondary Community law of rights in favour of individuals

54.
However, the question of environmental protection in the Community legal order is not exhaustively
treated by pointers to the existence of a general obligation imposed on Community and national
authorities. Secondary Community law which, directly or indirectly, gives effect to the Communitys
environmental objectives, has created in favour of individuals certain specific rights of real substance,
which are guaranteed by the national and Community judicial authorities 30. Those rights, as a general
rule, flow from the direct effect of directives 639 whose subject matter is concerned with environmental
protection. For example, the court held in EC Commission v Germany Case C-131/88 [1991] ECR I-825
(para 7) that Council Directive (EEC) 80/68 on the protection of groundwater against pollution caused by
certain dangerous substances (OJ 1979 L20 p 43)
30
Certainly, in the context of environmental protection, individuals may avail themselves of the other rights conferred on
them by the Community legal order. Following adoption by the Commission of Commission Decision (ECSC, EC,
Euratom) 94/90 on public access to Commission documents, certain environmental associations, the same ones which
opposed the building of a nature observatory in Ireland, requested access to all the Commission documents concerning
examination of the construction project in point and the question whether structural funds could be used for that project
(see footnote 19, above). The Commission refused access, relying on protection of the public interest and its interest in
the confidentiality of its deliberations. WWF (UK) brought an action against that refusal before the Court of First Instance
which annulled the refusal on the ground that it did not satisfy the requirements as to reasoning laid down in art 190 of
the Treaty, having first found that Decision 94/90 constituted an act capable of conferring on third parties rights which
the Commission was obliged to observe (see WWF UK v European Commission Case T-105/95 [1997] All ER (EC)
300, [1997] ECR II-313 (para 55)).

seeks to protect the Communitys groundwater in an effective manner by laying down specific
and detailed provisions requiring the Member States to adopt a series of prohibitions, authorization
schemes and monitoring procedures in order to prevent or limit discharges of certain substances.
The purpose of those provisions of the directive is thus to create rights and obligations for
individuals.
Advocate General Van Gerven observed in his opinion in that case that the
clear and precise implementation of the directives provisions may also be important for third
parties (for instance environmental groups or neighbourhood residents) seeking to have the
prohibitions and restrictions contained in the directive enforced as against the authorities or other
individuals. (See [1991] ECR I-825 (para 7).)
55.
The tendency of the case law in favour of conferral on individuals of rights in connection with
environmental protection or, at least, of the possibility of relying before national courts on the obligations
imposed by secondary Community law on national authorities, is also confirmed by the recent judgment of
the court in its judgment in Associazione Italiana per il World Wildlife Fund v Regione Veneto Case C-
118/94 [1996] ECR I-1223 (see also the judgment in R v Secretary of State for the Environment, ex p
Royal Society for the Protection of Birds Case C-44/95 [1996] ECR I-3805). That case concerned a
preliminary question which arose in the context of an action brought before the Italian courts by certain
associations for nature 640 protection, against a national decision determining the hunting calendar.
Those associations maintained, inter alia, that there had been a breach of the principles contained in
Council Directive (EEC) 79/409 on the conservation of wild birds (OJ 1979 L103 p 1). The question raised
concerned the conditions under which art 9 of the directive permits member states to derogate from the
general prohibition on hunting protected species. In the final analysis, the court conferred direct effect on
that article, relying on its settled case law according to which
where by means of a directive the Community authorities have placed member states under a
duty to adopt a certain course of action, the effectiveness of such a measure would be diminished if
persons were prevented from relying on it in proceedings before the courts, and national courts
were prevented from taking it into consideration as an element of Community law. (See the
judgment in Associazione Italiana per il World Wildlife Fund v Regione Veneto Case C-118/94
[1996] ECR I-1223 (para 19).)31

31
It is worth noting that the court has consistently taken the view the judgments in EC Commission v Netherlands Case
236/85 [1987] ECR 3989, EC Commission v Belgium Case 247/85 [1987] ECR 3029, EC Commission v France Case
252/85 [1988] ECR 2243, EC Commission v Germany Case C-288/88 [1990] ECR I-2721, EC Commission v Spain
Case C-355/90 [1993] ECR I-4221 and Association pour la Protection des Animaux Sauvages v Prfet de Maine et
Loire Case C-435/92 [1994] ECR I-67) that Community legislation on the protection of wild birds seeks to preserve the
common heritage, the management of which is entrusted to the member states. In that sense the directives in question
serve to protect the environment which exists as a Community public interest, in accordance with the reasoning
elaborated hitherto, and do not appear to have in direct contemplation the securing of a right in favour of individuals.
Moreover, notwithstanding the fact that the relevant directives have been in issue before the court, the court has not
adjudicated on the question whether they produce direct effect in the domestic legal orders.

56.
I regard that judgment as constituting a typical example of the manner in which a general objective,
such as environmental protection, when implemented by means of a series of provisions of secondary
Community legislation, may result in securing for individuals the possibility of seeking judicial assistance
in conserving the environment, even where such a right is not provided for directly and expressly by the
Community legislature32.
32
Also of interest is the judgment of the court in EC Commission v Germany Case C-361/88 [1991] ECR I-2567
concerning the mandatory effect of Council Directive (EEC) 80/779 on air quality limit values for sulphur dioxide and
suspended particulates. The court appears to link the existence of a public obligation on the part of the public authorities
with regard to protection of human health with a corresponding right on the part of individuals. At para 16 of the
judgment, it is stated as follows: In that respect, it should be pointed out that the obligation imposed on the member
states to prescribe limit values not to be exceeded within specified periods and in specified circumstances, laid down in
Article 2 of the Directive, is imposed in order to protect human health in particular. It implies, therefore, that whenever
the exceeding of the limit values could endanger human health, the persons concerned must be in a position to rely on
mandatory rules in order to be able to assert their rights It is worth observing that it is human health which is taken to
be the ground justifying the conferral of rights on individuals. Nevertheless, I consider that in that case public health
constitutes a special objective directly related to environmental protection. Moreover, the link between environment and
health is highlighted by the formulation of art 130r of the Treaty.

57.
That reasoning is also followed by the appellants when they maintain that Directive 85/337 created
certain rights in their favour which are affected by the Commission decision at issue. It should be noted
that that directive constitutes one of the most significant achievements by the Community legislature in
favour of environmental protection. Its objective is, as stated in the sixth recital in the preamble, to
establish the principle of prior assessment of the significant effects which certain interferences with the
environment are likely to have. Under art 3 of the directive member states are obliged to assess the direct
and indirect impact of a project on human beings, fauna and flora, soil, water, air, climate, landscape,
641material assets and the cultural heritage. The appellants consider that arts 2, 3, 6 and 8 of that
directive enshrine certain rights in favour of a group of individuals who are described in art 6 of the
directive under the term public concerned.
58.
Indeed, the court has upheld the direct effect of arts 2, 3 and 8 of Directive 85/337 (see the judgment
in EC Commission v Germany Case C-431/92 [1995] ECR I-2189 (para 39)). As regards the mandatory
effect of art 6(2) of the directive, Advocate General Elmer, in his opinion in Aannemersbedrijf PK
Kraaijeveld BV v Gedeputeerde Staten van Zuid-Holland Case C-72/95 [1995] All ER (EC) 134, [1996]
ECR I-5431 (para 70), observed as follows:
According to art 6(2) of the directive, it is for the member states to ensure that any request for
development consent and any information gathered are made available to the public and that the
public concerned is given an opportunity to express an opinion before the project is initiated. The
directive thus requires the member states to introduce a consultation procedure to give individuals
a right to express their opinion. Where a member states implementation of the directive is such
that projects which are likely to have significant effects on the environment are not made the
subject of an environmental impact assessment, the citizen is prevented from exercising his right to
be heard. The member states own negligent implementation of the directive thus deprives the
citizen of a right under the directive. I doubt that Kraaijeveld forms part of the public concerned,
since the project, it is stated, has serious consequences for the business operated by Kraaijeveld.
In those circumstances my view is that arts 2(1) and 4(2) of the directive, in conjunction with art
6(2), confer rights on individuals.33
 33
However, the court did not expressly form a view on that issue because, on the basis of the factual circumstances of the
case pending before the court making the reference, the parties had not raised a claim concerning their rights under
Directive 85/337. On that ground, the court confined itself to replying that art 2(1) of Directive 85/337 has direct effect
and that a national court which, under national law, has the power to apply of its own motion rules of law not relied on
must apply the above-mentioned provision, even if it has not been invoked by the party concerned.

59.
I believe that, if the appellants are to be regarded as belonging to the public which may be concerned
by the projected construction of the two power stations in the Canary Islands 34, then they are entitled to
the rights conferred by Directive 85/ 642337. Those rights are the following: first, the requirement that the
projected interference with the environment be submitted for an appraisal of its impact before
authorisation is granted; secondly, projects for works impacting on the environment, namely those
mentioned in Annex I to the directive or, under certain conditions, in Annex II thereto, must be drawn up in
accordance with the procedure described in arts 5 to 10 of the directive. In particular, that procedure gives
the right, first, for information gathered to be made available to the public, secondly, to provide the public
with the possibility of expressing its opinion before the project is started and, thirdly, to require that opinion
to be taken into account in the context of the procedure for granting authorisation. Accordingly, if the
Commissions contested decision, under which it was decided to continue financing the works in Spain at
issue in these proceedings, infringes those rights, then that decision did indeed concern the appellants
within the meaning of the fourth paragraph of art 173 of the Treaty and the appellants were entitled, as a
matter of principle, to challenge it35.
34
It should be noted that Directive 85/337 does not itself determine the public to which the relevant provisions refer, which
is not without its relevance. See para 96, below.
35
Of course, the fact that the Commission decision concerns the appellants does not necessarily mean that it also affects
them individually (see para 66, below).

60.
It follows from the foregoing that secondary Community law creates special attendant rights to
environmental protection which may naturally be safeguarded by recourse to the courts. However, if the
rights to be vindicated, or the legal interests of the citizen in connection with the environment, go no
further than this, then the protection provided by the Community legal order remains incomplete and
fragmentary. First, judicially protected rights or legal interests exist only where rules of secondary
Community law having direct effect in the domestic legal order have been adopted 36, and cannot therefore
arise when they involve the imposition of an obligation on an individual 37. Secondly, reliance as against
Community institutions on rights arising as described above, in annulment proceedings against acts of
those institutions before the Court of First Instance, is in practice fraught with particularly severe
obstacles38.
36
It should also be emphasised that it is not always the rule that rights in favour of individuals may be derived from
directives concerning environmental protection. It is instructive to note the position taken by the court in its judgment in
Comitato di Coordinamento per la Difesa della Cava v Regione Lombardia Case C-236/92 [1994] I-483. In that case the
question arose as to whether rights are conferred on individuals by art 4 of Council Directive (EEC) 75/442 on waste (OJ
1975 L194 p 39) which is in the following terms: Member States shall take the necessary measures to ensure that
waste is disposed of without endangering human health and without harming the environment and, in particular, without
risk to water, air, soil and plants and animals, without causing a nuisance through noise or odours, and without
adversely affecting the countryside or places of special interest. Certain individuals had relied on the provision in
question before the national courts, seeking annulment of the decision of a national authority involved in the application
of that provision, on the ground that the domestic rules did not provide for the measures necessary to encourage
processing and recycling of waste. The court considered in the end that art 4 of the directive indicated a programme to
be followed and set out the objectives to be followed by the member states in the performance of the specific obligations
imposed on them by other articles of the directive. Thus, the provision at issue must be regarded as defining the
framework for the action to be taken by the member states regarding the treatment of waste and not as requiring, in
itself, the adoption of specific measures or a particular method of waste disposal. It is therefore neither unconditional nor
sufficiently precise and thus is not capable of conferring rights on which individuals may rely as against the state (see
[1994] ECR I-483 (para 14)). See also the judgment in Ministre public v Traen Joined cases 372374/85 [1987] ECR
2141.
37
A characteristic example is afforded by the judgment of the Court of Justice in Criminal proceedings against Arcaro
Case C-168/95 [1997] All ER 82, [1996] ECR I-4705 concerning Council Directive (EEC) 76/464 on pollution caused by
certain dangerous substances discharged into the aquatic environment of the Community. After reiterating its settled
position on the horizontal and non-horizontal effect of directives, the court decided that provisions of directives which
require a person discharging cadmium to seek and obtain authorisation for that purpose cannot in and by themselves,
and without the national legislation applicable in connection with the application of those provisions being taken into
account, justify an aggravation or diminution of the dischargers criminal liability for infringement of those provisions.
38
The principal obstacle is the fact that the infringement of rights enshrined in a directive as a rule stems from a decision
made or action taken by a national body which, even if it relates to an act of a Community body, cannot be the subject of
judicial review by the Community judicature. See judgment in Borelli [1992] ECR I-6313. See also footnote 63, below.

(c) Existence of a specific obligation on the Commission to monitor whether the works financed
comply with Community environmental legislation
61.
In addition to rights which Community law may expressly confer on individuals as against (primarily)
national authorities, I consider that in certain specific cases Community rules relating to environmental
protection directly impose on Community bodies specific and clear obligations whose observance may be
judicially enforced by individuals affected. The Commission has such an obligation under primary and
secondary legislation whenever it proposes to finance works likely to have an impact on the environment,
which is the case here. In the case of the works at issue, moreover, and as was rightly pointed out by the
appellants, the Commission acknowledged the existence of those obligations in Decision C(91)440 and,
moreover, bound itself to observe them. Let us now examine this issue in greater detail.
 643
62.
I will begin by observing that the Treaty provisions concerning the environment are not mere
proclamations of principle. Paragraph 1 of art 130r may refer in general terms to the pursuit of objectives
to which Community policy must contribute in the sector of the environment, and para 2 may trace the
general principles of that policy, but the last sentence of the first sub-paragraph of para 2 of art 130r of the
Treaty appears to impose on the Community institutions a specific and clear obligation which could be
deemed to produce direct effect in the Community legal order. It expressly states that: Environmental
protection requirements must be integrated into the definition and implementation of other Community
policies.
63.
It should also be pointed out that that obligation has not remained a dead letter but has been
imported into secondary Community law, and indeed in the context of legislation on the financing by the
Community of certain actions having an impact on the environment. As referred to above, art 7 of
Regulation 2052/88 provides: Measures financed by the Funds shall be in keeping with Community
policies, including those concerning environmental protection. Thus, in so far as the financing of the
works for the construction of the two power stations in the Canary Islands is concerned, the Commission
was obliged to take into consideration the aspect of environmental protection and it was right to bind itself
to do so in Decision C(91)44039.
39
Notwithstanding the assertions of the Spanish government, I consider that the Commission did indeed bind itself in that
way. It is stated in the preamble to Decision C(91)440: Whereas given the characteristics of this investment and its
impacts it is compulsory to comply with Community law in this matter and, above all, with Directive 85/337. Article 5 of
the decision goes on to add that Lack of compliance with any of the conditions mentioned in this Decision shall
authorise the Commission to reduce or cancel the assistance granted. It is further stated at point 2 of Section C of
Annex III to the decision that: If the Commission observed that a certain operation has not complied with or is not
complying with Community policy, it shall retain payment of Community funds for that operation and shall notify this to
the authority responsible in the Member State for implementing the operation.

64.
I consider that the above obligations imposed on the Commission flow from a framework of rules
which may in a proper case give rise for a certain category of individuals to the right to seek the
assistance of the courts to secure compliance with those obligations. Moreover, it is not by chance that
the Community judicature has already discerned the need to link the obligations imposed on national
authorities by clear rules of secondary Community law relating to environmental protection with the
possibility for individuals to secure compliance with those obligations through the courts. That is the
conclusion to be inferred from the aforementioned case law of the court on the direct effect of Directives
80/68, 79/409 and 80/779. More specifically, there is a discernible intention in that case law on the part of
the Community judicature not to leave monitoring of observance of that secondary Community law in the
sole care of the privileged parties mentioned in the second 644 paragraph of art 173 of the Treaty
because the legislation in question is perceived to be closely interrelated with the protection of a certain
group of individuals. I am of the view that the same philosophy should inform the court in cases, such as
the present one, in which the Community rule does not merely impose obligations on member states but
directly concerns the activity of Community bodies. Moreover, the fact that the Commission does not itself
intervene in the environment but merely finances an intervention in it does not necessarily mean that its
decisions in that connection may not be likely to affect certain individuals, and indeed in a direct and
individual manner40.
40
Particularly revealing in that connection is the recent judgment in Geotronics v European Commission Case C-395/95 P
[1997] ECR I-2271. In accordance with the settled case law of the court in the matter of public works contracts financed
by the European Development Fund, measures adopted by bodies acting on behalf of the Commission, whether
approvals or refusals to approve, endorsements or refusals to endorse, are intended solely to establish whether or not
the conditions for Community financing have been met, and are notand cannot beintended to interfere with the
principle that the contracts in question remain national contracts for which the beneficiary states alone are responsible
(see also judgments in STS Consorzio per Sistemi di Telecommunicazione via Satellite SpA v EC Commission Case
126/83 [1984] ECR 2769, CMC Cooperativa muratori e cementisti v EC Commission Case 118/83 [1985] ECR 2325 and
Italsolar SpA v EC Commission Case C-257/90 [1993] ECR I-9).

65.
It must, I believe, be accepted that observance, in the present case, of the Commissions specific and
clear obligation to take into account the safeguarding of environmental interests at the same time as
adhering to the relevant Community legislation during the financing of the relevant works in the Canary
Islands, is not of concern solely to the Commission but is also of relevance for certain individuals.
None the less, the court held that, in light of the specific nature of the Geotronics v European
Commission Case C-395/95 P [1997] ECR I-2271 case, a Commission decision refusing a tender from a
company, in the context of the PHARE programme, for the supply of electronic equipment to Romania,
concerned that company individually. On that ground it set aside the judgment of the Court of First
Instance which had held to the contrary. Thus, the fact that the Commission simply finances an activity,
which it does not itself directly undertake, does not in principle preclude its decisions in that connection
from affecting certain persons, and indeed individually.
The latter may seek judicial protection in the event of non-compliance with that obligation, provided of
course that they satisfy the procedural requirements in that connection.
66.
Finally, I consider that in the context of the present case the appellants were entitled to plead before
the Court of First Instance, on the one hand, infringement by the Commission of their rights under
Directive 85/337 and, on the other, the damage they claimed to have suffered as a result of the fact that
the Commission 645 failed, as they maintain, to check, prior to continuation of the funding, whether the
relevant works were being carried out in accordance with Community environmental legislation. Certainly,
the foregoing considerations do not necessarily lead to the conclusion that the appellants action before
the Court of First Instance was admissible. The admissibility of their action depends on whether they
satisfied the preconditions of the fourth paragraph of art 173 of the Treaty and, in particular, the
requirement that the contested decision must concern them individually. I now turn to analyse that issue.

B Whether the contested decision was of individual concern to the appellants

67.
As stated above, the Court of First Instance, in the order appealed against, came to the conclusion
that the contested decision was not of individual concern to the applicants and, on that ground, adjudged
the action to be inadmissible. It is that view of the matter taken by the Court of First Instance that the
appellants are contesting in this appeal.
68.
The appellants also challenge the adherence by the Court of First Instance to settled case law to the
effect that a decision of a Community institution can be of individual concern to persons other than the
addressee thereof only if that decision affects them by reason of certain attributes peculiar to them, or by
reason of factual circumstances which differentiate them from all other persons and thereby distinguish
them individually in the same way as the person addressed. That is the major premiss of the legal
reasoning followed by the Court of First Instance.
69.
The appellants also developed their view before the Court of First Instance that the Community
judicature, particularly in their case, should circumvent the strict procedural restrictions imposed by the
aforementioned case law. In that connection they rely on the specific nature of the environmental legal
interest, on developments at national and international level in the direction of more effective judicial
protection of legal interests of an environmental nature and on the danger of the legal vacuum in their
legal protection created by the order appealed against. Indeed, they advance a series of criteria which, in
their view, the Community judicature must assess when it is called upon to decide upon the admissibility
of an action involving a question of damage to the environment.
70.
1 shall next examine the question whether the order of the Court of First Instance appealed against
was correct. In so doing, I shall distinguish between (b) the position which it took as against applicant
individuals and (c) that which it maintained as against the applicant environmental associations. However,
before embarking on those issues, I consider it essential to refer to a problem discussed by the parties
which of course argued the point from opposing positionstouching on the quality and completeness of
judicial protection afforded by the national (Spanish) courts (a).
646
(a) Judicial protection afforded by the national courts as a ground for denying locus standi to the
appellants
71.
The extent to which a natural person may seek and obtain judicial assistance within the national legal
order is generally one of the matters examined when inquiry is made into the correct interpretation and
application of the procedural preconditions laid down in the fourth paragraph of art 173 of the Treaty. Even
though, from a theoretical viewpoint, I am opposed to the consideration of that factor because the
possibility of seeking redress before the national courts does not preclude the possibility of directly calling
in question before the Community judicature the legality of a decision of a Community institution, where
the procedural requirements of the fourth paragraph of art 173 of the Treaty are satisfied 41, I do consider it
necessary to stress the following further matters:
41
Consequently, the quality of judicial protection afforded under national law cannot be used as a criterion in reviewing the
correct interpretation of those procedural preconditions and whether on each occasion the relevant factual
circumstances may be brought within them, particularly since it is not mentioned expressly in the fourth paragraph of art
173 of the Treaty. Moreover, I can scarcely imagine the Community judicature embarking on a detailed appraisal of that
quality, merely to compare it with the judicial protection which it itself affords. For that reason, the Court of First Instance
was in my view correct not to draw any inferences from that issue, though invited to do so by the parties.

72.
(i) First, I do not agree with the view put forward by the Spanish government to the effect that the
appellants were seeking a decision by the Community judicature on the legality of a decision of a national
authority and for that reason had contrived the present dispute on the Commissions decision, which in
reality does not concern them. It is my belief that the appellants are solely contesting the decision of a
Community institution, and are alleging an infringement by that institution of a rule of Community law
which, according to them, affects their legally protected rights and interests 42. The fact that the contested
Commission decision relates to construction works in Spain does not necessarily mean that persons
desirous of contesting it have to proceed before the Spanish courts. Nor does it mean, at least in theory,
that that decision may not, in and by itself, affect the rights and legal interests of certain persons,
irrespective of whether any national administrative acts concerning those construction works subsist
against which only the means of redress afforded by national law are available to the persons concerned.
42
That is also why, moreover, the reliance placed by Spain on the judgment in Borelli [1992] ECR I-6313 is not on this
point felicitous. The Court of First Instance was not requested by the appellants to adjudicate on the legality of a
decision of the Spanish authorities adopted within the framework of the Community decision making process, as was
 the case in Borelli. Thus, the appellants did not maintain that the contested Commission decision was unlawful on the
ground that it was based on an unlawful decision adopted by or on irregularities on the part of the Spanish authorities.
They are maintaining that the Commission did not check, as it ought to have done, the legality of the acts or omissions
of the national authorities, solely in the light of Community law. Consequently, their argument is on a different footing
from that of the Borelli judgment. In that case, the action was brought against a Community act which in law required the
concurrence of the national authorities, and the applicants arguments referred exclusively to the illegality of the opinion
given at national level. Notwithstanding that fact, the Borelli judgment is not entirely irrelevant to the present case (see
footnote 63, below).

73.
(ii) Both the Commission and the Spanish government further argue that the applicants could have
secured adequate judicial protection if they had confined themselves to seeking redress before the
national courts; in the proceedings for that purpose they could have raised the issue of the legality of the
Commission decision to continue financing, whereupon, in the event of doubt, the national court could
have referred a question in that connection to the Court of Justice for a preliminary 647 ruling. They go on
to stress that that solution was the most appropriate if it is borne in mind that the national court is the
ordinary court having jurisdiction to apply Community law.
74.
I think it useful to point out that, as is apparent from the documents in the case-file and from the
submissions on behalf of the parties at the hearing before the court, certain of the appellants sought to
prevent the construction works in question by seeking the appropriate remedies before the competent
national courts. Those proceedings are still pending. On the other hand, there is no question pending
before the Court of Justice submitted for a preliminary ruling by a Spanish court on the issue of the
legality of the relevant Commission decision to continue financing of the works on the Canary Islands. Nor
do I see in what way the issue of the legality of that decision could be raised in the context of national
proceedings. Those proceedings can concern only the lawfulness of the administrative authorisations
granted for construction of the electricity-generating power stations, or of the environmental impact
assessment. But, even if any such supplementary issue could, exceptionally, be raised 43, the protection
likely to be afforded by the national court would certainly not be as far reaching and comprehensive as
that which would have been secured by the appellants, had their action before the Court of First Instance
been successful. The judgment of the national court could not extend to cover the issue of legality of the
financing per se or, a fortiori, lead to the setting aside of the Commission decision to continue financing.
43
The only theoretical situation which I could imagine is one in which the Commission decision is relied on by the Spanish
authorities before the national courts in order to justify commencement of the construction of the electricity generating
plants in the Canary Islands without an environmental impact assessment. In other words the Spanish authorities could
argue that they did not proceed to draw up an impact assessment in accordance with Directive 85/337 because that was
not essential in the present case, as is corroborated, moreover, by the position adopted by the Commission which
continued to finance the works. And then again, of course, in order to reject that argument, the national court (on the
supposition that, under the directive, such assessment was mandatory) would not need to decide first on the legality of
the Commission decision or, a fortiori, to refer a question to the court for a preliminary ruling.

75.
Finally, if the appellants satisfy the procedural requirements for contesting the Commission decision by
means of an action under the fourth paragraph of art 173 of the Treaty, then, whatever may be the
possibilities of judicial protection under national law, those possibilities do not in the slightest act as a bar
to their having locus standi before the Court of First Instance.
(b) Locus standi of the appellant individuals
76.
First and foremost, I do not consider that the appellants are entitled to challenge the major premiss of
the reasoning followed by the Court of First Instance, by arguing that in their case the fourth paragraph of
art 173 of the Treaty, as consistently interpreted by the Court of Justice, should not be applied. The strict
and indeed restrictive criterion of admissibility according to which the contested act must be of individual
concern to the applicants was expressly laid down by primary Community law. Properly construing that
provision, the Court of First Instance is requiring persons who are not addressees of a decision of a
Community institution to plead and prove the existence of certain attributes peculiar to them, or of factual
circumstances which differentiate them and individually distinguish them from all other persons 44. That
view of the matter cannot be called in question either by the specific nature of the legal interest in
environmental protection or by the matters 648 relied on by the appellants concerning modern
developments in national and international law in the matter of environmental protection 45.
44
In any event I will not dwell on the further element as to the manner in which the applicants are distinguished in the
same way as the person addressed. That specific requirement, which is not expressly laid down in the relevant
provision of the fourth paragraph of art 173 of the Treaty, is correctly demanded by the Community judicature in cases of
an economic nature but cannot play the same role in cases such as the present one because, otherwise, it would
deprive that provision of all effect. The result would be a reductio ad absurdum in which individuals or environmental
associations, in proceedings against a decision of the Commission concerning the payment of financial aid to a member
state for works having an impact on the environment, are required to show that they are in law and in fact identical with
the addressee of the decision, that is to say may be assimilated to the state to which the financing is allocated. In other
respects the distinguishing requirement in regard to applicants remains entirely legitimate and current.
45
The procedural restrictions contained in the fourth paragraph of art 173 of the Treaty cannot be called in question even
where the interest at stake is the protection of the environment. The importance of preserving the environment,
underlined by the Rio Declaration, Agenda 21 and other associated texts, cannot exempt the appellants from the
aforementioned procedural restrictions. Furthermore, the court recently held that the Fifth Environmental Action
Programme, approved by the Council and representatives of the member states on 1 February 1993, aims to create a
framework for determining and giving effect to Community policy in the environmental sector, but does not contain
binding rules of law (see the judgment in Rovigo [1995] ECR II-455 (para 32)).

77.
 None the less, the appellants further submit that the Court of First Instance erred in law by
misinterpreting the procedural rule contained in the fourth paragraph of art 173 of the Treaty and drawing
erroneous conclusions from the case law of the Court of Justice on that rule. That ground of appeal is
formally admissible and I now turn to examine it.
(i) Existing case law
78.
(aa) It is, I think, essential to refer to the courts case law in which the above mentioned procedural
requirement has been interpreted and particularised. In the first place, it should be emphasised that,
notwithstanding the apparent homogeneity of its judgments, at least in their formulation, the court does
not profess to adhere to an entirely immutable point of view 46. Indeed, it is prepared to ease the
procedural obstacles, when the specific nature of a case so requires, for the sake of affording more
comprehensive judicial protection47.
46
I refer eg to the alteration in the case law brought about by the court in its judgment in Sofrimport SARL v EC
Commission Case C-152/88 [1990] ECR I-2477 in relation to an earlier authority. On that case, see para 86, below.
47
Eg the court interpreted the relevant procedural criterion broadly and in a manner favourable to the applicants in its
judgment in Parti cologiste Les Verts v European Parliament Case 294/83 [1986] ECR 1339 in which it held that not
only the political groups in existence when the decision of the President of the Parliament was taken on the allocation of
credits authorised to cover the expenses of the political groups for the 1982 elections, but also groups unknown at that
time, were entitled to bring proceedings against that decision.

79.
In the first place, locus standi in favour of individuals presents no particular difficulties in cases where
they have participated in the preparation of the contested act 48 or where Community law has made
provision for a special procedure prior to adoption of the contested act, in which those individuals may
participate and submit their observations. On the basis of that reasoning, an action brought by a
company, which complained to the Commission of an agreement contrary to art 85 of the Treaty between
its competitors, against the Commission decision exempting that agreement under art 85(3) from the
prohibition laid down in art 85(1), was held to be admissible 49. Also held to have standing have been
persons who complained to 649 the Commission of unlawful state aid and submitted observations in the
procedure under art 93(2) of the Treaty (see the judgments in Compagnie franaise de lazote (Cofaz) SA
v EC Commission Case 169/84 [1986] ECR 391 and William Cook plc v EC Commission Case C-198/91
[1993] ECR I-2487)50.
48
See the judgment in Timex Corp v EC Council Case 264/82 [1985] ECR 849 on the adoption of an anti-dumping
regulation.
49
The court in that case based itself on the significance of the procedure under art 3 of Regulation 17/62 under which
certain persons are authorised to request the Commission to establish an infringement of art 85 of the Treaty.
Accordingly, it is logical for the same persons to have access to the courts in order to protect their legal interests
recognised by Regulation 17/62 (see the judgment in Metro SB Gromrkte GmbH & Co KG v EC Commission Case
26/76 [1977] ECR 1875).
50
The same applies, where the procedure under that para was not followed by the Commission, to persons who could
have submitted observations, had the art 93(2) procedure been implemented judgment in Matra [1993] ECR I-3203). In
that judgment the court emphasised the special nature of the procedure under art 93(2) of the Treaty and the right to a
hearing afforded by that procedure to interested parties. It is to be noted that a competitor undertaking may likewise
challenge a Commission decision finding that a merger decision between undertakings is compatible with the common
market, where that undertaking submitted observations under the procedure laid down in Regulation 4064/89 and
received the reply that its observations would be fully taken into consideration. A competing undertaking may even
contest a statement by the Commission that the merger in question does not have a Community dimension and thus
does not come within the scope of Regulation 4064/89. That is also the position hitherto taken by the Court of First
Instance in Air France [1994] ECR II-323 and Air France [1994] ECR II-121.

80.
The justification, in the final analysis, for the courts upholding the admissibility of the actions in the
above cases was elucidated in the judgment in Cofaz [1986] ECR 391 (paras 2223) as follows:
(22) It is clear from a consistent line of decisions of the Court that persons other than those to
whom a decision is addressed may claim to be concerned within the meaning of the second
paragraph of Article 173 only if that decision affects them by reason of certain attributes which are
peculiar to them, or by reason of circumstances in which they are differentiated from all other
persons, and by virtue of these factors distinguishes them individually just as in the case of the
person addressed
(23) More particularly, as regards the circumstances referred to in that judgment, the Court has
repeatedly held that where a regulation accords to applicant undertakings procedural guarantees
entitling them to request the Commission to find an infringement of Community rules, those
undertakings should be able to institute proceedings in order to protect their legitimate interests
(My emphasis.)
81.
Conversely, where such a procedure is not provided for under Community law, it is difficult for
individuals in an analogous situation to those mentioned above to secure protection by the Community
judicature51.
51
Eg a Commission decision granting EAGGF financing to certain undertakings cannot be challenged by competitors of
those undertakings (see the judgment in Societ Eridania Zuccherifici Nazionali v EEC Commission Joined cases
10/68 and 18/68 [1969] ECR 459). For the same reason, moreover, persons in receipt of state aid are not deemed to be
individually affected by a Commission decision declaring such aid incompatible with the common market (see the
judgment in Van der Kooy [1988] ECR 219).
82.
(bb) In other cases actions brought by individuals frequently run up against the courts settled case law
to the effect that a measure cannot concern applicants individually when it applies to objectively defined
situations and produces legal effects on categories of persons determined in a general and abstract
manner. It should also be noted that whether a measure challenged is objective in nature must 650 not be
confused with the number of persons affected by it. A single person may be affected by a decision, but
that does not mean that such person is individually affected by it for the purposes of the fourth paragraph
of art 173 of the Treaty52. Again
52
See eg the judgment in Spijker. In that judgment it was held that the decision whereby the Commission permitted
Belgium, Luxembourg and the Netherlands to exclude products such as brushes, brooms and similar products imported
from China did not concern the applicant in spite of the fact that it was at the time the sole importer into those member
states of the relevant products. The court considered that the Commission decision affected the applicant qua importer
of those products in the same way as any other undertaking actually or potentially in the same situation.

the possibility of determining more or less precisely the number or even the identity of the
persons to whom a measure applies by no mean implies that it must be regarded as being of
individual concern to them as long as it is established that such application takes effect by virtue of
an objective legal or factual situation defined by the measure in question (See the judgment in
Buralux SA v EU Council Case C-209/94 P [1996] ECR I-615 (para 24).)53

53
See also the judgment in Abertal SAT Ltda v EC Council Case C-264/91 [1993] ECR I-3265 (para 16).

83.
(cc) Furthermore, the court appears to accept that an act is of individual concern to the applicant only
where he belongs to a closed class of persons. I nevertheless believe that in that regard the courts case
law shows evidence of significant developments and fluctuations. (1) Initially, as for example in the
Toepfer v EEC Commission Joined cases 106107/63 [1965] ECR 405 and Bock v EC Commission Case
62/70 [1971] ECR 897 cases, the court deemed it decisive, in favour of the admissibility of an action, that
the number and identity of importers concerned in this way was already fixed and ascertainable before
the date of adoption of the decision. It also considered that the Commission was in a position to know that
its decision would affect the interests and situation of those importers alone. It went on to conclude that
the factual situation thus created differentiates the latter from all other persons and distinguishes them
individually just as in the case of the person addressed (see Bock [1971] ECR 897 (para 10)). (2)
Subsequently, the court appears to have determined the closed class of potential applicants to whom a
contested decision is of individual concern in a manner more favourable to individuals than was originally
the case. In Piraiki-Patraiki [1985] ECR 207 (para 19), a case concerning an action for the annulment of a
Commission decision authorising France to adopt protective measures against imports of cotton yarn of
Greek provenance, brought by a group of Greek exporters, the court stated:
it must be held that the fact that, before the adoption of the decision at issue, they had
entered into contracts which were to be carried out during the months to which the contracts
applied constitutes a circumstance which distinguishes them from any other person concerned by
the decision, in so far as the execution of their contracts was wholly or partly prevented by the
adoption of the decision.
It is noteworthy that in that case the Commission called in question the admissibility of the actions in
reliance on the fact that, when it adopted its decision, it was unaware of the number of contracts which
had been entered into in respect of the period covered by the decision. The court held ([1985] ECR 207
(para 21)):
in that respect it must be observed that the reply to be given to the question whether and to
what extent the Commission was aware which Greek 651 exporters had entered into contracts
covering the period of application of the contested decision depends on the interpretation given to
Article 130 of the Act of Accession, and in particular on the question whether the Commission,
before authorising a protective measure under that provision, is obliged to make appropriate
enquiries as to the economic effects of the decision to be taken and the undertakings which would
be affected by it (My emphasis.)
I would observe that in that judgment the court distanced itself from the condition originally laid down
concerning definition of the closed class, namely that the persons comprising that closed class had of
necessity to be known to the authority adopting the contested act at the time of its adoption. In coming to
the view which it took on the matter, the court took account of the special nature of the Commissions
obligations when adopting the contested decision and the interrelation between those obligations and
protection of the applicants legal interests.
84.
That judicial trend was reaffirmed in Sofrimport SARL v EC Commission Case C-152/88 [1990] ECR I-
2477. The case concerned an action brought by an importer of apples from Chile against the Commission
regulation suspending the issue of import licences for that fruit and determining import quotas.
Notwithstanding the Advocate Generals opinion to the contrary, the court considered importers of Chilean
apples which were in transit when the regulation at issue was adopted to constitute a restricted group
which is sufficiently well defined in relation to any other importer of Chilean apples and cannot be
extended after the suspensory measures in question take effect (see [1990] ECR I-2477 (para 11)). It
may be seen that the court no longer refers to persons defined in number and identity but to a restricted
group, sufficiently well defined.
85.
 In reaching that conclusion, the court had regard, as in Piraiki-Patraiki [1985] ECR 207, to the
Commissions specific obligations, under the applicable Community law, in regard to the framing of its
contested decision. In actual fact, a Council regulation had imposed on the Commission the requirement
to take into consideration, when adopting protective measures, the specific situation of goods in transit to
the Community. It is also noteworthy that in Sofrimport [1990] ECR I-2477 it is expressly stated that,
where a Community rule affords special protection to a specific group of importers, they must therefore
be able to enforce observance of that protection and bring legal proceedings for that purpose (see [1990]
ECR I-2477 (para 12)).
86.
The judgment in Sofrimport is also of interest from another point of view. In that judgment the court
accepted that it was possible for there to be a closed class of personsentitled to bring an action under
art 173 of the Treatywithin an open class of persons to whom that procedural possibility is not
available. Thus, the Community act does not need to concern solely the members of that closed class (in
Sofrimport a specific group of importers), but may also affect, in an objective manner of course, persons
comprising the open class of actual or potential importers 54. None the less, it is essential for that closed
class to be sufficiently well defined (see [1990] ECR I-2477 (para 11)). For the purposes of such
definition, the court attached particular importance to the nature and extent of the Commissions
obligations under the applicable legislation.
54
The court appears to have abandoned the opposing viewpoint which it had adopted in Moksel Import-Export GmbH &
Co Handels KG v EC Commission Case 45/81 [1982] ECR 1129. In that case an exporter of beef meat had brought an
action against a Regulation suspending advance fixing of export refunds for beef meat, relying for locus standi on the
fact that it belonged to a closed class, known in advance and fully ascertained, of undertakings which had lodged
requests for refunds before the entry into force of the regulation and whose requests were still pending. The court did
not accept the applicants submissions, notwithstanding the Advocate Generals opinion to the contrary, taking the view
that the class of persons in the aggregate concerned by the regulation was not closed. Since Article 1 of Regulation No
3318/80 concerns both earlier applications and those lodged during the period of suspension, the nature of the
contested measure as a regulation is not called in question merely by the fact that it may be possible to determine the
number or even the identity of certain traders concerned, especially where such a possibility by definition did not exist
for other traders also covered by Regulation No 3318/80 (see [1982] ECR 1129 (para 17)).

(3) I wish now to draw attention to two judgments of the court which highlight the importance attached
by the Community judicature to the effects which the contested decision is likely to have on the applicant.
Those effects may give rise to a factual situation which sufficiently distinguishes the applicant.
652
87.
In Extramet [1991] ECR I-2501, an importer sought the annulment of a regulation imposing anti-
dumping duties on imports into the Community of certain products from China and the Soviet Union. On
the basis of the case law as it then stood, a distinction was made between producers, exporters and the
complainants, on the one hand, and independent importers, on the other. As regards the latter, the issue
of admissibility was dealt with by the court in a particularly succinct manner 55.
55
Regulations imposing temporary and definitive antidumping duties have been held to constitute, as regards independent
importers, measures of general application because they apply to objectively determined situations and entail legal
effects for categories of persons regarded generally and in the abstract (see the judgment in Alusuisse Italia SpA v EC
Council Case 307/81 [1982] ECR 3463 (para 9)). In accordance with its settled case law, the court had refused to
recognise a right of action in favour of independent importers even where the case concerned the sole importer within a
state of the product on which the duty was imposed (see the orders in Sermes (SA) v EC Commission Case 279/86
[1987] ECR 3109 and Nuova Ceam Srl v EC Commission Case 205/87 [1987] ECR 4427) on the ground that a
measure does not cease to be a regulation because it is possible to determine the number or even the identity of the
persons to whom it applies at any given time as long as it is established that such application takes effect by virtue of an
objective legal or factual situation defined by the measure in relation to its purpose (see the judgments in Zuckerfabrik
Watenstedt GmbH v EC Council Case 6/68 [1968] ECR 409 and Alusuisse [1982] ECR 3463 (para 11)).

Consequently, Extramet, as an independent importer of calcium metal, could not be deemed to be

affected by the contested regulation having regard to its status, which did not differentiate it from any
other undertaking already carrying on the same activities or capable of doing so in the future 56.
56
In other words, the regulation, as regards independent importers, could not be regarded as equivalent to a decision
concerning them individually but contained general and abstract rules of law which by their nature did not affect persons
individually.

88.
None the less, Advocate General Jacobs proposed that the court should remove that obstacle, stating
as follows ([1991] ECR I-2501 (paras 7576)):
(75) I am accordingly of the opinion that the Court should recognize that a measure imposing
an anti-dumping duty is of direct and individual concern to any undertaking which is able to
establish that it is identified, explicitly or implicitly, by the measure in question The Court
should in my view clarify the case-law by expressly acknowledging that, at least in the anti-dumping
field, it is not necessary for an applicant, in order to establish standing, to address the additional
question of whether the contested measure constitutes a regulation or a decision.
(76) This approach accords with the purpose of Article 173, which is designed to enable persons
to challenge measures having a particular impact on them, while limiting the right to challenge
regulations so that there is no risk of their annulment being sought by an unlimited class of
applicants. (My emphasis.)
653
89.
For its part the court, after recalling that an anti-dumping regulation may affect undertakings owing to
certain attributes peculiar to them and which differentiate them from all other persons, held ([1991] ECR I-
2501 (para 17)):
the applicant has established the existence of a set of factors constituting such a situation
which is peculiar to the applicant and which differentiates it, as regards the measure in question,
from all other traders. The applicant is the largest importer of the product forming the subject-matter
of the anti-dumping measure and, at the same time, the end-user of the product. In addition, its
business activities depend to a very large extent on those imports and are seriously affected by the
contested regulation in view of the limited number of manufacturers of the product concerned and
of the difficulties which it encounters in obtaining supplies from the sole Community producer,
which, moreover, is its main competitor for the processed product. (My emphasis.)
90.
In Codorniu SA v EU Council Case C-309/89 [1994] ECR I-1853 the court was called upon to
adjudicate on an action brought by a Spanish producer of sparkling wine against a provision of a
regulation laying down the conditions for use of the term crmant for sparkling wines. The Council raised
an objection of inadmissibility, contending that that measure was of a purely legislative nature and did not
concern the applicant, except in its capacity as a producer using the term cremant, that is to say in the
same way as it concerned any other producer in the same situation. In his opinion in that case, Advocate
General Lenz found first, as a matter of undisputed fact, that the contested measure was of a legislative
nature. Notwithstanding that fact, he did not consider that the action should be dismissed without more
ado as inadmissible, or that detailed examination was required of the question whether the measure in
point, though overall of a legislative nature, nevertheless had the effect of a decision as regards the
applicant. Instead, he examined whether the measure was of individual concern to the applicant. In order
to reply to that question, the Advocate General first considered whether, amongst the undertakings to
which the prohibition on use of the term crmant was of concern, there could be identified a category
comprising a fixed number of persons which could not be enlarged after adoption of the measure at issue.
The applicant undertaking did in fact come within that clearly defined category 57. However, that factor is
not sufficient because, in accordance with the case law, there must be a specific connection between the
applicants situation and the contested measure. A particular link of that kind, which also enables the
class of potential applicants under the fourth paragraph of art 173 of the Treaty to be determined, may be
that identified in the judgment in Extramet . The relevant criterion is the effects which a measure has on a
certain undertaking and which may be differentiated from the effects of that measure on other persons 58.
Finally, the Advocate General, 654having regard to all the evidence adduced by the applicant in
connection with the impact on it of the contested measure, concluded that the applicant is also
distinguished from the category of traders affected by the effects which the measure has on its
undertaking, and the measure is thus of individual concern to it (see the opinion in Codorniu [1994] ECR
I-1853 (para 64)). The court reached the same conclusion on the basis of evidently more elliptical
reasoning59.
57
That is to say, producers using the term crmant before entry into force of the regulation at issue.
58
The criterion of effects on the situation of the applicant is not applied only in Extramet and Codorniu. Also in cases
where the applicants participated in the preparation of the contested act and thus appear to be individually affected by it
(see para 79, above), the court accepts that the preconditions of art 173 are met only if the damage which the applicants
are likely to sustain is of a certain seriousness. Thus, undertakings which participated in the procedure under art 93(2)
of the Treaty in order to oppose state aid have standing to bring an action before the Community judicature provided,
however, that their position on the market is significantly affected by the aid which is the subject of the contested
decision (see Cofaz [1986] ECR 391 (para 25)).
59
Codorniu registered the graphic trade mark Gran Cremant de Codorniu in Spain in 1924 and traditionally used that
mark both before and after registration. By reserving the right to use the term crmant to French and Luxembourg
producers, the contested provision prevents Codorniu from using its graphic trade mark. It follows that Codorniu has
established the existence of a situation which from the point of view of the contested provision differentiates it from all
other traders (see [1994] ECR I-1853 (paras 2122)).

91.
The significance and extent of mitigation by the court, in Extramet and Codorniu, of the rigour of the
case law should not, however, be overstated. At least I infer as much from the judgment in Buralux SA v
EU Council Case C-209/94 P [1996] ECR I-615. In that case the court, on appeal, upheld as correct an
order of the Court of First Instance which had dismissed as inadmissible an action for annulment of
provisions of a regulation concerning the monitoring and control of shipments of waste within the
Community. The action had been brought by undertakings engaged in the collection, shipment and
dumping of household waste. The Advocate General proposed that the court should declare the action
brought by one of the undertakings to be admissible under the Extramet criteria60. The court did not follow
that proposal, regarding as decisive the fact that the legal effects which the contested regulation was
likely to produce concerned categories of persons in a general and abstract manner.
60
On the basis of these criteria it may very well be assumed that in this case the appellants by reason of special personal
circumstances are differentiated from all other persons concerned. Buralux, together with its partners, is the largest
importer, at least in the France/Germany area, and as it cannot fulfil its continuing contracts it is affected particularly
seriously by the regulation and the import prohibition envisaged therein. These contracts are almost all valid beyond the
date on which the regulation was to become applicable. In my view it may therefore be stated that in this case it is of
individual concern to the appellants (see the opinion of Advocate General Lenz: [1994] ECR I-1853 (para 33)).

92.
I do not believe, however, that that judgment constitutes a departure from Extramet. The court was
seeking to exclude cases in which an increasingly broad interpretation and application of the procedural
provisions of the fourth paragraph of art 173 of the Treaty would lead to recognition of a right of action by
individuals against rules of Community law and not against individual decisions. In other words, in
Buralux, the court wished to safeguard the legislative nature of the regulation, as it did in Deutz und
Geldermann, Sektkellerei Breisach/Baden GmbH v EC Council Case 26/86 [1987] ECR 941 (see paras
100 and 101, below).
(ii) The procedural restrictions in the fourth paragraph of art 173 of the Treaty and the particular
nature of the present dispute
93.
(aa) I shall first examine whether, on the basis of the inferences to be drawn from the case law
referred to above, the contested decision of the Commission to continue financing of the construction
works for the two power stations in the Canary Islands may be deemed to be of individual concern to the
appellants.
94.
As I stated above, the appellants have relied in support of their action both on the specific obligations
imposed on the Commission on adoption of the contested act, requiring it to monitor whether the works
were progressing in conformity with Community environmental law, and on the rights which they consider
are 655 conferred on them by Directive 85/337 in connection with the environmental impact assessment
for works such as those in the present case. However, they do not call in question the correctness of the
lower courts reasoning to the effect that the fact that certain of the appellant individuals lodged
complaints with the Commission or exchanged correspondence with it does not in itself suffice for the
contested act to be deemed to concern them individually (see [1995] ECR II-2205 (para 56)).
Consequently, I shall not dwell on that issue.
95.
It could nevertheless be argued that, whilst for the adoption of the contested decision no direct
provision is made for participation by interested parties 61, none the less, the procedure laid down in art
6(2) of Directive 85/337 should be regarded as such a participatory procedure in relation to the adoption
of the contested decision (see paras 58 and 59, above). Thus, Community law may not have included
individuals in the procedure for preparation of Commission decisions on financing of works having an
environmental impact, but it does require the Commission to verify prior to continuation of financing
whether the relevant works are being carried out in conformity with Community provisions, including the
provision contained in Directive 85/337 allowing the public concerned to participate in the drawing up of
the environmental impact assessment. It could be inferred from the combination of those obligations on
the Commission and from the rights in favour of the public concerned conferred by Directive 85/337 that
those persons comprising the public concerned are sufficiently differentiated in relation to the contested
act. They are therefore distinguished from all other persons because Directive 85/337 has afforded them
concrete procedural guarantees analogous to those in Cofaz, Metro SB Gromrkte GmbH & Co KG v
EC Commission Case 26/76 [1977] ECR 1875 and Matra SA v EC Commission Case C-225/91 [1993]
ECR I-3203. In other words, by dint of reasoning by analogy, it might be inferred from the Commissions
specific obligation to monitor whether the works financed are progressing on the basis of Community
environmental law and, thus, in accordance with the provisions of Directive 85/337 that the persons on
whom rights are conferred by Directive 85/337 constitute a closed class and thus have locus standi by
analogy with the Piraiki-Patraiki and Sofrimport cases.
61
For a discussion of an analogous problem see Rovigo [1995] ECR II-455.

96.
I cannot adhere to that view or, at least, cannot regard it as founded on existing case law. The
procedural guarantees relied on by the appellants are contained in the text of a directive and not a
regulation, unlike the situation before the court in Cofaz and Metro. That difference is not without
importance. Regulations and directives do not have the same binding effect as to their content.
Furthermore, the cases in which the court founded itself, in conferring locus standi on applicants, on the
existence of procedural guarantees were concerned purely with Community law procedures, implemented
exclusively by Community bodies. In contrast, the procedure for drawing up an environmental impact
assessment is national in character, is governed by rules of both Community and national origin and is a
matter for national authorities. But, irrespective of those differences 62, it would be difficult to equate the
procedure provided for in Directive 85/337 with the procedure for the adoption of the contested
Commission decision, in respect of which there is no direct provision for participation by interested
parties. Finally, it 656 does not follow from the nature of the obligation imposed on the Commission in
connection with the adoption of the contested act that a group of individuals has locus standi to bring
proceedings in the event of any infringement of that obligation. That obligation consists in the monitoring
of the application of Community legislation by the national authorities which implement the funded
projects and does not expressly concern the protection of certain persons. But even if the monitoring
obligationbecause it extends to cover the correct application of Directive 85/337is deemed to include
those provisions of that directive which provide for participation of the public concerned in the drawing up
of the environmental impact assessment, that does not automatically mean that those persons comprising
the public concerned are individually affected by the contested Commission decision. The public
concerned under Directive 85/337 cannot be regarded as constituting a closed class as defined in the
courts case law. Moreover, the concept of public concerned is not sufficiently defined by Directive
85/337. The definition of that term is left to the national courts. Thus, in Directive 85/337 Community law
dictates the protection of a group of persons which, however, it does not precisely define 63.
62
Those differences are not so crucial if it is accepted that, for the purposes of Directive 85/337, national authorities are in
essence exercising a Community competence conferred on them by the directive, that is to say in the context of rules of
 Community law. That observation cannot, however, detract from the national character of acts adopted by national
bodies pursuant to Directive 85/337.
63
Moreover, if, finally, a Commission decision, such as the one at issue, were deemed individually to affect those persons
constituting the public to which Directive 85/337 refers, then, whenever an environmental impact assessment is required
for infrastructure works benefiting from financing (most frequently the case), a particularly wide category of persons
could bring proceedings before the Community judicature against decisions of the Commission concerning financing of
the works, basing their locus standi on the non-existence of or defects in the environmental impact assessment. A
development of that kind would run directly counter to the judgment in Borelli, according to which the court does not
have jurisdiction to decide upon the lawfulness of acts of a national authority, even where the national act is adopted in
the framework of the Community decision-making procedure. In any event, however, the majority of such proceedings
would be inadmissible for lack of any legal interest. We would thus arrive at the paradoxical situation in which the
procedural conditions laid down in the fourth paragraph of art 173 of the Treatythat the contested act must individually
affect the applicantwould be met more readily than the requirement of the existence of a legal interest. It might, finally,
be objected that relaxation to such an extent of the locus standi conditions is justified in extreme cases, such as the
present one, that is to say where the Commission declines to perform its supervisory function in order to bring to an end
a substantial irregularity, such as failure to obtain an environmental impact assessment. I cannot adopt that view of the
matter, even though I acknowledge its expediency, because I would then be interpreting the procedural requirements as
to admissibility after first appraising the substance of the dispute, and thus in a manner which would be methodologically
incorrect. However, I would also refer to the view which I develop in footnote 75, below.

97.
Accordingly, no inference may be drawn directly, either from the obligations imposed on the
Commission on adoption of the contested act or from the provisions of Directive 85/337, to the effect that
the contested act of the Commission individually concerned the appellant individuals within the meaning
of the fourth paragraph of art 173 of the Treaty, as interpreted in the courts case law.
98.
(bb) In light of the foregoing, the Court of First Instance in fact faithfully followed the case law
established hitherto. If the sole criterion of correct interpretation of the fourth paragraph of art 173 of the
Treaty is the reiteration of the viewpoint hitherto judicially accepted, then the order appealed against is
unimpeachable.
99.
Nevertheless, I take the view that the court should examine the possibility of taking a further step
forward from certain of its hitherto settled positions. The point on which I consider it expedient to submit to
particular examination the solution adopted by the Court of First Instance, and on which the need for
advancement in the case law is brought into sharp relief, is that part of the lower courts reasoning in
which the Commissions contested decision is deemed not to concern the appellants 657 individually on
the ground that it affected them in the same way as any other person living, staying or carrying on
business on Gran Canaria and Tenerife. According to the Court of First Instance, the contested decision
for them is a measure whose effects are likely to impinge on, objectively, generally and in the abstract,
various categories of person (see [1995] ECR II-2205 (para 54)).
100.
In the first place, I consider it relevant to emphasise that the general and abstract nature of the
impingement on the appellant individuals, on which the order of the Court of First Instance dismissing
their action is based, is not due to the legislative nature of the Commissions contested decision. The
significance of that point may, of course, be doubted. In accordance with existing case law, if a person is
affected by an act, whether of an individual nature or in the nature of a regulation, in a general and
abstract manner, that finding suffices to deny that person locus standi, without the need for further
distinction as between the legislative or non-legislative nature of that act 64
64
See eg the judgment in Union Deutsche Lebensmittelwerke GmbH v EC Commission Case 97/85 [1987] ECR 2265.
Proceedings brought by sellers of margarine against the Commission decision in favour of Germany in connection with
the promotion of sales of butter in the West Berlin market were held to be inadmissible. According to the courts
judgment, although the contested decision affects the applicants, that is only because of the effects it produces on their
position on the market. In that regard, the decision is of concern to the applicants just as it was to any other person
supplying margarine on the West Berlin market while the contested operation was in progress, and it is not therefore of
individual concern to them for the purposes of the second paragraph of Article 173 of the EEC Treaty (see [1987] ECR
2265 (para 11)).

101.
Yet I do not believe that the nature of the contested act is entirely without importance. When that act is
of a legislative nature, the court is particularly strict in denying individuals a right of action, precisely in
order to safeguard the legislative nature of the act (see para 92, above). In accordance with the expressly
formulated intention of the framers of the Treaty, a rule of Community law cannot form the subject matter
of proceedings brought by persons other than those mentioned in the second paragraph of art 173 of the
Treaty. That specific bar does not apply in cases where the contested act does not contain rules of law,
thus where the general and objective nature of the results produced by the act are due not to its
legislative nature but to its subject matter. In this problematical context, I believe the appellants
submission as to the particular nature of the consequences which an intervention in the environment has,
or is likely to have, to be worthy of consideration.
102.
For environmental protection is indeed a matter of general interest. Conservation of the environment is
a legal interest theoretically shared by all natural persons; it thus has a communal dimension 65.
Furthermore, the more significant is the intervention in or impingement on the environment, the greater is
the number of persons affected thereby.
65
For that reason, moreover, it has been recognized in certain member states as a fundamental social right.
103.
That unexceptionable statement cannot of course lead, on account of the special nature of the
environmental legal interest, to the setting aside of the procedural requirements laid down in the fourth
paragraph of art 173 of the Treaty. The conferral of a right of action before the Court of First Instance on
every person whose interest in conservation of the environment is affected by an act of a Community
institution would be tantamount to acceptance of an actio popularis in all cases having an environmental
dimension. As I stated at paras 53 and 76, above, a departure in the case law in that direction is
impossible because, apart from the practical obstacles which would be encountered, it would run counter
to the letter of the fourth paragraph of art 173 of the Treaty. It would equally be impossible to 658 lay
down specifically for cases raising issues of environmental protection requirements as to locus standi
which differ from those contained in the above mentioned provision. In the final analysis, the starting-point
for appraisal of locus standi cannot but be the same in all cases, irrespective of the subject matter of the
dispute: the applicant must be individually affected by the act which he contests.
104.
The points which I have made concerning the particular characteristics of the legal interest in the
environment and its protection are not intended to overturn the settled positions described above, but to
secure what is, in my view, the correct application of the fourth paragraph of art 173 of the Treaty 66. In
light of the particular nature of the issue of environmental protection, I consider that the Community
judicature, in particular in cases where that protection is likely to have been undermined by a non-
legislative act of a Community body, must not rest content with the self-evident finding that the likely harm
to the environment by its very nature affects categories of persons in a general, objective and abstract
manner, and must not, on that ground alone, dismiss the action brought. That is particularly so when the
Community body was under a specific and clear obligation to take account, on adoption of the contested
decision, of the factor of conservation of the environment (see paras 62 to 65, above). On this point, then,
I would draw the attention of the court to a possible and, in my view, appropriate easing of the
requirements laid down in the case law. Allow me to explain:
66
I believe, moreover, that judicial protection within a legal order governed by the rule of law must seek to safeguard the
rights and interests conferred by that legal order on persons subject to it. The procedural dimension of the legal interest,
as the precondition of admissibility of a legal action, cannot be distinguished in an absolute manner from the substantive
dimension of the legal interest, which the legal order seeks to safeguard in favour of the person entitled. Thus, account
should be taken of the particular nature of each interest protected by the Community legal order in determining the
specific procedural conditions under which the person in whose favour the rules enshrining that interest were created
may seek judicial assistance in upholding those principles. I further believe that to interpret the written procedural rules
in such a manner as entirely to debar a person entitled to a right or legal interest (in the substantive sense of that term)
from access to justice to defend his interests under the legal order, first, renders nugatory recognition by substantive law
of those rights and legal interests and, secondly, must be regarded as wrong in law, inasmuch as the procedural
provisions are laid down by the legal order in order to provide a structural framework, in so far as practicable, to give
effect to the rights or legal interests conferred on persons subject to it, and notabsolutely and uniformlyin order
entirely to debar them from judicial protection. Otherwise, if persons enjoying the protection of certain rules of law are
completely debarred by procedural rules from obtaining judicial protection, the legal order is simply abdicating its
function.

A decision which has an impact on the environment does indeed affect, or may affect, large categories
of citizens in a general and abstract manner. None the less, it is not impossible for one or more of the
persons affected, who constitute a closed class, to be particularly affected, and thus to be distinguished
from any other person, that is to say to be differentiated for the purposes of the fourth paragraph of art
173 of the Treaty. An intervention in the environment, such as that in point in the present case, is located
in a specific geographical area, and the extent of its impact is lessened the further away one is from the
area of the intervention67. Accordingly, persons close to the construction works suffer its consequences in
a different, more 659 intense manner than persons farther away, the latter being at a greater radius from
the epicentre of the intervention in the environment. By way of logical extension, it may be argued that
persons near the epicentre comprise a particularly closed and defined class, who find themselves in a
situation which differentiates them from any other person. It is then the task of the courts to determine, on
the basis of the appropriate criteria, the breadth of that closed class, the width of its radius. By logical
extension, persons within that class should be regarded as having locus standi to bring an action against
the decision occasioning consequences for the environment.
67
This phenomenon may be likened to the throwing of a stone into a lake which creates on the surface of the water a
series of concentric circles. Etymologically, moreover, the notion of a circle is inherent in the term environment. For
example, in Greek, the word for environment (perivallon) comes from pen (around) and vallo (throw); in French or
English from the word envirum; and the German equivalent Umwelt is made up of um (around) and Welt (world).
Thus, that term refers to an object which encompasses something else, that is to say encloses and embraces it.

105.
The criteria may not only be geographical. Geographical proximityreferred to in art 130r(2) of the
Treaty68is certainly useful, particularly in cases such as the present one, but will have to be weighed
together with the nature of the consequences which the intervention in the environment will or is likely to
have69, and principally having regard to the extent, that is to say the gravity, of those consequences 70.
Those are, moreover, the matters chiefly taken into account by the national courts in the member states in
disputes of this kind71.
68
Community policy shall be based on the precautionary principle and on the principles that preventive action should
be taken, that environmental damage should as a priority be rectified at source (my emphasis).
69
Eg construction of a conventional power station would be dealt with differently from that of a nuclear power station.
70
The gravity of the consequences was, moreover, one of the basic criteria taken into account in drawing up the annexes
to Directive 85/337 and distinguishing the works for which a mandatory environmental impact assessment is required
from those for which such assessment is optional.
 71
See eg: English law: in R v Secretary of State for Trade and Industry, ex p Duddridge [1995] ELR 151 the decision of a
public authority not to limit by regulation electromagnetic emissions from electricity cables may be challenged before the
courts by parents residing in the area in which new electricity cables are placed, relying solely on the increased danger
of leukaemia to which their children are exposed as a specific consequence of high electromagnetic levels. Belgian law:
Conseil dtat, ville de Lige et Heze, 20.9.1991, No 37.676 Proceedings by neighbour to quash decision approving
installation of plant using substances harmful to environment held to be admissible. Netherlands law: Raad Van State,
Afdeling Bestuursrechtspraak (Council of State, Administrative Law Section), 18.6.96, AB 1996, 313. Inhabitants of a
village may invoke expected reduction in road safety in their village in order to challenge projected works. German law:
Bundesverwaltungsgericht (Fed Admin Ct), 1.128Z BVerwGE 66 p 307 (the crab-fishermen case): proceedings held to
be admissible brought by fishermen against decision approving dumping at sea of liquid toxic waste on ground of
reduction in fish population as a result of dumping of waste. Italian law: TAR Lazio, 20.1.95, No 62, Foro Italiano 1995,
11-460. Inhabitants of an area may invoke their right to quality of life (interesse di vita) in order to challenge permission
to build shopping and trading centre in their area. Greek law: Simvoulio tis Epikratias (Council of State) 2281/1992:
Inhabitant of the centre of a large town held to have locus standi to seek the quashing of decisions authorising
clearance of wooded area on edge of town. Held that the town and threatened woods belonged to same geographical
basin which constituted an unbroken ekistic whole with very few green spaces in constant diminution. Thus, the
unfavourable consequences for the ecological balance and for the quality of life of its inhabitants of the decisions
leading to clearance of a wooded area in that basin are experienced not only by those in its immediate vicinity but also
by those in more distant and lower-lying areas, and indeed in some cases more intensely by the latter. French law:
Vicinity constitutes the principal criterion of locus standi for natural persons in planning cases (Conseil dtat, 22.10.86,
Reynaud, Lebon, p 652). In determining vicinity regard is had, in addition to distance from proposed works, to nature
and gravity of consequences arising. Thus, an applicant challenging building permit for large shopping centre (Conseil
dtat, 24.6.91, Soc Interprovence Cte dAzur Lebon, p 1110) does not need to be in such close proximity to the works
as an applicant challenging construction works having less significant environmental impact (CE 17.6.91, Renauld,
Lebon, p 1110). See also R Chapus, Droit du Contentieur Administratif, LGDJ, 6th edn, 1996, No 438).

106.
That reasoning is not entirely foreign to existing case law. As I have said, it is possible for an act to
concern both an open class of persons (in environmental matters that class is especially wide) without
locus standi to challenge it and a closed class of persons who do have that procedural possibility (see my
observations on the Sofrimport [1990] ECR I-2477 at paras 84ff, above). Furthermore, the gravity of the
impact which a measure has or may have on a person may bring about a situation which distinguishes
that person from all others, as was held to be the case in Extramet and Codorniu72.
72
Nevertheless, it cannot of course be maintained that the solution here advocated may be directly drawn from dicta in the
courts existing case law. Those dicta are merely indicative of the interpretative possibilities open to the Community
judicature in the context of the application of the fourth paragraph of art 173 of the Treaty.

107.
The interpretative approach I am advocating constitutes, in my view, the appropriate way, in cases
such as this, in which to particularise the procedural condition that the contested act must concern the
applicant individually. In my opinion, it cannot be maintained that this approach no longer seeks to
differentiate the applicant but to establish whether the applicant is affected personally by the act 73. The
applicants individual relationship to the contested act continues to be the decisive criterion also under the
solution now advanced. Admittedly, on this view of the matter, the requirement of differentiation is
assimilated to that of the existence of an individual legal interest and is perhaps identical with it, albeit that
the Community judicature has hitherto not been accustomed to that assimilation. I do not believe,
however, that, in particular in cases such as the present one, to approximate the courts reasoning on the
individual nature of the effect on the applicant to the reasoning followed by the majority of national courts
in determining the legal interest of the individual, runs counter to the formulation and spirit of the fourth
paragraph of art 173 of the Treaty or to the underlying rationale of the courts existing case law, even if it
does not constitute a linear continuation of that case law.
73
On that point I refer to the relevant arguments of the Commission at para 33, above.

108.
Moreover, I do not believe that the above class of persons having locus standi which will emerge from
the application of the interpretative approach advocated is not sufficiently closed and defined, especially
if it comprises persons who already enjoy, before the entry into force of the contested act, the
environmental interests which that act is likely to affect (that criterion is also to be 660 found in Sofrimport,
Codorniu and Extramet). Therefore, protection should be afforded to natural persons who had previously
secured, perhaps even over a long period of time, a quality of life which is likely to be particularly severely
affected by the act of the Community institution. It is worth remembering that the court, in its case law to
date, appears to accept the locus standi of non-privileged parties in cases where they invoke the
protection of an acquired right. At least, that interpretation may be inferred from the judgments in Bock,
Piraiki-Patraiki and Sofrimport. I consider therefore that the situation of a natural person who was already
enjoying environmental protection of a certain quality, before the alteration likely to be brought about by
the contested act of the Community institution, is analogous to the situation of the applicants in the above
cases and is equally deserving of judicial protection.
109.
Again, the adequacy of the definition of the closed class of natural persons with locus standi depends
on the criteria which the Community judicature will apply. I have already said that, in my view, the severity
with which the quality of life of the applicant, or any other interest connected with the environment, is likely
to be affected will be particularly considerable so as to bring him within the class of persons having locus
standi, regard being had to the nature of the environmental intervention and the applicants situation.
However, the criterion cannot be a mathematical one. A Community act concerning construction works,
such as an electricity-generating station, irrespective of whether it is built on an island with many
inhabitants, such as Gran Canaria or Tenerife, or on an island with very few, affects them either generally
and in an abstract mannerin which case no question of locus standi arisesor individually, irrespective
of their number. The criteria by which it is to be judged whether an applicant has locus standi remain the
same.
661
110.
I now come to the application of the above observations to the present case. The natural persons who
brought proceedings before the Court of First Instance are not relying on precisely the same legal and
factual situations. Certain of them say that they reside in areas which are a short distance from the works
in question, others that they are owners of real estate in those areas, and others that they carry on some
occupational activity there. Others refer to the negative impact which the construction works in question
will have on the health of inhabitants, on tourism, on fishing, on farming, education of the young, local
flora and fauna or occupations in connection with windsurfing. Finally, some of the parties cite health
problems associated with the impact on the environment of the construction works.
111.
The above submissions and the factual evidence relating to them were taken into consideration by the
Court of First Instance, which examined them at paras 35 and 36 of the order appealed against and
concluded that the action should be dismissed. The legal classification of the above submissions and
evidence falls within the purview of appellate review. However, contrary to the appellants assertions, and
again in light of the restrictive interpretation, as set out above, of the fourth paragraph of art 173 of the
Treaty, I do not consider that the Court of First Instance erred on this point. It did not have before it
sufficient evidence for it to accept that the probable impingement on the environment individually affects
the natural persons belonging to the closed class and, specifically, that the applicants before it are
affected, by virtue of the legal and factual situation peculiar to them, in such a way as to differentiate them
from all other persons. Thus, the Court of First Instancecorrectlydid not deduce from the matters
relied on by the applicants that the latter were differentiated in regard to the contested act and thus had
locus standi to bring an action under art 173.
112.
In particular, certain of the applicants vaguely74 relied on the fact that they reside very close to the
units under construction, without, however, its being possible to deduce from that unsupported
submission that their situation differs from that of other persons. Nor is it possible to discern the reason
why arable production, fishing, tourism or other activities are affected by the construction works in
question, and to what extent, or whether the likely impact on such activities especially affects the
applicants with particular severity, so as to justify granting them locus standi. Equally general and abstract
is the invocation of harm to health likely to be occasioned by the carrying out of the works in question.
74
With the sole exception of one appellant who, without giving further particulars, merely stated that she is the owner of
real estate ten kilometres away from the construction works in question.

113.
Subject to those essential clarifications, I consider that the view taken by the Court of First Instance in
para 54 of the order appealed against is correct 75.
75
Nevertheless, there are also arguments in favour of precisely the opposite viewpoint, which have as the major premiss
of their legal reasoning the interpretative approach which I advocate above. In fact, in the present case, the Commission
is under a clear and specific obligation to monitor whether the works financed are being carried out in accordance with
Directive 85/337. Under the applicable legislation, commencement and continuation of the construction of the electricity
generating stations, such as the ones in question, depend upon the prior obtaining of an environmental impact
assessment and, hence, it may be presumed that those works may have an unfavourable impact on the environment.
The importance attached by both Community and national law to that assessment should indeed be emphasised.
Whether those works can be implemented and under what specific terms and conditions will in the end be determined
by that assessment. Prior to completion of the assessment, it is not possible accurately to determine the impact which
the works in question will have on the environment or, a fortiori, whether those works should be carried out and, by
logical extension, whether they should be financed by Community funds. Accordingly, it is perhaps excessively strict for
individuals desirous of challenging a Commission decision continuing financing of the works in question to be required
fully to prove the damage they are likely to sustain as a result of those works, since, precisely because there is no
environmental impact assessment, the consequences of the works in question for the environment remain essentially
unknown. A special rule should thus be implied that, when there is no environmental impact assessment for a given set
of works, persons challenging the Commission decision granting financial aid for the carrying out of those works, when
required to show why the contested act affects them individually in accordance with the provisions of the fourth
paragraph of art 173 of the Treaty, should not be obliged to provide full and concrete proof of the effects on their
individual situations owing to the actual or potential environmental impact of the works financed. It may then suffice for
them to invoke their status as residents of the wider area in which the works in question are being carried out or as
persons carrying on occupations in that area. Under that interpretation, the matters relied on by the appellants in the
present case were in principle sufficient for them to be deemed to be affected individually by the contested act, in which
case the order of the Court of First Instance is wrong in law and should be set aside.

(c) Locus standi of the appellant environmental associations

114.
As stated above, the Court of First Instance held that the environmental associations which had
brought proceedings before it, that is to say Greenpeace, TEA and CIC, are not affected individually by
the contested act within the meaning of the fourth paragraph of art 173 of the Treaty and thus do not have
locus standi. In particular, according to the reasoning of the Court of First Instance, the environmental
associations did not have that procedural possibility because, first, its 662 members could not
independently bring annulment proceedings and, secondly, those associations did not play any special
role in the procedure which led to the adoption of the contested decision such as to give them locus
standi under the courts existing case law (see the judgments in Van der Kooy and CIRFS).
115.
 The appellant associations do not call in question the second limb of the lower courts reasoning 76. I
will dwell no longer on this issue. Again, in my view correctly, the Court of First Instance followed the
settled case law of the Court of Justice which makes the locus standi of an association dependent on that
of its members77. In light of that case law and of the fact that the locus standi of the appellant individuals
in the present case must be precluded in accordance with the foregoing analysis, the Court of First
Instance, I think correctly, dismissed the action brought before it by the environmental associations. At
any rate, that is the solution dictated by the existing case law of the Court of Justice.
76
Neither the complaints to the Commission nor the communication which those environmental associations had with it
may be said to amount to participation by them in a special Community procedure thus differentiating them in relation to
the contested act.
77
See eg judgments in Fdration nationale de la boucherie en gros et du commerce en gros des viandes [1962] ECR
491, Union syndicale-Service public europen v EC Council Case 72/74 [1975] ECR 401, Comit de dveloppement et
de promotion du textile et de lhabillement (DEFI) v EC Commission Case 282/85 [1986] ECR 2469 and EEC Seed
Crushers and Oil Producers Federation (Fediol) v EC Commission Case 191/82 [1983] ECR 2913.

116.
None the less, the appellant associations argue that, in general, legal persons which represent the
interests of a group of persons should be recognised as having the right to bring proceedings against a
Community act where one or more of their members would have locus standi to do so or where that legal
person can show the existence of a specific right or interest. As regards the second alternative basis of
locus standi which they put forwardwhich, it is to be noted, finds no support in existing case lawthe
appellant associations make the following observations: in their view, the court ought to recognise the
existence of a specific interest on the part of environmental associations to bring annulment proceedings
against acts which jeopardise environmental protection, even where their members or other natural
persons cannot show themselves to be individually affected by such an act. In that way, it is contended on
behalf of the appellants, it is possible to ensure adequate judicial protection in favour of natural persons
who, though affected by a Community act having negative consequences on the environment, do not 663
comprise a closed class of persons, in accordance with the courts case law cited above, and thus cannot
directly challenge such an act78.
78
The appellant associations also stress that to confer on them locus standi would enable judicial review of Community
measures concerning the environment to be exercised in a more consistent and coherent manner.

117.
The court is thus invited to embark on a further extension of its case law. For my part, I consider it
desirable to set out the following observations. First, I believe that a relaxation by the court, to the extent
sought, of the criteria of admissibility could be abused and lead to aberrant consequences. Natural
persons without locus standi under the fourth paragraph of art 173 of the Treaty could circumvent that
procedural impediment by setting up an environmental association. Moreover, whilst the number of
natural persons, that is to say citizens of the European Union, however high it may be, none the less
remains limited, the number of environmental associations capable of being created is, at least in theory,
infinite. But, even if that obstacle could be overcome, for example, by conferring locus standi only on
associations constituted prior to adoption of the contested measure, account would have to be taken of
the fact that, within the European Union, the number of legal persons which have as their object the
protection and conservation of the environment is today particularly high. If the court were ultimately to
follow the proposal of the appellant associations, in future every measure of a Community institution
concerning the environment or having an impact on it could be expected, on each occasion, to form the
subject matter of proceedings brought by a plethora of environmental associations.
118.
For those reasons and notwithstanding the recent developments in national and international law, I
continue to have significant reservations concerning the expediency of the departure from existing case
law sought by the appellant associations. Moreoverand this I think is the decisive argumentto give to
environmental associations the possibility of challenging Community measures concerning the
environment would, notwithstanding the positive consequences which might well flow from this, run
counter to the letter of the law, in this case of art 173 of the Treaty. The Community legislature has divided
applicants into two categories, namely those mentioned in the second and those mentioned in the fourth
paragraphs of art 173. To accept the interpretative approach advocated by the appellant associations
would be tantamount to creating a third category extra legem. In other words, between the privileged
parties mentioned in the second para, who are not required to invoke any legal interest, and the parties
mentioned in the fourth paragraph, who must be directly and individually affected by the contested
measure, there would be interposed the environmental association, which would be presumed to have
locus standi whenever the contested measure concerns the environment or any impact on it. In
conclusion, I believe that the departure from case law advocated, irrespective of whether or not it is
desirable, remains unfeasible79 on the basis of existing written law.
79
Moreover, I do not see why locus standi should be granted in that way as a privilege only to environmental associations
and not to other legal persons of a representative nature. Invocation of the special nature of environmental protection,
extensively canvassed above, cannot, in my view, justify different treatment of environmental associations in relation to
associations which have other, analogous objects.

119.
From all the foregoing, I conclude that the order of the Court of First Instance appealed against is
unimpeachable and there can be no question of its being 664 set aside. Moreover, there is no need to
examine the remaining objections of inadmissibility raised by the Commission and Spain, that is to say
whether a decision to continue financing infrastructure works can directly affect the rights or interests of
the appellants or whether that measure, by its very nature, can constitute the subject matter of
proceedings under the fourth paragraph of art 173 of the Treaty.

VIICONCLUSION
120.
In light of the foregoing I propose that the Court of Justice should: dismiss the appeal in its entirety,
and order the appellants to pay the costs.

2 April 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By application lodged at the registry of the Court of Justice of the European Communities on 16
October 1995, Stichting Greenpeace Council (Greenpeace International) and others brought an appeal
under art 49 of the EC Statute of the Court of Justice against the order of the Court of First Instance
(Case T-585/93 [1995] ECR II-2205) in so far as it declared inadmissible their action for annulment of the
Commissions decision allegedly taken between 7 March 1991 and 29 October 1993 to disburse to the
Kingdom of Spain ECU 12m, or other amounts of that order, pursuant to Decision C(91)440 concerning
financial assistance provided by the European Regional Development Fund for the construction of two
power stations in the Canary Islands (Gran Canaria and Tenerife).
2.
According to the contested order, the factual background to the dispute is as follows ([1995] ECR II-
2205 (paras 112)):
(1) On 7 March 1991, on the basis of Council Regulation (EEC) No 1787/84 of 19 June 1984 on
the European Regional Development Fund (OJ 1984 L 169, p. 1, the basic regulation), as
amended by Council Regulation (EEC) No 3641/85 of 20 December 1985 (OJ 1985 L 350, p. 40),
the Commission adopted Decision C(91)440 granting the Kingdom of Spain financial assistance
from the European Regional Development Fund (the ERDF) up to a maximum of ECU 108 578
419, for infrastructure investment. The project concerned was for the building of two power stations
in the Canary Islands, on Gran Canaria and on Tenerife, by Unin Elctrica de Canarias SA
(Unelco).
(2) The Community finance for the construction of the two power stations was to be spread over
four years, from 1991 to 1994, and to be paid in yearly tranches (Articles 1 and 3 of, and Annexes II
and III to, the decision). The financial commitment for the first year (1991), for ECU 28 953 000
(Article 1 of the decision), was payable on the defendants adoption of the decision (Annex III,
paragraph A4, of the decision). Subsequent disbursements, based on the financial plan for the
operation and on the progress of its implementation, were to cover expenditure relating to the
operations in question, legally approved in the member state concerned (Articles 1 and 3 of the
decision). Under Article 5 of the decision, the Commission could reduce or suspend the aid granted
to the operation in issue if an examination were to reveal an irregularity and in particular a
significant change affecting the way in which it was carried out for which the Commissions
approval had not been requested (see also paragraphs A20, A21 and C2 of Annex III to the
decision).
665
(3) By letter dated 23 December 1991, Aurora Gonzlez Gonzlez and Pedro Melin Castro, the
fifth and sixth applicants, informed the Commission that the works carried out on Gran Canaria
were unlawful because Unelco had failed to undertake an environmental impact assessment study
in accordance with Council Directive 85/337/EEC of 27 June 1985 on the assessment of the effects
of certain public and private projects on the environment (OJ 1985 L 175, p. 40) and asked it to
intervene to stop the works. Their letter was registered as No 4084/92.
(4) By letter dated 23 November 1992, Domingo Viera Gonzlez, the second applicant, sought
the Commissions assistance on the ground that Unelco had already started work on Gran Canaria
and Tenerife without the Comisin de Urbanismo y Medio Ambiente de Canarias (Canary Islands
Commission for Planning and the Environment, Cumac) having issued its declaration of
environmental impact in accordance with the applicable national legislation. That letter was
registered as No 5151/92.
(5) On 3 December 1992, Cumac issued two declarations of environmental impact relating to
the construction of the power stations on Gran Canaria and Tenerife, published in the Boletn
Oficial de Canarias on 26 February and 3 March 1993 respectively.
(6) On 26 March 1993, Tagoror Ecologista Alternativo (TEA), the 18th applicant, a local
environmental protection association based on Tenerife, lodged an administrative appeal against
Cumacs declaration of environmental impact relating to the project for the construction of a power
station on Tenerife. On 2 April 1993, the Comisin Canaria contra la Contaminacin (Canary
Islands Commission against Pollution, hereinafter CIC), the 19th applicant, a local environmental
protection association based on Gran Canaria, also brought administrative proceedings against
Cumacs declaration of environmental impact relating to the two construction projects on Gran
Canaria and Tenerife.
 (7) On 18 December 1993, Greenpeace Spain, an environmental protection association
responsible at the national level for the achievement at local level of the objectives of Stichting
Greenpeace Council (Greenpeace), the first applicant, a nature conservancy foundation having its
head office in the Netherlands, brought legal proceedings challenging the validity of the
administrative authorisations issued to Unelco by the Canary Island Regional Ministry of Industry,
Commerce and Consumption.
(8) By letter of 17 March 1993 addressed to the Director-General of the Commissions
Directorate-General for Regional Policies (DG XVI), Greenpeace asked the Commission to
confirm whether Community structural funds had been paid to the Regional government of the
Canary Islands for the construction of two power stations and to inform it of the timetable for the
release of those funds.
(9) By letter of 13 April 1993, the Director-General of DG XVI recommended that Greenpeace
read the Decision C(91)440 which, he said, contained details of the specific conditions to be
respected by Unelco in order to obtain Community support and the financing plan.
(10) By letter of 17 May 1993, Greenpeace asked the Commission for full disclosure of all
information relating to measures it had taken with regard to the construction of the two power
stations in the Canary Islands, in accordance with Article 7 of Council Regulation (EEC) No
2052/88 of 24 June 1988 on the tasks of the Structural Funds and their effectiveness and on
coordination of their 666 activities between themselves and with the operations of the European
Investment Bank and the other existing financial instruments (OJ 1988 L185 p 9), which provides:
Measures financed by the Funds or receiving assistance from the EIB or from another existing
financial instrument shall be in keeping with the provisions of the Treaties, with the instruments
adopted pursuant thereto and with Community policies, including those concerning
environmental protection.
(11) By letter dated 23 June 1993, the Director-General of DG XVI wrote as follows to
Greenpeace: I regret to say that I am unable to supply this information since it concerns the
internal decision making procedures of the Commission but I can assure you that the
Commissions decision was taken only after full consultation between the various services
concerned.
(12) On 29 October 1993 a meeting took place at the Commissions premises in Brussels
between Greenpeace and DG XVI, concerning the financing by the ERDF of the construction of the
power stations on Gran Canaria and Tenerife.
3.
It was against that background that, by application lodged at the registry of the Court of First Instance
on 21 December 1993, the applicants brought an action seeking annulment of the decision allegedly
taken by the Commission to disburse to the Spanish government, in addition to the first tranche of ECU
28,953,000, a further ECU 12m in reimbursement of expenses incurred in the construction of two power
stations in the Canary Islands (Gran Canaria and Tenerife). That decision was allegedly taken between 7
March 1991, when Decision C(91)440 was adopted, and 29 October 1993, when the Commission, at the
above mentioned meeting with Greenpeace, whilst refusing to provide Greenpeace with detailed
information regarding the financing of the construction of the two power stations in the Canary Islands,
confirmed that a total of ECU 40m had already been disbursed to the Spanish government pursuant to
Decision C(91)440.
4.
By separate document lodged at the registry of the Court of First Instance on 22 February 1994, the
Commission raised an objection of inadmissibility in support of which it raised two pleas, one concerning
the nature of the contested decision and the other the applicants lack of locus standi.
5.
By the contested order, the Court of First Instance upheld the objection and declared the action
inadmissible.
6.
As to the pleas raised by the Commission in support of its objection of inadmissibility, the Court of First
Instance stated that it was first necessary to examine whether the applicants had locus standi to bring an
action, before considering whether the act which they were challenging constituted a decision within the
meaning of art 173 of the EC Treaty (see [1995] ECR II-2205 (para 46)).
7.
As regards, first, the locus standi of the applicants who are private individuals, the Court of First
Instance ([1995] ECR II-2205 (para 48)) referred first to the settled case law of the Court of Justice
according to which persons other than the addressees may claim that a decision is of direct concern to
them only if that decision affects them by reason of certain attributes which are peculiar to them, or by
reason of factual circumstances which differentiate them from all other persons and thereby distinguish
them individually in the same way as the person addressed (see the judgments in Plaumann & Co v EEC
Commission Case 25/62 [1963] ECR 95, Spijker Kwasten BV v EC Commission Case 231/82 [1983] ECR
2559, Union Deutsche Lebensmittelwerke GmbH v EC Commission Case 97/85 [1987] ECR 2265,
William Cook plc v EC Commission Case C-198/91 [1993] ECR I-2487, Matra SA v EC Commission Case
C-225/91 [1993] ECR I-3203, Air France (Socit Anonyme Participation Ouvrire 667 Cie) v European
Commission Case T-2/93 [1994] ECR II-323 and Consorzio Gruppo di Azione Locale Murgia Messapica
v EC Commission Case T-465/93 [1994] ECR II-361).
8.
 The Court of First Instance then decided to examine the applicants argument that the court should not
be constrained by the limits imposed by that case law and should concentrate on the sole fact that third-
party applicants had suffered or would potentially suffer loss or detriment from the harmful environmental
effects arising out of unlawful conduct on the part of the Community institutions ([1995] ECR II-2205 (para
49)).
9.
In that regard, the Court of First Instance held that whilst the settled case law of the Court of Justice
concerned essentially cases involving economic interests, the essential criterion which it applied (namely,
a combination of circumstances sufficient for the third-party applicant to be able to claim to be affected by
the contested decision in a manner which differentiated him from all other persons) remained applicable
whatever the nature, economic or otherwise, of the applicants interests which were affected (see [1995]
ECR II-2205 (para 50)).
10.
The Court of First Instance accordingly held ([1995] ECR II-2205 (para 51)) that the criterion proposed
by the applicants for appraising their locus standi, namely the existence of harm suffered or to be
suffered, was not in itself sufficient to confer locus standi on an applicant; this was because such harm
might affect, in a general abstract way, a large number of persons who could not be determined in
advance in such a way as to distinguish them individually just like the addressee of a decision, as
required under the case law cited above. The Court of First Instance added that, in view of the conditions
laid down in the fourth paragraph of art 173 of the Treaty, that conclusion could not be affected by the
practice of national courts whereby locus standi might depend merely on applicants having a sufficient
interest.
11.
The Court of First Instance therefore concluded ([1995] ECR II-2205 (para 52)) that the applicants
argument that the question of their locus standi in this case should be determined in the light of criteria
other than those already laid down in the case law could not be accepted, and went on to hold that their
locus standi had to be assessed in the light of the criteria laid down in that case law (see para 53).
12.
In this regard, the Court of First Instance stated first of all that the objective status of local resident,
fisherman or farmer or of persons concerned by the impact which the building of two power stations
might have on local tourism, on the health of Canary Island residents and on the environment, relied on
by the applicants, did not differ from that of all the people living or pursuing an activity in the areas
concerned and that the applicants thus could not be affected by the contested decision otherwise than in
the same manner as any other local resident, fisherman, farmer or tourist who was, or might be in the
future, in the same situation (see paras 5455).
13.
Finally, the Court of First Instance held ([1995] ECR II-2205 (para 56)) that the fact that certain of the
applicants had submitted a complaint to the Commission could not confer locus standi under art 173 of
the Treaty, since no specific procedures were provided for whereby individuals might be associated with
the adoption, implementation and monitoring of decisions taken in the field of financial assistance granted
by the ERDF. The Court of Justice had held that, although a person who asked an institution not to take a
decision in respect of him, but to open an inquiry with regard to third parties, might be considered to have
an indirect interest, such a person was nevertheless not in the precise legal position of the actual or
potential addressee of a measure which might be annulled under art 173 of the Treaty (see William v EC
Commission Case 246/81 [1982] ECR 2277).
668
14.
Second, as regards the locus standi of the applicant associations, the Court of First Instance recalled
([1995] ECR II-2205 (para 59)) that it had consistently been held that an association formed for the
protection of the collective interests of a category of persons could not be considered to be directly and
individually concerned, for the purposes of the fourth paragraph of art 173 of the Treaty, by a measure
affecting the general interests of that category, and was therefore not entitled to bring an action for
annulment where its members could not do so individually (see Fdration nationale de la boucherie en
gros et du commerce en gros des viandes v EEC Council Joined cases 1922/62 [1962] ECR 491, Union
syndicale-Service public europen v EC Council Case 72/74 [1975] ECR 401, Fdration Nationale des
Producteurs de Vins de Table et Vins de Pays v EC Commission Case 60/79 [1979] ECR 2429, Comit
de dveloppement et de promotion du textile et de lhabillement (DEFI) v EC Commission Case 282/85
[1986] ECR 2469, Union de Federaciones Agrarias de Espaa (UFADE) v EC Council Case 117/86
[1986] ECR 3255 (para 12) and Associazione Tecnico Economica del Cemento v European Commission
Joined cases T-447449/93 [1995] ECR II-1971 (paras 5859)). Since the Court of First Instance had
held that the applicants who were private individuals could not be considered to be individually concerned
by the contested decision, it therefore concluded that the members of the applicant associations, as local
residents of Gran Canaria and Tenerife, likewise could not be considered to be individually concerned
(see [1995] ECR II-2205 (para 60)).
15.
The Court of First Instance went on to observe ([1995] ECR II-2205 (para 59)) that special
circumstances, such as the role played by an association in a procedure which led to the adoption of an
act within the meaning of art 173 of the Treaty, might justify treating as admissible an action brought by an
association whose members were not directly and individually concerned by the contested measure (see
Kwekerij Gebroeders Van der Kooy BV v EC Commission Joined cases 67/85, 68/85 and 70/85 [1988]
ECR 219 and Comit International de la Rayonne et des Fibres Synthtiques (CIRFS) v EC Commission
Case C-313/90 [1993] ECR I-1125).
16.
However, the Court of First Instance came to the conclusion ([1995] ECR II-2205 (para 62)) that the
exchange of correspondence and the discussions which Greenpeace had with the Commission
concerning the financing of the project for the construction of two power stations in the Canary Islands did
not constitute special circumstances of that kind since the Commission did not, prior to the adoption of the
contested decision, initiate any procedure in which Greenpeace participated. Nor was Greenpeace in any
way the interlocutor of the Commission with regard to the adoption of the basic Decision C(91)440 and/or
of the contested decision.
17.
In their appeal the appellants submit that, in determining whether they were individually concerned by
the contested decision within the meaning of art 173 of the Treaty, the Court of First Instance erred in its
interpretation and application of that provision and that, by applying the case law developed by the Court
of Justice in relation to economic issues and economic rights, according to which an individual must
belong to a closed class in order to be individually concerned by a Community act, the Court of First
Instance failed to take account of the nature and specific character of the environmental interests
underpinning their action.
18.
In particular, the appellants argue, first, that the approach adopted by the Court of First Instance
creates a legal vacuum in ensuring compliance with Community environmental legislation, since in this
area the interests are, by their very nature, common and shared, and the rights relating to those interests
are liable to be held by a potentially large number of individuals so that there could never be 669 a closed
class of applicants satisfying the criteria adopted by the Court of First Instance.
19.
Nor can that legal vacuum be filled by the possibility of bringing proceedings before the national
courts. According to the appellants, such proceedings have in fact been brought in the present case, but
they concern the Spanish authorities failure to comply with their obligations under Council Directive
(EEC) 85/337 on the assessment of the effects of certain public and private projects on the environment
(OJ 1985 L175 p 40), and not the legality of the Commission measure, that is to say the lawfulness under
Community law of the Commissions disbursement of structural funds on the ground that that
disbursement is in violation of an obligation for protecting the environment.
20.
Second, the appellants submit that the Court of First Instance was wrong to take the view, at para 51
of the contested order, that reference to national laws on locus standi was irrelevant for the purposes of
art 173. The solution adopted by the Court of First Instance appears to conflict with that required by
national judicial decisions and legislation as well as by developments in international law. According to the
appellants, it is clear from the Final Report on Access to Justice (1992), prepared by the KO-Institut for
the Commission, which describes the position concerning locus standi on environmental issues, that, if
they had been required to bring proceedings before a court of a member state, actions brought by some
or all of the applicants would have been declared admissible. The appellants add that the above
mentioned developments have been influenced by American law, the United States Supreme Court
holding in Sierra Club v Morton 405 US 727, (1972) 31 Led 2d 636 at 643, that
Aesthetic and environmental well-being, like economic well-being, are important ingredients of
the quality of life in our society, and the fact that particular environmental interests are shared by
the many rather than the few does not make them less deserving of legal protection through the
judicial process.
21.
Third, the appellants submit that the approach adopted by the Court of First Instance in the contested
order is at odds with both the case law of the Court of Justice and declarations of the Community
institutions and governments of the member states on environmental matters. As regards case law, they
rely on the holding that environmental protection is one of the Communitys essential objectives (see the
judgments in Procureur de la Rpublique v Association de dfense des brleurs dhuiles usages Case
240/83 [1985] ECR 531 (para 13) and EC Commission v Denmark Case 302/86 [1988] ECR 4607 (para
8)) and submit that Community environmental legislation can create rights and obligations for individuals
(see the judgments in EC Commission v Germany Case C-131/88 [1991] ECR I-825 (para 7) and EC
Commission v Germany Case C-361/88 [1991] ECR I-2567 (paras 1516)). Furthermore, in the present
case, the appellants submit that their arguments relating to individual concern are based essentially on
their individual rights conferred by Directive 85/337, arts 6(2) and 8 of which provide for participation in
the environmental impact assessment procedure in relation to certain projects, and that they are singled
out by virtue of those rights which are recognised and protected in Commission Decision C(91)440 (see
the judgment in EC Commission v Germany Case C-431/92 [1995] ECR I-2189 (paras 3740)).
22.
The appellants go on to refer to the Fifth Environmental Action Programme (OJ 1993 C138, p 1), to
Principle 10 of the Rio Declaration on Environment and Development, ratified by the Community at the
1992 United Nations Conference on 670 Environment and Development, to Agenda 21, adopted at the
same conference, to the Council of Europe Convention on Civil Liability for Damage resulting from
Activities Dangerous to the Environment (Lugano, 21 June 1993; ETS No 150), and to the system of
administrative review introduced by the World Bank to allow review of its acts where they have negative
effects on the environment (see World Bank, Resolution No 93-10, Resolution No IDA93-6, 22 September
1993, para 12).
23.
Fourth, the appellants propose a different interpretation of the fourth paragraph of art 173 of the Treaty.
In order to determine whether a particular applicant is individually concerned by a Community act
involving violations of Community environmental obligations, that applicant should be required to
demonstrate that: (a) he has personally suffered (or is likely personally to suffer) some actual or
threatened detriment as a result of the allegedly illegal conduct of the Community institution concerned,
such as a violation of his environmental rights or interference with his environmental interests; (b) the
detriment can be traced to the act challenged; and (c) the detriment is capable of being redressed by a
favourable judgment.
24.
The appellants contend that they satisfy those three conditions. As regards the first condition, they
state that they submitted statements describing the detriment which they have suffered as a result of the
Commissions acts. As regards the second condition, they point out that, by disbursing to the Kingdom of
Spain the funds granted under Decision C(91)440 for the construction of projects carried out in breach of
Community environmental law, the Commission directly contributed to the detriment caused to their
interests since the Spanish authorities had no discretion as to the use to which those funds were to be
put. As regards the third condition, the appellants consider that, if the Court of First Instance had annulled
the contested decision, the Commission would not have continued to finance work on construction of the
power stations which would then have probably been suspended until completion of the environmental
impact procedure.
25.
The appellants submit further that environmental associations should be recognised as having locus
standi where their objectives concern chiefly environmental protection and one or more of their members
are individually concerned by the contested Community decision, but also where, independently, their
primary objective is environmental protection and they can demonstrate a specific interest in the question
at issue.
26.
Referring to the judgment in Plaumann, the appellants conclude that art 173 must not be interpreted
restrictively; its wording does not expressly require an approach based on the idea of a closed class, as
affirmed in the case law of the Court of Justice and the Court of First Instance (see the judgments in
Piraiki-Patraiki v EC Commission Case 11/82 [1985] ECR 207, Extramet Industrie SA v EC Council Case
C-358/89 [1991] ECR I-2501, Codorniu SA v EU Council Case C-309/89 [1994] ECR I-1853 and Antillean
Rice Mills NV v EC Commission Joined cases T-480/93 and T-483/93 [1995] ECR II-2305). Rather, it
must be interpreted in such a way as to safeguard fundamental environmental interests and protect
individual environmental rights effectively (see the judgments in Procureur de la Rpublique [1985] ECR
531 (para 13), Johnston v Chief Constable of the Royal Ulster Constabulary Case C-222/84 [1986] 3 All
ER 135, [1986] ECR 1651 (paras 1321) and Union nationale des entraneurs et Cadres techniques
professionnels du football (Unectef) v Heylens Case 222/86 [1987] ECR 4097 (para 14)).
671

Findings of the court

27.
The interpretation of the fourth paragraph of art 173 of the Treaty that the Court of First Instance
applied in concluding that the appellants did not have locus standi is consonant with the settled case law
of the Court of Justice.
28.
As far as natural persons are concerned, it follows from the case law, cited at both para 48 of the
contested order and at para 7 of this judgment, that where, as in the present case, the specific situation of
the applicant was not taken into consideration in the adoption of the act, which concerns him in a general
and abstract fashion and, in fact, like any other person in the same situation, the applicant is not
individually concerned by the act.
29.
The same applies to associations which claim to have locus standi on the basis of the fact that the
persons whom they represent are individually concerned by the contested decision. For the reasons given
in the preceding para, that is not the case.
30.
In appraising the appellants arguments purporting to demonstrate that the case law of the Court of
Justice, as applied by the Court of First Instance, takes no account of the nature and specific
characteristics of the environmental interests underpinning their action, it should be emphasised that it is
the decision to build the two power stations in question which is liable to affect the environmental rights
arising under Directive 85/337 that the appellants seek to invoke.
31.
In those circumstances, the contested decision, which concerns the Community financing of those
power stations, can affect those rights only indirectly.
32.
As regards the appellants argument that application of the courts case law would mean that, in the
present case, the rights which they derive from Directive 85/337 would have no effective judicial
protection at all, it must be noted that, as is clear from the file, Greenpeace brought proceedings before
the national courts challenging the administrative authorisations issued to Unelco concerning the
construction of those power stations. TEA and CIC also lodged appeals against CUMACs declaration of
environmental impact relating to the two construction projects (see [1995] ECR II-2205 (paras 67)).
33.
Although the subject matter of those proceedings and of the action brought before the Court of First
Instance is different, both actions are based on the same rights afforded to individuals by Directive
85/337, so that in the circumstances of the present case those rights are fully protected by the national
courts which may, if need be, refer a question to this court for a preliminary ruling under art 177 of the
Treaty.
34.
The Court of First Instance did not therefore err in law in determining the question of the appellants
locus standi in the light of the criteria developed by the Court of Justice in the case law set out at para 7 of
this judgment.
35.
In those circumstances the appeal must be dismissed.

Costs
36.
Under art 69(2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if
they have been asked for in the successful partys pleadings. Since the appellants have been
unsuccessful in their appeal, they must be ordered to pay the costs. Under art 69(4) of the Rules of
Procedure, the Kingdom of Spain, which intervened in these proceedings, is to bear its own costs.
On those grounds, the Court of Justice hereby: (1) dismisses the appeal; (2) orders the appellants to
pay the costs; (3) orders the Kingdom of Spain to bear its own costs.

672

[1998] All ER (EC) 673

Decker v Caisse de Maladie des Employs Privs (Case C-120/95)

Kohll v Union des Caisses de Maladie (Case C-158/96)

EUROPEAN COMMUNITY; Free movement of goods, Freedom to provide services

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN AND RAGNEMALM (RAPPORTEUR)
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, KAPTEYN, MURRAY, EDWARD,
PUISSOCHET, HIRSCH AND JANN ADVOCATE GENERAL TESAURO
2 JULY 1996, 16 SEPTEMBER 1997, 28 APRIL 1998
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN, RAGNEMALM (RAPPORTEUR) AND
WATHELET (PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, KAPTEYN,
MURRAY, EDWARD, PUISSOCHET, HIRSCH, JANN AND SEVN ADVOCATE GENERAL TESAURO
15 JANUARY, 16 SEPTEMBER 1997, 28 APRIL 1998
European Community Freedom of movement Goods Spectacles National rules making
reimbursement of medical expenses incurred in another member state conditional on prior authorisation
Luxembourg plaintiff purchasing spectacles in Belgium without prior approval Plaintiffs application for
reimbursement refused Whether prior authorisation compatible with Community provisions on free
movement EC Treaty, arts 30, 36.
European Community Freedom of movement Services Dental care National rules making
reimbursement of medical expenses incurred in another member state conditional on prior authorisation
Authorisation for plaintiffs daughter to receive treatment from an orthodontist in Germany refused on
ground that proposed treatment not urgent and could be provided in Luxembourg Whether prior
authorisation compatible with Community provisions on free movement Council Regulation (EEC)
1408/71, art 22 EC Treaty, arts 59, 60.

In the first case, the plaintiff, a Luxembourg national, purchased a pair of spectacles from an optician
established in Belgium, on a prescription from an opthalmologist established in Luxembourg. However, in
accordance with the Luxembourg Code des Assurances Sociales, which provided that treatment abroad
had to be authorised by the competent sickness fund, the defendant insurance fund refused to reimburse
the plaintiff, since the spectacles had been purchased without its prior authorisation. The plaintiff, relying
on art 30 (free movement of goods) of the EC Treaty, contested that decision before the Social Insurance
Arbitration Council, which stayed the proceedings and referred to the Court of Justice of the European
Communities the question whether the relevant measures were compatible with arts 30 1 and 362 of the
Treaty.
1
Article 30, so far as material, provides: Quantitative restrictions on imports and all measures having equivalent effect
shall be prohibited between Member States.
 2
Article 36, so far as material, provides: Articles 30 to 34 shall not preclude prohibitions or restrictions on imports,
exports or goods in transit justified on grounds of the protection of health and life of humans

In the second case, the plaintiffs sickness insurance fund, UCM, rejected a request made in
accordance with the Luxembourg Code des Assurances Sociales 673 by a doctor established in
Luxembourg for authorisation for the plaintiffs daughter, a minor, to receive treatment from an
orthodontist in Germany on the ground that the proposed treatment was not urgent and that it could be
provided in Luxembourg. That decision was confirmed by the UCM board and the plaintiff appealed to the
Social Insurance Arbitration Council, contending that the provisions relied on were contrary to art 59 3 (free
movement of services) of the EC Treaty. His appeal was dismissed and he appealed to the Higher Social
Insurance Council, which upheld the contested decision on the ground that the national provisions relied
on were consistent with art 224 of Council Regulation (EEC) 1408/71 on the application of social security
schemes to employed persons, to self-employed persons and to members of their families moving within
the Community. The plaintiff appealed to the Cour de Cassation, which stayed the proceedings and
referred to the Court of Justice the question whether the relevant measures were compatible with arts 59
and 605 of the Treaty.
3
Article 59, so far as material, provides: restrictions on freedom to provide services within the Community shall be
abolished in respect of nationals of Member States who are established in a State of the Community other than that of
the person for whom the services are intended.
4
Article 22, so far as material, is set out at p 710 h to p 711 b, post
5
Article 60, so far as material, provides: Services shall, in particular include (d) activities of the professions.

Before the Court of Justice the Luxembourg government contended in both cases (i) that the relevant
measures did not fall within the scope of the Treaty since they concerned social security and therefore
should be examined solely from the point of view of art 22 of Regulation 1408/71, and (ii) that any barriers
to free movement presented by the measures were objectively justified under arts 36 and 56 6 since their
purpose was to control health expenditure and to guarantee the protection of public health.
6
Article 56, so far as material, provides: (1) The provisions of this Chapter shall not prejudice the applicability of
provisions providing for special treatment for foreign nationals on grounds of public health.

Held (1) Measures adopted by member states in social security matters which might affect the
marketing of medical products and indirectly influence the possibilities of importing those products were
subject to the Community rules on the free movement of goods; and the special nature of certain services
did not remove them from the ambit of the fundamental principle of freedom of movement. Consequently,
the fact that the rules at issue fell within the sphere of social security could not exclude the application of
arts 30 and 59. Further, the fact that a national measure was consistent with a provision of secondary
legislation, such as art 22 of Regulation 1408/71, did not have the effect of removing that measure from
the scope of the provisions of the Treaty. Moreover, on its true construction, art 22 was not intended to
regulate and therefore did not in any way prevent the reimbursement by member states, at the tariff in
force in the competent state, of costs incurred in connection with treatment provided in another member
state, even without prior authorisation (see p 712 d g to j, p 717 h and p 718 c to f, post); Duphar BV v
Netherlands Case 238/ 67482 [1984] ECR 523 and Criminal proceedings against Webb Case 279/80
[1981] ECR 3305 applied.
(2) Articles 30 and 36 of the Treaty precluded national rules under which a social security institution of
a member state refused to reimburse an insured person, on a flat-rate basis, the cost of a pair of
spectacles with corrective lenses purchased from an optician established in another member state, on the
ground that prior authorisation was required for the purchase of any medical product abroad. Such rules
encouraged insured persons to purchase those products in Luxembourg rather than in other member
states and were liable to curb the import of spectacles assembled in those states and therefore
constituted a barrier to the free movement of goods. Moreover, they could not be justified by purely
economic objectives and it was clear that reimbursement at a flat rate of the cost of spectacles purchased
in other member states had no effect on the financing or balance of the social security system.
Furthermore, the art 36 derogation on grounds of public health was not applicable since the conditions for
taking up and pursuing a profession had been the subject of regulation at Community level, so that the
purchase of spectacles from an optician established in another member state provided guarantees
equivalent to those afforded by a sale on national territory (see p 713 b to d j to p 714 a d to j to p 715 a,
post); EC Commission v France Case 18/84 [1985] ECR 1339 and EC Commission v Germany Case C-
62/90 [1992] ECR I-2575 applied.
(3) Articles 59 and 60 of the Treaty precluded national rules under which reimbursement, in
accordance with the scale of the insuring state, of the cost of dental treatment provided by an orthodontist
established in another member state was subject to authorisation by the insured persons social security
institution, since such rules deterred insured persons from approaching providers of medical services
established in another member state and constituted, for them and their patients, a barrier to freedom to
provide services. Moreover, they were not justified on the grounds of public health, since it had not been
shown that they were necessary to provide a balanced medical and hospital service accessible to all or
that they were indispensable for the maintenance of an essential treatment facility or medical service on
national territory. Nor were they necessary to protect the quality of medical services provided in other
member states, since doctors and dentists established in other members states had to be afforded all
guarantees equivalent to those accorded to doctors and dentists established on national territory, for the
purposes of freedom to provide services (see p 719 b to d j, p 720 f to p 721 e, post) Gl v
Regierungsprsident Dsseldorf Case 131/85 [1986] ECR 1573, Syndesmos ton en Elladi Touristikon kai
Taxidiotikon Grafeion v Ergasias Case C-398/95 [1997] ECR I-3091, Bachmann v Belgium Case C-
204/90 [1992] ECR I-249 and European Commission v France Case C-381/93 [1994] ECR I-5145,
applied.

Notes
For exceptions to the principle of free movement of goods, see 52 Halsburys Laws (4th edn) paras 1298,
12112.
For exceptions to the principle of freedom to provide services, see ibid, para 1623.
For the EC Treaty, arts 30, 36, 56, 59 and 60, see 50 Halsburys Statutues (4th edn) 276, 278, 286,
287, 288.

675

Cases cited
Aragonesa de Publicidad Exterior SA v Departamento de Sanidad y Seguridad de la Generalitat de
Catalua Joined cases C-1/90 and C-176/90 [1991] ECR I-4151.
Bachmann v Belgium Case C-204/90 [1992] ECR I-249.
Barassi v EC Commission Case T-41/90 [1992] ECR II-159.
Belgium v Humbel Case C-263/86 [1988] ECR 5365.
Bestuur van het Algemeen Zieekenfonds Drenthe-Platteland v Pierik Case 117/77 [1978] ECR 825.
Bestuur van het Algemmen Ziekenfonds Drenthe-Platteland v Pierik Case 182/78 [1979] ECR 1977.
Bond van Adverteerders v Netherlands Case 352/85 [1988] ECR 2085.
Borowitz v Bundesversicherungsanstalt fr Angestellte Case 21/87 [1988] ECR 3715.
Campus Oil Ltd v Minister for Industry and Energy Case 72/83 [1984] ECR 2727.
Coenen v Sociaal-Economische Raad Case 39/75 [1975] ECR 1547.
Coonan v Insurance Officer Case 110/79 [1980] ECR 1445.
Customs and Excise Comrs v Schindler Case C-275/92 [1994] 2 All ER 193, [1994] QB 610, [1994] 3
WLR 103, [1994] ECR I-1039, ECJ.
Delattre (Criminal proceedings against) Case C-369/88 [1991] ECR I-1487.
Du Pont de Nemours Italiana SpA v Unit Sanitaria Case C-21/88 [1990] ECR I-889.
Duphar BV v Netherlands Case 238/82 [1984] ECR 523.
EC Commission v Belgium Case 221/85 [1987] ECR 719.
EC Commission v Belgium Case C-249/88 [1991] ECR I-1275.
EC Commission v Belgium Case C-211/91 [1992] ECR I-6757.
EC Commission v France Case 18/84 [1985] ECR 1339.
EC Commission v Germany Case 205/84 [1986] ECR 3755.
EC Commission v Germany Case C-62/90 [1992] ECR I-2575.
EC Commission v Italy Case 7/61 [1961] ECR 317.
EC Commission v Netherlands Case C-353/89 [1991] ECR I-4069.
EC Commission v Spain Case C-45/93 [1994] ECR I-911.
EC Commission v UK Case 124/81 [1983] ECR 203.
EC Commission v Ireland Case 113/80 [1981] ECR 1625.
European Commission v France Case C-381/93 [1994] ECR I-5145.
European Commission v Italy Case C-101/94 [1996] ECR I-2691.
Federacin de Distribuidores Cinematogrficos v Estado Espaol Case C-17/92 [1993] ECR I-2239.
Fdration Franaise Socits dAssurances v Ministre de lAgriculture et de la Pche Case C-244/94
[1995] ECR I-4013.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Futura Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471.
Garcia v La Mutuelle de Prvoyance Sociale Aquitaine Case C-238/94 [1996] ECR I-1673.
Gebhard v Consiglio dellOrdine degli Avvocati e Procuratori di Milano Case C-55/94 [1996] All ER (EC)
189, [1995] ECR I-4165, ECJ.
Gravina v Landesversicherungsanstalt Schwaben Case 807/79 [1980] ECR 2205.
Gl v Regierungsprsident Dsseldorf Case 131/85 [1986] ECR 1573.
Hoekstra (ne Unger) v Bestuur der Bedrijfsvereniging voor Detailhandel en Ambachten Case 75/63
[1964] ECR 177.
Hfner v Macrotron GmbH Case C-41/90 [1991] ECR I-1979.
Hubbard v Hamburger Case C-20/92 [1993] ECR I-3777.
676
Jordens-Vosters v Bestuur van de Bedrijfsvereniging voor de Leder- en Lederwerkende Industrie Case
69/79 [1980] ECR 75.
Laboratoire de Prothses Oculaires (LPO) v Union Nationale des Syndicats dOpticiens de France Case
C-271/92 [1993] ECR I-2899.
Lepore v Office National de Pensions (ONP) Joined cases C-4546/92 [1993] ECR I-6497.
Luisi v Ministero del Tesoro Joined cases 286/82 and 26/83 [1984] ECR 377.
Monteil (Criminal proceedings against) Case C-60/89 [1991] ECR I-1547.
Mnchen v Luxembourg (Case No 28895/95) (16 April 1996, unreported), E Com HR.
Paraschi v Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501.
Pinna v Caisse dAllocations Familiales de la Savoie Case 41/84 [1986] ECR 1.
Pistre (Criminal proceedings against) Joined cases C-321/94 and C-324/94 [1997] ECR I-2343.
Poucet v Assurances Gnrales de Frances (AGF), Pistre v Caisse Autonome Nationale de
Compensation de lAssurance Viellesse des Artisans (Cancaval) Joined cases C-159160/91 [1993]
ECR I-637.
Procureur du Roi v Dassonville Case 8/74 [1974] ECR 837.
R v Secretary of State for Health, ex p Richardson Case C-137/94 [1995] All ER (EC) 865, [1995] ECR I-
3407, ECJ.
Rewe-Zentral AG v Bundesmonopolverwaltung fr Branntwein Case 120/78 [1979] ECR 649.
Rijkinstituut voor de Sociale Verzekering der Zelfstandingen v Wolf Case 154/87 [1988] ECR 3897.
Rnfeldt v Bundesversicherungsanstalt fr Angestellte Case C-227/89 [1991] ECR I-323.
Sger v Dennemeyer & Co Ltd Case C-76/90 [1991] ECR I-4221.
Schumacher v Hauptzollamt Frankfurt am Main-Ost Case 215/87 [1989] ECR 617.
Socit Civile Immobilire Parodi v Banque H Albert de Bary et Cie Case C-222/95 [1997] All ER (EC)
946, [1997] ECR I-3899, ECJ.
Society for the Protection of Unborn Children Ireland Ltd v Grogan Case C-159/90 [1991] ECR I-4685.
Sodemare SA v Regione Lombardia Case C-70/95 [1997] ECR I-3395.
Spruyt v Bestuur van de Sociale Verzekeringsbank Case 284/84 [1986] ECR 685.
Stanton v Inasti (Institute national dassurances sociales) Case 143/87 [1988] ECR 3877.
Stichting Collectieve Antennevoorziening Gouda v Commissariaat voor de Media Case C-288/89 [1991]
ECR I-4007.
Stber v Bundesanstalt fr Arbeit Joined cases C-45/95 [1997] ECR I-511.
Svensson v Ministre du Logement et de lUrbanisme Case C-484/93 [1995] ECR I-3955.
Syndesmos ton en Elladi Touristikon kai Taxidiotikon Grafeion v Ergasias Case C-398/95 [1997] ECR I-
3091.
Triches v Casse de Compensation pour Allocations Familales de la Rgion Ligeoise Case 19/76 [1976]
ECR 1243.
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER
(EC) 97, [1995] ECR I-4921, ECJ.
van Binsbergen v Bestuur van de Bedrijfsvereniging voor de Metaalnijverheid Case 33/74 [1974] ECR
1299.
Webb (Criminal proceedings against) Case 279/80 [1981] ECR 3305.
Wirth v Landeshauptstadt Hannover Case C-109/92 [1993] ECR I-6447.
677

References

Decker v Caisse de Maladie des Employs Privs

By decision of 5 April 1995, the Conseil Arbitral des Assurances Sociales, Luxembourg, referred to the
Court of Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty a
question (set out at p 711 d e, post) on the interpretation of arts 30 and 36 of that treaty. The question
arose in proceedings between Mr Decker, a Luxembourg national, and the Caisse de Maladie des
Employs Privs, concerning a request for reimbursement of the cost of a pair of spectacles purchased
from an optician established in Belgium, on a prescription from an ophthalmologist established in
Luxembourg. Written observations were submitted on behalf of: Mr Decker, by A Braun and S Wagner, of
the Luxembourg Bar; the Luxembourg government, by C Ewen, Social Security Inspector, First Class, in
the Ministry of Social Security, acting as agent; the Belgian government, by J Devadder, Director of
Administration in the Ministry of Foreign Affairs, Foreign Trade and Development Cooperation, acting as
agent; the German government, by E Rder, Ministerialrat in the Federal Ministry of Economic Affairs, and
G Thiele, Assessor in that ministry, acting as agents; the Spanish government, by A Navarro Gonzlez,
Director General of Community Legal and Institutional Coordination, and G Calvo Daz, Abogado del
Estado, acting as agents; the French government, by C de Salins, Deputy Director in the Legal Affairs
Directorate of the Ministry of Foreign Affairs, and P Martinet, Foreign Affairs Secretary in that directorate,
acting as agents; the Netherlands government, by A Bos, Legal Adviser, acting as agent; the United
Kingdom government, by L Nicoll, of the Treasury Solicitors Department, acting as agent, and P Watson,
Barrister; and the European Commission, by H van Lier, Legal Adviser, and J-F Pasquier, a national civil
servant seconded to the Legal Service, acting as agents. Oral observations were made by: Mr Decker,
represented by S Wagner; the Caisse de Maladie des Employs Privs, represented by A Rodesch, of the
Luxembourg Bar; the Luxembourg government, represented by C Ewen, the German government;
represented by E Rder, the Spanish government, represented by G Calvo Daz; the French government,
represented by P Martinet; the UK government, represented by P Watson; and the Commission,
represented by J-F Pasquier. The language of the case was French. The facts are set out in the opinion
of the Advocate General.

Kohll v Union des Caisses de Maladie

By judgment of 25 April 1996, the Luxembourg Cour de Cassation referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty two questions (set out at p 716 h j, post) on the
interpretation of arts 59 and 60 of that treaty. Those questions arose in proceedings between Mr Kohll, a
Luxembourg national, and the Union des Caisses de Maladie (UCM), with which he was insured,
concerning a request for authorisation for his daughter to receive treatment from an orthodontist
established in Germany. Written observations were submitted on behalf of: Mr Kohll, by J Hoss and P
Santer, of the Luxembourg Bar; UCM, by A Rodesch, of the Luxembourg Bar; the Luxembourg
government, by C Ewen, Social Security Inspector, First Class, in the Ministry of Social Security, acting as
agent; the German government, by E Rder, Ministerialrat in the Federal Ministry of Economic Affairs, and
S Maa, Regierungsrtin in that ministry, acting as agents; the Greek government, by 678 V
Kondolaimos, Assistant Legal Adviser in the state Legal Service, and S Vodina, specialist technical
assistant in the Community Legal Affairs Department, Ministry of Foreign Affairs, acting as agents; the
French government, by C de Salins, Deputy Director in the Legal Affairs Directorate of the Ministry of
Foreign Affairs, and P Martinet, Foreign Affairs Secretary in that directorate, acting as agents; the Austrian
government, by M Potacs, of the Federal Chancellors Office, acting as agent; the United Kingdom
government, by S Ridley, of the Treasury Solicitors Department, acting as agent, and D Pannick QC and
P Watson, Barrister; and the European Commission, by M Patakia, of its Legal Service, acting as agent.
Oral observations were made by: Mr Kohll, represented by J Hoss and P Santer; UCM, represented by A
Rodesch; the Luxembourg government, represented by C Ewen; the Greek government, represented by
V Kondolaimos; the French government, represented by J-F Dobelle, Deputy Director in the Legal Affairs
Directorate of the Ministry of Foreign Affairs, acting as agent, and P Martinet; the UK government,
represented by R Plender QC and P Watson; and the Commission, represented by J-C Sch, of its
Legal Service, acting as agent. The language of the case was French. The facts are set out in the opinion
of the Advocate General.

16 September 1997.

The Advocate General (G Tesauro)

delivered the following joint opinion in Decker v Caisse de Maladie des Employs Privs Case C-120/95
and Kohll v Union des Caisses de Maladie Case C-158/96 (translated from the Italian).
1.
The implementation of the single market, that area without internal borders within which the free
movement of goods, persons, services and capital is ensured, ought by now to have been completed and
to constitute one of the cornerstones and hallmarks of the Community. However, the questions referred to
the court by two Luxembourg judicial bodies, the Conseil Arbitral des Assurances Sociales (the Social
Insurance Arbitration Council) (Case C-120/95) and the Cour de Cassation (the Court of Cassation) (Case
C-158/96), cast some doubt on whether this is so, by highlighting the fact that to this day citizens of the
Community, unless they forgo reimbursement of the costs incurred, are obliged, as a rule, to obtain
medical services in the state in which they are insured.
Reimbursement of medical expenses incurred in another member statewhether to purchase
medicinal products or to obtain medical treatmentis conditional on prior authorisation, the granting of
which by the competent social security institution is moreover subject to particularly restrictive conditions.
This state of affairs cannot but discourage the free movement of patients, or of the less affluent among
them at least, which of itself is a disadvantage for the patients concerned of course. It is this very state of
affairs which lies at the root of the disputes pending before the Conseil Arbitral des Assurances Sociales
and the Cour de Cassation. The issue to be decided by the former is whether the sickness fund acted
lawfully in refusing to reimburse the cost of a pair of spectacles purchased in another member state by an
insured person who had not applied for and obtained the prescribed authorisation in advance. The case
before the latter, meanwhile, involves the refusal of a request for authorisation sought by an insured
person on behalf of his daughter, who is a minor, for the purpose of claiming the cost of orthodontic
treatment to be carried out in another member state.
2.
The questions referred by the two bodies require the court to rule essentially on the compatibility with
Community law of national rules, of 679 Luxembourg in this case, which make the reimbursement of
medical expenses incurred outside the national territory subject to the condition that the medical
treatment or the purchase of the medical products and/or accessories in question were duly authorised by
the competent social security institution. Those rules are attacked on two different grounds: that they limit
the free movement of medical products and accessories, namely a pair of spectacles, and thus
contravene arts 30 and 36 of the Treaty (Case C-120/95); and that they may fetter the freedom to provide
medical services, namely orthodontic treatment, by reason of the restrictions they impose on the
recipients of such services, and are thus contrary to arts 59 and 60 of the Treaty (Case C-158/96).
The issue, therefore, is whether the requirement of prior authorisation, as a condition for
reimbursement, is such as to constitute a barrier to the free movement of goods (Case C-120/95) or the
freedom to provide services (Case C-158/96) and, if so, whether it may none the less be regarded as a
barrier which is justified in view of the special requirements inherent in a national health service.
3.
 Although these two cases have been referred by two different bodies and, at least at first sight, turn on
the interpretation of different provisions, I none the less consider it appropriate to deal with them together,
since the contested national measure is the same in both instances and the arguments adduced by the
parties and by the governments which have submitted observations are essentially the same. My decision
is also influenced by the fact that the Community rules in the field of social security may have a significant
bearing on the assessment of the restrictive effects on trade, whether in goods or in services, complained
of by the plaintiffs in the main proceedings, and this impact would not in principle be any different
according to whether it is art 30 or arts 59 and 60 which come into play.

RELEVANT LEGISLATION
4.
Article 20(1) of the Luxembourg Social Insurance Code, which was enacted on 27 July 1992 and
entered into force on 1 January 1994, provides that, with the exception of emergency care received in the
event of accident or illness abroad, insured persons may receive medical treatment abroad or have
recourse to a treatment centre or a centre providing ancillary facilities abroad only after obtaining the prior
authorisation of the competent social security institution. The terms and conditions for granting
authorisation are laid down by arts 25 to 27 of the statutes of the Union des Caisses de Maladie (the UCM
statutes), in the version which entered into force on 1 January 1995. In particular, the statutes provide that
authorisation may not be given for services which are not eligible for reimbursement under the national
rules (art 25), that the cost of duly authorised treatment is to be reimbursed in accordance with the tariffs
applicable to persons insured in the state in which the treatment takes place (art 26), and that
authorisation will be granted only after a medical assessment and on production of a written request from
a doctor established in Luxembourg indicating the doctor or hospital centre recommended to the insured
person and the criteria and facts which make it impossible for the treatment in question to be carried out
in Luxembourg (art 27).
It should also be noted, since those national rules were not yet in force at the time material to Case C-
120/95, that the relevant provisions of the former code wereso far as is relevant hereessentially
identical. In particular, the question of treatment abroad and prior authorisation thereof was governed by
art 60(3), 680which is essentially the same as art 20(1) of the code currently in force. Furthermore, the
rules governing reimbursement of the cost of spectacles were at the material time contained in art 78 of
the UCM statutes, which referred to the relevant collective agreement. For present purposes it is sufficient
to note that then, as now, reimbursement was on a flat-rate basis with a ceiling of Lf 1,600 for frames (see
art 119 of the current UCM statutes).
5.
As regards the relevant Community legislation, apart from the provisions on the movement of goods
and provision of services, art 22 of Council Regulation (EEC) 1408/71 on the application of social security
schemes to employed persons, to self-employed persons and to members of their families moving within
the Community7, is of particular significance, as we shall see below.
7
See the version consolidated by Council Regulation (EC) 118/97 (OJ 1997 L28 p 1).

The relevant parts of that article provide:

(1) An employed or self-employed person who satisfies the conditions of the legislation of the
competent state for entitlement to benefits, taking account where appropriate of the provisions of
Article 18, and (c) who is authorized by the competent institution to go to the territory of another
Member State to receive there the treatment appropriate to his condition, shall be entitled: (i) to
benefits in kind provided on behalf of the competent institution by the institution of the place of stay
or residence in accordance with the provisions of the legislation which it administers, as though he
were insured with it; the length of the period during which benefits are provided shall be governed,
however, by the legislation of the competent state; (ii) to cash benefits provided by the competent
institution in accordance with the provisions of the legislation which it administers. However, by
agreement between the competent institution and the institution of the place of stay or residence,
such benefits may be provided by the latter institution on behalf of the former, in accordance with
the provisions of the legislation of the competent state
(2) The authorisation required under paragraph 1(c) may not be refused where the treatment
in question is among the benefits provided for by the legislation of the member state on whose
territory the person concerned resided and where he cannot be given such treatment within the
time normally necessary for obtaining the treatment in question in the Member State of residence
taking account of his current state of health and the probable course of the disease.
(3) The provisions of paragraphs 1 and 2 shall apply by analogy to members of the family of an
employed or self-employed person.8

8
Article 22a, inserted by Council Regulation (EC) 3095/95 (OJ 1995 L335 p 1), provides as follows: Notwithstanding
Article 2 of the Regulation, Article 22(1)(a) and (c) shall also apply to persons who are nationals of a Member State and
are insured under the legislation of a Member State and to the members of their families residing with them. Following
that amendment, therefore, it is no longer necessary to be a worker or a member of a workers family in order to rely on
art 22; it is enough to be insured, irrespective of the basis on which one is insured.

This article therefore, like the Luxembourg legislation challenged in this case, makes reimbursement of
medical expenses incurred in another member state subject to the condition that the insured person (who
received the benefits in question) obtained prior authorisation from the competent social security 681
institution. Only if that is the case will the competent institution bear the costs incurred 9.
 9
Article 36(1) of the regulation provides that benefits in kind provided in accordance with the provisions of this chapter by
the institution of one Member State on behalf of the institution of another Member State shall be fully refunded.

FACTS AND QUESTIONS SUBMITTED FOR A PRELIMINARY RULING

Case C-120/95
6.
The dispute in Case C-120/95 is between Mr Decker, a Luxembourg national, and the Caisse de
Maladie des Employs Privs (the fund) and arises from the latters refusal to reimburse the cost of a pair
of spectacles purchased from an optician in Arlon (Belgium) on production of a prescription from an
ophthalmologist in Luxembourg. Arguing that this refusal, based on the ground that he had failed to seek
prior authorisation as required under the relevant legislation, was contrary to the Community rules on the
free movement of goods, Mr Decker complained to the fund and subsequently appealed to the Conseil
Arbitral des Assurances Sociales10.
10
The Conseil Arbitral des Assurances Sociales dismissed the appeal by order of 24 August 1993. It likewise dismissed Mr
Deckers appeal against that order by decision of 20 October 1993. It was after the setting aside of that decision by the
Cour de Cassation that the case was remitted to the Conseil Arbitral des Assurances Sociales.

7.
For the purpose of resolving the dispute before it, the latter body decided to refer the following
question to the Court of Justice for a preliminary ruling:
Is Article 60 of the Luxembourg Code des Assurances Sociales, under which a social security
institution of Member State A refuses to reimburse to an insured person, who is a national of
Member State A, the cost of spectacles with corrective lenses, prescribed by a doctor established
in Member State A but purchased from an optician established in Member State B, on the ground
that all medical treatment abroad must be authorised in advance by the above social security
institution, compatible with Articles 30 and 36 of the EEC Treaty, in so far as it penalises in general
the importation by private individuals of medicinal products or, as in this case, spectacles from
other Member States?

Case C-158/96
8.
The plaintiff in Case C-158/96 is another Luxembourg national, a Mr Kohll, who, unlike Mr Decker, did
request prior authorisation from the Union des Caisses de Maladie (the UCM), with which he is insured, to
enable his daughter Aline, who is a minor, to receive orthodontic treatment in Trier (Germany). His request
was rejected, however, by decision of 7 February 1994, on the grounds, first, that the treatment sought
was not urgent and, secondly, that suitable treatment of that kind could be obtained in Luxembourg.
The Conseil Arbitral des Assurances Sociales, before which Mr Kohll challenged the refusal, since
confirmed by the UCM Management Board, dismissed his appeal by decision of 6 October 1994. Mr Kohll
appealed against that decision to the Conseil Supeneur des Assurances Sociales which, by decision of 17
July 1995, upheld the contested decision on the ground that art 20 of the code and arts 25 and 27 of the
UCM statutes, on which the refusal was based, were fully in accordance with art 22 of the regulation.
682
9.
Finding that the appellate body had considered only whether the Luxembourg rules were consistent
with the regulation, and had not taken any account of the provisions on freedom to provide services, the
Cour de Cassation, before which Mr Kohll challenged the decision of the appellate body, decided that it
was necessary to refer the following questions to the Court of Justice for a preliminary ruling:
(1) Are Articles 59 and 60 of the Treaty establishing the EEC to be interpreted as precluding
rules under which reimbursement of the cost of benefits is subject to authorisation by the insured
persons social security institution if the benefits are provided in a Member State other than the
State in which that person resides?
(2) Is the answer to Question 1 any different if the aim of the rules is to maintain a balanced
medical and hospital service accessible to everyone in a given region?

THE ISSUES RAISED AND OUTLINE OF THIS OPINION

10.
As they are worded, the questions referred by the Conseil Arbitral des Assurances Sociales and the
Cour de Cassation turn solely on the interpretation of the Treaty provisions relating to the free movement
of goods and services: the court has been asked to determine whether the requirement of prior
authorisation for reimbursement of medical expenses incurred in a member state other than the state of
residence is contrary to arts 30 and 59 of the Treaty.
In the course of the procedure, however, there was some discussion, in depth, of the applicability to
the case of the Community rules on social security, and specifically art 22(1)(c)(i) of the regulation. In
particular, it was argued that it is this provision which regulates the issue of prior authorisation and
reimbursement and that, since the disputed rules are fully in accordance with it, the possibility of conflict
with arts 30 and 59 does not arise. This is closely linked to another argument put forward during the
course of the procedure, according to which the fact that the disputed measure falls within the sphere of
social security means that, even if the regulation were inapplicable to the case at hand, the measure still
could not be examined from the point of view of the Community rules providing for the free movement of
goods and services. This, the argument basically runs, is because the social security sector is even today,
at least where it is not governed by specific Community rules, a matter within the purview of the member
states.
11.
In those circumstances, I believe the first task is to ascertain whether the national rules in question fall
within the scope of arts 30 and 59. To do so, it will be necessary to determine, in the light inter alia of the
courts case law in this area, to what extent and subject to what limits the fact that the measure in
question is in the field of social security precludes any inquiry as to its compatibility with the fundamental
freedoms guaranteed by the Treaty.
It will likewise be necessary to determine whether art 22 of the regulation applies to the case and, if
so, whether this fact of itself takes the matter outside the scope of arts 30 and 59. On this point, I cannot
refrain from observing right away that a provision of secondary legislation, which is what art 22 of the
regulation is, cannot in any event be regarded as capable of excluding an examination into whether or not
Treaty provisions, such as arts 30 and 59, have been infringed. I therefore take the view that even if, after
examination, the Luxembourg rules are found to come within the scope of the regulation and to 683 be
consistent with it, it does not follow that arts 30 and 59 are inapplicable to this case.
12.
Secondly, once it has been established that neither the social security nature of the national measure
nor the existence of a Community provision having essentially the same content has the effect of
precluding the application of arts 30 and 59, it will then be necessary to determine whether the measure is
compatible with the Community provisions in question. In other words, it will be necessary to ascertain
whether the requirement of prior authorisation for reimbursement of medical expenses incurred in a
member state other than the state of residence amounts, at least in principle, to a barrier to trade in goods
and/or services.
Needless to say, the conclusion arrived at will apply equally to art 22 of the regulation, assuming of
course that the disputed national rules are consistent with it. In other words, if the barrier to the movement
of goods and the provision of services is the requirement to seek prior authorisation in the absence of
which the competent institution will refuse to reimburse expenses incurred by an insured person in
another member stateit is quite clear that the barrier is created in like manner by both the national rules
and the regulation.
13.
Lastly, we shall have to consider whether any restrictive effects caused by the national measure, and,
by the same token, by art 22 of the regulation, may nevertheless be justified. To that end, it will first be
necessary to establish whether the disputed rule is discriminatory or indistinctly applicable, since the
justifications which come into play in the two cases are different: express derogations (under arts 36 and
56) in the former, and overriding requirements or, if one prefers, reasons relating to the general interest, in
the latter. It is only in the latter case that account may be taken of special requirements, including
economic requirements where appropriate, inherent in the existence, operation and maintenance of a
national health service which is accessible to everyone.
Let me add that the analysis of the justifications relied upon and its outcome will inevitably imply a
view on the validity or otherwise of art 22 of the regulation. If the disputed rules are found to be
incompatible with arts 30 and 59, this may result ipso facto in the relevant Community provision being
declared invalidto the extent that both measures (national and Community) lay down the same rules,
follow the same logic and pursue the same aim.
14.
Finally, let me point out that this reasoning process could undoubtedly proceed more rapidly and less
laboriously if some of the steps mentioned above were omitted. I feel, however, that, at least in an
opinion, it is not permissible to dispense with the analytical rigour which a case of this importance
demands.

IThe applicability of arts 30 and 59

15.
Although the points in dispute in the two cases are many, perhaps too many, there is fortunately no
dispute about some fundamental matters which are essential to a correct definition of the issues involved.
It is common ground that medical products constitute goods within the meaning of art 30 of the Treaty,
from which it follows that any unjustified limitation on the importation of this category of products,
including importation by an individual for personal use, is contrary to that article (see Schumacher v
Hauptzollamt Frankfurt am Main-Ost Case 215/87 [1989] ECR 617 and EC Commission v Germany Case
C-62/90 [1992] ECR I-2575). Likewise, it is not disputed that medical activities constitute services, as is
expressly provided for by art 6011, nor, as was made clear by the court, that
11
See also the judgment in Society for the Protection of Unborn Children Ireland Ltd v Grogan Case C-159/90 [1991] ECR
I-4685 (para 21), in which it was held that medical termination of pregnancy, performed in accordance with the law of
the state in which it is carried out, constitutes a service within the meaning of Article 60 of the Treaty.

684
the freedom to provide services includes the freedom, for the recipients of services, to go to
another member state in order to receive a service there and that tourists, persons receiving
medical treatment and persons travelling for the purpose of education or business are to be
regarded as recipients of services. (See the judgment in Luisi v Ministero del Tesoro Joined cases
286/82 and 26/83 [1984] ECR 377 (para 16).)
 That being the case, it is equally indisputable, in my view, that rules which make reimbursement of
medical expenses conditional on the medical products and/or services upon which they were incurred
having been purchased within the national territory are quite capable, at least in theory, of having an
adverse effect on trade in goods and services. Suffice it for now to point out that such rules for the very
reason that they deny insured persons reimbursement of medical expenses incurred abroad, unless prior
authorisation has been obtainedmay discourage those concerned from purchasing such products or
from seeking treatment in a member state other than the one in which they reside, which may, in certain
cases, act as a restriction on imports of the relevant products or on the freedom to provide the services in
question.
16.
The applicability of the Treaty rules on the free movement of goods and freedom to provide services
has, however, been called into question, as indicated above, on the grounds that the national measure in
question concerns social security and is alleged to be in accordance with a precise and specific provision
of the regulation. It has been argued that those two considerations, which, as we shall see, are
interrelated, have the effect of taking the contested measure outside the scope of arts 30 and 59. The
following arguments seek to refute that contention, which was supported by the majority of the member
states which submitted observations in the two cases.
The fact that the contested measure concerns social security
17.
Let me begin by remarking that the fact that the national rules in question concern social security by
no means has the effect of removing them, at least not automatically, from the scope of the Community
rules on the movement of goods and provision of services. And I would add straight away that the courts
consistent view that Community law does not detract from the powers of the member states to organise
their social security systems by no means implies that the social security sector constitutes an island
beyond the reach of Community law and that, as a consequence, all national rules relating to social
security fall outside its scope (see the judgment in Duphar BV v Netherlands Case 238/82 [1984] ECR
523 (para 16))12.
12
See also, more recently, the judgment in Sodemare SA v Regione Lombardia Case C-70/95 [1997] ECR I-3395 (para
27). For a comprehensive review of the case law in this area, and its implications, see the opinion of Advocate General
Fennelly, delivered in Sodemare SA [1997] ECR I-3395 (paras 2330).

18.
It is of course true that as Community law stands at present it is for the legislature of each member
state to lay down the conditions creating the right or the obligation to become affiliated to a social security
scheme13, with the result that those subject to such legislation, if it makes affiliation compulsory, have no
13
It is not by chance that this statement is a Constant feature of the case law on the interpretation of Regulation 1408/71.
See, inter alia, the judgments in Coonan v Insurance Officer Case 110/79 [1980] ECR 1445 (para 12) and Paraschi v
Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501 (para 15).

685
choice as to which scheme to join. It is also true, as the court held in Poucet v Assurances Gnrales
de Frances (AGF), Pistre v Caisse Autonome Nationale de Compensation de lAssurance Viellesse des
Artisans (Cancaval) Joined cases C-159160/91 [1993] ECR I-637 (para 19), that the activity of sickness
funds, and of the organisations involved in the management of the state social security system, is not an
economic activity and, therefore, the organizations to which it is entrusted are not undertakings within the
meaning of Articles 85 and 86 of the Treaty.
However, the scope of that case law is much narrower and more confined than is claimed by the
member states in arguing that Community law does not apply to social security. It is therefore necessary
to place those decisions in their proper context and to assess their implications correctly.
19.
First of all, it is quite clear that, in the absence of harmonisation at Community level, it is for the
member states to lay down the conditions governing affiliation to the social security system and, at least
in principle, the conditions governing entitlement to social security benefits (see Stber v Bundesanstalt
fr Arbeit Joined cases C-45/95 [1997] ECR I-511 (para 36)). However, the court has also made clear
that the freedom thus left to the member states must not give rise to discrimination between nationals of
the host state and nationals of the other Member States (see Coonan v Insurance Officer Case 110/79
[1980] ECR 1445 (para 12) and Paraschi [1991] ECR I-4501 (para 15)).
In other words, while it is true that the organisation of the social security system remains a matter for
the member states and that the relationship between the social security institutions and their members is
governed by national law, this does not mean that member states may contravene with impunity a
fundamental principle established by the Treaty to secure the free movement of persons, namely the
prohibition of discrimination on grounds of nationality 14.
14
Suffice it to recall that one of the cardinal principles of Regulation 1408/71, although its purpose is limited merely to
coordinating national legislation in this area, is in fact equality of treatment, laid down in art 3(1), between a member
states own nationals and those of other member states resident in the state.

20.
Secondly, the point must be made that while the court did indeed conclude that social security
institutions are not subject to the rules on competition, this applies only in so far as the activity of those
institutions is based on the principle of solidarity, in the sense that the benefits provided are statutory
benefits bearing no relation to the amount of the contributions (see the judgment in Poucet [1993] ECR I-
637 (para 18)). In other words, it is clear that the determining factor in the outcomealthough the court
also referred, in the same judgment, to the exclusively social function performed by those institutions,
whose activity is entirely non-profit-makingwas precisely the fact that the system in question was based
on the principle of solidarity15. This was confirmed in a later judgment in which the court held that where
those (or similar) institutions operate instead a supplementary insurance scheme, based on the funding
principle and with the level of benefits directly linked to the level of
15
A point confirmed, albeit a contrario, in Hfner v Macrotron GmbH Case C-41/90 [1991] ECR I-1979 (para 21), where
the court stated that, in the context of competition law, the concept of an undertaking encompasses every entity
engaged in an economic activity, regardless of the legal status of the entity and the way in which it is financed.

686
contributions (the solidarity principle thus not applying), they are to be regarded as undertakings within
the meaning of the Treaty rules on competition (see the judgment in Fdration Franaise Socits
dAssurances v Ministre de lAgriculture et de la Pche Case C-244/94 [1995] ECR I-4013 (paras 17
22)).
In essence, therefore, compulsory affiliation to a particular scheme, although it has the effect of
eliminating all potential competition by other entities or at any rate reducing their activity to a purely
residual sector, is a necessary condition for the operation of a social security scheme based on the
solidarity principle, which depends on everybodys contribution in order that each may benefit. It is
therefore only such institutions as operate social security schemes of this nature that are exempted from
the application of the Community rules on competition.
21.
It is true also that in its judgment in Garcia v La Mutuelle de Prvoyance Sociale Aquitaine Case C-
238/94 [1996] ECR I-1673 (para 13) the court stated that arts 57(2) and 66 of the Treaty, which were the
legal basis for the adoption of the coordinating directives intended to facilitate the taking-up and pursuit of
self-employed activities, could not regulate the field of social security, which is covered by different
provisions of Community law. But this statement, I believe, cannot be interpreted as meaning that the
Treaty rules on the right of establishment and provision of services are inapplicable, by definition, to
national rules on social security16, and it should be read in the light of the particular features of that case.
16
Since the statement was taken, albeit couched in more absolute terms, from my opinion in that case, I must point out
that it was certainly not my intention to arrive at such an result (see [1996] ECR 1-1675 (para 9)). Rather, as may be
seen from the opinion, I was seeking to make the point that in the present state of Community law, none of the
provisions of the Treaty, including those directly concerning the social security sector or social affairs in general, can
serve as a basis for the adoption of measures aimed at dismantling the national social security systems (see footnote
6), where the term dismantling means precisely the destruction of the different systems which exist at present in the
member states. There is no doubt, on the other hand, that if those systems were one day to be harmonised, the proper
legal basis for undertaking the liberalisation (inter alia) of the activity of the institutions which administer them would
indeed be arts 57(2) and 66 of the Treaty.

In view of the fact that that case concerned a challenge to the requirement of compulsory affiliation to
a statutory social security scheme in reliance on the liberalisation introduced by a directive coordinating
national rules on direct insurance other than life assurance, it need only be pointed out that the exclusion
from liberalisation of the activities of bodies operating statutory social security schemes, which is,
moreover, an express provision of that directive, was a necessary corollary to the exemption, already
established, of such activities from the competition rules. Otherwise the obligation to join such schemes
would inevitably have been removed, whereas the court emphasised that the obligation was necessary in
order to ensure that the principle of solidarity is applied and that their financial equilibrium is maintained,
pointing out that, if the obligation was removed, the schemes in question would thus be unable to survive
(see the judgment in Garcia [1996] ECR I-1673 (para 14)).
22.
In the light of the foregoing, it is abundantly clear that the statement that Community law does not
detract from the powers of member states to organise their social security systems means, quite simply,
that Community law does not regulate this area directly, nor does it impinge on it, if its application would
have the effect of jeopardising the survival of social security schemes of the kind described above.
Otherwise, however, member states are bound to comply with 687 Community law even when exercising
the powers reserved to them in the field of social security.
There is no basis for any other conclusion. The fact remains, therefore, that no discrimination based
on nationality is permitted on the territory of a member state, and that, with the exception of the case
where national rules are closely bound up with the operation and survival of the relevant social security
scheme, social security is not exempt from the application of Community law (see para 19, above). In
particular, for the purposes of the present case, we may note that unjustified restrictions on the free
movement of persons (establishment and services) and of goods are not permitted merely because the
benefit conferred on individuals by the relevant Community rules conflicts with a national measure which
in some way relates to social security.
23.
The courts case law on the matter confirms this. For example, the court has acknowledged that, as
Community law stands at present, a member state may consider that the social welfare system it has put
in place, whose implementation is in principle entrusted to the public authorities, [and which] is based on
the principle of solidarity, necessarily requires, in order to achieve its objectives, that the admission of
private operators to the system be subject to the condition that they are non-profit-making; the court
accordingly came to the conclusion that national rules laying down such a requirement are not contrary to
art 52 of the Treaty (see Sodemare [1997] ECR I-3395 (paras 29, 32, 34))17. Clearly, the outcome was
determined by the importance attributed, rightly or wrongly, to the fact that the system in question was
based on the principle of solidarity.
17
The court none the less felt it necessary, in that judgment, to state that the system in question cannot however place
profit-making companies from other Member States in a less favourable factual or legal situation than profit-making
companies from the Member State in which they are established (see [1997] ECR I-3395 (para 33)).

The court has taken an entirely different approach, however, in cases where the national measures,
albeit falling within the sphere of social security, were not of such a kind, on the face of it, as to be
capable of having an adverse effect on the survival of social security schemes based on the principle of
solidarity. For example, in upholding as compatible with art 52 a Belgian measure which denied
laboratories operated by legal persons, whose members were also legal persons, reimbursement of the
cost of clinical biology services performed by them, the court not only pointed out that the legislation in
question applied without distinction to Belgian nationals and those of other member states, but also stated
that there was no evidence that it had been adopted for discriminatory purposes or that it produced
discriminatory effects (see the judgment in EC Commission v Belgium Case 221/85 [1987] ECR 719
(para 11)). Still on the subject of art 52, but on a more general note, the court has stated that member
states are under an obligation to observe its provisions even though, in the absence of Community
legislation on social security for self-employed persons, they retained legislative jurisdiction in this field
(see the judgment in Stanton v Inasti (Institute national dassurances sociales) Case 143/87 [1988] ECR
3877 (para 10))18.
18
See also the judgments of the same date in Rijkinstituut voor de Sociale Verzekering der Zelfstandingen v Wolf Case
154/87 [1988] ECR 3897 (para 10).

24.
The court has also had occasion to state, in relation to the rules on the free movement of goods, that
even measures adopted in the field of social security and which do not relate directly to importation,
depending on the manner of their application and the use made of them may affect the possibilities of
688 marketing the preparations and, to that extent may indirectly influence the possibilities of
importation, thus, potentially, falling foul of art 30 (see the judgment in Duphar [1984] ECR 523 (para
18))19. This would be the case, for instance, of national rules which had the effect of making only domestic
products eligible for reimbursement.
19
See also the judgment in EC Commission v Belgium Case C-249/88 [1991] ECR I-1275 (paras 38, 42), where the court
held that national rules which gave preference, as regards eligibility for reimbursement, only to domestic pharmaceutical
products, were contrary to art 30.

Lastly, in the judgment in Sodemare SA v Regione Lombardia Case C-70/95 [1997] ECR I-3395 (paras
3640), where a possible breach of the rules on the freedom to provide services was also in issue, the
court took pains to explain that in the case in question there was no provision of services within the
meaning of the Treaty. It accordingly ruled:
(40) Article 59 of the Treaty does not cover the situation of a company which, having
established itself in a member state in order to run old peoples homes there, provides services to
residents who, for that purpose, reside in those homes permanently or for an indefinite period.
That statement demonstrates, however, lest there be any remaining doubt, that social security does not
per se fall outside the scope of the Community rules.
25.
In short, the fact that, in the present state of Community law, member states powers in the field of
social security, as in other areas20, remain intact, by no means leaves them free to enact in that field rules
which are contrary to Community law. It follows that the social security nature of the disputed rules does
not, of itself, by any means have the effect of precluding any review of their compatibility with arts 30 and
59 of the Treaty.
20
A similar approach is to be seen, for example, in the area of direct taxation. The court, in its case law in that field, has
consistently held that although, as Community law stands at present, direct taxation does not as such fall within the
purview of the Community, the powers retained by the member states must nevertheless be exercised consistently with
Community law (see Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR
I-225 (para 21); see also, more recently, Futura Participations SA v Administration des Contributions Case C-250/95
[1997] ECR I-2471 (para 19)).

Article 22 of Regulation 1408/71

26.
It now falls to be considered whether the above conclusion may be called into question by reason of
the existence of a provision of Community law, namely art 22 of Regulation 1408/71, governing the very
matter which concerns us. Let me say straight away that the answer to that question can only be in the
negative, although the possibility of the said provision itself being invalid cannot be excluded.
But let us take things in order. We need first to ascertain whether the provision in question governs the
matter in hand, then to verify whether the contested national measure is consistent with it and, finally, if
that is indeed the case, to examine whether that fact alone may be regarded as precluding any further
inquiry into the question whether the fundamental freedoms guaranteed to individuals by the Treaty, which
undoubtedly include the free movement of goods and services, have been observed.
27.
 Let me begin by observing that art 22(1)(c) of the regulation, by making it possible for individuals
albeit subject to authorisation by the competent institutionto receive the medical services they require in
another member 689 state, constitutes one of the measures intended to permit a worker who is a national
of one of the member states of the Community, without regard to the national institution to which he is
affiliated or the place of his residence, to receive benefits in kind provided in any other member state, and
to do so, of course, without having to forgo reimbursement of the expenses incurred (see Bestuur van het
Algemeen Zieekenfonds Drenthe-Platteland v Pierik Case 117/77 [1978] ECR 825 (para 14) (Pierik No
1)). Clearly, Mr Kohll and Mr Decker are seeking to be allowed to avail themselves of this possibility I
need hardly add, in this regard, that these individuals, although they are Luxembourg nationals and have
not exercised their right to freedom of movement, are persons to whom the regulation applies. As we
know, the regulation applies not only to migrant workers and members of their families, but also to
(sedentary) workers and to members of their families moving within the Community for reasons other than
work (see Bestuur van het Algemmen Ziekenfonds Drenthe-Platteland v Pierik Case 182/78 [1979] ECR
1977 (para 4) (Pierik No 2) and Hoekstra (ne Unger) v Bestuur der Bedrijfsvereniging voor Detailhandel
en Ambachten Case 75/63 [1964] ECR 177 esp at 186)21.
21
Moreover, by virtue of the insertion of art 22a in the regulation, it is no longer even necessary, for the purposes of the
application of art 22 at least, to be a worker or a member of a workers family; it is now enough to be insured (see
footnote 8, above)

28.
That being so, it would appear, at least on the face of it, that cases such as this do indeed come within
the scope of art 22 of the regulation and are accordingly governed by it. That conclusion, which is not in
dispute in respect of cases where the benefits sought by the insured person consist, for example, in
specialist consultations or treatment (as in the Kohll case), is however disputed in respect of cases where
the benefits sought consist instead of the purchase of medical products and accessories (as in the
Decker case).
Specifically, the European Commission argues that the term benefits in art 22 covers only medical
services in the strict sense and does not extend to medical products and accessories, such as medicines
and spectacles. By contrast, the member states which have intervened (in the Decker case) maintain that
the term covers all benefits necessary for the treatment of a particular ailment and hence all products
necessary for that purpose as well. They further argue that the applicability of art 22 to medical products
and accessories is confirmed by art 19 of Council Regulation (EEC) 574/72 laying down the procedure for
implementing Regulation 1408/7122, which expressly provides that frontier workerswho are entitled to
medical treatment in both their state of residence and their state of employmentmay purchase such
items as medicines and spectacles only in the member state in which those products were prescribed 23.
22
See the amended and updated version published in OJ 1997 L28 p 1.
23
Article 19 of Regulation 574/72 provides: In the case of frontier workers or members of their families, medicinal
products, bandages, spectacles and small appliances may be issued, and laboratory analyses and tests carried out,
only in the territory of the Member State in which they were prescribed, in accordance with the provisions of the
legislation of that Member State, except where the legislation administered by the competent institution or an agreement
concluded between the Member States concerned or the competent authorities of those Member States is more
favourable.

29.
There is no doubt that sickness and maternity benefits, within the meaning of Ch 1 in Title III of the
regulation, of which art 22 forms part, include benefits under legislation concerning invalidity which are in
the nature of medical or surgical benefits (see the judgment in Jordens-Vosters v Bestuur van de 690
Bedrijfsvereniging voor de Leder- en Lederwerkende Industrie Case 69/79 [1980] ECR 75 (para 9))24.
Furthermore, I am not convinced by the Commissions argument that art 19 of Regulation 574/72 has no
application to the present case. While it is true that the provision in question applies only to frontier
workers, whom it requires to purchase medical products and accessories (and to have laboratory
analyses carried out) in the member state where the prescription was made, this is simply because those
workers are the only ones to have access, without prior authorisation, to the benefits in question in two
different member states. It would be illogical, however, to infer that what frontier workers are not allowed
to do, to prevent them from purchasing the relevant products in whichever of the two member states has
more favourable reimbursement terms, other workers in any of the other 14 member states are permitted
to do.
24
The particular case involved the refusal by the Competent social security institution, that of the Netherlands, to
reimburse to a Belgian national expenses which she had incurred in Belgium on pharmaceutical products and
medicines.

That said, I readily concede that where the benefit sought consists, as it does in the Decker case, in
the purchase of a pair of spectacles, or in the purchase of medical products in general, it can hardly be
argued that the treatment in question could not be provided in sufficient time to prevent any deterioration
in the insured persons state of health, in which case art 22 provides that authorisation may not be
refused. But this fact is not conclusive, as it means simply that authorisation will virtually never be granted
for the purchase of medical products and accessories25, save in the case of products (for example, a
particular kind of prosthesis or appliance) which cannot be found in the member state in question.
25
Of course, prior authorisation granted for medical treatment in another member state cannot but also cover expenses
incurred in that state on the purchase of medical products and accessories; this fact is, I believe, indisputable and,
indeed, has not been disputed.
30.
In the final analysis, it is my view that art 22 is intended to apply not only to medical services, in the
narrow sense, as the Commission maintains, but to all benefits involved in a particular type of treatment
or health care, and thus also to medical products and, for our purposes, to the purchase of a pair of
spectacles. I therefore conclude that the cases under consideration do indeed fall within the scope of art
22(1)(c) of the regulation.
31.
I now turn to the question whether the contested measure is in conformity with art 22 of the regulation.
It is not in dispute that both measures (Community and national) make the reimbursement of medical
expenses incurred in another member state subject to prior authorisation. Both measures also require, in
order for such authorisation to be granted, that the benefits sought by the insured person are among
those eligible for reimbursement under the legislation of the member state in question. I would also recall
that member states are bound to grant authorisation, under art 22(2), only where the treatment sought
cannot be provided within such time as to ensure its effectiveness, thereby leaving all other eventualities
to the member states discretion. For their part, the national rules in force in Luxembourg, specifically art
25 of the UCM statute, provide that authorisation is to be granted only if the treatment sought is not
available in Luxembourg or if the standard of the health care provided is inadequate for the particular
ailment from which the insured person is suffering.
691
Accordingly, there can be no doubt that the contested rules are consistent with art 22 of the regulation.
It is quite clear that those rules, at least in terms of their wording, do not go beyond the limits laid down by
the latter provision26.
26
It would be otherwise only if the interpretation given by the court to the phrase treatment appropriate to his condition, in
art 22(1)(c), were still valid today. The court had held that the phrase in question meant that benefits in kind for which
the worker is authorized to go to another Member State cover all treatment calculated to be effective for the sickness or
disease from which the person concerned suffers (see Pierik No 1 [1978] ECR 825 (para 15) and Pierik No 2 [1979]
ECR 1977 (para 10); my emphasis). Unfortunately, that interpretation must now be regarded as having been
superseded, because art 22(2) was amended, in fact as a consequence of the Pierik judgments, and made more
restrictive so that, in particular, it now sets out expressly and without ambiguity the one case in which authorisation may
not be refused.

32.
That finding, however, as already indicated, does not mean, contrary to the contention of a number of
governments during the proceedings, that there is no possibility of the contested rules conflicting with arts
30 and 59 and that, consequently, there is no need to examine whether the former are compatible with
the latter. Furthermore, the arguments put forward in support of this view are, even on the face of it,
devoid of substance.
In particular, we need not be detained by the argument that if the contested rules were found to be
incompatible with arts 30 and 59 of the Treaty, the fact that this would mean that art 22 of the regulation
was unlawful (as well) would have the effect of giving those Treaty provisions precedence over art 51, the
legal basis of the regulation, thereby setting up a hierarchy of norms for which there is no basis in the
Treaty itself. Suffice it to say that the fact that a particular national rule is compatible with an article of the
Treaty can under no circumstances constitute sufficient reason for the rule in question to be immune from
the application of other (relevant) Treaty provision27. Similarly, it is difficult even to conceive that a
regulation might violate Treaty provisions with impunity merely because (and as long as) it complied with
the Treaty provision which constitutes its legal basis, in this case art 51 28
27
On this point see eg the judgment in Du Pont de Nemours Italiana SpA v Unit Sanitaria Case C-21/88 [1990] ECR I-
889 (paras 2021), in which the court ruled that the possible classification of a national measure as aid within the
meaning of art 92 did not exempt it from the prohibition in art 30.
28
I would also recall that the court has made it clear that the discretionary power vested in the Community legislature by
art 51 must be exercised by means which are objectively justified (see the judgment in Triches v Casse de
Compensation pour Allocations Familales de la Rgion Ligeoise Case 19/76 [1976] ECR 1243 (para 18)). This
statement can only be interpreted as meaning that measures adopted pursuant to art 51 may not unjustifiably restrict
the scope of the rights conferred by the Treaty on the citizens of the Community.

33.
Furthermore, the regulation does not set up a common scheme of social security but allows different
national schemes to exist and its sole objective is to coordinate those national schemes 29, whence it
follows that the substantive and procedural differences between the systems of individual Member
States, and hence in the rights of persons working in the Member States, are unaffected by Article 51 of
the Treaty (see the judgment in Pinna v Caisse dAllocations Familiales de la Savoie Case 41/84 [1986]
ECR 1 (para 20)). It would therefore be perverse, in the absence of common rules on the matter, to hold
that a national measure is exempt from any review of its compatibility with the Treaty provisions merely
because it is one of those covered by the Communitys coordinating legislation adopted on the basis of art
51.
29
See the judgments in Gravina v Landesversicherungsanstalt Schwaben Case 807/79 [1980] ECR 2205 (para 7),
Borowitz v Bundesversicherungsanstalt fr Angestellte Case 21/87 [1988] ECR 3715 (para 23) and Rnfeldt v
Bundesversicherungsanstalt fr Angestellte Case C-227/89 [1991] ECR I-323 (para 12)

692
This appears also to be the view taken by the court in its decisions in point. I refer, in particular, to a
judgment in which the court interpreted the provisions of the regulation relied upon in the case in question
as meaning that entitlement to family benefits provided in respect of children resident in another member
state may not also be conferred on self-employed persons affiliated to a voluntary social insurance
scheme where, as in that case, the competent national institution for the payment of such benefits is
German (see the judgment in Stber [1997] ECR I-511 (paras 3234)). That conclusion, however, did not
prevent the court from inquiring into whether the national rules under which the self-employed are eligible
for the allowances in question only if they are affiliated to a statutory social security scheme were
compatible with art 5230. Thus, having pointed out that the rules in question treat nationals who have not
exercised their right to free movement and migrant workers differently, to the detriment of the latter, since
it is primarily the latters children who do not reside in the territory of the Member State granting the
benefits in question, the court found that such treatment was not objectively justified and held that the
rules in question must be regarded as discriminatory and hence as incompatible with Article 52 of the
Treaty (see [1997] ECR I-511 (paras 3839)).
30
This approach confirms, moreover, that the courts statement quoted above to the effect that observance of a directly
effective provision, such as art 52 of the Treaty, is binding on member states even though, in the absence of Community
legislation on social security for self-employed persons, they retained legislative jurisdiction in this field, has not lost any
of its significance merely because of the fact that the regulation has since been extended to cover self-employed
workers (see the judgments in Stanton [1988] ECR 3877 (para 10) and Wolf [1988] ECR 3897 (para 10)).

34.
In short, the court held in that case, on the one hand, that the relevant provisions of the regulation did
not entitle the plaintiffs to the benefits sought and upheld the lawfulness of such exclusion 31 and, on the
other, that the plaintiffs were entitled to those benefits by virtue of art 52 of the Treaty, with which the
relevant national rules were therefore in conflict32. Clearly, this is an outcome which confirms, at the risk of
stating the obvious, that there is no basis for the argument that the court is precluded from inquiring into
whether national rules are compatible with Treaty provisions having direct effect, solely because such
rules are in conformity with the relevant regulation or, at least, are not outlawed by it.
31
In this connection, the court noted in that judgment that there is nothing to prevent Member States from restricting
entitlement to family benefits to persons belonging to a solidarity system constituted by an old-age insurance scheme.
The Member States are at liberty to determine the conditions for entitlement to social security benefits, since Regulation
No 1408/71 merely plays a coordinating role (see [1997] ECR I-511 (para 36)).
32
In that respect, it is puzzling that the court did not find it necessary to question the validity of the provision of the
regulation which permitted workers not affiliated to the statutory social security scheme to be excluded from eligibility for
family allowancesparticularly if it is borne in mind that, in support of its conclusion that the national rules were
incompatible with art 52, the court pointed out that such exclusion had the effect of placing at a disadvantage nationals
who had exercised their right to free movement. But if that is the case, it must follow that the provision of the regulation
which permits such exclusion is invalid. I need hardly recall that the court has consistently held that the aim of Articles
48 to 51 would not be attained if, as a consequence of the exercise of their right to freedom of movement, workers were
to lose the advantages in the field of social security guaranteed to them by the law of a single Member State (see, inter
alia, the judgments in Spruyt v Bestuur van de Sociale Verzekeringsbank Case 284/84 [1986] ECR 685 (para 19) and
Lepore v Office National de Pensions (ONP) Joined cases C-4546/92 [1993] ECR I-6497 (para 21)). Clearly, the same
must apply to self-employed persons who exercise their right to freedom of movement, since the regulation is now also
applicable to them.

693

IIThe restrictive effects of the contested measure

35.
Having established the relevance to the case of the fundamental freedoms enshrined in the Treaty, I
now propose to examine whether the contested national rules may act as a barrier to the movement of
goods and the provision of services. I need hardly add that the outcome of this examination will be equally
valid in relation to art 22 of Regulation 1408/71, now that we have shown that the national rules in
question are in conformity with it. In other words, any restrictive effects on the movement of goods and/or
the provision of services will be capable of being traced back to both the national rules and Regulation
1408/71.
36.
The disputed rules, it will be recalled, makes the reimbursement of medical expenses incurred by an
insured person in a member state other than his or her state of residence whether on the purchase of
medical products and accessories or on medical and hospital servicessubject to the condition that the
insured person in question has obtained prior authorisation from the competent social security institution.
It therefore needs to be determined whether, as claimed by the plaintiffs in both the main proceedings, the
requirement of prior authorisation is in conflict, at least in principle, with art 30 and/or art 59 of the Treaty.
On that point, all the member states which submitted observations in the two cases were at one in
dismissing the possibility of any barrier to intra-Community trade. In their view, the rules in question have
neither the object nor the effect of restricting trade flows but in fact do no more than set out the terms and
conditions for reimbursement of medical expenses. The rules are therefore concerned only with the
relationship between the insured person and the social security institution to which he is affiliated, the
member states maintain. That line of reasoning, which is, to say the least, based on a simplistic view of
the issue, must be rejected as regards both the movement of goods and the provision of services.
The movement of goods
37.
I have already recalled that, according to the courts case law in this area, any rules which unjustifiably
restrict an individuals ability to import medicinal products for personal use is contrary to art 30 (see para
15, above). I need hardly add that this must also apply to a pair of spectacles and, in general, to all
medical or surgical accessories.
 That said, it is true that the rules under discussion by no means amount to a ban on importing the
products in question, nor do they even directly affect the possibility of purchasing them outside the
national territory. Insured persons remain free to purchase such products wherever they please, including
in a member state other than their state of residence 33. But from that it does not follow, at least not of
itself, that the rules in question do not create any barrier to the importation of such products.
33
Suffice it to note that Mr Decker had no difficulty whatsoever in purchasing the pair of spectacles giving rise to the
dispute in a member state other than his state of residence. The problems began, however, when he applied to the
competent social security institution for reimbursement of the cost, which was duly refused.

38.
In the first place, it is quite clear that the rules in question, by requiring prior authorisation only for
purchases made outside the national territory, involve unequal treatment based on the place of purchase
of the products 694 concerned34. Even assuming, for the sake of argument, that such differentiated
treatment is in itself of no relevance for the purposes of the application of the Community rules on goods,
it is nevertheless the case that the rules in question, by denying insured persons not having obtained prior
authorisation reimbursement of medical expenses incurred outside the national territory, have the effect of
deterring those concerned from buying medicinal products, or even a pair of spectacles, in another
member state. In other words, such rules constitute a clear disincentive to purchase the products
concerned in a member state other than the state of residence: and this results, or at any rate may result,
in a limitation on imports of such products.
34
In this regard, it should however be emphasised that the authorisation in question, although it is required only for the
purchase of products abroad, cannot be equated with other prior authorisations struck down by the court (see eg EC
Commission v UK Case 124/81 [1983] ECR 203 (para 18)). This is essentially because in the case which concerns us
importation as such is not subject to authorisation.

Moreover, the authorisation in question, in view of the particularly restrictive conditions to which its
grant is subject, is very unlikely to be accorded in a case where the benefit sought by the insured person
is simply the purchase of a pair of spectacles and/or, in general, any products prescribed in the member
state in which he or she is resident35. In those circumstances, clearly, the rules in question regulate the
reimbursement process in such a way that only products purchased in the national territory are eligible 36.
In view of the point made above regarding the deterrent effect of reimbursement not being available, it is
therefore indisputable that such rules inhibit, albeit indirectly, imports of medical products and accessories
by private individuals for their personal use.
35
In circumstances such as these, it is obvious that the competent social security institution will refuse to grant prior
authorisation, since it must be assumed, first, that a pair of spectacles (even of a particular kind), or a medicinal product,
prescribed respectively by an ophthalmologist and a doctor established in a given member state, are available in the
territory of that state and, secondly, that the state of health of the insured person is not going to deteriorate if he does
not purchase those products outside the national territory (on this point, see also para 29, above). If anything, the
insured persons state of health could deteriorate if he decided, instead of buying the products he needs as soon as
possible, to seek prior authorisation and to await the result (virtually certain to be negative) of a series of medical and
administrative formalities.
36
A very clear corollary of the decision in Duphar [1984] ECR 523 (paras 1822), interpreted a contrario, is that if national
rules were to regulate the reimbursement process in such a way that only domestic products were eligible, they would
be contrary to art 30. The fact that in the case which concerns us it is only products purchased in the national territory
which are eligible for reimbursement should not, to my mind, produce a different outcome.

39.
To my mind, this is sufficient to warrant the conclusion that the national rules in question meet the
criteria for a measure having equivalent effect to a quantitative restriction inasmuch as they are capable,
in accordance with the well-known Dassonville formula, of directly or indirectly, actually or potentially,
hindering intra-Community trade (see the judgment in Procureur du Roi v Dassonville Case 8/74 [1974]
ECR 837 (para 5)). It follows that, without prejudice to any possible justification which may be available,
they must be regarded as contrary to art 30 of the Treaty. This conclusion also holds, it goes without
saying, in relation to art 22 of the regulation.
The provision of services
40.
The same rules are in dispute and the barrier is of the same type. However, the relevant provisions of
Community law, arts 59 and 60 of the Treaty in this case, are different, as is the position of the central
figure, who, as 695 always in such cases, is still the patient. In this case, the patient derives standing in
Community law not indirectly, via the products he imports, but rather as a recipient of services, which is a
factor of some importance (see the judgment in Luisi [1984] ECR 377 (para 16))37. Given then that the
patient is covered by the provisions on freedom to provide services when travelling to another state in
order to avail himself or herself of medical treatment appropriate to his or her state of health, what are we
to make of the rules in question?
37
In that connection, moreover, I cannot overcome the suspicion that the rules on the provision of services may also be of
relevance to cases involving the purchase of products (as in the Decker case). In view of the special nature of the
products in question and also the fact that the court has acknowledged that the presence of qualified personnel is
required both for the sale of spectacles and contact lenses and for that of medicinal products (see the judgments in
Criminal proceedings against Delattre Case C-369/88 [1991] ECR I-1487 and Criminal proceedings against Monteil
Case C-60/89 [1991] ECR I-1547, both concerning pharmacists, and Laboratoire de Prothses Oculaires (LPO) v Union
Nationale des Syndicats dOpticiens de France Case C-271/92 [1993] ECR I-2899, which concerned opticians), it could
very well be argued that the rules applicable in such cases also come within the scope of arts 59 and 60 of the Treaty. In
short, I do not believe the possibility can be ruled out, at least not categorically, that the rules on the provision of
services may apply to the circumstances of Mr Decker travelling to another member state to purchase a pair of
spectacles: by virtue of the fact that the journey is undertaken not for the purpose of simply purchasing a product but to
 take advantage of the professional skills of an optician established in another member state. That said, the points made
in the text in relation to the Kohll case would clearly also be valid in relation to the Decker case if it were found that the
rules on the provision of services were applicable to that case as well.

Let me say straight away that those rules must also be regarded as being in conflict with arts 59 and
60, for reasons broadly similar to those adduced in considering their compatibility with art 30. Before
examining more closely the restrictions they place on freedom to provide services, however, I believe it
worthwhile to clear up some of the misunderstandings which have emerged during the course of the
proceedings.
41.
It has been argued that the rules in question merely lay down the terms and conditions for
reimbursement of medical expenses and that, accordingly, they are concerned only with the relationship
between the insured person and the social security institution to which he or she is affiliated, with the
result that a dispute over reimbursement of the expenses in question is purely an internal matter. The
relevant service in such circumstances, according to this argument, is the benefit provided by the social
security institution to the insured person, and not a provision of services within the meaning of arts 59 and
60 of the Treaty. In any event, it should be acknowledged that an activity financed by the state using
public funds does not involve any provision of services within the meaning of those articles.
That argument, in my opinion, betrays some confusion about the nature of the issue under
consideration. What needs to be determined is whether national rules which make reimbursement of
medical expenses incurred in a member state other than the state of residence conditional upon prior
authorisation have the effect of discouraging and hence restricting the freedom to provide services in the
sector in question. The fact that the rules in question are in the field of social security and are expressly
concerned with the terms and conditions of reimbursement of medical expenses is, in this regard, wholly
irrelevant38. Nor does the mere fact that the state is involved in financing the benefit in question mean that
there is no provision of services: the fact remains that the medical 696 treatment is supplied for
consideration39 and that the insured person bears a significant portion of the cost through his health
insurance contributions40.
38
The case law on services provides many examples of national measures which, although they were not concerned with
the provision of a service themselves, were found to be contrary to art 59 as they were liable to have an adverse effect
on the provision of the service in question. For instance, the court has held that national rules which made the grant of a
housing benefit subject to the requirement that the relevant loan was obtained from a credit institution established in the
member state in question were incompatible with art 59, for the very reason that such a requirement was liable to
dissuade borrowers from approaching banks established in other member states in order to obtain loans intended to
finance the construction, acquisition or improvement of housing (see the judgment in Svensson v Ministre du Logement
et de lUrbanisme Case C-484/93 [1995] ECR I-3955). Needless to say, in that case just as in the present case, the
rules under scrutiny did not directly concern the provision of the services in question.
39
The court has in fact made clear that it is not necessary that the service should be paid for directly by those receiving it
(see the judgment in Bond van Adverteerders v Netherlands Case 352/85 [1988] ECR 2085 (para 16)), hence the nature
of the issue does not change even if the competent social security institution pays for the service directly.
40
Medical practice is thus not comparable to public instruction in this respect (see the judgments in Belgium v Humbel
Case C-263/86 [1988] ECR 5365 (paras 46) and Wirth v Landeshauptstadt Hannover Case C-109/92 [1993] ECR I-
6447).

42.
That said, it is clear that the rules in question do not prohibit insured persons (who are recipients of
services in this case) from using a provider of services established in another member state and, in
general, do not, at least not directly, make their access to medical care in other member states subject to
conditions. Moreover, since the requirement of prior authorisation applies to all persons resident in the
member state in question wishing to travel to another member state in order to obtain the health care they
require, it is equally clear that those rules do not discriminate on grounds of nationality among the
recipients of the services in question.
None the less, the fact of the matter is that the authorisation referred to is required only where it is
sought to use a provider of services established in another member state, and this constitutes, albeit
indirectly, a difference in the treatment of insured persons based on the country of origin of the service.
Furthermore, reimbursement is denied only to insured persons who obtain a medical service in another
member state but without having secured the requisite authorisation; this constitutes a difference in the
treatment of insured persons according to whether they choose to avail themselves of services supplied
by providers established in the national territory or in another member state.
43.
In short, it is abundantly clear that the rules in question, because they make reimbursement of medical
expenses incurred in another member state subject to prior authorisation and because they deny
reimbursement of such expenses to insured persons who have not obtained such authorisation, are a
highly deterrent factor and thus entail a restriction on the freedom to provide services. Furthermore, it
cannot be denied that a situation of that kind is bound inevitably to have an adverse effect on providers of
the service in question who are not established in the state concerned 41. Save for the limited number of
instances in which authorisation is granted, they can supply only benefits which are not eligible for
reimbursement.
41
It may be recalled that art 59 requires the abolition of any restriction when it is liable to prohibit or otherwise impede
the activities of a provider of services established in another Member State where he lawfully provides similar services
(see the judgment in Sger v Dennemeyer & Co Ltd Case C-76/90 [1991] ECR I-4221 (para 12)). I need hardly add that
the nature of the issue does not change even if the rules in question do not apply to all providers of services established
in the territory but only to those approved for social security reimbursement purposes. It is settled case law that the fact
that national rules do not give preference to all domestic providers of a service is immaterial for the purposes of the
application of art 59 (see eg the judgment in EC Commission v Netherlands Case C-353/89 [1991] ECR I-4069 (para
25)).
 In the final analysis, the disputed national measure and, by the same token, art 22 of the regulation,
entail restrictions based, albeit indirectly, on where the provider of services is established 42. Both
provisions reduce very substantially the freedom of health care consumers (recipients of services) to use
(also)
42
In that regard, the General Programme for the abolition of restrictions on freedom to provide services (OJ 1974 S edn
(II) IX, p 3) included among the restrictions to be abolished those which impinge indirectly on providers of services, for
example via their effect on the recipient of the service.

697
providers of services established in other member states and thereby hinder the latters cross-border
trade. Clearly, such restrictions are manifestly contrary, at least in principle, to art 59 of the Treaty 43.
43
This conclusion, that art 22 of the regulation is also, on the face of it, contrary to the Treaty rules on services, is one
which commands widespread support among academic writers. See, inter alia, Bosscher La seguridad social de los
trabajadores migrantes en la perspectiva del establecimiento del mercado interior in Los sistemas de seguridad social y
el mercado nico europeo (1993) p 23 esp at pp 31ff; and Cornelissen The Principle of Territoriality and the Community
Regulations on Social Security (1996) CML Rev 439 esp at 463466.

IIIThe reasons advanced to justify the contested measure

44.
Given that it is the same measure which creates barriers of the same kind to both the free movement
of goods and the freedom to provide services, it is hardly surprising that the reasons advanced to justify it
are essentially the same in the case of both freedoms. To be precise, the Luxembourg government and
most of the governments which have submitted observations in the two cases contend that the
restrictions entailed by the contested rules are necessary on public health grounds and, in particular, to
ensure the maintenance of a balanced medical and hospital service accessible to everyone.
As already indicated, appraisal of those justifications necessarily entails taking a view on the validity of
art 22 of the regulation. Once we have established that this provision, by imposing the requirement of
prior authorisation, produces the same restrictive effects as the national measure in issue, we must
perforce ascertain whether or not, like the national measure, it is compatible with arts 30 and 59 of the
Treaty44. The outcome would be different only if the two measures (national and Community), while laying
down essentially the same rules, had different aims and/or followed different schemes.
44
For the sake of completeness, I would point out that the European Commission on Human Rights has given an opinion
on the relevant Luxembourg rules and on art 22 of Regulation 1408/71, in a case in which the plaintiffa Luxembourg
national who had been refused by the health insurance fund reimbursement of medical expenses incurred in another
member state (Belgium) and for which she had not obtained prior authorisationalleged a violation of art 6(1) of the
Convention for the Protection of Human Rights and Fundamental Freedoms (Rome, 4 November 1950; TS 71 (1953);
Cmd 8969) on account of the excessive length of the procedure and the fact that the dismissal of her appeal on the
ground of inadmissibility by the Cour de Cassation, a court bound by the obligation to make a reference for a preliminary
ruling under art 177 of the EC Treaty, had deprived her of the possibility of a ruling by the Court of Justice on the
interpretation of art 59 of the Treaty, with which, she contended, the national rules were manifestly in conflict. The
Commission on Human Rightshaving noted that the health insurance fund and the Luxembourg courts had dismissed
her action on the ground that the plaintiff could not reasonably claim, under art 60(3) of the Code des Assurances
Sociales, art 51 of the Treaty and art 22 of Regulation 1408/71, reimbursement of medical expenses incurred in another
member state without having sought and obtained the prescribed authorisation in advancestated that the legislation
applicable to the case did not confer on the plaintiff an entitlement to be reimbursed for health care received in Belgium.
Essentially, the Commission on Human Rights, having interpreted art 22 of the regulation as not conferring entitlement
to reimbursement, drew the conclusion that the plaintiff did not hold a right under art 6(1) of the 1950 convention and
accordingly held the action to be inadmissible (see Mnchen v Luxembourg (Case No 28895/95) (16 April 1996,
unreported).

45.
That said, in order to determine what type of justification is permissible it first needs to be decided
whether the contested measure is to be categorised as formally discriminatory or as indistinctly
applicable. In the former case it can be justified, and hence upheld as compatible with Community law,
only if it comes within the scope of art 36 (goods) and art 56, referred to by art 66 (services), in other
words, if it comes within one of the derogations expressly provided for by the Treaty 45, with the further
consequence that no consideration may be given to economic aims pursued by means of the restrictive
measure46 but only, in the present instance, to the protection of public health.
45
In relation to goods, see EC Commission v Ireland Case 113/80 [1981] ECR 1625 (paras 8, 11), and, more recently,
Criminal proceedings against Pistre Joined cases C-321/94 and C-324/94 [1997] ECR I-2343 (para 52). In relation to
services, it was only in its judgment in Bond van Adverteerders [1988] ECR 2085 (para 32) that the court for the first
time clearly stated that national rules which are discriminatory are compatible with Community law only if they can be
brought within the scope of an express derogation; on the same point, see also the more recent case of Svensson
[1995] ECR I-3955 (para 15).
46
As regards goods, see Duphar [1984] ECR 523 (para 23) and EC Commission v Italy Case 7/61 [1961] ECR 317 at 329.
As to services, see, inter alia, Bond van Adverteerders [1988] ECR 2085 (para 34), and, more recently, Federacin de
Distribuidores Cinematogrficos (Fedicine) v Estado Espaol Case C-17/92 [1993] ECR I-2239 (paras 16, 21).

In the latter case, on the other hand, there is a broader range of requirements pertaining to the general
interest capable of justifying the measure. The court has held that in order to be compatible with
Community law national measures liable to hinder or make less attractive the exercise of fundamental
freedoms 698 guaranteed by the Treaty must fulfil four conditions: they must be applied in a non-
discriminatory manner; they must be justified by imperative requirements in the general interest; they
must be suitable for securing the attainment of the objective which they pursue; and they must not go
beyond what is necessary in order to attain it 47. This means, first, that even measures applicable without
distinction may be incompatible with Community law if they are restrictive and not justified by overriding
requirements or reasons relating to the general interest 48; and, secondly, as we shall see below, that in
this latter case consideration may also be given to economic aims pursued by means of the restrictive
measure, such as the safeguarding of the financial stability of the health care system.

47
See the judgment in Gebhard v Consiglio dellOrdine degli Avvocati e Procuratori di Milano Case C-55/94 [1996] All
ER (EC) 189, [1995] ECR I-4165 (para 37), in which, significantly, the court referred without distinction to all the
fundamental freedoms enshrined in the Treaty, thus highlighting the unitary nature, for the purposes which concern us
here, of the relevant rules.
48
This approach, which when first adopted, in the judgment in Rewe-Zentral AG v Bundesmonopolverwaltung fr
Branntwein Case 120/78 [1979] ECR 649 (Cassis de Dijon), was confined to the free movement of goods, was
subsequently extended to the other fundamental freedoms enshrined in the Treaty. As far as services are concerned,
that approach was expressly adopted, in particular, in Stichting Collectieve Antennevoorziening Gouda v Commissariaat
voor de Media Case C-288/89 [1991] ECR I-4007 (paras 1115), EC Commission v Netherlands Case C-353/89 [1991]
ECR I-4069 (paras 1519), and Sger [1991] ECR I-4221 (para 15). Given, however, that since its very first decisions
on the subject the court has upheld as justified in the general interest measures restricting freedom to provide services
(see eg van Binsbergen v Bestuur van de Bedrijfsvereniging voor de Metaalnijverheid Case 33/74 [1974] ECR 1299), it
is clear that the Gouda, Commission v Netherlands and Sger judgments, far from constituting a new departure, merely
clarified the existing approach and provided it with a fuller theoretical analysis.

46.
However, the parties and the governments which submitted observations have, in justifying the
contested measure, referred without distinction both to the protection of public health, and thus to one of
the derogations provided by arts 36 and 56, and to the safeguarding of the financial stability of the health
care system, which is a reason relating to the general interest. The Commission, for its part, submits that
the measure is discriminatory as regards goods, since prior authorisation is not required to purchase the
relevant products in the national territory, but indistinctly applicable as regards services, since the
authorisation requirement, although it applies only where medical services are sought abroad, applies in
the same manner to nationals and non-nationals alike. It would seem, therefore, that under the rules laid
down for goods even measures which, while not prescribing any particular formalities for the purposes of
importation as such, are likely to discourage imports, are discriminatory, whereas under the rules
governing services only measures which involve different treatment based on nationality are
discriminatory49.
49
This view has in fact been superseded by developments in the case law. The court also regards as discriminatory, and
hence as justifiable only in accordance with one of the derogations set out in art 56, national rules which treat services
differently according to their origin (see, inter alia, the judgments in Bond van Adverteerders [1988] ECR 2085 (paras
26, 29), EC Commission v Belgium Case C-211/91 [1992] ECR I-6757 (paras 911) and Fedicine [1993] ECR I-2239
(para 14)).

699
I cannot subscribe to this analysis. The measure in question does not directly hinder trade either in
goods or in services. It may, however, have the effect of limiting such trade by reason of the obvious
disadvantage (no reimbursement) in purchasing products or medical services abroad, without
authorisation. What needs to be determined, therefore, is whether restrictions on the free movement of
goods and on the freedom to provide services based, albeit indirectly, on where the seller of the goods or
provider of the service is established, fall to be classified as discriminatory or as indistinctly applicable.
47.
So far as concerns the rules on goods, let me begin by pointing out that this is the first time that the
court has been called upon to determine whether a measure which imposes an additional burden (prior
authorisation) only on those intending to purchase the products in question outside the national territory is
discriminatory for the purposes and within the meaning of art 30 of the Treaty. Given that authorisation is
required solely for the purpose of conferring a benefit (total or partial reimbursement of the cost of a
particular product) and not for the purpose of importation 50, it must be recognised, however, that the
measure in question does not lay down different rules for imported products, but rather entails a
difference in the treatment of persons (the insured), all resident in the same member state, according to
whether they have chosen to purchase a particular product in their state of residence or in some other
member state. Let me add that the fact that the difference in the treatment of insured persons depends,
albeit indirectly, on the place where the optician or pharmacist who supplies the products is established, is
of no significance for the purposes of the rules on goods, even if it may be considered to constitute formal
discrimination based on the place of establishment51.
50
There still is complete freedom as regards actual importation. See para 37, above, and esp footnote 34.
51
This would be a relevant factor, however, if it were decided that the rules on the provision of services were also
applicable to the Decker case. See footnote 37, above.

700
Any restriction on imports stems from the deterrent effect, on insured persons decisions, of the fact
that the cost of products purchased in another member state is not reimbursed. It is my view, therefore,
that the measure in question, even though it favours the purchase of products sold in the national territory,
is not formally discriminatory. Leaving aside all other considerations, it requires no specific formality to be
complied with at the time of and for the purpose of importation, nor does it favour domestic products over
imported products.
48.
Turning to the scope of the contested national measure in relation to the provisions on services, I
would first recall that the court has consistently held that measures are formally discriminatory only where
they prescribe different rules for non-nationals52 and/or for the provision of services originating in other
member states (see footnote 49, above). It regards as indistinctly applicable, on the other hand,
measures capable of applying to all those who carry on a particular activity in the territory of a particular
member state, even if such measures expressly impose a requirement of residence 53 or establishment54
which effectively makes it impossible for service providers established in another member state to carry
on the activity in question55.
52
See the judgments in Hubbard v Hamburger Case C-20/92 [1993] ECR I-3777 (paras 1415) and EC Commission v
Spain Case C-45/93 [1994] ECR I-911 (paras 910).
53
See eg the judgments in van Binsbergen [1974] ECR 1299 (para 14) and Coenen v Sociaal-Economische Raad Case
39/75 [1975] ECR 1547 (paras 7/8, 9/10).
54
See esp the judgments in EC Commission v Germany Case 205/84 [1986] ECR 3755 (paras 5257) and European
Commission v Italy Case C-101/94 [1996] ECR I-2691 (para 31).
55
See, most recently, the judgment in Socit Civile Immobilire Parodi v Banque H Albert de Bary et Cie Case C-222/95
[1997] All ER (EC) 946, [1997] ECR I-3899 (para 31), in which the court again emphasised that the requirement of
establishment has the result of depriving art 59 of the Treaty of all effectiveness, a provision whose very purpose is to
abolish restrictions on the freedom to provide services of persons who are not established in the state in which the
service is to be provided. The court added, however, that if such a requirement is to be accepted, it must be shown that
it constitutes a condition which is indispensable for attaining the objective pursued.

The measure under consideration, it is worth noting, does not entail any discrimination based on
nationality nor does it prescribe, at least not directly, separate rules for service providers established in
another member state. The difference in treatment concerns, at least formally, all those insured under the
social security scheme in question. Bearing in mind, however, that different treatment depends on the
choice of doctor or of hospital, it is quite clear that the difference in the way insured persons are treated
depends on the place where the provider of the service is established. Is this sufficient to warrant a
finding that the measure in question, in so far as it results in different rules applying in respect of the
provision of services originating in another member state, is in any event to be categorised as
discriminatory?
49.
A negative answer to this question could be deduced from the fact, referred to above, that, under the
courts case law on services, rules requiring a person to be established in the member state in question in
order to carry on a given activity are justifiable on grounds relating to the general interest, by reason of
the fact, however unsatisfactory this result may seem, that the requirement is imposed on all those who
carry on a particular activity in the territory of a particular member state and is thus, formally at least, not
in any way discriminatory. From the same perspective, a measure which does not prescribe, at least not
formally and directly, different rules for providers of services established in another member state, should
likewise be regarded as non-discriminatory.
701
On this very point, however, the case law on measures of the kind at issue in this case, in other words
measures by which the conferring of a benefit is made conditional upon a provider of services established
in the national territory being used, is not free from uncertainties and contradictions. For example, in
Bachmann v Belgium Case C-204/90 [1992] ECR I-249 the court held that the contested measure, which
made the deductibility of certain contributions subject to the condition that they were paid in the member
state itself, was justifiable in the interests of the cohesion of the national tax system, thereby classifying
the measure as indistinctly applicable56. In the later case of Svensson [1995] ECR I-3955 (para 15), on
the other hand, the court ruled that a measure which specified as a condition for the grant of a housing
benefit that the loan should be obtained from a lending institution established in the national territory was
discriminatory on the ground of the place of establishment, and thus justifiable only on the basis of the
derogations expressly provided for in the Treaty57. Nor did it help to clarify matters that the court then
considered it necessary, in the same judgment, to rebut the proposition that the measure in question was
necessary in order to preserve the cohesion of the tax system: in so doing, the court was also verifying
whether that measure could be justified on the basis of requirements likely to be taken into account only
in the case of measures applicable without distinction (see [1995] ECR I-3955 (paras 1618)).
56
To this end, it is worth pointing out, the court confined itself to noting that the requirement of an establishment is
compatible with Article 59 of the Treaty where it constitutes a condition which is indispensable to the achievement of the
public-interest objective pursued (see [1992] ECR I-249 (para 32)). It would appear from that statement that since the
general interest can justify even a prohibition, imposed by means of a requirement of establishment, on exercising
particular activities in the territory of the state concerned, there is all the more reason to regard as justified rules which
entail some disadvantages, or at any rate do not confer certain advantages, for those who choose to avail themselves of
providers of services not established in the state in question.
57
More precisely, the court stated that the rule in question entails discrimination based on the place of establishment and
that accordingly it can only be justified on the general interest grounds referred to in Article 56(1) of the Treaty, to which
Article 66 refers, and which do not include economic aims.

50.
I do not believe that the compatibility of a national measure with Community law may be assessed
without distinction and/or cumulatively with regard to derogations expressly provided for in the Treaty and
to grounds relating to the general interest58, particularly as the justifications in question apply in different
situations which are in part governed by different rules. While acknowledging that it may not be an easy
matter to establish whether or not a particular measure is discriminatoryeither because the
discriminatory effect (in fact) is all too obvious, or because the particular features of a given sector induce
a degree of cautionit is essential none the less to place the measure in the proper category.
58
On this point, it should be emphasised that Svensson is not the only case, nor the first, in which the court took into
consideration, for the purpose of determining whether a national measure restricting freedom to provide services was
 compatible with art 59, both the requirements set out in art 56 and those relating to the general interest (see eg the
judgment in European Commission v Italy Case C-101/94 [1996] ECR I-2691 (paras 3132)).

In short, it is my view that, for the purposes of this case and in order to dispel the aforementioned
ambiguities in the case law, the court should either follow Svensson and rule that discriminatory measures
include those which indirectly give rise to unequal treatment as between providers of services established
in a given member state and those not so established or, alternatively, confirm the 702 approach which
appears to have been adopted in Bachmann, namely that measures which do not formally lay down
different rules for providers of services not established in the member state are still deemed to be
indistinctly applicable. I would merely add that the latter approach would appear at present to be more in
keeping with the case law in this area considered as a whole. This does not of course alter the fact that a
reappraisal would be timely.
51.
Assuming that the national measure under consideration is to be regarded as indistinctly applicable,
with respect to the Community rules on both goods and services, I now turn to the requirements relied
upon to justify it, namely the protection of public health and the maintenance of a balanced medical and
hospital system accessible to everyone in a given region. The latter requirement, it may be recalled, is
expressly mentioned in the second question submitted by the national court in Case C-158/96 (Kohll),
while the member states which submitted observations have referred instead to preserving the systems
financial stability. The two aims are in fact interrelated in that the latter, more immediate, objective
constitutes, according to those member states, the means by which the former is to be achieved.
One further aspect, I believe, needs to be made clear. The points raised and the result arrived at in
examining whether the contested measure is justifiable on public health grounds would be equally valid,
given that it is a requirement expressly provided for in arts 36 and 56 of the Treaty, if the contested
measure were assumed to be discriminatory (see esp the judgments in Aragonesa de Publicidad Exterior
SA v Departamento de Sanidad y Seguridad de la Generalitat de Catalua Joined cases C-1/90 and C-
176/90 [1991] ECR I-4151 (para 13)). But the same cannot be said for the requirement of maintaining a
balanced medical and hospital system accessible to everyone in a given region. Even granting that this
aim comes within the concept of public health, in the broad sense of the term 59, the fact remains that it is
based entirely on the financial stability of the system and thus on an economic aim, which, as such, can
under no circumstances justify a discriminatory measure60.
59
But the court has so far given a narrow interpretation to the concept of public health and has allowed it to be used to
justify only measures aimed at preventing risks to the health of humans and animals.
60
See the judgments cited in footnote 46. I would merely note that in Duphar [1984] ECR 523 (para 23) the court made
clear that because art 36 relates to measures of a non-economic nature, it cannot therefore justify a national measure
intended to reduce the operating costs of a sickness insurance scheme.

52.
That said, let me add straight away that the proposition, also put forward by the Luxembourg
government, that the contested rules are necessary to ensure the quality of medical products and
serviceswhich, where travel abroad is involved, can only be checked when authorisation is requested
is devoid of any foundation. The conditions for taking up (and pursuing) the activities with which we are
concerned here, like the rules governing the release of medicinal products on to the market, have been
the subject of Community coordinating and harmonising directives 61. Indeed, the court has already cited
those directives in judgments where it stated that the prescription of a medicinal product by a doctor in
another member state and the purchase of such a product in that state provide guarantees equivalent to
those afforded where prescription is by a doctor in the importing state or the product is sold through a
pharmacy in the member state into which it is imported by a private individual (see the judgments in
Schumacher [1989] ECR 617 (para 20) and EC Commission v Germany Case C-62/90
61
I refer, in particular, to Council Directive (EEC) 93/16 to facilitate the free movement of doctors and the mutual
recognition of their diplomas, certificates and other evidence of formal qualifications (OJ 1993 L165 p 1) and, among the
specific directives, to Council Directives (EEC) 78/686 and 78/687 concerning the mutual recognition of diplomas and
the coordination of national provisions on dentists (OJ 1978 L233 pp 1, 10). I would also refer to Council Directives
(EEC) 85/432 and 85/433 on the coordination of national provisions and the mutual recognition of diplomas in relation to
certain activities in the pharmaceutical sector (OJ 1985 L253 pp 34, 37). Finally, I would recall that the free movement of
pharmaceutical products is ensured at Community level, with due account taken of the protection of public health, by
Council Directive (EEC) 65/65 (OJ S edn 196566 p 20).

[1992] ECR I-2575 (para 18)). That case law, which of necessity also applies to the purchase of all
medical and surgical accessories, including a pair of spectacles, presupposes that guarantees equivalent
to those provided by doctors, pharmacists and opticians established in the national territory must be
assumed to be provided by doctors, pharmacists and opticians established in other member states as
well.
In those circumstances, I do not believe it is reasonable to argue that the health of insured persons is
not adequately protected when they obtain health care in another member state 62. It follows, of course,
that the measure in question cannot be regarded as necessary for the protection of health.
62
Since the Luxembourg government itself stated that insured persons are free to obtain health care in another member
state, the question has to be asked why the aforesaid public health concerns no longer arise when reimbursement is not
sought.

53.
So far as preserving the systems financial stability is concerned, let me begin by stating that this is, in
my view, a requirement worthy of protection by Community law. While it is true that the court has on
occasion categorically dismissed economic aims put forward to justify indistinctly applicable measures (as
well)63, it is also apparent, on a closer reading of the relevant judgments, that economic aims are indeed
justifiable, where far from being an end in themselves, they are crucial to the operation of the system in
question64 or affect interests of vital importance to the state 65.
63
See the judgment in Gouda [1991] ECR I-4007 (para 11). See also, more recently, the judgment in Syndesmos ton en
Elladi Touristikon kai Taxidiotikon Grafeion v Ergasias Case C-398/95 [1997] ECR I-3091 (para 23), in which the court
stated that maintaining industrial peace as a means of bringing a collective labour dispute to an end and thereby
preventing any adverse effects on an economic sector, and consequently on the economy of the state, must be
regarded as an economic aim which cannot constitute a reason relating to the general interest that justifies a restriction
of a fundamental freedom guaranteed by the Treaty.
64
I recall, for example, that in the judgment in Customs and Excise Comrs v Schindler Case C-275/92 [1994] 2 All ER
193, [1994] ECR I-1039 (para 60), the court stated that a ground which is not without relevance, although it cannot in
itself be regarded as an objective justification, is that lotteries may make a significant contribution to the financing of
benevolent or public interest activities such as social works, charitable works, sport or culture (my emphasis). I would
further refer to the judgment in Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-
415/93 [1996] All ER (EC) 97, [1995] ECR I-4921 (paras 106107), in which the court upheld as legitimate the aim
of maintaining a financial and competitive balance between football clubs as its purpose was to preserve a degree of
equality and uncertainty as to results.
65
This is one possible interpretation of the courts acceptance of the cohesion of the tax system, which is undoubtedly an
economic aim, as a ground related to the general interest (see the judgment in Bachmann [1992] ECR I-249 (paras 21
28)).

From that point of view, I believe it is beyond dispute that the preservation of the financial stability of
the social security system, which is indeed the essential aim of the measure in question, is not an end in
itself but a means which contributes (at least) to providing insured persons with services of a certain
standard in terms of both quantity and quality. If the financial balance of the system were upset, the level
of health protection could deteriorate with obvious and inevitable adverse repercussions, particularly for
insured persons belonging 703 to the weakest strata of society. Moreover, the court has already
recognised, for example in the matter of equal treatment of men and women in the social security
sector66, and also in interpreting the relevant provisions of the Staff Regulations of Officials, that the
financial stability of social security systems constitutes a legitimate objective and, accordingly, one worthy
of protection (see, most recently, Barassi v EC Commission Case T-41/90 [1992] ECR II-159 (paras 32
35)).

66
See the judgment in R v Secretary of State for Health, ex p Richardson Case C-137/94 [1995] All ER (EC) 865,
[1995] ECR I-3407 (paras 1829).

54.
I now turn to consider whether the contested measure is necessary for and suited to preserving the
financial stability, and thereby the continued operation, of a medical and hospital service accessible to
everyone. All the governments which submitted observations have argued, on this issue, that prior
authorisation is essential in order to avoid the social security institutions having to reimburse the cost of
the benefits in question in accordance with the legislation of the member state in which each individual
insured person chooses to avail himself or herself of medial care and/or to purchase medical products.
They add that the member state chosen will normally be the one in which the benefit in question is
considered most advanced and is, accordingly, the most expensive.
Now it must of course be acknowledged that an obligation to reimburse, in accordance with the
legislation of the different member states, the cost of benefits which insured persons have chosen to avail
themselves of outside the national territory, could indeed upset the financial stability and adversely affect
the continued operation of a balanced medical and hospital system accessible to everyone. In that regard,
suffice it to note that, at present, the costs of health care and the rules governing the financing of health
insurance systems differ appreciably from one member state to another 67. Unrestricted freedom of
movement for patients could therefore lead to imbalances which would be highly detrimental to those
member states in which the cost of benefits is significantly lower than the cost which their social security
institutions would have to reimburse to insured persons who chose to avail themselves of such benefits in
other member states.
67
For an overview of current differences, see Le Grand La asistencia sanitaria y la construccin del mercado nico:
perspectiva y problemtica in Los sistemas de seguridad social y el mercado nico europeo (1993) pp 321ff. See also,
in relation to the more general issue of the funding of social protection, Euzeby Le financement de la protection sociale
dans les pays de la CEE: problmes et perspectives in Quel avenir pour lEurope sociale: 1992 et aprs? (1990) pp
133ff, and, by the same author, Financement de la protection sociale, efficacit conomique et justice sociale in (1997)
Revue du March commun et de lUnion europenne 253ff.

55.
In view of this, the contested measure must be regarded as necessary and proportionate to the
attainment of the aim pursued. The requirement of prior authorisation is the only means which enables
the social security institutions to assume the burden of expenses incurred by insured persons in another
member state, on the terms laid down in that member state, only where it is adjudged necessary because
of the state of health of the persons insured and thus to contain excessive expenditure which could upset
the financial balance of the system.
The Commission, however, arguing that art 22 of Regulation 1408/71 imposes an obligation not to
refuse authorisation in a particular case but is silent as to other situations which might arise, maintains
that each case has to be looked 704 at individually to determine whether the grant of authorisation for a
particular benefit, and the attendant financial burden, would entail upsetting the financial balance 68. As
attractive as this approach may appear, it is lacking in rigour. It is clearly the case, and rightly so, that
member states could (and indeed can) be more generous than is required of them by art 22 of the
regulation; I wonder, however, on the basis of what criteria national rules which authorise insured persons
to travel to other member states to avail themselves of medical care, at least in the cases in which art 22
so requires, can be regarded as compatible with that provision, which the Commission itself deems valid,
but contrary to the Treaty rules on services69.
68
This point, it should be noted, would apply only to services. Having maintained that the contested measure was
discriminatory in relation to the rules on goods, the Commission was consistent and submitted that the measure in
question was not justifiable on the ground of protection of public health set out in art 36 of the Treaty.
69
While I do not share the authors view (see para 31 and footnote 36, above), I find more convincing the view put forward
by da Mavridis Le citoyen europen peut-il se faire soigner dans ltat de son choix? (1996) Droit social 1086ff,
according to which it is art 22 itself, as interpreted by the court in Pierik No 1 and Pierik No 2, which requires
authorisation to be granted for all care liable to ensure effective treatment of the disease from which the person
concerned is suffering.

56.
Rather, it is quite true that art 22 in no way requires, or implies, that in cases in which authorisation is
not granted on the ground that the conditions it lays down are not satisfied, insured persons must bear the
full cost of the benefits concerned or are not entitled to any form of reimbursement. The purpose of the
provision in question is to ensure that the right of the insured to certain benefits may, by derogation from
the principle that a person is subject to a single body of legislation, which is the basis for the coordination
brought about by the regulation, also be exercised in another member state (at least) in cases where a
transfer is necessary because of the state of health of the person concerned. In other words, the provision
in question is intended to guarantee to insured persons the right to receive appropriate care in another
member state without suffering financially as a result and without endangering the stability of the systems
in operation in the various member states.
57.
But if this is true it follows, first, that the restrictions imposed by the relevant provisions of art 22,
precisely because they are aimed at ensuring that insured persons may avail themselves of medical
benefits in another member state on the terms laid down by the legislation of that state, are objectively
justified and hence compatible with arts 30 and 59 of the Treaty and, secondly, that the preservation of
financial stability, the aim pursued by the national measure, is relevant only if and to the extent that it is
assumed that reimbursement by the competent social security institution of benefits which insured
persons choose to obtain in another member state must be based on the rules and criteria applied by the
member state in which the benefits are provided.
That being so, it must therefore be considered whether the risk of upsetting the financial balance, and
hence the need for prior authorisation, still subsists if the competent social security institution is obliged to
make full reimbursement in accordance not with the legislation of the member state in which the insured
person received the care in question, but with the legislation and the tariffs in force in the member state in
which the recipient of the medical care is insured. In other words, the question is whether reimbursing Mr
Decker, as he claims, on the same terms as those to which he would have been entitled had he bought
the spectacles from an optician established in Luxembourg, or reimbursing Mr Kohll 705 on the basis of
the Luxembourg tariff, would really have the effect of undermining the financial balance of the system.
58.
Viewed in those terms, it is clear that differences in financing methods and in health care costs
between the various systems are immaterial for as long as the competent social security institution does
not have to reimburse benefits received by insured persons in another member state in accordance with
the legislation applicable in that state. It is self-evident that reimbursement to Mr Decker of the sum of Lf
1,600, to which he would have been entitled had he purchased the spectacles from an optician
established in Luxembourg, would have no effect on the financial balance of the system, any more than
would be the case in the event of reimbursement of orthodontic treatment provided in Germany to a
person insured by the Luxembourg health insurance fund in accordance with the Luxembourg cost of
such treatment70.
70
The only effect I can conceive of is that one optician established in Luxembourg will have sold one less pair of
spectacles and the only orthodontist established in the same state will have lost one patient. It is therefore the individual
practitioners who are adversely affected and not the system itself.

Nor, in consequence, would the maintenance of a medical service accessible to everyone in a given
region be placed in jeopardy. While it is true that those living in border regions may cross the border more
frequently, inter alia for medical treatment, the fact remains that the traffic is not all one-way, at least in
principle, and, especially, that the benefits we have been considering do not affect the financial balance to
a greater extent than if they had been obtained in the member state of residence of the persons
concerned.
59.
I believe, however, that the position in relation to hospital infrastructure is quite different and requires a
different answer. Unlike the benefits provided by individual practitioners, the reality in the case of hospitals
is, first, that their location and number is determined by forward-planning and, secondly, that the cost of
one persons stay in a hospital cannot be separated from that of the hospital as a whole. Clearly, if a large
number of insured persons chose to avail themselves of hospital facilities located in another member
state, their domestic hospitals would be under-utilised but would have the same staff and equipment
overheads as if they were being used to full capacity71.
71
Of course, this does not mean I wish to associate myself with the view advanced by some member states that, if
freedom to choose ones doctor and hospital were given, there would be an uncontrolled and uncontrollable flow of
 patients from one member state to another. The fact remains that travelling to a state other than the state of residence
entails significant inconvenience, often including language difficulties, and additional costs, if only for those persons
accompanying the patient.

In other words, reimbursement, by the competent social security institution, of medical benefits
received by insured persons in hospitals of other member states, even on the basis of a flat-rate
equivalent to the Luxembourg cost of the benefits in question, would none the less result in an additional
financial burden for the system concerned. In the hospital sector, therefore, it remains essential, in order
not to disrupt the financial balance of the system and to ensure the maintenance of a service which is
accessible, both financially and logistically, to everyone, including therefore those who do not intend to
travel but wish to receive the health care they require in the place nearest to them, to seek and obtain
prior authorisation72.
72
On this point, however, the question arises whether, and to what extent, the situation is otherwise, in relation to the aim
of maintaining a hospital system accessible to everyone, in those member states where the competent social security
institution provides reimbursement, albeit on a partial and flat-rate basis, of the cost of benefits which insured persons
choose to obtain in private clinics.

60.
In short, it is my view that the contested measure is justified in relation to all benefits which must be
provided to insured persons in hospitals and, in 706 general, to all benefits which the insured person
wishes to have paid or reimbursed in full by the competent social security institution. On the other band,
the measure is not justifiable in relation to the purchase of products or medical services provided by
private practitioners, consisting of consultations and visits to specialists, in respect of which a flat-rate
reimbursement is requested at the tariff in force in the member state in which the person concerned is
insured.
Action by the Community legislature aimed at harmonising the area in question so as to allow genuine
and effective freedom of movement for patients, which would be a significant factor in the creation of a
single integrated market, would be welcome. Conscious of the fact that this is an ambitious goal which is
difficult to achieve at present, I believe that the Community legislature should at least act, and do so
promptly, to broaden the range of circumstances in which authorisation may not be refused. There is no
doubt that it would be advantageous in many respects for authorisation to be granted in all cases in which
the insured person could receive more effective treatment in another member state or where, as was
explained in the course of the hearing in the Kohll case, there is just one specialist in the member state of
residence capable of providing the service required.
61.
In the light of the foregoing considerations, I suggest that the Court of Justice answer the questions
referred to it by the Luxembourg Conseil Arbitral des Assurances Sociales and the Cour de Cassation
respectively, as follows:

CONCLUSION

(a) Case C-120/95 (Decker)

Articles 30 and 36 of the Treaty are to be interpreted as precluding national rules under which a
social security institution refuses to reimburse to an insured person, on the ground that prior
authorisation is required for any health care abroad, the cost of a pair of spectacles, prescribed by
a doctor established in the member state in which the insured person resides but purchased from
an optician established in another member state, in accordance with the legislation applicable in
the former state.

(b) Case C-158/96 (Kohll)

Articles 59 and 60 of the Treaty are to be interpreted as precluding national rules under which
reimbursement of the cost of benefits is subject to authorisation by the insured persons social
security institution if the benefits are provided in a member state other than the state in which that
person resides, on condition that and in so far as such benefits are provided outside the hospital
infrastructure and the cost is reimbursed in accordance with the rules not of the state in which they
are provided but of the state of insurance; on the other hand, arts 59 and 60 of the Treaty are to be
interpreted as not precluding such rules where their purpose is the preservation of the financial
balance along with the maintenance of a 707 hospital service accessible to everyone in a given
region, where the benefits in question must be provided in hospitals and, in general, in respect of
all benefits the cost of which is to be reimbursed in accordance with the rules of the member state
in which they are provided, rather than the member state in which the insured person resides.

28 April 1998.

The COURT OF JUSTICE

delivered the following judgment in Decker v Caisse de Maladie des Employs Privs Case C-120/95.
1.
By decision of 5 April 1995, received at the Court of Justice of the European Communities on 7 April
1995, the Conseil Arbitral des Assurances Sociales (the Social Insurance Arbitration Council),
Luxembourg, referred to the Court of Justice for a preliminary ruling under art 177 of the EC Treaty a
question on the interpretation of arts 30 and 36 of that treaty.
2.
That question was raised in proceedings between Mr Decker, a Luxembourg national, and the Caisse
de Maladie des Employs Privs (the fund) concerning a request for reimbursement of the cost of a pair
of spectacles with corrective lenses purchased from an optician established in Arlon, Belgium, on a
prescription from an ophthalmologist established in Luxembourg.
3.
By letter of 14 September 1992, the fund informed Mr Decker that it would not reimburse him the cost
of those spectacles, on the ground that they had been purchased abroad without its prior authorisation.
4.
Mr Decker contested that decision, relying in particular on the Treaty rules on the free movement of
goods. Upon hearing his complaint, the fund maintained its position by decision of its managerial
committee of 22 October 1992 and so rejected his claim.
5.
Mr Decker submitted an application to the Conseil Arbitral des Assurances Sociales, which rejected it
by order of 24 August 1993.
6.
By application of 8 September 1993, Mr Decker appealed against that order to the Conseil Arbitral des
Assurances Sociales, which dismissed the appeal by decision of 20 October 1993, on the ground in
particular that the matter was connected not with the free movement of goods but with social security law,
that is, Council Regulation (EEC) 1408/71 on the application of social security schemes to employed
persons, to self-employed persons and to members of their families moving within the Community (see
the version amended and updated by Council Regulation (EC) 118/97; OJ 1997 L28 p 1).
7.
Mr Decker appealed to the Cour de Cassation (the Court of Cassation). By judgment of 12 January
1995, the contested decision was set aside and the case was remitted to the Conseil Arbitral des
Assurances Sociales. By judgment of 5 April 1995, it held that art 60 of the Code des Assurances
Sociales (the Social Insurance Code) and art 58 of the statutes of the Union des Caisses de Maladie des
Salaris (the UCM) applied to the dispute.
8.
Article 60 of the Luxembourg Code des Assurances Sociales, in the version in force at the material
time, provided in particular:
Insured persons shall be entitled to approach the doctor, dentist, pharmacist, hospital or
medical auxiliary of their choice. Only the following may provide treatment and services on the
territory of the Grand Duchy: (1) doctors, dentists, pharmacists, hospitals, midwives, medical
auxiliaries authorised to practise their profession in all or part of the Grand Duchy; (2) foreign
doctors consulted in the Grand Duchy with the agreement of the attending doctor and the medical
adviser, without prejudice to wider 708 international arrangements. However, insured persons may
obtain treatment abroad only with the consent of their sickness fund, except in the case of initial
treatment in the event of accident or illness occurring abroad. The sickness fund may not refuse
consent if the treatment abroad is recommended by the doctor attending the insured person and a
medical adviser, or if the treatment needed is not available in the Grand Duchy.
9.
Reimbursement of the cost of spectacle frames and corrective lenses was governed at the material
time by art 78 of the UCM statutes and by the collective agreement of 30 June 1975 concluded pursuant
to art 308 bis of the Code des Assurances Sociales between the UCM and the professional grouping
representing opticians.
10.
Article 78 of the UCM statutes states:
The cost of spectacles and other visual aids shall be borne by the sickness fund up to the
amounts stated in the tariffs and in accordance with the conditions determined in the agreements or
decisions in lieu thereof in accordance with Article 308 bis of the Code des Assurances Sociales.
11.
Article 2 of the collective agreement of 30 June 1975 provides that, without prejudice to Community
and international provisions concerning social security of migrant workers and persons treated as such,
spectacles are to be supplied to insured persons, in so far as they are permanently or actually resident in
Luxembourg, by opticians who are registered in the Luxembourg register of trades and established in the
Grand Duchy.
12.
Under those provisions, reimbursement was on a flat-rate basis with a ceiling of Lf 1,600 for frames.
13.
For corrective lenses, the reimbursement tariffs were fixed in Annex A to the collective agreement of 30
June 1975. Under art 12 of that agreement, the amounts capable of reimbursement for corrective lenses
fixed in Annex A were to be adjusted up or down by reference solely to the price lists of the firms Zeiss
and American Optical.
14.
The Code des Assurances Sociales and the UCM statutes were substantially amended in 1992.
However, the principle set out in the old art 60 of the Code des Assurances Sociales, relating to prior
authorisation by the sickness fund for all medical treatment abroad, was incorporated in the new art 20 of
the code.
15.
Article 22 of Regulation 1408/71 provides in particular:
(1) An employed or self-employed person who satisfies the conditions of the legislation of the
competent state for entitlement to benefits, taking account where appropriate of the provisions of
Article 18, and (c) who is authorised by the competent institution to go to the territory of another
Member State to receive there the treatment appropriate to his condition, shall be entitled: (i) to
benefits in kind provided on behalf of the competent institution by the institution of the place of stay
or residence in accordance with the provisions of the legislation which it administers, as though he
were insured with it; the length of the period during which benefits are provided shall be governed,
however, by the legislation of the competent state; (ii) to cash benefits provided by the competent
institution in accordance with the provisions of the legislation which it administers. However, by
agreement between the competent institution and the institution of the place of stay or residence,
such benefits may be provided by the latter institution on behalf 709 of the former, in accordance
with the provisions of the legislation of the competent state.
(2) The authorization required under paragraph 1(c) may not be refused where the treatment
in question is among the benefits provided for by the legislation of the Member State on whose
territory the person concerned resides and where he cannot be given such treatment within the
time normally necessary for obtaining the treatment in question in the Member State of residence
taking account of his current state of health and the probable course of the disease.
(3) The provisions of paragraphs 1 and 2 shall apply by analogy to members of the family of an
employed or self-employed person.
16.
Since it was uncertain whether those national provisions were compatible with Community law, more
particularly with arts 30 and 36 of the Treaty, the Conseil Arbitral des Assurances Sociales stayed the
proceedings and referred the following question to the Court of Justice for a preliminary ruling:
Is Article 60 of the Luxembourg Code des Assurances Sociales, under which a social security
institution of Member State A refuses to reimburse to an insured person, who is a national of
Member State A, the cost of spectacles with corrective lenses, prescribed by a doctor established
in Member State A but purchased from an optician established in Member State B, on the ground
that all medical treatment abroad must be authorised in advance by the above social security
institution, compatible with Articles 30 and 36 of the EEC Treaty in so far as it penalises in general
the importation by private individuals of medicinal products or, as in this case, spectacles from
other Member States?
17.
Mr Decker and the European Commission submit that national rules under which an insured person is
denied reimbursement of the cost of products normally reimbursed, unless prior authorisation has been
granted by the insured persons social security institution, constitutes an unjustified barrier to the free
movement of goods.
18.
The Luxembourg, Belgian, French and UK governments, on the other hand, submit that rules such as
those at issue in the main proceedings do not fall within the scope of arts 30 and 36 of the Treaty, in that
they concern social security. They submit, in the alternative, that those provisions do not in any event
preclude such rules from being maintained. The German, Spanish and Netherlands governments agree
with the alternative submission.
19.
Having regard to the observations submitted, the questions to be considered concern first the
application of the principle of freedom of movement in the field of social security, then the effect of
Regulation 1408/71, and finally the application of the provisions on the free movement of goods.

Application of the fundamental principle of freedom of movement in the field of social

security
20.
The Luxembourg, Belgian, French and UK governments submit, primarily, that the rules at issue in the
main proceedings, which concern reimbursement of the cost of treatment, do not fall within the scope of
art 30 of the Treaty, in that they concern a particular branch of social security.
21.
It must be observed, first of all, that, according to settled case law, Community law does not detract
from the powers of the member states to 710 organise their social security systems (see the judgments in
Duphar BV v Netherlands Case 238/82 [1984] ECR 523 (para 16) and Sodemare SA v Regione
Lombardia Case C-70/95 [1997] ECR I-3395 (para 27)).
22.
In the absence of harmonisation at Community level, it is therefore for the legislation of each member
state to determine, first, the conditions concerning the right or duty to be insured with a social security
scheme (see Coonan v Insurance Officer Case 110/79 [1980] ECR 1445 (para 12) and Paraschi v
Landes- versicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501 (para 15)) and, second,
the conditions for entitlement to benefits (see Stber v Bundesanstalt fr Arbeit Joined cases C-45/95
[1997] ECR I-511 (para 36)).
23.
As Advocate General Tesauro observes in his opinion, the member states must nevertheless comply
with Community law when exercising those powers (see paras 1725, above).
24.
The court has held that measures adopted by member states in social security matters which may
affect the marketing of medical products and indirectly influence the possibilities of importing those
products are subject to the Treaty rules on the free movement of goods (see the judgment in Duphar
[1984] ECR 523 (para 18)).
25.
Consequently, the fact that the national rules at issue in the main proceedings fall within the sphere of
social security cannot exclude the application of art 30 of the Treaty.

Effect of Regulation 1408/71

26.
The Luxembourg government submits that art 22 of Regulation 1408/71 lays down the principle that
prior authorisation is required for any treatment in another member state. In that governments view, to
challenge the national provisions relating to reimbursement of the cost of benefits obtained abroad
amounts to calling into question the validity of the corresponding provision in the regulation (see paras 17
to 25, above).
27.
It must be stated that the fact that a national measure may be consistent with a provision of secondary
legislation, in this case art 22 of the regulation, does not have the effect of removing that measure from
the scope of the provisions of the Treaty.
28.
Moreover, as Advocate General Tesauro observes in his opinion (see paras 55 and 57, above), art
22(1) is intended to allow an insured person, authorised by the competent institution to go to another
member state to receive there treatment appropriate to his condition, to receive sickness benefits in kind,
on account of the competent institution but in accordance with the provisions of the legislation of the state
in which the services are provided, in particular where the need for the transfer arises because of the
state of health of the person concerned, without that person incurring additional expenditure.
29.
On the other hand, art 22, interpreted in the light of its purpose, is not intended to regulate and hence
does not in any way prevent the reimbursement by member states, at the tariffs in force in the competent
state, of the cost of medical products purchased in another member state, even without prior
authorisation.
30.
Consequently, the court must examine the compatibility of national rules such as those at issue in the
main proceedings with the Treaty provisions on the free movement of goods.
711

Application of the provisions on the free movement of goods

31.
It is necessary to examine whether rules such as those at issue in the main proceedings are capable
of hindering, directly or indirectly, actually or potentially, intra-Community trade (see the judgment in
Procureur du Roi v Dassonville Case 8/74 [1974] ECR 837 (para 5)).
32.
Mr Decker and the Commission submit that a system under which reimbursement of the cost of
medical products, in accordance with the conditions laid down by the state of insurance, is subject to prior
authorisation by the competent institution of that state where the products are supplied in another
member state constitutes a restriction on the free movement of goods within the meaning of art 30 of the
Treaty.
33.
The member states which have submitted observations argue essentially that rules such as those at
issue do not have the purpose or effect of restricting trade flows, but merely lay down the conditions for
the reimbursement of medical expenses. Such rules do not have the effect of prohibiting the import of
spectacles, nor do they have any direct influence on the possibility of purchasing them outside the
national territory.
They do not prohibit Luxembourg opticians from importing spectacles and corrective lenses from other
member states, processing them and selling them.
34.
It must be observed that the rules at issue encourage persons insured under the Luxembourg social
security scheme to purchase their spectacles from, and have them assembled by, opticians established in
Luxembourg rather than in other member states.
35.
While the national rules at issue in the main proceedings do not deprive insured persons of the
possibility of purchasing medical products in another member state, they do nevertheless make
reimbursement of the costs incurred in that member state subject to prior authorisation, and deny such
reimbursement to insured persons who have not obtained that authorisation. Costs incurred in the state of
insurance are not, however, subject to that authorisation.
36.
Such rules must be categorised as a barrier to the free movement of goods, since they encourage
insured persons to purchase those products in Luxembourg rather than in other member states, and are
thus liable to curb the import of spectacles assembled in those states (see the judgment in EC
Commission v France Case 18/84 [1985] ECR 1339 (para 16)).
37.
The Luxembourg government submits, however, that the free movement of goods is not absolute and
that the rules at issue, the purpose of which is the control of the health expenditure which must
necessarily be taken into consideration, are justified on that basis.
38.
Mr Decker, on the other hand, claims that if his purchase were reimbursed, the financial burden on the
funds budget would be the same, as it reimburses only a flat-rate sum for both frames and corrective
lenses sold by an optician. Since that flat rate is fixed independently of the costs actually incurred, there is
no objective reason why the fund should refuse reimbursement if the purchase is made from an optician
established in another member state. The rules at issue therefore cannot be justified by the need to
control health expenditure.
39.
It must be recalled that aims of a purely economic nature cannot justify a barrier to the fundamental
principle of the free movement of goods. However, it cannot be excluded that the risk of seriously
undermining the financial balance 712 of the social security system may constitute an overriding reason in
the general interest capable of justifying a barrier of that kind.
40.
But, as the Luxembourg government acknowledged in reply to a question from the court, it is clear that
reimbursement at a flat rate of the cost of spectacles and corrective lenses purchased in other member
states has no effect on the financing or balance of the social security system.
41.
The Belgian, German and Netherlands governments have also submitted that the right of insured
persons to have access to quality treatment constitutes a justification for the rules at issue, on the ground
of the protection of public health, as provided for by art 36 of the Treaty. The Belgian government adds
that spectacles must be supplied by persons authorised by law to pursue the profession. If they are
supplied in another member state, supervision to ensure that this has been carried out properly is
seriously called into question, or even impossible.
42.
It must be observed that the conditions for taking up and pursuing regulated professions have been
the subject of Council Directive (EEC) 92/51 on a second general system for the recognition of
professional education and training to supplement Council Directive (EEC) 89/48 (OJ 1992 L209 p 25)
and Commission Directive (EC) 95/43 (OJ 1995 L184 p 21), which amended Annexes C and D to
Directive 92/51.
43.
This means that the purchase of a pair of spectacles from an optician established in another member
state provides guarantees equivalent to those afforded on the sale of a pair of spectacles by an optician
established in the national territory (see, with reference to the purchase of medicinal products in another
member state, the judgments in Schumacher v Hauptzollamt Frankfurt am Main-Ost Case 215/87 [1989]
ECR 617 (para 20) and EC Commission v Germany Case C-62/90 [1992] ECR I-2575 (para 18)).
44.
Furthermore, in the present case the spectacles were purchased on a prescription from an
ophthalmologist, which guarantees the protection of public health.
45.
It follows that rules such as those applicable in the main proceedings are not justified on grounds of
public health in order to ensure the quality of medical products supplied in other member states.
46.
In those circumstances, the answer must be that arts 30 and 36 of the Treaty preclude national rules
under which a social security institution of a member state refuses to reimburse to an insured person on a
flat-rate basis the cost of a pair of spectacles with corrective lenses purchased from an optician
established in another member state, on the ground that prior authorisation is required for the purchase of
any medical product abroad.

Costs
47.
The costs incurred by the Luxembourg, Belgian, German, Spanish, French, Netherlands and UK
governments and by the European Commission, which have submitted observations to the Court of
Justice, are not recoverable. Since these proceedings are, for the parties to the main proceedings, a step
in the action pending before the national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice in answer to the question referred to it by the Conseil Arbitral
des Assurances Sociales by decision of 5 April 1995, hereby rules: arts 30 and 36 of the EC Treaty
preclude national rules under which 713 a social security institution of a member state refuses to
reimburse to an insured person on a flat-rate basis the cost of a pair of spectacles with corrective lenses
purchased from an optician established in another member state, on the ground that prior authorisation is
required for the purchase of any medical product abroad.

28 April 1998.

The COURT OF JUSTICE

delivered the following judgment in Kohll v Union des Caisses de Maladie Case C-158/96.
1.
By judgment of 25 April 1996, received at the Court of Justice of the European Communities on 9 May
1996, the Luxembourg Cour de Cassation (the Court of Cassation) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty two questions on the interpretation of arts 59 and 60 of
that Treaty.
2.
Those questions arose in proceedings between Mr Kohll, a Luxembourg national, and the Union des
Caisses de Maladie (UCM), with which he is insured, concerning a request by a doctor established in
Luxembourg for authorisation for his daughter, who is a minor, to receive treatment from an orthodontist
established in Trier (Germany).
3.
By decision of 7 February 1994 following a negative opinion of the social security medical supervisors,
the request was rejected on the grounds that the proposed treatment was not urgent and that it could be
provided in Luxembourg. That decision was confirmed on 27 April 1994 by a decision of the UCM board.
4.
Mr Kohll appealed against that decision to the Conseil Arbitral des Assurances Sociales (the Social
Insurance Arbitration Council), arguing that the provisions relied on were contrary to art 59 of the Treaty.
The appeal was dismissed by decision of 6 October 1994.
5.
Mr Kohll appealed against the latter decision to the Conseil Suprieur des Assurances Sociales (the
Higher Social Insurance Council), which by judgment of 17 July 1995 upheld the contested decision on
the ground that art 20 of the Luxembourg Codes des Assurances Sociales (the Social Insurance Code)
and arts 25 and 27 of the UCM statutes were consistent with Council Regulation (EEC) 1408/71 on the
application of social security schemes to employed persons, to self-employed persons and to members of
their families moving within the Community (see the version amended and updated by Council Regulation
(EC) 118/97; OJ 1997 L28 p 1).
6.
It appears from art 20(1) of the Code des Assurances Sociales, as amended by the Law of 27 July
1992, which entered into force on 1 January 1994, that with the exception of emergency treatment
received in the event of illness or accident abroad, insured persons may be treated abroad or approach a
treatment centre or centre providing ancillary facilities abroad only after obtaining the prior authorisation of
the competent social security institution.
7.
The terms and conditions for granting authorisation are laid down by arts 25 to 27 of the UCM statutes,
in the version which entered into force on 1 January 1995. Article 25 prescribes in particular that
authorisation may not be given for services which are not reimbursable under the national rules. Article 26
states that the cost of duly authorised treatment is to be reimbursed in accordance with the tariffs
applicable to persons insured under the social security system of the state in which the treatment is
provided. Under art 27, finally, authorisation will be granted only after a medical assessment and on
production of a written 714 request from a doctor established in Luxembourg indicating the doctor or
hospital centre recommended and the facts and criteria which make it impossible for the treatment in
question to be carried out in Luxembourg.
8.
Article 22 of Regulation 1408/71 provides in particular:
(1) An employed or self-employed person who satisfies the conditions of the legislation of the
competent state for entitlement to benefits, taking account where appropriate of the provisions of
Article 18, and (c) who is authorised by the competent institution to go to the territory of another
Member State to receive there the treatment appropriate to his condition, shall be entitled: (i) to
benefits in kind provided on behalf of the competent institution by the institution of the place of stay
or residence in accordance with the provisions of the legislation which it administers, as though he
were insured with it; the length of the period during which benefits are provided shall be governed,
however, by the legislation of the competent state; (ii) to cash benefits provided by the competent
institution in accordance with the provisions of the legislation which it administers. However, by
agreement between the competent institution and the institution of the place of stay or residence,
such benefits may be provided by the latter institution on behalf of the former, in accordance with
the provisions of the legislation of the competent state.
(2) The authorisation required under paragraph 1(c) may not be refused where the treatment in
question is among the benefits provided for by the legislation of the Member State on whose
territory the person concerned resides and where he cannot be given such treatment within the
 time normally necessary for obtaining the treatment in question in the Member State of residence
taking account of his current state of health and the probable course of the disease.
(3) The provisions of paragraphs 1 and 2 shall apply by analogy to members of the family of an
employed or self-employed person.
9.
Mr Kohll appealed against the judgment of the Conseil Suprieur des Assurances Sociales, arguing in
particular that it had considered only whether the national rules were consistent with Regulation 1408/71,
and not whether they were consistent with arts 59 and 60 of the Treaty.
10.
Since it considered that that argument raised a question concerning the interpretation of Community
law, the Cour de Cassation stayed the proceedings and referred the following two questions to the Court
of Justice for a preliminary ruling:
(1) Are Articles 59 and 60 of the Treaty establishing the EEC to be interpreted as precluding
rules under which reimbursement of the cost of benefits is subject to authorisation by the insured
persons social security institution if the benefits are provided in a Member State other than the
state in which that person resides?
(2) Is the answer to Question 1 any different if the aim of the rules is to maintain a balanced
medical and hospital service accessible to everyone in a given region?
11.
By those questions, which should be taken together, the national court essentially asks whether arts
59 and 60 of the Treaty preclude the application of social security rules such as those at issue in the main
proceedings.
715
12.
Mr Kohll submits that arts 59 and 60 of the Treaty preclude such national rules which make
reimbursement, in accordance with the scale of the member state of insurance, of the cost of dental
treatment provided by an orthodontist established in another member state subject to authorisation by the
insured persons social security institution.
13.
UCM and the Luxembourg, Greek and UK governments contend that those provisions are not
applicable, or, in the alternative, do not preclude the rules in question from being maintained. The
German, French and Austrian governments agree with the alternative submission.
14.
The Commission submits that the rules constitute a barrier to the freedom to provide services but may
be justified, under certain conditions, by overriding reasons relating to the general interest.
15.
Having regard to the observations submitted, the questions to be considered concern first the
application of the principle of freedom of movement in the field of social security, then the effect of
Regulation 1408/71, and finally the application of the provisions on freedom to provide services.
Application of the fundamental principle of freedom of movement in the field of social security
16.
The Luxembourg, Greek and UK governments submit that the rules at issue in the main proceedings
do not fall within the scope of the Community provisions on freedom to provide services, in that they
concern social security, and so should be examined solely from the point of view of art 22 of Regulation
1408/71.
17.
It must be observed, first of all, that, according to settled case law, Community law does not detract
from the powers of the member states to organise their social security systems (see the judgments in
Duphar BV v Netherlands Case 238/82 [1984] ECR 523 (para 16) and Sodemare SA v Regione
Lombardia Case C-70/95 [1997] ECR I-3395 (para 27)).
18.
In the absence of harmonisation at Community level, it is therefore for the legislation of each member
state to determine, first, the conditions concerning the right or duty to be insured with a social security
scheme (see the judgments in Coonan v Insurance Officer Case 110/79 [1980] ECR 1445 (para 12) and
Paraschi v Landesversicherungsanstalt Wrttemberg Case C-349/87 [1991] ECR I-4501 (para 15)) and,
second, the conditions for entitlement to benefits (see the judgment in Stber v Bundesanstalt fr Arbeit
Joined cases C-45/95 [1997] ECR I-511 (para 36)).
19.
As Advocate General Tesauro observes in his opinion, the member states must nevertheless comply
with Community law when exercising those powers (see paras 1725, above).
20.
The court has held that the special nature of certain services does not remove them from the ambit of
the fundamental principle of freedom of movement (see the judgment in Criminal proceedings against
Webb Case 279/80 [1981] ECR 3305 (para 10)).
21.
 Consequently, the fact that the national rules at issue in the main proceedings fall within the sphere of
social security cannot exclude the application of arts 59 and 60 of the Treaty.

Effect of Regulation 1408/71.

22.
UCM and the Luxembourg government submit that art 22 of Regulation 1408/71 lays down the
principle that prior authorisation is required for any treatment in another member state. To challenge the
national provisions relating to reimbursement of the cost of services obtained abroad amounts to 716
calling into question the validity of the corresponding provision in the regulation.
23.
In the proceedings before the court, Mr Kohll submitted that he sought reimbursement by UCM of the
amount he would have been entitled to if the treatment had been carried out by the only specialist
established in Luxembourg at the material time.
24.
On that point, UCM considers that the principle that a person is subject to one social security tariff only
would indeed be complied with if the Luxembourg tariff were applied, but claims that Regulation 1408/71
would compel it to reimburse expenditure according to the tariffs in force in the state in which the service
was provided.
25.
It must be stated that the fact that a national measure may be consistent with a provision of secondary
legislation, in this case art 22 of the regulation, does not have the effect of removing that measure from
the scope of the provisions of the Treaty.
26.
Moreover, as the Advocate General observes in his opinion (see paras 55 and 57, above), art 22(1) of
the regulation is intended to allow an insured person, authorised by the competent institution to go to
another member state to receive there treatment appropriate to his condition, to receive sickness benefits
in kind, on account of the competent institution but in accordance with the provisions of the legislation of
the state in which the services are provided, in particular where the need for the transfer arises because
of the state of health of the person concerned, without that person incurring additional expenditure.
27.
On the other hand, art 22, interpreted in the light of its purpose, is not intended to regulate and hence
does not in any way prevent the reimbursement by member states, at the tariffs in force in the competent
state, of costs incurred in connection with treatment provided in another member state, even without prior
authorisation.
28.
Consequently, the court must examine the compatibility of national rules such as those at issue in the
main proceedings with the Treaty provisions on freedom to provide services.

Application of the provisions on freedom to provide services

29.
The dispute before the national court concerns treatment provided by an orthodontist established in
another member state, outside any hospital infrastructure. That service, provided for remuneration, must
be regarded as a service within the meaning of art 60 of the Treaty, which expressly refers to activities of
the professions.
30.
It must therefore be examined whether rules such as those at issue in the main proceedings constitute
a restriction on freedom to provide services, and if so, whether they may be objectively justified.
Restrictive effects of the rules at issue
31.
Mr Kohll and the European Commission submit that the fact that reimbursement of the cost of medical
services, in accordance with the legislation of the state of insurance, is subject to prior authorisation by
the institution of that state where the services are provided in another member state constitutes a
restriction on freedom to provide services within the meaning of arts 59 and 60 of the Treaty.
32.
The member states which have submitted observations consider, on the contrary, that the rules at
issue do not have as their purpose or effect to restrict 717 freedom to provide services, but merely lay
down the conditions for the reimbursement of medical expenses.
33.
It should be noted that, according to the courts case law, art 59 of the Treaty precludes the application
of any national rules which have the effect of making the provision of services between member states
more difficult than the provision of services purely within one member state (see the judgment in
European Commission v France Case C-381/93 [1994] ECR I-5145 (para 17)).
34.
While the national rules at issue in the main proceedings do not deprive insured persons of the
possibility of approaching a provider of services established in another member state, they do
nevertheless make reimbursement of the costs incurred in that member state subject to prior
authorisation, and deny such reimbursement to insured persons who have not obtained that authorisation.
Costs incurred in the state of insurance are not, however, subject to that authorisation.
35.
Consequently, such rules deter insured persons from approaching providers of medical services
established in another member state and constitute, for them and their patients, a barrier to freedom to
provide services (see the judgments in Luisi v Ministero del Tesoro Joined cases 286/82 and 26/83 [1984]
ECR 377 (para 16) and Bachmann v Belgium Case C-204/90 [1992] ECR I-249 (para 31)).
36.
The court must therefore examine whether a measure of the kind at issue in this case may be
objectively justified.
Justification of the rules at issue
37.
UCM and the governments of the member states which have submitted observations submit that
freedom to provide services is not absolute and that reasons connected with the control of health
expenditure must be taken into consideration. The requirement of prior authorisation constitutes the only
effective and least restrictive means of controlling expenditure on health and balancing the budget of the
social security system.
38.
According to UCM, the Luxembourg government and the Commission, the risk of upsetting the
financial balance of the social security scheme, which aims to ensure a balanced medical and hospital
service available to all its insured, constitutes an overriding reason in the general interest capable of
justifying restrictions on freedom to provide services.
39.
The Commission adds that the refusal of the national authorities to grant prior authorisation must be
justified by a genuine and actual risk of upsetting the financial balance of the social security scheme.
40.
On the latter point, Mr Kohll submits that the financial burden on the budget of the Luxembourg social
security institution is the same whether he approaches a Luxembourg orthodontist or one established in
another member state, since he asked for medical expenses to be reimbursed at the rate applied in
Luxembourg. The rules at issue therefore cannot be justified by the need to control health expenditure.
41.
It must be recalled that aims of a purely economic nature cannot justify a barrier to the fundamental
principle of freedom to provide services (see, to that effect, the judgment in Syndesmos ton en Elladi
Touristikon kai Taxidiotikon Grafeion v Ergasias Case C-398/95 [1997] ECR I-3091 (para 23)). However, it
cannot be excluded that the risk of seriously undermining the financial balance of the social security
system may constitute an overriding reason in the general interest capable of justifying a barrier of that
kind.
718
42.
But, contrary to the submissions of UCM and the Luxembourg government, it is clear that
reimbursement of the costs of dental treatment provided in other member states in accordance with the
tariff of the state of insurance has no significant effect on the financing of the social security system.
43.
The Luxembourg government also relies on grounds based on the protection of public health, arguing,
first, that the rules at issue are necessary to guarantee the quality of medical services, which in the case
of persons going to another member state can be ascertained only at the time of the request for
authorisation, and, second, that the Luxembourg sickness insurance system aims to provide a balanced
medical and hospital service open to all insured persons.
44.
Mr Kohll submits, on the other hand, that there is no scientific reason to conclude that treatment
provided in Luxembourg is more effective, now that the pursuit of the medical professions is the subject of
mutual recognition between member states. He further submits that the reference to a balanced medical
and hospital sector open to all must above all be categorised as an economic aim intended to protect
UCMs financial resources.
45.
It should be noted, first of all, that under arts 56 and 66 of the EC Treaty member states may limit
freedom to provide services on grounds of public health.
46.
However, that does not permit them to exclude the public health sector, as a sector of economic
activity and from the point of view of freedom to provide services, from the application of the fundamental
principle of freedom of movement (see Gl v Regierungsprsident Dsseldorf Case 131/85 [1986] ECR
1573 (para 17)).
47.
The conditions for taking up and pursuing the profession of doctor and dentist have been the subject
of several coordinating or harmonising directives (see Council Directive (EEC) 78/686 concerning the
mutual recognition of diplomas, certificates and other evidence of the formal qualifications of practitioners
of dentistry, including measures to facilitate the effective exercise of the right of establishment and
freedom to provide services (OJ 1978 L233 p 1), Council Directive (EEC) 78/687 concerning the
coordination of provisionslaid down by law, regulation or administrative action in respect of the activities of
dental practitioners (OJ 1978 L233 p 10) and Council Directive (EEC) 93/16 to facilitate the free
movement of doctors and the mutual recognition of their diplomas, certificates and other evidence of
formal qualifications (OJ 1993 L165 p 1)).
48.
It follows that doctors and dentists established in other member states must be afforded all guarantees
equivalent to those accorded to doctors and dentists established on national territory, for the purposes of
freedom to provide services.
49.
Consequently, rules such as those applicable in the main proceedings cannot be justified on grounds
of public health in order to protect the quality of medical services provided in other member states.
50.
As to the objective of maintaining a balanced medical and hospital service open to all, that objective,
although intrinsically linked to the method of financing the social security system, may also fall within the
derogations on grounds of public health under art 56 of the Treaty, in so far as it contributes to the
attainment of a high level of health protection.
51.
Article 56 of the Treaty permits member states to restrict the freedom to provide medical and hospital
services in so far as the maintenance of a treatment facility or medical service on national territory is
essential for the public health 719 and even the survival of the population (see, with respect to public
security within the meaning of art 36 of the Treaty, the judgment in Campus Oil Ltd v Minister for Industry
and Energy Case 72/83 [1984] ECR 2727 (paras 3336)).
52.
However, neither UCM nor the governments of the member states which have submitted observations
have shown that the rules at issue were necessary to provide a balanced medical and hospital service
accessible to all. None of those who have submitted observations has argued that the rules were
indispensable for the maintenance of an essential treatment facility or medical service on national
territory.
53.
The conclusion must therefore be drawn that the rules at issue in the main proceedings are not
justified on grounds of public health.
54.
In those circumstances, the answer must be that arts 59 and 60 of the Treaty preclude national rules
under which reimbursement, in accordance with the scale of the state of insurance, of the cost of dental
treatment provided by an orthodontist established in another member state is subject to authorisation by
the insured persons social security institution.

Costs
55.
The costs incurred by the Luxembourg, German, Greek, French, Austrian and UK governments and by
the European Commission, which have submitted observations to the Court of Justice, are not
recoverable. Since these proceedings are, for the parties to the main proceedings, a step in the action
pending before the national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Luxembourg
Cour de Cassation by judgment of 25 April 1996, hereby rules: arts 59 and 60 of the EC Treaty preclude
national rules under which reimbursement, in accordance with the scale of the state of insurance, of the
cost of dental treatment provided by an orthodontist established in another member state is subject to
authorisation by the insured persons social security institution.

720

[1998] All ER (EC) 722

Hill and another v Revenue Comrs and another

(Case C-243/95)

EUROPEAN COMMUNITY; Social policy: EMPLOYMENT; Discrimination

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE CHAMBER), SCHINTGEN, MANCINI, MURRAY
(RAPPORTEUR) AND HIRSCH ADVOCATE GENERAL LA PERGOLA
10 DECEMBER 1996, 20 FEBRUARY, 17 JUNE 1998
European Community Equality of treatment of men and women Pay Civil service job-sharing
scheme Women applicants job-sharing for two years Applicants later taking up full-time employment
Employer moving applicants down scale of annual pay increases on grounds that two years job share
equivalent to one years full-time employment Whether application of criterion of service calculated by
length of time actually worked constituting discrimination Council Directive (EEC) 75/117 EC Treaty,
art 119.
The applicants were clerical assistants in the Irish civil service, both of whom had exercised their right
under Irish rules governing public service to job share. Under the relevant scheme two civil servants
shared one full-time job and its benefits equally and each was paid 50% of the applicable full-time salary;
the rate of annual salary increases for job sharers was parallel to that for full-time workers; and
participants in the scheme were entitled to take up or return to full-time employment, provided that
suitable vacancies existed. After job-sharing for two years both of the applicants, each of which had
moved one point up the pay scale with each year of service, returned to/took up full-time employment.
They were initially assimilated to the same point on the full-time pay scale as that which they had
occupied on the job-sharers scale, but were subsequently reclassified at lower points, on the grounds that
two years job-sharing was equivalent to only one years full-time employment. The applicants challenged
that decision before the Equality Officer, pursuant to s 7(1) of the Anti-Discrimination (Pay) Act 1974, who
found in their favour. On appeal, the Labour Court referred to the Court of Justice of the European
Communities for a preliminary ruling the question whether such pay-related reclassification of workers
moving from job-sharing to full-time work breached the principle of equal pay as defined in Council
Directive (EEC) 75/117 on the approximation of the laws of the member states relating to the application
of the principle of equal pay for men and women.

Held Discrimination could arise only through the application of different rules to comparable situations
or the application of the same rules to different situations. Moreover, Directive 75/117 precluded the
application of national measures which, although formulated in neutral terms, worked to the disadvantage
of far more women than men, unless that measure was based on objective factors unrelated to any
discrimination on grounds of sex. In the instant case, workers entering full-time work were paid less than
double what they would have earned had they been job-sharing and, as a result, their hourly rate of pay
was reduced without taking account of the fact that a job-sharer could acquire the same experience and
performed the same functions as a full-time worker in respect of the quality and the quantity of work. It
followed that 722 full-time employees who had previously job-shared were treated differently from those
who had always worked on a full-time basis; and since 992% of clerical assistants and 98% of all civil
servants who job shared were women, there was indirect discrimination. Accordingly, where far more
women were affected than men, art 1191 of the EC Treaty and Directive 75/117 precluded national
measures which provided that workers who converted from job-sharing to full-time work regressed on
their salary scale due to the application of the criterion of service calculated by the length of time actually
worked in a post, unless such measures could be objectively justified (see p 740 c to g, p 741 e to g and p
743 b c to g, post).
1
Article 119, so far as material, is set out at p 737 f, post

Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319 and Megner v
Innungskrankenkasse Vorderpfalz Case C-444/93 [1996] All ER (EC) 212 applied.

Notes
For the EC Treaty, art 119, see 50 Halsburys Laws (4th edn) 306.

Cases cited
Almelo (Municipality of) v NV Energiebedrijf Ijsselmij Case C-393/92 [1994] ECR I-1477.
Bilka-Kaufhaus GmbH v Weber von Hartz Case 170/84 [1986] ECR 1607.
Corbiau v Administration des contributions Case C-24/92 [1993] ECR I-1277.
De Weerd, ne Roks v Bestuur van de Bedrijfsvereniging voor de Gezondheid, Geestelijke en
Maatschappelijke Belangen Case C-343/92 [1994] ECR I-571.
Defrenne v Sabena Case 43/75 [1981] 1 All ER 122, [1976] ECR 455, ECJ.
Enderby v Frenchay Health Authority Case C-127/92 [1993] ECR I-5535.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Gerster v Freistaat Bayern Case C-1/95 [1997] ECR I-5253.
Handels- og Kontorfunktionrernes Forbund i Danmark v Dansk Arbejdsgiverforening acting on behalf of
Danfoss Case 109/88 [1989] ECR 3199.
Jenkins v Kingsgate (Clothing Productions) Ltd Case 96/80 [1981] 1 WLR 972, [1981] ECR 911, ECJ.
Kording v Senator fr Finanzen Case C-100/95 [1997] ECR I-5189.
Kowalska v Freie und Hansestadt Hamburg Case C-33/89 [1990] ECR I-2591.
Kuratorium fr Dialyse und Nierentransplantation eV v Lewark Case C-457/93 [1996] ECR I-243.
Megner v Innungskrankenkasse Vorderpfalz Case C-444/93 [1996] All ER (EC) 212, [1995] ECR I-4741,
ECJ.
Nimz v Freie und Hansestadt Hamburg Case C-184/89 [1991] ECR I-297.
Pardini (Fratelli) SpA v Ministero del Commercio Case 338/85 [1988] ECR 2041.
Rinner-Khn v FWW Spezial Gebudereinigung GmbH Case 171/88 [1989] ECR 2743.
Secretary of State for Social Security v Thomas Case C-328/91 [1993] 4 All ER 556, sub nom Thomas v
Chief Adjudication Officer [1993] QB 747, [1993] 3 WLR 581, [1993] ECR I-1247, ECJ.
Stadt Lengerich v Helmig Joined cases C-399/92, C-409/92, C-425/92, C-34/93, C-50/93 and C-78/93
[1994] ECR I-5727.
Vaassen v Management of the Beambtenfonds voor het Mijnbedrijf Case 61/65 [1966] ECR 261.
723
X (Criminal proceedings against) Joined cases C-74/95 and C-129/95 [1996] ECR I-6609.

Reference
By order of 5 April 1995, the Labour Court referred to the Court of Justice of the European Communities
for a preliminary ruling under art 177 of the EC Treaty three questions (set out at p 739 c to f, post) on the
interpretation of Council Directive (EEC) 75/117 on the approximation of the laws of the member states
relating to the application of the principle of equal pay for men and women. Those questions arose in a
dispute between Ms Hill and Ms Stapleton, who had previously worked in a job-sharing capacity, and the
Revenue Commissioners and Department of Finance concerning the latters decision to place them, on
their conversion to full-time employment, on a point of the full-time pay scale lower than that of the job-
sharing pay scale which they had previously occupied. Written observations were submitted on behalf of:
Kathleen Hill and Ann Stapleton, by M Redmond, Solicitor, and J OReilly, SC; the Revenue
Commissioners and the Department of Finance, by M A Buckley, Chief state Solicitor, M Finlay, SC, and F
Flanagan, of the Office of the Attorney General, acting as agents; and the European Commission, by M
Wolfcarius and C Docksey, of its Legal Service, acting as agents. Oral observations were made by:
Kathleen Hill and Ann Stapleton, represented by M Redmond and J OReilly; the Revenue
Commissioners and the Department of Finance, represented by M Finlay; the United Kingdom
government, represented by L Nicoll, the Treasury Solicitors Department, acting as agent, and C Lewis,
Barrister; and the Commission, represented by M Wolfcarius and C Docksey. The language of the case
was English. The facts are set out in the opinion of the Advocate General.

20 February 1997.

The Advocate General (A La Pergola)

delivered the following opinion (translated from the Italian).
1.
In this case, the Court of Justice of the European Communities is once more called upon to give a
ruling on the compatibility with art 119 of the Treaty of a system of remuneration applicable to civil
servants. The point in issue here is whether the criteria set by the Irish rules for calculating job-sharing
service for the purposes of pay when the employee concerned exercises the right to return to full-time
employment, are consistent with the principle of equal treatment for men and women laid down in the
Community legal order.

FACTS
2.
The facts underlying the dispute may be briefly recapitulated as follows: Kathleen Hill and Ann
Stapleton (the claimants)recruited to the Irish civil service from open competitions for the grade of
Clerical Assistantwere assigned to the office of the Revenue Commissioners.
3.
Both claimants exercised their right under the Irish rules governing public service to work in a job-
sharing capacity2. More specifically, Ms Hill, who was recruited in July 1981, began job-sharing in May
1988 when she was on the
2
As the national court makes clear (order for reference, p 3), the job-sharing scheme was introduced in 1984 as a job-
creation measure. Two officers share one full-time job: since the salary is divided, the cost to management remains the
same. Under the rules, staff recruited on a full-time basis may participate in the scheme, in which case they retain the
right to return to full-time work (Ms Hills position); staff recruited on a job-sharing basis are entitled to be appointed to
full-time positions provided that suitable vacancies exist (Ms Stapletons position).

724
seventh point of the scale of pay for Clerical Assistant posts; Ms Stapleton, for her part, was directly
recruited to a job-sharing post in April 1986.
The claimants each job-shared for two years (Ms Hill from May 1988 to June 1990; Ms Stapleton from
April 1986 to April 1988). During that period of service they worked half full-time hours on a one week
on/one week off basis3.
3
This is one of four job-sharing options operated under the Irish system. The others are: day on/day off;
morning/afternoon, and alternate three day/two day week.

4.
As regards pay, under instructions contained in a circular of 27 February 1984 issued by the
Department of Finance, the scale of pay applicable to job-sharing staff is a scale each point of which
represents 50% of the corresponding point on the scale of pay appropriate to full-time staff 4. Employment
while job-sharing is therefore paid on a strictly pro-rata basis.
4
It seems worth adding a few brief comments on the pay classification system used in the Irish administration. There are
two different scales, subdivided into eleven salary grades: one for full-time employees and the other for job-sharers. The
 salaries in the second scale represent, at each grade, half of the corresponding point on the scale of pay appropriate to
full-time staff. Under the rules governing employment within the administration, increments on that scale are granted
annually if the officers services are considered satisfactory by the head of the department.

5.
The rules laid down in the administrative circular in question were incomplete. There was no provision
concerning the criteria for calculating periods of service in a job-sharing capacity for the purposes of
progression on the pay scale when the employee exercised the right to return to full-time work (or, in the
case of employment in a job-sharing capacity ab initio, the right to take up full-time employment).
6.
The competent administrative authoritythe Department of Finance addressed that question in a
circular of 31 March 1987 which stated that as each years job-sharing service is reckonable as 6
monthsfull-time service, an officer who has served for two years in a job-sharing capacity should be
placed on the second point of the full-time scale (equivalent to one years full-time service) (emphasis
added). Where officers had been job-sharing for more than two years, their position on the full-time scale
was to be adjusted on a strictly pro-rata basis (see Circular of 31 March 1987 from the Civil Service
Training Centre of the Department of Finance).
7.
Those criteria for calculating years of service were not, however, immediately applied to the claimants.
In calculating their point on the pay scale, the administration initially reckoned the period spent in job-
sharing as if it had been full-time employment. On that basis, Ms Hill, who returned to full-time work in
June 1990, was placed on the ninth point on the corresponding scale (that is to say, at the same point as
she had reached on the job-sharing scale); Ms Stapleton, on the other hand, was placed when she
secured full-time workhaving completed two years job-sharing serviceon the third point of the pay
scale and, since she worked satisfactorily throughout 1989 and 1990, moved up to the fifth grade of the
full-time scale.
8.
It was only subsequently that the administration realised its mistake and undertook the necessary
adjustments. Ms Hill was regressed from the ninth to the eighth point on the pay scale. For her part, Ms
Stapleton was not permitted to progress to the sixth point on the scale, even though from the point of view
of the quality of her work, she satisfied the requirements for such advancement.
9.
Under s 7(1) of the Anti-Discrimination (Pay) Act 1974, the claimants submitted a complaint to an
Equality Officer in which they argued that the rules governing the recognition of pay increments accrued
while job-sharing were 725 unlawful as contrary to art 119 of the Treaty and Council Directive (EEC)
75/117 on the approximation of the laws of the member states relating to the application of the principle of
equal pay for men and women (OJ 1975 L45 p 19).
10.
The Equality Officer, relying on the 1974 Act and the courts judgment in Nimz v Freie und Hansestadt
Hamburg Case C-184/89 [1991] ECR I-297, found in the claimants favour. In his opinion, equating two
years job-sharing service to one years full-time service for the purposes of progression on the
incremental scale constituted discrimination against them.
11.
Following the recommendation of the Equality Officer, the Revenue Commissioners and the
Department of Finance (the appellants) appealed to the Labour Court. The claimants, for their part, also
appealed to that court for an order requiring the administration to implement the recommendation.
12.
The Labour Court found that a case of direct discrimination was not made out. However, on the
grounds that the majority of job-sharing staff are women and that the period of service spent job-sharing
may give the officer experience equivalent to that gained by a full-time worker, it has referred the following
questions to the Court of Justice in order to ascertain whether indirect discrimination may be made out
against female civil employees in this case.
In circumstances in which far more female workers than male workers spend part of their
working lives in a job-sharing capacity: (a) Does a prima facie case of indirect discrimination arise
where job-sharing workers who convert to full-time work are given credit for incremental
progression on the scale of pay for full-time staff by reference to actual time worked such that,
while the benefits awarded to them are fully prorated to those awarded to staff who have always
worked full-time, they are placed at lower points on the full-time scale than comparators who are in
all respects similar to them except that they have worked continuously on a full-time basis? In other
words, is the principle of equal pay, as defined in Directive 75/117/EEC, contravened if employees
who convert from job-sharing to full-time work, regress on the incremental scale and hence on their
salary scale, due to the application by the employer of the criterion of service calculated by time
worked in a job? (b) If so, does the employer have to provide special justification for recourse to the
criterion of service, defined as actual time worked, in awarding incremental credit? (c) If so, can a
practice of incremental progression by reference to actual time worked be objectively justified by
reference to factors other than the acquisition of a particular level of skill and experience over
time?
Apart from the parties, the European Commission has also submitted written observations. The
parties, the UK government and the Commission participated in the hearing.
 The relevant provisions
13.
The provisions quoted below are relevant to consideration of the case.
The second paragraph of art 119 of the Treaty defines pay as follows:
For the purpose of this Article, pay means the ordinary basic or minimum wage or salary and
any other consideration, whether in cash or in kind, which the worker receives, directly or indirectly,
in respect of his employment from his employer.
Article 1 of Directive 75/117 provides:
726
The principle of equal pay for men and women outlined in Article 119 of the Treaty, hereinafter
called principle of equal pay, means, for the same work or for work to which equal value is
attributed, the elimination of all discrimination on grounds of sex with regard to all aspects and
conditions of remuneration.

Legal analysis
14.
Before examining the substance of the questions referred by the Labour Court, I shall start by tackling
a matter which the Commission alone raised in its observations. May the Labour Court be regarded as a
court or tribunal within the meaning of art 177 of the Treaty? The Commission considers that it may. The
other parties did not state any views in the written procedure. At the hearing, however, the representative
of the appellant administrative authorities considered it necessary to state that the Labour Court may be
regarded as a court or tribunal within the meaning of art 177 solely in respect of certain areas of
jurisdiction, including that involved in this dispute.
15.
It seems to me that a few observations will suffice on this point. The Labour Court was first constituted
by the Industrial Relations Act 19465, which lays down detailed rules concerning its powers and
procedure. Its members are appointed by the Minister for Trade and Industry on recommendations from
trade union and employers organisations6. Whilst its original jurisdiction was exclusively in the sphere of
arbitration, the functions of the Labour Court have changed over time. The alteration introduced by the
Anti-Discrimination (Pay) Act 19747 particularly significant for the purposes of this case. Section 8 of the
Act provides that the Labour Court is to hear and determine appealsas in this caseconcerning
Recommendations of the Equality Officer. In such cases there is provision on the one hand for a private
or public hearing (s 8(1)(c)(i) and (ii)), and on the other hand for a possible appeal on a point of law to the
High Court against the Labour Courts rulings, which does not exist for other cases within the Labour
Courts jurisdiction (s 8(3))8.
5
See Acts of the Oireachtas (1946) No 26, p 1019, ss 1025.
6
The procedure for appointing the chairman does not involve consultation with employers and workers representatives,
but is a matter for the minister alone. See s 10(3) of the 1946 Act.
7
See Acts of the Oireachtas (1974) No 15, p 211.
8
See s 17 of the Industrial Relations Act 1946. It is to be noted that s 8(4) of the 1974 Act lays down the procedure to be
followed where the employer fails to implement a decision of the Labour Court. In that case, there is a second stage
before that court, at the end of which the employer may be ordered to take certain action. If the employer fails to comply,
fines may be imposed.

16.
To my mind, the requirements which a decision-making body must satisfy, according to the courts
case law, in order to constitute a court or tribunal within the meaning of art 177 are fulfilled in this case: it
must be established by law and be independent, have a permanent existence, exercise binding
jurisdiction, be bound by rules of adversary procedure and apply rules of law 9. The Labour Court, which
was established and is governed by law, is a permanent, independent body and also exercises binding
jurisdiction inasmuch as it has sole jurisdiction to hear and determine equal-treatment disputes (see s
7(1) of the 1974 Act). It takes its decisions in accordance with a procedure laid 727 down by statute, with
the safeguard of an inter partes hearing, and is called upon to apply rules of law. An appeal against its
decisions can be brought before the High Court on a point of law (see Vaassen v Management of the
Beambtenfonds voor het Mijnbedrijf Case 61/65 [1966] ECR 261). Last, I would point out that the court
held in Handels- og Kontorfunktionrernes Forbund i Danmark v Dansk Arbejdsgiverforening acting on
behalf of Danfoss Case 109/88 [1989] ECR 3199 (Danfoss), a case much more open to doubt than this,
that an industrial arbitration board was entitled to refer questions for a preliminary ruling under art 177 10.
9
See, most recently, Criminal proceedings against X Joined cases C-74/95 and C-129/95 [1996] ECR I-6609;
Municipality of Almelo v NV Energiebedrijf Ijsselmij Case C-393/92 [1994] ECR I-1477 (para 21); Corbiau v
Administration des contributions Case C-24/92 [1993] ECR I-1277; Handels- og Kontor- funktionrernes Forbund i
Danmark v Dansk Arbejdsgiverforening acting on behalf of Danfoss Case 109/88 [1989] ECR 3199 and Fratelli Pardini
SpA v Ministero del Commercio Case 338/85 [1988] ECR 2041.
10
I would point out that in that case neither the establishment of the board nor the procedure to be observed was
determined in detail by statute; the arbitration board was in general composed on an ad hoc basis, so that doubts arose
as to its nature as a permanent body; finally, the legal rules which it had to apply were those of collective agreements.
On this point, see the opinion of Advocate General Lenz in Danfoss [1989] ECR 3209 (paras 1624).

17.
I would ask the court to bear with me while I make a few preliminary observations regarding this
dispute. The case before the court resembles those considered in Gerster v Freistaat Bayern Case C-
1/95 [1997] ECR I-5253 and Kording v Senator fr Finanzen Case C-100/95 [1997] ECR I-5189on
which I delivered a joint opinion on 22 October 1996and in the case of Nimz. Yet this is true only in
certain legal and factual respects. In other respects the case now before the court differs.
18.
The points of similarity with the Gerster and Kording cases are as follows. In those proceedings it was
alleged that the calculation method used by the German government in regard to advancementin terms
of career in those cases as against in terms of pay scale in this casegave rise to indirect discrimination
against part-time workers. Having said that, it should be noted that the circumstances under consideration
in Gerster and Kording fell within the ambit of Community rules different from those in point in this dispute.
In the cases in question it was necessary to determine whether the German administrations conduct with
regard to access to a post of higher grade was lawful under Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to employment,
vocational training and promotion, and working conditions (OJ 1976 L39 p 40).
19.
In relation to Nimz [1991] ECR I-297, the similarity lies in the fact that, in that case and in this, what
has to be determined is whether national rules laying down the criteria which part-time workers have to
satisfy in order to acquire the right to pay increases are compatible with the principles of equal treatment.
Unlike the provision at issue in Nimz, however, the Irish provision under consideration in this case
evaluates work carried out over a reduced number of hours on the basis of strictly pro-rata parameters:
there is a direct correlation between the work carried out and progression on the scale 11.
11
I note that the provisions of Bundesangestelltentarifvertrag (the BAT) (the collective wage agreement scale for federal
employees; see Nimz [1991] ECR I-297 (para 3)) which were at issue in Nimz provided that full account should be taken
of the period of service of workers employed for at least three-quarters of normal working time, but that only one-half of
such period of service should be taken into account in the case of workers whose working hours were between one-half
and three-quarters of normal working time.

20.
To my mind, this case falls within the scope of the Community rules on equal pay: art 119 of the Treaty
and the relevant implementing provisions, especially those laid down in Directive 75/117. I have more
than one reason for taking this view.
728
In its decisions the court has repeatedly held that art 119 is mandatory and applies to individual and
collective agreements, as well as to laws, regulations and administrative provisions (see Rinner-Khn v
FWW Spezial Gebudereinigung GmbH Case 171/88 [1989] ECR 2743, Kowalska v Freie und
Hansestadt Hamburg Case C-33/89 [1990] ECR I-2591 and Enderby v Frenchay Health Authority Case
C-127/92 [1993] ECR I-5535 (para 21).).
On top of this, there isas the Commission observesspecific clarification afforded by the judgment
in Nimz [1991] ECR I-297 (para 10), which is directly relevant to the subject matter of this case: the rules
governing what is practically an automatic reclassification in a higher salary grade come in principle within
the concept of pay as defined in Article 119 of the Treaty. In the case now before the court, movement
from one salary grade to another is automatic, which accords with the test laid down in Nimz. The
question arises precisely because employees who initially worked in a job-sharing capacity automatically
suffer a reduction in hourly pay at the time when they convert their employment relationship into a full-
time arrangement.
The first question
21.
As we know, two kinds of appraisal have to be carried out in discrimination cases. First of all, it must
be ascertained whether discrimination is present and then whether it is objectively justified (see Jenkins v
Kingsgate (Clothing Productions) Ltd Case 96/80 [1981] 1 WLR 972, [1981] ECR 911 (paras 1011)).
From the latter angle, the Labour Court asks in its second and third questions whether, in order for the
employer to have recourse to the criterion of service, defined as actual time worked, it must provide
special justification and, if so, whether it may do so by reference to factors other than the proven
acquisition of a particular level of skill.
22.
So, let us see whether discrimination is present in this case. One point is certain already: there is no
direct discrimination. The national court itself has said so. The rules in question apply without distinction
to men and women.
It must, however, be pointed out that the majority of officers in the grade of Clerical Assistant are
women. What is more relevant for present purposes is that the overwhelming majority of job-sharers are
women12. This explains why the question of possible discrimination between employees on the ground of
sex came to be referred to the court.
12
From this point of view, the figures supplied during the proceedings speak for themselves. Of clerical assistants working
in a job-sharing capacity, 992% are women; more generally, 98% of Irish Civil Servants who have chosen to job-share
are women. The claimants assertion, made both in their observations (point 1.2) and at the hearing, that workers
choose that scheme in order to combine work and family responsibilities, seems to be borne out by the experience of
the Civil and Public Service Union that 83% of job-sharers do so to combine family and work responsibilities. This
invariably involves caring for children (quoted in point 2.9 of the claimants observations).

23.
What arguments have been put before the court concerning the hypothesis of indirect discrimination?
 As I have said, under the calculation rules used in the Irish provisions, when workers who have
exercised the right to job-share return to full-time work, their pay is less than that received by workers who
have always worked full-time. In the claimants view, this results in differential treatment which has no
objective justification and is disadvantageous to women, who form the majority of job-sharers. The
practical upshot is indirect discrimination contrary to the principles of Community law. Furthermore, the
particular character of job-sharing 729 distinguishes it from part-time work: job-sharing, they say, is the
sharing of work and the related responsibilities between two employees. That difference in kind between
job-sharing and part-time work provides justification for taking account of job-sharing for the purposes of
calculating length of service in accordance with the same rules as apply to full-time work.
24.
The government concerned, and also that of the United Kingdom, consider that the present case does
not exhibit the essential features of discrimination. They cite the courts judgment in Stadt Lengerich v
Helmig Joined cases C-399/92, C-409/92, C-425/92, C-34/93, C-50/93 and C-78/93 [1994] ECR I-5727 in
order to argue that it is lawful for a job-sharer to be placed at a lower level on the pay scale on returning
to full-time work. They argue that such regression occurs as a result of the use of a strictly pro-rata
criterion in reckoning service in a job-sharing capacity, and that that criterion is fully justified for the
purposes also of progression on the pay scale: the claimants, they say, are placed, as regards pay, on a
substantially equal footing with workers who have worked an equivalent number of hours. The Irish
authorities maintain that that solution is not discriminatory. Conversely, the solution proposed by the
claimants would be. Once equal treatment in terms of pay is claimed for classes of employees who have
carried out quantitatively different work, those who have worked reduced hours will be unjustifiably
advantaged over those who have worked full-time.
25.
I shall now consider the arguments set out above. Under the rules governing job-sharing, just as under
the rules for full-time work, progression on the pay scale depends on an assessment of both the quality
and quantity of the work performed (see, to this effect, Circular No 9/87 of the Department of Finance).
These are two inseparable assessment criteria, which operate together in forming the decision on
advancement which the administration has to take. Given the same qualitative assessment of quality of
two workers, the employee working in a job-sharing capacity will progress on the pay scale whilst he is
job-sharing in parallel with the full-time worker. Each incremental step corresponds to half of the pay of
the full-time employee. Accordingly, the hourly pay of the two categories of worker is the same at every
level of the scale. Yet, when the job-sharer converts to full-time work, his position is automatically
reviewed, with the result that he is placed at a lower level on the full-time scale than he occupied on the
job-sharing scale.
26.
That method of calculation is based on the number of hours actually worked and on strictly pro-rata
criteria. Can this constitute indirect discrimination? In my view, it can. The regression provided for in the
rules has direct effects on pay. Workers receive less, in real terms, than twice what they would have
earned job-sharing13: they therefore suffer a reduction in their hourly rate of pay. On close consideration,
this situation is the mirror-image of that created by the German rules in issue in Nimz. In that case, by
providing that account should be taken of only half of the hours worked, the national legislation prevented
(or, more properly, delayed) part-time workers from gaining access to a higher level in the pay scale. By
contrast, in this case, under the rules in question workers returning to full-time work are unable to
preserve the level on the 730 pay-scale which they had reached while job-sharing. In both cases,
therefore, employees working part-time (here the expression is to be understood in the broad sense) are
denied pay benefits which are, however, granted to those doing the same work full-time. This difference of
treatment in terms of pay within the category of full-time workers affects those who have previously
worked in a job-sharing capacity and are regressed in relation to the position which they had already
achieved on the pay scale.
13
In this regard it might be interesting to note the figures supplied by the claimants in their observations for loss of hourly
pay. As a result of regression Ms Hills hourly pay went down from Ir 618 at the ninth level to Ir 600 at the eighth
level; in Ms Stapletons case, the hourly loss was the 18 pence difference between the rates at the second and third
levels (see the claimants observations, points 5.5 and 5.6).

On examination, this difference in treatment is based solely on the mechanism provided in order for
the time work variable to be taken into account in calculating pay when the job-sharer converts to full-time
work. I cannot but conclude, if I may paraphrase Advocate General Darmon, that there can be no doubt:
employees converting from job-sharing are subject to discrimination in relation to full-time employees;
they require twice the length of service in order to remain, on moving to full-time work, at the level of pay
which they had reached on the scale for part-time work (see Advocate General Darmons opinion in Nimz
[1991] ECR I-297 (para 7)).
27.
In my view, this first conclusion is not contradicted by the courts decision in Helmig [1994] ECR I-
5727. In that case the issue was the right of part-time workers to overtime payments on the same basis
as full-time workers when they had completed hours in excess of the set part-time working hours. The
court found that the rules in question were not discriminatory. Part-time workers could not claim the right
to overtime supplements except under conditions in substance equal to the conditions applicable to full-
time workers, and provided always that they had actually worked in excess of the normal full-time working
hours. As is evident, the dispute then concerned the possibility of treating employees assigned to the
same work in a different manner for the purposes of calculating pay. Giving part-time workers the right to
overtime supplements as from the first hour worked over and above their working hours would have led to
the unacceptable result of paying full-time workers proportionately less than part-timers. The court held
that there is unequal treatment wherever the overall pay of full-time employees is higher than that of part-
time employees for the same number of hours worked on the basis of an employment relationship (see
[1994] ECR I-5727 (paras 2630)). That is the correct view, which holds good, as the Commission notes,
in this case as far as concerns the amount of pay due to the employee while job-sharing. As long as the
persons concerned are employed in a job-sharing capacity, their pay is calculated on the basis of the
work actually performed, and in no other way.
28.
In the case now before the court, however, the pay aspect is different. On close study, the court is not
faced, as it then was, with the calculation of pay for employees required to work according to different
systems. In our case, the court is called upon to ascertain whether or not persons working in a job-sharing
capacity have the right to preserve their hourly rate of pay, once they have exercised their right to return
to full-time work. In essence, if the strictly pro-rata criterion adopted in Helmig were to apply in the case in
point, as the Irish and UK governments propose, that would signify, to all intents and purposes, calling in
question workers hourly pay, thus making some of their financial rights nugatory 14. For my part, I do not
believe that I can accept such a result.
14
Here, I consider it useful to recall a subsequent measure which bears out the confusion created in this area of the Irish
system: the circular of the Department of Finance of 14 July 1994 dealing with Job-Sharing Staff and Increments. In
paras 3 and 4 the circular tackled various problems connected with the system of calculation which had arisen where a
person returned to job-sharing after resuming full-time duties for a period (para 3) and where a person was requested to
resume full-time work for a period (para 4). In both cases, the administration found it necessary to adopt mark-time
provisions to ensure that the workers preserved their financial entitlements. This is a significant example of the
consequences for workers rights which may ensue from the rules.

Let me dwell on this point. If it is held to be lawful to reduce the hourly pay of workers moving from job-
sharing, this will be tantamount to reducing ex post the pay they received during the years in which they
worked reduced hours15. To my mind, this disparity in the hourly pay of a full-time worker compared with a
731
part-time worker is directly contrary to the principle of equal treatment upon which the judgment in
Helmig is based.
15
An example in figures: a worker job-sharing at the first level earns Ir 447 an hour; at the second level Ir 468 an hour
and at the third level Ir 486 an hour. Her average hourly pay over three years is Ir 467 an hour; by contrast, if the
Irish governments calculation method is used, her average pay per hour is Ir 460. It can therefore be seen that the
calculation used leads to an ex post reduction in the workers hourly earnings.

29.
Similar kinds of objections are encountered by the Irish administrations argument that if full-time
employees who had formerly chosen to share jobs were to be treated in the same way as those who have
always worked full-time, the first category of workers would be unjustifiably advantaged in relation to the
second category. That is not in fact the case. I repeat: we are talking about calculating pay when
someone leaves job-sharing; that method of calculation incorporates as a variable the fact that the person
employed in a job-sharing capacity has worked for double the period required of a full-time worker. That is
where the discrimination is to be found. In order to refute the governments argument suffice it to observe
that workers who work, and while they work, in a job-sharing capacity, are paid in proportion to the work
actually performed. That is not treating different situations in the same way: pay for job-sharing is different
from pay for full-time work.
30.
Extension of such different treatment, still on account of the pro-rata criterion, to employees moving
from job-sharing to full-time work results in a downward valuation of the work they previously did when
they were employed part-time. To my mind, that cannot be squared with the judgment of the court in
Nimz. The important thing to consider is the effect of such a valuation. Recourse to the criterion of hours
worked during the period of part-time employment, as provided for by the Irish rules, introduces a
retrospective disparity in the overall pay of workers performing the same duties in their employment both
in terms of quality and quantity: workers are treated as belonging to different categories. Hence the
conclusion that there is unequal treatment. It is apparent from the order for reference that the majority of
the employees concerned are women: the calculation criterion used by the Irish government does
therefore constitute discrimination against them (see the judgment in Nimz [1991] ECR I-297 (para 12)).
The second question
31.
I must now begin to consider the second question. The national court asks the court to specify whether
the employer is required to provide special justification for recourse to the criterion of service, defined as
actual time worked.
32.
In essence, the court is called upon to clarify what it held in its judgment in Danfoss [1989] ECR 3199
(para 24). There, it laid down the principle that since length of experience goes hand in hand with
experience and since experience generally enables the employee to perform his duties better, the
employer need not provide special justification for recourse to the criterion of length of service.
732
33.
To my way of thinking, such a statement can only be considered in the light of the facts of the case
then before the court. Let me briefly recall them. Under a clause in a collective employment agreement,
Danfoss paid individual supplements to employees to reward flexibility, training and length of service 16.
The system used by the employer to calculate those supplements was absolutely lacking in transparency,
with the result that it was impossible for employees to find out the actual break-down of the supplements
paid to them. Recognising the need for workers to know the detailed break-down of their pay, the court
went on to assess the individual criteria for the supplements, and in respect of the criterion of length of
service it made the statement quoted above.
16
Under the collective agreement the employer was free to grant supplements on the basis of certain individual
characteristics of the employee (see Advocate General Lenzs opinion in Danfoss [1989] ECR 3199 (paras 4243)).

34.
That judicial opinion, the meaning of which the national court now wishes the court to define, cannot
be understood in isolation from the facts set out above. Some analysis is called for here. When the
employer takes an individual decision it is certainly lawful to take length of service into account as one
factor in granting the employee a differential pay supplement. It may be that greater experience in the
post enables the employee to work more efficiently. What does, however, give rise to doubts is the use of
a criterion which generalises recourse to length of service, so as to extend it indiscriminately even to
cases in which it may be unjustified.
This is an idea which I have already discussed in my joint opinion in Gerster and Kording but to which I
believe I must now return. The criterion of length of service must be supported by adequate justification
where it is applied to a series of work relationships in respect of which it is far from proved that length of
service can be equated with competence. Different treatment of part-time and full-time workers is not
justified where it is assumed, in a general way and merely on the basis of strictly proportional criteria, that
workers in the first category are per se less deserving of pay supplements. Thus it is in our case. The
result is the discrimination we have just seen between men and women, given the fact that the great
majority of part-time workers are women17.
17
On the other hand, I must indicate the lines along which some learned writers have interpreted the relationship between
Danfoss and Nimz. In particular, More Seniority pay for part-time workers (1991) European Law Review p 826 (sic) has
written: This (the finding on that point in Nimz) appears to contradict the view expressed by the Court of Justice in
Danfoss in 1989, when it stated that seniority goes hand in hand with experience which generally places a worker in
a better position to carry out his duties and that it was permissible for an employer to reward seniority without having to
establish the importance which it takes on for the performance of the specific duties to be entrusted to the worker. The
latter statement, in particular, appears to be at variance with the view of the court in Nimz that there should be
identifiable connection between rewarding seniority and an improved performance of the task in hand. This suggests
that the courts view of justifying seniority payments as expressed in Danfoss can no longer be regarded as valid. This is
a welcome development and can, perhaps, be regarded as the major contribution of the Nimz judgment.

35.
I must now consider whether the solution adopted by the court in Danfoss may profitably be
transposed to the case in point. I do not think it can. Indeed, it seems clear to me that the principle laid
down in that case must not only be construed having regard to the above-mentioned caveats but must
also apply only to length of service reckoned in terms of years and not of hours actually worked. In that
decision the court gave an important explanation of the way in 733 which the criterion of length of service
must be used in order to avoid unlawful discrimination between men and women:
as regards the criterion of length of service, it is also not to be excluded that it may
involve less advantageous treatment of women than of men in so far as women have entered the
labour market more recently than men or more frequently suffer an interruption of their career.
(See Danfoss [1989] ECR 3199 (para 24); my emphasis.)
To my mind, this explanation leads us to draw a distinction between length of service reckoned in years
which the employer can take into consideration in deciding on promotions without having to establish the
importance it has in the performance of specific tasks entrusted to the employee (para 24)and length of
service reckoned in hours worked, whose relevance for the purposes of progression to a higher rate of
pay must, in contrast, be proved by objective evidence.
36.
To round off this reasoning, and the better to explain it, I shall add a few other observations which will
help to distinguish this case from Nimz and to bear out the conclusion I have proposed. The job-sharing
scheme undoubtedly possesses specific features differentiating it from part-time work. Each of the
workers who share the job is responsible for the work carried out by the entire teamif I may so
describe itmade up of the two officers sharing the job 18. In addition, employees who have opted for job-
sharing may not perform any work other than their work for the administration.
18
See, to this effect, the analysis of the tasks performed by the claimants as set out in their observations, highlighting the
high degree of integration and communication in both job-sharing partners (point 3.7). More generally, compare the
statements in point 2.3 of the observations concerning the fact that the job-sharers were not on fixed-term contracts:
There are no part-timers in the [Irish] Civil Service and no job-sharers on temporary contracts.

In the light of both those aspects of job-sharing, the employment relationship can be seen to have
special characteristics different from those of full-time work. Employees working in a job-sharing capacity
are in a position halfway between part-time and full-time work. Their employment presents features of
both categories: it resembles full-time work from the point of view of the responsibility for the work in its
entirety and part-time work because the hours to be worked are half the full-time hours. In my view, all of
this must necessarily be taken into account in the assessment required as to the work performed and
experience acquired by those who have worked as job-sharers. That is one more reason for considering
that in the circumstances of this case it is unjustifiable to have recourse to length of service reckoned in
hours worked19. Moreover, that argument of mine is indirectly, but significantly, confirmed by the
statements of the national court itself. It says that it has found that the job-sharing period may give the
employee the equivalent experience of a full-time employee.
19
On this point it would be apt to recall the words of Advocate General Darmon in his opinion in Nimz [1991] ECR I-297
(para 14), where, when considering the United Kingdoms argument, he stated that: Certainly it could be considered
necessary to take account of length of service not in years but in hours of work performed, as proposed by the United
Kingdom, but on condition that it is to facilitate progression to different duties, entailing new responsibilities and
consequently presupposing a certain amount of experience. In the present case, however, the length of service required
is merely intended to enable access to a higher wage grade without any alteration in the nature of the duties performed.

37.
It follows that the criterion of length of service, defined as time actually worked, needs objective
justification. The criterion will be justified where it is 734 clear that full-time employment is the only
solution to be adopted in order for the employee to obtain the requisite experience for his intended duties.
Otherwise, a different sort of justification will have to be found that will in any event satisfy the criteria laid
down by the court in Bilka-Kaufhaus GmbH v Weber von Hartz Case 170/84 [1986] ECR 1607. It is clear
from that judgment, as far as concerns this dispute, that the objective pursued by the administration in
regulating its employees incremental steps must correspond to a real need on its part. Moreover, the
means chosen with a view to achieving it must be appropriate and necessary (see [1986] ECR 1607 (para
36)).
The third question
38.
That last area of inquiry is precisely that raised by the national courts third and last question. Is it, or is
it not, possible to regard the practice of making incremental steps depend on time actually worked as
objectively justified on the basis of factors other than the acquisition over time of a given level of skills and
experience?
39.
With regard to that question, the claimants argue that there is no justification whatsoever which is both
objective and necessary as required by the courts case law. For its part, the administration argues that
the Irish system is objectively justifiable on an implied series of grounds which satisfy the requirements
set out in the case law of the court20.
20
As stated in the order for reference (point 21), the administration has put forward the following reasons to justify its
rules: (a) the rules are in keeping with the standard practice of the civil service whereby incremental progression is
related to service and only actual paid service is credited; (b) this practice is valuable to the employer in that it gives
incentives to improve the quality of work performed; (c) to make an exception of job-sharing service would lead to
arbitrary and inequitable situations; (d) to make such an exception because the majority of job-sharers are women
would amount to discrimination in favour of women; (e) the present practice ensures that the incremental cost of job-
sharing staff is the same as that of full-time staff, thus making the cost of work done by job-sharers the same as the cost
of work done by full-time staff.

40.
It must be stated that the court has consistently held that
it is for the national court to determine whether and to what extent the grounds put forward
by an employer to explain the adoption of a pay practice may be regarded as objectively justified
economic grounds.
It is therefore for the national court to assess the grounds adduced in support of the criteria adopted by
the administration in the case pending before it. Once more referring to settled case law, the court may,
however, supply general and other more specific relevant guidance concerning the reasons put forward
by the administration during the proceedings21.
21
See Rinner-Khn and Nimz. See also the opinion of Advocate General Darmon in Nimz [1991] ECR I-297 (paras 10
11); the judgments in Secretary of State for Social Security v Thomas Case C-328/91 [1993] 4 All ER 556, [1993]
ECR I-1247 and Kuratorium fr Dialyse und Nierentransplantation eV v Lewark Case C-457/93 [1996] ECR I-243 (para
32); and the joint opinion in Gerster [1997] ECR I-5253 and Kording [1997] ECR I-5289 (para 43).

41.
With regard to the first justification relied onthat it is standard practice in the Irish civil service to
credit only actual paid serviceI fail to see how it can serve in the circumstances of the case. As the
court has held, generalisations about certain categories of workers are unacceptable unless they are
supported by objective criteria and are unrelated to any discrimination on grounds of sex (see Rinner-
Khn [1989] ECR 2743 (para 14)). The argument put forward by the Irish government has already been
refuted, in substance, in the judgment in 735 Nimz22. There the court rejected the proposition that
classification of part-time workers in a higher salary grade should depend solely on the number of hours
worked. That argument did not prevail in Nimz and it does not deserve to be upheld in this case.
22
See the opinion of Mr Darmon in Nimz [1991] ECR I-297 (para 14).

42.
To my mind, the second reason given by the administration is equally unconvincing. It argues that the
need to give staff sufficient incentives is objective justification. To maintain motivation and a sense of
vocation in the long term, or to raise the morale of staff, is a valid and laudable aim. All the more so when,
as the representative of the appellant administration observed at the hearing, the career of clerical
assistants is long and needs, if I may so express it, to be invigorated by continuing stimuli. However, it
seems to me that such an objective could be pursued even if the disparity in pay were abolished and
workers converting from job-sharing to full-time work could claim the length of service they had acquired
in their previous work experience. While it is of course for the national court to make a definitive
determination with regard to that point I for one cannot see how the equal treatment as regards pay
claimed by the workers concerned can possibly conflict with the administrations declared purpose. Nor,
therefore, do I understand how the requirement of need laid down in the case law referred to can be
regarded as satisfied in order to justify the discriminatory measure adopted in this case. In any event, it is
for the authority concerned to supply actual proof that there is a basis for the alleged justification.
43.
What is to be said of the discriminatory consequences which, according to the administration, would
be produced by the adoption of an ad hoc solution for a group, such as that of job-sharers, the majority of
whom are women? I have already explained that there is no discrimination in favour of women. On the
contrary. Those who have worked, and continue to work, full-timea group of which women do not form
the majorityreceive twice the pay. The disadvantaged group is the group, notable for the preponderance
of women over men, of those who have job-shared and move on to full-time work. The disadvantage can
only be removed by recognising their right to the incremental steps which they have previously accrued
while job-sharing, as I have already explained.
44.
Nor, lastly, am I persuaded by the justification concerning the administrations financial needs. That,
under the present system, the incremental cost of job-sharers is equal to that of full-time employees does
not strike me as a meritorious argument. It does not prove that the means chosen is necessary in order to
satisfy a real need on the authoritys part, as required by the judgment in Bilka. I cannot for my part
discover any reasons why the economic needs which the administration asserts cannot be appropriately
satisfied in a different manner, without discrimination as regards pay between one group and the group of
full-time workers. In short, the authority concerned must prove that the criterion adopted is justified, in
preference to and to the exclusion of that put forward by the claimants, by the expenditure effects of the
claimants increments.
45.
For the reasons set out above I propose that the Court of Justice should reply as follows to the
questions referred by the national court:
(1) Article 119 of the Treaty and the rules regarding equal pay must be interpreted as meaning
that, where a greater percentage of women than men are employed on a job-sharing basis, the pay
increments awarded by 736 reference to time actually worked to those exercising the right to move
from job-sharing to full-time work may not be organised in such a way that the individuals
concerned are placed at a lower grade in the pay scale than other workers on full time who have
the same length of service measured in years.
(2) Where an employer awards a pay increment on the basis of hours actually worked, it must
prove that that criterion corresponds to a real need on its part and is effective, necessary and
appropriate with a view to achieving the objectives pursued by the employer.
(3) It is for the national court to determine whether the practice of correlating pay increments
with time actually worked is objectively justified.

17 June 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
By order of 5 April 1995, received on 12 July 1995, the Labour Court referred to the Court of Justice of
the European Communities for a preliminary ruling under art 177 of the EC Treaty three questions on the
interpretation of Council Directive (EEC) 75/117 on the approximation of the laws of the member states
relating to the application of the principle of equal pay for men and women (OJ 1975 L45, p 19).
2.
Those questions have arisen in a dispute between Ms Hill and Ms Stapleton, who had previously
worked in a job-sharing capacity, and the Revenue Commissioners and Department of Finance
concerning the latters decision to place them, on their conversion to full-time employment, on a point of
the full-time pay scale lower than that of the job-sharing pay scale which they had previously occupied.
3.
Article 119 of the EC Treaty sets out the principle that men and women should receive equal pay for
equal work. The second paragraph of art 119 provides:
For the purpose of this article, pay means the ordinary basic or minimum wage or salary and
any other consideration, whether in cash or in kind, which the worker receives, directly or indirectly,
in respect of his employment from his employer.
4.
Article 1 of the directive refers to the principle of equal pay, which means, for the same work or for
work to which equal value is attributed, the elimination of all discrimination on grounds of sex with regard
to all aspects and conditions of remuneration.
5.
Job-sharing was introduced in Ireland by a government decision in 1984, primarily with a view to
creating employment.
6.
Job-sharing, established by Circular 3/84, involved an arrangement under which two civil servants
shared equally one full-time job, such that the benefits were shared equally by both persons concerned
and the costs of the post to the administration remained the same. Staff recruited on a full-time basis
could decide to participate in that scheme and retained the right to return to full-time employment at the
end of the period for which they had opted to job-share, provided that suitable vacancies existed. Staff
recruited on a job-sharing basis between 1986 and 1987 were entitled to be appointed to full-time
positions within two years of appointment, again provided that suitable vacancies existed. Staff choosing
to job-share were required to give a written undertaking that they would not engage in outside
employment. This scheme was amended in 1988, 737and since then persons taken on to job-share are
recruited on temporary contracts and have no entitlement to full-time employment.
7.
Circular 3/84 provided that:
For each grade in which job-sharing arrangements operate, the scale of pay applicable to job-
sharing staff shall be a scale each point of which shall represent 50% of the corresponding point on
the scale of pay appropriate to full-time staff. Increments on this scale will be granted annually if the
officers services are satisfactory.
8.
Circular 3/84 did not define the rules regarding incremental credit for job-sharing service on conversion
to full-time work. This matter was, however, clarified in a letter sent to departments on 31 March 1987 by
the Department of Finance, stating as follows:
The position is that, as each years job-sharing service is reckonable as six months full-time
service, an officer who has served for two years in a job-sharing capacity should be placed on the
second point of the full-time scale (equivalent to one years full-time service). In cases where
officers have been job-sharing for more than two years, the incremental date should be adjusted on
a pro-rata basis.
9.
Circular 9/87, which is the first comprehensive circular on increments, replaces all previous circulars
on the subject. Article 2 of Circular 9/87 provides as follows: An increment is an increase in pay for which
provision is made in a pay scale. As a general rule, increments are granted annually provided an officers
services are satisfactory.
10.
Ms Hill and Ms Stapleton were recruited to the Irish Civil Service through open competitions for the
grade of Clerical Assistant and were assigned to the office of the Revenue Commissioners. Ms Hill was
recruited in July 1981 and began job-sharing in May 1988. Ms Stapleton was recruited in a job-sharing
capacity in April 1986. Ms Hill and Ms Stapleton were employed in a job-sharing capacity for two years.
They worked exactly half the time which a full-time employee would have worked, on a one week on/one
week off basis. During their respective job-sharing periods of employment, each moved one point up the
incremental scale with each year of service and was paid at the rate of 50% of the salary for clerical
assistants, according to the point each had reached on the scale.
11.
Ms Hill returned to full-time employment in June 1990. At that time she had reached the ninth point on
the incremental job-sharing scale. Initially, on her return to full-time work, she was assimilated to the ninth
point of the corresponding scale, but was subsequently placed on the eighth point, on the ground that two
years job-sharing were equivalent to one years full-time service.
12.
Ms Stapleton secured a full-time post in April 1988. She had at that time reached the third point on the
incremental job-sharing scale. She continued to move up the scale in 1989 and 1990 to the fourth and
fifth points respectively, but was informed in April 1991 that there had been a mistake in her classification,
with the result that she was unable to progress to the sixth point. She was informed that her two years
job-sharing service should have been counted as one years full-time service.
13.
Ms Hill and Ms Stapleton challenged the decision reclassifying them before an Equality Officer,
pursuant to s 7(1) of the Anti-Discrimination (Pay) Act 1974. Relying on the courts judgment in Nimz v
Freie und Hansestadt Hamburg Case 738 C-184/89 [1991] ECR I-297, the Equality Officer found in their
favour on the ground that their employer was precluded from applying a provision to the effect that only
paid service should be taken into account for progression on the incremental scale.
14.
The Revenue Commissioners and the Department of Finance appealed to the Labour Court against
that recommendation. Ms Hill and Ms Stapleton appealed for the implementation of that recommendation.
15.
Taking the view that the outcome of the dispute depended on the interpretation of Community law, the
Labour Court, exercising the powers conferred on it by s 8 of the Anti-Discrimination (Pay) Act 1974,
referred the following three questions to the Court of Justice for a preliminary ruling:
 In circumstances in which far more female workers than male workers spend part of their
working lives in a job-sharing capacity, (a) Does a prima facie case of indirect discrimination arise
where job-sharing workers who convert to full-time work are given credit for incremental
progression on the scale of pay for full-time staff by reference to actual time worked such that,
while the benefits awarded to them are fully pro-rated to those awarded to staff who have always
worked full-time, they are placed at lower points on the full-time scale than comparators who are in
all respects similar to them except that they have worked continuously on a full-time basis? In other
words, is the principle of equal pay, as defined in Directive 75/117/EEC, contravened if employees,
who convert from job-sharing to full-time work, regress on the incremental scale, and hence on
their salary scale, due to the application by the employer of the criterion of service calculated by
time worked in a job? (b) If so, does the employer have to provide special justification for recourse
to the criterion of service, defined as actual time worked, in awarding incremental credit? (c) If so,
can a practice of incremental progression by reference to actual time worked be objectively justified
by reference to factors other than the acquisition of a particular level of skill and experience over
time?
16.
By those three questions, which it is appropriate to examine together, the national tribunal is in
substance asking whether there is discrimination in a case where workers who convert from job-sharing
to full-time work regress on the incremental scale, and hence on their salary scale, due to the application
by the employer of the criterion of service calculated by the length of time actually worked in a post. If
application of that criterion gives rise to indirect discrimination, the national tribunal asks whether this can
be justified.
17.
It is apparent from the case file that the national system in question places workers who convert from
job-sharing to full-time work at a disadvantage compared with those who have worked on a full-time basis
for the same number of years in so far as, when converting to full-time work, a job-sharing worker is
placed, on the full-time pay scale, at a level below that which he or she previously occupied on the pay
scale applicable to job-sharing staff and, consequently, at a level lower than that of a full-time worker
employed for the same period of time.
18.
It must be recalled at the outset that art 119 of the Treaty sets out the principle that men and women
should receive equal pay for equal work. As the court has already held in Defrenne v Sabena Case 43/75
[1981] 1 All ER 122, [1976] ECR 455 (para 12), that principle forms part of the foundations of the
Community.
739
19.
Furthermore, the court also held in Jenkins v Kingsgate (Clothing Productions) Ltd Case 96/80 [1981]
1 WLR 972, [1981] ECR 911 that art 1 of the directive, which is essentially designed to facilitate the
practical application of the principle of equal pay outlined in art 119 of the Treaty, in no way alters the
scope or content of that principle as defined in the latter provision.
20.
In order to provide the national tribunal with a helpful answer, it is necessary first of all to ascertain
whether the system for classifying workers converting from job-sharing to full-time employment comes
within the scope of art 119 and, consequently, within the scope of the directive.
21.
It should be noted in this regard that the system in question determines the progression of pay due to
those workers. It follows that the system comes within the concept of pay for the purposes of art 119.
22.
As the court held in Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319,
[1995] ECR I-225 (para 30), discrimination can arise only through the application of different rules to
comparable situations or the application of the same rule to different situations.
23.
It should be noted in that regard, as indeed the Labour Court has found, that it has not been
established that the unfavourable treatment applied to Ms Hill and Ms Stapleton constitutes direct
discrimination on grounds of sex. It is thus necessary to examine whether that treatment may amount to
indirect discrimination.
24.
According to settled case law, art 4(1) of the directive precludes the application of a national measure
which, although formulated in neutral terms, works to the disadvantage of far more women than men,
unless that measure is based on objective factors unrelated to any discrimination on grounds of sex (see,
to that effect, De Weerd, ne Roks v Bestuur van de Bedrijfsvereniging voor de Gezondheid, Geestelijke
en Maatschappelijke Belangen Case C-343/92 [1994] ECR I-571 (para 33) and Megner v
Innungskrankenkasse Vorderpfalz Case C-444/93 [1996] All ER (EC) 212, [1995] ECR I-4741 (para 24)).
25.
It is apparent from the case-file that 992% of clerical assistants who job-share are women, as are
98% of all civil servants employed under job-sharing contracts. In those circumstances, a provision which,
without objective justification, adversely affects the legal position of those workers coming within the
category of job-sharers has discriminatory effects based on sex.
26.
 The Labour Court takes the view that job-sharing is in a unique category as it does not involve a break
in service. The specific feature which distinguishes part-time work from job-sharing is that in the case of
the latter the work and responsibilities are shared between two employees. According to the case-file, full-
time commitment and coordination of responsibilities related to the job being shared can be required of
job-sharers.
27.
It must be borne in mind that the Labour Court found that Ms Hill and Ms Stapleton were doing the
same work as their colleagues who were working on a full-time basis and were in a situation comparable
to theirs. As has been pointed out in para 6 of the present judgment, staff choosing to job-share were
required to give a written undertaking that they would not engage in outside employment. The Labour
Court is satisfied that a job-sharer can acquire the same experience as a full-time worker. The only
difference between a job-sharer and a colleague working full-time lies in the time actually worked during
the period of job-sharing.
28.
It should also be remembered that, during the entire period for which they job-shared, Ms Hill and Ms
Stapleton progressed by one incremental point each 740 year and were paid at the rate of 50% of the
salary for clerical assistants, according to the point each had reached on the scale.
29.
Under the rules applicable to job-sharing, as well as those applicable to the scheme for full-time
workers, incremental pay progression depends on an assessment of both the quality and quantity of the
work performed. On the basis that the qualitative assessment of the two categories of worker in question
is identical, job-sharers, when performing their work in a job-sharing capacity, progress along the pay
scale in parallel to full-time workers. A figure equivalent to one-half of the pay provided for full-time
workers corresponds to each point of incremental progression. The hourly pay for the two categories of
worker is thus identical at each point along the scale.
30.
In those circumstances, when workers convert from job-sharing, under which they will have worked for
50% of full time, receiving 50% of the salary corresponding to that point on the pay scale for full-time
work, they should expect both their hours of work and the level of their pay to increase by 50%, in the
same way as workers converting from full-time work to job-sharing would expect those factors to be
reduced by 50%, unless a difference in treatment can be justified.
31.
However, there is no such progression in this case. When job-sharing workers convert to full-time
work, their situation is automatically reviewed in such a way that they are placed, on the full-time pay
scale, at a level lower than that which they occupied on the pay scale applicable to job-sharing.
32.
The regression to which workers are subject when entering or returning to full-time work directly
affects their pay. They are in fact paid less than double what they would have earned had they been job-
sharing. Consequently, their hourly rate of pay is reduced. Reference to the criterion of hours worked
during the period of job-sharing employment, as provided for under the scheme applicable here, fails to
take account either of the fact that job-sharing, as pointed out in para 26 of the present judgment, is in a
unique category as it does not involve a break in service, or of the fact, stated in para 27 of the present
judgment, that a job-sharer can acquire the same experience as a full-time worker. Furthermore, a
disparity is retroactively introduced into the overall pay of employees performing the same functions so far
as concerns both the quality and the quantity of the work performed. The result of this disparity is that
employees working full time but who previously job-shared are treated differently from those who have
always worked on a full-time basis.
33.
Within the category of full-time workers, therefore, there is unequal treatment, as regards pay, of
employees who previously job-shared, and who regress in relation to the position which they already
occupied on the pay scale.
34.
In such a case, provisions of the kind at issue in the main proceedings result in discrimination of
female workers vis--vis male workers and must in principle be treated as contrary to art 119 of the Treaty
and therefore contrary to the directive. It would be otherwise only if the difference of treatment found to
exist between the two categories of worker were justified by objective factors unrelated to any
discrimination based on sex (see, along these lines, the judgments in Bilka-Kaufhaus GmbH v Weber von
Hartz Case 170/84 [1986] ECR 1607 (para 29), Rinner-Khn v FWW Spezial Gebudereinigung GmbH
Case 171/88 [1989] ECR 2743 (para 12) and Kuratorium fr Dialyse und Nierentransplantation eV v
Lewark Case C-457/93 [1996] ECR I-243 (para 31)).
35.
It is for the national court, which has sole jurisdiction to assess the facts and interpret the national
legislation, to determine whether and to what extent a 741 legislative provision which, though applying
independently of the sex of the worker, actually affects a greater number of women than men is justified
by objective reasons unrelated to any discrimination on grounds of sex (see the judgments in Jenkins
[1981] 1 WLR 972, [1981] ECR 911 (para 14), Bilka-Kaufhaus [1986] ECR 1607 (para 36) and Rinner-
Khn [1989] ECR 2743 (para 15)).
36.
 However, although in preliminary-ruling proceedings it is for the national court to establish whether
such objective factors exist in the particular case before it, the Court of Justice, whose task is to provide
the national court with helpful answers, may provide guidance based on the documents in the main
proceedings and on the written and oral observations which have been submitted to it, in order to enable
the national court to give judgment (see the judgments in Secretary of State for Social Security v Thomas
Case C-328/91 [1993] 4 All ER 556, [1993] ECR I-1247 (para 13) and Lewark [1996] ECR I-243 (para
32)).
37.
In the view of the Revenue Commissioners and the Department of Finance, a method of pay
progression based on the duration of work actually performed is objectively justified by criteria which
satisfy the conditions laid down by the court in its case law.
38.
In this connection, neither the justification provided by the Revenue Commissioners and the
Department of Finance to the effect that there is an established practice within the Civil Service of
crediting only actual service, nor that stating that this practice establishes a reward system which
maintains staff motivation, commitment and morale, is relevant. The first justification is no more than a
general assertion unsupported by objective criteria. With regard to the second, the system of
remuneration for employees working on a full-time basis cannot be influenced by the job-sharing scheme.
39.
So far as concerns the justification that, if an exception were to be made in favour of job-sharing, this
would result in arbitrary or inequitable situations or would amount to impermissible discrimination in favour
of women, it should be pointed out, as is clear from para 29 of the present judgment, that to grant to
workers who convert to full-time employment the same point as that which they had under their job-
sharing contract does not constitute discrimination in favour of female workers.
40.
So far as the justification based on economic grounds is concerned, it should be noted that an
employer cannot justify discrimination arising from a job-sharing scheme solely on the ground that
avoidance of such discrimination would involve increased costs.
41.
It must be borne in mind that all the parties to the main proceedings, and the national tribunal, agree
that almost all job-sharing workers in the Irish public sector are women. It is apparent from the case-file
that approximately 83% of those who chose that option did so in order to be able to combine work and
family responsibilities, which invariably involve caring for children.
42.
Community policy in this area is to encourage and, if possible, adapt working conditions to family
responsibilities. Protection of women within family life and in the course of their professional activities is,
in the same way as for men, a principle which is widely regarded in the legal systems of the member
states as being the natural corollary of the equality between men and women, and which is recognised by
Community law.
43.
The onus is therefore on the Revenue Commissioners and the Department of Finance to establish
before the Labour Court that the reference to the criterion of service, defined as the length of time actually
worked, in the assessment of the incremental credit to be granted to workers who convert from 742 job-
sharing to full-time work is justified by objective factors unrelated to any discrimination on grounds of sex.
If such evidence is adduced by those authorities, the mere fact that the national legislation affects far
more women than men cannot be regarded as constituting a breach of art 119 of the Treaty and,
consequently, a breach of the directive.
44.
The answer to the questions submitted must therefore be that art 119 of the Treaty and the directive
are to be interpreted as precluding legislation which provides that, where a much higher percentage of
female workers than male workers are engaged in job-sharing, job-sharers who convert to full-time
employment are given a point on the pay scale applicable to full-time staff which is lower than that which
those workers previously occupied on the pay scale applicable to job-sharing staff due to the fact that the
employer has applied the criterion of service calculated by the actual length of time worked in a post,
unless such legislation can be justified by objective criteria unrelated to any discrimination on grounds of
sex.

Costs
45.
The costs incurred by the UK government and the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the action pending before the national tribunal, the decision on costs is a
matter for that tribunal.
On those grounds, the Court of Justice (Sixth Chamber), in answer to the questions referred to it by
the Labour Court by order of 5 April 1995, hereby rules: art 119 of the EC Treaty and Council Directive
(EEC) 75/117 on the approximation of the laws of the member states relating to the application of the
principle of equal pay for men and women are to be interpreted as precluding legislation which provides
that, where a much higher percentage of female workers than male workers are engaged in job-sharing,
job-sharers who convert to full-time employment are given a point on the pay scale applicable to full-time
staff which is lower than that which those workers previously occupied on the pay scale applicable to job-
sharing staff due to the fact that the employer has applied the criterion of service calculated by the actual
length of time worked in a post, unless such legislation can be justified by objective criteria unrelated to
any discrimination on grounds of sex.

743

[1998] All ER (EC) 744

Customs and Excise Comrs v First National Bank of Chicago

(Case C-172/96)
EUROPEAN COMMUNITY; Taxation: TAXATION; VAT and Customs and Excise
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FIFTH CHAMBER)
JUDGES GULMANN (PRESIDENT OF THE CHAMBER), WATHELET, MOITINHO DE ALMEIDA, JANN
AND SEVN (RAPPORTEUR)
ADVOCATE GENERAL LENZ
25 JUNE, 16 SEPTEMBER 1997, 14 JULY 1998
European Community Value added tax Supply of goods or services Supply for a consideration
Consideration Direct link between supply and consideration Legal relationship between provider and
recipient of service Reciprocal performance Foreign exchange transactions Bank not charging
transaction fees or commission Whether transactions constituting supply of goods or services effected
for consideration Council Directive (EEC) 77/388, art 2(1).
European Community Value added tax Value of supply of goods or services Determination of value
Foreign exchange transactions Bank not charging transaction fees or commission Whether value of
consideration full value of currency received in exchange for that supplied by bank Whether
consideration amount that bank could actually apply to its own use Whether consideration net result of
transaction over given period Council Directive (EEC) 77/388, art 11A(1)(a).

The plaintiff bank carried on a wide range of banking activities, including foreign exchange dealing. The
bid price at which the bank offered to buy a currency was typically slightly below the offer price at which
the bank would sell that currency, the difference between the prices being known as the spread. No
money was delivered physically in the foreign exchange transactions; what was delivered was the
availability of drawing on an account opened with a bank in the currency delivered. Moreover, the bank
did not charge transaction fees or commissions, but sought to make a profit, over a period of time, as a
result of the spread between its bid and offer quotes. The bank was partially exempt for value added tax
purposes, but did have the right to deduct input tax corresponding to transactions completed with counter-
parties outside the Community. In its return for the period from 1 May to 31 July 1994, the bank therefore
took into account the foreign exchange transactions which it had entered into with such counter-parties
over the relevant period and claimed a large input tax credit. However, the commissioners reduced the
input tax credit claimed by disallowing the portion corresponding to the foreign exchange transactions. On
appeal, the High Court stayed the proceedings and referred two questions to the Court of Justice of the
European Communities for a preliminary ruling to determine: (i) whether foreign exchange transactions,
as defined by the British Bankers Association1, constituted a supply of goods or services effected for
consideration within the meaning of art 2(1)2 of Council Directive 77/388 on the harmonisation of the rules
of the member states relating to turnover taxescommon system of 744 value added tax: uniform basis
of assessment; and (ii) if so, what was the taxable amount of such supplies within art 11A(1) 3.
1
The definition is set out at p 764 h, post
2
Article 2(1), so far as material, is set out at p 762 f, post
3
Article 11A(1), so far as material, is set out at p 763 j, post

Held (1) Since currencies exchanged against other currencies in a foreign exchange transaction were
used as legal tender, they could not be regarded as tangible property within art 5 of Directive 77/388.
They were not therefore supplies of goods, but were supplies of services within art 6 (see p 765 d, post).
(2) A supply of services was effected for consideration within art 2(1) and, as such, taxable only if
there was a legal relationship between the provider and recipient of the service pursuant to which there
was reciprocal performance, the remuneration received by the provider constituting the value actually
given in return for the service supplied to the recipient. In the instant case, it was clear that a bilateral
legal relationship existed between the bank and its counter-party under which the two parties gave
reciprocal undertakings to transfer amounts in a given currency and to receive the countervalue in
another currency. Apart from the actual exchange transaction, the service provided by the bank was
characterised by the banks preparedness to conclude such transactions in the currencies in which it
specialised; and the fact that no fees or commission were charged by the bank on a specific transaction
did not mean that no consideration was given. Accordingly, foreign exchange transactions, even when
performed without commission or direct fees, were supplies of services effected for consideration within
the meaning of art 2(1) of the directive (see p 765 e g to p 766 c and p 768 d, post); Tolsma v Inspecteur
der Omzetbelasting Leeuwarden Case C-16/93 [1994] STC 509 applied.
(3) Although they were the subject of a supply, the currencies transferred to the bank by its counter-
party could not be regarded as remuneration for the service of exchanging currencies for other currencies
or consequently as constituting remuneration for that service. Accordingly, in order to determine the
consideration, it was necessary to determine what the bank could actually take for itself. However, since
the bank carried out a large number of transactions relating to different amounts and involving different
currencies, the rates of which were in constant fluctuation, the amount which it could actually apply to its
own use had to be regarded as consisting of the net result of its transactions over a given period of time.
It followed, for the purposes of art 11(A)(1)(a), that the taxable amount was the net result of the
transactions of the supplier of the services over a given period of time (see p 767 c to g and p 768 a e,
post); H J Glawe Spiel- und Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt
Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994] STC 543 applied.

Cases cited
Apple and Pear Development Council v Customs and Excise Comrs Case 102/86 [1988] 2 All ER 922,
[1988] STC 221, [1988] ECR 1443, ECJ.
Argos Distributors Ltd v Customs and Excise Comrs Case C-288/94 [1996] STC 1359, [1997] QB 499,
[1996] ECR I-5311, ECJ.
Fischer v Finanzamt Donaueschingen Case C-283/95 [1998] All ER (EC) 567, [1998] STC 708, ECJ.
Glawe (H J) Spiel- und Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt
Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994] STC 543, [1994] ECR I-1679, ECJ.
745
Muys en De Winters Bouw-en Aannemingsbedrijf BV v Staatssecretaris van Financin Case C-281/91
[1997] STC 665, [1993] ECR I-5405, ECJ.
Naturally Yours Cosmetics Ltd v Customs and Excise Comrs Case 230/87 [1988] STC 879, [1988] ECR
6365, ECJ.
Staatssecretaris van Financin v Coperatieve Aardappelenbewaarplaats GA Case 154/80 [1981] ECR
445.
Tolsma v Inspecteur der Omzetbelasting Leeuwarden Case C-16/93 [1994] STC 509, [1994] ECR I-743,
ECJ.

Reference
By order of 13 May 1996, the Queens Bench Division of the High Court) referred for a preliminary ruling
two questions (set out at p 764 f, post) on the interpretation of Council Directive (EEC) 77/388 on the
harmonisation of the rules of the member states relating to turnover taxescommon system of value
added tax: uniform basis of assessment. Those questions were raised in proceedings between First
National Bank of Chicago (the bank) and the Commissioners of Customs and Excise concerning
deduction of input tax on certain foreign exchange transactions. Written observations were submitted on
behalf of: the bank, by P Lasok QC; the United Kingdom government, by S Ridley, of the Treasury
Solicitors department, acting as agent, N Pleming QC and C Vajda, barristers; the French government,
by C de Salins, Deputy Director in the Legal Affairs Directorate of the Ministry of Foreign Affairs, and G
Mignot, Foreign Affairs Secretary in that directorate, acting as agents; and the European Commission, by
P Oliver and E Traversa, of its legal service, acting as agents. Oral observations were submitted on behalf
of: the bank, represented by D Goy QC; the UK government, represented by J E Collins, Assistant
Treasury Solicitor, acting as agent, N Pleming QC and C Vajda; and the Commission, represented by P
Oliver. The language of the case was English. The facts are set out in the opinion of the Advocate
General.

16 September 1997.

The Advocate General (C O Lenz)

delivered the following opinion (translated from the German).

AINTRODUCTION
1.
In this reference for a preliminary ruling, the Queens Bench Division of the High Court has referred
questions to the Court of Justice of the European Communities concerning the interpretation of Council
Directive (EEC) 77/388 on the harmonisation of the rules of the member states relating to turnover taxes
common system of value added tax: uniform basis of assessment (OJ 1977 L145 p 1) (the Sixth
Directive). The questions are concerned with the taxation of foreign exchange transactions and with
whether the London branch of the First National Bank of Chicago (the bank) is entitled to deduct input tax.

IBackground to the main proceedings

2.
According to the information provided by the national court, the dispute presents itself as follows: the
bank, which is registered for value added tax (VAT) and partly exempt therefrom, agreed with the
Commissioners of Customs and Excise (the commissioners) a special partial exemption method. The
recoverable portion of input tax allocated by the agreed method to the bank, which includes the
department carrying out the foreign exchange transactions, is determined on the basis of the number of
transactions carried out by that department during the 746 period in question in the ratio of the fraction in
which the numerator is the number of transactions with counterparties outside the European Union (the
EU) and the denominator is the total number of transactions.
3.
In its VAT return for the period 1 May 1994 to 31 July 1994, which included its annual adjustment for
the period April 1993 to April 1994, the bank took into account the foreign exchange transactions entered
into by it in the period April 1993 to July 1994. It calculated that the input tax credit to which it was entitled
over that extended period of 15 months attributable to foreign exchange transactions with counterparties
belonging in countries outside the EU amounted to 251,45490.
4.
By decision of the commissioners of 26 September 1994, the input tax credit claimed was readjusted
by disallowing that part reflecting foreign exchange transactions conducted with counterparties from
outside the Community. According to that decision, the tax authority took the view that the whole amount
of input tax had diminished.
5.
The bank appealed to the London Value Added Tax Tribunal, which considered the case on the agreed
limited issue of whether or not the relevant foreign exchange transactions constituted supplies of goods or
services for VAT purposes. The tribunal allowed the appeal ([1995] V&DR 306), whereupon the
commissioners appealed on a point of law to the High Court. That court takes the view that it is material to
this case to determine whether the foreign exchange transactions constitute a supply of goods or services
effected for consideration within the meaning of the Sixth Directive.

IIQuestions referred by the High Court

6.
Accordingly, the High Court has referred the following questions to the Court of Justice for a
preliminary ruling pursuant to art 177 of the EC Treaty:
On the proper interpretation of the Sixth Directive and in relation to transactions of foreign
exchange as defined by the British Bankers Association [see below]: (1) Do such foreign exchange
transactions constitute the supply of goods or services effected for consideration? (2) If there has
been a supply of goods or services effected for consideration, what is the nature of the
consideration in relation to such transaction?

IIIThe banks foreign exchange transactions

7.
The expression foreign exchange transaction has been defined by the British Bankers Association as

any transaction between the parties for the purchase by one party of an agreed amount in one
currency against the sale by it to the other of an agreed amount in another currency, both such
amounts being deliverable on the same value date, and in respect of which transaction the parties
have agreed (whether orally, electronically or in writing): the currencies involved, the amounts of
such currencies to be purchased and sold, which party will purchase which currency and the value
date. (See International Foreign Exchange Master Agreement (1993).)
8.
The London branch of the bank, a national banking association organised with limited liability under
federal laws of the United States of America, carries on a wide range of banking activities including
foreign exchange dealing. At the time of the order for reference, it employed approximately 440 staff, of
whom about 40 747were employed in its foreign exchange trading department, with further staff in the
back office providing support.
9.
The bank is a market maker. It is willing at all times to provide and receive those currencies in which
it specialises. It provides and receives currencies in transactions which are commonly described as those
of purchase and sale. In common with other market makers, the bank will quote prices at which it is willing
to trade as bid or offer prices. The banks bid rate is the exchange rate at which the bank is willing to
buy a currency. The bank at any one specific time will bid, that is to say offer to buy, at one price
expressed as a rate of exchange and at the same time will offer, that is to say offer to sell, the currency in
the same denomination and the same amount at a slightly higher price. The difference between the two
rates is known as the spread.
10.
The banks customers for its foreign exchange transactions fall into three categories. The first includes
corporate customers seeking to manage their foreign currency risks and needs through spot and forward
contracts and hedging. The second category covers fund managers, such as pension funds. Customers
in this group are typically organisations which manage other peoples money. The third category includes
other financial institutions.
11.
All three categories of customer enter into essentially the same types of foreign exchange transaction
with confirmatory documentation including similar information. These foreign exchange transactions
include spot and forward transactions. 65% of customer transactions entered into by the bank are spot
transactions, the remaining 35% being forward transactions.
12.
 A spot transaction is the purchase of one currency against the sale of another currency, with the
delivery and sale normally being completed on the second following business day, which is known as the
settlement date or value date. Following agreement for a spot transaction, the bank will supply the other
party to the transaction with documentary confirmation of the terms of the transaction and how it is to be
effected. The confirmation will include statements of: the name and address of the customer; the
production date of the confirmation, which will be typically the date on which the deal was agreed; the
deal date, being the date on which the transaction was agreed; the currency and amount agreed to be
purchased by the bank from the customer; the value date for settlement of the transaction; the rate of
exchange applicable to the transaction; the foreign currency and amount agreed to be sold by the bank to
the customer; the bank account to which the customer will transfer the currency to be delivered by it to the
bank; and the bank account to which the bank will transfer the currency to be delivered by it to the
customer. The confirmation will show the one agreed rate of exchange for the particular transaction. It will
not show the two rates of bid and offer. However, the bid and offer rates will generally be known to the
customer, since he will commonly ask the bank to quote them.
13.
A customer for a spot transaction may, for example, be a manufacturer in the United States who has
shipped a product manufactured there to a customer in Germany and receives Deutschmarks abroad as
payment. As a rule, the customer will wish to exchange the Deutschmarks for US dollars. He will then
telephone the bank and ask for a price to sell Deutschmarks for US dollars for spot.
14.
In contrast a forward transaction is the purchase of one currency against the sale of another currency,
with delivery and sale being completed on a future value date. The amounts are fixed by reference to the
rate of exchange agreed on the deal date. Following agreement for a forward transaction, the bank will
748 supply the other party with similar documentary confirmation including the same information as that
included in a confirmation for a spot transaction. The essential difference from a spot transaction is that
the value date confirmed will be a future date more than two business days after the deal date.
15.
In the foreign exchange transactions entered into by the bank, no money is delivered physically in the
form of coin, bank notes or other chattels. What is delivered is the availability of drawing on a credit
opened with the bank in the currency delivered.
16.
These spot and forward transactions may be effected in a number of ways. On the one hand, a
computerised system is used, whereby the prices for the currency amounts to be exchanged are agreed
by the dealers for both parties by telephone and subsequently confirmed in writing. The confirmation
takes the form of a computer-printed note. The details required are keyed into the computer at the time
the transaction is agreed. Confirmation is given by pressing a single designated button on the keyboard.
In the case of transactions with corporate customers worldwide, a system is employed whereby the
dealers receive or provide their details and confirmations by telex. The bank also provides currency to
private customers. In this case the currency is obtained from a bank using the telex system. Confirmation
is then given by post.
17.
No transaction fee or commission is charged for or invoiced by the bank for any foreign exchange
transaction. Like any other market maker, the bank looks to make a profit out of its foreign exchange
dealings at least in part as a result of the spread between its bid and offer quotes. Generally speaking, the
greater the number of foreign exchange transactions the bank can make for the purchase and sale of
currencies at its bid and offer prices, the greater will be the possibility of profit on its foreign exchange
transactions. Each of its traders will have his or her own book of particular currencies and will be
expected to make a profit over appropriate periods. This profit is the result of all his or her dealings over
the period. Each transaction is entered into in the belief that it has value to the bank, but it is not the
banks practice to value each transaction individually.
18.
In any foreign exchange transaction, and in particular a forward contract, the bank will run at least two
risks. The first risk is that of default by the other party. A more significant risk is the risk that the market
rates will move against any position taken by the bank. The bid and offer rates quoted by the bank are
liable to change rapidly in the course of a business day. Thus, for example, where the bank has
contracted to pay Deutschmarks in a forward exchange contract against settlement by the bank in dollars
and the Deutschmark depreciates against the dollar, the bank runs the risk of a loss expressed in dollars.
The bank will therefore seek to limit its potential risk by seeking counterparties at relevant rates, value
dates and amounts. A significant proportion of these transactions will be initiated by other financial
institutions seeking the same protection for themselves.
19.
In order to maintain and increase the banks goodwill in the foreign exchange market, it publishes
circulars and information sheets which it circulates to its approved customers free of charge. For similar
reasons, it offers free advice to its corporate and fund manager customers.

IVRelevant provisions of the Sixth Directive

20.
Article 13B of the Sixth Directive provides as follows as regards foreign exchange transactions:
749
Other exemptions
 Without prejudice to other Community provisions, Member States shall exempt the following
under conditions which they shall lay down for the purpose of ensuring the correct and
straightforward application of the exemptions and of preventing any possible evasion, avoidance or
abuse (d) the following transactions (4) transactions, including negotiation, concerning
currency, bank notes and coins used as legal tender, with the exception of collectors items
21.
However, it is provided under art 13C:
Options
Member States may allow taxpayers a right of option for taxation in cases of (b) the
transactions covered in B(d) Member States may restrict the scope of this right of option and
shall fix the details of its use.
22.
Accordingly, by virtue of art 17(1) of the Sixth Directive, no deduction of input tax is possible with
regard to foreign exchange transactions, since that provision provides that the right to deduct shall arise
when the deductible tax becomes chargeable. One of the exceptions provided for in art 17(3) allows
input tax to be deducted in so far as the goods and services are used for the purposes of (c) any of the
transactions exempted under Article 13B(a) and (d), paragraphs 1 to 5, when the customer is established
outside the Community.
23.
According to art 6(1), supply of services means any transaction which does not constitute a supply of
goods within the meaning of Article 5. Under art 5(1) of the Sixth Directive, supply of goods means the
transfer of the right to dispose of tangible property as owner.

BOPINION

IQuestion 1
24.
The wording of this question refers to art 2 of the Sixth Directive, which prescribes what activities are
subject to VAT. According to art 2(1), the activities in question are supplies of goods or services effected
for consideration within the territory of the country by a taxable person acting as such. The national
courts first question seeks to ascertain whether the aforementioned foreign exchange transactions of the
bank fall within the scope of the directive.
1. Tax exemptions under the Sixth Directive
25.
As has already been mentioned, such foreign exchange transactions normally have no bearing on
taxation, since, under art 13B(d)(4) of the Sixth Directive, they are exempt from VAT. However, under art
17(3)(c), input tax may be deducted in connection with such foreign exchange transactions where the
customer is resident outside the Community. The bank is seeking the benefit of such deduction of input
tax in this case. In order for it to be able to do so, it is first necessary, of course, that the banks foreign
exchange transactions should come within the scope of VAT and hence of the Sixth Directive.
26.
In my view, this is clear already from the fact that, under art 13B(d)(4), transactions concerning
currency are expressly exempted from VAT. Such exemption would be necessary and make sense only if
it were indeed possible to tax such transactions, that is to say if they fell within the scope of VAT. This is
further supported by the fact that, under art 13C(b), the member states may allow taxable persons a right
of option in respect of the taxation of such transactions.
750
The upshot is that, in certain circumstances, such foreign exchange transactions are in fact subject to
VAT. Lastly it would be completely incomprehensible if art 17(3)(c) were to grant the right to deduct input
tax in respect of transactions which did not fall within the scope of VAT. Accordingly, the bank and the
European Commission point out that the provisions in question would be completely superfluous and
meaningless if the foreign exchange transactionsas the United Kingdom arguesfell completely
outwith the scope of the Sixth Directive.
2. Concept of consideration
27.
In the United Kingdoms view, the relevant provisions of arts 13 and 17 of the directive do not apply to
the transactions at issue on the ground that no consideration is paid for the banks service. As has already
been described, the bank does not charge a fee for exchanging foreign currency, but obtains a profit at
least partly as a result of its fixing different bid and offer prices. The difference between the two prices,
that is to say the spread, does not, the United Kingdom maintains, constitute consideration within the
meaning of the Sixth Directive. For their part, the bank, the French government and the Commission take
a different view.
2.1 Comparison between commission and spread
28.
In the United Kingdoms view, the bank would be working for consideration only if it charged a
commission for exchanging foreign currency. In other words, where a bank made a charge of, say, 2% for
exchanging money, it would, in the United Kingdoms opinion, without doubt be supplying a service for
consideration within the meaning of the directive. At the hearing, the United Kingdom explained this using
the example of a bureau de change. There is no difference in principle between a bureau de change and
the bank, except that in the latter case the foreign exchange transactions are bigger and more complex.
29.
The United Kingdom contends that if the bureau de change or the bank makes no charge but seeks to
make a profit by buying and selling foreign exchange at different rates, it does not receive any
consideration within the meaning of the directive. It does not follow from the fact that over a given period
the bank makes a profit from various foreign exchange transactions that it supplies a service for
consideration in the case of each individual foreign exchange transaction.
30.
The United Kingdom further argued at the hearing, again using the example of a bureau de change,
that, even if the bureau de change were to make a charge for exchanging the money, it could effectuate
that exchange only if it were to offer to purchase foreign currencies at particular rates and to sell them at
other rates, in order thus to obtain the corresponding currencies. The selling price is invariably higher than
the purchase price and, as a result, a profit is made over a certain period of time. The bureau de change
is trading. It carries out foreign exchange transactions in the normal course of its economic activities. This
correspondson a small scaleto the activities of the bank.
31.
The United Kingdom goes on to argue that if that bureau de change were now to decide no longer to
charge commission, it would receive no consideration for exchanging currencies and hence would not be
supplying a service within the meaning of the directive.
32.
In such a case, the United Kingdom argues, the bureau de changeand also the bank in this case
would be working free of charge. In the Commissions contention, this is very improbable. The United
Kingdom itself points out that the 751 bank and the bureau de change are seeking to make a profit even
from these general foreign exchange transactions.
33.
If we now consider the two cases described by the United Kingdomthe general foreign exchange
transactions of the bank or of the bureau de change, on the one hand, and the additional charge of
commission as consideration for the exchange of foreign currencies, on the otherit proves that those
cases do not differ as fundamentally as the United Kingdom maintains they do. In its example of a bureau
de change which charges 2% commission for exchanging foreign currencies, the United Kingdom has
already affirmed that changing amounts of money into another currency constitutes a service within the
meaning of the Sixth Directive and hence not a supply of goods.
34.
Nothing changes where no commission is charged for this operation. It remains the case that the
customer goes to the bank or the bureau de change and asks for means of payment to be made available
to him or her in a particular currency in return for means of payment in another currency. Even in the
event that the bank charges no commission, it operates for the customer and supplies him with the
means of payment by enabling him or her to have recourse to a credit opened at a bank in the currency
supplied. Consequently the bank again supplies a service. Even ifas the United Kingdom argues
there is no consideration for that service, it nevertheless remains a service. It may possibly no longer fall
within the scope of the Sixth Directive. In any event, the bank tries to sell the means of payment at a
somewhat higher rate or price than the one at which it purchases it.
35.
As the United Kingdom itself concedes, the bank pays less than it hopes to obtain when it sells on the
currency. However, this means simply that it supplies commensurately less money in the foreign
currency and hence makes a profit. So, also in this case, the customer pays for the banks service. In the
course of this currency transaction, which, as has just been shown, still constitutes a service for the
customer, the bank is endeavouring to make a profit; this means that it is seeking to recover the costs of
the service and more besides. The costs of large-scale foreign exchange transactions as carried out by
the bank are very much higher than the costs incurred by a small bureau de change. As has already been
mentioned, the use of computers and technology on an extensive scale is necessary.
36.
In any event, the fact remains that the bank has to set its rates in such a way that it receives payment
for its service. This means that also where the bank seeks to make a profit only through the purchase and
sale of foreign currencies, it does not work free of charge contrary to the United Kingdoms contention but
makes the customer pay for its service and that payment takes the form of a smaller amount of
consideration being paid for currencies purchased and a higher amount of consideration for currencies
sold.
2.2 Case law of the Court of Justice
37.
The fact that income is actually received for an operation does not mean in every case that that
operation is effected for consideration within the meaning of the Sixth Directive (see my opinion in Tolsma
v Inspecteur der Omzetbelasting Leeuwarden Case C-16/93 [1994] STC 509, [1994] ECR I-743 (para
13)). Whether the requirement for consideration is satisfied can be ascertained from the case law of the
court, which has had to pronounce on this question on several occasions. Thus, in the judgment in
Tolsma the court held, referring to its judgments in Staatssecretaris van Financin v Coperatieve
Aardappelenbewaarplaats GA Case 154/ 75280 [1981] ECR 445 (para 12) and Naturally Yours
Cosmetics Ltd v Customs and Excise Comrs Case 230/87 [1988] STC 879, [1988] ECR 6365 (para 11)
that a provision of services is taxable only if there is a direct link between the service provided and the
consideration received (see Tolsma [1994] STC 509, [1994] ECR I-743 (para 13); see also Apple and
Pear Development Council v Customs and Excise Comrs Case 102/86 [1988] 2 All ER 922, [1988] STC
221, [1988] ECR 1443 (paras 1112)).
38.
The court concluded from this that
a supply of services is effected for consideration within the meaning of art 2(1) of the Sixth
Directive, and hence is taxable, only if there is a legal relationship between the provider of the
service and the recipient pursuant to which there is reciprocal performance, the remuneration
received by the provider of the service constituting the value actually given in return for the service
supplied to the recipient. (See Tolsma [1994] STC 509, [1994] ECR I-743 (para 14).)
39.
The Commission and the bank rightly argue that all those criteria are fulfilled in this case.
40.
There is a legal relationship between the provider of the service and the recipient pursuant to which
there is reciprocal performance. It appears from the information provided by the national court that, in the
course of negotiations on a given exchange, the customer and the bank agree that the customer should
place a specific amount of money in a particular currency in a precisely specified account, whilst the bank
commits itself for its part to deposit a certain sum of money in another currency in an account specified by
the customer. Consequently, the customer and the bank commit themselves to reciprocal performance.
41.
The question here is whether the remuneration received from the supplier of the service, here the
bank, constitutes the actual value given in return for the service supplied to the recipient.
42.
In Tolsma it was held that this was not the case. The question there was whether the receipts of a
musician who played a musical instrument on the public highway could be regarded as consideration for
the service provided by him of playing music. The court considered that there was no agreement in that
case between the parties, since the passers-by voluntarily made a donation, whose amount they
determined as they wished. In addition, the court found that there was no necessary link between the
musical service and the payments to which it gave rise, since the passers-by did not request music to be
played for them. Moreover, they paid sums which depended not on the musical service but on subjective
motives which might bring feelings of sympathy into play (see [1994] STC 509, [1994] ECR I-743 (para
17)).
43.
In the instant case the situation is different. It is the customer who approaches the bank and asks for a
service, namely the exchange of a foreign currency. According to the bank, the customer is aware that
that service will not be performed free of charge. This, moreover, is contested only by the United
Kingdom, which considers that the spread between the bid and offer price does not constitute
consideration for the service. On the other hand, the United Kingdom also states that customers generally
inquire at the bank about the two rates, that is to say also about the spread. Customers therefore know by
how much the selling price of foreign currencies exceeds the purchase price. Consequently, customers
know that they are paying for the service and are aware of how much they are paying.
753
44.
It is also absolutely clear to the bank itself, which constitutes the other party to the reciprocal
relationship, that its payment for the service of exchanging currencies results from its spread. This means
that there is no doubt as between the supplier and the recipient of the service that the service is effected
for consideration and that the consideration relates to the transaction in question.
45.
It remains to be noted therefore that, in the case of the rates at which the bank is prepared to
purchase currencies from customers and to sell currencies to customers, the spread resulting from the
difference in rates constitutes the payment for the service supplied by the bank. The bank fixes the rates
with that aim in mind. The United Kingdom itself confirmed at the hearing that where a charge is
concealed in the difference between the bid and the offer rate, but is capable of being identified, a
service is effected for consideration. In this case the charge is concealed in the spread in so far as it
constitutes the payment for the service and, to that extent, constitutes a charge. Accordingly, the charge is
also capable of being identified.
46.
For this reason, it is possible, as the Commission suggests, to divide the amount which the customer
pays in a given currency to the bank into the amount corresponding to the other currency supplied by the
bank and the consideration for the service, that is to say the spread.
47.
In the United Kingdoms view, however, it is impossible to determine a countervalue for the amount
paid by the bank on the ground that there is no corresponding market price which can be used to make
such a determination. There is just the bid and offer prices set by the bank.
48.
 The Commission disagrees. It takes the view that there is a market price whose value is between the
bid and the offer price.
49.
To my mind, it is certainly conceivable that there are other possibilities for determining the value of a
given amount of money in terms of a countervalue in another currency apart from the bid and offer prices
fixed by the bank for its customers. I would point out in this connection simply that the individual
currencies are also traded on the stock exchange and that prices are fixed there. To what extent it is
actually feasible to determine a countervalue in another currency is a matter for the national court.
50.
Even if it should not be feasible to determine an exact countervalue on the basis of a market price, this
would not alter the fact that the banks service is paid for through the spread. As has been made clear
above, the bank calculates its prices in such a manner that its service is paid for in the case of each
transaction. This is because the spread is calculated for each exchange, which means that on each
transaction the customer receives less back from the bank than he pays. The bank stated at the hearing
that the customer paid a higher price for a foreign currency than he would obtain were he to sell that
currency back to the bank immediately.
51.
Consequently, the customer pays for the banks service in respect of each individual exchange via the
spread fixed between the bid and the offer price, with the result that whenever the customer makes an
exchange transaction he receives less back from the bank than he gave the bank. It is irrelevant in this
connection whether he may make ultimately a profit on a transaction as a result of currency fluctuations
occurring in the meantime. I shall explain this further (see para 73, below).
754
2.3 Necessity for a second transaction in order to obtain consideration
52.
The United Kingdom goes on, however, to name other reasons why, in its view, the banks receipts
derived from the spread cannot be regarded as consideration for the individual exchange. It argues first
that the profit from the different purchase and selling prices is invariably not realised until the next
transaction, that is to say when the bank sells on the money purchased from the customer to another
customer.
53.
As has already been made clear, however, in the case of every transaction the bank purchases a
particular currency. In so doing it supplies less to the customer than it receives from him. The bank and
the Commission also agree that for each transaction the customer does not receive the full countervalue
of the sum exchanged by him on account of the spread. The Commission cites in this connection the
decision of the value added tax tribunal ([1995] V&DR 306 at 325), which also assumed that the rate at
which the bank sells the currency to the customer includes the costs of the exchange, that is to say the
service.
2.4 Direct link between the service provided and the consideration received (individual valuation)
54.
The United Kingdom submits, as an additional argument in support of its view that the spread cannot
constitute the consideration for the service consisting of the exchange, that consideration within the
meaning of the Sixth Directive must be capable of being determined for each individual transaction. It
refers in this connection to the case law of the court, which has held that in order for a service to be
taxable, there must be a direct link between the service provided and the consideration received (see the
judgments in Naturally Yours Cosmetics Ltd [1988] STC 879, [1988] ECR 6365 (para 11) and
Coperatieve Aardappelenbewaarplaats [1981] ECR 445 (para 12)).
55.
It is clear from the parties submissions and from the order for reference that each of the banks
transactions is conducted in the belief that it will yield an advantage for the bank. In addition, however, it
emerges that it is not the banks practice to value each transaction individually. In other words, the bank
calculates its profit over a particular period. The United Kingdom argues that this is too imprecise to
ground the assumption that the banks service is effected for consideration in the case of each individual
transaction. It refers in this connection, inter alia, to the opinion in H J Glawe Spiel- und
Unterhaltungsgerte Aufstellungsgesellschaft mbH & Co KG v Finanzamt Hamburg-Barmbek-Uhlenhorst
Case C-38/93 [1994] STC 543 at 544, [1994] ECR I-1679 at 1681. Advocate General Jacobs stated in
that case that gaming transactions are ill-suited to VAT (see [1994] STC 543, [1994] ECR I-1679 (para
16)). Elsewhere in the opinion the Advocate General expressed the view that there may be some
theoretical difficulty in viewing, for example, a bookmakers net winnings as the consideration for services
(see para 22). In the United Kingdoms view, those difficulties and the lack of suitability to VAT apply a
fortiori in the present case of foreign exchange transactions, since here it is not only a question of
consideration being difficult to determine, but of no consideration at all.
56.
As has already been shown, it cannot be considered here that there is no consideration for the service
performed by the bank in relation to foreign exchange transactions. Nevertheless, it cannot be denied that
it is not entirely an easy matter to determine the consideration. As the United Kingdom rightly submits, the
banks receipts are the outcome of its involvement in a series of transactions, all of which are concluded
at different rates and under different 755 market conditions. Even if foreign exchange transactions are ill-
suited to VAT, this is perhaps the reason why they are exempt from tax under the Sixth Directive.
However, even after the opinion in Glawe, those difficulties in determining the consideration do not lead to
the conclusion that there is no consideration within the meaning of the directive and that foreign exchange
transactions consequently do not come within the scope of VAT. It is worth emphasising here once again
that the bank fixes a spread for each exchange. The spread is the difference between the rate agreed for
the transaction and the offer price (or, if one exists, the market price). However, the bank does not value
each transaction individually and hence also does not value every spread. The United Kingdom takes the
view that this is too imprecise, since the customer is not charged the spread. It further considers that in
principle the banks profit cannot be regarded as consideration for the purposes of the Sixth Directive.
57.
As to this, I would first say that it cannot be concluded simply from the fact that the bank does not
value each individual transaction that it is impossible to effect such an individual valuation. Perhaps the
bank does not carry out an individual valuation becauseand this is beyond doubtit would be very
complicated and is unnecessary for the bank. It is so complicated because, in order to determine how
much profit the bank has ultimately made, it is not sufficient to determine when which sum of money was
exchanged at which rate, the market situation at the relevant time must also be taken into account and
in the case of forward transactionsregard has to be paid to the subsequent development of the market.
For this reason, an individual valuation would only be possible after the event, if at all. The court does not
have sufficient information in order to judge whether it would be possible for the bank to carry out such an
individual valuation. It should be left to the national court to assess this if necessary.
2.5 Need for individual valuation (judgments in Glawe and Fischer)
58.
It must be considered, however, in the light of the judgment of the Court of Justice in the Glawe case
whether such an individual valuation is necessary for the purposes of charging VAT.
59.
The Glawe case was concerned with charging in respect of gaming machines in bars and restaurants.
The operation of such machines was regulated by law. They incorporated a reserve compartment holding
a stock of coins from which winnings were paid out, and a cash box compartment. If, following the
payment of winnings, the reserve was no longer full, the coins inserted by the players did not fall into the
cash box but entered the reserve. The machines were required to be set in such a way that they paid out
as winnings at least 60% of coins inserted by players (the stakes), with the remainder, some 40%, being
retained in the cash box.
60.
The court followed the proposal of Advocate General Jacobs and regarded the stakes as being divided
into two parts: one served to replenish the reserve, and thus to pay out winnings, and the remainder
entered the cash box (see [1994] STC 543, [1994] ECR I-1679 (para 11)).
61.
In his opinion, the Advocate General defined that remainder somewhat more precisely. He considered
that that remaining portion was the price paid for the services provided by the operator. He further
expressed the view that, over a given period, the two components would correspond to the amounts
collected respectively by the cash box and the reserve of the machine (see [1994] STC 543, [1994] ECR
I-1679 (para 29)).
756
62.
The court held that the proportion of the stakes which was paid out as winnings could not be regarded
as forming part of the consideration for the provision of the machine to the players, nor as the price for
any other service provided to the players, since it was mandatorily fixed in advance (see [1994] STC 543,
[1994] ECR I-1679 (para 12)). Consequently, the taxable amount was the takings of the operator of the
gaming machine, that is to say the coins contained in the cash box. Consequently, in that case, too, each
individual game was not evaluated in accordance with whether the gaming machine or the player had
won, but the takings of the operator, calculated over a period of time, were regarded as the taxable
amount.
63.
These issues have also been raised in Fischer v Finanzamt Donaueschingen Case C-283/95 [1998]
All ER (EC) 567, [1998] STC 708, which concerned the taxation of a game amounting to the game of
roulette. Similarly, players purchased chips which they could place on a table resembling a roulette table.
Here too, it was possible for a player to win several times his stake, the winnings being paid out in chips
after every game. Players wishing to discontinue the game could exchange their remaining chips for cash.
64.
In this case too, Advocate General Jacobs considered that each chip placed on the table comprised,
as a matter of legal analysis, two components: (a) the wager and (b) the consideration for the organisers
service, ie the price paid by players for the right to participate in the game and obtain the chance of
winning. That price, consisting in the advantage which the house reserved to itself by virtue of the odds
being in its favour, could be calculated precisely and was a standard percentage varying according to the
version of roulette played. It was paid by each player each time he placed a chip on the table. It would be
perfectly possible for an organiser to separate the two components by eliminating the advantage for the
house and replacing it with a separate charge to cover his costs and provide him with a profit (see [1998]
All ER (EC) 567, [1998] STC 708 (para 47)).
65.
 The Advocate General ultimately reached the conclusion that in practice individual calculations based
on each chip placed on the table were unnecessary. The total of the amounts received by way of
consideration for individual transactions corresponded to the organisers net takings (after payment of
winnings) during a given period. Over a period the organisers net takings necessarily corresponded to
the advantage which he reserved to himself. The Advocate General further stated that the fact that there
was in practice an easier method of determining the taxable amount did not, however, mean that tax was
not levied on individual transactions (see [1998] All ER (EC) 567, [1998] STC 708 (para 49)).
66.
What of the present case? Here too it is possible to divide what the customer pays the bank into two
components. As we have already seen, one component is the countervalue of the amount of money
provided by the bank, whilst the second component is the consideration, that is, the price, for the service
of making the exchange. In the cases of Glawe and Fischer, that component was determined,
respectively, by the statutory percentage of winnings or by the odds determined by the house. In the
present case, the corresponding component is the spread. As in the case of Fischer that componentthe
spreadcould also have taken the form of a charge. It can therefore be considered, on the same lines as
in Glawe and Fischer, that, in the case of each individual transaction, part of what the customer pays may
also be regarded as the consideration for a service, and that that portion can be precisely determined.
757
67.
It must be considered, however, whether that price component is as precisely determined in this case
as it was in the cases of Glawe and Fischer. In the case of Glawe, for instance, it was clear from the
outset that the operator of the gaming machines would receive, in terms of his net takings, a given
percentage of the amounts inserted in the machines. It was in that case impossible to reconstruct after a
given period how much money had been inserted in the machines. Yet it was clear that the amount in the
cash box after a certain period of players winning and losing corresponded to a certain percentage of the
stakes. In other words, the percentage was fixed from the outset, but the exact amount could only be
determined after a certain period of time.
68.
In the instant case, the consideration is determined by the spread. At the time of the transaction the
spread is determinate, since it consists of the difference between the individual rates. However, its exact
amount may possibly not be realised until later, as, for example in the case of forward transactions. In my
view, it is irrelevant that the spread may vary from one transaction to another, provided that it may be
clearly determined for each transaction. This was also the view reached by the Advocate General in his
opinion in Glawe [1994] STC 543, [1994] ECR I-1679 (para 22) when he found that in the case of a
bookmaker, for example, the price which he receives for that service varies and depends partly on
chance and partly on his skill in setting the odds. This does not mean, however, that that service has to be
excluded from the scope of the directive.
69.
It should therefore be considered that the price component in the instant case is no less precisely
predetermined than it was in the cases of Glawe and Fischer. This means that it can be assumed, as in
those cases, that individual transactions are being taxed in this case. At the same time, I can see no
reason why it should not be possible to effect the calculation over a period of time, as was necessary in
the cases of Glawe and Fischer and as in fact is the banks practice in this case. It should therefore be
considered that in this case individual transactions are being taxed and that the banks calculation is
sufficient for the purposes of the taxation. Accordingly, it is clear that, in carrying out foreign exchange
transactions, the bank effects a service for consideration within the meaning of the Sixth Directive.
2.6 Comparison with typical cases of VAT
70.
Consequently, it is clear that in this casecontrary to the opinion of the United Kingdomthe banks
profit can be regarded as consideration for a service. As the Advocate General explained his opinion in
Fischer [1998] All ER (EC) 567, [1998] STC 708 (para 45), this approach produces results most closely
resembling typical cases of VAT. If, for example, a manufacturer sells a product for a given price plus VAT,
the amount remaining after deduction of VAT constitutes the amount covering his profit margin, his
material costs and all other overheads. The tax is precisely proportional to the price, since the ratio
between the price, ie the aggregate takings, and the VAT correspond to the statutory rate of VAT. In the
instant case, the banks profit, that is to say its takings, constitute the amount covering the profit-margin,
the costs of carrying out the transaction and the costs of operating the bank and the foreign exchange
department. It should be noted in this connection that it is not the banks pure profit which is to be taken
into account here, but everything which it receives by way of spread.
71.
I would further point out that, also in typical service transactions, what is taxed is what the supplier
obtains as consideration for his service. As the Commission states, it is also compatible with the case
hardly conceivable in 758 practicein which the bank makes a loss over a given period and then has to
pay no tax.
72.
In a case in which a bank made both a charge and purchased and sold foreign currencies at different
rates, the consideration for the banks service would consist, not only of the charge, but also of the
spread, as the Commission rightly suggests.
2.7 Exchange of means of payment distinguished
73.
Neither can the United Kingdoms submission that a foreign exchange transaction is nothing more than
the exchange of one means of payment for another alter the fact that the spread is to be regarded as the
banks consideration. Exchanging dollars for Deutschmarks, for example, is more than exchanging a
banknote for coins of the same currency. In exchanging different currencies, an exchange rate must be
fixed. Even if in the case of a fixed exchange rate the exchange of currencies ceases to differ from the
transaction of exchanging a banknote for coins, it must be borne in mind in this connection that in the
case of the banks foreign exchange transactions the exchange rate must first be agreed, taking into
account the situation on the foreign exchange market, and that the exchange rate is then confirmed
electronically. The United Kingdom itself refers to buying and selling in connection with foreign exchange
transactions, which points to the fact that more is being done than merely exchanging means of payment.
2.8 Consideration in the event of losses on the part of the bank
74.
The outcome is not altered either by the fact that the bank may make losses on individual transactions.
Also in the case of games of chance, the house may incur very heavy losses. Yet this does not alter the
fact that as mentioned above one component of each individual stake represents the payment for the
house. This can be illustrated for the purposes of the present case by the fact that even if the bank made
losses on transactions, the losses would be even higher if the bank had not calculated a spread but
instead had paid the full countervalue. If the bank has charged a spread, it does not take the full
countervalue as the basis for determining its losses.
2.9 Simple games of chance distinguished
75.
Lastly, I would consider a further argument put forward by the United Kingdom, which once again
refers to the opinion in Glawe. In that case, the Advocate General stated that gambling for money in its
simplest form does not give rise to consumption of goods or services, even though it does entail
expenditure by gamblers. This would be the case with a private bet, where both gamblers place their
stakes on the table. The placing of the bets, although it involves the outlay of money, does not constitute,
the Advocate General considered, the consumption of goods or services which is the taxable event under
the VAT system (see [1994] STC 543, [1994] ECR I-1679 (para 20)). Referring to that passage of the
opinion, the United Kingdom submits that also in this case there is only a movement of money from the
bank to the customer and from the customer to the bank. It cannot be concluded from this that there is a
consumption of services within the meaning of the system.
76.
The Advocate General further considered that commercial gambling is different in so far as the person
organising the gambling arranges matters in such a way that on average his winnings are sufficient to
meet his costs in organising 759 the gambling and to provide him with a reasonable profit. He gave as an
example a bookmaker who sets the odds for bets on horse-racing at a level intended to ensure that he
makes an overall profit on bets placed. To that extent the person organising the gambling may perhaps be
regarded as not only taking part in the gambling himself but as also providing a service to the other
gamblers consisting in organising the gambling (see [1994] STC 543, [1994] ECR I-1679 (para 21)). The
position is no different in this case. The bank will fix the rates for its foreign exchange transactions in such
a way as to ensure that it obtains a profit from the transactions which it concludes, taken as a whole. The
United Kingdom does not contest this. On this ground, it should be considered that the banks activities in
connection with foreign exchange transactions cannot be compared to gambling in a simple form, which
does not involve any consumption of services within the meaning of the system. Moreover, the bank takes
this view.
77.
It must therefore be considered that, in engaging in foreign exchange transactions, the bank effects
services for consideration within the meaning of the Sixth Directive. Consequently, those services come
within the scope of the directive and mayeven if they are exempt from taxentitle the bank to deduct
input tax in accordance with art 17(3)(c) in connection with transactions with counterparties resident
outside the Community. The consideration therefore may be precisely determined and imputed to
individual transactions, even if it is not calculated for each individual transaction.

IIQuestion 2
1. Need to answer the question
78.
Given that in the course of answering the first question referred for a preliminary ruling it was already
necessary precisely to determine the consideration, the second question has already been answered.
However, it was not unnecessary to answer the second questionas the bank suggestedsince, in my
view, it cannot be proved that there is consideration within the meaning of the Sixth Directive without
precisely defining that consideration.
2. Consideration of the counter-arguments
79.
Since my answer to the second question does not accord with the approach proposed by the bank, I
would now briefly consider the counter-arguments put forward by the bank. In the banks view the total
amount of foreign currency paid by the customer should be taxed as the consideration. Its grounds for
arguing this are that VAT is a tax on turnover and not a tax on profit. It refers in this connection to art
11A(1)(a) of the Sixth Directive, which provides that the taxable amount in respect of supplies of services
is everything which constitutes the value of the consideration which the provider of services receives for
the transaction from the recipient of the services. The bank infers from this that the taxable amount is
everything which the bank receives from the customer.
2.1 Wording of art 11A(1)(a)
80.
As the Commission and the United Kingdom rightly submit, art 11 does not support this argument. It
provides merely that the taxable amount is everything obtained by way of consideration. That cannot be
equated with everything which the provider of services receives. Consequently, the amount of the
consideration still has to be determined.
760
2.2 VAT as turnover tax
81.
As far as concerns the submission that VAT as a turnover tax may not be charged on a taxable
amount consisting of the banks profit, the bank itself refers to the judgment in Glawe and concludes that
the approach taken there, that is to say that the takings of the operator of the gambling machines
constitute the taxable amount, may be regarded as correct. As I have already mentioned, it appears from
the opinion in Fischer [1998] All ER (EC) 567, [1998] STC 708 (para 45) that the approach taken in Glawe
is the one which most closely approximates to the normal case of VAT.
2.3 Practical effects of the banks approach
82.
This becomes clear if the approach advocated by the bank is taken a stage further. If the bank had to
tax everything which it received from the customerthat is to say the total amount of foreign currencyit
would have to pay disproportionately high tax in relation to its earningswhich are constituted only by the
spread. As I have already shown above, in the case of VAT the price of a service, that is to say that which
the provider receives, is taxed. The banks approach would therefore distort the VAT system and, in this
case, would result in the bank being able to claim a disproportionately high amount of input tax 4.
4
Whether that would be the actual consequence in this case is questionable sinceas the bank statesunder the
special method of partial exemption agreed with the commissioners, not turnover, but only the number of foreign
transactions are used in order to calculate VAT and the input tax deducted.

2.4 Examples proving the opposite

83.
The United Kingdom rightly points out that if the service consisting of the exchange were to be paid for
by a charge, it is clear that only that charge would be taxed and not the charge together with the amount
of foreign currency exchanged by the customer. The Commission takes a similar view, constructing an
example in which in the context of one currencyie notes to coina charge was also levied. In this case
too, there would be no reason for taxing the amount of money to be exchanged in addition to the charge.
At the hearing, the Commission gave another example relating to the taxation of a service in general. The
example was cleaning a coat for a given price. In this case, too, it is clear that only the price of the
cleaning and not the value of the coat plus the price of cleaning is taxed.
2.5 Proposal for a Nineteenth VAT Directive
84.
Lastly, the Commission refers to its proposal for a Nineteenth Directive (see the Proposal for a
Nineteenth Council Directive on the harmonisation of the laws of the member states relating to turnover
taxes, amending the Sixth Directive (OJ 1984 C347 p 5) containing provisions on foreign exchange
transactions, in which an approach was chosen whereby only the charge or the costs demanded by the
purchaser as payment for the service were to be regarded as consideration, which came to nought, not
because of this approach but for other reasons.
85.
Consequently, it only remains to confirm that the consideration for the exchange of foreign currencies
is the spread.
761
2.6 Foreign exchange transactions as exchange transactions?
86.
I do not intend to consider further the submission made by the bank that foreign exchange
transactions should be regarded as exchange transactions. Admittedly, a currency is exchanged against
another, yet not by the bank handing over to the customer money physically in the form of coins or
banknotes and at the same time receiving coins or banknotes from the customer. Instead, the bank
enables the customer to use a credit opened at a bank in the currency sought by the customer. Here, the
banks interest lies principally in the amount of the margin and less in the type of currency supplied by the
customer. The bank itself points out that the exchange is paid for by the spread. It can be seen from this
that the transaction is not an exchange but a provision of services, namely the changing of foreign
currencies.

CCONCLUSION
87.
 I therefore propose that the questions referred for a preliminary ruling be answered as follows:
(1) In relation to foreign exchange transactions as defined by the British Bankers Association
(see para 7, above), the bank effects a service for consideration within the meaning of Council
Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to turnover
taxescommon system of value added tax: uniform basis of assessment, where that service is not
paid for by a charge but by the spread between the bid and offer rates.
(2) The consideration for the service is what the bank receives by way of spread between the
bid and offer rates.

14 July 1998.

The COURT OF JUSTICE (Fifth Chamber)

delivered the following judgment.
1.
By order of 13 May 1996, received at the Court of Justice of the European Communities on 20 May
1996, the Queens Bench Division of the High Court referred for a preliminary ruling under art 177 of the
EC Treaty two questions on the interpretation of EC Council Directive (EEC) 77/388 on the harmonisation
of the laws of the member states relating to turnover taxescommon system of value added tax: uniform
basis of assessment (OJ 1977 L145 p 1) (the Sixth Directive).
2.
Those questions have been raised in proceedings between First National Bank of Chicago (the bank)
and the Commissioners of Customs and Excise concerning deduction of input tax on certain foreign
exchange transactions.
3.
Article 2 of the Sixth Directive provides as follows:
The following shall be subject to value added tax: (1) the supply of goods or services effected
for consideration within the territory of the country by a taxable person acting as such; (2) the
importation of goods.
4.
Article 5(1) defines the supply of goods in these terms: Supply of goods shall mean the transfer of
the right to dispose of tangible property as owner.
5.
The supply of services is defined in art 6(1) as follows: Supply of services shall mean any
transaction which does not constitute a supply of goods within the meaning of Article 5.
6.
Article 11A(1)(a) is worded as follows:
762
The taxable amount shall be: (a) in respect of supplies of goods and services other than those
referred to in (b), (c) and (d) below, everything which constitutes the consideration which has been
or is to be obtained by the supplier from the purchaser, the customer or a third party for such
supplies including subsidies directly linked to the price of such supplies.
7.
Article 13B provides as follows:
Without prejudice to other Community provisions, Member States shall exempt the following
under conditions which they shall lay down for the purpose of ensuring the correct and
straightforward application of the exemptions and of preventing any possible evasion, avoidance or
abuse (d) the following transactions (4) transactions, including negotiation, concerning
currency, bank notes and coins used as legal tender, with the exception of collectors items;
collectors items shall be taken to mean gold, silver or other metal coins or bank notes which are
not normally used as legal tender or coins of numismatic interest.
8.
Article 13C(b), however, makes it possible for member states to allow their taxpayers a right of option
for taxation in respect of the transactions covered by, inter alia, art 13B(d).
9.
Article 17(3) of the Sixth Directive provides as follows:
Member States shall also grant to every taxable person the right to a deduction or refund of the
value added tax referred to in paragraph 2 in so far as the goods and services are used for the
purposes of (c) any of the transactions exempted under Article 13B(a) and (d), paragraphs 1 to
5, when the customer is established outside the Community or when these transactions are directly
linked with goods intended to be exported to a country outside the Community.
10.
According to the order for reference, the bank is registered for value added tax (VAT) in the United
Kingdom and carries on a wide range of banking activities, including foreign exchange dealing. It is a
market maker, being at all times willing to provide and receive those currencies in which it specialises.
11.
The bank quotes prices at which it is willing to trade in currencies as bid or offer prices. At any one
specific time it will bid, that is to say offer to buy a currency, at one price expressed as a rate of exchange
and at the same time will offer to sell the currency in the same denomination and the same amount at a
slightly higher price expressed as a rate of exchange, the difference between the two prices being known
as the spread.
12.
Foreign exchange transactions consist of either spot or forward transactions. A spot transaction is
the purchase of one currency against the sale of another currency, with the delivery and sale normally
being completed on the second subsequent business day, which is known as the settlement date or value
date. A forward transaction differs from a spot transaction in that the delivery and sale of currencies are
completed only on a future value date, the amounts, however, being fixed by reference to the rates of
exchange agreed on the deal date.
13.
The national court points out that no money is delivered physically in the form of coin, note or other
chattel in the foreign exchange transactions entered into by the bank. What is delivered is the availability
of drawing on an account opened with a bank in the currency delivered.
763
14.
No transaction fee or commission is charged for or invoiced by the bank for the transactions at issue in
the main proceedings. The bank seeks to make a profit out of its foreign exchange dealings as a result of
the spread between its bid and offer quotes. Each of its traders will have his or her book of particular
currencies and will be expected to make a profit over appropriate periods. This profit is a result of all of
their dealings over a particular period.
15.
The bank is partly exempt for VAT purposes. It does, however, have the right to deduct input tax
corresponding to transactions completed with counterparties established outside the Community. In order
to determine the deductible amount, the bank has agreed with the commissioners a special partial
exemption method under reg 31 of the Value Added Tax (General) Regulations 1985, SI 1985/886 (see
now s 102 of the Value Added Tax Regulations 1995, SI 1995/2518). The recoverable proportion of input
tax which the agreed method allocates to the bank is determined by reference to the number of foreign
exchange transactions carried out as represented by the fraction in which the numerator is the number of
transactions with counterparties outside the European Union and the denominator is the total number of
transactions.
16.
In its return for the period from 1 May 1994 to 31 July 1994, which included its annual adjustment for
the period from April 1993 to April 1994, the bank took into account, in determining the numerator and
denominator of the relevant fraction, the foreign exchange transactions into which it had entered over the
period from April 1993 to July 1994. It calculated that the input tax credit to which it was entitled over that
extended 15-month period attributable to foreign exchange transactions with counterparties established in
countries outside the Community amounted to 251,45490.
17.
By decision of 26 September 1994, the commissioners reduced the input tax credit which the bank
was claiming by disallowing the portion corresponding to the foreign exchange transactions concluded
with those counterparties.
18.
The bank appealed to the London Value Added Tax Tribunal. The appeal was heard on the agreed
limited issue of whether or not the relevant foreign exchange transactions were supplies of services or
goods for VAT purposes. By a decision of 12 September 1995, the tribunal (see [1995] V&DR 306)
allowed that appeal.
19.
The commissioners appealed to the High Court against that decision.
20.
Taking the view that resolution of the case depended on an interpretation of the Sixth Directive, the
High Court decided to stay proceedings and refer the following questions to the Court of Justice:
On the proper interpretation of the Sixth Directive and in relation to transactions of foreign
exchange as defined by the British Bankers Association [see below]: (1) Do such foreign exchange
transactions constitute the supply of goods or services effected for consideration? (2) If there has
been a supply of goods or services effected for consideration, what is the nature of the
consideration in relation to such transaction?
21.
The definition of foreign exchange transaction referred to in the question reads as follows
any transaction between the parties for the purchase by one party of an agreed amount in one
currency against the sale by it to the other of an agreed amount in another currency, both such
amounts being deliverable on the same value date, and in respect of which transaction the parties
have agreed 764(whether orally, electronically or in writing): the currencies involved, the amounts
of such currencies to be purchased and sold, which party will purchase which currency and the
 value date. (See the British Bankers Association International Foreign Exchange Master
Agreement (1993).)

The first question

22.
By its first question, the High Court is asking essentially whether transactions between parties for the
purchase by one party of an agreed amount in one currency against the sale by it to the other party of an
agreed amount in another currency, both such amounts being deliverable on the same value date, and in
respect of which transactions the parties have agreed (whether orally, electronically or in writing) the
currencies involved, the amounts of such currencies to be purchased and sold, which party will purchase
which currency and the value date, constitute supplies of goods or services effected for consideration
within the meaning of art 2(1) of the Sixth Directive.
23.
The bank, the French government and the European Commission take the view that foreign exchange
transactions constitute supplies of services. Since they are effected for consideration, they come within
the Sixth Directive.
24.
The UK government, on the other hand, considers that, in the absence of consideration, a foreign
exchange transaction entered into without the charging of a commission or a fee does not constitute a
supply of goods or services for consideration within the meaning of the Sixth Directive but is simply the
exchange of one means of payment for another.
25.
On this question, the court observes first of all that the currencies which are exchanged against other
currencies in a foreign exchange transaction cannot be regarded as tangible property within the meaning
of art 5 of the Sixth Directive, since money used as legal tender is involved. Foreign exchange
transactions are thus supplies of services within the meaning of art 6 of the Sixth Directive.
26.
With regard, second, to the question whether services are supplied for consideration, the court has
already held that a supply of services is effected for consideration within the meaning of art 2(1) of the
Sixth Directive, and is therefore taxable, only if there is a legal relationship between the provider of the
service and the recipient pursuant to which there is reciprocal performance, the remuneration received by
the provider of the service constituting the value actually given in return for the service supplied to the
recipient (see the judgment in Tolsma v Inspecteur der Omzetbelasting Leeuwarden Case C-16/93 [1994]
STC 509, [1994] ECR I-743 (para 14)).
27.
Only where a persons activity consists exclusively in providing services for no direct consideration is
there no basis of assessment and the services are therefore not subject to VAT (see [1994] STC 509,
[1994] ECR I-743 (para 12)).
28.
In the present case, it cannot be disputed that a bilateral legal relationship exists between the bank
and its counterparty under which the two parties to the transaction give reciprocal undertakings to transfer
amounts in a given currency and to receive the countervalue in another currency.
29.
Apart from the actual exchange transaction, the service provided by the bank is characterised by the
banks preparedness to conclude such transactions in the currencies in which it specialises.
30.
From the mere fact that no fees or commission are charged by the bank upon a specific foreign
exchange transaction it does not follow that no consideration is given.
765
31.
Moreover, any technical difficulties which exist in determining the amount of consideration cannot by
themselves justify the conclusion that no consideration exists.
32.
In addition, it is apparent from the case file that the rates at which the bank is prepared to sell or
purchase currencies are different and are separated by a spread. The conclusion must therefore be that,
in return for the service which it provides, the bank takes for itself a consideration which it includes in the
calculation of those rates.
33.
To hold that currency transactions are taxable only when effected in return for payment of a
commission or specific fees, which would thus allow a trader to avoid taxation if he sought to be
remunerated for his services by providing for a spread between the proposed transaction rates rather
than by charging such sums, would be a solution incompatible with the system put in place by the Sixth
Directive and would be liable to place traders on an unequal footing for purposes of taxation.
34.
It must therefore be held that foreign exchange transactions, performed even without commission or
direct fees, are supplies of services provided in return for consideration, that is to say supplies of services
effected for consideration within the meaning of art 2(1) of the Sixth Directive.
35.
 The answer to be given to the first question must therefore be that transactions between parties for the
purchase by one party of an agreed amount in one currency against the sale by it to the other party of an
agreed amount in another currency, both such amounts being deliverable on the same value date, and in
respect of which transactions the parties have agreed (whether orally, electronically or in writing) the
currencies involved, the amounts of such currencies to be purchased and sold, which party will purchase
which currency and the value date, constitute supplies of services effected for consideration within the
meaning of art 2(1) of the Sixth Directive.

The second question

36.
By its second question, the High Court essentially seeks to ascertain the precise nature of the
consideration. The question must therefore be understood as seeking to determine the taxable amount.
37.
The bank submits that the consideration is everything which is received in the course of foreign
exchange transactions, that is to say the turnover representing the total value of the currencies supplied
in the course of foreign exchange transactions.
38.
The Commission and the French government, on the other hand, take the view that the consideration
is the amount of exchange profit realised and the other remuneration obtained by the supplier.
39.
The Commission points out that it had prepared a proposal for a directive containing a provision
specifically relating to foreign exchange transactions (see the Proposal for a Nineteenth Council Directive
on the harmonisation of the laws of the member states relating to turnover taxes, amending the Sixth
Directive (OJ 1984 C347 p 5). The proposed amendment would have added the following sentences to
the second indent of art 19(1) (OJ 1984 C347 p 7):
The amount to be included in the denominator shall be reduced by the purchase price of
transfers of currency and securities exempted pursuant to Article 13B(d)(4) and (5); this amount
shall include, where appropriate, commission and expenses incurred by the purchaser. Where the
taxable 766 person cannot determine the purchase price he may substitute therefor the purchase
price of currency or securities acquired during the same period, provided those currencies or
securities are identical with those sold.
40.
The Commission explains that it withdrew this proposal for reasons unconnected with that provision.
41.
The UK government considers that, should the court take the view that the foreign exchange
transactions at issue are a service provided for consideration, any valuation based on the spread between
the bid and the offer prices would be incorrect for two reasons. First, the bank does not charge any
customer that spread. Second, such valuation would be tantamount to levying VAT on profit rather than
turnover. The UK government also submits that it is impossible to identify any consideration in foreign
exchange transactions, since the profits or receipts of the bank arise from its participation in a series of
transactions, all at different exchange rates, and not from profit on any individual transaction. Finally, the
currencies exchanged do not constitute consideration one for the other.
42.
It should be borne in mind that art 11A(1)(a) of the Sixth Directive provides that the taxable amount is,
in respect of supplies of services, that which constitutes the consideration which has been or is to be
obtained by the supplier from the purchaser for such supplies.
43.
While they are the subject of a supply, the currencies transferred to a trader by his counterparty in the
course of a foreign exchange transaction cannot be regarded as constituting remuneration for the service
of exchanging currencies for other currencies or consequently as constituting consideration for that
service.
44.
Determining the consideration therefore comes down to determining what the bank receives for foreign
exchange transactions, that is to say the remuneration on foreign exchange transactions which it can
actually take for itself (see, in this regard, the judgment in H J Glawe Spiel- und Unterhaltungsgerte
Aufstellungsgesellschaft mbH & Co KG v Finanzamt Hamburg-Barmbek-Uhlenhorst Case C-38/93 [1994]
STC 543, [1994] ECR I-1679 (para 9)).
45.
In this regard, the spread representing the difference between the bid price and the offer price is only
the notional price which the bank would obtain if it were to conclude, at the same instant and on similar
conditions, two corresponding purchase and sale transactions for the same amounts and the same
currencies.
46.
However, these are simply theoretical considerations, since the bank carries out a large number of
transactions relating to different amounts and involving different currencies, the rates of which are in
constant fluctuation. A trader cannot normally foresee, when concluding one particular transaction, at
what moment and at what price he may subsequently effect one or more transactions enabling him to
eliminate or fix, at a specific amount, the risk of a change in rate to which he is exposed following the first
transaction.
47.
So, the consideration, that is to say the amount which the bank can actually apply to its own use, must
be regarded as consisting of the net result of its transactions over a given period of time.
48.
It should be borne in mind in this regard that, in the case of transactions which are effected for
consideration but the actual consideration for which depends on future factors such as passage of time,
the court has already ruled that the taxable amount must be defined on the basis of, in particular, the
interest accrued over a deferred payment period, which was not yet known when the 767 taxable
transaction was concluded (see the judgment in Muys en De Winters Bouw- en Aannemingsbedrijf v
Staatssecretaris van Financin Case C-281/91 [1997] STC 665, [1993] ECR I-5405 (para 18)).
49.
Nor is it necessary for either the taxable person supplying the goods or performing the service or the
other party to the transaction to know the exact amount of the consideration serving as the taxable
amount in order for it to be possible to tax a particular type of transaction (see the judgment in Argos
Distributors Ltd v Customs and Excise Comrs Case C-288/94 [1996] STC 1359, [1996] ECR I-5311
(paras 2122)). Consequently, it does not matter that when the transaction is concluded the parties do not
know the basis on which VAT will be charged and that it remains unknown, even afterwards, to the
recipient of the service.
50.
The answer to be given to the second question must therefore be that art 11A(1)(a) of the Sixth
Directive is to be construed as meaning that, in foreign exchange transactions in which no fees or
commission are calculated with regard to certain specific transactions, the taxable amount is the net result
of the transactions of the supplier of the services over a given period of time.

Costs
51.
The costs incurred by the UK and French governments and the European Commission, which have
submitted observations to the Court of Justice, are not recoverable. Since these proceedings are, for the
parties to the main proceedings, a step in the proceedings pending before the national court, the decision
on costs is a matter for that court.
On those grounds, the Court of Justice (Fifth Chamber), in answer to the questions referred to it by the
Queens Bench Division of the High Court, by order of 13 May 1996, hereby rules: (1) Transactions
between parties for the purchase by one party of an agreed amount in one currency against the sale by it
to the other party of an agreed amount in another currency, both such amounts being deliverable on the
same value date, and in respect of which transactions the parties have agreed (whether orally,
electronically or in writing) the currencies involved, the amounts of such currencies to be purchased and
sold, which party will purchase which currency and the value date, constitute supplies of services effected
for consideration within the meaning of art 2(1) of Council Directive (EEC) 77/388 on the harmonisation of
the laws of the member states relating to turnover taxescommon system of value added tax: uniform
basis of assessment. (2) Article 11A(1)(a) of Directive 77/388 must be construed as meaning that, in
foreign exchange transactions in which no fees or commission are calculated with regard to certain
specific transactions, the taxable amount is the net result of the transactions of the supplier of the
services over a given period of time.

768

[1998] All ER (EC) 769

Silhouette International Schmied GmbH & Co KG v Hartlauer

Handelsgesellschaft mbH
(Case C-355/96)

EUROPEAN COMMUNITY; Free movement of goods: INTELLECTUAL PROPERTY; Trade Marks

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN (RAPPORTEUR), WATHELET AND
SCHINTGEN (PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, MURRAY,
EDWARD, JANN, SEVN AND IOANNOU.
ADVOCATE GENERAL JACOBS
14 OCTOBER 1997, 29 JANUARY, 16 JULY 1998
European Community Freedom of movement Goods Trade mark exhaustion Trademarked goods
sold by Austrian trade mark proprietor to buyer outside EEA, with instruction only to sell in Bulgaria and
former Soviet Union Goods obtained in Bulgaria by third party and sold in Austria without trade mark
proprietors consent Whether release onto non-EEA market rendering trade mark exhausted within EEA
Council Directive (EEC) 89/104, arts 5, 7.
S manufactured spectacles in the higher price ranges and marketed them worldwide under the trade mark
Silhouette, which was registered in Austria and most other countries of the world. H sold low-priced
spectacles in Austria and was not supplied by S because it considered that distribution of its products by
H would be harmful to its image as a manufacturer of top quality fashion spectacles. However, in
December 1995 H obtained a quantity of Ss frames, which S had supplied to a Bulgarian buyer in Sofia,
and offered them for sale in Austria under its trade mark. S, which had instructed a representative to
inform the Bulgarian buyer not sell them other than in Bulgaria or the former Soviet Union, brought an
action before the Landesgericht Steyr seeking an injunction restraining H from offering the frames for sale
in Austria. S contended that it had not exhausted its trade mark rights since, under art 7(1) 1 of Council
Directive (EEC) 89/104 to approximate the laws of the member states relating to trade marks, those rights
were exhausted only when the products had been put on the market in the European Economic Area (the
EEA) under that mark by the proprietor or with his consent. That action was dismissed by the
Landesgericht Steyr and, on appeal, by the Oberlandesgericht Linz. S appealed to the Oberster
Gerichtshof, which stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling the questions: (i) whether national rules providing for exhaustion of
trade mark rights, in respect of products put on the market outside the EEA under that mark by the
proprietor or with his consent, were contrary to art 7(1) of Directive 89/104; and (ii) whether a proprietor of
a trade mark was entitled, on the basis of art 7(1), to obtain an order restraining a third party from using
its trade mark for products which had been put on the market in a non-EEA state under that mark by the
proprietor or with his consent.
1
Article 7(1), so far as material, is set out at p 773 g, post

Held (1) Articles 52 to 7 of Directive 89/104 were to be construed as embodying a complete

harmonisation of the rules relating to the rights conferred 769 by a trade mark. The directive therefore
could not be interpreted as leaving it open to member states to provide in their domestic law for
exhaustion of the rights conferred by a trade mark in respect of products put on the market in non-EEA
states. That was the only interpretation capable of ensuring that the purpose of the directive, to safeguard
the functioning of the internal market, was achieved. Indeed, a situation in which some member states
could provide for international exhaustion while others provided for Community exhaustion only would
inevitably give rise to barriers to the free movement of goods and the freedom to provide services (see p
788 f to h and p 790 f, post).
2
Article 5, so far as material, is set out at p 773 d to f, post

(2) The directive required member states to implement provisions enabling a proprietor whose trade
mark rights were infringed to obtain an order restraining third parties from making use of his mark.
However, that requirement was imposed, not by art 7, but by art 5 of the directive. Accordingly, subject to
the national courts duty to interpret, as far as possible, domestic law in conformity with Community law,
art 7(1) of the directive could not be interpreted as meaning that the proprietor of a trade mark was
entitled, on the basis of that provision alone, to obtain an order restraining a third party from using his
trade mark for products which had been put on the market outside the EEA under that mark by the
proprietor or with his consent (see p 789 j and p 790 c g, post).

Notes
For the free movement of goods and trade marks, see 52 Halsburys Laws (4th edn) para 12108.

Cases cited
Centrafarm BV v Winthrop BV Case 16/74 [1974] ECR 1183.
EMI Records Ltd v CBS UK Ltd Case 51/75 [1976] ECR 811.
Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325, ECJ.
Factortame Ltd v Secretary of State for Transport (No 2) Case C-213/89 [1991] 1 All ER 70, [1991] 1 AC
603, [1990] 3 WLR 818, [1990] ECR I-2433, ECJ.
Herms International v FHT Marketing Choice BV Case C-53/96 (1997) Transcript (opinion), 13
November.
Javico International v Yves Saint Laurent Parfums SA Case C-306/96 (1997) Transcript (opinion) 6
November, ECJ.
Mag Instrument Inc v California Trading Co Norway, Ulsteen Case E-2/97 [1997] Report of the EFTA
Court 129.
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135.
MPA Pharma GmbH v Rhne-Poulenc Pharma GmbH Case C-232/94 [1996] ECR I-3671.
Opinion 1/91 [1991] ECR I-6079.
Opinion 1/94 [1994] ECR I-5267.
Polydor Ltd v Harlequin Record Shops Ltd Case 270/80 [1982] ECR 329.
Van Schijndel v Stichting Pensioenfonds voor Fysiotherapeuten Joined cases C-430431/93 [1996] All
ER (EC) 259, [1995] ECR I-4705, ECJ.
Reference
By order of 15 October 1996, the Oberster Gerichtshof referred to the Court of Justice of the European
Communities for a preliminary ruling under art 177 of the EC Treaty two questions (set out at p 787 b,
post) on the interpretation of art 7 of 770 Council Directive (EEC) 89/104 to approximate the laws of the
member states relating to trade marks, as amended by the Agreement on the European Economic Area
(Oporto, 2 May 1992; TS26 (1995); Cm 2847; OJ 1994 L1 p 3). Those questions arose in proceedings
between two Austrian companies, Silhouette International Schmied GmbH & Co KG and Hartlauer
Handelsgesellschaft mbH. Written observations were submitted on behalf of: Silhouette, by K Haslinger,
Rechtsanwalt, Linz; Hartlauer, by W Mller, Rechtsanwalt, Linz; the Austrian government, by W Okresek,
Ministerialrat in the Chancellors Office, acting as agent; the German government, by A Dittrich,
Regierungsdirektor in the Federal Ministry of Justice, and B Kloke, Oberregierungsrat in the Federal
Ministry of Economic Affairs, acting as agents; the French government, by C de Salins, Head of Sub-
directorate in the Legal Affairs directorate of the Ministry of Foreign Affairs, and P Martinet, Secretary in
the same directorate, acting as agents; the Italian government, by U Leanza, Head of the Legal Service in
the Ministry of Foreign Affairs, acting as agent, and O Fiumara, Avvocato dello Stato; the Swedish
government, by E Brattgrd, Departementsrd in the Foreign Trade Department of the Ministry of Foreign
Affairs, T Norstrm, Kanslird in the same ministry, and I Simfors, Hovrtts- assessor in the same
ministry, acting as agents; the UK government, by L Nicoll, of the Treasury Solicitors Department, acting
as agent, and by M Silverleaf, Barrister; and the European Commission, by J Grunwald, Legal Adviser,
and B J Drijber, of its Legal Service, acting as agents. Oral observations were made by Silhouette,
represented by K Haslinger; Hartlauer, represented by W Mller; the Italian government, represented by
O Fiumara; and the Commission, represented by J Grunwald. The language of the case was German.
The facts are set out in the opinion of the Advocate General.

29 January 1998.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
The Court of Justice of the European Communities case law on arts 30 and 36 of the EC Treaty
established for trade marks, as well as for other forms of intellectual property, a principle of Community
wide exhaustion3: thus the sale in the Community of trademarked goods, by or with the consent of the
trade mark owner, exhausts the trade mark rights throughout the Community, and he cannot, other than in
exceptional circumstances, oppose the use of the mark by others in subsequent transactions anywhere in
the Community.
3
The principle was established, in relation to trade marks, by the judgment in Centrafarm BV v Winthrop BV Case 16/74
[1974] ECR 1183.

2.
Article 7(1) of Council Directive (EEC) 89/104 to approximate the laws of the member states relating to
trade marks (OJ 1989 L40 p 1) (the trade marks directive) gives effect to the principle of Community
exhaustion as developed by the courts case law. It provides that a trade mark does not entitle the
proprietor to prohibit its use in relation to goods which have been put on the market in the Community
under that trade mark by the proprietor or with his consent. Subsequently the principle was extended, by
virtue of the Agreement on the European Economic Area (the EEA agreement) (Oporto, 2 May 1992; TS
26 (1995); Cm 2847; OJ 1994 L1, p 3), to the territory of the EEA, now consisting of the Community on
the one hand and Iceland, Liechtenstein and Norway on the other hand. But can the trade mark owner
prevent a third party from using the mark in the Community or in the EEA for goods which have been put
on the 771 market under that mark, by or with the consent of the owner, outside the EEA? The question
comes by way of a request for a preliminary ruling from the Oberster Gerichtshof (the Supreme Court),
Austria.
3.
The issue therefore is whether Community law requires member states to provide for exhaustion only
when the goods have been marketed in the EEA, or whether member states may (or perhaps even must)
provide for exhaustion when the goods have been marketed in a third countrya principle of international
(ie worldwide) exhaustion.

THE TRADE MARKS DIRECTIVE

4.
The trade marks directive was adopted under art 100a of the EC Treaty. Its aim was not to undertake
full-scale approximation of the trade mark laws of the Member States but simply to approximate those
national provisions of law which most directly affect the functioning of the internal market (see the third
recital of the preamble).
5.
The first, third, and ninth recitals of the preamble to the directive state, respectively:
Whereas the trade mark laws at present applicable in the Member States contain disparities
which may impede the free movement of goods and freedom to provide services and may distort
competition within the common market; whereas it is therefore necessary, in view of the
 establishment and functioning of the internal market, to approximate the laws of Member States
Whereas it does not appear to be necessary at present to undertake full-scale approximation of the
trade mark laws of the Member States and it will be sufficient if approximation is limited to those
national provisions of law which most directly affect the functioning of the internal market
Whereas it is fundamental, in order to facilitate the free circulation of goods and services, to ensure
that henceforth registered trade marks enjoy the same protection under the legal systems of all the
Member States; whereas this should however not prevent the Member States from granting at their
option extensive protection to those trade marks which have a reputation.
6.
In summary, the directive harmonises the general conditions for obtaining and continuing to hold a
registered trade mark (seventh recital) and the rights conferred by a trade mark (arts 5, 6 and 7). Thus it
specifies signs of which a trade mark may consist (art 2), the grounds for refusing to register or
invalidating a trade mark (arts 3 and 4), the consequences of acquiescence in the use of a later trade
mark (art 9) and of failure to use a registered trade mark (arts 10 to 12), and the grounds for revocation of
a trade mark (art 12).
7.
However, in certain areas member states are given a discretion to decide whether to adopt the rules
provided for in the directive: for example, there are certain optional grounds for refusing to register or
invalidating a trade mark (arts 3(2) and 4(4)) and an option as to whether or not to provide protection in
certain specified circumstances for a trade mark with a reputation concerning its use in relation to
dissimilar goods or services (art 5(2))4. In addition, the seventh recital specifies:
4
See also arts 3(4), 9(2) and 15(2).

Member States will be able to maintain or introduce into their legislation grounds of refusal or
invalidity linked to conditions for obtaining 772 and continuing to hold a trade mark for which there
is no provision of approximation, concerning, for example, the eligibility for the grant of a trade
mark, the renewal of the trade mark or rules on fees, or related to the non-compliance with
procedural rules.
The directive also leaves to the member states matters such as the procedure concerning the
registration, revocation and invalidity of trade marks (fifth recital), the protection of unregistered trade
marks (fourth recital) and provisions relating to unfair competition, civil liability and consumer protection
(sixth recital).
8.
The most important provisions in relation to the present case are arts 5 and 7, entitled, respectively,
Rights conferred by a trade mark and Exhaustion of the rights conferred by a trade mark.
9.
Article 5 provides:
(1) The registered trade mark shall confer on the proprietor exclusive rights therein. The
proprietor shall be entitled to prevent all third parties not having his consent from using in the
course of trade: (a) any sign which is identical with the trade mark in relation to goods or services
which are identical with those for which the trade mark is registered; (b) any sign where, because of
its identity with, or similarity to, the trade mark and the identity or similarity of the goods or services
covered by the trade mark and the sign, there exists a likelihood of confusion on the part of the
public, which includes the likelihood of association between the sign and the trade mark (3) The
following, inter alia, may be prohibited under paragraphs 1 and 2: (a) affixing the sign to the goods
or to the packaging thereof; (b) offering the goods, or putting them on the market or stocking them
for these purposes under that sign, or offering or supplying services thereunder; (c) importing or
exporting the goods under the sign; (d) using the sign on business papers and in advertising.
10.
Article 7 provides:
(1) The trade mark shall not entitle the proprietor to prohibit its use in relation to goods which
have been put on the market in the Community under that trade mark by the proprietor or with his
consent. (2) Paragraph 1 shall not apply where there exist legitimate reasons for the proprietor to
oppose further commercialization of the goods, especially where the condition of the goods is
changed or impaired after they have been put on the market.
11.
Provisions on exhaustion of similar effect to those contained in art 7 have been included in other
Community instruments on intellectual property rights5. The most relevant in that respect is Council
Regulation (EC) 40/94 on the Community trade mark (OJ 1994 L11 p 1), which I consider below.
5
See eg art 9(2) of Council Directive (EC) 92/100 on rental right and lending right and on certain rights related to
copyright in the field of intellectual property (OJ 1992 L346 p 61). The European Commission considers that those
provisions also have the effect of excluding international exhaustion: see its answer to a written question in the
European Parliament (OJ 1994 C340 p 37).

THE EEA AGREEMENT

12.
 Although art 7(1) of the trade marks directive refers to marketing in the Community, the principle of the
exhaustion of rights, as previously mentioned, 773was extended for certain purposes to the EEA. The
directive was one of the legislative acts incorporated into EEA law by the EEA agreement, which entered
into force on 1 January 1994 (1 May 1995 in relation to Liechtenstein). Annex XVII to the agreement
amends art 7(1) of the directive for the purposes of the Agreement so as to refer to marketing within the
EEA rather than the Community: it replaces the words in the Community with the words in a Contracting
Party (see OJ 1994 L1 p 482, point 3). Moreover, a protocol to the agreement, Protocol 28 on intellectual
property, contains an article, art 2, headed Exhaustion of rights (see p 194). Article 2(1) provides:
To the extent that exhaustion is dealt with in Community measures or jurisprudence, the
Contracting Parties shall provide for such exhaustion of intellectual property rights as laid down in
Community law. Without prejudice to future developments of case-law, this provision shall be
interpreted in accordance with the meaning established in the relevant rulings of the Court of
Justice of the European Communities given prior to the signature of the Agreement.
13.
No issue under that protocol arises in the present case, the facts of which occurred after Austria
(previously an EEA member state) acceded to the Community on 1 January 1995.

THE FACTS
14.
The plaintiff, Silhouette International Schmied Gesellschaft mbH & Co KG (Silhouette), is an Austrian
company which produces fashion spectacles in the higher price ranges. It distributes the spectacles
worldwide under the word and picture trade mark Silhouette, which is registered in Austria and in most
countries of the world, as well as internationally. In Austria Silhouette supplies the spectacles to specialist
opticians; in other countries it has subsidiary companies or distributors.
15.
The defendant, Hartlauer Handelsgesellschaft mbH (Hartlauer), sells spectacles in numerous
branches in Austria and solicits customers mainly by its low prices. It is not supplied by Silhouette
because Silhouette considers sales by Hartlauer to be harmful to the image which Silhouette has created
for its products as fashionable spectacles of special quality.
16.
In October 1995 Silhouette sold 21,000 spectacle frames of an outdated model which had expired to a
firm called Union Trading for $261,450. The transaction was arranged by Silhouettes sales representative
for the Middle East. Silhouette directed him to instruct the purchaser to sell the frames in Bulgaria or the
states of the former Soviet Union only and not to export them to other countries. The sales representative
informed Silhouette that he had instructed the purchaser accordingly. The Oberster Gerichtshof observes
that it has not been possible to ascertain whether that actually happened.
17.
Silhouette delivered the goods to Union Trading in Sofia in November 1995. Hartlauer subsequently
acquired the goods (according to the Oberster Gerichtshof, it has not been possible to ascertain from
whom) and offered them for sale in Austria from December 1995. It announced in a press campaign that,
although it had not been supplied by Silhouette, it had succeeded in purchasing 21,000 Silhouette frames
from abroad. In its observations, Hartlauer maintains that when it acquired the products it was assured
that there would be no obstacle to importing them into Austria.
774
18.
Silhouette objects to the sale of its frames by Hartlauer in Austria and seeks an order prohibiting
Hartlauer from marketing under its trade mark spectacles or spectacle frames which were not put on the
market in the EEA by it or with its consent. It argues that it has not exhausted its trade mark rights
because the directive provides that such rights can be exhausted only by reason of marketing within the
EEA by the trade mark owner or with his consent. It bases its claim not only on para 10a of the
Markenschutzgesetz (the law on protection of trade marks), but also on paras 1 and 9 of the Gesetz
gegen den unlauteren Wettbewerb (the UWG) (the law on unfair competition), and art 43 of the
Allgemeines Brgerliches Gesetzbuch (the ABGB) (the Civil Code).
19.
Hartlauer contends that Silhouette did not sell the frames subject to the instruction that any import into
the Community was excluded and that Silhouettes application should be dismissed.
20.
Silhouettes action failed before the Landgericht Steyr (the Steyr Regional Court) and on appeal to the
Oberlandesgericht Linz (the Linz Higher Regional Court). The current reference is made in the context of
an appeal by Silhouette to the Oberster Gerichtshof against the decision of the Oberlandesgericht Linz.
21.
Article 7 of the trade marks directive was implemented into Austrian law almost word for word by the
1992 amendments to the Markenschutzgesetz. Paragraph 10a of the Markenschutzgesetz provides that
the trade mark does not entitle the owner of the trade mark to prohibit a third party from using the mark for
goods which have been put on the market in the EEA under that mark by the owner or with his consent.
22.
The Oberster Gerichtshof explains that, prior to the implementation of the trade marks directive, the
principle of international exhaustion had been applied by the Austrian courts. It refers to the decision of
the Oberster Gerichtshof in the case of Agfa6. The position subsequent to implementation of the directive,
however, is unclear. According to the explanatory memorandum it was intended that the question of the
validity of the principle of international exhaustion should be settled by legal practice 7.
6
See (1971) SZ 43/219.
7
See 669 BlgNR 18 GP5.

23.
The Oberster Gerichtshof accordingly wishes to ascertain whether the directive allows member states
to apply a rule of international exhaustion. It also poses a second question concerning the remedies
which should be made available to the trade mark owner under the directive.
24.
The Oberster Gerichtshof has submitted the following questions to the Court of Justice:
Is Article 7(1) of the First Council Directive 89/104/EEC of 21 December 1988 to approximate
the laws of the Member States relating to trade marks (the trade marks directive) to be interpreted
as meaning that the trade mark entitles its proprietor to prohibit a third party from using the mark for
goods which have been put on the market under that mark in a State which is not a Contracting
State? May the proprietor of the trade mark on the basis of Article 7(1) of the Trade Marks Directive
alone seek an order that the third party cease using the trade mark for goods which have been put
on the market under that mark in a State which is not a Contracting State?
The reference to a contracting state is to be understood as referring to a contracting party to the EEA
agreement, ie on the European Free Trade 775 Association (EFTA) side, those EFTA states which are
parties to the agreement (currently Iceland, Liechtenstein and Norway), and on the Community side, the
Community and/or the EC member states (see art 2(c)). Hence the questions are concerned with goods
put on the market outside the EEA. It is unnecessary to consider what the position would be in relation to
goods marketed within the EEA and subsequently imported into the Community. For convenience I shall
refer in what follows to importing into the Community goods marketed outside the EEA.
25.
Written observations have been submitted by Silhouette, Hartlauer, the Austrian, French, German,
Italian, Swedish, and UK governments, and the European Commission. At the hearing oral submissions
were made by Silhouette, Hartlauer, the Italian government, and the Commission.

QUESTION 1
26.
By its first question, the Oberster Gerichtshof asks whether art 7(1) of the trade marks directive is to
be interpreted as meaning that the proprietor of a trade mark is entitled to prohibit a third party from using
the mark for goods which have been put on the market under that mark in a state which is not a member
of the EEA. It is not specified in the question whether the trade mark owner consented to such marketing
in the non-EEA state. However, it is clear from the order for reference that Silhouette did consent to
marketing in Bulgaria since it is stated that Silhouette gave directions for the sale of the goods there and
delivered them to the purchaser in Sofia. The question should accordingly be addressed on the
assumption that the trade mark owner consented to the marketing of his products outside the EEA.
27.
It should also be assumed for present purposes that Silhouette did not consent to its products being
resold within the EEA. That is so even though the national court expresses some doubt as to whether the
restrictions upon resale were passed on to the purchaser. If Silhouette had consented to marketing in the
EEA, the answer to the first question referred would clearly be that Silhouette could not oppose the import
of its products into Austria.
28.
The Oberster Gerichtshof has not suggested that there are any legitimate reasons within the meaning
of art 7(2) of the directive for Silhouette to oppose the resale of its spectacles in Austria.
29.
Thus, in the present case, the Court of Justice is faced squarely with the question whether the trade
marks directive, in referring to the exhaustion of trade mark rights following marketing in the Community,
requires member states to allow a trade mark owner to oppose the import into the Community of products
placed on the market outside the EEA by him or with his consent simply because he has not consented to
the marketing of those products within the Community: ie whether it precludes member states from
adopting the principle of international exhaustion.

The terms of the directive

30.
Article 7(1) of the directive provides for exhaustion only where the goods have been put on the market
in the Community: it provides therefore only for Community wide, not for international exhaustion.
31.
It is accepted on all sides, and with good reason in my view, that the directive does not require
member states to provide for international exhaustion: at most, it leaves that open as an option for
member states. If the directive had 776 sought to impose international exhaustion, art 7(1) would not have
referred only to marketing in the Community.
32.
 That the directive did not intend to impose international exhaustion is confirmed by the legislative
history of the directive. The Commissions original proposal would have imposed international exhaustion
(see OJ 1980 C351 p 1)8. The Commission subsequently changed its view, and its amended proposal
explicitly limited the exhaustion principle to goods which had been put on the market in the Community
(see p 4).
8
For the Explanatory Memorandum, see COM (80) 635 final.

33.
As to whether the directive precludes international exhaustion, or leaves that open, the language of art
7(1) inclines me to the former view. Article 7(1) spells out the circumstances in which the trade mark rights
are exhausted: it is naturally read as doing so exhaustively. In providing that the rights are exhausted
when the goods are marketed in the Community, art 7(1) is naturally understood as meaning that the
rights are not exhausted when the goods are marketed in a third country. It is true that the directive does
not specifically preclude international exhaustion, but that effect can reasonably be inferred from the
language. I accept that there are arguments which go the other way, but those arguments derive little
support from the language of the directive.
34.
My view of the effect of the language of art 7(1) is supported by the structure of the directive. Article
7(1) is a derogation from the rights conferred on the trade mark owner by art 5(1). In general derogations
should not be construed broadly. Here art 7(1) cannot be construed more broadly than as providing for
Community exhaustion. It would be necessary to read into the directive a further, implied derogation
leaving open the possibility of provision for international exhaustion, which seems contrary to the structure
of the directive.

The aims and scope of the directive

35.
Since the terms of the directive are not conclusive, the aims and scope of the directive are of crucial
significance in interpreting its provisions. The indications in the preamble, however, do not all point in the
same direction. On the one hand, it will be recalled that the directive does not purport to undertake full-
scale approximation of the trade mark laws of the Member States but aims to approximate those national
provisions of law which most directly affect the functioning of the internal market. On the other hand, the
directive seeks to ensure, with certain limited exceptions, that trade marks enjoy the same protection
under the legal systems of all the Member States.
36.
Those who favour international exhaustion point to the limited nature of the harmonisation attempted
by the directive and contend that the reference to Community exhaustion in art 7(1) should be regarded
only as a minimum standard.
37.
Moreover, they argue that the intention of art 7 was simply to codify the courts existing case law on
the exhaustion of rights since the court has stressed that art 7 is to be interpreted in the same way as the
courts case law on arts 30 and 36. They contend that, prior to implementation of the directive, member
states had a discretion as to whether or not to adopt the principle of international exhaustion; and that, in
the absence of express language to the contrary, that should remain the position under the directive.
777
38.
The opponents of international exhaustion, relying on the wording of the third recital of the preamble to
the directive, argue that, whilst it is true that the directive is not a measure of total harmonisation, the
application by a member state of the principle of international exhaustion is one of the provisions which
most directly affect the functioning of the internal market and is accordingly the type of issue which the
directive sought to harmonise. Moreover, the purpose of the directive was to ensure that trade marks
enjoy the same protection under the legal systems of all the Member States. Although the protection
afforded by the directive does not impose a totally uniform system since certain areas are left to the
discretion of the member states, those areas are very limited and the choice carefully specified (see para
7, above).
39.
As regards the scope and effects of the directive, it can in my view be argued that the directive has
transformed the impact of Community law on trade mark protection. Previously the only issue under
Community law was that of the impact of arts 30 to 36 of the Treaty on national trade mark law. The
directive harmonises the essential conditions and consequences of trade mark protection. Although in an
internal Community context the court has treated art 7 of the directive as codifying the previous case law,
it cannot be assumed that that is the sole function of art 7. The directive regulates the substance of trade
mark rights, and its provisions are designed to be substituted for the diverse national laws across the
whole range of its provisions.
40.
If the directive is seen as establishing the essential terms and effects of trade mark protection, it is
difficult to argue that it leaves member states free to opt for international exhaustion. The scope of the
exhaustion principle is after all central to the content of trade mark rights.
41.
 But even if one takes a narrower view of the character of the directive, it seems clear that international
exhaustion is one of the matters which most directly affect the functioning of the internal market and
which the directive therefore seeks to harmonise. If some member states practise international exhaustion
while others do not, there will be barriers to trade within the internal market, which it is precisely the object
of the directive to remove.
42.
It is above all on that ground that the Austrian, French, German, Italian and UK governments and the
Commission all submit that the directive should be interpreted as precluding the principle of international
exhaustion. Essentially they argue that, if the member states were free to determine whether trade mark
owners could prevent imports from third countries, then the same products could be the subject of parallel
imports into one member state but not into another, a result incompatible with the internal market. It is of
course no answer to that submission to suggest that once goods were imported into one member state
which did provide for international exhaustion they could then benefit from free movement throughout the
Community, since that suggestion would have the effect of imposing international exhaustion on all
member states, which, as has been seen above, would be contrary to the directive. The submission of the
five member states and of the Commission has in my view much force.
43.
A similar argument was advanced by some member states (France, Germany and the United
Kingdom) and by the Commission before the EFTA Court in Mag Instrument Inc v California Trading Co
Norway, Ulsteen Case E-2/97 [1997] Report of the EFTA Court 129. That case concerned the
interpretation of art 7(1) of the directive and in particular the issue of international exhaustion in relation to
the EFTA states. As mentioned above, art 7(1) was extended, for the purposes of the EEA agreement, to
goods marketed throughout the EEA. The 778 EFTA Court responded as follows ([1997] Report of the
EFTA Court 129 at 135 (paras 2526)):
(25) This argumentation has to be rejected in so far as it concerns the EFTA States. Unlike the
EC Treaty, the EEA Agreement does not establish a customs union. The purpose and the scope of
the EC Treaty and the EEA Agreement are different (see Opinion 1/91 of the ECJ regarding the
Draft Agreement between the Community, on the one hand, and the countries of the European
Free Trade Association, on the other, relating to the creation of the European Economic Area [1991]
ECR I-6079). Thus, the EEA Agreement does not establish a customs union, but a free trade area.
(26) The above-mentioned differences between the Community and the EEA will have to be
reflected in the application of the principle of exhaustion of trade mark rights. According to Article 8
EEA, the principle of free movement of goods as laid down in Articles 11 to 13 EEA applies only to
goods originating in the EEA, while in the Community a product is in free circulation once it has
been lawfully placed on the market in a Member State. In general, the latter applies in the context
of the EEA only in respect of products originating in the EEA. In the case at hand, the product was
manufactured in the United states and imported into Norway. Accordingly, it is not subject to the
principle of the free movement of goods within the EEA.
44.
The EFTA Court concluded that it was for the EFTA states, ie their legislators or courts, to decide
whether to introduce or maintain the principle of international exhaustion with regard to goods originating
outside the EEA. However the EFTA Court did not consider the question of goods originating within the
EEA.

Article 100a of the EC Treaty

45.
In the present case the Swedish government, in contrast to the other governments, contends that the
directive leaves the issue of international exhaustion to be resolved by national law. It argues that a
directive based solely on art 100a of the Treaty could not regulate the question of international
exhaustion. The Swedish government contends that that question concerns the relations between the
member states and third countries; moreover according to Opinion 1/94 [1994] ECR I-5267 on the
Agreement establishing the World Trade Organisation Agreement (the WTO agreement) (Marakesh, 15
April 1994; TS 12 (1996); Cm 3045; OJ 1994 L336 p 3), the external competence in matters of intellectual
property is not exclusive to the Community.
46.
It seems to me however that a distinction has to be made between measures of commercial policy on
the one hand and provisions governing the effects of trade mark rights within the Community on the other.
Although to preclude international exhaustion clearly has an effect on external trade, it is less clear that it
actually regulates such trade: contrary to the suggestion of the Swedish government, the directive, if
interpreted as precluding international exhaustion, would not regulate relations between Member States
and third States. Rather, the directive lays down the rights of trade mark owners in the Community. It
provides for the conditions under which the trade mark owner can take action against the marketing of
certain goods, which may or may not be imported from third countries. Moreover it is inevitable that
internal market measures will affect imports from third countries. Thus measures harmonising 779
technical standards will affect goods from third countries, but can properly be based on art 100a of the
Treaty.
47.
With regard to Opinion 1/94 and the external competence of the Community, that issue would arise
only if negotiations were to be undertaken with third countries to deal with international exhaustion. No
doubt considerations of commercial policy and concern about the possible lack of reciprocity were among
the reasons why the provision for international exhaustion which featured in the Commissions original
proposal was not maintained. But the existence of such underlying policy considerations does not limit the
material scope of a measure based on art 100a. It does not lead to the conclusion that a measure based
on art 100a cannot be construed as having dealt with the subject of international exhaustion. It seems to
me that the Community can regulate under art 100a the rights of trade mark owners within the Community
in respect of goods bearing their mark whether they have been marketed inside or outside the
Community.

The origin function of trade marks

48.
The Swedish government also relies on the Court of Justices case law on the function of trade marks.
That function is essentially to guarantee the consumer the possibility of identifying the origin of the
product. It is no part of the function of a trade mark to enable the owner to divide up the market and to
exploit price differentials. The adoption of international exhaustion would bring substantial advantages to
consumers, and would promote price competition.
49.
I confess to finding those arguments extremely attractive. However it must be remembered that the
courts case law on the function of trade marks was developed in the context of the Community, not the
world market. Indeed in EMI Records Ltd v CBS UK Ltd Case 51/75 [1976] ECR 811 the court held that
its case law under arts 30 to 36 could not be transposed to imports from third countries. Circumscribing
the protection of trade mark rights by defining their essential function was considered necessary to
prevent restrictions on trade between member states.
50.
Such compelling considerations do not apply to imports from third countries. On the contrary, to allow
member states to opt for international exhaustion would itself, as has been seen, result in barriers
between member states.
51.
There is of course a powerful argument based on the concern for free trade at the international level.
To some commentators the exclusion of international exhaustion will appear protectionist and therefore
harmful9. Commercial policy considerations may however be more complex than they allow for. I have
already alluded to concern about the possible lack of reciprocity if the Community were unilaterally to
provide for international exhaustion. In any event it is no part of the courts function to seek to evaluate
such policy considerations.
9
Advocates of international exhaustion (to a greater or lesser extent) include: Beier (1990) IIC 131; Rasmussen The
Principle of Exhaustion of Trade Mark Rights Pursuant to Directive 89/104 (and Regulation 40/94) [1995] 4 EIPR 174;
Shea Does the First Trade Mark Directive Allow International Exhaustion of Rights? [1995] 10 EIPR 463; and Verkade
Extra-communautaire parallelimport en rechten van intellectuele eigendom (1997) SEW p 304.

52.
As regards price competition and the benefit to consumers, such benefits again have to be set against
the threat to the integrity of the internal market. That 780 integrity would be severely prejudiced if one
member state provided for international exhaustion while another did not. Only consumers in the first
state would benefit from the lower prices of imports from third countries. Price competition within the
internal market would be distorted.
53.
As regards the Communitys competition policy, the ruling to be given by the court on international
exhaustion will in no way limit the possible application of the competition rules of the Treaty. It will not
exclude the possibility that arts 85 and 86 of the Treaty may apply to agreements between undertakings,
or to unilateral behaviour by a dominant undertaking, seeking to divide up the markets 10.
10
For the possible application of art 85, see eg EMI Records Ltd v CBS UK Ltd Case 51/75 [1976] ECR 811, and most
recently the opinion of Advocate General Tesauro in Javico International v Yves Saint Laurent Parfums SA Case C-
306/96 (1997) ECJ Transcript (opinion), 6 November.

54.
Finally, it should be recalled that some member states, and some third countries, do not practise
international exhaustion, and that that has not been held to be contrary to the General Agreement on
Tariffs and Trade (the GATT). The situation is not changed in that respect by the WTO agreement. Annex
1C, the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs) (Marakesh, 15 April
1994; TS 10 (1996); Cm 3046; OJ 1994 L336 p 213), provides by art 6 that, for the purposes of dispute
settlement under that agreement, nothing in the agreement (subject to certain provisions) shall be used to
address the issue of the exhaustion of intellectual property rights 11.
11
On art 50 of TRIPs, which relates to provisional measures, see the opinion of Advocate General Tesauro in Herms
International v FHT Marketing Choice BV Case C-53/96 (1997) ECJ Transcript (opinion), 13 November.

The Community trade mark regulation

55.
Further guidance on the interpretation of the directive is provided by Regulation 40/94 (the trade mark
regulation). The regulation, which provides for a single Community mark valid throughout the Community,
was drafted concurrently with the directive and it contains a virtually identical provision on exhaustion.
56.
Article 1(2) provides that a Community trade mark has a unitary character and that
It shall have equal effect throughout the Community: it shall not be registered, transferred or
surrendered or be the subject of a decision revoking the rights of the proprietor or declaring it
invalid, nor shall its use be prohibited, save in respect of the whole Community. This principle shall
apply unless otherwise provided in this Regulation.
57.
Article 13, entitled Exhaustion of the rights conferred by a Community trade mark, provides as
follows:
(1) A Community trade mark shall not entitle the proprietor to prohibit its use in relation to goods
which have been put on the market in the Community under that trade mark by the proprietor or
with his consent. (2) Paragraph 1 shall not apply where there exist legitimate reasons for the
proprietor to oppose further commercialization of the goods, especially where the condition of the
goods is changed or impaired after they have been put on the market.
781
58.
Thus, except for the reference to the Community trade mark, art 13 of the regulation is in identical
terms to art 7 of the directive.
59.
As in the case of the directive, the Commissions original proposal for the regulation would have
provided for international exhaustion, but again the proposal was amended and the regulation provides
for exhaustion only for goods which have been put on the market in the Community. Once again,
therefore, it is impossible to read the regulation as imposing international exhaustion. The choice would
then be between precluding international exhaustion or leaving the issue to the member states.
60.
In the case of the regulation, however, it seems scarcely possible to contend that the member states
have a discretion. Whereas the directive, as has been seen, is a partial measure of harmonisation of
national laws, the regulation governs comprehensively the incidents and effects of a Community trade
mark. Moreover art 14(1) provides: The effects of Community trade marks shall be governed solely by
the provisions of this Regulation; leaving only infringement actions to be governed by national law in
accordance with Title X of the regulation, which is concerned only with jurisdiction and procedure in legal
actions relating to Community trade marks.
61.
It therefore seems impossible to contend that the trade mark regulation confers a discretion on
member states to opt for international exhaustion. The question then is whether the provisions on
exhaustion in the regulation and directive, notwithstanding their common origin and their identical
wording, can be construed differently. There are of course well-known examples of identical provisions
being construed differently in different contexts, notably in the context of the Treaty on the one hand and
in the context of a free trade agreement on the other, as in Polydor Ltd v Harlequin Record Shops Ltd
Case 270/80 [1982] ECR 329. The advisory opinion of the EFTA Court in the Mag Instrument case
provides a further illustration of the reasons for adopting a different approach, justified by the different
context, to the very provision in issue in the present case. In the present case however the context is,
both for the regulation and for the directive, the Communitys internal market. Although an argument can
be advanced that the objectives of the two instruments are different, since the directive only aims to
achieve a limited measure of harmonisation, it must be accepted that the regulation provides at least
some further support for the view that the directive precludes international exhaustion.
62.
I accordingly conclude, having regard to the wording and purpose of the directive, its legislative history,
the identical wording in the trade mark regulation, and the undesirable effects of leaving the question to
the discretion of the member states, that art 7(1) of the directive precludes member states from adopting
the principle of international exhaustion.
63.
Thus, in answer to the first question, art 7(1) of the directive is to be interpreted as meaning that the
proprietor of a trade mark is entitled to prevent a third party from using the mark for goods which have
been put on the market under that mark outside the territory of the EEA. Member states are accordingly
precluded from adopting the principle of international exhaustion.

QUESTION 2
64.
By its second question the Oberster Gerichtshof seeks to ascertain whether art 7(1) of the directive
can alone constitute the basis for the grant of an injunction in support of a trade mark owner seeking to
prohibit the sale of his goods which are circulating in the Community without his consent. It appears from
the order 782 for reference and from a subsequent communication from the referring court that that
question arises for the following reasons.
65.
Under Austrian trade mark law there is no right to a prohibitory injunction in respect of trade mark
infringement: an injunction to cease a trade mark infringement can however be sought under para 9 of the
UWG (the law on unfair competition). Under para 9(1) of the UWG, a prohibitory order may be made
against a person who in the course of business uses a name, a trading name or the particular designation
of an undertaking in a manner which is liable to cause confusion with the name, trading name or particular
designation which another person makes use of with authority. A registered trade mark qualifies as a
particular designation within the meaning of that provision (para 9(3) UWG). However, it appears that
under Austrian law injunctions cannot be obtained on the basis of art 9 of the UWG to prevent parallel
imports since the Oberster Gerichtshof states that the marketing of genuine goods cannot cause
confusion within the meaning of that provision when the goods in issue are original products of the trade
mark owner.
66.
Although there are two other bases in Austrian law upon which a prohibitory injunction may be based,
the Oberster Gerichtshof appears to consider that neither is applicable to the present case. Those
additional provisions are para 1 of the UWG and art 43 of the ABGB (the Civil Code). Pursuant to the
former, a prohibitory order may be made against a person who in the course of business for purposes of
competition performs acts which are against public policy. A breach of the law may be against public
policy within the meaning of that provision. However, the breach must be subjectively reprehensible and
liable to give the person acting illegally an advantage over his law-abiding competitors. Article 43 of the
ABGB allows a claim for a prohibitory injunction where a persons right to use his name is disputed or
where he is adversely affected by the unauthorised use of his name (or pseudonym).
67.
Since the Oberster Gerichtshof considers that none of the above provisions provides the foundation
for an injunction in the present case, it believes that it will be unable to grant an injunction to Silhouette
unless the right to an injunction flows from the wording of art 7(1) of the trade marks directive. It reasons
that, if art 7(1) can be interpreted in that way, it will be able to issue an injunction on the basis of para 10a
of the Markenschutzgesetz since the two provisions are in virtually identical terms. It observes that the
issue is not whether that provision of the directive can have direct effect, since the provision has been
copied almost verbatim into Austrian law. Rather it wishes to elucidate the correct interpretation of that
provision. Although it is art 5 of the directive, rather than art 7, which confers the substantive rights upon
the trade mark owner, it appears that the Oberster Gerichtshof is considering only the interpretation of the
latter provision since art 5(1)(a) has not been implemented in Austrian law.
68.
It is well established that, whether or not the specific provisions of a directive have been implemented
in national law, and independently of the possible direct effect of those provisionswhich, if not
implemented, can be invoked only against the state or a public bodythe national courts are under a duty
to take account of all provisions of national law so as to ensure, wherever possible, that the result
prescribed by the directive is attained (see the judgment in Marleasing SA v La Comercial Internacional
de Alimentacin SA Case C-106/89 [1990] ECR I-4135). That duty applies as regards not only national
legislation specifically introduced in order to implement a directive but also other provisions 783 of
national law, including those adopted before the directive. As the Court of Justice stated in relation to the
very provision in issue in the present case (MPA Pharma GmbH v Rhne-Poulenc Pharma GmbH Case
C-232/94 [1996] ECR I-3671 (para 12)):
when applying national law, whether adopted before or after the directive, the national court
that has to interpret that law must do so, as far as possible, in the light of the wording and the
purpose of the directive so as to achieve the result that it has in view and thereby comply with the
third paragraph of Article 189 of the EC Treaty.
69.
It follows that, whether or not the national legislation relating to trade marks has been amended to give
effect to all the provisions of the directive, that legislation must be interpreted consistently with the
directive. Provided that the legislation is capable of being interpreted in that way, the national courts are
under a duty to give trade marks the same protection as if each of the provisions of the directive had been
specifically and explicitly transposed into national law.
70.
Thus although in the present case it appears that art 5(1)(a) of the directive, which confers exclusive
rights on the trade mark owner, has not been implemented in Austrian law, nevertheless the courts are
required, as far as possible, to interpret the Austrian legislation in the light of art 5(1)(a). Article 5(1) states
that the proprietor of the mark shall be entitled to prevent all third parties not having his consent from
using the mark. The terms of art 5(1) thus envisage that he shall be entitled to a court order prohibiting
the use of the mark. Where such an order is available under national legislation, whether under the
legislation on trade marks or under other legislation such as that on unfair competition, it must therefore
also be made available in the case of infringement of trade mark rights as defined by art 5(1) of the
directive.
71.
Moreover the case law of the Court of Justice recognises, as a general principle of law, that the
national courts must provide effective remedies for the enforcement of Community rights. The case law
has established two principles in particular: first that national rules governing remedies for the exercise of
Community rights must not be less favourable than those available for rights arising under national law;
and secondly that the exercise of Community rights must not in any event be rendered impossible in
practice or excessively difficult (see eg Van Schijndel v Stichting Pensioenfonds voor Fysiotherapeuten
Joined cases C-430431/93 [1996] All ER (EC) 259, [1995] ECR I-4705). It may well be that the refusal of
the remedy which is in issue in the present case, namely an injunction, would, in the circumstances of this
case, contravene both those requirements.
72.
So far as interlocutory injunctions are concerned, the court has held that national courts may be
required to afford interim relief for the protection of Community rights even in cases where they would be
unable to do so under national law (see the judgment in Factortame Ltd v Secretary of State for Transport
(No 2) Case C-213/89 [1991] 1 All ER 70, [1990] ECR I-2433). The court stated that the full effectiveness
of Community law would be impaired if a rule of national law could prevent a national court from granting
interim relief pending the national courts final judgment (see [1991] 1 All ER 70, [1990] ECR I-2433 (para
21)). It seems clear that the same applies to a final injunction: that remedy also must be ensured by the
national court where it is necessary to ensure the effective protection of the rights conferred by
Community law.
784

CONCLUSION
73.
Accordingly I am of the opinion that the questions referred by the Oberster Gerichtshof should be
answered as follows:
(1) Article 7(1) of Council Directive (EEC) 89/104 to approximate the laws of the member states
relating to trade marks must be interpreted as meaning that the proprietor of a trade mark is entitled
to prevent a third party from using the mark for goods which have been put on the market under
that mark outside the European Economic Area.
(2) Even where art 7(1) alone of the relevant provisions of the directive has been specifically
transposed into national law, the proprietor of the trade mark is entitled to obtain an order
prohibiting the third party from using the mark for goods which have been put on the market under
that mark outside the European Economic Area.

16 July 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 15 October 1996, received at the Court of Justice of the European Communities on 30
October 1996, the Oberster Gerichtshof referred to the Court of Justice for a preliminary ruling under art
177 of the EC Treaty two questions on the interpretation of art 7 of Council Directive (EEC) 89/104 to
approximate the laws of the member states relating to trade marks (OJ 1989 L40 p 1), as amended by the
Agreement on the European Economic Area (the EEA agreement) (Oporto, 2 May 1992; TS26 (1995);
Cm 2847; OJ 1994 L1 p 3).
2.
Those questions were raised in proceedings between two Austrian companies, Silhouette International
Schmied GmbH & Co KG (Silhouette) and Hartlauer Handelsgesellschaft mbH (Hartlauer).
3.
Article 7 of Directive 89/104, concerning exhaustion of the rights conferred by a trade mark, provides:
(1) The trade mark shall not entitle the proprietor to prohibit its use in relation to goods which
have been put on the market in the Community under that trade mark by the proprietor or with his
consent. (2) Paragraph 1 shall not apply where there exist legitimate reasons for the proprietor to
oppose further commercialization of the goods, especially where the condition of the goods is
changed or impaired after they have been put on the market.
4.
In accordance with art 65(2), in conjunction with Annex XVII, point 4, of the EEA agreement, art 7(1)
has been amended for the purposes of the agreement so that the expression in the Community has
been replaced by in a Contracting Party (see OJ 1994 L1 p 483).
5.
Article 7 of the directive was transposed into Austrian law by para 10a of the Markenschutzgesetz (the
Law on the Protection of Trade Marks), the first sub-paragraph of which provides:
The right conferred by the trade mark shall not entitle the proprietor to prohibit a third party from
using it in relation to goods which have been put on the market in the European Economic Area
under that trade mark by the proprietor or with his consent.
6.
Silhouette produces spectacles in the higher price ranges. It markets them worldwide under the trade
mark Silhouette, registered in Austria and most 785 countries of the world. In Austria, Silhouette itself
supplies spectacles to opticians; in other states it has subsidiary companies or distributors.
7.
Hartlauer sells inter alia spectacles through its numerous subsidiaries in Austria, and its low prices are
its chief selling point. It is not supplied by Silhouette because that company considers that distribution of
its products by Hartlauer would be harmful to its image as a manufacturer of top quality fashion
spectacles.
8.
 In October 1995 Silhouette sold 21,000 out-of-fashion spectacle frames to a Bulgarian company,
Union Trading, for the sum of $261,450. It had directed its representative to instruct the purchasers to sell
the spectacle frames in Bulgaria or the states of the former Soviet Union only, and not to export them to
other countries. The representative assured Silhouette that it had so instructed the purchaser. However,
the Oberster Gerichtshof noted that it had not proved possible to ascertain whether that had actually been
done.
9.
In November 1995 Silhouette delivered the frames in question to Union Trading in Sofia. Hartlauer
bought those goodsit has not, according to the Oberster Gerichtshof, been possible to find out from
whomand offered them for sale in Austria from December 1995. In a press campaign Hartlauer
announced that, despite not being supplied by Silhouette, it had managed to acquire 21,000 Silhouette
frames abroad.
10.
Silhouette brought an action for interim relief before the Landesgericht Steyr, seeking an injunction
restraining Hartlauer from offering spectacles or spectacle frames for sale in Austria under its trade mark,
where they had not been put on the market in the European Economic Area (the EEA) by Silhouette itself
or by third parties with its consent. It claims that it has not exhausted its trade mark rights, since, in terms
of the directive, trade mark rights are exhausted only when the products have been put on the market in
the EEA by the proprietor or with his consent. It based its claim on para 10a of the Markenschutzgesetz
and on paras 1 and 9 of the Gesetz gegen den Unlauteren Wettbewerb (the UWG) (the Law against
Unfair Competition) and para 43 of the Allgemeines Brgerliches Gesetzbuch (the ABGB) (the General
Civil Code).
11.
Hartlauer contended that the action should be dismissed since Silhouette had not sold the frames
subject to any prohibition of re-importation into the Community. In its view para 43 of the ABGB was not
applicable. Moreover, it observed that the Markenschutzgesetz does not grant a right to seek prohibitory
injunctions and that, given that the legal position was unclear, its conduct was not contrary to established
customs.
12.
Silhouettes action was dismissed by the Landesgericht Steyr and, on appeal, by the
Oberlandesgericht Linz. Silhouette appealed to the Oberster Gerichtshof on a point of law.
13.
The Gerichtshof noted, first, that the case before it concerned the re-importation of goods originally
produced by the proprietor of the trade mark and put on the market by the proprietor in a non-member
country. It went on to point out that before para 10a of the Markenschutzgesetz entered into force Austrian
courts applied the principle of international exhaustion of the right conferred by a trade mark (the principle
that the proprietors rights are exhausted once the trademarked product has been put on the market, no
matter where that takes place). Finally, the Oberster Gerichtshof stated that the explanatory
memorandum to the Austrian law implementing art 7 of the directive indicated that it was intended to
leave the resolution of the question of the validity of the principle of international exhaustion to judicial
decision.
786
14.
Accordingly, the Oberster Gerichtshof decided to stay proceedings and refer the following questions to
the Court of Justice for a preliminary ruling:
(1) Is Article 7(1) of the First Council Directive 89/104/EEC of 21 December 1988 to
approximate the laws of the Member States relating to trade marks (OJ 1989 L40 p 1) to be
interpreted as meaning that the trade mark entitles its proprietor to prohibit a third party from using
the mark for goods which have been put on the market under that mark in a State which is not a
Contracting State?
(2) May the proprietor of the trade mark on the basis of Article 7(1) of the Trade Marks Directive
alone seek an order that the third party cease using the trade mark for goods which have been put
on the market under that mark in a State which is not a Contracting State?

Question 1
15.
By its first question the Oberster Gerichtshof is in substance asking whether national rules providing
for exhaustion of trade mark rights in respect of products put on the market outside the EEA under that
mark by the proprietor or with his consent are contrary to art 7(1) of the directive.
16.
It is to be noted at the outset that art 5 of the directive defines the rights conferred by a trade mark
and art 7 contains the rule concerning exhaustion of the rights conferred by a trade mark.
17.
According to art 5(1) of the directive, the registered trade mark confers on the proprietor exclusive
rights therein. In addition, art 5(1)(a) provides that those exclusive rights entitle the proprietor to prevent
all third parties not having his consent from use in the course of trade of, inter alia, any sign identical with
the trade mark in relation to goods or services which are identical to those for which the trade mark is
registered. Article 5(3) sets out a non-exhaustive list of the kinds of practice which the proprietor is entitled
to prohibit under art 5(1), including, in particular, importing or exporting goods under the trade mark
concerned.
18.
Like the rules laid down in art 6 of the directive, which set certain limits to the effects of a trade mark,
art 7 states that, in the circumstances which it specifies, the exclusive rights conferred by the trade mark
are exhausted, with the result that the proprietor is no longer entitled to prohibit use of the mark.
Exhaustion is subject first of all to the condition that the goods have been put on the market by the
proprietor or with his consent. According to the text of the directive itself, exhaustion occurs only where
the products have been put on the market in the Community (in the EEA since the EEA agreement
entered into force).
19.
No argument has been presented to the court that the directive could be interpreted as providing for
the exhaustion of the rights conferred by a trade mark in respect of goods put on the market by the
proprietor or with his consent irrespective of where they were put on the market.
20.
On the contrary, Hartlauer and the Swedish government have maintained that the directive left the
member states free to provide in their national law for exhaustion, not only in respect of products put on
the market in the EEA but also of those put on the market in non-member countries.
21.
The interpretation of the directive proposed by Hartlauer and the Swedish government assumes,
having regard to the wording of art 7, that the directive, like the courts case law concerning arts 30 and
36 of the EC Treaty, is limited to requiring the member states to provide for exhaustion within the
Community, but that art 7 does not comprehensively resolve the question of exhaustion of 787 rights
conferred by the trade mark, thus leaving it open to the member states to adopt rules on exhaustion going
further than those explicitly laid down in art 7 of the directive.
22.
As Silhouette, the Austrian, French, German, Italian and UK governments and the European
Commission have all argued, such an interpretation is contrary to the wording of art 7 and to the scheme
and purpose of the rules of the directive concerning the rights which a trade mark confers on its
proprietor.
23.
In that respect, although the third recital in the preamble to the directive states that it does not appear
to be necessary at present to undertake full-scale approximation of the trade mark laws of the Member
States, the directive none the less provides for harmonisation in relation to substantive rules of central
importance in this sphere, that is to say, according to that same recital, the rules concerning those
provisions of national law which most directly affect the functioning of the internal market, and that that
recital does not preclude the harmonisation relating to those rules from being complete.
24.
The first recital in the preamble to the directive notes that the trade mark laws applicable in the
member states contain disparities which may impede the free movement of goods and freedom to provide
services and may distort competition within the common market, so that it is necessary, in view of the
establishment and functioning of the internal market, to approximate the laws of member states. The ninth
recital emphasises that it is fundamental, in order to facilitate the free movement of goods and services, to
ensure that registered trade marks enjoy the same protection under the legal systems of all the member
states, but that this should not prevent member states from granting at their option extensive protection to
those trade marks which have a reputation.
25.
In the light of those recitals, arts 5 to 7 of the directive must be construed as embodying a complete
harmonisation of the rules relating to the rights conferred by a trade mark. That interpretation, it may be
added, is borne out by the fact that art 5 expressly leaves it open to the member states to maintain or
introduce certain rules specifically defined by the Community legislature. Thus, in accordance with art
5(2), to which the ninth recital refers, the member states have the option to grant more extensive
protection to trade marks with a reputation.
26.
Accordingly, the directive cannot be interpreted as leaving it open to the member states to provide in
their domestic law for exhaustion of the rights conferred by a trade mark in respect of products put on the
market in non-member countries.
27.
This, moreover, is the only interpretation which is fully capable of ensuring that the purpose of the
directive is achieved, namely to safeguard the functioning of the internal market. A situation in which some
member states could provide for international exhaustion while others provided for Community exhaustion
only would inevitably give rise to barriers to the free movement of goods and the freedom to provide
services.
28.
Contrary to the arguments of the Swedish government, it is no objection to that interpretation that
since the directive was adopted on the basis of art 100a of the EC Treaty, which governs the
approximation of the laws of the member states concerning the functioning of the internal market, it
cannot regulate relations between the member states and non-member countries, with the result that art 7
is to be interpreted as meaning that the directive applies only to intra-Community relations.
 788
29.
Even if art 100a of the Treaty were to be construed in the sense argued for by the Swedish
government, the fact remains that art 7, as has been pointed out in this judgment, is not intended to
regulate relations between member states and non-member countries but to define the rights of
proprietors of trade marks in the Community.
30.
Finally, the Community authorities could always extend the exhaustion provided for by art 7 to
products put on the market in non-member countries by entering into international agreements in that
sphere, as was done in the context of the EEA agreement.
31.
In the light of the foregoing, the answer to be given to the first question must be that national rules
providing for exhaustion of trade mark rights in respect of products put on the market outside the EEA
under that mark by the proprietor or with his consent are contrary to art 7(1) of the directive, as amended
by the EEA agreement.

Question 2
32.
By its second question the Oberster Gerichtshof is in substance asking whether art 7(1) of the
directive can be construed as meaning that the proprietor of a trade mark is entitled, on the basis of that
provision alone, to obtain an order restraining a third party from using its mark for products which have
been put on the market outside the EEA under that mark by the proprietor or with his consent.
33.
In its order for reference, as clarified subsequently by letter, the Oberster Gerichtshof has pointed out:
that the second question was put because the Markenschutzgesetz does not provide for any right to
obtain a prohibitory injunction, nor does it contain any provision corresponding to art 5(1)(a) of the
directive. A prohibitory injunction may be sought in respect of a trade mark infringement only if there is at
the same time a breach of para 9 of the UWG, the application of which presupposes the risk of confusion,
which is not the case where the original products of the trade mark proprietor are concerned; in Austrian
law, at least according to current academic legal writing, the proprietor of a trade mark has no right to
obtain a prohibitory injunction against a person who makes parallel imports or re-imports of trademarked
goods, unless the right to a prohibitory injunction is already available under para 10a(1) of the
Markenschutzgesetz. The question thus arises, under Austrian law, whether art 7(1) of the trade marks
directive, which has the same content as para 10a(1) of the Markenschutzgesetz, provides for such a
right to apply for a prohibitory injunction and whether the proprietor of the trade mark can therefore seek,
solely on the basis of that provision, an order that a third party cease using the trade mark for goods
which have been put on the market under that mark outside the EEA.
34.
Under the scheme of the directive the rights conferred by a trade mark are defined by art 5, while art 7
contains an important qualification with respect to that definition, in that it provides that the rights
conferred by art 5 do not entitle the proprietor to prohibit the use of the trade mark where the conditions
laid down in that provision are satisfied.
35.
Accordingly, while it is undeniable that the directive requires member states to implement provisions
on the basis of which the proprietor of a trade mark, when his rights are infringed, must be able to obtain
an order restraining third parties from making use of his mark, that requirement is imposed, not by art 7,
but by art 5 of the directive.
789
36.
That being so, it is to be remembered, first, that, according to settled case law of the court, a directive
cannot of itself impose obligations on an individual and cannot therefore be relied upon as such against
an individual. Second, according to the same case law, when applying domestic law, whether adopted
before or after the directive, the national court that has to interpret that law must do so, as far as possible,
in the light of the wording and the purpose of the directive so as to achieve the result it has in view and
thereby comply with the third paragraph of art 189 of the Treaty (see, inter alia, the judgments in
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135
(paras 6, 8) and Faccini Dori v Recreb Srl Case C-91/92 [1995] All ER (EC) 1, [1994] ECR I-3325 (paras
20, 26)).
37.
The answer to be given to the second question must therefore be that, subject to the national courts
duty to interpret, so far as possible, domestic law in conformity with Community law, art 7(1) of the
directive cannot be interpreted as meaning that the proprietor of a trade mark is entitled, on the basis of
that provision alone, to obtain an order restraining a third party from using his trade mark for products
which have been put on the market outside the EEA under that mark by the proprietor or with his consent.

Costs
38.
The costs incurred by the Austrian, French, German, Italian, Swedish and UK governments and by the
European Commission, which have submitted observations to the Court of Justice, are not recoverable.
Since these proceedings are, for the parties to the main proceedings, a step in the action pending before
the national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Oberster
Gerichtshof by order of 15 October 1996, hereby rules: (1) National rules providing for exhaustion of trade
mark rights in respect of products put on the market outside the European Economic Area under that
mark by the proprietor or with its consent are contrary to art 7(1) of Council Directive (EEC) 89/104 to
approximate the laws of the member states relating to trade marks, as amended by Agreement on the
European Economic Area (Oporto, 2 May 1992; TS26 (1995); Cm 2847; OJ 1994 L1 p 3). (2) Article 7(1)
of Directive 89/104 cannot be interpreted as meaning that the proprietor of a trade mark is entitled, on the
basis of that provision alone, to obtain an order restraining a third party from using his trade mark for
products which have been put on the market outside the European Economic Area under that mark by the
proprietor or with his consent.

790

[1998] All ER (EC) 791

Brown v Rentokil Ltd

(Case C-394/96)

EUROPEAN COMMUNITY; Social Policy: EMPLOYMENT; Discrimination

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES GULMANN (PRESIDENT OF THE THIRD AND FIFTH CHAMBERS, ACTING AS PRESIDENT),
RAGNEMALM, WATHELET AND SCHINTGEN (PRESIDENTS OF CHAMBERS), MANCINI, KAPTEYN
(RAPPORTEUR), MURRAY, EDWARD, PUISSOCHET, JANN AND SEVN
ADVOCATE GENERAL RUIZ-JARABO COLOMER
16 DECEMBER 1997, 5 FEBRUARY, 30 JUNE 1998
European Community Equality of treatment of men and women Employment and working conditions
Employment contract making 26 consecutive weeks absence through illness a ground for dismissal
Employee suffering pregnancy-related illnesses rendering her unable to work for over 26 weeks duly
dismissed Whether dismissal constituting discrimination on grounds of sex Council Directive (EEC)
76/207, arts 2(1), 5(1).

In August 1990 the applicant informed her employer that she was pregnant. She subsequently suffered a
succession of pregnancy-related disorders and did not work again after mid-August. The applicants
contract of employment provided that employees could be dismissed if they were continuously absent
from work due to sickness for more than 26 weeks and she had not yet gained the two years continuous
employment necessary to benefit from statutory maternity leave, which allowed a pregnant employee to
absent herself from work without risk of dismissal from the beginning of the eleventh week before her
expected date of confinement. In February 1991, therefore, the applicant was duly dismissed; and her
child was born in March. The applicant brought a sex discrimination action before the Industrial Tribunal,
which held that where absence through pregnancy-related illness began long before the statutory
maternity provisions could apply and subsisted continuously thereafter, subsequent dismissal did not fall
into the category of dismissals which had to be automatically considered discriminatory because they
were due to pregnancy. On appeal from the dismissal of her appeal to the Employment Appeal Tribunal,
the Court of Session held that since there was a clear distinction between pregnancy and illness due to
pregnancy, the applicant, whose absence was due to illness and who had been dismissed on that basis,
could not succeed. On a further appeal, the House of Lords stayed the proceedings and referred to the
Court of Justice of the European Communities for a preliminary ruling the questions, inter alia, whether
arts 2(1)1 and 5(1)2 of Council Directive (EEC) 76/207 on the implementation of the principle of equal
treatment for men and women as regards access to employment, vocational training and working
conditions precluded the dismissal of a female employee at any time during her pregnancy, as a result of
absence through illness arising from that pregnancy and whether the answer was affected by a
contractual provision such as that at issue, which 791 applied irrespective of gender to employees who
were continuously for more than a specified period.
1
Article 2(1), so far as material, is set out at p 795 f, post
2
Article 5(1), so far as material, is set out at p 795 gh, post

Held The dismissal of a pregnant worker for absences due to incapacity for work resulting from her
pregnancy was linked to the occurrence of risks inherent in pregnancy and had therefore to be regarded
as essentially based on the fact of pregnancy. Such a dismissal therefore constituted direct discrimination
on grounds of sex. It followed that arts 2(1) and 5(1) of Directive 76/207 precluded the dismissal of a
female worker at any time from the beginning of her pregnancy to the end of her maternity leave for
absences due to incapacity for work caused by illness resulting from that pregnancy. Moreover, the fact
that a female worker had been dismissed on the basis of a contractual term allowing the employer to
dismiss employees after a stipulated period of continuous absence did not alter that position (see p 823 j
to p 824 b and p 825 d e, post).
 Larsson v Ftex Supermarked A/S Case C-400/95 [1997] ECR I-2757 overruled; Dekker v Stichting
Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990] ECR I-3941
applied.
Per curiam. Where pathological conditions caused by pregnancy or childbirth arise after the end of
maternity leave, they are covered by the general rules applicable in the event of illness, and any resulting
absence should be treated in the same way as a mans absence of the same duration through incapacity
for work (see p 824 c, post); Handels- og Kontor- funktionaerernes Foribund i Danmark v Dansk
Arbejdsgiverforening Case C-179/88 [1990] ECR I-3979 applied.

Notes
For equal treatment in access to employment, see 52 Halsburys Laws (4th edn) para 2113.

Cases cited
Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990]
ECR I-3941.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Gillespie v Northern Health and Social Services Board Case C-342/93 [1996] All ER (EC) 284, [1996]
ECR I-475, ECJ.
Habermann-Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb/Opf-eV Case C-421/92 [1994] ECR I-
1657.
Handels- og Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiver- forening Case C-179/88
[1990] ECR I-3979.
Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047.
Larsson v Ftex Supermarked A/S Case C-400/95 [1997] ECR I-2757.
Lavery v Plessey Telecommunications Ltd [1982] ICR 373, [1982] IRLR 180.
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135.
Wagner Miret v Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911.
Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115, [1994] QB 718, [1994] 1 WLR 941,
[1994] ECR I-3567, ECJ.
792

Reference
By order of 28 November 1996, the House of Lords referred to the Court of Justice of the European
Communities for a preliminary ruling under art 177 of the EC Treaty two questions (set out at p 821 j to p
822 b, post) on the interpretation of arts 2(1) and 5(1) of Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to employment,
vocational training and promotion, and working conditions. Those questions arose in proceedings brought
by Mary Brown against Rentokil Ltd in connection with her dismissal whilst pregnant. Written observations
were submitted on behalf of: Mrs Brown, by C McEachran QC, and I Truscott, Advocate, instructed by
Mackay Simon, Solicitors; Rentokil, by J Hand QC, and Gerard F McDermott, Barrister, instructed by G T
Brown, Solicitor; the United Kingdom government, by S Ridley, of the Treasury Solicitors Department,
acting as agent, and D Rose, Barrister; and the European Commission, by P J Kuyper, Legal Adviser, and
M Wolfcarius, of its Legal Service, acting as agents. Oral observations were made by Mrs Brown,
Rentokil, the UK government and the Commission. The language of the case was English. The facts are
set out in the opinion of the Advocate General.

5 February 1998.

The Advocate General (D Ruiz-Jarabo Colomer)

delivered the following opinion (translated from the Spanish).
1.
Is it contrary to Community law, and specifically to Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to employment,
vocational training and promotion, and working conditions (OJ 1976 L39 p 40), for an employer to dismiss
a pregnant worker on account of incapacity for work caused by her pregnancy, the period of such
incapacity having exceeded that which, under the contract of employment, entails dismissal of workers on
grounds of sickness? That, in general terms, is the question on which the House of Lords seeks a
preliminary ruling in this case.

ITHE DISPUTE BEFORE THE NATIONAL COURTS

2.
The facts, as described in the order for reference, are as follows: Mrs Brown, the appellant in the main
proceedings, was employed as a driver for Rentokil Ltd, which has now become Rentokil Initial UK Ltd,
the respondent in those proceedings. Her job was mainly to transport and change Sanitact units in shops
and other centres. She became pregnant and informed Rentokil in August 1990.
Her pregnancy later became complicated through a number of interrelated causes, of which details are
not given in the documents before the Court of Justice of the European Communities. As from 16 August
1990 she submitted a succession of four-week medical certificates mentioning various diagnoses such as
symptoms of pregnancy, bleeding in pregnancy or pregnant backache. From that time, until her
dismissal, the appellant remained unable to work.
3.
The respondent included a clause in the contract of employment under which any employee, man or
women, who was incapable of work for more than 26 weeks without interruption would be dismissed. On
9 November 1990 Mrs Brown had a meeting with two executives of the company, who informed her that
half of the 26-week period had passed and reminded her that her 793 employment contract would be
terminated on 8 February 1991 if she was not back at work by that time, following an independent medical
examination confirming that she was able to work. Those details were confirmed to her by letter of the
same date.
The appellant did not go back to work after receiving that letter. The parties agree that there was never
any question of her being able to return to work prior to the expiry of the 26-week period.
4.
In accordance with the above-mentioned clause, the appellant was dismissed while she was still
pregnant, by letter dated 30 January 1991 with effect from 8 February 1991. She gave birth on 22 March
1991.
5.
The national court states in its order for reference that, on the basis that 22 March 1991 was also the
expected date of confinement, then, had the appellant been employed for two years on 30 December
1990, she would have been entitled to absent herself from work from the beginning of the eleventh week
before confinement and would have had the right to return to work at any time in the 29 weeks following
her confinement (pursuant to s 33 of the Employment Protection (Consolidation) Act 1978). Owing to the
appellants length of service, she was not so entitled, although, if she had not been dismissed, she would
have been entitled to maternity pay from her employer.
6.
The Industrial Tribunal dismissed the proceedings brought by the appellant, in which she alleged
discrimination on grounds of sex, stating:
It is plain that where as in this case an absence through illness related to reasons of pregnancy
but beginning long before the applicability of the statutory maternity provisions and subsisting
continuously thereafter followed by dismissal does not fall to be put into the automatic category of
being discriminatory because the dismissal was due to pregnancy.
7.
The Employment Appeal Tribunal dismissed Mrs Browns appeal ([1992] IRLR 302). That tribunal
considered itself bound by the decision of the Court of Appeal in Webb v EMO Air Cargo (UK) Ltd Case
C-32/93 [1994] 4 All ER 115, [1994] ECR I-3567 and held that the industrial tribunal had, as the law stood
at that time, reached the correct conclusion.
8.
Lord Allanbridge, giving the judgment of the Extra Division of the Court of Session ([1995] IRLR 211),
held that a preliminary conclusion was that in Mrs Browns case there was no discrimination under the
Sex Discrimination Act 1975. He further held that the clear distinction established by the Court of Justice
in Handels- og Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiverforening Case C-179/88
[1990] ECR I-3979 (the Hertz case) between pregnancy and illness attributable to pregnancy was
applicable to Mrs Browns appeal. Therefore, the appellant, whose absence was due to illness and who
was dismissed on account of that illness in February 1991, could not succeed on the facts relevant to her
case.

IITHE PRELIMINARY QUESTIONS

9.
In order to determine the dispute, which came before it by way of appeal from the judgment of 18
January 1995, the House of Lords, after hearing submissions from the parties, referred the following
questions to the Court of Justice for a preliminary ruling:
(1)(a) Is it contrary to Articles 2(1) and 5(1) of Directive 76/207 of the Council of the European
Communities (The Equal Treatment Directive) to dismiss a female employee at any time during
her 794 pregnancy as a result of absence through illness arising from that pregnancy? (b) Does it
make any difference to the answer given to Question 1(a) that the employee was dismissed in
pursuance of a contractual provision entitling the employer to dismiss employees, irrespective of
gender, after a stipulated number of weeks of continuous absence?
(2)(a) Is it contrary to Articles 2(1) and 5(1) of the Equal Treatment Directive to dismiss a female
employee as a result of absence through illness arising from pregnancy who does not qualify for
the right to absent herself from work on account of pregnancy or childbirth for the period specified
by national law because she has not been employed for the period imposed by national law, where
dismissal takes place during that period? (b) Does it make any difference to the answer given to
Question 2(a) that the employee was dismissed in pursuance of a contractual provision entitling the
 employer to dismiss employees, irrespective of gender, after a stipulated number of weeks of
continued absence?

IIITHE COMMUNITY LEGISLATION

The Community provisions of which an interpretation is needed for judgment to be given in this dispute
are all contained in Directive 76/207, namely art 2(1) and (3) and art 5(1) and (2), which provide as
follows:
Article 2
(1) For the purposes of the following provisions, the principle of equal treatment shall mean that
there shall be no discrimination whatsoever on grounds of sex either directly or indirectly by
reference in particular to marital or family status (3) This Directive shall be without prejudice to
provisions concerning the protection of women, particularly as regards pregnancy and maternity
Article 5
(1) Application of the principle of equal treatment with regard to working conditions, including the
conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex. (2) To this end, Member States shall take the
measures necessary to ensure that: (a) any laws, regulations and administrative provisions
contrary to the principle of equal treatment shall be abolished; (b) any provisions contrary to the
principle of equal treatment which are included in collective agreements, individual contracts of
employment, internal rules of undertakings or in rules governing the independent occupations and
professions shall be, or may be declared, null and void or may be amended
10.
On 19 October 1992 the EC Council adopted Council Directive (EEC) 92/85 on the introduction of
measures to encourage improvements in the safety and health at work of pregnant workers and workers
who have recently given birth or are breastfeeding (tenth individual directive within the meaning of art
16(1) of Directive 89/391) (OJ 1992 L348 p 1), in order to protect the health and safety of pregnant
workers who have given birth or are breastfeeding, which requires the member states to adopt, before 19
October 7951994, among others provisions needed to ensure that female workers enjoy a continuous
period of maternity leave of at least 14 weeks, allocated before and/or after confinement, including two
weeks compulsory leave. It also prohibits dismissal of a pregnant worker, in the following terms:
Article 10
(1) Member States shall take the necessary measures to prohibit the dismissal of workers
[who are pregnant, have recently given birth or are breastfeeding] during the period from the
beginning of their pregnancy to the end of the maternity leave save in exceptional circumstances
not connected with their condition which are permitted under national legislation and/or practice
and, where applicable, provided that the competent authority has given its consent
11.
However, since Mrs Brown was dismissed at the beginning of 1991, there is no need to interpret those
provisions.

IVTHE NATIONAL LEGISLATION

12.
The national provisions applicable to the main proceedings are contained in the Sex Discrimination Act
1975, the Employment Protection (Consolidation) Act 1978 and the Social Security Act 1986.
Sections 1 and 5 of the 1975 Act provide that a woman is discriminated against on grounds of sex if,
for the purposes of that act, she receives less favourable treatment than a man. A comparison must be
made between the cases of persons of different sex or marital status such that the circumstances in the
one case are the same, or not materially different, in the other.
As regards rights of employees in connection with pregnancy and maternity, the 1978 Act provided, at
the time of Mrs Browns dismissal, that an employee who is absent from work because of pregnancy or
confinement will be entitled to return to work, provided that the following conditions are met: she
continues to be employed, whether or not at work, at the beginning of the eleventh week before the
expected date of confinement; at the beginning of that eleventh week she has been employed for a period
of not less than two years; she informs her employer in writing at least 21 days before stopping work, that
her absence will be due to pregnancy or confinement, that she intends to return to work, of the expected
week of confinement or, if the confinement has occurred, the date on which she gave birth; she submits,
at the employers request, a medical certificate indicating the expected week of confinement.
As regards dismissal, the 1978 Act provided at the material time that an employee was to be treated
as unfairly dismissed if the direct or indirect reason for her dismissal was that she was pregnant, unless,
on the date of the dismissal, she was incapable, because of her pregnancy, of adequately performing her
work or if, because of her pregnancy, she could not continue to work after that date without contravention,
either by her or her employer, of a duty or restriction imposed by or under any enactment. The national
court makes it clear that, at the material time, Mrs Brown would have had to have completed two years
continuous employment in order to avail herself of the right not to be unfairly dismissed, a requirement
which she did not meet.
796
The subsequent amendment of the provisions of that Act which govern unfair dismissal, made in order
to incorporate the provisions of Directive 92/85, entered into force on 10 June 1994.
 Finally, so far as is relevant here, the 1986 Act provides that a female employee will be entitled to
maternity pay from her employer if she meets the following conditions: she has been in the employers
service for a continuous period of at least 26 weeks, ending with the week immediately preceding the
fourteenth week before the expected date of confinement, but has ceased to work wholly or partly
because of pregnancy or confinement; she has received weekly earnings for the period of eight weeks
ending with the week immediately preceding the fourteenth week before the expected date of
confinement which are not less than the lower earnings limit in force before the commencement of that
fourteenth week; she has reached the beginning of the eleventh week prior to the expected date of
confinement or has given birth before that time; and she has notified her employer that she is ceasing
work because of her pregnancy or confinement.
Maternity pay will be payable for a maximum period of 18 weeks, normally commencing in the
eleventh week preceding the expected confinement date, but no later than the sixth week. Provision is
made for two types of maternity pay, referred to as the higher rate and the lower rate. The higher rate is
equivalent to nine-tenths of a womans normal weekly earnings for the period of eight weeks preceding
the fourteenth week before the expected week of confinement; the lower rate, the amount of which is
fixed, applies when it is found to exceed the higher rate. A woman who, for a continuous period of at least
two years ending at the beginning of the fourteenth week prior to the expected confinement date, has
worked for an employer who is liable to pay the allowance to her will receive it at the higher rate for the
first six weeks and at the lower rate for the remaining period. A woman working for her employer under an
employment contract involving employment for less than 16 hours weekly will not be entitled to receive
the allowance at the higher rate. A woman entitled to maternity pay who is unable to claim the higher rate
will receive the lower rate.
According to the documents before the Court of Justice, in 1996 the lower rate of maternity pay was, in
1996, 5455 per week3.
3
See Social Security Benefits Up-rating Order 1996, SI 1996/599, arts 1(2)(c) and 10, as from 7 April 1996. The previous
amount was 5250 (see Social Security Benefits Up-rating Order 1995, SI 1995/559, art 10). There is no indication of
the amount payable at the time of the events in the main proceedings.

VTHE OBSERVATIONS SUBMITTED TO THE COURT OF JUSTICE

13.
Written observations were submitted within the period laid down by art 20 of the EC Statute of the
Court of Justice, and oral argument was presented at the hearing by the appellant, the respondent, the
UK government and the European Commission.
14.
The appellant, on the basis of the case law of this court, which I shall examine in due course,
considers that the dismissal of a pregnant woman on account of incapacity for work attributable to
pregnancy is contrary to Directive 76/207 and constitutes direct discrimination on grounds of sex, since
pregnancy is a condition which affects only women. She considers that it is also contrary to that directive
to dismiss a pregnant worker on account of incapacity for work attributable to her condition, when she is
not entitled to 797 absent herself from work on account of pregnancy or childbirth for the period specified
by national law because of insufficient length of service, if the dismissal takes place during that period.
Finally, she maintains that the fact that the dismissal occurred under a contractual rule enabling the
employer to dismiss a worker, regardless of sex, for continuous absence for a given number of weeks has
no bearing on that position.
15.
The respondent states that it is necessary to go back to a date falling eleven weeks before the
expected date of confinement in order to arrive at the date on which Mrs Browns employment rights
should be considered. Since on 30 December 1990 she had not worked for her employer for a continuous
period of two years she was not entitled to the benefit of the right to return to work provisions, nor could
she rely on the provisions concerning unfair dismissal intended to protect pregnant women against
dismissal on account of their pregnancy. She was, however, entitled to maternity pay at the lower rate,
payable by the state.
It submits that, in view of the circumstances in which Mrs Brown found herself, the reason for her
being dismissed was not that she became pregnant but rather that her pregnancy was abnormal,
preventing her from working from about the eighth week. The duration of her incapacity for work was
longer than might normally be anticipated, falling entirely outside the period generally provided for on
account of maternity. The reason for Mrs Browns dismissal was that she became ill during her pregnancy
and her unavailability for work exceeded 26 weeks.
On the basis of the same case law as that relied on by the appellant to reach the opposite conclusion,
it submits that the rule in the employment contracts concluded with its employees, which allowed Mrs
Brown to be dismissed after 26 weeks absence, is not discriminatory. In support of that view, it gives the
example of an employer who employs three pregnant women. The first, who is robust, has no time off
until she is able to absent herself from work by reason of maternity; the second has some time off due to
pregnancy-related health problems before the beginning of her absence; the third is Mrs Browns case:
her pregnancy-related incapacity for work exceeded 26 weeks and she was dismissed on the same terms
as a male employee after absence through illness for precisely the same period. Why should such an
approach be described as sex discrimination? On the contrary, if Mrs Brown were treated in a manner
different from other pregnant women or male employees, that would be equivalent to introducing positive
discrimination applicable only to a section of pregnant employees rather than in all cases of pregnant
employees.
 It maintains that a distinction must be drawn between an illness like that suffered by Mrs Brown, which
led to her being unfit for work practically throughout the whole of her pregnancy, and the normal risks of
pregnancy, such as morning sickness at the start of the pregnancy and tiredness towards the end,
problems which, moreover, do not arise in all pregnancies. The answer to be given to the House of Lords
must, in its opinion, be based on that distinction.
As regards the fact that Mrs Brown was dismissed when, if her length of service had been sufficient,
she would have been entitled to absent herself from work on account of pregnancy, the respondent
considers that, as the applicable Community law stood at that time, it was exclusively for the member
states to determine the conditions under which a woman would be entitled to such right of absence and to
set the duration thereof. Exercising 798 that prerogative, the United Kingdom imposed the legal
requirement of two years service. In any event, Mrs Brown was dismissed for being absent on sick leave
for the number of weeks laid down in her contract of employment, and not for her absences in the short
period for which, had she been entitled to absent herself from work on account of pregnancy, she could
have done so.
16.
The UK government stresses that the period for which Mrs Brown was unfit for work, as a result of
pregnancy, far exceeded the period for which a woman may absent herself on account of pregnancy and
confinement under national law. It states that, in order to answer the preliminary questions, it is necessary
to distinguish dismissal arising out of normal risks and disorders normally inherent in pregnancy and
maternity, for example absence from work for routine ante-natal appointments, minor absences for
morning sickness and absences during the period immediately preceding and following her confinement,
from dismissal of a woman arising out of sickness, whether pregnancy-related or not. In the first case, the
dismissal will constitute direct sex discrimination, whereas in the second case it will be lawful, provided
that a male employee who was unfit for work for a similar period would have been treated no more
favourably. In its view, it is for the national court to determine whether the period for which Mrs Brown was
absent was justified by a normal risk of pregnancy.
17.
The Commission states that there is no objective justification for making a distinction during
pregnancy, in order to apply different treatment to them, between the period before the worker can
exercise the right to absent herself from work and the period from that time until the birth. Nor does it
consider that there is any justification for distinguishing between dismissal for reasons of pregnancy and
dismissal for reasons of pregnancy-related incapacity for work, provided that dismissal takes place when
the worker is still pregnant, since health problems associated with pregnancy are due to a specifically
female condition both before and during her absence. The application to women in Mrs Browns situation
of the rule applicable to workers in general, in case of illness, constitutes discrimination against women
since the situations involved are different.
It adds that that view is confirmed by art 10 of Directive 92/85, which prohibits dismissal of workers in
the period from the beginning of their pregnancy to the end of their maternity leave. Despite the fact that
the period within which the member states were to incorporate the directive in their national law had not
expired at the material time, its provisions must be taken into account in interpreting Directive 76/207.

VIEXAMINATION OF THE PRELIMINARY QUESTIONS

A. Opening remarks
18.
Before considering the problems raised by the questions, I wish to make a number of observations.
19.
The first concerns the order for reference, which is particularly laconic in its account of the facts of the
main proceedings. There is no indication, for example, of how long Mrs Brown had been pregnant when
she became unfit for work or of the expected date of confinement or whether her pregnancy was in any
way subject to high risks. I also consider that the reasons which, it appears, were given in the successive
medical certificates, namely symptoms of pregnancy, bleeding in pregnancy and pregnant backache, if
they were 799 the only ones, seem, especially in the case of the first and third, not to constitute very
cogent reasons for a doctor to have repeatedly certified unfitness for work for a continuous period of 26
weeks. Nor are we told whether Mrs Brown underwent the independent medical examination scheduled
for the end of the 26 weeks or, if so, what the result was.
I imagine that the purpose of that examination was for the company to obtain an independent opinion
as to the workers chances of recovery, with a view to her returning to work sooner or later, dismissal
being the appropriate course if the result of the examination suggested that she would not return to work
in the short or medium term. If that were not the case, the dismissal would appear to be a disciplinary
dismissal on grounds of illness rather than dismissal justified by a situation in which the employer had to
bear an excessive burden.
If Mrs Browns unfitness for work was actually caused by her pregnancy, and there is no reason to
think that the position was otherwise, it would seem logical to assume that she would cease to be unfit
when she gave birth. Why, therefore, was she dismissed one-and-a-half months before she gave birth?
Simply because she had by then been away for 26 weeks, or was it rather because national law gave her
no right to be re-instated after her confinement because her length of service with the company was less
than two years?
Those are, of course, matters of fact to be assessed by the national court in the light of the
interpretation of Community law given by the Court of Justice. I must point out, however, that it would
have been useful for those facts to have been disclosed.
20.
The second observation relates to the need to define, once and for all, the respective fields of
application of Directives 76/207 and 92/85, in the light of their provisions and their purposes 4. The
respondent states that the repercussions of the judgment to be given in this case will be very limited since
Directive 92/85 requires the member states to prohibit, with effect from no later than 19 October 1994,
dismissal of a worker who is pregnant, has recently given birth or is breastfeeding. In its opinion, Mrs
Brown was merely unlucky since, when she was dismissed, that directive had not been adopted.
4
That need was pointed out when Directive 92/85 was merely planned, by Shaw Pregnancy discrimination in sex
discrimination (1991) ELRev 313320 esp at 318.

I cannot agree with that view, since the provisions of the two directives differ and the aims which they
pursue are also different.
21.
Directive 76/207, whose legal basis is art 235 of the EC Treaty, represents Community action
considered necessary, at the relevant time, in order to achieve the principle of equal treatment for men
and women in respect of access to employment, vocational training and promotion and in respect of other
working conditions (third recital in the preamble). The only exceptions allowed to the principle of equal
treatment are to be found in art 2(2), (3) and (4), specifically relating to occupational activities for which,
by reason of their nature, the sex of the worker constitutes a determining factor; provisions relating to
protection of women, particularly in relation to pregnancy and maternity; and measures to promote equal
opportunity for men and women, in particular by removing any existing inequalities which affect womens
opportunities in the areas covered by the directive. In all cases, the measures involved, as Community
law stands at present, are a matter for the member states.
800
22.
Directive 92/85, on the other hand, has as its legal basis art 118a of the Treaty, which requires the
Council to adopt, by means of directives, minimum provisions to encourage improvements, especially in
the working environment, as regards the health and safety of workers; it is the tenth individual directive
adopted under art 16(1) of Council Directive (EEC) 89/91 on the introduction of measures to encourage
improvements in the safety and health of workers at work (OJ 1989 L183 p 1), art 15 of which provides
that particularly sensitive risk groups must be protected against the dangers which specifically affect
them.
The purpose of Directive 92/85 is to apply measures to encourage improvements to health and safety
at work for workers who are pregnant, have recently given birth or are breastfeeding. It is thus clearly
concerned with protection, providing for different treatment for women in any of those situations. The
legislature itself recognises, in the ninth recital in the preamble, that such protection should not treat
women on the labour market unfavourably or work to the detriment of directives concerning equal
treatment for men and women.
As the Court of Justice stated in its judgment in Webb [1994] 4 All ER 115, [1994] ECR I-3567 (paras
2122), in view of the effects which the risks of dismissal may have on the physical and mental state of
women who are pregnant, have recently given birth or are breastfeeding, including the particularly serious
risk that pregnant women may be prompted voluntarily to terminate their pregnancy, the Community
legislature provided, pursuant to art 10 of Directive 92/85, for special protection to be given to women by
prohibiting dismissal during the period from the beginning of their pregnancy to the end of their maternity
leave, and that there is to be no exception to that prohibition during that period save in exceptional cases
not connected with the condition of the person concerned.
23.
It is true that, once the prohibition of dismissal laid down in art 10 of Directive 92/85 is incorporated in
the domestic law of the member states, it will become unnecessary to apply art 5 of Directive 76/207,
which lays down the principle of equal treatment for men and women with regard to working conditions,
including the conditions governing dismissal, to cases in which a woman has been dismissed whilst
pregnant.
However, the prohibition of dismissal in art 10 of Directive 92/85 does not resolve the problem of a
woman who, having returned to work on an entirely normal basis after her maternity leave, is then
dismissed for having been intermittently unfit for work during the previous year for the number of weeks
laid down by national legislation; in such circumstances, is the woman dismissed on the same terms as a
man who has been ill for the same number of weeks, when for the female worker account is taken of the
period for which she was confined to bed on doctors orders because, for example, her pregnancy
displayed a risk of miscarriage?
That is why art 5 of Directive 76/207 and the courts case law interpreting it will continue to be
essential in order to clarify the question whether the periods for which a pregnant woman has been
prevented from working because of her pregnancy may be added to the periods preceding her pregnancy
and following her maternity leave, for the purpose of calculating absences from work such as to justify
dismissal.
24.
My third observation relates to the case law of this court interpreting the principle of equal treatment
laid down by Directive 76/207, where the 801 employment rights of workers who are pregnant or have
recently given birth are involved. It might not at first sight seem extraordinary that the only four judgments
on this subject delivered before the expiry date of the period for submitting written observations in this
case, namely the judgments in Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-
Centrum) Plus Case C-177/88 [1990] ECR I-3941, the Hertz case, Habermann-Beltermann v
Arbeiterwohlfahrt, Bezirksverband Ndb/Opf-eV Case C-421/92 [1994] ECR I-1657 and the Webb case,
enabled the parties to the main proceedings, the UK government and the Commission to defend such
conflicting positions. What is surprising, and indeed worrying, is that all rely on the same paragraphs in
support of their views.
I believe therefore that, as well as answering the preliminary questions from the House of Lords, the
court must clarify the existing case law, in order to ensure legal certainty. From my part, I shall review
those judgments and suggest how I think they should be interpreted.

B. The case law of the Court of Justice on the application of employment rights governed by
Directive 76/207 to workers who are pregnant or who have recently given birth
(1) The judgments in the Dekker and Hertz cases, in which the court held that it was
discriminatory to refuse to engage a pregnant woman but not to dismiss her because of
incapacity for work commencing after maternity leave, even though the illness was caused by the
confinement
25.
On 8 November 1990, this court delivered two judgments which have had a considerable impact in the
area of Community social law concerned with application of the principle of equal treatment for men and
women in matters of employment. They were the first in which the interpretation of Directive 76/207 was
in issue and, more specifically, the right of access to employment or the right to maintain a post in relation
to situations as exclusive to women as pregnancy and maternity. Both illustrate well the thesis of Lucinda
M Finley, which can be summarised in the following statement: The fact that women bear children and
men do not has been the major impediment to women becoming fully integrated into the public world of
the workplace (see Finley Transcending Equality Theory: A Way out of the Maternity and the Workplace
Debate 86 Columbia Law Review 1118 at 1119).
26.
They are the judgments in the Dekker and Hertz cases. In the first, the court gave a preliminary ruling
on a question from the Hoge Raad der Nederlanden (the Supreme Court of the Netherlands), dealing with
a case in which Mrs Dekker, a candidate for a post of instructor in a training centre for young adults, who
had informed the selection committee that she was three-months pregnant and who had been
recommended by the committee as the most suitable candidate for the duties involved, objected to the
employers refusal to engage her on the ground that she was already pregnant when she submitted her
application. That refusal came after the employer established that its insurer would not reimburse the
daily benefit that it would have to pay the worker during her absence on account of pregnancy and
confinement. The applicable national law allowed the insurer to refuse wholly or in part to refund daily
benefits to the employer where the employees incapacity for 802 work had arisen within the six months
following recruitment, provided that, at that time, the workers state of health was such that the
emergence of incapacity within that period could be foreseen. Since no exception was made for
pregnancy, incapacity for work by reason of maternity was treated in the same way as incapacity for work
because of a foreseeable sickness.
The national courts question, so far as relevant to this case, was whether the employers refusal to
engage the worker in those circumstances was contrary to the principle of equal treatment laid down in
Directive 76/207.
27.
In the second case, the preliminary ruling was sought by the Hjesteret (the Supreme Court),
Denmark, to enable it to give judgment in the proceedings before it, brought by Mrs Hertz against her
former employer, Aldi-Marked K/S. Mrs Hertz had been recruited in July 1982 as a part-time cashier and
saleswoman. In June 1983, after a pregnancy marked by complications, for most of which, with the
consent of her employer, she was on sick leave, she gave birth to a son. Towards the end of 1983, at the
end of her maternity leave, Mrs Hertz returned to work and did not take any more sick leave until June
1984. In the course of the next year she was on sick leave for 100 working days, for which reason she
was dismissed: the company claimed that it was normal practice to dismiss workers who were ill
frequently. It was common ground between the parties that the plaintiffs absences between June 1984
and June 1985 were a consequence of her confinement in 1983.
The national court asked whether art 5(1), in conjunction with art 2(1), of Directive 76/207 covered
dismissals as a consequence of absences due to illness attributable to pregnancy or confinement and, if
so, whether protection against dismissal due to illness caused by pregnancy or confinement was
unlimited in time.
28.
Advocate General Darmon, who delivered an opinion covering both cases, carried out a general and
invaluable review of the question of motherhood and the place which it should occupy, in the light of the
principle of equal treatment for men and women, in the economic and social life of European societies,
which is still up to date today (see the joint opinion of Advocate General Darmon in the Hertz and Dekker
cases: [1990] ECR I-3956 (esp para 21)). After considering whether there was any event more closely
connected to the specific nature of womanhood, and whether it was conceivable to treat female workers
on an equal footing with their male counterparts without taking account of motherhood, he took the view,
in the Dekker case, that the refusal to employ the candidate on account of forthcoming motherhood,
thereby taking into consideration an event which affected only female workers, constituted direct
discrimination on grounds of sex.
29.
 The Hertz case, he said, involved in a perhaps more pressing way the difficult task of reconciling the
principle of equal treatment with the demands of economic life. And he added ([1990] ECR I-3956 (para
40)):
How are periods of sickness occurring after maternity leave but directly attributable to
pregnancy and confinement to be viewed? Should it or should it not be subject to what might be
termed the ordinary rules governing absences for reasons of health?
The conclusion which he reaches in that case is the opposite of that in Dekker, namely that the dismissal
of a worker outside periods of maternity leave 803 because of absences due to illness attributable to
pregnancy or confinement does not constitute discrimination directly based on sex.
30.
I must make it clear, however, that that bare statement is the consequence of reasoning which
incorporates throughout the idea that the workers absences were justified by illness originating from
pregnancy or confinement, which occur after the expiry of the maternity leave. In fact he says ([1990]
ECR I-3956 (para 43)):
I have to confess that I was tempted to propose a solution whereby medical conditions which
were directly, definitely and preponderantly due to pregnancy or confinement would enjoy
immunity, in the sense that the principle of equality of treatment would restrain the employer from
dismissing his employee for a reasonable period after the event in question. (My emphasis.)
In para 45, he says If complications resulting from a confinement are severe, a female worker may
remain unable to work for several years, without her employers being able to dismiss her. In para 46, he
explains that an expedient protecting a few women affected by severe post-natal problemsin statistical
terms, fortunately a minute percentage of casesmay jeopardize the chances of all women wishing to
enter the labour market (my emphasis).
In the next paragraph, when he considers what criteria the court could adopt in order to decide in
which cases a medical condition must fall within the protection demanded by maternity and what the
duration of such protection should be, the underlying idea continues to be that the principle of equal
treatment applies once the woman has come to the end of her maternity leave.
And, in para 48, the solution he suggests is to draw a distinction between the normal risks of
pregnancy and confinement, which he defines as the common attendant complications sometimes
leading to the grant of additional maternity leaverisks which, in his opinion, should qualify for
Community protection inasmuch as they are specific to motherhoodand medical conditions which are
not associated with the ordinary risks of pregnancy and should therefore be treated on the same footing
as ordinary sickness, adding that
in the absence of national legal provisions conferring special protection on women, the
employer must be able to dismiss his employee at the end of maternity leave Thus, once a
female worker has exhausted her entitlement to the various types of maternity leave, her periods of
absence for reasons of sickness, even if those reasons can be traced back to pregnancy or
confinement, cannot be attributed to the normal risks of maternity and must accordingly be viewed
in the same light as the absences of any other worker, unless the national legislature provides
special protection pursuant to Article 2(3) of the Directive. (My emphasis.)
31.
Both judgments were delivered by the full court, which adopted the solution proposed by the Advocate
General. The hearings were held on 3 October 1989 and the opinion was delivered on 14 November of
the same year. The fact that the judgments came a year later is indicative, in my opinion, of the difficulties
involved in those cases, in particular the Hertz case, as the court itself acknowledges, which may account
for the sparse nature of the reasoning of both judgments (see [1990] ECR I-3979 (para 7)).
804
32.
In the judgment in Dekker [1990] ECR I-3941 (para 21), the court took the view that only women can
be refused employment on account of pregnancy and such refusal therefore constitutes direct
discrimination on grounds of sex, that a refusal of employment on account of the financial consequences
of absence due to pregnancy is essentially based on the fact of pregnancy, and that such discrimination
cannot be justified on grounds relating to the financial loss which an employer who appointed a pregnant
woman would suffer for the duration of her maternity leave.
33.
The grounds and operative part of the judgment in Hertz, however, deserve special attention. It should
be remembered that the positions of the parties were that some claimed that the dismissal of a woman on
account of pregnancy, confinement or repeated periods of absence due to an illness attributable to
pregnancy or confinement is, irrespective of the time when the illness occurs, contrary to the principle of
equal treatment, since a male worker is not subject to such disorders and hence cannot be dismissed on
that ground. The others maintained that an employer cannot be prohibited from dismissing a female
worker on account of her frequent periods of sick leave solely because her illness is attributable to
pregnancy or confinement. Such a prohibition, which would affect the employer for many years after the
confinement, would be liable to entail not only administrative difficulties and unfair consequences for
employers but also negative repercussions on the employment of women (see [1990] ECR I-3979 (paras
89)).
34.
 That judgment is of fundamental importance when it comes to reconciling the principle of equal
treatment for men and women in matters of employment with the role of women in the reproductive
process. Indeed, it states ([1990] ECR I-3979 (paras 1315)):
(13) It follows from the provisions of the Directive that the dismissal of a female worker on
account of pregnancy constitutes direct discrimination on grounds of sex, as is a refusal to appoint
a pregnant woman
(14) On the other hand, the dismissal of a female worker on account of repeated periods of sick
leave which are not attributable to pregnancy or confinement does not constitute direct
discrimination on grounds of sex, inasmuch as such periods of sick leave lead to the dismissal of a
male worker in the same circumstances.
(15) The Directive does not envisage the case of an illness attributable to pregnancy or
confinement. It does, however, admit of national provisions guaranteeing womens specific rights
on account of pregnancy and maternity, such as maternity leave. During the maternity leave
accorded to her pursuant to national law, a woman is accordingly protected against dismissal due
to absence. It is for every Member State to fix periods of maternity leave in such a way as to enable
female workers to absent themselves during the period in which the disorders inherent in
pregnancy and confinement occur.
35.
It is paras 16 and 17, which, in my opinion, indicate the direction to be taken in interpreting that
judgment. In them, the court expressly states that:
(16) In the case of an illness manifesting itself after the maternity leave, there is no reason to
distinguish an illness attributable to pregnancy or confinement from any other illness. Such a
pathological condition is therefore covered by the general rules applicable in the event of illness.
805
(17) Male and female workers are equally exposed to illness. Although certain disorders are, it is
true, specific to one or other sex, the only question is whether a woman is dismissed on account of
absence due to illness in the same circumstances as a man; if that is the case, then there is no
direct discrimination on grounds of sex.
36.
However, the difficulty which arises in interpreting the scope of that judgment lies in the fact that the
clarification concerning the time at which the illness emerges does not appear in the operative part, in
which an answer is given that corresponds precisely to the question, as submittedand the court adopts
the exact wording thereofin the following terms ([1990] ECR I-3979 (operative part)):
Without prejudice to the provisions of national law adopted pursuant to Article 2(3) of Council
Directive 76/207 Article 5(1) of that directive, in conjunction with Article 2(1) thereof, does not
preclude dismissals which are the result of absences due to an illness attributable to pregnancy or
confinement.
That apparent incongruity, which is easily accounted for by the fact that, in paras 16 and 17, the court
is very mindful of the circumstances in which the dispute before the national court arose, whereas, in the
operative part, the answer strictly matches the preliminary question, as phrased by the national court 5, is
what has given rise to the divergent interpretations underlying the views of the parties in these
proceedings.
5
The question is worded as follows: Do the provisions of Article 5(1), in conjunction with Article 2(1), of Directive
76/207 encompass dismissal as a consequence of absence due to illness which is attributable to pregnancy or
confinement?

37.
I am in no doubt whatsoever that that judgment cannot be interpreted solely by reference to the
operative part: it must be borne in mind that, in para 16, the court states in absolutely clear terms that, if
the illness manifests itself after the maternity leave, there is no reason to distinguish an illness attributable
to pregnancy or confinement from any other illness.
That is certainly the construction adopted by academic authors commenting on that judgment: all the
commentaries I have read stress, more or less disapprovingly, the fact that the court, when stating that art
5(1), in conjunction with art 2(1), of Directive 76/207 does not preclude dismissals which are the result of
absences due to an illness attributable to pregnancy or confinement, is referring to absences occurring
after maternity leave6.
6
See Jaeqmain Chroniques de droit social (1991) pp 4950 (para 4); Devos (1991) Journal des tribunaux du travail 121
122 (para 3); Shaw Pregnancy discrimination in sex discrimination (1991) ELRev 313 at 317; Traversa (1991) Revue
trimestrielle de droit europen 425 at 436; Kilpatrick How long is a piece of string? Regulation of the Post-Birth Period
in Sex Equality Law in the European Union (1996) pp 81ff, esp at p 84; Burrows & Mair European Social Law (1996) p
155; Nielsen (1992) CMLRev 160 at 164; More Reflections on pregnancy discrimination under European Community
law (1992) The Journal of Social Welfare and Family Law 48 at 5354; Bolger Discrimination on Grounds of Pregnancy
as Sex Discrimination in Gazette of the Incorporated Law Society of Ireland p 383; Rodrguez-Piero Discriminacin
por razn de sexo y embarazo de la trabajadora (1991) I Relaciones Laborales 3 at 8; Flynn Pregnancy and Dismissal:
Rejecting the Sick Male comparison (1994) ILT 257; Van Kraay The Difference Between Pregnant Women and Sick
Men (1995) 29 Law Teacher 92ff, esp at 93; McGlynn Webb v EMO: a hope for the Future (1995) Northern Ireland
Legal Quarterly 50ff, esp at 54; Boch (1995) CML Rev 547ff, esp at 558; Hervey Justifications for Sex Discrimination in
Employment (1993) p 67; Szyszczak Community Law on Pregnancy and Maternity Sex Equality Law in the European
Union (1996) pp 51ff, esp p 54 and Bolger (1994) ILT Jo 6566.
 806
38.
That is also the interpretation of the judgment in Hertz adopted by Advocate General Tesauro in his
opinion in Webb [1994] 4 All ER 115, [1994] ECR I-3567. In para 13, he states that the Court of Justice
considered, in that judgment, that the dismissal of a female employee on account of repeated absences
through illness, attributable to pregnancy or confinement, does not constitute direct discrimination on
grounds of sex if those absences occur after the period of maternity leave 7 and he adds that:
7
Advocate General Tesauro expresses the same view in his opinion in Habermann-Beltermann [1994] ECR I-1657.

That judgment may certainly not be construed as meaning that the court has recognised as
permissible (or even justifiable) the dismissal of a woman who is absent from work for a reason
(illness) connected with pregnancy. Closer examination reveals that the courts decision turned on
the fact that Mrs Hertzs illness began after her return to work at the end of her maternity leave.
And he lays stress, in para 14, on the fact that
To the extent to which that judgment holds that it is not discriminatory to dismiss an employee
on account of absences through an illness which, while it may be attributable to pregnancy or
confinement, began after the end of the maternity leave, it follows a fortiori that the pregnancy may
not be equated with illness.
39.
The contradiction between the judgments in the Dekker and Hertz cases, emphasised by some
authors8, lies in the fact that, in the former, the court described as direct discrimination a refusal to make
an appointment owing to pregnancy, since that can apply only to women, whereas, in the latter, it stated
that, for the purposes of dismissal, if an illness manifests itself after maternity leave, there is no reason to
distinguish an illness attributable to pregnancy or confinement from any other illnesswhen it is clear that
only women can suffer an illness which is attributable to pregnancy or confinementand that
contradiction can be resolved in my opinion only if it is concluded that the reason for that distinction lies in
the fact that, once a woman has given birth and returned from maternity leave, her physiological state is
no different from that of male workers and, from that time, it is inappropriate to draw a distinction by
reference to the origin of the illness. If, as a result of having given birth at any time in their lives, women
could claim what amounted to a sort of insurance against dismissal for the rest of their working life, as a
result of which no account would be taken for such purposes of periods of incapacity for work following
maternity leave, whose origin might be attributable to their pregnancy or confinement, that would amount
to a privilege contrary to the principle of equal treatment.
8
See Jaeqmain Chroniques de droit social (1991) p 50; More (1992) The Journal of Social Welfare and Family Law 48 at
55; Shaw (1991) ELRev 313 at 320; Traversa (1991) Revue trimestrielle de droit europen 425 at 436; McGlynn (1995)
Northern Ireland Legal Quarterly 50 at 5455; Boch (1995) CML Rev 547 at 559; Kilpatrick How long is a piece of
string? Regulation of the Post-Birth Period in Sex Equality Law in the European Union (1996) p 85; Szyszczak
Community Law on Pregnancy and Maternity Sex Equality Law in the European Union (1996) p 54; and Burrows &
Mair European Social Law (1996) p 155.

807
(2) The judgment in the Habermann-Beltermann case, in which the court considered it
discriminatory for a contract of employment to be declared void or avoided by virtue of the legal
prohibition of night work for pregnant women
40.
In its judgment of 5 May 1994 in Habermann-Beltermann [1994] ECR I-1657, the Court of Justice
answered preliminary questions from the Arbeitsgericht (the Labour Court) Regensburg concerning an
employment contract of indeterminate duration entered into by an employer and a pregnant employee.
When they signed the contract, which required night work, both parties were unaware of the pregnancy.
The national court wished to know whether art 2(1), in conjunction with art 3(1) and art 5(1), of Directive
76/207 precluded that contract from being declared void as a result of the prohibition of night work which
applied, under national law, during pregnancy and breast-feeding, and from being avoided by the
employer on account of a mistake as to the essential personal characteristics of the other contracting
party when the contract was concluded.
The Court of Justice observed that, in the case of a contract of indeterminate duration, the prohibition
of night work by pregnant women takes effect only for a limited period in relation to the total length of the
contract, and reached the conclusion that, in those circumstances, it would be contrary to the objective of
protection pursued by art 2(3) of Directive 76/207 and would deprive that provision of its effectiveness to
allow a contract to be declared void or be avoided because the pregnant worker was temporarily
prevented from performing the night work for which she was taken on. The specific answer which it gave
to the national court was that the provisions of Directive 76/207 preclude an employment contract for an
indefinite period for the performance of night-time work concluded between an employer and a pregnant
employee, both of whom were unaware of the pregnancy, from being held to be void on account of the
statutory prohibition on night-time work which applies, by virtue of national law, during pregnancy and
breast-feeding, or from being avoided by the employer on account of a mistake on his part as to the
essential personal characteristics of the woman at the time when the contract was concluded.
(3) The judgment in the Webb case, in which the court held that the situation of a pregnant
woman, who is unfit to carry out work for which she was recruited, cannot be compared with that
of a man who suffers the same incapacity for medical or other reasons
41.
On 14 July 1994 the Court of Justice gave judgment in Webb [1994] 4 All ER 115, [1994] ECR I-3567,
in response to a request submitted by the House of Lords for a preliminary ruling on the interpretation of
Directive 76/207 to resolve a dispute before it between an employer and a female worker, recruited for an
indeterminate period, who had been dismissed when the employer discovered she was pregnant on the
ground that he had engaged her for the specific purposeinitiallyof replacing another employee during
the latters maternity leave. In that judgment the Court of Justice stated that it was not appropriate to
consider, as requested by the national court, whether the situation of a woman who is incapable of
performing the task for which she was recruited, owing to a pregnancy of which she became aware very
soon after signing the employment contract, can be compared with that of a man 808 who suffers the
same incapacity for medical or other reasons. It states ([1994] 4 All ER 115, [1994] ECR I-3567 (para
25)):
pregnancy is not in any way comparable with a pathological condition, and even less so with
unavailability for work on non-medical grounds, both of which are situations that may justify the
dismissal of a woman without discriminating on grounds of sex. Moreover, in the Hertz case the
court drew a clear distinction between pregnancy and illness, even where the illness is attributable
to pregnancy but manifests itself after the maternity leave.
The conclusion reached by the court was that, in circumstances such as those of Mrs Webb,
termination of a contract for an indefinite period on account of the womans pregnancy cannot be justified
by the fact that she is prevented, on a purely temporary basis, from performing the work for which she has
been engaged (see [1994] 4 All ER 115, [1994] ECR I-3567 (para 27)).
(4) The judgment in the Larsson case, in which the court appears to have taken the view that it is
not discriminatory to take account, for the purposes of dismissal, of a womans incapacity for
work, attributable to pregnancy before commencement of maternity leave.
42.
Against that background, on 29 May 19979 the Court of Justice gave judgment in Larsson v Ftex
Supermarked A/S Case C-400/95 [1997] ECR I-2757. The court was again asked, this time by the S- og
Handelsretten (the Maritime and Commercial Court), Denmark, to interpret art 5(1) and art 2(1) of
Directive 76/207, as applied to a worker who had been dismissed after her maternity leave owing to
relatively long periods of sick leave attributable to her condition, most of which occurred during pregnancy
and were covered by medical certificates. I shall consider that judgment in some detail.
9
At which time the period for submitting observations in the present case had expired.

Mrs Larsson had been employed by Ftex Supermarked A/S in March 1990. In August the following
year she informed her employer that she was pregnant. During her pregnancy she took sick leave twice.
The first absence was for 18 days. The second was on account of a pelvic prolapse attributable to the
pregnancy, and lasted about four-and-a-half months. The birth took place on 2 April 1992. Her maternity
leave, of 24 weeks, ended on 18 September 1992. She then took annual leave until 16 October. During
her maternity leave and holidays, Mrs Larsson continued to receive treatment for the pelvic prolapse. She
remained on sick leave at the end of her holidays and was not declared fit for work until 4 January 1993.
On 10 November 1992less than a month after the end of her annual holidaysher employer informed
her by letter that she was to be dismissed with effect from the end of December. The reason for dismissal
was the long periods of absence from work, together with the fact that it is highly unlikelyfor health
reasonsthat in the future you will recover the ability to perform your work in a satisfactory manner.
43.
The national court asked whether the above-mentioned provisions of Directive 76/207 covered
dismissals on account of absence following the end of maternity leave if the absence was attributable to
an illness which arose during pregnancy and after maternity leave, it being assumed that the dismissal
took place after the end of the maternity leave. It should be noted 809 that the national courts question
concerned dismissals for absences following the end of maternity leave.
44.
In the opinion which I delivered in that case I suggested that the Court of Justice declare, on the basis
of the judgments in Hertz, Webb and Habermann-Beltermann, that, for application of the principle of equal
treatment for men and women in relation to conditions of employment, including those governing
dismissal, laid down in art 5(1) of Directive 76/207, a line should be drawn at the end of the maternity
period (see [1997] ECR I-2759). Thereafter, any illness suffered by a woman, whether or not obstetric in
origin, will be covered by the general regime applicable to all workers. Conversely, periods of sick leave
for health problems attributable to pregnancy, before confinement, cannot be equated, for the purposes of
dismissal, with a mans absences through sickness.
My proposed answer to the question from the national court was as follows: art 5(1), in conjunction
with art 2(1) of Directive 76/207, does not preclude a woman from being dismissed as a result of periods
of absence subsequent to her maternity leave, where those periods of absence are due to an illness
which arose during pregnancy and has continued both during and after the maternity leave, unless, when
calculating the periods of absence for the purposes of dismissal, account is taken of periods of absence
on medical grounds for health problems attributable to pregnancy prior to childbirth.
45.
In response to the arguments of the plaintiff in the main proceedings, who maintained that, in the Hertz
judgment, a distinction had been established by reference to the moment of onset or first appearance of
the illness, the Court of Justice stated in Larsson [1997] ECR I-2757 (para 17) that on that occasion
 It merely held that, in the factual situation submitted to it on that occasion, there was no reason
to distinguish, from the point of view of the principle of equal treatment enshrined in the directive,
between an illness attributable to pregnancy or confinement and any other illness. That
interpretation is confirmed, moreover, by the absence of any reference in the operative part of the
Hertz judgment to the moment of onset or first appearance of the illness.
46.
Later on in the judgment, details are given of the position of the plaintiff in the main proceedings, the
Danish government and the Commission, to the effect that it was contrary to Community law for an
employer, in calculating the period justifying dismissal under national law, to be able to take account both
of absences occurring between the beginning of the pregnancy and the beginning of maternity leave and
also the duration of the maternity leave. Indeed, it was apparent from the documents in the main
proceedings that, if those periods and the four weeks which she took as leave were discounted, Mrs
Larsson had been unfit for work for less than four weeks when she was dismissed.
47.
As regards maternity leave taken under national law, the court considered that a woman is protected
against dismissal due to absence and to allow absence during such a period to be taken into account as
grounds for a subsequent dismissal would be contrary to the objective of permitting national measures
concerning the protection of women, particularly as regards pregnancy and maternity, pursued by art 2(3)
of the directive (see [1997] ECR 810 I-2757 (para 22)). Outside the period of maternity leave provided for
by the member states and in the absence of national or Community provisions affording women specific
protection, female workers do not enjoy, under Directive 76/207, protection against dismissal for absence
due to an illness originating in pregnancy.
48.
The court inferred from those statements that
The principle of equal treatment enshrined in the directive does not, therefore, preclude account
being taken of a womans absence from work between the beginning of her pregnancy and the
beginning of her maternity leave when calculating the period providing grounds for her dismissal
under national law,
and stated in reply to the national court that, without prejudice to the provisions of national law adopted
pursuant to art 2(3) of the directive, art 5(1), in conjunction with art 2(1), of that Directive does not
preclude dismissals which are the result of absences due to an illness attributable to pregnancy or
confinement, even where that illness arose during pregnancy and continued during and after maternity
leave (see [1997] ECR I-2757 (paras 24, 26)).
49.
I confess that I find it rather difficult to reconcile those statements with the line taken in the case law
developed by the Court of Justice up to the time of that judgment, as set out in the foregoing paragraphs.
The court appears not only to maintain in Larsson a position contrary to what is to be inferred from a brief
examination of its earlier judgments, but also directly to contradict the reading of those judgments which
has been propounded over time both by its Advocates General and by the numerous authors who have
commented on the judgments.
I note, once again with a degree of preoccupation, that in para 20, where it states that the directive
does not preclude dismissal on the ground of periods of absence due to an illness attributable to
pregnancy or confinement, even where such illness first appeared during pregnancy and continued during
and after the period of maternity leave, the court does not indicate at what time the absences occurred.
Furthermore, where, in para 23, it states that, outside the period of maternity leave, in the absence of
national Community provisions, a female worker does not enjoy under art 76/207 any protection against
dismissal due to absences attributable to an illness arising during pregnancy, the court likewise fails to
make clear at what time the absences occurred.
Similarly, in para 24, when it concludes that the principle of equal treatment enshrined in the directive
does not preclude account being taken, for calculation of the period justifying dismissal, of absences
occurring between commencement of the pregnancy and commencement of maternity leave, the court
does not say that the absences were caused by an illness attributable to pregnancy.
Finally, the operative part of the judgment does not give a complete answer to the question from the
national court, which was concerned with dismissals arising as a result of absence following the end of
maternity leave if the absence is attributable to an illness which arose during pregnancy and continued
during and after maternity leave, it being assumed that the dismissal took place after the end of the
maternity leave, whereas the court answered that the provisions of the 811 directive do not preclude
dismissals which are the result of absences due to an illness attributable to pregnancy or confinement,
even where the illness arose during pregnancy and continued during and after maternity leave. Thus, the
answer does not mention either the time at which the absences occurred or the time at which the
dismissal occurred.
50.
I consider that, in view of its shortcomings, the judgment in Larsson does not provide a sufficient basis
for the view that the Court of Justice wished to make a U-turn in its case law. I believe, therefore, that
there is very good reason, for the sake of legal certainty, for the court to give a clear ruling on the principle
of equal treatment for men and women in relation to dismissals of female workers occurring during
pregnancy or after maternity leave, where account was taken of periods of unavailability for work caused
by pregnancy, before the commencement of maternity leave. That was precisely what happened in Mrs
Browns case.
51.
What is involved here, ultimately, is the duty incumbent upon us all of progressively removing all traces
of the discrimination which women have suffered over the centuries 10, a duty to which the institutions of
the European Union are so deeply committed.
10
For example, Miguel de Cervantes Saavedra, in Ch LI of the first part of his work The Adventures of Don Quijote of la
Mancha (1950) translation by J M Cohen, pp 449450), has the goatherd refer, when relating the story of Leandra, to
frivolity and failings natural to woman-kind, who are generally ill-balanced and unsteady. He then takes an easier and,
according to him, more proper course, which is to curse the fickleness of women, their inconstancy, their double-
dealing, their unkept promises, their broken faith, and last of all, the lack of judgment they show in their choice of objects
for their desires and affections.

C. The first question from the House of Lords: dismissal of a pregnant woman through
incapacity for work attributable to pregnancy
52.
By this question, the national court wishes to ascertain whether it is contrary to the principle of equal
treatment laid down by Directive 76/207 to dismiss a worker, while she is pregnant, for absence due to
incapacity for work attributable to pregnancy and whether the answer to the question is affected by the
fact that the dismissal is based on a contractual provision entitling the employer to dismiss employees,
irrespective of gender, after a stipulated number of weeks of continued absence.
53.
There is no doubt that the application of Community law would have precluded Mrs Browns dismissal
if, when complications arose in her pregnancy, the period prescribed for transposition of Directive 92/85
into national law had ended. However, that directive had not even been adopted. As a result, Directive
76/207 contains the only applicable Community law.
54.
It will be remembered that art 5 of Directive 76/207 provides that application of the principle of equal
treatment with regard to working conditions, including the conditions governing dismissal, means that men
and women are to be guaranteed the same conditions without discrimination on grounds of sex.
Starting from the premiss that equality, as defined by the Constitutional Court of one of the member
states, is not a reality or an abstract mathematical concept but rather unequal treatment of that which is
unequal or equal treatment of that which is similar or alike11 and having regard to the settled 812 case law
of the Court of Justice to the effect that discrimination can arise only through the application of different
rules to comparable situations or the application of the same rule to different situations 12, I shall now
consider whether the dismissal of a pregnant woman on grounds of incapacity for work arising from her
condition occurs under the same conditions as dismissal of a man for incapacity for work of the same
duration, arising from an illness.
11
See the judgment of the Spanish Constitutional Court 29/1987 of 6 March, para 5(b) (BOE of 24 March 1987).
12
See the judgment in Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-
225 (para 30).

55.
By contrast with the situation of Mrs Hertz, whose unfitness for work, which was of obstetric origin,
commenced some time after the end of her maternity leave, the complications attendant upon Mrs
Browns pregnancy, which prevented her from working for much of that period, became apparent at a very
early stage. It is clear from the documents before the court that she was unfit for work for the period of 26
weeks which allowed her to be dismissed pursuant to the condition included by Rentokil in employment
contracts and that she did not give birth until six weeks later. In view of the fact that the average term of
human gestation is 38 weeks, I calculate that she was able to work for only the first five or six.
56.
I wonder whether it is still necessary, at this stage, to repeat the self-evident fact that pregnancy is a
situation which affects only women, since only they can become pregnant. Pregnancy, besides being a
biological situation pertaining exclusively to women, is a period limited in time, during which there may
occur not only the well-known phenomenon of morning sickness but also complications such as risks of
miscarriage or premature contractions associated with stress, which may compel the woman to rest
absolutely for periods which may extend from two or three months to the whole of the period of
pregnancy.
57.
This court emphatically stated, in the judgment in Hertz [1990] ECR I-3979 (para 14), that
the dismissal of a female worker on account of repeated periods of sick leave which are not
attributed to pregnancy or confinement does not constitute direct discrimination on grounds of sex,
inasmuch as such periods of sick leave would lead to the dismissal of a male worker in the same
circumstances. (My emphasis.)
58.
Can it be said, however, that the dismissal of a pregnant woman on account of repeated periods of
sick leave attributable to pregnancy occurs under the same conditions as the dismissal of a man who has
been on sick leave for the same period? In my opinion the answer is no.
59.
 Without wishing to meddle in matters which are the province of doctors, I should make it clear that,
although, as this court stated in Webb [1994] 4 All ER 115, [1994] ECR I-3567 (para 25), pregnancy is not
in any way comparable with a pathological condition, no one is unaware of the existence of high-risk
pregnancies which occurthe following being examples, not an exhaustive listwhen there is a history
of premature or still-births, when the placenta is lower than normal, when the woman has undergone in
vitro fertilisation treatment, and in cases where the woman suffers from a heart condition or diabetes. The
main characteristic of such pregnancies is not that they cause the woman to be ill but that, normally, they
require her to remain 813 under strict medical supervision and, in some of the cases mentioned above, to
rest absolutely for several months or, sometimes, throughout her pregnancy 13.
13
Women whose pregnancy causes unfitness for work of such a duration are, fortunately, few in number: it is calculated
that they represent 10 to 15% of cases.

I cannot share the view that, in situations of that kind, where the woman is not suffering from any
illness but is simply pregnant, it can be said that, in the event of dismissal for repeated absences, she is
dismissed under the same conditions as a man who has been absent through illness for the same period
of time. The same reasoning will apply where the incapacity for work derives from the fact that pregnancy
has aggravated an existing illness or brought about conditions which may be classified as real illness.
60.
Whilst the situation of a pregnant worker whose pregnancy prevents her from working and that of a
male worker who is ill coincide in so far as neither of them can, for a period, carry out the tasks involved in
her or his employment, important differences distinguish them: only women may find themselves at some
time during their working life in a situation where they are prevented from working because of incapacity
arising from a pregnancy; and, in most cases, a womans incapacity for work arising from pregnancy will
end on a date known in advance with more or less accuracy, when she gives birth.
61.
These factors appear to have been taken into account by the domestic law applicable in most member
states at the time of the material events in this case, which was fairly similaronly the legislation in the
United Kingdom and Ireland differed radically.
Thus, in Germany there was specific protection for women from the start of pregnancy until after
confinement, whereby in that period dismissal was subject to administrative authorisation; in Denmark,
the Ministry of Employment considered it discriminatory to take account, for the purposes of dismissal, of
absences due to incapacity for work attributable to pregnancy, before a woman gave birth; in France, the
employer could not dismiss a woman during pregnancy, during maternity leave or during the four weeks
thereafter; in Greece, it was prohibited to dismiss a worker during pregnancy or within a year after she
gave birth and she could not be dismissed for absences due to unfitness for work attributable to
pregnancy or confinement; in Italy, the prohibition of dismissal extended from the start of the pregnancy
until the end of the first year of the childs life, the rule having been interpreted by case law to the effect
that, during that period, a woman likewise could not be dismissed for exceeding the maximum permitted
number of days sick leave; in the Netherlands, it was prohibited to dismiss a woman during pregnancy or
maternity leave or within six weeks thereafter; in Portugal, it was only possible to dismiss a worker
through incapacity for work in cases of absolute and definitive incapacity and, as regards the rules
applicable to pregnant women, dismissal was prohibited where their incapacity for work was attributable
to pregnancy or an illness associated with it. Finally, although in Spain there was no specific legislation for
the protection of pregnant women against dismissal, the effect of the case law was that a woman could
not be dismissed for absences arising from unfitness for work attributable to difficult pregnancies or
illnesses associated with pregnancy.
As regards the three new member states, in Austria specific protection existed for women from the
start of the pregnancy until after confinement, 814whereby during that period dismissal was subject to
judicial authorisation; in Finland, for a worker to be dismissed on account of incapacity for work, there had
to be a substantial and permanent reduction of her ability to work; finally, in Sweden, incapacity for work
through illness did not justify dismissal if it was probable that the worker would recover and a woman
could not be dismissed on account of unfitness for work resulting from pregnancy.
62.
The respondent in the main proceedings contends that Mrs Browns dismissal was due to the
abnormal nature of her pregnancy. It maintains that to treat her, on account of the origin of her unfitness
for work, differently from pregnant women whose pregnancy did not prevent them from working or from
other employees who were ill would constitute positive discrimination applicable only to a particular
section of women.
I cannot agree with that assertion for two reasons. First, because it must be borne in mind that
Directive 76/207 establishes the principle of equal treatment for men and women regarding working
conditions, including conditions governing dismissal, and that, in seeking a basis of comparison, it is
inappropriate to draw parallels or distinctions between two pregnant women experiencing more or less
easy or problematical pregnancies the point of reference continues to be the male worker. Secondly,
because, as I indicated in the foregoing paragraph, for the purposes of dismissal the situation of a
pregnant woman whose pregnancy prevents her from working and that of a man who is unwell are not
comparable, in the light of the principle of equal treatment laid down by that directive.
63.
Nor do I agree with the position of the United Kingdom, which proposes distinguishing between, on the
one hand, dismissal on account of ordinary risks and disorders normally inherent in pregnancy and
confinement, the examples given by it being absence from work for routine medical examinations and
minor absences for ailments such as morning sickness, and, on the other hand, dismissal of a woman on
account of an illness, whether or not linked with pregnancy. I think the origin of that suggested
differentiation is Advocate General Darmons opinion in the Dekker and Hertz cases, where he
considered, in the Hertz case, the possibility of determining the duration of protection linked with maternity
(see [1990] ECR I-3956 (paras 4748)).
64.
In my opinion, quite apart from the fact that it is doubtful whether a woman would be dismissed on
account of absences from work attributable to routine medical examinations or minor absences due to
morning sickness14, the position taken by the United Kingdom was arrived at by interpreting that part of
the opinion out of context. As I have already stated in para 30, above, Mr Darmon in fact suggested
distinguishing between, on the one hand, normal risks of pregnancy and confinement, namely the usual
complications accompanying such events, which sometimes result in the grant of an additional period of
maternity leave, and should qualify for Community protection inasmuch as they are specific to
motherhood, and, on the other, medical conditions which are not associated with the ordinary risks of
pregnancy and should be treated in the same way as ordinary sickness.
14
Morning sickness may be violent and extend beyond the first three months, requiring hospitalisation in some cases.

Now, it must not be forgotten that Mrs Hertz had been dismissed on account of incapacity for work
which, although caused by her confinement, did not prevent her from working until one year after the end
of her maternity 815 leave. To my mind, that is why the Advocate General adds, at a later stage, that,
where a female worker has exhausted her entitlement to all legally prescribed types of maternity leave,
periods of absence through sickness, even though they may be traced back to her pregnancy or
confinement, cannot be regarded as included among the normal risks of maternity.
65.
I consider, therefore, that, if one wishes to draw a distinction between the normal risks of pregnancy
and confinement and medical conditions which do not reflect the ordinary risks of pregnancy, it is
necessary to take a chronological approach: the former will, necessarily, be those which arise while the
woman is in one of those situations, that is to say whilst she is pregnant or on maternity leave, whereas
the latter will be all those which arise after maternity leave, even though their cause may be traceable
back to the pregnancy or confinement.
66.
For the reasons given above, I consider that the dismissal of a woman whilst she is pregnant, on
account of unfitness for work caused by her pregnancy, by taking into consideration a situation in which
only women can find themselves, constitutes direct discrimination contrary to art 5(1) of Directive 76/207.
67.
I wish to make it clear that the interpretation I propose is, in my opinion, the only possible interpretation
of art 5(1), in conjunction with art 2(1), of Directive 76/207 and of the line followed in the case law of the
Court of Justice until the judgment in Larsson. If that were not the case, and the future of an employed
woman, on becoming pregnant, could depend on whether or not her pregnancy involved excessive
complications, I would be obliged to state, paraphrasing Papinian, that despite the passage of the
centuries there are many points in our [Community] law in which the condition of females is inferior to that
of males (see Papinianus libro trigensimo primo quaestionum. In multis iuris nostri articulis deterior est
condicio feminarum quam masculorum. The Digest of Justinian, Vol I, Book One (Human Status), p 16).
68.
The national court also asks whether that answer is affected by the fact that the dismissal is based on
a contractual provision entitling the employer to dismiss a worker, regardless of gender, after a specific
number of weeks of continued absence.
69.
I believe it can be inferred from the foregoing reasoning that the fact that the dismissal is based on a
contractual provision entitling the employer to dismiss a worker, regardless of gender, after a specific
period of sick leave, has no impact on the foregoing answer.
That contractual provision, which applies to men and women alike, by simply assimilating incapacity
for work attributable to pregnancy to incapacity for work attributable to an illness, introduces direct
discrimination on grounds of sex, by applying the same rule to different situations, in that it takes into
consideration, for calculation of periods of incapacity for work justifying dismissal, an obstacle to work
which can affect only women.
70.
The fact that the provision is contained in the employment contract is irrelevant for the present
purposes. First, art 5(2)(b) of Directive 76/207 requires the member states to ensure that any provisions
contrary to the principle of equal treatment which are included, inter alia, in individual contracts of
employment are, or may be, declared null and void. Secondly, the Court of Justice has held that when it
interprets and applies national law, every national court must presume that the state had the intention of
fulfilling 816 entirely the obligations arising from the directive concerned (see Wagner Miret v Fondo de
Garanta Salarial Case C-334/92 [1993] ECR I-6911 (para 20)). Finally, the member states obligation, by
virtue of a directive, to achieve the result pursued by it, and its duty under art 5 of the Treaty to take all
appropriate measures, whether general or particular, to ensure fulfilment of that obligation, apply to all the
authorities of the member states, including, within their jurisdiction, judicial authorities. It follows that, in
applying national law, whether the provisions in question were adopted before or after the directive, the
national court called upon to interpret it is required to do so, as far as possible, in the light of the wording
and the purpose of the directive in order to achieve the result pursued by the latter and thereby comply
with the third paragraph of art 189 of the Treaty (see Marleasing SA v La Comercial Internacional de
Alimentacin SA Case C-106/89 [1990] ECR I-4135 (para 8)).
71.
In view of the answer which I propose be given to the national court in response to the first question, it
is unnecessary to answer the second. I shall nevertheless examine it, in case the court does not share my
view.

D. The second question from the House of Lords: dismissal of a pregnant woman during the
period in which, had she fulfilled the requirements laid down by national law, she could
have absented herself from work on account of pregnancy and confinement
72.
By its second question, the national court asks whether it is contrary to the principle of equal treatment
laid down by Directive 76/207 to dismiss a female employee as a result of absence through illness arising
from pregnancy who does not qualify for the right to absent herself from work on account of pregnancy or
confinement for the period specified by national law because of insufficient length of service, where
dismissal takes place during that period, and whether it makes any difference to the answer to be given to
that question that the employee was dismissed under a contractual provision entitling the employer to
dismiss any employee, irrespective of gender, after a stipulated number of weeks of continued absence.
73.
From the information in the order for reference, I infer that what the national court wishes to ascertain,
by asking that question, is whether Directive 76/207 must be construed as having imposed on member
states the obligation to provide, in domestic law, that female workers must be entitled to a period of
absence from work merely because they have given birth, and that no other conditions may be imposed
on them in that regard.
74.
The purpose of Directive 76/207 is to apply in the member states the principle of equal treatment for
men and women regarding access to employment, vocational training and promotion and working
conditions. However, by virtue of art 2(3) of the directive, its provisions are to be without prejudice to the
grant by member states of specific rights for women as regards pregnancy or maternity. It thus only
contemplates the adoption of measures for the protection of women in such circumstances as an
exception to that principle; it does not oblige the member states to legislate to that effect, nor does it
specify what the rights in question must be or the conditions which may be imposed for them to be
exercised.
75.
The course followed by this court has been to take the view that art 2(3) of Directive 76/207, by
reserving to member states the right to retain or introduce provisions intended to protect women in
connection with pregnancy and maternity, recognises the legitimacy, in terms of the principle 817 of
equal treatment, of protecting a womans biological condition during pregnancy and thereafter, and of
protecting the special relationship between a woman and her child over the period which follows
pregnancy and childbirth (see the judgments in Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR
3047 (para 25), Habermann-Beltermann [1994] ECR I-1657 (para 21) and Webb [1994] 4 All ER 115,
[1994] ECR I-3567 (para 20)).
76.
It is also important to bear in mind that the court held in Hertz [1990] ECR I-3979 (para 15) that
Directive 76/207
does not envisage the case of an illness attributable to pregnancy or confinement. It does,
however, admit of national provisions guaranteeing women specific rights on account of pregnancy
and maternity, such as maternity leave. During the maternity leave accorded to her pursuant to a
national law, a woman is accordingly protected against dismissal due to absence. It is for every
Member State to fix periods of maternity leave in such a way as to enable female workers to absent
themselves during the period in which the disorders inherent in pregnancy and confinement occur.
77.
The United Kingdom legislation applicable at the material time gave a worker the right to absent
herself from work from the beginning of the eleventh week before confinement and the right to return to
work within the 29 weeks following childbirth, if she fulfilled certain conditions, namely, that she had been
employed in the company until the beginning of that eleventh week; that at that time she had been
continuously employed for a period of not less than two years; that she had observed the prescribed
procedure for notification to her employer; and, at the latters request, that she had supplied a medical
certificate indicating the expected week of confinement 15.
15
The legislation concerned is extraordinarily complex, so much so that the Employment Appeal Tribunal exclaimed, in
Lavery v Plessey Telecommunications Ltd [1982] ICR 373 at 379, [1982] IRLR 180 at 182: These statutory provisions
[on maternity rights] are of inordinate complexity exceeding the worst excesses of a taxing statute; we find that
especially regrettable bearing in mind that they are regulating the every-day rights of ordinary employers and
employees. We feel no confidence that, even with the assistance of detailed arguments from skilled advocates, we have
now correctly understood them: it is difficult to see how an ordinary employer or employee is expected to do so.

According to the documents before the court, if Mrs Brown had been working for Rentokil for two years
at the beginning of the eleventh week before the expected week of her confinement, she would have
been able to absent herself from work on account of her approaching maternity before accumulating an
absence totalling 26 weeks which, in her case, gave rise to dismissal. It must therefore be assumed that if
her right to absent herself had not been conditional upon fulfilment of the requirements mentioned above,
she would have given birth and, within the 29 weeks following her confinement, would have been able to
return to work.
Moreover, it has become apparent in the course of the proceedings that Mrs Brown was entitled to
receive maternity benefit from the state at the lower rate.
78.
I wonder whether it must be considered that a woman who is close to giving birth is protected from
dismissal on account of absence because, under national legislation, her contract is suspended for a
specific period of time, which necessarily implies that she is entitled to return to her previous job, or 818
whether it is sufficient that she can cease working and be entitled to receive a benefit from the state,
without any entitlement to re-engagement.
79.
In my opinion, the only period that can be regarded as maternity leave within the meaning of the case
law of the court, that is to say the period in which a woman is in a specific situation requiring her to be
afforded special protection, which is not comparable either with that of a man or with that of a woman who
is actually working16 and in which she is protected against dismissal on account of absence, is the period
laid down by national law as a specific employment right enabling the woman to cease working for a
period of a specified duration, during which she is entitled to receive all or part of her remuneration or to
receive certain income in the form of a social security benefit, without thereby losing her job.

16
See the judgment in Gillespie v Northern Health and Social Services Board Case C-342/93 [1996] All ER (EC)
284, [1996] ECR I-475 (para 17).

80.
However, the simple right available to Mrs Brown or to any other female worker who, like her, did not
fulfil the necessary conditions to secure suspension of her employment contract, to receive for a
maximum period of 18 weeks a maternity benefit from the state if she stopped work in order to give birth,
without the possibility of re-engagement, does not constitute, in relation to the principle of equality, a
specific employment right intended to protect the biological condition of a woman or the relationship
between mother and child in the period following confinement.
Indeed, the right to cease working for an indefinite time, receiving for a limited period a financial
benefit from the state, is not reserved to women, nor is it necessary to be pregnant or to have given birth
in order to exercise it.
81.
Therefore, the eleven-week period preceding the expected date of confinement, in which, had she
fulfilled the requisite conditions, Mrs Brown could have exercised her right to absent herself from work,
cannot in my view be regarded as a period during which she was protected against dismissal on account
of absence.
82.
Furthermore, Directive 92/85, which lays down specific measures of protection within the Community
for workers who are pregnant, have recently given birth or are breast-feeding, and requires the member
states to provide in their legislation for a continuous period of maternity leave of at least 14 weeks, of
which at least two weeks must be compulsory, being allocated before or after confinement, had not been
adopted at the material time.
83.
For those reasons, in the event of this court considering that it is not discriminatory to dismiss a
pregnant woman on account of absences arising from an illness attributable to pregnancy, it will have to
be concluded that Directive 76/207 does not preclude dismissal of such a person when she is not entitled
to absent herself from work on account of pregnancy or confinement for the period specified by national
law because of insufficient length of service, where dismissal takes place during that period.
84.
Finally, the fact that the womans dismissal is based on a contractual provision entitling the employer
to dismiss any worker, regardless of gender, after a stipulated number of weeks of continued absence
makes no difference to the answer which I suggest for the second question.
819

VIICONCLUSION
In view of the foregoing considerations, I propose that the Court of Justice answer the first question
from the House of Lords as follows:
(1)(a) Dismissal of a woman whilst she is pregnant, on account of her incapacity for work as a
result of pregnancy, is contrary to art 5(1) in conjunction with art 2(1) of Council Directive (EEC)
76/207 on the implementation of the principle of equal treatment for men and women as regards
access to employment, vocational training and promotion, and working conditions. (b) The fact that
the dismissal is based on a contractual provision entitling the employer to dismiss an employee,
regardless of gender, after a stipulated number of weeks of continued absence does not affect the
foregoing answer in any way.
In the event that, on the contrary, the Court of Justice considers that it is not discriminatory to dismiss a
pregnant woman on account of incapacity for work attributable to her pregnancy, I suggest the following
answer to the second question:
 (2)(a) Article 5(1), in conjunction with art 2(1), of Directive 76/207 does not preclude dismissal of
a pregnant woman who is not entitled to absent herself from work on account of pregnancy or
confinement for the period specified by national law because of insufficient length of service where
dismissal takes place during that period. (b) The fact that the dismissal is based on a contractual
provision entitling the employer to dismiss an employee, regardless of gender, after a stipulated
number of weeks of continued absence does not affect the foregoing answer in any way.

30 June 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 28 November 1996, received at the registry of the Court of Justice of the European
Communities on 9 December 1996, the House of Lords referred to the Court of Justice for a preliminary
ruling under art 177 of the EC Treaty two questions on the interpretation of arts 2(1) and 5(1) of Council
Directive (EEC) 76/207 on the implementation of the principle of equal treatment for men and women as
regards access to employment, vocational training and promotion, and working conditions (OJ 1976 L39 p
40).
2.
Those questions have been raised in proceedings brought by Mary Brown against Rentokil Ltd in
connection with her dismissal whilst pregnant.
3.
According to the order for reference, Mrs Brown was employed by Rentokil as a driver. Her job was
mainly to transport and change Sanitact units in shops and other centres. In her view, it was heavy work.
4.
In August 1990, Mrs Brown informed Rentokil that she was pregnant. Thereafter she had difficulties
associated with the pregnancy. From 16 August 1990 onwards, she submitted a succession of four-week
certificates mentioning various pregnancy-related disorders. She did not work again after mid-August
1990.
5.
Rentokils contracts of employment included a clause stipulating that, if an employee was absent
because of sickness for more than 26 weeks continuously, he or she would be dismissed.
6.
On 9 November 1990, Rentokils representatives told Mrs Brown that half of the 26-week period had
run and that her employment would end on 8208 February 1991 if, following an independent medical
examination, she had not returned to work by then. A letter to the same effect was sent to her on that
date.
7.
Mrs Brown did not go back to work following that letter. The parties agree that there was never any
question of her being able to return to work before the end of the 26-week period. By letter of 30 January
1991, which took effect on 8 February 1991, she was accordingly dismissed while pregnant. Her child
was born on 22 March 1991.
8.
At the time when Mrs Brown was dismissed, s 33 of the Employment Protection (Consolidation) Act
1978 provided that an employee who was absent from work wholly or partially because of pregnancy or
confinement would, subject to certain conditions, be entitled to return to work. In particular, the employee
had to have been in employment until immediately before the start of the eleventh week before the
expected date of confinement and, at the beginning of the eleventh week, have been continuously
employed for a period of not less than two years.
9.
According to the order for reference, on the assumption that the date on which Mrs Browns child was
born was also the expected date of delivery, she was not entitled, because she had not been in
employment for two years as at 30 December 1990, to absent herself from work from the beginning of the
eleventh week before delivery pursuant to s 33 of the Employment Protection (Consolidation) Act 1978, or
to return to work at any time during the 29 weeks following delivery. She was, however, entitled to
statutory maternity pay under ss 46 to 48 of the Social Security Act 1986.
10.
By order dated 5 August 1991, the Industrial Tribunal dismissed Mrs Browns application under the Sex
Discrimination Act 1975 concerning her dismissal. The tribunal held that, where absence through
pregnancy-related illness, but which began long before the statutory maternity provisions could apply and
subsisted continuously thereafter, is followed by dismissal, that dismissal does not fall into the category of
dismissals which must automatically be considered discriminatory because they are due to pregnancy.
11.
By order of 23 March 1992, the Employment Appeal Tribunal dismissed Mrs Browns appeal ([1992]
IRLR 302).
12.
 By judgment of 18 January 1995, the Extra Division of the Court of Session reached the preliminary
conclusion ([1995] IRLR 211) that in this case there was no discrimination within the meaning of the Sex
Discrimination Act 1975. It pointed out that, since the Court of Justice had drawn a clear distinction
between pregnancy and illness attributable to pregnancy, Mrs Brown, whose absence was due to illness
and who had been dismissed on account of that illness, could not succeed (see the judgment in Handels-
og Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiverforening Case C-179/88 [1990] ECR
I-3979 (the Hertz case)).
13.
Mrs Brown appealed to the House of Lords, which referred the following questions to the Court of
Justice for a preliminary ruling:
(1)(a) Is it contrary to Articles 2(1) and 5(1) of Directive 76/207 of the Council of the European
Communities (the Equal Treatment Directive) to dismiss a female employee, at any time during
her pregnancy, as a result of absence through illness arising from that pregnancy? (b) Does it make
any difference to the answer given to Question 1(a) that the 821 employee was dismissed in
pursuance of a contractual provision entitling the employer to dismiss employees, irrespective of
gender, after a stipulated number of weeks of continued absence?
(2)(a) Is it contrary to Articles 2(1) and 5(1) of the Equal Treatment Directive to dismiss a female
employee as a result of absence through illness arising from pregnancy who does not qualify for
the right to absent herself from work on account of pregnancy or childbirth for the period specified
by national law because she has not been employed for the period imposed by national law, where
dismissal takes place during that period? (b) Does it make any difference to the answer given to
Question 2(a) that the employee was dismissed in pursuance of a contractual provision entitling the
employer to dismiss employees, irrespective of gender, after a stipulated number of weeks of
continued absence?

The first part of the first question

14.
It should be noted at the outset that the purpose of Directive 76/207, according to art 1(1), is to put into
effect in the member states the principle of equal treatment for men and women as regards access to
employment, vocational training and promotion, and working conditions.
15.
Article 2(1) of the directive provides that
the principle of equal treatment shall mean that there shall be no discrimination whatsoever on
grounds of sex either directly or indirectly by reference in particular to marital or family status.
According to art 5(1) of the directive:
Application of the principle of equal treatment with regard to working conditions, including the
conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex.
16.
According to settled case law of the Court of Justice, the dismissal of a female worker on account of
pregnancy, or essentially on account of pregnancy, can affect only women and therefore constitutes direct
discrimination on grounds of sex (see the judgments in Dekker v Stichting Vormingscentrum voor Jong
Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990] ECR I-3941 (para 12), Hertz [1990] ECR I-3979
(para 13), Habermann- Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb/Opf-eV Case C-421/92
[1994] ECR I-1657 (para 15) and Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115,
[1994] ECR I-3567 (para 19)).
17.
As the court pointed out in its judgment in Webb [1994] 4 All ER 115, [1994] ECR I-3567 (para 20), by
reserving to member states the right to retain or introduce provisions which are intended to protect
women in connection with pregnancy and maternity, art 2(3) of Directive 76/207 recognises the
legitimacy, in terms of the principle of equal treatment, first, of protecting a womans biological condition
during and after pregnancy and, second, of protecting the special relationship between a woman and her
child over the period which follows pregnancy and childbirth.
18.
It was precisely in view of the harmful effects which the risk of dismissal may have on the physical and
mental state of women who are pregnant, women who have recently given birth or women who are
breastfeeding, 822including the particularly serious risk that pregnant women may be prompted
voluntarily to terminate their pregnancy, that the Community legislature, pursuant to art 10 of Council
Directive (EEC) 92/85 on the introduction of measures to encourage improvements in the safety and
health at work of pregnant workers and workers who have recently given birth or are breastfeeding (tenth
individual Directive adopted within the meaning of art 16(1) of Directive 89/391 (OJ 1992 L348 p 1), which
was to be transposed into the laws of the member states no later than two years after its adoption,
provided for special protection to be given to women, by prohibiting dismissal during the period from the
beginning of their pregnancy to the end of their maternity leave. Article 10 of Directive 92/85 provides that
there is to be no exception to, or derogation from, the prohibition of dismissal of pregnant women during
that period, save in exceptional cases not connected with their condition (see, in this regard, the judgment
in Webb [1994] 4 All ER 115, [1994] ECR I-3567 (paras 2122)).
19.
 In replying to the first part of the first question, which concerns Directive 76/207, account must be
taken of that general context.
20.
At the outset, it is clear from the documents before the court that the question concerns the dismissal
of a female worker during her pregnancy as a result of absences through incapacity for work arising from
her pregnant condition. As Rentokil points out, the cause of Mrs Browns dismissal lies in the fact that she
was ill during her pregnancy to such an extent that she was unfit for work for 26 weeks. It is common
ground that her illness was attributable to her pregnancy.
21.
However, dismissal of a woman during pregnancy cannot be based on her inability, as a result of her
condition, to perform the duties which she is contractually bound to carry out. If such an interpretation
were adopted, the protection afforded by Community law to a woman during pregnancy would be
available only to pregnant women who were able to comply with the conditions of their employment
contracts, with the result that the provisions of Directive 76/207 would be rendered ineffective (see [1994]
4 All ER 115, [1994] ECR I-3567 (para 26)).
22.
Although pregnancy is not in any way comparable to a pathological condition ([1994] 4 All ER 115,
[1994] ECR I-3567 (para 25)), the fact remains, as the Advocate General stresses in para 56 of his
opinion, above, that pregnancy is a period during which disorders and complications may arise compelling
a woman to undergo strict medical supervision and, in some cases, to rest absolutely for all or part of her
pregnancy. Those disorders and complications, which may cause incapacity for work, form part of the
risks inherent in the condition of pregnancy and are thus a specific feature of that condition.
23.
In its judgment in Hertz [1990] ECR I-3979 (para 15), the court, on the basis of art 2(3) of Directive
76/207, also pointed out that that directive admits of national provisions guaranteeing women specific
rights on account of pregnancy and maternity. It concluded that, during the maternity leave accorded to
her under national law, a woman is protected against dismissal on the grounds of her absence.
24.
Although, under art 2(3) of Directive 76/207, such protection against dismissal must be afforded to
women during maternity leave ([1990] ECR I-3979 (para 15)), the principle of non-discrimination, for its
part, requires similar protection throughout the period of pregnancy. Finally, as is clear 823 from para 22
of this judgment, dismissal of a female worker during pregnancy for absences due to incapacity for work
resulting from her pregnancy is linked to the occurrence of risks inherent in pregnancy and must therefore
be regarded as essentially based on the fact of pregnancy. Such a dismissal can affect only women and
therefore constitutes direct discrimination on grounds of sex.
25.
It follows that arts 2(1) and 5(1) of Directive 76/207 preclude dismissal of a female worker at any time
during her pregnancy for absences due to incapacity for work caused by an illness resulting from that
pregnancy.
26.
However, where pathological conditions caused by pregnancy or childbirth arise after the end of
maternity leave, they are covered by the general rules applicable in the event of illness (see, to that effect,
Hertz [1990] ECR I-3979 (paras 1617)). In such circumstances, the sole question is whether a female
workers absences, following maternity leave, caused by her incapacity for work brought on by such
disorders, are treated in the same way as a male workers absences, of the same duration, caused by
incapacity for work; if they are, there is no discrimination on grounds of sex.
27.
It is also clear from all the foregoing considerations that, contrary to the courts ruling in Larsson v
Ftex Supermarked A/S Case C-400/95 [1997] ECR I-2757 (para 23), where a woman is absent owing to
illness resulting from pregnancy or childbirth, and that illness arose during pregnancy and persisted
during and after maternity leave, her absence not only during maternity leave but also during the period
extending from the start of her pregnancy to the start of her maternity leave cannot be taken into account
for computation of the period justifying her dismissal under national law. As to her absence after maternity
leave, this may be taken into account under the same conditions as a mans absence, of the same
duration, through incapacity for work.
28.
The answer to the first part of the first question must therefore be that arts 2(1) and 5(1) of Directive
76/207 preclude dismissal of a female worker at any time during her pregnancy for absences due to
incapacity for work caused by illness resulting from that pregnancy.

The second part of the first question

29.
The second part of the first question concerns a contractual term providing that an employer may
dismiss workers of either sex after a stipulated number of weeks of continuous absence.
30.
It is well settled that discrimination involves the application of different rules to comparable situations
or the application of the same rule to different situations (see esp Gillespie v Northern Health and Social
Services Board Case C-342/93 [1996] All ER (EC) 284, [1996] ECR I-475 (para 16)).
31.
 Where it is relied on to dismiss a pregnant worker because of absences due to incapacity for work
resulting from her pregnancy, such a contractual term, applying both to men and to women, is applied in
the same way to different situations since, as is clear from the answer given to the first part of the first
question, the situation of a pregnant worker who is unfit for work as a result of disorders associated with
her pregnancy cannot be considered to be the same as that of a male worker who is ill and absent
through incapacity for work for the same length of time.
32.
Consequently, application of that contractual term in circumstances such as the present constitutes
direct discrimination on grounds of sex.
824
33.
The answer to the second part of the first question must therefore be that the fact that a female worker
has been dismissed during her pregnancy on the basis of a contractual term providing that the employer
may dismiss employees of either sex after a stipulated number of weeks of continuous absence cannot
affect the answer given to the first part of the first question.

The second question

34.
In view of the answer given to the first question, it is unnecessary to answer the second question.

Costs
35.
The costs incurred by the UK government and by the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the proceedings pending before the national court, the decision on costs
is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the House of Lords
by order of 28 November 1996, hereby rules: arts 2(1) and 5(1) of Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to employment,
vocational training and promotion, and working conditions, preclude dismissal of a female worker at any
time during her pregnancy for absences due to incapacity for work caused by illness resulting from that
pregnancy. The fact that a female worker has been dismissed during her pregnancy on the basis of a
contractual term providing that the employer may dismiss employees of either sex after a stipulated
number of weeks of continuous absence does not affect the answer given.

825

[1998] All ER (EC) 826

Gilly and another v Directeur des Services Fiscaux du Bas-Rhin

(Case C-336/96)

EUROPEAN COMMUNITY; Free movement of workers

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), WATHELET (RAPPORTEUR) AND SCHINTGEN
(PRESIDENTS OF CHAMBERS), MANCINI, MOITINHO DE ALMEIDA, KAPTEYN, MURRAY,
PUISSOCHET, HIRSCH, SEVN AND IOANNOU
ADVOCATE GENERAL RUIZ-JARABO COLOMER
23 OCTOBER, 20 NOVEMBER 1997, 12 MAY 1998
European Community Freedom of movement Workers Direct taxation Applicants married and
resident in France where married couples income tax payable on joint income German wife working in
Germany and having dual nationality Franco-German convention on avoidance of double taxation
applying criteria of, inter alia, nationality to determine tax jurisdiction Convention making wifes income
taxable in Germany but entitling her to tax credit in France French tax credit less than tax actually paid
in Germany Whether constituting discrimination of grounds of nationality EC Treaty, art 48.

The applicants, Mr and Mrs G, lived in France, near the German border. Since Mrs G, a German national
who had also acquired French nationality by marriage, taught in a state school across the border in
Germany, her income was taxed in accordance with a Franco-German convention for the avoidance of
double taxation. Under arts 131 and 142 of that convention Mrs Gs teaching income was taxable in
Germany, since she was a public sector employee in that state and also had German as well as French
nationality. Under art 203 of the convention, the fact that Mrs Gs income was taxed in Germany entitled
her to a tax credit in France (where she and her husband were taxed as a couple) equal to the amount of
the French tax on the relevant income. Since the German tax rates were more progressive than French
ones, it was possible for the amount of the French tax credit to be less than the tax actually paid in
Germany. Mrs Gs income was taxed in accordance with those rules from 1989 to 1993. Thereafter, she
and her husband brought proceedings before the Tribunal Administratif, Strasbourg, contending that the
application of the convention had led to unjustified, discriminatory and excessive taxation, which was
incompatible with, inter alia, art 484 of the EC Treaty (free movement of workers). The tribunal stayed the
proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling the
questions whether art 48 precluded the application to frontier workers of (i) a tax regime which differed
depending on whether they worked in the public or private sector, where they worked in the public sector
and whether or not they had only the nationality of the state employing them or not; and (ii) a tax credit
mechanism such as that provided for in art 20 of the convention.
1
Article 13, so far as material, is set out at p 848 e to h, post
2
Article 14, so far as material, is set out at p 848 j to p 849 a, post
3
Article 20, so far as material, is set out at p 849 d e, post
4
Article 48, so far as material, is set out at p 832 f, post

826

Held (1) Although the abolition of double taxation within the Community was one of the objectives of the
Treaty, no unifying or harmonising measure for the elimination of double taxation had been adopted at
Community level, nor had the members states yet concluded any multilateral convention to that effect
under art 2205 of the Treaty. Member states were therefore competent to determine the criteria for
taxation on income and wealth with a view to eliminating double taxation by means of international
agreement. It followed that art 48 of the Treaty did not preclude the application of the provisions such as
those in arts 13 and 14 of the convention (see p 851 j to p 853 b, p 853 f and p 856 c d, post).
5
Article 220, so far as material, is set out at p 832 g, post

(2) The object of a convention such as that at issue was to prevent the same income from being taxed
in each of the two states and not to ensure that the tax to which the taxpayer was subject in one state was
no higher than that to which he would be subject in the other. Indeed, if the state of residence were
required to accord a tax credit greater than the fraction of its national tax corresponding to the income
from abroad, it would have to reduce its tax in respect of the remaining income, which would entail a loss
of tax revenue for it and would thus encroach on its sovereignty in matters of direct taxation. Moreover,
the fact that, in allocating powers of taxation between them, the contracting parties had chosen various
connecting factors, in particular nationality with regard to public service remuneration received in the state
other than the state of residence, could not in itself constitute discrimination prohibited by Community law.
It followed that art 48 of the Treaty did not preclude the application of a tax credit mechanism such as that
provided for in art 20 of the convention (see p 854 j, p 855 b h j and p 856 e, post).

Notes
For the prohibition of discrimination on grounds of nationality, see 52 Halsburys Laws (4th edn) 1513.
For the EC Treaty, art 48, see 50 Halsburys Statutes (4th edn) 283.

Cases cited
Asscher v Staatssecretaris van Financien Case C-107/94 [1996] All ER (EC) 757, [1996] I-3089, ECJ.
Biehl v Administration des Contributions du Grand-Duch de Luxembourg Case C-175/88 [1991] STC
575, [1990] ECR I-1779, ECJ.
Dias v Director da Alfndega do Porto Case C-343/90 [1992] ECR I-4673.
Durighello v Istituto Nazionale Della Previdenza Sociale (INPS) Case C-186/90 [1991] ECR I-5773.
EC Commission v France Case 270/83 [1986] ECR 273.
EC Commission v Luxembourg Case C-111/91 [1993] ECR I-817.
Esso Espaola v Administracin de la Comunidad Autnoma de Canarias Case C-134/94 [1995] ECR I-
4223.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Foglia v Novello Case 244/80 [1981] ECR 3045.
Futura Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471.
827
Gullung v Conseils de lordre des avocats du barreau de Colmar et de Saverne Case 292/86 [1988] ECR
111.
Kontogeorgas v Kartonpak AE Case C-104/95 [1996] ECR I-6643.
Matteucci v Communaut franaise de Belgique Case 238/87 [1988] ECR 5589.
Meilicke v ADV/ORGA AG Case C-83/91 [1992] ECR I-4871.
Ministre public v Mutsch Case 137/84 [1985] ECR 2861.
Mora Romero v Landesversicherungsanstalt Rheinprovinz Case C-131/96 [1997] ECR I-3659.
Pigs Marketing Board v Redmond Case 83/78 [1978] ECR 2347.
Robards v Insurance Officer Case 149/82 [1983] ECR 171.
Ruiz Bernldez (Criminal proceedings against) Case C-129/94 [1996] All ER (EC) 741, [1996] ECR I-
1829, ECJ.
Salonia v Poidomani Case 126/80 [1981] ECR 1563.
Scholz v Opera Universitaria di Cagliari Case C-419/92 [1994] ECR I-505.
Skanavi (Criminal proceedings against) Case C-193/94 [1996] All ER (EC) 435, [1996] ECR I-929, ECJ.
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER
(EC) 97, [1995] ECR I-4921, ECJ.
USSL No 47 di Biella v Istituto Nazionale per lAssicurazione contro gli Infortuni sul Lavoro (Inail) Case C-
134/95 [1997] ECR I-195.
Wielockx v Inspecteur der Directe Belastingen Case C-80/94 [1995] All ER (EC) 679, [1995] ECR I-2493,
ECJ.

Reference
By judgment of 10 October 1996, the Tribunal Administratif (the Administrative Court), Strasbourg,
referred to the Court of Justice of the European Communities for a preliminary ruling under art 177 of the
EC Treaty six questions (set out at p 850 b to j, post) on the interpretation of arts 6, 48 and 220 of that
Treaty. Those questions were raised in the context of various proceedings brought by Mr and Mrs Gilly
against the Director of Tax Services for the dpartement of Bas-Rhin, concerning the calculation of their
personal income tax pursuant to the Convention between the French Republic and the Federal Republic
of Germany for the avoidance of double taxation and the establishment of rules for mutual legal and
administrative assistance in the field of income and wealth tax and in the field of business tax and land tax
(Paris, 21 July 1959). Written observations were submitted by: Mr and Mrs Gilly, the applicants in the
main proceedings; the French government, represented by C de Salins, Head of Sub-directorate in the
Legal Directorate of the Ministry of Foreign Affairs, and G Mignot, Foreign Affairs Secretary in the same
directorate, acting as agents; the Belgian government, represented by J Devadder, General Adviser in the
Ministry of Foreign Affairs, Trade and Co-operation with Developing Countries, acting as agent; the
Danish government, represented by P Biering, Legal Adviser, Head of Division in the Ministry of Foreign
Affairs, acting as agent; on behalf of the German government, by E Rder, Ministerialrat in the Federal
Ministry of Economic Affairs, acting as agent; the Italian government, represented by Prof U Leanza,
Head of the Legal Department of the Ministry of Foreign Affairs, acting as agent, assisted by G De Bellis,
Avvocato dello Stato; the Finnish government, represented by H Rotkirch, Ambassador, Head of the Legal
Department of the Ministry of Foreign Affairs, acting as agent; the Swedish government, represented by E
Brattgrd, Departmentsrd in the Foreign Trade Department of the Ministry of Foreign Affairs, acting as
agent; 828the United Kingdom government, represented by J E Collins, Assistant Treasury Solicitor,
acting as agent; and the European Commission, represented by H Michard and E Traversa, of its Legal
Service, acting as agents. Oral observations were made by: Mr Gilly; the French government,
represented by G Mignot; the Danish government, represented by J Molde, Legal Adviser, Head of
Division in the Ministry of Foreign Affairs, acting as agent; the Italian government, represented by G De
Bellis; the Netherlands government, represented by M Fierstra, Deputy Legal Adviser in the Ministry of
Foreign Affairs, acting as agent; the UK government, represented by R Singh, of the Treasury Solicitors
Department, acting as agent; and of the Commission, represented by H Michard. The language of the
case was French. The facts are set out in the opinion of the Advocate General.

20 November 1998.

The Advocate General (D Ruiz-Jarabo Colomer)

delivered the following opinion (translated from the Spanish).
1.
The Tribunal Administratif, Strasbourg, has referred to the Court of Justice of the European
Communities, pursuant to art 177 of the EC Treaty, various questions concerning the interpretation of arts
6, 48 and 220 of the EC Treaty in order to give judgment in actions brought by Mr and Mrs Gilly against
decisions of the Directeur des Services Fiscaux du Bas-Rhin requiring them to pay additional personal
income tax for the years 1986, 1988, 1990, 1991, 1992 and 1993.
2.
According to the findings of the national court in the order for reference, Mr Gilly is a French national
and works as a teacher in the French state education system. His wife, who was originally a German
national and acquired French nationality by marriage, is employed as a teacher in the German state
education system. The couple reside in France.
3.
Mrs Gillys income tax liability is governed by the Convention between the French Republic and the
Federal Republic of Germany for the avoidance of double taxation and the establishment of rules for
mutual legal and administrative assistance in the field of income and wealth tax and in the field of
business tax and land tax (Paris, 21 July 1959) (the Franco-German convention), by the 1969 and 1989
Additional Protocols (Bonn, 9 June and 28 September, respectively) and by Instruction 14-B-2-93 of the
French tax authorities containing detailed rules for application.
4.
In the present case, Mrs Gillys income from employment, which was paid by Land Baden-
Wrttemberg, was taxed in Germany in accordance with the first sentence of art 14(1) of the Franco-
German convention because it was public service remuneration and the recipient was a German national.
5.
 To avoid double taxation, while art 20(a) in the version of the 1969 Additional Protocol was in force,
France did not include Mrs Gillys income in the couples taxable income but took it into account when
calculating the rate of tax applying to income received in France. When this provision was amended by
the 1989 Protocol, such income was also taxed in France, although in respect of tax paid abroad there
was a right to a tax credit equal to the amount of the French tax on the relevant income.
6.
In their actions Mr and Mrs Gilly contend that application of the Franco-German convention entails in
their case excessive, unjustified and discriminatory taxation which is incompatible with arts 3(c), 6, 48,
73d and 220 of the Treaty. They claim that the additional tax assessments by the French tax authorities
should be annulled and that Mrs Gilly should be granted the status of 829 a frontier worker for tax
purposes. Alternatively, they seek an order that the tax credit granted in France in respect of tax paid
abroad should be equal to the amount of the actual tax paid in Germany and, if not, that Mrs Gillys
income in Germany should not be taken into account in calculating the couples tax in France. Finally,
they seek repayment of the tax wrongfully paid.

The questions referred

7.
Taking the view that the outcome of the proceedings depended on the interpretation of arts 6, 48 and
220 of the Treaty, the Tribunal Administratif, Strasbourg, stayed the proceedings and requested the Court
of Justice to give a preliminary ruling:
(1) on whether the principle of freedom of movement for workers, as embodied in the Treaty of
Rome and the implementing legislation, is contravened by a tax regime, applicable to frontier
workers, of the kind provided for by the Franco-German Convention, in so far as the latter lays
down taxation arrangements which are different for people whose remuneration is paid by a public
entity as compared with those whose remuneration is paid by private persons and as a result is
liable to have an impact on access to posts in the public or private sectors depending on residence
in one State or another;
(2) as to the compatibility, in view of the Courts interpretation of the Treaty, with the principle of
the freedom of movement and the abolition of all discrimination on grounds of nationality of a rule
under which a frontier worker receiving remuneration from a State or an agency thereof governed
by public law is taxable in that State whereas, if the frontier worker has the nationality of the other
State but is not at the same time a national of the first State, his remuneration is taxable in the
State where the frontier worker resides;
(3) as to the compatibility with Article 7 [now Article 6]6 of the Treaty of a tax provision which lays
down for frontier workers employed by persons governed by public law and residing in one of the
Member States a tax regime which differs according to whether they are nationals only of that State
or have dual nationality;

6
Amended by art G(8) of the Treaty on European Union (Maastricht, 7 February 1992; TS 12 (1994); Cmnd 2485).

(4) on whether the principle of freedom of movement for workers, as embodied in the Treaty, is
contravened by tax rules which are liable to affect the choice made by teachers in the contracting
states as to whether to work on a more or less long-term basis in another State having regard to
the differences, based on the duration of employment, in the tax regimes of the States in question;
(5) on whether the objective of abolishing double taxation laid down in Article 220 of the Treaty
must be regarded, in view of the time which the Member States have had to implement it, as now
having the status of a directly applicable rule under which double taxation may no longer take place
and, secondly, whether the objective of avoiding double taxation assigned to the Member States by
Article 220 is contravened by a tax convention under which the tax regime applicable to frontier
workers of States party to the convention varies according to their nationality and the 830 public or
private nature of the post held and whether a tax credit regime applicable to a household living in
one State which does not take into account the exact amount of the tax paid in another State but
only a tax credit, which may be lower, meets the objective assigned to the Member States of
abolishing double taxation; and
(6) on whether Article 48 must be interpreted as meaning that nationals of a member state who
are frontier workers in another Member State may not, by reason of a tax credit mechanism of the
type provided for by the Franco-German Convention, be taxed more heavily than persons whose
occupational activity is pursued in their State of residence.

The contested provisions of the Franco-German convention

8.
Article 13(1) lays down the basic principle that income from employment is taxable only in the
contracting state where the personal activity giving rise to the income is carried on. This rule does not
apply to what is referred to as public service remuneration.
9.
Article 13(5)(a) provides for an exception to the above-mentioned rule in that income from employment
earned by persons who work in the frontier area of one contracting state and who have their permanent
home in the frontier area of the other contracting state to which they normally return each day is taxable
only that other state.
10.
 Article 14(1) sets out the criteria governing the taxation of public service remuneration. The first
sentence lays down the general rule that remuneration paid by one of the contracting states or by a Land
or by a legal person of that state governed by public law to natural persons resident in the other state in
consideration for military or administrative services is taxable only in the first state.
There is also an exception to this rule, set out in the second sentence, which is to the effect that where
remuneration is paid to persons having the nationality of the other state without being at the same time
nationals of the first state, the remuneration is taxable only in the state where they reside.
11.
Article 16 contains a special rule applying to teachers who go from one state to work in the other for a
limited period, in which case they remain taxable in the state in which they habitually reside. According to
this provision, teachers habitually residing in one of the contracting states who, in the course of a
temporary stay not exceeding two years in the other state, receive remuneration for teaching activity in a
university, college, school or other teaching establishment are taxable on that remuneration only in the
first state.
12.
Article 20(2) lays down detailed rules for the avoidance of double taxation of persons residing in
France. As worded by the 1969 Additional Protocol, it provided as follows:
(a) Subject to the provisions of (b) and (c), income arising in the Federal Republic which, under
this Convention, is taxable in the Federal Republic shall be excluded from the basis of assessment
in France. However, this rule shall not limit the right of France to take account of the income thus
excluded when determining its rates of taxation.
Since the entry into force of the 1989 Additional Protocol, the wording of the provision for the
avoidance of double taxation of persons residing in France has been as follows:
831
(a) Profits and other positive income arising in the Federal Republic and taxable there under the
provisions of this Convention shall also be taxable in France where they accrue to a person
resident in France. The German tax shall not be deductible for calculation of the taxable income in
France. However, the recipient shall be entitled to a tax credit which may be set against the French
tax charged on the taxable amount which includes that income. That tax credit shall be equal
(cc) for all other income, to the amount of the French tax on the relevant income. This provision
shall apply in particular to the income referred to in Articles 13(1) and (2) and 14.
13.
Article 21(1) provides for equal treatment of taxpayers in that nationals of one contracting state are not
to be liable in the other contracting state for any tax or obligation relating thereto which is different from or
more onerous than the taxes or obligations relating thereto to which nationals of that other state are or
may be liable in the same situation.

The Community legislation

14.
The provisions which are the subject of the national courts request for interpretation all form part of
the EC Treaty and are as follows:
Article 6
Within the scope of application of this Treaty, and without prejudice to any special provisions
contained therein, any discrimination on the grounds of nationality shall be prohibited
Article 48
(1) Freedom of movement for workers (2) shall entail the abolition of any discrimination
based on nationality between workers of the Member States as regards employment, remuneration
and other conditions of work and employment
Article 220
Member States shall, so far as is necessary, enter into negotiations with each other with a view
to securing for the benefit of their nationals the abolition of double taxation within the
Community
15.
In addition, art 7 of Council Regulation (EEC) 1612/68 on freedom of movement for workers within the
Community, provides as follows (OJ S Edn 1968 (II) p 475):
(1) A worker who is a national of a Member State may not, in the territory of another Member
State, be treated differently from national workers by reason of his nationality in respect of any
conditions of employment and work, in particular as regards remuneration, dismissal, and should
he become unemployed, reinstatement or re-employment. (2) He shall enjoy the same social and
tax advantages as national workers.

The different views put forward in the proceedings on the reference

16.
Within the time limit laid down for the purpose by art 20 of the EC Statute of the Court of Justice,
written observations were submitted by the 832 plaintiffs in the main proceedings, the governments of
Belgium, Denmark, Germany, France, Italy, Finland, Sweden and the United Kingdom, and the European
Commission. During the oral procedure observations were submitted by the plaintiffs in the main
proceedings and by the representatives of the governments of Denmark, France, Italy, the Netherlands
and the United Kingdom, and by the Commission.
17.
The plaintiffs consider Mrs Gillys tax situation to be discriminatory for three reasons: first, because art
14(1) of the Franco-German convention provides for treatment which differs according to the nationality of
the taxpayer in so far as that nationality determines whether the income from employment paid by a
public agency is taxable in the state which pays it or in the state of residence; secondly, because the
same provision differentiates between frontier workers according to whether they work in the public or the
private sector; and, thirdly, because art 16 of the Franco-German convention distinguishes between
teachers residing in France according to whether they go to teach in Germany for a period of less or more
than two years.
In addition, they assert, Mrs Gilly suffers double taxation in so far as, under the Franco-German
convention, her income from employment is taxable both in Germany, where she is deemed to be a single
taxpayer without children (whereas in fact she is married and has two dependent children) and in France,
where the income she receives in Germany is added to her husbands income for the purpose of
calculating the total taxable income of the household. In this connection the plaintiffs add that the tax
credit for tax paid abroad, laid down by the convention for income from work as employees, reduces the
double taxation slightly but does not eliminate it.
18.
All the member states which have submitted observations agree that the provisions of the Franco-
German convention, which are regarded by the plaintiffs in the main proceedings as discriminatory and
contrary to art 48 of the Treaty, are in reality entirely compatible with it and that art 220 does not have
direct effect.
19.
Article 48, they maintain, does not prevent two member states from applying, in order to eliminate
double taxation on the income from employment of persons residing in one state and working in the other,
different tax criteria depending on whether the taxpayers are employed in the public or the private sector,
nor does it prevent such states from applying, for the same purpose, different tax criteria to public sector
employees of one of them, depending on whether the taxpayer is or is not a national of that state or
whether teachers residing in one state go to the other to work there for a period longer than two yes or
not.
Nor does art 48 preclude, in the framework of a convention between two states for the avoidance of
double taxation, the state where the taxpayer resides from taxing all his income, including that received in
the other member state, and granting him, in relation to the latter, a tax credit for tax paid abroad equal to
the amount of the national tax on the relevant income.
Several member states observe that the problem confronting the plaintiffs in the main proceedings
does not arise from discriminatory treatment under French tax law, but from the difference between tax
rates in the two countries, the rate being higher in Germany than in France. Some states draw the courts
attention to the repercussions which would ensue from a judgment interpreting art 48 as precluding the
provisions in question of the Franco-German convention, because all those provisions conform with the
model convention of 833 the Organisation for Economic Co-operation and Development (the OECD) for
the avoidance of double taxation, on which most bilateral conventions signed by the member states
among themselves are based.
Article 220, they consider, does not have direct effect because it is not sufficiently clear and
unconditional and does not confer upon individuals a right to the abolition of all double taxation within the
Community.
20.
The Commission begins with a detailed examination of the consequences of applying the Franco-
German convention to the tax situation of Mr and Mrs Gilly.
In France, where they reside, income tax is payable on the entire income of the couple (who constitute
a household for tax purposes) irrespective of where the income is received. The French scale of tax rates
and the French system of progressive tax are applied, and spouses cannot opt to be taxed separately.
When calculating allowances and deductions for family commitments, account is taken of the taxable
income in France, that is to say, the couples total income. In the present case, as the income received in
France is less than one half of the total income (Mr Gillys income being 45%), he ends up by paying
more income tax than if he were taxed separately.
In Germany, Mrs Gilly, whose income is approximately 55% of the couples total income, is not entitled
to the preferential scale for married couples, which is known as the Splittingtarif 7. She is automatically
deemed to be single because her husband does not reside in Germany. In her case, application of the
preferential scale would have resulted in reducing her tax liability in Germany because her income more
than one half of the couples total income.
7
Splitting consists in aggregating the spouses income and notionally attributing 50% to each. If the income of one
spouse is greater than that of the other, this system levels out the taxable income and reduces the progressive increase
in the scale of the tax.

On that basis, the Commission considers that in Germany Mrs Gillys liability to tax should have regard
to her marital status, so that her husbands income in France should be taken into account. That would
certainly ensure consistency in each state with regard to application of the progressive scale of its tax.
21.
 The Commission goes on to discuss art 20(2)(a)(cc) of the Franco-German convention in the light of
art 220 of the Treaty. It submits that art 220 imposes on the member states an obligation to act, namely to
enter into negotiations if necessary, but not an obligation to achieve a specific result, and that bilateral
conventions for the avoidance of double taxation meet the objective of art 220. The Commission adds
that, in its opinion, the machinery of the convention avoids double taxation and that Community law does
not prevent Mr and Mrs Gilly from being subject to a higher tax burden in so far as this is due to the higher
rate of tax in Germany.
22.
It concludes that the application of French law to the couples total income and of German law to Mrs
Gillys income in Germany constitutes, by reason of the way in which her marital status is taken into
account, an obstacle which is incompatible with the principles governing the freedom of movement of
workers.

Preliminary observations
23.
Before discussing the questions referred by the Tribunal Administratif, Strasbourg, I think it is
necessary to comment on the following points. (A) The courts jurisdiction, in the framework of the
procedure laid down by art 177, to 834 give a ruling on the compatibility of the Franco-German convention
with Community law. (B) The admissibility of the fourth question from the national court, asking whether
the principle of freedom of movement for workers, as embodied in the Treaty, is contravened by tax rules
which are liable to affect the choice made by teachers in the contracting states as to whether to work on a
more or less long-term basis in another state. (C) The provisions of Community law applicable to the main
proceedings in relation to the prohibition of discrimination on the ground of nationality with regard to
freedom of movement for workers.
24.
I shall examine these questions in that order.
A. The court s jurisdiction, in the framework of the procedure laid down by art 177, to give a
ruling on the compatibility of the Franco-German convention with Community law
25.
On this point I take the view that, just as the court does not, within the framework of these
proceedings, have jurisdiction to give a ruling on the compatibility of a national measure with Community
law (see the judgment in USSL No 47 di Biella v Istituto Nazionale per lAssicurazione contro gli Infortuni
sul Lavoro (Inail) Case C-134/95 [1997] ECR I-195 (para 17)), nor can it give a ruling on the compatibility
with Community law of the provisions of an international treaty concluded by two member states for the
avoidance of double taxation.
Moreover, as the treaty in question is a bilateral convention on a matter such as direct taxation which
is outside the Communitys competence and which is regulated exclusively by the member states, the
court could not even undertake to interpret it.
However, the rules governing freedom of movement for workers are within the ambit of Community law
and the parties to the main proceedings are the tax authority of one of the member states and a
Community national who has exercised her freedom of movement and who considers herself a victim of
discrimination arising from the provisions of a bilateral convention for the avoidance of double taxation. In
those circumstances, the court may give the national court guidance on the interpretation of the area of
Community law which will enable it to give judgment in the main proceedings (see the judgment in
Matteucci v Communaut franaise de Belgique Case 238/87 [1988] ECR 5589 (para 14)).
I therefore propose that the Court of Justice reformulate the questions from the national court.
B. The admissibility of the fourth question from the national court, asking whether art 48
precludes a provision of the nature of art 16 of the Franco-German convention
26.
Under art 16 of the Franco-German convention, teachers habitually residing in one of the contracting
states who, in the course of a temporary stay not exceeding two years in the other state, receive
remuneration for teaching in a university, college, school or other teaching establishment are taxable on
that remuneration only in the first state.
Mr and Mrs Gilly consider that this provision gives rise to discrimination in tax matters between
teachers, contrary to the freedom to move between France and Germany, because teachers who reside
in one of those states and teach in the other for a limited period have the status of frontier workers without
having to reside or work in the frontier area, and thus pay less tax than teachers who, 835like Mrs Gilly,
reside in France and decide to teach in Germany for more than two years.
27.
In my opinion, the plaintiffs purpose in advancing this argument, which is echoed in the national
courts observation, in the order for reference, that art 16 may influence the choice made by teachers in
the contracting states as to whether to work on a more or less long-term basis in another state, is not so
much to support a particular interpretation of the principle of freedom of movement for workers within the
Community, as to obtain from the court a ruling against art 16 of the Franco-German convention, which
has not been applied to them.
28.
In actual fact, it is not clear from the order for reference whether Mr and Mrs Gilly were in the situation
to which art 16 refers. According to the national court, Mr Gilly teaches in the French state education
system and it does not appear that he has taught in Germany for a period either greater or less than two
years. His wife teaches in the German state education system, in a post which she has held continuously
for more than two years and again there is no indication that she previously taught in France for less than
two years while resident in Germany. In reply to my question at the hearing, the plaintiffs confirmed these
points.
29.
There is a body of settled case law concerning the respective roles of the national courts and the
Court of Justice in the framework of the co-operation procedure provided for by art 177 of the Treaty.
According to that case law, the national court, which alone has direct knowledge of the facts of the case,
is in the best position to assess, having regard to the particular features of the case, whether a
preliminary ruling is necessary to enable it to give judgment and the relevance of the questions to be put
to the court8, whereas it is a matter for the Court of Justice, in order to determine whether it has
jurisdiction, to examine the conditions under which the case is referred to it by the national court. The
spirit of co-operation which must prevail in the preliminary ruling procedure requires the national court to
have regard to the function entrusted to the Court of Justice, which is to assist in the administration of
justice in the member states and not to deliver opinions on general or hypothetical questions (see the
judgment in Robards v Insurance Officer Case 149/82 [1983] ECR 171 (para 17)).
8
See the judgments in Pigs Marketing Board v Redmond Case 83/78 [1978] ECR 2347 (para 25), Durighello v Istituto
Nazionale Della Previdenza Sociale (INPS) Case C-186/90 [1991] ECR I-5773 (para 8) and Dias v Director da
Alfndega do Porto Case C-343/90 [1992] ECR I-4673 (para 15).

30.
Taking account of this function, the Court of Justice has held that it could not give a ruling on a
question from a national court where the request for interpretation or examination of the validity of a rule
of Community law bore no relation to the actual nature of the case or to the subject matter of the main
proceedings9, or where it was asked to give a ruling on a hypothetical problem without having before it the
matters of fact or law necessary to give a useful answer to the questions submitted to it (see the judgment
in Meilicke v ADV/ORGA AG Case C-83/91 [1992] ECR I-4871 (paras 3233)).
9
See the judgments in Salonia v Poidomani Case 126/80 [1981] ECR 1563 (para 6), Durighello [1991] ECR I-5773 (para
9), Criminal proceedings against Ruiz Bernldez Case C-129/94 [1996] All ER (EC) 741, [1996] ECR I-1829 (para
7) and Kontogeorgas v Kartonpak AE Case C-104/95 [1996] ECR I-6643 (para 11).

The Court of Justice has held that it

836
does not have jurisdiction to reply to questions of interpretation which are submitted within the
framework of procedural devices arranged by the parties in order to induce the Court to give its
views on certain problems of Community law which do not correspond to an objective requirement
inherent in the resolution of a dispute. (See the judgment in Foglia v Novello Case 244/80 [1981]
ECR 3045 (para 18).)
On this point the Court of Justice has added that it is essential for the national court to explain the
reasons why it considers that a reply to its questions is necessary for it to give judgment, in order to
enable the Court of Justice to ascertain whether the interpretation of Community law which is sought is
related to the actual nature and subject matter of the main proceedings. If it appears that the question
raised is manifestly irrelevant for the purposes of deciding the case, the court must declare that there is
no need to proceed to judgment (see the judgment in Dias v Director da Alfndega do Porto Case C-
343/90 [1992] ECR I-4673 (paras 1920)).
31.
In the light of this case law, and as it appears that neither of the plaintiffs before the national court has
been in the situation covered by art 16 of the Franco-German conventionbecause neither of them has
worked for less than two years in the contracting state other than the one in which they resideI consider
that a reply from the court interpreting the Community law on freedom of movement for workers in the
Community would be of no use to the national court in giving judgment in the case before it. For this
reason I propose that the court rule the fourth question inadmissible.
C. The prohibition of discrimination on the ground of nationality with regard to the freedom of
movement of workers
32.
The national court seeks interpretation of art 6 of the Treaty, which prohibits all discrimination on the
grounds of nationality. In this connection it must be observed that the Court of Justice has consistently
held that this principle applies independently only to situations governed by Community law in respect of
which the Treaty lays down no specific prohibition of discrimination (see the judgments in Scholz v Opera
Universitaria di Cagliari Case C-419/92 [1994] ECR I-505 (para 6) and Criminal proceedings against
Skanavi Case C-193/94 [1996] All ER (EC) 435, [1996] ECR I-929 (para 20)).
However, with regard to freedom of movement for workers, the principle of equal treatment has been
given specific application by art 48(2) of the Treaty, which provides for the abolition of all discrimination as
regards employment, remuneration and other conditions of work. Furthermore, art 7 of Regulation
1612/68, which provides that a worker who is a national of a member state is to enjoy, in the territory of
other member states, the same social and tax advantages as national workers, is a specific expression of
the general principle of non-discrimination against workers by means of tax measures.
In this case, therefore, it is unnecessary to refer to art 6 of the Treaty in order to reply to the questions
from the Tribunal Administratif, Strasbourg.
 Reformulation and discussion of the questions referred to the Court of Justice
33.
So far as the other questions are concerned, I conclude from the statements of the national court in
the order for reference that, in requesting a preliminary ruling under art 177, it seeks clarification of the
following points: (1) Firstly, whether art 220 of the Treaty is directly applicable. (2) Secondly, 837whether
the provisions of art 13(1) and (5) and art 14 of the Franco-German convention are contrary to art 48 of
the Treaty and art 7(2) of Regulation 1612/68 in so far as they lay down criteria for the taxation of income
from work as employees in one or the other state: by reference to the place where the work is done;
according to whether the worker fulfils the conditions for being regarded as a frontier worker for the
purpose of the convention; according to whether the worker receives public service remuneration and, if
so, depending on whether he is a national of the state other than the one paying the remuneration,
without at the same time being a national of the latter. (3) Thirdly, whether the provisions of art 20(2)(a)
(cc) of the Franco-German convention are contrary to arts 48 and 220 of the Treaty and art 7(2) of
Regulation 1612/68 in so far as, to avoid double taxation of the income from employment of residents of
one of the contracting states where such income is taxed in the other state, a procedure is laid down
whereby a tax credit equal to the amount of the national tax on the relevant income is granted,
irrespective of the amount of tax paid in the other state which, under certain circumstances, may mean
that the taxpayer pays more income tax than he would have had to pay if the income in question had
been earned in the state of residence, or more than he would have had to pay if the income had been
earned in the other contracting state but taxed only in the state of residence.

First question: the possible direct effect of art 220, second indent, of the Treaty
34.
The national courts question here concerns the direct applicability of this provision, which states that
the member states are to enter, so far as is necessary, into negotiations with each other with a view to
securing for the benefit of their nationals the abolition of double taxation within the Community.
I think the very wording of this provision shows that it is not sufficiently clear and unconditional for
direct effect to be attributed to it and that it thus cannot give rise to rights in favour of individuals which the
national courts must safeguard. I agree with the Commission that the second indent of art 220 imposes
on the member states an obligation to act, namely to enter into negotiations so far as is necessary, but
not an obligation to achieve a specific result.
35.
I consider that the Court of Justices case law relating to the first indent of art 220which requires the
member states, so far as is necessary, to enter into negotiations with each other with a view to securing
for the benefit of their nationals the protection of persons and the enjoyment and protection of rights under
the same conditions as those accorded by each state to its own nationalsshould be applied to the
second indent. The court has stated that art 220 is not intended to lay down a legal rule directly applicable
as such, but merely defines a number of matters on which the member states are to enter into
negotiations with each other so far as is necessary (see the judgment in Ministre public v Mutsch Case
137/84 [1985] ECR 2861 (para 11)).
Furthermore the provision, as worded, does not lay down an absolute obligation but leaves the
member states a wide discretion to decide whether to enter into negotiations. France and Germany
exercised that discretion when, in 1959, they signed the Franco-German convention, which abrogated the
1934 convention on the same subject, and also when they amended it by successive additional protocols
in 1969 and 1989. By signing such a convention, France and 838 Germany shared between them the
power to tax income received by their respective residents which is earned in or paid by the other
contracting state.

Second question: equal treatment of workers as regards taxation and the provisions of the
Franco-German convention laying down criteria for the taxation of income from
employment
36.
To reply to this question, I must consider in some detail art 13(1) and (5) and art 14 of the Franco-
German convention, which lay down the criteria for the taxation of income from employment which may
affect Mrs Gillys tax situation directly or indirectly and, by extension, that of members of her household in
so far as, because they reside in France, their aggregate income is taken into account under French law
and they cannot be taxed separately.
37.
In referring the questions, the national court appears to assume that Mrs Gilly must be deemed to
have the status of a frontier worker because she is domiciled in the French frontier area and her place of
work is situated in the German frontier area.
The term frontier worker is defined in art 13(5)(a) as covering those who work in the frontier area of
one contracting state and have their permanent home in the frontier area of the other contracting state to
which they normally return each day. Frontier area is defined in art 13(5)(b) and (c).
Under art 13(5)(a), the income received by a frontier worker in the state of employment is taxed only in
the state where he or she resides. Mrs Gilly seeks the status of a frontier worker and, in that way, her
salary would not be taxed in Germany, but in France, where tax rates are lower.
38.
However, a systematic examination of the contested provisions of the convention shows that, firstly,
the general rule is laid down by art 13(1) that income from employment is taxable in the state of
employment, and taxation of the income of frontier workers only in the state where they reside is an
exception to that general rule.
Secondly, art 14(1) of the convention lays down a lex specialis applying to public service
remuneration. That is to say, remuneration paid by a state, a local or regional authority or a public entity.
This lex specialis consists in turn of a general rule and an exception. The general rule is set out in the first
sentence of art 14(1) and states that, if the employer is a legal person governed by public law and if the
employee resides in the other state, remuneration and retirement pensions paid in consideration for
administrative or military services are taxable in the state paying them. The exception is given in the
second sentence, which is to the effect that this above-mentioned rule does not apply where the
remuneration is paid to persons who have the nationality of the other state but are not at the same time
nationals of the first state, in which case the remuneration is taxable only in the state where they are
resident.
For example, the remuneration paid by the German state to a person residing in France is taxable in
Germany. If, in the same situation, the recipient possesses French nationality, the remuneration will be
taxable in France. If, like Mrs Gilly, the recipient is a national of both states, the right of the paying state to
tax the remuneration prevails.
39.
Community law does not require the member states to lay down a different criterion for taxing the
income from employment of frontier workers from that applying to other employed persons, state
employees and persons treated as such, nor is this customary when concluding conventions for the 839
avoidance of double taxation10. Moreover, a recent comparative study of the bilateral tax conventions
concluded by France shows that, where a special tax regime is laid down for frontier workers, income is
not always taxed in the state where the worker resides11.
10
According to the report on frontier workers presented to the French National Assembly on 22 January 1997 (Rapport
dinformation No 3307) by Mr D Jacquat, a deput, on behalf of the Committee for Cultural, Family and Social Affairs,
apart from Greece, which has no common frontiers with other member states, and France, only six of the conventions
concluded between the other member states contain provisions relating to the remuneration of frontier workers. These
are the convention of 11 April 1967 between Germany and Belgium, that of 4 October 1954 between Germany and
Austria, that of 29 June 1981 between Austria and Italy, that of 26 October 1993 between Spain and Portugal, that of 16
November 1973 between Sweden and Denmark, and that of 27 June 1993 between Sweden and Finland.
11
Ibid p 31. Among the bilateral tax conventions signed by France, only five contain special provisions for frontier workers.
These are the conventions with Germany, Belgium, Spain, Italy and Switzerland. The first four provide that income is to
be taxed in the state of residence whereas, in the case of Switzerland, for the canton of Geneva, where a frontier area
has not been demarcated, income is taxable at the place where the work is done and, for the other cantons, at the place
of residence.

40.
Nor does Community law prohibit the member states from laying down rules for frontier workers which
differ from those applying to workers in general, employees of public authorities and persons treated as
such, provided that this does not entail discrimination against workers who are nationals of other member
states, by comparison with national workers. Furthermore, such a practice is not unknown in Community
law itself.
For example, in Council Regulation (EEC) 1408/71 on the application of social security schemes to
employed persons, to self-employed persons and to members of their families moving within the
Community, as amended and updated by Council Regulation (EEC) 2001/83 (OJ 1983 L230 p 6) adopted
by the EC Council pursuant to its obligation under art 51 of the Treaty, workers in general are, apart from
certain exceptions, subject to the social security legislation of the state where they work, whereas civil
servants and persons treated as such are subject to the social security legislation of the state to which the
authority employing them belongs. For frontier workers, the regulation has a whole catalogue of special
measures in chapters such as those relating to sickness insurance, accidents at work and occupational
illnesses, unemployment and family benefits.
41.
There is a special criterion for the taxation of public service remuneration in art 19 of the 1992 version
of the OECDs model double taxation convention on income and on capital (the model convention), which
forms the basis of the bilateral conventions for the avoidance of double taxation concluded by the
member states12.
12
As a recent example, I may cite art 19 of the Convention between the Kingdom of Spain and the French Republic for the
avoidance of double taxation and the prevention of fraud and evasion with regard to taxes on income and assets
(Madrid, 10 October 1995; (1997) Boletin Oficial del Estado 140), which contains the same rule for the taxation of public
remuneration as art 14 of the Franco-German convention.

The member states which have submitted written observations in this case point out that this provision
is based on the comity of nations and mutual respect for the sovereignty of each state. The commentary
on art 19 of the model convention points out that the principle of giving the exclusive right of taxation to
the paying state is contained in so many of the existing conventions between OECD member states that it
can be said to be already internationally 840 accepted. With regard to the exception, the same
commentary adds that it originates from the Vienna Convention on Diplomatic Relations (Vienna, 18 April
1961; Cmnd 2565) and the Vienna Convention on Consular Relations (Vienna, 24 April 1963; TS 14
(1973); Cmnd 5219), under which the host state has the right to tax the remuneration of the members of
certain categories of the staff of foreign consular and diplomatic missions who reside permanently in that
state or who are nationals thereof.
42.
 It is not disputed that, as it stands at present, Community law does not regulate direct taxation. The tax
provisions in arts 95 to 99 of the EC Treaty relate only to indirect taxation. In secondary legislation, very
few Community measures dealing with the direct taxation of natural persons have been adopted hitherto.
Of these, the only one which is binding is Council Directive (EEC) 77/799 concerning mutual assistance
by the competent authorities of the member states in the field of direct taxation (OJ 1977 L336 p 15). The
others are merely a proposal for a Council directive concerning the harmonisation of income taxation
provisions with respect to freedom of movement for workers within the Community (OJ 1980 C21 p 6) and
Commission Recommendation (EC) 94/79 on the taxation of certain items of income received by non-
residents in a member state other than that in which they are resident (OJ 1994 L39 p 22).
43.
Therefore direct taxation still fails within the competence of the member states, as the court has found
in its past judgments on the subject. However, the states must exercise that competence consistently with
Community law, refraining from any overt or covert discrimination on grounds of nationality 13. In the
present case it is necessary to ascertain whether, in signing the convention, the member states exercised
that competence consistently with Community law and, in particular, with the provisions governing the
freedom of movement of workers within the Community.
13
See the judgments in EC Commission v France Case 270/83 [1986] ECR 273 (para 24), Finanzamt Kln-Altstadt v
Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225 (para 21), Wielockx v Inspecteur der
Directe Belastingen Case C-80/94 [1995] All ER (EC) 679, [1995] ECR I-2493 (para 16), Asscher v
Staatssecretaris van Financien Case C-107/94 [1996] All ER (EC) 757, [1996] I-3089 (para 36) and Futura
Participations SA v Administration des Contributions Case C-250/95 [1997] ECR I-2471 (para 19).

44.
By signing a bilateral convention for the avoidance of double taxation, the two states concerned agree
to limit their fiscal sovereignty and to waive part of it. It is not surprising that, when sharing the power to
tax the income of their respective residents from employment in the other state, they use criteria such as
those in art 13(1) and (5) and art 14 of the Franco-German convention, namely the place where the work
is performed, fulfilment of the conditions necessary for being deemed a frontier worker, whether the
employer is a public law entity and, if so, whether the taxpayer is a national of one state or the other. Nor
are there many other possibilities.
In my opinion, those criteria, which have the purpose only of determining the power to tax certain
income, are neutral with regard to freedom of movement for workers because, in the two states
concerned, they do not, in tax matters, treat workers of other member states less favourably than or
differently from their own nationals who are in the same situation.
45.
It cannot be branded as discriminatory to provide that the remuneration of an employed person is
taxable in the state where he works or in the state where he resides or by the state paying the
remuneration, even if, in the case of public service remuneration, it is necessary in the final analysis to
refer to the 841 criterion of the recipients nationality in order to decide which of the two states is to tax it,
because for this purpose that is as neutral a criterion as the others. At the stage of applying the criteria for
taxation, it is only necessary to decide in each case which of the two states is to tax the income. Then the
state where the taxpayer resides and to which it falls to tax his total income will apply the legal procedure
agreed upon with the other state in order, when calculating the tax payable under its own law, to avoid
taxing once again the income from employment which has already been taxed in the other state. I shall
discuss the legal procedure applied by France in more detail when examining the third question.
46.
There is no doubt that, once the state which is to tax the income from employment has been
determined, the tax payable will vary depending on which state it is. In Mrs Gillys case, the tax she pays
in Germany on the income from her employment is more than she would pay if she received the income
in France or if, although received in Germany, it were taxable in France.
47.
However, that difference does not arise from the taxation criteria in the Franco-German convention,
but from German tax law, which specifies a higher rate of tax on such income than the rate in France,
while in addition the tax systems and the progressive nature of the tax differ considerably in the two
states.
Such differences will remain until the Council adopts directives for harmonising fiscal provisions
applying to direct taxes. As art 100a(2) of the Treaty provides that art 100a(1), which regulates the
adoption of measures by a qualified majority, is not to apply to, inter alia, fiscal provisions, this is unlikely
to happen in the short or medium term.

Third question: whether art 20(2)(a)(cc) of the Franco-German convention is contrary to arts
48 and 220 of the Treaty and art 7(2) of Regulation 1612/68
48.
This question refers to the procedure laid down in the convention, whereby France avoids taxing once
again income originating in Germany received by French residents after having borne tax in Germany.
The procedure applies, inter alia, to remuneration covered by the general rule of taxation in the state
where the work is performed, in accordance with art 13(1), and public service remuneration within the
meaning of art 14. Under art 20(2)(a)(cc), the procedure consists, firstly, in aggregating the income from
employment earned in Germany with the taxable income calculated in accordance with French tax law,
and then granting a tax credit in respect of the tax paid abroad, equal to the amount of the French tax on
the relevant income.
49.
Mrs Gilly contends that the application of this provision to the circumstances of her own case gives
rise to discrimination on grounds of nationality, contrary to art 48 of the Treaty and art 7(2) of Regulation
1612/68, because the income tax which she has to pay in France, which is where all the income of her tax
household is taxed, is greater than she would have had to pay if it had fallen to France to tax her income
received in Germany.
That would be the case if, for example, she had the status of a frontier worker or if she possessed
French nationality only and not dual nationality. She considers that the above-mentioned provisions mean
that a worker who exercises the right to freedom of movement should not be penalised by a greater tax
liability than if that right had not been exercised.
50.
As interpreted by the court in accordance with art 48(2) of the Treaty, freedom of movement for
workers entails the abolition of all discrimination 842 based on nationality between workers of the member
states, particularly with regard to remuneration. The principle of equal treatment with regard to
remuneration would be rendered ineffective if it could be undermined by discriminatory national provisions
on income tax. For that reason the Council laid down, in art 7 of Regulation 1612/68, that workers who
are nationals of a member state are to enjoy, in the territory of another member state, the same tax
advantages as national workers (see the judgment in Biehl v Administration des Contributions du Grand-
Duch de Luxembourg Case C-175/88 [1991] STC 575, [1990] ECR I-1779 (paras 1112)).
The court has also repeatedly held that art 48 of the Treaty prohibits not only overt discrimination by
reason of nationality but also all covert forms of discrimination which, by the application of other
distinguishing criteria, lead in fact to the same result (see the judgments in EC Commission v
Luxembourg Case C-111/91 [1993] ECR I-817 (para 9) and Scholz [1994] ECR I-505 (para 9)).
51.
To find the existence of discrimination within the meaning of Community law in Mrs Gillys case, it
would be necessary to show that, by applying to her the provision in question, France, where she resides
and where all her income is taxable together with that of her husband, treats her less favourably in
respect of tax than a worker in the same situation who has French nationality. In this connection it is
irrelevant that Mrs Gilly is also a national of the state to which she attributes discriminatory treatment
because, as the court has observed, any Community national who, irrespective of his place of residence
and his nationality, has exercised the right to freedom of movement for workers and who has been
employed in another member state, falls within the scope of art 48 of the Treaty (see the judgment in
Scholz [1994] ECR I-505 (para 9)).
52.
It must also be borne in mind that, according to the definition given by the court, discrimination can
arise only through the application of different rules to comparable situations or the application of the same
rule to different situations (see the judgment in Finanzamt Kln-Altstadt v Schumacker Case C-279/93
[1995] All ER (EC) 319, [1995] ECR I-225 (para 30)).
53.
When must a worker of French nationality be deemed to be in a situation comparable, for tax
purposes, to that of Mrs Gilly? It all depends on what is meant by comparable situation. Of course, it is
possible to proceed by assuming that all the circumstances are the same and changing only the
nationality of the person who is claimed to be the victim of discrimination. On that basis, the equivalent for
comparison with Mrs Gilly would be a French worker who is not at the same time a German national, who
receives remuneration from the German state for work in Germany, who resides in France, who is married
and, who in relation to his or her spouse, contributes to the family income in the same proportion as Mrs
Gilly. In practice, the public service remuneration of such a person would be taxable in France, there
would be no reason for using the procedure for avoiding double taxation and the amount paid by the
couple as income tax in France would be less than the total tax paid by Mrs Gilly in Germany and her tax
household in France. That is the reasoning of the plaintiffs in the main proceedings to show that Mrs Gilly
suffers discrimination.
54.
However, there is another method, which consists in making the comparison with a worker of French
nationality whose remuneration is taxed in Germany under art 13(1) of the Franco-German convention
and who resides in France, where he or she is liable to income tax on all his or her income. Such 843
would be the case, for example, of a married worker employed in a private undertaking in Germany,
residing in France and with a spouse who resides and works in France, provided that the two spouses
contribute to the family income in the same proportion as Mr and Mrs Gilly. In that situation, even after
being taxed in Germany, the income from employment will be aggregated with the French taxable income
and then a credit will be given for the tax paid in Germany, equal to the French tax appropriate to that
income.
55.
In my opinion, there are a number of reasons why the second method is the correct one for
determining whether a provision such as that in question here is contrary to art 48: (a) Firstly, because,
before applying art 20(2)(a)(cc) of the convention, which lays down the procedure whereby France, as the
state to which it falls to tax the total income of its residents, avoids the double taxation of income which
has already been taxed in Germany, it is necessary, as a first and essential step, to determine, by
applying the taxation criteria of the convention, which of the two states should tax income from
employment. As I have mentioned when discussing the national courts second question, these criteria,
which are the place where the work is done, fulfilment of the conditions for being deemed a frontier
worker, whether the remuneration is public service remuneration and, if so, whether the taxpayer is a
national of one state or the other, are neutral from the viewpoint of the Community law relating to freedom
of movement for workers. (b) Secondly, because, after the different items of the taxpayers income have
been taxed in one or the other state, the state of residence must apply the procedure agreed upon with
the other state in order to avoid a further tax charge on that income. The result will be more or less
favourable to the employee, depending on the national tax legislation and the personal and family
situation of each taxpayer.
56.
If the second method is used, a perfect parallel is obtained from the outset: income from work as an
employee received in Germany by a person resident in France and taxed in Germany, the person being in
the same family situation as Mrs Gilly. In this way it can be shown that her tax treatment in France is the
same as that of a French person in the same situation.
If the first method is used, however, the starting-point is the result obtained after application of the
system for avoiding double taxation. In other words, Mrs Gilly pays more income tax in France than she
would if she had French nationality, all else being equal, and from this finding one works back to the
beginning.
57.
Does Mrs Gilly suffer covert discrimination because, as the Commission maintains, her family situation
is not taken into account either in Germany, since her husband does not reside there and consequently
they cannot have the benefit of the preferential scale for married couples, or in France, since the fact that
the couples family situation is taken into account there affects only Mr Gillys income?
58.
I consider that the reply in this case must be in the negative. As the court has previously observed, in
relation to direct taxes, the situations of residents and non-residents are not, as a rule, comparable and
the fact that a member state does not grant a non-resident certain tax benefits which it grants to a
resident is not, as a rule, discriminatory since those two categories of taxpayer are not in a comparable
situation. Accordingly art 48 of the Treaty does not in principle preclude the application of rules of a
member state under which a non-resident working as an employed person in that state is taxed more
heavily on his or her income than a resident in the same employment.
844
To reach that conclusion, the court reasoned as follows ([1995] All ER (EC) 319, [1995] ECR I-225
(paras 3233)):
(32) Income received in the territory of a member state by a non-resident is in most cases only
a part of his total income, which is concentrated at his place of residence. Moreover, a non-
residents personal ability to pay tax, determined by reference to his aggregate income and his
personal and family circumstances, is more easy to assess at the place where his personal and
financial interests are centred. In general, that is the place where he has his usual abode.
Accordingly, international tax law, and in particular the Model Double Tax Convention on Income
and on Capital of the Organisation for Economic Cooperation and Development, recognises that in
principle the overall taxation of taxpayers, taking account of their personal and family
circumstances, is a matter for the State of residence.
The court added that
(33) The situation of a resident is different in so far as the major part of his income is normally
concentrated in the state of residence. Moreover, that state generally has available all the
information needed to assess the taxpayers overall ability to pay, taking account of his personal
and family circumstances.
59.
The only cases in which the court has not followed this reasoning are those involving workers who did
not receive significant income in their state of residence so that their tax liability in that country was not
sufficient for their personal and family situation to be taken into account, and who received most of their
income, and almost all their household income, in a different member state. In such cases, there is
discrimination against the worker in so far as his personal and family situation is not taken into account in
the state of residence or the state of employment (the court found discrimination of this kind in the
judgments in the Schumacker case, Wielockx v Inspecteur der Directe Belastingen Case C-80/94 [1995]
All ER (EC) 679, [1995] ECR I-2493 and Asscher v Staatssecretaris van Financien Case C-107/94 [1996]
All ER (EC) 757, [1996] I-3089).
However, this cannot be said to apply to Mrs Gilly who, although as an individual she receives almost
all her income in Germany in the form of her salary, has her personal and family situation taken into
account in France, where her salary is aggregated with the taxable income of her tax household and
where she is granted the tax reliefs, rebates and deductions laid down by French tax law.
60.
The remaining point to consider is whether the legal procedure used by France to avoid double
taxation under art 20(2)(a)(cc) of the Franco-German convention is an obstacle, prohibited by art 48 of the
Treaty, to the freedom of movement of workers. For this purpose the question is whether, although it
applies irrespective of nationality, that procedure in fact has an adverse effect on persons who have
exercised their freedom of movement by treating them less favourably than those who have not done so.
In that connection the court has observed that
the provisions of the Treaty relating to freedom of movement for persons are intended to
facilitate the pursuit by Community citizens of occupational activities of all kinds throughout the
 Community, and preclude [national] 845measures which might place Community citizens at a
disadvantage when they wish to pursue an economic activity in the territory of another member
state. (See the judgment in Union Royale Belge des Socits de Football Association ASBL v
Bosman Case C-415/93 [1996] All ER (EC) 97, [1995] ECR I-4921 (para 94).)
61.
Here again, the reply must be in the negative because if Germany were simply to make a sufficient
reduction in its present rate of tax on income from employment, the procedure which is now criticised for
adversely affecting workers exercising their freedom of movement would have the opposite effect. If the
German tax rate were lower than the French rate on the same income, with Mrs Gilly receiving, in respect
of tax paid abroad, a tax credit equal to the amount of the French tax on the relevant income, the tax
credit would be greater than the tax already paid in Germany and she would end up by paying less tax on
that income than if she had received it in France or if, having been received in Germany, it had been
taxed in France in accordance with the criteria described above. A similar situation could arise if France
were to decide to increase its rate of tax on such income to a rate higher than the German rate.
Whether the consequences of a provision such as that in question here are unfavourable to workers
depends, in the final analysis, on the tax rates charged in each member state on certain income and I
therefore consider that those consequences are too uncertain and indirect for the provision to be
regarded as being capable of deterring a worker from exercising his or her freedom of movement
between the two member states in question (see the judgment in Esso Espaola v Administracin de la
Comunidad Autnoma de Canarias Case C-134/94 [1995] ECR I-4223 (para 24)).
62.
Finally, Mrs Gilly contends that the second indent of art 220 of the Treaty precludes her German
income which has already been taxed in Germany from being taken into account in France in the
calculation of the taxable income of her household for income tax purposes, because the French tax
credit for tax paid abroad does not take account of the exact amount of tax paid in Germany, which
means that the procedure does not avoid double taxation, but only reduces
63.
Articles 23A and 23B of the model convention for the avoidance of double taxation envisage two
methods for achieving this result which, for the sake of clarity, are worth examining in some detail.
64.
Article 23A lays down the rules for the exemption with progression method. Where a resident of a
contracting state receives income which, according to the convention, may be taxed in the other state, the
first state exempts the income from tax. However, it may take the income into account when determining
the rate applying to the income received in the state of residence. In other words, although the income
received in the source state does not form part of the taxable income in the state of residence, the total
tax payable in the latter is increased by the effect of progressive taxation in the country of residence.
Article 23B covers the full credit or ordinary credit method. Unlike the method described above, this
aggregates the taxpayers total income, whether of national origin or not, in the taxable income of the
state of residence. Where a resident of one state receives income which, according to the convention,
may be taxed in the other state, the first state deducts from the tax charged on the residents income an
amount equal to the income tax paid in the other state However, such deduction may not exceed that part
of its own income tax, 846calculated before deduction, which is attributable to the income taxable in the
other state.
65.
The legal procedure used by France, since the entry into force of the 1989 Additional Protocol, to
avoid the further taxation in France of income from employment which has already been taxed in
Germany under the convention, consists in granting a tax credit equal to the amount of tax on the relevant
income. The method previously used consisted in exempting income already taxed in Germany, although
it was taken into account for determining the rate applying to income received in France. In the final
analysis these two methods, which at first appear to be different, produce the same result
66.
The object of a bilateral double taxation convention is to prevent income which is taxed in one state
from being taxed again in the other. The object is not, therefore, to ensure that the tax paid by the
taxpayer in one state is not more than would be payable in the other, regardless of where the income was
received and whatever its specific source. In actual fact, bilateral taxation conventions, in accordance with
art 24 of the model convention, lay down a rule of equal treatment between nationals and citizens of the
other state.
67.
In no case, therefore, does the right to freedom of movement for workers confer upon them a right to
be granted, in their state of residence, the tax status which is most favourable to them in particular. They
have a right only to the same tax treatment as nationals of that state. Moreover, I consider, this result is
achieved by using the procedure laid down in the provision in question.

Conclusion
68.
In the light of the foregoing, I propose that the court should: (A) rule that the fourth question referred by
the Tribunal Administratif, Strasbourg, is inadmissible; (B) reply as follows to the other questions:
(1) The second indent of art 220 of the EC Treaty does not have direct effect.
 (2) Articles 48 and 220 of the Treaty and art 7 of Council Regulation (EEC) 1612/68 on freedom
of movement for workers within the Community must be interpreted as meaning that they do not
preclude a convention between two member states for the avoidance of double taxation from laying
down criteria for the taxation of income from employment in one or the other state: by reference to
the place where the work is done; or according to whether the worker fulfils the conditions for being
regarded as a frontier worker; or according to whether the worker receives public service
remuneration and, if so, depending on whether he is a national of the state other than the one
paying the remuneration, without at the same time being a national of the latter.
(3) Articles 48 and 220 of the Treaty and art 7 of Regulation 1612/68 must be interpreted as
meaning that they do not preclude a convention between two member states for the avoidance of
double taxation from laying down that, in one of them, double taxation is avoided by means of a
procedure of the kind referred to in art 20(2)(a)(cc) of the Franco-German convention, under which
income received in Germany by a resident in France, which has already been taxed in Germany, is
taken into consideration for the purpose of calculating the taxpayers taxable income in France by
granting a tax credit in respect of the tax paid abroad, equal to the amount of the French tax on the
relevant income.

847
12 May 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By judgment of 10 October 1996, received at the Court of Justice of the European Communities on 11
October 1996, the Tribunal Administratif (the Administrative Court), Strasbourg, referred to the Court of
Justice for a preliminary ruling under art 177 of the EC Treaty six questions on the interpretation of arts 6,
48 and 220 of that Treaty.
2.
Those questions were raised in the context of various proceedings brought by Mr and Mrs Gilly
against the Directeur des Services Fiscaux du Bas-Rhin (the Director of Tax Services for the dpartement
of Bas-Rhin) concerning the calculation of their personal income tax for the tax years 1989, 1990, 1991,
1992 and 1993, pursuant to the Convention between the French Republic and the Federal Republic of
Germany for the avoidance of double taxation and the establishment of rules for mutual legal and
administrative assistance in the field of income and wealth tax and in the field of business tax and land tax
(the Franco-German convention) (Paris, 21 July 1959), as amended by the 1969 and 1989 Additional
Protocols (Bonn, 9 June 1969 and 28 September 1989, respectively).
3.
Mr and Mrs Gilly reside in France, near the German border. Mr Gilly, a French national, teaches in a
state school in France. Mrs Gilly, who is a German national having also acquired French nationality by
marriage, teaches in a state primary school in Germany, in the frontier area.
4.
As regards taxation of income from employment, art 13(1) of the Franco-German convention lays
down the following basic principle:
Subject to the provisions of the following paragraphs, income from dependent work shall be
taxable only in the Contracting State in which the personal activity in respect of which it is received
is carried out. In particular, salaries, wages, pay, gratuities or other emoluments shall be deemed to
constitute income from dependent work, together with all similar benefits paid or awarded by
persons other than those referred to in Article 14.
5.
Article 13(5)(a) contains an exception to the rule that income is to be taxed in the country where the
work is carried out in the case of frontier workers, who are to be taxed in their state of residence:
By way of exception to paragraphs 1, 3 and 4, income from dependent work earned by persons
who work in the frontier area of one Contracting State and who have their permanent home in the
other Contracting State, to which they normally return each day, shall be taxable only in that other
State.
6.
However, taxpayers receiving remuneration and pensions from the public sector are in principle
taxable, under art 14(1) of the convention, in the paying state:
Salaries, wages and similar remuneration, and retirement pensions, paid by one of the
Contracting States, by a Land or by a legal person of that State or Land governed by public law to
natural persons resident in the other state in consideration for present or past administrative or
military services shall be taxable only in the first State. However, that provision shall not be
applicable where the remuneration is paid to persons having the nationality of the other State
without being at the same time nationals of the first State; 848in such cases, the remuneration
shall be taxable only in the State in which such persons are resident.
7.
 Article 16 of the convention lays down a special rule applicable to teachers who are temporarily
resident, under which taxation remains with the state of their original employment:
Teachers resident in one of the Contracting States who, in the course of a period of temporary
residence not exceeding two years, receive remuneration in respect of teaching in a university,
college, school or other teaching establishment in the other State shall be taxable in respect of that
remuneration only in the first State.
8.
As regards double taxation, art 20(2)(a)(cc) of the convention, as amended by the 1989 Additional
Protocol, reads as follows:
2. Double taxation of persons resident in France shall be avoided in the following manner: (a)
Profits and other positive income arising in the Federal Republic and taxable there under the
provisions of this Convention shall also be taxable in France where they accrue to a person
resident in France. The German tax shall not be deductible for calculation of the taxable income in
France. However, the recipient shall be entitled to a tax credit to be set against the French tax
charged on the taxable amount which includes that income. That tax credit shall be equal (cc)
for all other income, to the amount of the French tax on the relevant income. This provision shall
apply in particular to the income referred to in Articles 13(1) and (2) and 14.
9.
It appears from the national courts judgment that under the effective rate rule, the amount of the tax
credit is equal to the amount of the net income taxed in Germany multiplied by the ratio between the tax
actually payable in respect of the total net income taxable under the French legislation and the amount of
that income.
10.
The national court states that the tax credit to be set against the French tax may prove to be less than
the tax actually paid in Germany because of the greater progressivity of the tax scale in the latter country.
French frontier workers taxed both in Germany on income received there and in France on their total
income after deduction of the tax credit mentioned above may thus be taxed more heavily than persons
receiving exactly the same income but only in France.
11.
In the present case, the public service remuneration received by Mrs Gilly in Germany in 1989, 1990,
1991, 1992 and 1993 was taxed in Germany in accordance with art 14(1) of the convention, because she
is a German national. That remuneration was also taxed in France pursuant to art 20(2)(a) of the
convention. However, under art 20(2)(a)(cc), the fact that it was taxed in Germany entitled her to a tax
credit equal to the amount of the French tax on the relevant income.
12.
In the proceedings which they initiated before the Tribunal Administratif, Strasbourg, on 8 July 1992
and 21 July 1995, Mr and Mrs Gilly submitted that application of the above-mentioned provisions of the
convention had led to unjustified, discriminatory and excessive taxation which was incompatible with arts
6, 48 and 220 of the EC Treaty. They therefore sought to be discharged from the contested tax obligations
and to have the amount of the tax which they 849 claimed to have been wrongly charged reimbursed to
them by the tax authorities.
13.
The Tribunal Administratif took the view that the dispute before it turned on the interpretation to be
given to arts 6, 48 and 220 of the Treaty. It therefore stayed proceedings and requested a preliminary
ruling by the Court of Justice on the following questions:
(1) Is the principle of freedom of movement for workers, as embodied in the Treaty of Rome
and the implementing legislation, contravened by a tax regime, applicable to frontier workers, of the
kind provided for by the Franco-German Convention, in so far as the latter lays down taxation
arrangements which are different for people whose remuneration is paid by a public entity as
compared with those whose remuneration is paid by private persons and as a result is liable to
have an impact on access to posts in the public or private sectors depending on residence in one
State or another?
(2) Is a rule under which a frontier worker receiving remuneration from a State or an agency
thereof governed by public law is taxable in that State, whereas, if the frontier worker has the
nationality of the other state but is not at the same time a national of the first State, his
remuneration is taxable in the State where the frontier worker resides, compatible with the principle
of freedom of movement and the abolition of all discrimination on grounds of nationality?
(3) Is a tax provision which lays down for frontier workers employed by persons governed by
public law a tax regime which differs according to whether they are nationals only of that State or
have dual nationality compatible with Article 7 [now Article 6] of the Treaty?
(4) Is the principle of freedom of movement for workers, as embodied in the Treaty, contravened
by tax rules which are liable to affect the choice made by teachers in the Contracting States as to
whether to work on a more or less long-term basis in another State having regard to the
differences, based on the duration of employment, in the tax regimes of the States in question?
(5) Must the objective of abolishing double taxation laid down in Article 220 of the Treaty be
regarded, in view of the time which the Member States have had to implement it, as now having the
status of a directly applicable rule under which double taxation may no longer take place? Is the
 objective of avoiding double taxation assigned to the Member States by Article 220 contravened by
a tax convention under which the tax regime applicable to frontier workers of States party to the
convention varies according to their nationality and the public or private nature of the post held?
Does a tax credit regime applicable to a household living in one State which does not take into
account the exact amount of the tax paid in another State but only a tax credit, which may be lower,
meet the objective assigned to the Member States of abolishing double taxation?
(6) Must Article 48 be interpreted as meaning that nationals of a Member State who are frontier
workers in another Member State may not, by reason of a tax credit mechanism of the type
provided for by the Franco-German Convention, be taxed more heavily than persons whose
occupational activity is pursued in their State of residence?
850

The fifth question

14.
By its fifth question, which should be addressed first, the national court raises the question whether the
second indent of art 220 of the Treaty is directly applicable.
15.
As the court has already held in Ministre public v Mutsch Case 137/84 [1985] ECR 2861 (para 11), art
220 is not intended to lay down a legal rule directly applicable as such, but merely defines a number of
matters on which the member states are to enter into negotiations with each other so far as is necessary.
Its second indent merely indicates the abolition of double taxation within the Community as an objective of
any such negotiations.
16.
Although the abolition of double taxation within the Community is thus included among the objectives
of the Treaty, it is clear from the wording of that provision that it cannot itself confer on individuals any
rights on which they might be able to rely before their national courts.
17.
The answer to the fifth question must therefore be that the second indent of art 220 of the Treaty does
not have direct effect.

The first, second and fourth questions

18.
By its first, second and fourth questions, the national court wishes to know whether, on a proper
construction, art 48 of the Treaty precludes the application of provisions such as those in arts 13(5)(a),
14(1) and 16 of the convention, under which the tax regime applicable to frontier workers differs
depending on whether they work in the private sector or the public sector and, where they work in the
public sector, on whether or not they have only the nationality of the state of the authority employing them
or not, and the regime applicable to teachers differs depending on whether their residence in the state in
which they are teaching is for a short period or not.
Applicability of art 48 of the Treaty
19.
The first matter to be determined is whether a situation such as that of the applicants in the main
proceedings falls within the Treaty rules relating to freedom of movement for workers.
20.
At the hearing, the French government expressed the view that Mrs Gilly has not exercised in France
the rights conferred upon her by art 48 of the Treaty, since she works in her state of origin, namely
Germany.
21.
It need merely be pointed out here that Mrs Gilly has acquired French nationality by her marriage and
works in Germany whilst residing in France. She must therefore be considered in France as a worker
exercising her right to freedom of movement, as guaranteed by the Treaty, in order to work in a member
state other than that in which she resides. The circumstance that she has retained the nationality of the
state in which she is employed in no way affects the fact that, for the French authorities, she is a French
national working in another member state (see, to the same effect, Gullung v Conseils de lordre des
avocats du barreau de Colmar et de Saverne Case 292/86 [1988] ECR 111 (para 12)).
22.
A situation such as that in issue in the main proceedings must therefore be held to fall within the scope
of art 48 of the Treaty.
Compatibility of the fiscal connecting factors with art 48 of the Treaty
23.
Whilst abolition of double taxation within the Community is, as stated in para 16 above, one of the
objectives of the Treaty, it must none the less be noted that, apart from the Convention on the elimination
of double taxation in 851 connection with the adjustment of profits of associated enterprises (Brussels, 23
July 1990; OJ 1990 L225 p 10), no unifying or harmonising measure for the elimination of double taxation
has yet been adopted at Community level, nor have the member states yet concluded any multilateral
convention to that effect under art 220 of the Treaty.
24.
 The member states are competent to determine the criteria for taxation on income and wealth with a
view to eliminating double taxationby means, inter alia, of international agreementsand have
concluded many bilateral conventions based, in particular, on the model conventions on income and
wealth tax drawn up by the Organisation for Economic Co-operation and Development (the OECD).
25.
That is the context in which the convention concluded between the French Republic and the Federal
Republic of Germany applies a number of connecting factors for the purpose of allocating jurisdiction
between the contracting parties as to the taxation of income from dependent work.
26.
Although as a general rule workers are taxed in accordance with art 13(1) of the convention in the
state in which the state in which the personal activity in respect of which the income is received is carried
out, frontier workers are taxed in their state of residence, under art 13(5)(a).
27.
However, the first sentence of art 14(1) of the convention provides that taxpayers receiving public
service remuneration are in principle to be taxed in the paying state. There is also an exception to the
latter rule (the paying state rule) in the second sentence of art 14(1), under which remuneration paid to a
person having the nationality of the other state without being at the same time a national of the first state
is taxable in the taxpayers state of residence.
28.
In addition, art 16 of the convention lays down a special connecting rule applicable to teachers
habitually resident in one of the contracting states who, in the course of a short period of residence not
exceeding two years in the other contracting state, receive remuneration for teaching in the latter state.
Taxpayers in that category are taxed in the state of original employment.
29.
It is thus clear that arts 13(1) and (5)(a), 14(1) and 16 of the convention lay down different connecting
factors depending on whether the taxpayer is a frontier worker or not, is a teacher in short-term residence
or not, or is employed in the private or the public sector. Taxpayers in the latter category are in principle
taxed in the paying state unless they have the nationality of the other contracting state without being at
the same time nationals of the first, in which case they are taxed in their state of residence.
30.
Although the criterion of nationality appears as such in the second sentence of art 14(1) for the
purpose of allocation of fiscal jurisdiction, such differentiation cannot be regarded as constituting
discrimination prohibited under art 48 of the Treaty. It flows, in the absence of any unifying or harmonising
measures adopted in the Community context under, in particular, the second indent of art 220 of the
Treaty, from the contracting parties competence to define the criteria for allocating their powers of
taxation as between themselves, with a view to eliminating double taxation.
31.
Nor, in the allocation of fiscal jurisdiction, is it unreasonable for the member states to base their
agreements on international practice and the model convention drawn up by the OECD, art 19(1)(a) of the
1994 version of which in particular provides for recourse to the paying state principle. According to the
commentary on that article, that principle is justified by the rules of international courtesy and mutual
respect between sovereign states and is 852 contained in so many of the existing conventions between
OECD member countries that it can be said to be already internationally accepted.
32.
In the present case, the first sentence of art 14(1) of the convention reproduces the tenor of art 19(1)
(a) of the OECD model convention. It is true that under the second sentence the paying state principle is
abandoned where the taxpayer has the nationality of the other contracting state without being at the same
time a national of the first state, but the same type of exception, based at least in part on the criterion of
nationality, is found in art 19(1)(b) of the model convention in cases where the services are rendered in
the other contracting state and the taxpayer is a resident of that state who (i) is a national of that State; or
(ii) did not become a resident of that State solely for the purpose of rendering the services.
33.
In any event, even if the second sentence of art 14(1), the legality of which is challenged by Mrs Gilly,
were to be ignored, her tax position would remain unchanged because the paying state principle would
still have to be applied to her income earned in Germany from teaching in the state education system.
34.
Nor is it established in the present case that the choice of the paying state as the state competent to
tax income earned in the public sector can of itself be to the disadvantage of the taxpayers concerned. As
has been pointed out by the governments of the member states which have submitted observations and
by the Commission, whether the tax treatment of the taxpayers concerned is favourable or unfavourable
is determined not, strictly speaking, by the choice of the connecting factor but by the level of taxation in
the competent state, in the absence of any Community harmonisation of scales of direct taxation.
35.
The answer to the first, second and fourth questions must therefore be that, on a proper construction,
art 48 of the Treaty does not preclude the application of provisions such as those in arts 13(5)(a), 14(1)
and 16 of the convention, under which the tax regime applicable to frontier workers differs depending on
whether they work in the private sector or the public sector and, where they work in the public sector, on
whether or not they have only the nationality of the state of the authority employing them, and the regime
applicable to teachers differs depending on whether their residence in the state in which they are teaching
is for a short period or not.

The third question

36.
By its third question, the national court wishes to know whether, on a proper construction, art 7 of the
EEC Treaty, now art 6 of the EC Treaty, precludes the application of a provision such as that contained in
the second sentence of art 14(1) of the convention, under which the tax regime applicable to frontier
workers employed in the public sector in a member state differs depending on whether they have the
nationality of that state or not.
37.
It is settled case law that art 6 of the Treaty, which lays down the general principle of the prohibition of
discrimination on grounds of nationality, applies independently only to situations governed by Community
law in respect of which the Treaty lays down no specific prohibition of discrimination (see esp Mora
Romero v Landesversicherungsanstalt Rheinprovinz Case C-131/96 [1997] ECR I-3659 (para 10)).
38.
In the field of freedom of movement for workers, the prohibition of discrimination has been specifically
implemented and embodied in art 48 of the Treaty and by acts of secondary legislation including, in
particular, Council 853 Regulation (EEC) 1612/68 on freedom of movement for workers within the
Community (OJ S Edn 1968 (II), p 475).
39.
It follows from the answer given to the first and second questions that a situation such as that in the
main proceedings falls within art 48 of the Treaty. It is therefore unnecessary to rule on the interpretation
of art 6 of the Treaty.

The sixth question

40.
By its sixth question, the national court wishes to know whether, on a proper construction, art 48 of the
Treaty precludes the application of a tax credit mechanism such as that provided for in art 20(2)(a)(cc) of
the convention.
41.
The purpose of the tax credit mechanism set up by art 20(2)(a)(cc) of the convention, which is based
on the arrangements envisaged for that purpose in the OECD model convention, is to avoid the double
taxation of French residents in receipt, in Germany, of profits or other income taxable in both Germany
and France.
42.
That mechanism involves first aggregating the income earned from work in Germany within the
taxable basis calculated in accordance with the French legislation and then giving a tax credit in respect
of the tax paid in Germany of the same amount, in particular for the income referred to in art 14 of the
convention, as that of the French tax on the relevant income. The latter amount is calculated on the basis
of the proportion of the total net income taxable in France constituted by the net income taxable in
Germany.
43.
It further appears from the case file that Mrs Gillys personal and family circumstances were not taken
into account when calculating the tax on her income from employment in Germany during the tax years in
issue, whereas those circumstances were taken into account in the calculation of the tax payable in
France, when determining the total household income and when granting various tax rebates and
reductions.
44.
In the submission of the applicants in the main proceedings, the tax credit mechanism in issue
penalises those who have exercised their freedom of movement in that it allows a degree of double
taxation to remain. In the present case, as a result of the greater degree of progressivity of the German
tax scale as compared with the French tax scale and having regard to the proportion of the total income of
the tax household taxable in France constituted by Mrs Gillys income from employment, the amount of
the tax credit is always lower than that of the tax actually paid in Germany. Moreover, because Mrs Gillys
personal and family circumstances are not taken into account in Germany, whereas they are taken into
account in France when calculating the tax on the total income, the tax credit accorded in the state of
residence is lower than the amount of tax actually paid in the state of employment, taking account of the
rebates and reductions accorded in the first state.
45.
Double taxation could be fully avoided, the applicants in the main proceedings submit, only by a tax
credit equal to the amount of tax charged in Germany.
46.
Here, it must be stressed, as has been done by the Advocate General at para 66 of his opinion, that
the object of a convention such as that in issue is simply to prevent the same income from being taxed in
each of the two states. It is not to ensure that the tax to which the taxpayer is subject in one state is no
higher than that to which he or she would be subject in the other.
47.
 However, it is common ground that any unfavourable consequences entailed in the present case by
the tax credit mechanism set up by the bilateral 854 convention, as implemented in the context of the tax
system of the state of residence, are the result in the first place of the differences between the tax scales
of the member states concerned, and, in the absence of any Community legislation in the field, the
determination of those scales is a matter for the member states
48.
Furthermore, as has been observed by the French, Belgian, Danish, Finnish, Swedish and UK
governments, if the state of residence were required to accord a tax credit greater than the fraction of its
national tax corresponding to the income from abroad, it would have to reduce its tax in respect of the
remaining income, which would entail a loss of tax revenue for it and would thus be such as to encroach
on its sovereignty in matters of direct taxation.
49.
As regards the effect on the amount of the tax credit of the fact that the taxpayers personal and family
circumstances are taken into account in the state of residence but not in the state of employment, it must
be pointed out that the disparity derives from the fact that, in relation to direct taxes, the situations of
residents and of non-residents are not, as a rule, comparable, since income received in the territory of a
state by a non-resident is in most cases only a part of his total income, which is concentrated at his place
of residence (see the judgment in Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER
(EC) 319, [1995] ECR I-225 (paras 3132)).
50.
Nor is that observation inapplicable to a situation such as that of Mrs Gilly; although her individual
income from employment is received in Germany, it is none the less aggregated within the basis for
assessing the personal income tax payable by her tax household in France, where she is therefore
entitled to the tax advantages, rebates and deductions provided for in the French legislation. The German
tax authorities, however, were not obliged to take account of her personal and family circumstances in
such a situation.
51.
The applicants in the main proceedings also submit that application of art 20(2)(a)(cc) of the
convention in their case gives rise to discrimination on grounds of nationality, contrary to art 48 of the
Treaty, since, if Mrs Gilly possessed only French nationality and not dual nationality, her tax situation
would be governed by art 13(5)(a) of the convention, under which the income of frontier workers is taxed
in their state of residence.
52.
Here, it must be borne in mind that the purpose of art 20(2)(a)(cc) of the convention is to prevent
double taxation arising, in a case such as that with which the main proceedings are concerned, from
taxation in Germany of the income received there by Mrs Gilly, in accordance with art 14(1) of the
convention, when the total household income, including that income from Germany, is taxable in France.
53.
As is clear from the answer to the first, second and fourth questions, the fact that in allocating powers
of taxation between them the contracting parties have chosen various connecting factors, in particular
nationality with regard to public service remuneration received in the state other than the state of
residence, cannot in itself constitute discrimination prohibited by Community law.
54.
In the light of the foregoing, the answer to the sixth question must be that, on a proper construction, art
48 of the Treaty does not preclude the application of a tax credit mechanism such as that provided for in
art 20(2)(a)(cc) of the convention.
855

Costs
55.
The costs incurred by the French, Belgian, Danish, German, Italian, Netherlands, Finnish, Swedish
and UK governments and by the European Commission, which have submitted observations to the Court
of Justice, are not recoverable. Since these proceedings are, for the parties to the main proceedings, a
step in the proceedings pending before the national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Tribunal
Administratif, Strasbourg, by judgment of 10 October 1996, hereby rules: (1) The second indent of art 220
of the EC Treaty does not have direct effect. (2) On a proper construction, art 48 of the Treaty does not
preclude the application of provisions such as those in arts 13(5)(a), 14(1) and 16 of the Convention
between the French Republic and the Federal Republic of Germany for the avoidance of double taxation
(Paris, 21 July 1959), as amended by the 1969 and 1989 protocols (Bonn, 9 June 1969 and 28
September 1989, respectively), under which the tax regime applicable to frontier workers differs
depending on whether they work in the private sector or the public sector and, where they work in the
public sector, on whether or not they have only the nationality of the state of the authority employing them,
and the regime applicable to teachers differs depending on whether their residence in the state in which
they are teaching is for a short period or not. (3) On a proper construction, art 48 of the Treaty does not
preclude the application of a tax credit mechanism such as that provided for in art 20(2)(a)(cc) of the said
convention.

856
 [1998] All ER (EC) 857

Viscido and others v Ente Poste Italiane

(Joined cases C-52/97, C-53/97 and C-54/97)

EUROPEAN COMMUNITY; Other European Community: EMPLOYMENT; Contract of service

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (FOURTH CHAMBER)
JUDGES RAGNEMALM (PRESIDENT OF THE CHAMBER), MURRAY (RAPPORTEUR) AND IOANNOU
ADVOCATE GENERAL JACOBS
29 JANUARY, 19 FEBRUARY, 7 MAY 1998
European Community State aid Conditions of employment National law relieving Post Office of
general obligation not to use fixed-term employment contracts Whether national provisions constituting
illegal state aid EC Treaty, art 92(1).

The applicants were recruited and employed on the basis of fixed-term contracts by the Italian Post
Office, which was exempt, under domestic legislation, from the general prohibition on the use of such
contracts. The applicants brought proceedings against their employer, claiming that their contracts should
be deemed to have been converted into contracts of indeterminate duration. They contended that the
exemption granted to the Post Office constituted an illegal grant of state aid contrary to art 92 1 of the EC
Treaty since it relieved the Post Office of a burden applicable to other undertakings under the general law.
The Pretura Circondariale stayed the proceedings and referred to the Court of Justice of the European
Communities for a preliminary ruling, inter alia, the question whether a national provision relieving only
one undertaking of the obligation of complying with the generally applicable legislation concerning fixed-
term contracts constituted state aid within the meaning of art 92(1) of the Treaty.
1
Article 92, so far as material, is set out at p 860 a, post

Held Only advantages granted directly or indirectly through state resources were to be considered as
state aid within the meaning of art 92(1). The distinction made in that provision between aid granted by a
Member State and aid granted through State resources did not signify that all advantages granted by a
state, whether financed through state resources or not, constituted aid, but was intended merely to bring
within that definition both advantages granted directly by the state and those granted by a public or
private body designated or established by the state. In the instant case, the non-application of generally
applicable legislation concerning fixed-term employment contracts to a single undertaking did not involve
any direct or indirect transfer of state resources to that undertaking. It followed that a provision of the kind
at issue did not constitute a means of directly or indirectly granting an advantage through state resources
(see p 863 f to j and p 864 b, post).
Openbaar Ministerie v van Tiggele Case 82/77 [1978] ECR 25, Sloman Neptun Schiffahrts v
Seebetriebsrat Bodo Ziesemer der Sloman Neptun Schiffahrts Joined cases C-72-73/91 [1993] ECR I-
887 and Kirsammer-Hack v Sidal Case C-189/91 [1993] ECR I-6185 applied.

857

Notes
For state aids which are incompatible with the common market, see 51 Halsburys Laws (4th edn) paras
702709.
For the EC Treaty, art 92(1), see 50 Halsburys Laws (4th edn) 295.

Cases cited
Kirsammer-Hack v Sidal Case C-189/91 [1993] ECR I-6185.
Openbaar Ministerie v van Tiggele Case 82/77 [1978] ECR 25.
Sloman Neptun Schiffahrts v Seebetriebsrat Bodo Ziesemer der Sloman Neptun Schiffahrts Joined cases
C-72-73/91 [1993] ECR I-887.

Reference
By three orders of 3 February 1997, the Pretura Circondariale (the District Magistrates Court), Trento,
referred to the Court of Justice of the European Communities for a preliminary ruling under art 177 of the
EC Treaty four questions (set out at p 862 j to p 863 b, post) on the interpretation of arts 92(1) and 93 of
that Treaty. Those questions arose in three sets of proceedings brought by Epifanio Viscido, Mauro
Scandella and others and Massimiliano Terragnolo and others, against the Italian Post Office, by which
they were employed. Written observations were submitted on behalf of: the Italian government, by
Professor U Leanza, Head of the Legal Service in the Ministry of Foreign Affairs, acting as agent, and D
del Gaizo, Avvocato dello Stato; the German government, by E Rder, Ministerialrat in the Federal
Ministry of Economic Affairs, and B Kloke, Oberregierungsrat in the same ministry, acting as agents; and
the European Commission, by F Santaolalla, Principal Legal Adviser, and D Triantafyllou, of its Legal
Service, and E Altieri, a national civil servant seconded to that service, acting as agents. Oral
observations were made by the Italian government, represented by D Del Gaizo, and the Commission,
represented by D Triantafyllou and L Pignataro, of its Legal Service. The language of the case was Italian.
The facts are set out in the opinion of the Advocate General.

19 February 1998.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
The issue raised by the Pretura Circondariale di Trento (the Magistrates Court, Trento) is whether a
national rule providing for the recruitment of staff by the Italian Post Office under fixed-term contracts, by
way of derogation from the general rule of Italian law that employment contracts should be of
indeterminate duration, constitutes state aid notifiable under the last sentence of art 93(3) of the EC
Treaty.
2.
Under Italian law, employment under fixed-term contracts is permitted only in a number of specified
exceptional cases. Article 1 of Law No 230 of 18 April 1962 provides that, subject to certain exceptions
specified in the Law, an employment contract is to be considered to be of indeterminate duration (a
number of additional exceptions were provided for the Law No 56 of 28 February 1987).
3.
However, art 9(21) of Decree-Law No 510 of 1 October 1996, converted into Law No 608 of 28
November 1996, laying down urgent measures in relation to work of social utility, provides:
Workers employed from 1 December 1994 under a fixed-term contract by Poste Italiane shall
have a right of priority, in accordance with the 858 contractual provisions and those of a specific
agreement with the trade unions, in the event of staff being taken on for an indeterminate period by
Poste Italiane for posts of the same level and/or involving the same duties until 31 December 1996;
the workers concerned must give notice of their wish to exercise that right by 30 November 1996.
Recruitment of staff under fixed-term contracts of employment by Poste Italiane, from the date on
which it was set up until 30 June 1997, shall not give rise to employment relationships of
indeterminate duration and shall lapse upon the expiry date of each contract.
4.
The above provision is linked to the transformation of the Italian Post and Telecommunications
Administration into a public undertaking with effect from 1 January 1994. Under art 6(2) of Law No
71/1994 the staff of the Post and Telecommunications Administration became employees of Poste Italiane
under private law contracts. According to the order for reference, the purpose of the second sentence of
art 9(21) was to lay down a transitional period at the end of which employment relationships were brought
into line with the private sector system.
5.
The applicants in the main proceedings, Mr Epifanio Viscido, Mr Mauro Scandella and Mr
Massimiliano Terragnolo, brought proceedings against Poste Italiane complaining that, since 1 January
1994, the undertaking had responded to staff shortages by recruiting workers on fixed-term contracts.
They contended that recruitment on that basis should be regarded as having given rise to employment
relationships of indeterminate duration. They argued that the disputed provision, in so far as it relieved
Poste Italiane of a burden applicable to other undertakings under the general law, entailed the grant of
state aid contrary to arts 92 and 93 of the Treaty.
6.
Against that background the national court has put the following questions to the Court of Justice:
(1) whether a legal provision which relieves a particular public economic entity from the
obligation of complying with the generally applicable legislation concerning fixed-term employment
contracts falls within the scope of aid granted by a Member State or through State resources in
any form whatsoever;
(2) whether, if question (1) is answered in the affirmative, an aid of that kind should be subject to
the preliminary examination procedure under Article 93(3) of the Treaty;
(3) whether, where that procedure has not been followed, the prohibition of an aid of that kind
can be regarded as directly applicable within the domestic law of the Italian State;
(4) whether, in the event of question (3) being answered in the affirmative, such a prohibition
may be relied on in a dispute between the public economic entity and an individual who complains
of failure to apply to him the general legislation concerning fixed-term employment in order to
secure conversion of his employment relationship into one of indeterminate duration and/or
compensation for damage.
7.
The applicants have not submitted written or oral observations to the court. The German and Italian
governments and the European Commission take the view that the contested provision does not involve
the grant of state aid within the meaning of art 92(1) of the Treaty. I share that view.
 859
8.
Article 92(1) provides:
Save as otherwise provided in this Treaty, any aid granted by a Member State or through State
resources in any form whatsoever which distorts or threatens to distort competition by favouring
certain undertakings or the production of certain goods shall, in so far as it affects trade between
Member States, be incompatible with the common market.
9.
It is clear from the case law of the court that a measure constitutes aid only if it involves the transfer of
state resources to an undertaking (or relief from financial obligations towards the state, such as tax or
social security charges).
10.
In Openbaar Ministerie v van Tiggele Case 82/77 [1978] ECR 25 the court held that the fixing by a
national authority of a minimum retail price for a product at the exclusive expense of consumers did not
constitute state aid since it did not entail the direct or indirect grant of state resources.
11.
Subsequently, in Sloman Neptun Schiffahrts v Seebetriebsrat Bodo Ziesemer der Sloman Neptun
Schiffahrts Joined cases C-72-73/91 [1993] ECR I-887, the court held that the partial non-application of
German employment law and social security law to foreign crews on ships flying the German flag did not
constitute state aid. Referring to its ruling in van Tiggele [1993] ECR I-887 (para 19) the court observed:
only advantages which are granted directly or indirectly through state resources are to be
regarded as state aid within the meaning of Article 92(1) of the EEC Treaty. The wording of this
provision itself and the procedural rules laid down in Article 93 of the EEC Treaty show that
advantages granted from resources other than those of the state do not fall within the scope of the
provisions in question. The distinction between aid granted by the state and aid granted through
state resources serves to bring within the definition of aid not only aid granted by the state, but also
aid granted by public or private bodies designated or established by the state.
12.
Turning to the German rules the court noted ([1993] ECR I-887 (para 21)):
The system at issue does not seek, through its object and general structure, to create an
advantage which would constitute an additional burden for the state or the abovementioned bodies,
but only to alter in favour of shipping undertakings the framework within which contractual relations
are formed between those undertakings and their employees. The consequences arising from this,
in so far as they relate to the difference in the basis for the calculation of social security
contributions, mentioned by the national court, and to the potential loss of tax revenue because of
the low rates of pay, referred to by the Commission, are inherent in the system and are not a
means of granting a particular advantage to the undertakings concerned.
13.
In Kirsammer-Hack v Sidal Case C-189/91 [1993] ECR I-6185 the court, applying its earlier rulings,
held that the exclusion of small businesses from a national system protecting workers against unfair
dismissal did not constitute state aid.
The court noted ([1993] ECR I-6185 (para 17)):
In the present case, the exclusion of a category of businesses from the protection system in
question does not entail any direct or indirect transfer of state resources to those businesses but
derives solely from the legislatures 860 intention to provide a specific legislative framework for
working relationships between employers and employees in small businesses and to avoid
imposing on those businesses financial constraints which might hinder their development.
14.
It is clear that the above case law applies to the present case. By relieving Poste Italiane for a
transitional period of the obligation to recruit staff under contracts of indeterminate duration the Italian
rules do not provide for any direct or indirect transfer of state resources to that undertaking. Instead by
suspending the ordinary rules of Italian employment law their purpose was to remove legal constraints
which might hinder the smooth transformation of the Italian Postal Administration into a public
undertaking.
15.
It might be argued that employment under fixed-term contracts could result in costs for the state in the
form of lost tax revenue or unemployment benefits. However, as the court put it in Sloman Neptun such
costs are inherent in the system and are not a means of granting a particular advantage to Poste
Italiane. In any event such costs are uncertain and unquantifiable since, in the absence of the flexibility
provided by the contested provision, Poste Italiane may not have employed, or may have employed
fewer, additional staff to cover short-term staff shortages.
16.
It might be asked why, given their potential effect on competition, art 92(1) does not cover all labour
and other social measures which by virtue of being selective in their impact might distort competition and
thereby have an equivalent effect to state aid. The answer is perhaps essentially a pragmatic one: to
investigate all such regimes would entail an inquiry on the basis of the Treaty alone into the entire social
and economic life of a member state (for a discussion of this issue see Davies Market Integration and
Social Policy in the Court of Justice (1995) Industrial Law Journal 49 esp at 58ff).
17.
I conclude therefore, in answer to the national courts first question, that a provision such as that in
issue does not entail the grant of aid within the meaning of art 92(1) of the Treaty. It is therefore
unnecessary to consider the national courts remaining questions.

Conclusion
18.
Accordingly, I am of the opinion that the questions referred by the Pretura Circondariale di Trento
should be answered as follows:
A national provision which relieves an undertaking from the obligation of complying with the
generally applicable legislation concerning the duration of employment contracts does not entail the
grant of state aid within the meaning of art 92(1) of the EC Treaty.

7 May 1998.

The COURT OF JUSTICE (Fourth Chamber)

delivered the following judgment.
1.
By three orders of 3 February 1997, received at the registry of the Court of Justice of the European
Communities on 7 February 1997, the Pretura Circondariale (the District Magistrates Court), Trento,
referred to the Court of Justice for a preliminary ruling under art 177 of the EC Treaty four questions on
the interpretation of arts 92(1) and 93 of that treaty.
2.
Those questions were raised in three sets of proceedings brought by Epifanio Viscido, Mauro
Scandella and others and Massimiliano Terragnolo and 861 others, against Ente Poste Italiane (the Italian
Post Office), by which they are employed.
3.
In the main proceedings, the plaintiffs object to the fact that Ente Poste Italiane recruited them on the
basis of fixed-term contracts. They claim that such contracts should be deemed to have been converted
into contracts of indeterminate duration.
4.
Under Italian law, employment under fixed-term contracts is permitted only in a number of specified
exceptional cases. Article 1 of Law No 230 of 18 April 1962 provides that, subject to certain exceptions
specified in the Law, an employment contract is to be considered to be of indeterminate duration. Article 5
of the same Law provides that a worker employed under a fixed-term contract is to enjoy all the
advantages accorded within the undertaking to workers employed under contracts of indeterminate
duration on a basis proportional to their length of service, provided that the result is not objectively
incompatible with the nature of a fixed-term contract.
5.
Law No 56 of 28 February 1987 laying down arrangements for the labour market introduced, in
relation to certain categories of workers, further exceptions to the principle that fixed-term contracts are
prohibited.
6.
Article 9(21) of Decree-Law No 510 of 1 October 1996, converted into Law No 608 of 28 November
1996, adopting urgent measures relating to work of social utility, income-support measures and social
welfare measures, provides:
Workers employed from 1 December 1994 under a fixed-term contract by Ente Poste Italiane
shall have a right of priority, in accordance with the contractual provisions and those of a specific
agreement with the trade unions, in the event of staff being taken on for an indeterminate period by
Ente Poste Italiane for posts of the same level and/or involving the same duties until 31 December
1996; the workers concerned must give notice of their wish to exercise that right by 30 November
1996. Recruitment of staff under fixed-term contracts of employment by Ente Poste Italiane, from
the date on which it was set up until 30 June 1997, shall not give rise to employment relationships
of indeterminate duration and shall lapse upon the expiry date of each contract.
7.
In the main proceedings, Ente Poste Italiane contends that, under art 9(21) of Decree-Law No 510, the
plaintiffs employment contracts are not subject to Laws Nos 230 and 56.
8.
For their part, the plaintiffs in the main proceedings maintain that the rules at issue constitute state aid,
which, as such, must be subject to the procedures and to the verification of compatibility provided for in
arts 92 and 93 of the Treaty.
9.
Considering that an interpretation of those provisions is needed for it to give judgment, the Pretura
Circondariale has stayed proceedings and referred the following four questions to the Court of Justice for
a preliminary ruling:
 (1) whether a legal provision which relieves a particular public economic entity from the
obligation of complying with the generally applicable legislation concerning fixed-term employment
contracts falls within the scope of aid granted by a Member State or through State resources in
any form whatsoever;
862
(2) whether, if question (1) is answered in the affirmative, an aid of that kind should be subject to
the preliminary examination procedure under Article 93(3) of the Treaty;
(3) whether, where that procedure has not been followed, the prohibition of an aid of that kind
can be regarded as directly applicable within the domestic law of the Italian State;
(4) whether, in the event of question (3) being answered in the affirmative, such a prohibition
may be relied on in a dispute between the public economic entity and an individual who complains
of failure to apply to him the general legislation concerning fixed-term employment in order to
secure conversion of his employment relationship into one of indeterminate duration and/or
compensation for damage.
10.
By order of the President of the Court of Justice of 25 February 1997, Cases C-52/97, C-53/97 and C-
54/97 were joined for the purposes of the oral and written procedure and judgment.
11.
By its first question the national court asks essentially whether a national provision relieving only one
undertaking of the obligation of complying with the generally applicable legislation concerning fixed-term
contracts constitutes state aid within the meaning of art 92(1) of the Treaty.
12.
The national court observes that, since Ente Poste Italiane is under no obligation to conclude
employment contracts of indeterminate duration, it enjoys a flexibility not available to other undertakings
operating in the same sector.
13.
In that connection, it should be pointed out that only advantages granted directly or indirectly through
state resources are to be considered as aid within the meaning of art 92(1). The distinction made in that
provision between aid granted by a Member State and aid granted through State resources does not
signify that all advantages granted by a state, whether financed through state resources or not, constitute
aid but is intended merely to bring within that definition both advantages which are granted directly by the
state and those granted by a public or private body designated or established by the state (see the
judgments in Openbaar Ministerie v van Tiggele Case 82/77 [1978] ECR 25 (paras 2425), Sloman
Neptun Schiffahrts v Seebetriebsrat Bodo Ziesemer der Sloman Neptun Schiffahrts Joined cases C-72-
73/91 [1993] ECR I-887 (para 19) and Kirsammer-Hack v Sidal Case C-189/91 [1993] ECR I-6185 (para
16)).
14.
In this case, it must be observed that non-application of generally applicable legislation concerning
fixed-term employment contracts to a single undertaking does not involve any direct or indirect transfer of
state resources to that undertaking.
15.
It follows that a provision of the kind at issue in the main proceedings does not constitute a means of
directly or indirectly granting an advantage through state resources.
16.
The answer to the first question must therefore be that a national provision which relieves only one
undertaking of the obligation of complying with the generally applicable legislation concerning fixed-term
employment contracts does not constitute state aid within the meaning of art 92(1) of the Treaty.
17.
In view of the answer given to the first question, it is unnecessary to answer the second, third and
fourth questions.
863

Costs
18.
The costs incurred by the Italian and German governments and by the European Commission, which
have submitted observations to the Court of Justice, are not recoverable. Since these proceedings are,
for the parties to the main proceedings, a step in the proceedings pending before the national court, the
decision on costs is a matter for that court.
On those grounds, the Court of Justice (Fourth Chamber), in answer to the questions referred to it by
the Pretura Circondariale, Trento, by order of 3 February 1997, hereby rules: a national provision which
relieves only one undertaking of the obligation of complying with the generally applicable legislation
concerning fixed-term employment contracts does not constitute state aid within the meaning of art 92(1)
of the EC Treaty.

864

[1998] All ER (EC) 865

 Coote v Granada Hospitality Ltd
(Case C-185/97)

EUROPEAN COMMUNITY; Social policy: EMPLOYMENT; Discrimination

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), RAGNEMALM AND SCHINTGEN (PRESIDENTS OF
CHAMBERS), MANCINI, MOITINHO DE ALMEIDA (RAPPORTEUR), MURRAY, EDWARD,
PUISSOCHET, HIRSCH, JANN AND IOANNOU
ADVOCATE GENERAL MISCHO
19 FEBRUARY, 2 APRIL, 22 SEPTEMBER 1998
European Community Equality of treatment for men and women Access to employment
Discrimination on grounds of sex Plaintiff bringing successful sex discrimination action against
employer Employer subsequently refusing to supply reference for plaintiff Whether member states
obliged under Community law to provide judicial protection against such refusal Sex Discrimination Act
1975 Council Directive (EEC) 76/207, art 6.

The plaintiff brought a sex discrimination claim against her employer, G Ltd, alleging that she had been
dismissed because of her pregnancy. That claim was settled and the plaintiffs employment ended by
mutual agreement. G Ltd subsequently failed to provide a reference for the plaintiff to one of the
employment agencies which she contacted when seeking new employment. The plaintiff therefore
brought a further claim against G Ltd, contending that she had been victimised by her former employers
refusal to supply a reference and alleging that the refusal was a reaction to the sex discrimination claim
which she had previously brought against the company. The Industrial Tribunal dismissed that action on
the grounds that the Sex Discrimination Act 1975 only prohibited retaliatory measures whose prejudicial
effect appeared during the employment relationship. The plaintiff appealed to the Employment Appeal
Tribunal, which stayed the proceedings and referred to the Court of Justice of the European Communities
for a preliminary ruling the question whether Council Directive (EEC) 76/207 on the implementation of the
principle of equal treatment of men and women as regards access to employment, vocational training and
promotion, and working conditions required member states to ensure judicial protection for workers
whose employer, after the employment relationship had ended, refused to provide references as a
reaction to proceedings brought to enforce compliance with the principle of equal treatment.

Held Under art 61 of Directive 76/207 everyone had the right to obtain an effective remedy in a
competent court against measures which they considered to interfere with the principle of equal treatment
and it was for the member states to ensure effective judicial control of compliance with the applicable
provisions of Community law and of national legislation intended to give effect to the rights for which the
directive provided. That principle would be deprived of an essential part of its effectiveness if the
protection provided by the directive did not cover measures which an employer might take as a reaction to
legal proceedings brought by an employee with the aim of enforcing compliance with the principle of
equal treatment. Moreover, if there was no legal remedy against such 865 measures, workers who
considered themselves to be the victims of discrimination might be deterred from pursuing their claims by
judicial process, which clearly would seriously jeopardise implementation of the aim pursued by the
directive. It followed that art 6 required member states to ensure judicial protection for workers whose
former employer refused to provide references as a reaction to proceedings brought to enforce
compliance with the principle of equal treatment (see p 877 c f and p 878 b e, post).
1
Article 6, so far as material, is set out at p 864 a, post

Johnston v Chief Constable of the Royal Ulster Constabulary Case C-222/84 [1986] 3 All ER 135,
[1986] ECR 1651 applied.

Notes
For equal treatment for men and women, see 52 Halsburys Laws (4th edn) paras 21112116.

Cases cited
Burton v British Rlys Board Case 19/81 [1982] 3 All ER 537, [1982] QB 1080, [1982] 3 WLR 387, [1982]
ECR 555, ECJ.
Garland v British Rail Engineering Ltd Case 12/81 [1982] 2 All ER 402, [1983] AC 751, [1982] 2 WLR 918,
[1982] ECR 359, ECJ.
Johnston v Chief Constable of the Royal Ulster Constabulary Case C-222/84 [1986] 3 All ER 135, [1987]
QB 129, [1986] 3 WLR 1038, [1986] ECR 1651, ECJ.
Kowalska v Freie und Hansestadt Hamburg Case C-33/89 [1990] ECR I-2591.
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135.
Marshall v Southampton and South West Hampshire Area Health Authority (No 2) Case C-271/91 [1993]
4 All ER 586, [1994] QB 126, [1993] 3 WLR 1054, [1993] ECR I-4367, ECJ.
Marshall v Southampton and South West Hampshire Health Authority (Teaching) Case 152/84 [1986] 2 All
ER 584, [1986] QB 401, [1986] 2 WLR 780, [1986] ECR 723, ECJ.
Meyers v Adjudication Officer Case C-116/94 [1995] All ER (EC) 705, [1995] ECR I-2131, ECJ.
P v S Case C-13/94 [1996] All ER (EC) 397, [1996] ECR I-2143, ECJ.
Von Colson v Land Nordrhein-Westfalen Case 14/83 [1984] ECR 1891.
Wagner Miret v Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911.

Reference
By order of 20 November 1996, the Employment Appeal Tribunal, London, referred to the court for a
preliminary ruling under art 177 of the EC Treaty two questions (set out at p 875 h to p 876 b, post) on the
interpretation of Council Directive (EEC) 76/207 on the implementation of the principle of equal treatment
for men and women as regards access to employment, vocational training and promotion, and working
conditions. Those questions arose in proceedings between Ms Coote and her former employer, Granada
Hospitality Ltd, concerning Granadas refusal to provide references to potential employers of Ms Coote.
Written observations were submitted on behalf of: Ms Coote, by D Rose, Barrister, instructed by P
Matthews, Principal Legal Officer, Equal Opportunities Commission; the United Kingdom government, by
L Nicoll, of the Treasury Solicitors Department, acting as agent, and S Moore, Barrister; the European
Commission, by M Wolfcarius and X Lewis, of its Legal Service, acting as agents. Oral observations were
made by Ms Coote, the UK government and 866 the Commission. The language of the case was English.
The facts are set out in the opinion of the Advocate General.

2 April 1998.

The Advocate General (J Mischo)

delivered the following opinion (translated form the French).
1.
In proceedings between Mrs Coote and Granada Hospitality Ltd, the Employment Appeal Tribunal has
referred the following questions to the Court of Justice of the European Communities for a preliminary
ruling:
(1) Does Council Directive 76/207/EEC of 9 February 1976 on the implementation of the
principle of equal treatment for men and women as regards access to employment, vocational
training and promotion, and working conditions require Member States to introduce into their
national legal systems such measures as are necessary to enable a complainant to pursue a claim
by judicial process where the following circumstances apply: (i) the complainant was employed by
the respondent; (ii) during her employment, the complainant brought a claim of sex discrimination
against the respondent which was compromised; (iii) following the end of her employment, the
complainant has made efforts to find full-time work, but has been unsuccessful; (iv) the respondent
has caused or contributed to the complainants difficulties in finding work by refusing to provide
references to potential employers when requested to do so; (v) the employers decision to refuse to
provide references was taken after the end of the complainants employment; (vi) the reason, or a
principal reason, for the employers decision to refuse to provide a reference to the complainant
was that she had previously brought a claim of sex discrimination against the respondent?
(2) Does Council Directive 76/207/EEC on the implementation of the principle of equal treatment
for men and women as regards access to employment, vocational training and promotion, and
working conditions require Member States to introduce into their national legal systems such
measures as are necessary to enable persons to pursue a claim by judicial process where the
circumstances set out under 1 above apply, save that: (i) the respondents decision to refuse to
provide a reference was taken before the termination of the complainants employment; but (ii) the
actual refusal or refusals to provide a reference took place after the termination of the
complainants employment?
2.
In short, the Employment Appeal Tribunal thus seeks to ascertain whether, on the basis of Council
Directive (EEC) 76/207 on the implementation of the principle of equal treatment for men and women as
regards access to employment, vocational training and promotion, and working conditions (OJ 1976 L39 p
40), member states are obliged to introduce the measures necessary to allow a worker to bring
proceedings against a former employer, if the worker considers that the motive for the employers refusal
to provide references of use in seeking new employment is that the employer resents having been made
the defendant in sex discrimination proceedings brought by the worker.
3.
That question from the Employment Appeal Tribunal requires the Court of Justice to look once again at
legislation with which it is familiar, as may be seen from the abundant case law on that legislation. Such
familiarity does not mean, however, that the answer is obvious, especially as an examination of the
reasons for the reference shows that the tribunals questions in fact cover two points, 867which I
consider it necessary to distinguish. I therefore find it useful to start by summarising the essential facts of
the proceedings before the national tribunal.
 The main proceedings and national law
4.
Mrs Coote was employed by Granada Hospitality Ltd (Granada) from December 1992 to September
1993. In 1993 she brought proceedings alleging sex discrimination on the ground that she had been
dismissed because of pregnancy. Those proceedings ended with a settlement between Mrs Coote and
her former employer. In July 1994 Mrs Coote tried to find a new job through employment agencies. Her
return to the employment market ran up against difficulties which she attributes to a refusal by her former
employer to provide a reference to one of the employment agencies; Granada denies this. Mrs Coote
then brought proceedings in the Industrial Tribunal, arguing that she had been wronged by Granadas
refusal to provide references. The proceedings were dismissed, the Industrial Tribunal holding that it had
no jurisdiction because the alleged discrimination complained of by Mrs Coote, assuming it to be proved,
had taken place after the end of the employment relationship.
5.
At this stage it is necessary to consider the provisions of United Kingdom law on the prohibition of all
discrimination against workers on grounds of their sex, and on the way in which those provisions have
been interpreted by the national courts. It is a feature of s 4 of the Sex Discrimination Act 1975 that any
less favourable treatment by an employer of an employee because the latter has brought proceedings
alleging sex discrimination is to be treated in all respects as sex discrimination. In other words, retaliation
by an employer against an employee who has relied on the prohibition of discrimination on grounds of sex
is itself regarded as an infringement of that prohibition.
6.
This is clearly a provision which gives a high level of protection. It ensures, by guaranteeing impunity,
that anyone who has dared to brave an employers wrath by alleging discrimination on grounds of sex will
not repent that boldness. It is likely, moreover, to dissuade employers from succumbing to the temptation
to retaliate. In view of that provision of national law, Mrs Coote should, assuming it to be established that
Granada had refused to give references and that its refusal constituted retaliation, have been able to
succeed in her claim, and the tribunal would not have had to consider the interpretation of the directive.
The reason why that was not so is that the national tribunal was faced with a problem of the scope ratione
temporis of the 1975 Act. That act, like the Race Relations Act 1976 which prohibits race discrimination,
has been interpreted by the courts in the United Kingdom as not applying to discrimination by an
employer against a former employee. Mrs Coote, who was dismissed in 1993, was therefore prevented
from relying on the Sex Discrimination Act 1975 to bring proceedings in 1994. And because it is uncertain
whether the directive has been properly implemented by a national law which prohibits discrimination only
on the commencement and during the course of the employment relationship, and allows the employee to
bring court proceedings only when the discrimination falls within that context, the national tribunal has
sought a preliminary ruling.
7.
The wording of the questions referred to the court confirms that the scope ratione temporis of the
prohibitions which the national legislature must lay down in order to implement the directive is central to
the dispute before the national tribunal. Those questions, although they relate solely to a situation in
which the reference was in fact refused after the termination of the employment, distinguish between the
alternative possibilities that the employers decision to 868 refuse to provide references was taken before
or after the termination of the employment of the person in need of the reference.

Does the provision of references fall within the scope of the directive?
8.
If the court is to provide the national tribunal with an answer that genuinely deals with its concerns, it
must first, I consider, address the question whether the provision of references falls within the scope of
the directive. I myself am convinced that it does, although I do not necessarily agree with everything the
European Commission has said on that point in its observations.
9.
The Commission submits that the provision of references falls within the scope of both arts 3 and 5 of
the directive.
10.
Article 3(1), it will be remembered, provides:
Application of the principle of equal treatment means that there shall be no discrimination
whatsoever on grounds of sex in the conditions, including selection criteria, for access to all jobs or
posts, whatever the sector or branch of activity, and to all levels of the occupational hierarchy.
The Commission argues, on the basis of Meyers v Adjudication Officer Case C-116/94 [1995] All ER (EC)
705, [1995] ECR I-2131, that, since it facilitates access to employment, the provision of references falls
within that definition.
11.
My own view is that, to be brought into play here, art 3 would have to be given an interpretation that is
not only highly creative but also of questionable expediency, since it is not necessary in order to bring the
provision of references within the scope of the directive. One is, I feel, on much surer ground if one brings
the provision of references within the scope of art 5 of the directive, which concerns working conditions,
including those governing dismissal.
12.
 No one would deny that an employers appraisal of the quality of the services renderedand it is this
that is the point in the case of a reference which may assist in finding new employmentfalls entirely
within the scope of the relationship between employee and employer. Whilst I would not go so far as to
claim that it constitutes, as it were, an appurtenance of salary, in that the employee is entitled, in
exchange for good and loyal service, to both a pecuniary reward and an intangible reward in the form of
praise, I do consider that the service rendered by the employer in providing the employee with the parting
gift of a reference cannot be severed from the employment relationship, and certainly not from the
conditions governing dismissal which, as is clear from the judgment in Burton v British Rlys Board Case
19/81 [1982] 3 All ER 537, [1982] ECR 555 (para 9), are to be construed broadly.

Does the protection afforded by the directive cease with the termination of the employment
relationship?
13.
The moment at which the employer takes the decision whether to provide references is, I consider,
irrelevant in the context of art 5. Whilst it is true that in most cases references are provided after the
termination of the employment relationshipfollowing resignation or dismissalit is by no means
impossible for that to happen during the period of the contract of employment. A simple example would be
that of an employee whose spouse has found a new job in another area, or has been transferred without
change of employment to a new posting in another area, and who therefore, in preparation for his or her
own move to that area, has begun to explore the employment market there. Such an 869 employee will
undoubtedly ask his or her present employer to provide references for any prospective employers.
14.
It would be totally unjustifiable for the application of the prohibition of discrimination on grounds of sex
to a decision on the provision or refusal of references to depend on the moment when that decision was
taken, or when it became effective by the actual provision of or explicit refusal to provide, a reference.
Other than in the situation referred to above, in which the employee makes advance plans to change
employer, the moment when a reference is really needed is when, following the termination of the contract
of employment, the employee sets out to look for another job. It would be particularly inappropriate and
wholly contrary to the spirit of the directive for the employee to be deprived at that moment of the
protection which the directive is intended to afford, on the ground that the discrimination is the work of a
former employer with whom there is no longer any contractual relationship. It may be borne in mind here
that art 3 of the directive seeks to protect workers against any intended discrimination on the part of
prospective employers with whom, by definition, they as yet have no contractual relationship.
15.
The court has, moreover, always held with regard to equal pay that the prohibition of discrimination by
an employer between employees on grounds of sex does not cease to have effect on termination of the
contract of employment. A particularly clear instance is the judgment in Kowalska v Freie und Hansestadt
Hamburg Case C-33/89 [1990] ECR I-2591, in which the court held that art 119 of the Treaty was
applicable to benefits paid after the termination of the employment relationship. There could be no
justification for taking a different course with Directive 76/207.
16.
I therefore consider that an employer may not discriminate on grounds of sex when providing an
employee with references, whatever the momentwhether during or after the period of the employment
relationshipat which the decision is taken in that regard or at which those references are requested.
17.
I shall add three qualifications, however, to dispel any misunderstanding. The first is that the
prohibition of discrimination in cases where the employer provides references obviously in no way
prejudges the question of the existence of an obligation to provide references. As the Commission
accepts, the directive itself does not create any such obligation. In other words, it is only where the
employer is required by law or by an express or implied term of the contract to provide references, or in
practice habitually accedes to requests for references, that the principle of equal treatment must be
observed.
18.
But, as was already stressed in the judgment in Garland v British Rail Engineering Ltd Case 12/81
[1982] 2 All ER 402, [1982] ECR 359, concerning travel facilities provided by an employer to former
employees in the absence of any contractual obligation, there can be no question of exempting employers
from the duty to respect the principle of equal treatment when they grant benefits to their employees on a
strictly voluntary basis, since the ban on any discrimination on grounds of sex applies throughout the
whole area covered by the employment relationship.
19.
To cover all eventualities, I should add that, where an employer has only one employee, a refusal to
provide that employee with references on the ground of his or her sex would still be in breach of the
prohibition of discrimination on grounds of sex.
20.
Secondly, it must be made clear that the fact that the provision of references for former employees
falls within the scope of the directive in no way 870 interferes with any rules of national law concerning the
extent of the employers obligation to provide such referencesfor example, a rule limiting that obligation,
for practical reasons, to a specific period by providing, say, that the right to obtain references may be
exercised only during the first year following the termination of the employment relationship.
21.
 The third and final qualification is that the employer remains entirely unfettered as regards the
assessment to be made of the quality of the services rendered, provided that it remains within the limits
imposed by the duty of objectivity.
22.
Were it merely a matter of dispelling the national tribunals doubts as to the scope ratione temporis of
the prohibition of discrimination on grounds of sex laid down by the directive, I could conclude my
reasoning at this point, having reached the view that the provision of references to an employeewhich
falls within the scope of the directive as defined in art 5remains subject to that prohibition whatever the
momentwhether during or after the period of the employment relationshipat which it occurs. But the
formulation of the national tribunals question does not allow me to do so, highlighting as it does the
circumstance that the refusal to provide references may have constituted retaliation for court proceedings
brought by the former employee with a view to enforcing compliance with the principle of equal treatment
and asking whether, in such an event, the member states are required to introduce into their legal
systems such measures as are necessary to enable employees who consider themselves wronged to
pursue their claims by judicial process.

Refusal to provide references as a measure of retaliation

23.
In other words, does the requirement to provide judicial remedies imposed by art 6 of the directive also
apply in cases where the employee claims to have suffered not discrimination on grounds of sex but
rather retaliation for having exercised the right to bring proceedings challenging alleged discrimination on
grounds of sex?
24.
The answer to that question can only, in my view, be no. It is clear from a reading of the provisions of
the directive that the Community legislature, whilst perfectly aware that claims for sexual equality may
irritate certain employers to the point of prompting them to engage in reprisals, intentionally took account
of only one form of retaliation, the most serious but perhaps not the least commondismissal.
25.
That intention was expressed in art 7 of the directive, under which:
Member States shall take the necessary measures to protect employees against dismissal by
the employer as a reaction to a complaint within the undertaking or to any legal proceedings aimed
at enforcing compliance with the principle of equal treatment.
26.
One may think that, in not requiring member states to provide protection against other forms of
retaliation to which employers, outraged at having had to answer in court for their conduct with regard to
the prohibition of discrimination on grounds of sex, might be tempted to have recourse, the Community
legislature has demonstrated a certain timorousness.
27.
It is not, however, possible, by any legal reasoning constructed on the basis of such a view and of the
regrets which may justifiably be felt as a result, to derive from the directive obligations on the member
states which it does not contain.
871
28.
A measure of retaliation other than dismissal does not give rise to any right to bring legal proceedings
unless, of course, it transpires that the employers recourse to retaliation was based on the sex of the
employee who has had the effrontery to claim the right to equal treatment.
29.
In such a case, the discrimination would again be based directly on sex, in the context of the
employment relationship, and art 6 of the directive would be applicable.
30.
Unfortunately for her, however, that does not appear to be Mrs Cootes situation, or at any rate she has
not stated that her former employers reprisals are aimed selectively at women only.
31.
I readily agree that, but for art 7which, as the United Kingdom government very rightly points out, is
the expression of a clear political choiceone might properly have wondered whether art 6 should not be
interpreted as requiring not only that, as the court held in Von Colson v Land Nordrhein-Westfalen Case
14/83 [1984] ECR 1891 (para 18)2, the legal proceedings must provide employees having suffered
discrimination with an effective remedy but also that the pursuit of that remedy must not be capable of
giving rise to reprisals.
2
It follows from [art 6] that Member States are required to adopt measures which are sufficiently effective to achieve the
objective of the Directive and to ensure that those measures may in fact be relied on before the national courts by the
persons concerned.

32.
Such a line of reasoning would have tallied with the richly-promising seam of case law which has
recourse to the concept of effet utile. It would have been possible to consider that the effectiveness of
the right to bring proceedings laid down by art 6 would be significantly strengthened were no threat of
retaliation to hang over the bold complainants head, and to conclude that the member states were under
a duty to make a remedy available to any victim of such retaliation. But the presence of art 7 leaves, in my
view, no scope for any such construction.
33.
Nor is there scope for any interpretation whereby retaliation might constitute indirect discrimination on
grounds of sex within the meaning of art 2(1) of the directive, which provides:
For the purposes of the following provisions, the principle of equal treatment shall mean that
there shall be no discrimination whatsoever on grounds of sex either directly or indirectly by
reference in particular to marital or family status.
34.
It is clear to me that the use of the word indirectly in that article refers to a situation in which, although
a particular rule or measure may not be explicitly directed at employees of a particular sex, it is in fact
possible to pierce the veil of appearances and identify with certainty the sex in question.
35.
The addition of the words by reference in particular to marital or family status seems to me to leave
no possible room for doubt in that regard. But it does not appear that Mrs Cootes situation is of that kind.
The questions on which a ruling is sought stress that if the references have in fact been refused, it is
because she brought proceedings against her employer and not because she is a woman.
36.
It is not therefore possible to consider that the directive requires member states to introduce into their
national legal systems such measures as are necessary to enable employees who consider themselves
wronged to pursue their claims in circumstances such as those of theclearly morally blameworthy
conduct alleged against Mrs Cootes former employer.
872

Conclusion
37.
I propose, finally, that the Court of Justice should phrase its answer to the questions on which the
Employment Appeal Tribunal seeks a preliminary ruling as follows:
(1) The provision of references for employees by an employer is covered by the prohibition of
any discrimination on grounds of sex laid down by Council Directive (EEC) 76/207 on the
implementation of the principle of equal treatment for men and women as regards access to
employment, vocational training and promotion, and working conditions. In that connection, it is
irrelevant whether the references were in fact refused during the period of employment or after its
termination, or whether the employer decided on the refusal before or after the termination of the
period of employment.
(2) Directive 76/207 does not, however, require member states to introduce into their national
legal systems such measures as are necessary to enable employees to bring legal proceedings
against former employers who have refused to provide references for them, where that refusal
constitutes retaliation for legal proceedings brought by the employee against the employer with a
view to enforcing compliance with the requirement of equal treatment for men and women.

22 September 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 20 November 1996, received at the Court of Justice of the European Communities on 12
May 1997, the Employment Appeal Tribunal, London, referred to the Court of Justice for a preliminary
ruling under art 177 of the EC Treaty two questions on the interpretation of Council Directive (EEC) on the
implementation of the principle of equal treatment for men and women as regards access to employment,
vocational training and promotion, and working conditions (OJ 1976 L39 p 40).
2.
Those questions were raised in proceedings between Ms Coote and her former private sector
employer, Granada Hospitality Ltd (Granada), concerning Granadas refusal to provide references to
potential employers of Ms Coote.

The directive
3.
According to art 1(1) of the directive, its purpose is
to put into effect in the Member States the principle of equal treatment for men and women as
regards access to employment, including promotion, and to vocational training and as regards
working conditions and, on the conditions referred to in paragraph 2, social security. This principle
is hereinafter referred to as the principle of equal treatment.
4.
Article 5(1) of the directive states:
 Application of the principle of equal treatment with regard to working conditions, including the
conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex.
5.
Under art 6 of the directive, member states are to
873
introduce into their national legal systems such measures as are necessary to enable all
persons who consider themselves wronged by failure to apply to them the principle of equal
treatment within the meaning of Articles 3, 4 and 5 to pursue their claims by judicial process after
possible recourse to other competent authorities.
6.
Pursuant to art 7 of the directive, member states are to
take the necessary measures to protect employees against dismissal by the employer as a
reaction to a complaint within the undertaking or to any legal proceedings aimed at enforcing
compliance with the principle of equal treatment.

National legislation
7.
The Sex Discrimination Act 1975, which transposed the directive into UK law, provides, in s 4:
(1) A person (the discriminator) discriminates against another person (the person victimised)
in any circumstances relevant for the purposes of any provision of this Act if he treats the person
victimised less favourably than in those circumstances he treats or would treat other persons, and
does so by reason that the person victimised has(a) brought proceedings against the
discriminator or any other person under this Act or the Equal Pay Act 1970
8.
The discrimination is unlawful only if committed in one of the situations referred to in Parts II and IV of
the 1975 Act; Part II concerns discrimination in the employment field. In this respect s 6 provides:
(1) It is unlawful for a person, in relation to employment by him at an establishment in Great
Britain, to discriminate against a woman(a) in the arrangements he makes for the purpose of
determining who should be offered that employment, or (b) in the terms on which he offers her that
employment, or (c) by refusing or deliberately omitting to offer her that employment. (2) It is
unlawful for a person, in the case of a woman employed by him at an establishment in Great
Britain, to discriminate against her(a) in the way he affords her access to opportunities for
promotion, transfer or training, or to any other benefits, facilities or services, or by refusing or
deliberately omitting to afford her access to them, or (b) by dismissing her, or subjecting her to any
other detriment.

The main proceedings

9.
According to the order for reference, Ms Coote was employed by Granada from December 1992 to
September 1993. In 1993 she brought a claim for sex discrimination against Granada, alleging that she
had been dismissed because of pregnancy. That claim was settled, and Ms Cootes employment with
Granada ended by mutual agreement on 7 September 1993.
10.
In July 1994 Ms Coote, seeking new employment, had recourse to two employment agencies. She
considers that her difficulties in finding employment were due to Granadas failure to provide a reference
to one of the employment agencies; Granada disputes this.
874
11.
Ms Coote then brought a further claim against Granada before the Industrial Tribunal, Stratford,
claiming that she had been victimised by Granadas refusal to supply a reference to the employment
agency. She alleged that the refusal was a reaction to the claim which she had previously brought against
her former employer.
12.
That claim was dismissed on the grounds that the alleged discrimination had taken place after her
employment with Granada had ended and that, in any event, the alleged detriment had arisen after that
date. The Industrial Tribunal considered that the Sex Discrimination Act 1975 was to be interpreted as
prohibiting only retaliatory measures whose prejudicial effect appears during the employment relationship.
It stated that it reached that decision with some regret, since it left Ms Coote without a remedy.
13.
Ms Coote appealed against that decision to the Employment Appeal Tribunal. In its order for reference
the Employment Appeal Tribunal observes that under s 4 of the 1975 Act it is discrimination inter alia if a
person is treated less favourably than others because of having brought proceedings under the Act;
however, victimisation by the employer is unlawful only in the circumstances set out in Part II of the Act.
14.
 The Employment Appeal Tribunal observes that the Race Relations Act 1976, which contains
provisions similar to ss 4 and 6 of the Sex Discrimination Act 1975, has been held to cover only acts of
discrimination which occur during the relevant employment relationship, so that it does not apply to cases
in which an employer causes harm to a former employee. The Employment Appeal Tribunal notes that it
was on that basis that the Industrial Tribunal decided against Ms Coote.
15.
The Employment Appeal Tribunal is, however, uncertain whether, having regard to the directive, the
Sex Discrimination Act 1975, unlike the Race Relations Act 1976, ought not to be interpreted as
prohibiting not only retaliatory measures which take the form of detrimental conduct during the
employment relationship but also those which are decided on or whose harmful effects are produced after
the employment has ended.
16.
In those circumstances, the Employment Appeal Tribunal stayed proceedings and referred the
following questions to the Court of Justice for a preliminary ruling:
(1) Does Council Directive 76/207/EEC on the implementation of the principle of equal
treatment for men and women as regards access to employment, vocational training and
promotion, and working conditions require Member States to introduce into their national legal
systems such measures as are necessary to enable a complainant to pursue a claim by judicial
process where the following circumstances apply: (i) the complainant was employed by the
respondent; (ii) during her employment, the complainant brought a claim of sex discrimination
against the respondent which was compromised; (iii) following the end of her employment, the
complainant has made efforts to find full-time work, but has been unsuccessful; (iv) the respondent
has caused or contributed to the complainants difficulties in finding work by refusing to provide
references to potential employers when requested to do so; (v) the employers decision to refuse to
provide references was taken after the end of the complainants employment; (vi) the reason, or a
principal reason, for the employers 875 decision to refuse to provide a reference to the
complainant was that she had previously brought a claim of sex discrimination against the
respondent?
(2) Does Council Directive 76/207/EEC on the implementation of the principle of equal treatment
for men and women as regards access to employment, vocational training and promotion, and
working conditions require Member States to introduce into their national legal systems such
measures as are necessary to enable persons to pursue a claim by judicial process where the
circumstances set out under 1 above apply, save that: (i) the respondents decision to refuse to
provide a reference was taken before the termination of the complainants employment; but (ii) the
actual refusal or refusals to provide a reference took place after the termination of the
complainants employment?
17.
It must be pointed out at the outset that it has been consistently held since Marshall v Southampton
and South West Hampshire Health Authority (Teaching) Case 152/84 [1986] 2 All ER 584, [1986] ECR
723 (para 48), that a directive cannot of itself impose obligations on an individual, in this case a private
sector employer, and thus cannot be relied upon as such against such a person.
18.
However, it has also been consistently held since Von Colson v Land Nordrhein-Westfalen Case 14/83
[1984] ECR 1891 (para 26), that the member states obligation arising from a directive to achieve the
result envisaged by the directive and their duty under art 5 of the EC Treaty to take all appropriate
measures, whether general or particular, to ensure the fulfilment of that obligation are binding on all the
authorities of member states including, for matters within their jurisdiction, the courts. As follows from
Marleasing SA v La Comercial Internacional de Alimentacin SA Case C-106/89 [1990] ECR I-4135 (para
8) and Wagner Miret v Fondo de Garanta Salarial Case C-334/92 [1993] ECR I-6911 (paras 2021), in
applying national law, in particular legislative provisions which, as in the present case, were specially
introduced in order to implement the directive, the national court is required to interpret its national law, so
far as possible, in the light of the wording and the purpose of the directive in order to achieve the result
pursued by the third paragraph of art 189 of the Treaty.
19.
In those circumstances, the questions put by the national court must be understood as seeking to
ascertain, for the purpose of interpreting national provisions transposing the directive, whether the
directive requires member states to introduce into their national legal systems such measures as are
necessary to ensure judicial protection for workers whose employer, after the end of the employment
relationship, refuses to provide references as a reaction to proceedings brought to enforce compliance
with the principle of equal treatment within the meaning of the directive.
20.
On this point, it should be noted that art 6 of the directive requires member states to introduce into
their national legal systems such measures as are necessary to enable all persons who consider
themselves the victims of discrimination to pursue their claims by judicial process. It follows from that
provision that the member states must take measures which are sufficiently effective to achieve the aim of
the directive and that they must ensure that the rights thus conferred can be effectively relied upon before
the national courts by the persons concerned (see esp Von Colson [1984] ECR 1891 (para 18), Johnston
v Chief Constable of the Royal Ulster Constabulary Case C-222/84 [1986] 3 All ER 135, [1986] ECR 1651
(para 17) and Marshall v Southampton and South West Hampshire 876 Area Health Authority (No 2)
Case C-271/91 [1993] 4 All ER 586, [1993] ECR I-4367 (para 22)).
21.
The requirement laid down by that article that recourse be available to the courts reflects a general
principle of law which underlies the constitutional traditions common to the member states and which is
also enshrined in art 6 of the Convention for the Protection of Human Rights and Fundamental Freedoms
(Rome, 4 November 1950; TS 71 (1953); Cmnd 8969) (see esp Johnston [1986] 3 All ER 135, [1986]
ECR 1651 (para 18)).
22.
By virtue of art 6 of the directive, interpreted in the light of the general principle stated above, all
persons have the right to obtain an effective remedy in a competent court against measures which they
consider to interfere with the equal treatment for men and women laid down in the directive. It is for the
member states to ensure effective judicial control of compliance with the applicable provisions of
Community law and of national legislation intended to give effect to the rights for which the directive
provides (see [1986] 3 All ER 135, [1986] ECR 1651 (para 19)).
23.
As the court has also held, art 6 of the directive is an essential factor for attaining the fundamental
objective of equal treatment for men and women, which, as the court has repeatedly held, is one of the
fundamental human rights whose observance the court has a duty to ensure (see the judgments in
Marshall v Southampton and South West Hampshire Area Health Authority (No 2) [1993] 4 All ER 586,
[1993] ECR I-4367 (para 34) and P v S Case C-13/94 [1996] All ER (EC) 397, [1996] ECR I-2143 (para
19), respectively).
24.
The principle of effective judicial control laid down in art 6 of the directive would be deprived of an
essential part of its effectiveness if the protection which it provides did not cover measures which, as in
the main proceedings in this case, an employer might take as a reaction to legal proceedings brought by
an employee with the aim of enforcing compliance with the principle of equal treatment. Fear of such
measures, where no legal remedy is available against them, might deter workers who considered
themselves the victims of discrimination from pursuing their claims by judicial process, and would
consequently be liable seriously to jeopardise implementation of the aim pursued by the directive.
25.
In those circumstances, it is not possible to accept the United Kingdom governments argument that
measures taken by an employer against an employee as a reaction to legal proceedings brought to
enforce compliance with the principle of equal treatment do not fall within the scope of the directive if they
are taken after the employment relationship has ended.
26.
It is true that, as the UK government also stresses, art 7 of the directive expressly requires member
states to take the necessary measures to protect employees against dismissal by the employer as a
reaction to any legal proceedings aimed at enforcing compliance with the principle of equal treatment.
27.
However, contrary to that governments submissions, having regard to the objective of the directive,
which is to arrive at real equality of opportunity for men and women (see Marshall v Southampton and
South West Hampshire Health Authority (Teaching) [1986] 2 All ER 584, [1986] ECR 723 (para 24)), and
to the fundamental nature of the right to effective judicial protection, it is not, in the absence of a clear
indication to the contrary, to be inferred from art 7 of the directive that the legislatures intention was to
limit the protection of workers against retaliatory measures decided on by the employer solely to cases of
dismissal, which, although an exceptionally serious measure, is not the only 877 measure which may
effectively deter a worker from making use of the right to judicial protection. Such deterrent measures
include inter alia those which, as in the present case, are taken as a reaction to proceedings brought
against an employer and are intended to obstruct the dismissed employees attempts to find new
employment.
28.
In those circumstances, the answer to the questions put by the national court must be that art 6 of the
directive requires member states to introduce into their national legal systems such measures as are
necessary to ensure judicial protection for workers whose employer, after the employment relationship
has ended, refuses to provide references as a reaction to legal proceedings brought to enforce
compliance with the principle of equal treatment within the meaning of the directive.

Costs
29.
The costs incurred by the UK government and by the European Commission, which have submitted
observations to the Court of Justice, are not recoverable. Since these proceedings are, for the parties to
the main proceedings, a step in the action pending before the national tribunal, the decision on costs is a
matter for that tribunal.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Employment
Appeal Tribunal, London, by order of 20 November 1996, hereby rules: art 6 of Council Directive (EEC)
76/207 on the implementation of the principle of equal treatment for men and women as regards access
to employment, vocational training and promotion, and working conditions requires member states to
introduce into their national legal systems such measures as are necessary to ensure judicial protection
for workers whose employer, after the employment relationship has ended, refuses to provide references
as a reaction to legal proceedings brought to enforce compliance with the principle of equal treatment
within the meaning of that directive.
 878

[1998] All ER (EC) 879

Boyle and others v Equal Opportunities Commission

(Case C-411/96)

EUROPEAN COMMUNITY; Social policy: EMPLOYMENT; Discrimination

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), KAPTEYN (RAPPORTEUR), PUISSOCHET AND
JANN (PRESIDENTS OF CHAMBERS), GULMANN, MURRAY, EDWARD, RAGNEMALM, SEVN,
WATHELET AND SCHINTGEN
ADVOCATE GENERAL RUIZ-JARABO COLOMER
13 JANUARY, 19 FEBRUARY, 27 OCTOBER 1998
European Community Equality of treatment of men and women Access to employment and working
conditions Maternity scheme Benefits conditional on agreement to return to employer Failure to
return entailing partial repayment of maternity pay Maternity leave precluding simultaneous sick leave
Accrual during maternity leave of annual leave entitlement limited Accrual of pension rights during
maternity leave conditional on entitlement to maternity pay from employer Whether constituting
discrimination on grounds of sex Council Directives (EEC) 75/117, 76/207 and 92/85 EC Treaty, art
119.

The six applicants were female employees of the EOC who had each completed at least one years paid
service on a permanent basis. Their contracts of employment comprised a staff handbook, which applied
to all workers, and a maternity scheme, which applied to female workers. Under the maternity scheme,
the applicants were entitled to just over three months maternity leave on full pay plus a limited period of
unpaid supplementary maternity leave. However, in order to benefit from those rights, an employee had to
state that she intended to return to work at the EOC after childbirth and agree that, should she fail to
return, she would repay any payment received under the maternity scheme which exceeded the statutory
maternity pay to which she would have been entitled. The applicants applied to an industrial tribunal for a
declaration that that requirement and certain other clauses in their contracts of employment were void or
unenforceable in so far as they discriminated against female employees, contrary to art 119 1 (equal pay
for equal work) of the EC Treaty and Council Directive (EEC) 75/117 on the approximation of the laws of
the member states relating to the application of the principle of equal pay for men and women, Council
Directive (EEC) 76/207 on the implementation of the principle of equal treatment for men and women as
regards access to employment, vocational training and promotion and working conditions and Council
Directive (EEC) 92/85 on the introduction of measures to encourage improvements in the safety and
health at work of pregnant workers and workers who have recently given birth or are breastfeeding. The
tribunal stayed the proceedings and referred to the Court of Justice of the European Communities for a
preliminary ruling questions as to the compatibility of such clauses with the relevant Community rules.
1
Article 119, so far as material, is set out at p 887 c d, post

879

Held (1) Article 119 of the Treaty, art 1 of Directive 75/117 and art 11 2 of Directive 92/85 did not preclude
a clause in an employment contract which made the payment, during the period of maternity leave
referred to by art 83 of Directive 92/85, of pay higher than the statutory payments in respect of maternity
leave conditional on the workers undertaking to return to work after the birth of the child for at least one
month, failing which she was required to repay the difference between the amount of the pay she would
have received during the period of maternity leave, on the one hand, and the amount of those payments,
on the other (see p 910 e f, p 911 d to g and p 916 j, post).
2
Article 11, so far as material, is set out at p 908 j to p 909 c, post
3
Article 8, so far as material, is set out at p 908 h j, post

(2) Article 8 of Directive 92/85 and art 5(1)4 of Directive 76/207 did not preclude a clause in an
employment contract from requiring an employee who had expressed her intention to commence her
maternity leave during the six weeks preceding the expected week of childbirth, and was on sick leave
with a pregnancy-related illness immediately before that date and gave birth during the period of sick
leave, to bring forward the date on which her paid maternity leave commenced either to the beginning of
the sixth week preceding the expected week of childbirth or to the beginning of the period of sick leave,
whichever was the later (see p 912 d to f j and p 917 b, post).
4
Article 5(1), so far as material, is set out p 908 g, post

(3) A clause in an employment contract which prohibited a woman from taking sick leave during the
minimum period of 14 weeks maternity leave to which a female worker was entitled pursuant to art 8(1) of
Directive 92/85, unless she elected to return to work and thus terminate her maternity leave, was not
compatible with Directive 92/85. By contrast, a clause in an employment contract which prohibited a
woman from taking sick leave during a period of supplementary maternity leave granted to her by the
employer, unless she elected to return to work and thus terminate her maternity leave, was compatible
with Directives 76/207 and 92/85 (see p 913 f to p 914 j and p 917 c d, post).
(4) Directives 92/85 and 76/207 did not preclude a clause in an employment contract from limiting the
period during which annual leave accrued to the minimum period of 14 weeks maternity leave to which
female workers were entitled under art 8 of Directive 92/85 and from providing that annual leave ceased
to accrue during any period of supplementary maternity leave granted to them by their employer (see p
914 g to j, p 915 d to h and p 917 e, post).
(5) Directive 92/85 precluded a clause in an employment contract from limiting, in the context of an
occupational scheme wholly financed by the employer, the accrual of pension rights during the period of
maternity leave referred to by art 8 of that directive to the period during which the woman received the
pay provided for by that contract or national legislation (see p 916 b to f and p 917 f, post).

Notes
For equal pay and equal treatment for men and women, see 52 Halsburys Laws (4th edn) paras 2111
2116.
For the EC Treaty, art 119, see 50 Halsburys Statutes (4th edn) 306.

880

Cases cited
Aprile Srl (in liq) v Amministrazione delle Finanze dello Stato Case C-125/94 [1995] ECR I-2919.
Barber v Guardian Royal Exchange Assurance Group Case C-262/88 [1990] 2 All ER 660, [1991] QB
344, [1991] 2 WLR 72, [1990] ECR I-1889, ECJ.
Brown v Rentokil Ltd Case C-394/96 [1998] All ER (EC) 791, ECJ.
Caisse Nationale dAssurance Viellesse des Travailleurs Salaris v Thibault Case C-136/95 [1998] All ER
(EC) 385, [1998] ECR I-2011, ECJ.
Coloroll Pension Trustees Ltd v Russell Case C-200/91 [1995] All ER (EC) 23, [1994] ECR I-4389, ECJ.
De Weerd, ne Roks v Bestuur van de Bedrijfsvereniging voor de Gezondheid, Geestelijke en
Maatschappelijke Belangen Case C-343/92 [1994] ECR I-571.
Defrenne v Belgium Case 80/70 [1971] ECR 445.
Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case C-177/88 [1990]
ECR I-3941.
Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225, ECJ.
Furlanis Costruzioni Generali SpA v Azienda Nazionale Autonoma Strade (ANAS) Case C-143/93 [1995]
ECR I-3633.
Gerster v Freistaat Bayern Case C-1/95 [1997] ECR I-5253.
Gillespie v Northern Health and Social Services Board Case C-342/93 [1996] All ER (EC) 284, [1996]
ECR I-475, ECJ.
Habermann-Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb/Opf-eV Case C-421/92 [1994] ECR I-
1657.
Handels- og Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiver- forening Case C-179/88
[1990] ECR I-3979.
Hofmann v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047.
Kording v Senator fr Finanzen Case C-100/95 [1997] ECR I-5289.
Larsson v Ftex Supermarked A/S Case C-400/95 [1997] ECR I-2757.
Megner v Innungskrankenkasse Vorderpfalz Case C-444/93 [1996] All ER (EC) 212, [1995] ECR I-4741,
ECJ.
Meilicke v ADV/ORGA AG Case C-83/91 [1992] ECR I-4871.
Pedersen v Kvickly Skive Case C-66/96 (1997) Transcript (opinion) 10 July, ECJ.
Union Royale Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER
(EC) 97, [1995] ECR I-4921, ECJ.
Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994] 4 All ER 115, [1994] QB 718, [1994] 1 WLR 941,
[1994] ECR I-3567, ECJ.

Reference
By order of 15 October 1996, the Industrial Tribunal, Manchester, referred to the Court of Justice of the
European Communities for a preliminary ruling under art 177 of the EC Treaty five questions (set out at p
970 g to p 908 a, post) on the interpretation of art 119 of the Treaty, Council Directive (EEC) 75/117 on the
approximation of the laws of the member states relating to the application of the principle of equal pay for
men and women (OJ 1975 L45 p 19), Council Directive (EEC) 76/207 on the implementation of the
principle of equal treatment for men and women as regards access to employment, vocational training
and promotion, and working conditions (OJ 1976 L39 p 40) and Council Directive (EEC) 92/85 on the
introduction of measures to encourage improvements in the safety and health at work of pregnant
workers and workers who have recently given birth or are breastfeeding (OJ 1992 L348 881p 1). Written
observations were submitted on behalf of: Mrs Boyle and others, by D Rose, Barrister, instructed by A
Eddy, Solicitor; the Equal Opportunities Commission, by P Duffy QC, instructed by A Lakin, Solicitor; the
United Kingdom government, by L Nicoll, of the Treasury Solicitors Department, acting as agent, and by
E Sharpston, Barrister; the Irish government, by M A Buckley, Chief State Solicitor, acting as agent, and
by N Hyland, BL; and the European Commission, by M Wolfcarius and C OReilly, of its Legal Service,
acting as agents. Oral observations were made by: Mrs Boyle and others, represented by D Rose; the
Equal Opportunities Commission, represented by P Duffy QC; the UK government, represented by J E
Collins, Assistant Treasury Solicitor, acting as agent, and E Sharpston; the Irish government, represented
by B Lenihan SC, and N Hyland; the Austrian government, represented by C Pesendorfer, Oberrtin im
Bundeskanzleramt, acting as agent; and the Commission, represented by M Wolfcarius and C OReilly.
The language of the case was English. The facts are set out in the opinion of the Advocate General.

19 February 1998.

The Advocate General (D Ruiz-Jarabo Colomer)

delivered the following opinion (translated from the Spanish).
1.
This is the first time that the Court of Justice of the European Communities has been called upon to
give a preliminary ruling on the interpretation of certain provisions of Council Directive (EEC) 92/85 on the
introduction of measures to encourage improvements in the safety and health at work of pregnant
workers and workers who have recently given birth or are breastfeeding (OJ 1992 L348 p 1) concerning
maternity leave and entitlement to the maintenance of employment rights during such leave, at the
request of a national court, specifically the Industrial Tribunal, Manchester (United Kingdom), which has
stayed proceedings in a number of cases before it pending a ruling under art 177 of the EC Treaty on five
questions, the wording of which has been agreed between the parties.

ITHE FACTS
2.
The proceedings before the Industrial Tribunal have been brought by Mrs Boyle and other female
employees of the Equal Opportunities Commission (the EOC), a public body set up under the Sex
Discrimination Act 1975 to promote equal treatment and equality of opportunity as between men and
women in the United Kingdom. The applicants wish to have declared void certain provisions of the EOC
Maternity Scheme (rules governing employees absences on account of pregnancy and maternity), which
are incorporated in their employment contracts, because, by providing for the application of certain
measures in the event of pregnancy or maternity and/or ultimately, by reference to sex, they discriminate
against women. The contested provisions coincide with those applied in the Civil Service in England and
Wales.
3.
The applicants are employees of the EOC and are all of childbearing age. They have all completed
one years service with the EOC. They were not employed on a casual or standby basis and none of them
was employed for a fixed term of less than two years. Mrs Boyle started working for the EOC for an
indefinite period around 1976. During her employment she has taken maternity leave on two occasions.
The dates of birth of the plaintiffs in the main proceedings are: Mrs Boyle, 23 June 1953, Mrs Taylor, 24
June 1961, and Mrs Mansley, 14 December 1961. The last two each took maternity leave 882 recently
and all three may wish to take further periods of maternity leave: hence their interest in having the
contested provisions in their employment contracts declared void or unenforceable.

IITHE APPLICABLE NATIONAL PROVISIONS

ALegal provisions concerning employment rights in the event of pregnancy and maternity
and the right to be paid when absent from work
4.
Sections 71 to 78 of the Employment Rights Act 1996 grant all female workers a general right to
maternity leave for a continuous period of 14 weeks (maternity leave), commencing either on the date
which the employee notifies to her employer or on the first day after the beginning of the sixth week
before the expected date of childbirth, whichever is earlier.
Sections 79 to 85 of that Act grant employees who meet certain conditions, including continuous
employment for the two preceding years, the right to return to work after their maternity leave, at any time
during the 29 weeks following the week in which childbirth occurredthe right to return to work. To
describe the period of time for which the worker may stop working on account of pregnancy and maternity
and retain the right to return to work, the duration of which may, by virtue of a concession from the
employer, exceed 29 weeks, I shall use the expression unpaid maternity leave.
5.
Sections 164 to 166 of the Social Security Contributions and Benefits Act 1992 govern the right of
workers who have been employed for a specified period 5, and those earnings are of a specified level6, to
receive statutory maternity pay from their employer for a maximum of 18 weeks, if they are absent from
work on account of pregnancy or maternity. There are two rates of statutory maternity pay: the higher rate
and the lower rate. The higher rate consists of nine-tenths of the womans normal weekly earnings in the
eight weeks preceding the fourteenth week before the expected week of confinement; the lower rate,
which is fixed and amounts at present to 5455, is paid where its amount exceeds the nine-tenths figure.
A woman who, for two continuous years ending at the start of the fourteenth week prior to the expected
week of confinement, has worked for an employer who is under an obligation to pay her the benefit will
receive it at the higher rate for the first six weeks and at the lower rate for the remainder of the period. A
woman who is entitled to statutory maternity pay, but does not qualify for the higher rate, will receive it at
the lower rate.
5
At least 26 weeks ending with the week preceding the fourteenth week before the expected week of confinement.
6
This must not be less than the lowest level of income taken into account for the payment of social security contributions.

Any pregnant women who do not meet the necessary conditions to receive statutory maternity pay are
entitled to receive, subject to a maximum of 18 weeks, a maternity allowance, which at present amounts
to 5455 per week.
Sections 151 to 163 of the same Act govern the right to benefits in respect of incapacity for work on
account of illness. In such circumstances, workers are entitled to receive statutory sick pay from their
employers for a maximum of 28 weeks, the present rate being 5455 per week.
6.
Paragraph 5 of Sch 5 to the Social Security Act 1989, which incorporates in domestic law the
provisions of Council Directive (EEC) 86/378 on the implementation of the principle of equal treatment for
men and women in 883 occupational social security schemes (OJ 1986 L225 p 40), governs the situation
of a woman on maternity leave regarding the accrual of pension rights under an occupational scheme. By
virtue of that provision, a woman must be allowed to continue to belong to an occupational pension
scheme as if she were working normally. She may not receive more favourable treatment than a woman
in active employment, but her contributions to the pension scheme will be based on the amount she
receives from her employer whilst on maternity leave by way of contractual remuneration or statutory
maternity pay.

BThe contested provisions of the Equal Opportunities Commission maternity scheme

7.
The rules relevant to the decision to be given in these proceedings are as follows:
3. PAID MATERNITY LEAVE
3.1. A member of staff will be allowed 3 months and 1 weeks paid maternity leave for the period
of continuous absence before and after childbirth, provided that she:states that she intends to
return to work in the EOC after childbirth and she agrees to repay any payment made during that
period if she fails to return. Such payment will exclude any Statutory Maternity Pay (SMP) to which
there is an entitlement:is in paid service with the EOC at the time her maternity leave begins and
has rendered at least one years such service;is not employed on a casual, standby or short
notice appointment;is not employed on a fixed-term appointment of less than two years.
4. UNPAID MATERNITY LEAVE
A member of staff who qualifies for paid maternity leave will also qualify for unpaid maternity
leave, subject to the following restrictions:the total period of paid and unpaid maternity leave
must not exceed 52 weeks;unpaid maternity leave cannot be terminated earlier than 41 weeks
from the actual date of childbirth, for those staff who qualify: except where the first restriction in 4.1
prevents this; with a member of staffs consent.
4.2 A permanent member of staff with less than one years service is entitled to a total of 26
weeks leave.
4.3 All other employees, regardless of type of contract and number of hours worked, are entitled
to 14 weeks leave.
5. TIMING OF LEAVE
5.1 Maternity leave may start on any day of the working week, subject to the following
restrictions:unpaid maternity leave cannot start earlier than 14 weeks before the expected week
of childbirth;paid maternity leave cannot start earlier than 11 weeks before the expected week of
childbirth; paid maternity leave must start from the actual date of childbirth if this occurs earlier
than: the expected week of childbirth, or the date a member of staff has specified for beginning her
maternity leave.
5.2 If a member of staff specifies that she wishes to begin her maternity leave in any of the six
weeks before the expected week of childbirth, the following restriction applies:If she is on a
pregnancy-related sick absence immediately before her specified date and childbirth occurs during
the period of sick absence, paid maternity leave can be brought forward to 884 whichever is the
later of: the beginning of the period of sick absence; the 6th week before the expected week of
childbirth.
6. RIGHT TO RETURN TO WORK
6.3 The minimum period of service that must be completed following paid maternity leave is one
calendar month. Where this requirement is not satisfied, the member of staff will be asked to repay
any salary or wages paid for the period of maternity leave (less any SMP)
7. SICK ABSENCE DURING OR FOLLOWING PREGNANCY
7.1 Paid sick leave is not allowed once paid maternity leave has begun or during a period of
unpaid maternity leave. There may, however, be an entitlement to Statutory Sick Pay (SSP) during
 unpaid maternity leave. Evidence of incapacity must be submitted to the Personnel and Payroll Unit
so that any eligibility for SSP can be determined.
7.2 Where a member of staff has provided at least 3 weeks notification of her intention to return
to work on a specified date, paid sick leave will be allowed from this date. Paid sick leave following
childbirth terminates the maternity leave arrangements.
7.3 The period of paid and/or unpaid maternity leave and unpaid leave will not reckon against
the normal sick leave limits.
7.4 A member of staff who resumes work after maternity leave is entitled to have the same job
and the same terms and conditions as if she had not been absent.
8. CONTRACTUAL RIGHTS AND BENEFITS
8.1 A member of staff not entitled to paid leave of absence retains her contractual rights and
benefits, except remuneration, during the first 14 weeks of leave. This means that annual leave will
continue to accrue. The period of absence only accrues for pension purposes if in receipt of SMP.

CProvisions of the Equal Opportunities Commission Staff Handbook

8.
In the order for reference, the Industrial Tribunal states that the parties agree that Annex 4 to the Staff
Handbook7 forms part of the applicants contract of employment. So far as is relevant here, the manual
provides as follows:
7
Annex 4 governs types of leave available to EOC employees, namely: Annual Leave, Public and Privilege Leave, Sick
Leave, Medical Appointments, Special Leave and Maternity Leave.

4.1.9 Effect on Annual Leave of Leave Without Pay

Any leave taken without pay (for example unpaid sick, special, or maternity leave) reduces the
annual leave entitlement for that year by a proportion of the amount of unpaid leave taken: for
example, one months unpaid leave reduces annual leave entitlement for that year by 1/12
4.1.11 Sick Leave and Annual Leave
Annual leave is not taken instead of sick leave. Staff who fall ill or are injured during annual
leave to the extent that they are unfit to work should inform their line manager as soon as possible,
and submit either a self-certificate or a doctors note (according to the duration of sick leave) as
soon as possible after that. The period of absence covered by the note will then count as sick leave

4.3.4 Sick Pay
Salary is paid in full for all sick leave of up to six months total duration in any twelve-month
period. Thereafter half-pay continues up to a total of 885 twelve months paid sick leave in four
years. Once entitlement to full sick pay and half sick pay is exhausted further sick leave will be
unpaid; or sick pay may be given at pension rate, equivalent to the amount to which the employee
would have been entitled had they been retired prematurely on ill-health grounds.
4.3.5 Statutory Sick Pay
All employers are required by law to pay Statutory Sick Pay for any period of sick leave totalling
up to 28 weeks provided that there are four or more consecutive days on which an employee is
unable to work and provided certain other state benefits are not being paid. SSP rates are
generally less than the EOCs own sick pay provisions and therefore the EOC normally fulfils its
legal requirements in respect of SSP by virtue of the sick pay arrangements described in 4.3.4
4.3.6 Adjustments to pay: sickness benefits
If entitlement to Statutory Sick Pay is exhausted in a particular case, the employee is informed
by Personnel, who will forward medical certificates covering continuing absence to their
Department of Health and Social Security (DHSS) office so that state benefits can be claimed. Any
state benefits that are due will be paid by the DHSS. The total of sick pay and state benefits must
not exceed normal pay, and sick pay will be adjusted to avoid this happening.

IIITHE PRELIMINARY QUESTIONS

9.
In order to resolve the disputes before it, the Industrial Tribunal, Manchester, seeks a preliminary ruling
from the Court of Justice on the following questions, the wording of which was proposed by the parties:
In circumstances such as those of the present cases, do any of the following matters infringe
the prohibition of unfair and/or unfavourable treatment of women because of pregnancy, childbirth,
maternity and/or sickness in relation thereto under EC law (in particular Article 119 of the Treaty of
Rome and/or Council Directive 75/117/EEC8 and/or Council Directive 76/207/EEC9 and/or Council
Directive 92/85/EEC):

8
Council Directive (EEC) 75/117 on the approximation of the laws of the member states relating to the application of the
principle of equal pay for men and women (OJ 1975 L45 p 19)
9
Council Directive (EEC) 76/207 on the implementation of the principle of equal treatment for men and women as
regards access to employment, vocational training and promotion, and working conditions (OJ 1976 L39 p 39).
 (1) A condition that maternity pay, beyond the Statutory Maternity Pay, is paid only if the woman
states that she intends to return to work and agrees to be liable to repay such maternity pay if she
does not return to work for one month on the conclusion of maternity leave.
(2) A condition that where a woman, who is absent on paid sick leave with a pregnancy-related
illness, gives birth during such absence, her maternity leave may be backdated to the later date of
either six weeks before the expected week of childbirth or when the sickness leave began.
(3) A prohibition on a woman, who is unfit for work for any reason whilst on maternity leave, from
taking paid sick leave, unless she elects to return to work and terminate her maternity leave.
886
(4) A condition limiting the time during which annual leave accrues to the statutory minimum
period of 14 weeks maternity leave and accordingly excluding any other period of maternity leave.
(5) A condition limiting the time in which pensionable service accrues during maternity leave to
when the woman is in receipt of contractual or Statutory Maternity Pay and accordingly excluding
any period of unpaid maternity leave.

IVTHE APPLICABLE COMMUNITY PROVISIONS

10.
Article 119 of the Treaty provides as follows:
Each Member State shall during the first stage ensure and subsequently maintain the
application of the principle that men and women should receive equal pay for equal work. For the
purpose of this article, pay means the ordinary basic or minimum wage or salary and any other
consideration, whether in cash or in kind, which the worker receives, directly or indirectly, in respect
of his employment from his employer. Equal pay without discrimination based on sex means (a)
that pay for the same work at piece rates shall be calculated on the basis of the same unit of
measurement. (b) that pay for work at time rates shall be the same for the same job.
11.
Directive 75/117 provides, so far as is relevant here:
Article 1
The principle of equal pay for men and women outlined in Article 119 of the Treaty, hereinafter
principle of equal pay, means, for the same work or for work to which equal value is attributed, the
elimination of all discrimination on grounds of sex with regard to all aspects and conditions of
remuneration
Article 3
Member States shall abolish all discrimination between men and women arising from laws,
regulations or administrative provisions which is contrary to the principles of equal pay.
Article 4
Member States shall take the necessary measures to ensure that provisions appearing in
collective agreements, wage scales, wages agreements or individual contracts of employment
which are contrary to the principle of equal pay shall be, or may be declared, null and void or may
be amended.
12.
For its part, Directive 76/207 provides:
Article 1
(1) The purpose of this directive is to put into effect in the member states the principle of equal
treatment for men and women as regards access to employment, including promotion, and to
vocational training and as regards working conditions this principle is hereinafter referred to as the
principle of equal treatment.
887
Article 2
(1) For the purposes of the following provisions, the principle of equal treatment shall mean that
there shall be no discrimination whatsoever on grounds of sex either directly or indirectly by
reference in particular to marital or family status.
3. This directive shall be without prejudice to provisions concerning the protection of women,
particularly as regards pregnancy and maternity
Article 5
(1) Application of the principle of equal treatment with regard to working conditions, including the
conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex.
(2) To this end, Member States shall take the measures necessary to ensure that: (a) any laws,
regulations and administrative provisions contrary to the principle of equal treatment shall be
abolished; (b) any provisions contrary to the principle of equal treatment which are included in
collective agreements, individual contracts of employment, internal rules of undertakings or in rules
 governing the independent occupations and professions shall be, or may be declared, null and void
or may be amended.
13.
Article 6 of Directive 86/378 provides:
(1) Provisions contrary to the principle of equal treatment shall include those based on sex,
either directly or indirectly, in particular by reference to marital or family status, for (g)
suspending the retention or acquisition of rights during periods of maternity leave or leave for family
reasons which are granted by law or agreement and are paid by the employer
14.
Finally, art 8 of Directive 92/85 provides with regard to maternity leave:
(1) Member States shall take the necessary measures to ensure that workers within the
meaning of Article 2 [who are pregnant, have recently given birth or are breastfeeding] are entitled
to a continuous period of maternity leave of at least 14 weeks allocated before and/or after
confinement in accordance with national legislation and/or practice.
(2) The maternity leave stipulated in paragraph 1 must include compulsory maternity leave of at
least two weeks allocated before and/or after confinement in accordance with national legislation
and/or practice.
As regards rights under employment contracts, art 11 provides:
In order to guarantee workers within the meaning of Article 2 the exercise of their health and
safety protection rights as recognized in this Article, it shall be provided that (2) in the case
referred to in Article 8, the following must be ensured: (a) the rights connected with the employment
contract of workers within the meaning of Article 2, other than those referred to in point (b) below;
(b) maintenance of a payment to, and/or entitlement to an adequate allowance for, workers within
the meaning of Article 2; (3) the allowance referred to in point 2(b) shall be deemed adequate if it
guarantees income at least equivalent to that which the worker concerned would receive in the
event of a break in her activities 888 on grounds connected with her state of health, subject to any
ceiling laid down under national legislation; (4) Member States may make entitlement to pay or the
allowance referred to in points 1 and 2(b) conditional upon the worker concerned fulfilling the
conditions of eligibility for such benefits laid down under national legislation. These conditions may
under no circumstances provide for periods of previous employment in excess of 12 months
immediately prior to the presumed date of confinement.

VOBSERVATIONS SUBMITTED TO THE COURT OF JUSTICE

15.
Written observations have been submitted in these proceedings, under art 20 of the EC Statute of the
Court of Justice, by the applicants in the main proceedings, jointly, the EOC, the governments of the
United Kingdom and Ireland and the Commission. The Austrian government also attended the hearing to
present oral observations.
16.
The applicants claim that the obligation which the EOC Maternity Scheme imposes on pregnant
women in order to receive full pay during maternity leave, demanding repayment of the difference
between that amount and statutory maternity pay if they do not return to the work at the end of that
period, constitutes discrimination in relation to pay on account of pregnancy, contrary to art 119 of the
Treaty. A commitment of that kind is not imposed on workers as a precondition for receiving full pay in
cases of absence on other grounds. They also consider that the obligation imposed on a woman who is
unfit for work on account of illness to commence paid maternity leave rather than remaining on sick leave
when she reaches the sixth week prior to the expected week of confinement constitutes either
discrimination in relation to pay or discrimination in relation to working conditions, contrary to art 5 of
Directive 76/207.
In their opinion, it is also discriminatory and thereby contrary to Community law to refuse to permit a
woman who falls ill after maternity leave has commenced to take paid sick leave. In such circumstances
an employee who is unfit for work is denied the right, on account of her pregnancy or the fact that she has
recently given birth, from exercising her contractual right to take sick leave, on full pay, and she is
required to take her maternity leave, at the end of which she must repay part of her remuneration if she
does not return to work.
Similarly, they consider that the fact of disregarding, for the purposes of accruing pension rights under
the occupational scheme financed entirely by the employer, periods of leave of absence or unpaid leave,
including unpaid maternity leave, constitutes indirect discrimination against women. They consider that
the condition under which only paid service constitutes reckonable service for the purposes of accruing
pension rights appears neutral at first sight but is, in reality, a condition which adversely affects women,
since a substantially higher proportion of women than of men take leave of absence or unpaid leave, only
women being able to take such leave of absence after maternity leave. For the same reason, they submit
that the fact that any leave of absence or unpaid leave taken brings about a reduction, in the year in
question, of the period of annual leave entitlement proportional to the duration of the unpaid leave
constitutes discrimination against women.
17.
The respondent considers it wrong to equate any condition of employment applicable to pregnancy or
maternity with direct discrimination on grounds of sex, contrary to Community law, since to do so is
incompatible with the refusal of the Court of Justice to compare pregnancy to a pathological 889
condition. In its view it is necessary for the court to clarify the circumstances in which different treatment
on grounds of pregnancy constitutes discrimination contrary to Community law and, in particular, to
indicate the correct test to be applied to establish where discrimination exists under the various applicable
Community provisions. In its view, if the correct approach is to treat pregnancy and maternity as situations
requiring protection for a specified period, which cannot be equated with other absences from work, it will
be necessary to ascertain where that period is to begin and end. In the absence of specific legislation and
in application of the principle of subsidiarity, the EOC submits that determination of the commencing and
ending date of the protected period and the possibility of those dates being brought forward in certain
circumstances are matters to be determined in the employment contract.
18.
The United Kingdom states, first, that in the main proceedings the issue is the compatibility with
Community law of the contractual scheme drawn up by a particular employer, which is partially based
upon provisions of national law which implemented Community measures and which partially represents
particular contractual arrangements between employer and employee, which go beyond the minimum
protection for workers guaranteed by Directive 92/85. It goes on to say that the principle of equal pay for
men and women must be interpreted in the light of the provisions designed to protect women who are
pregnant or who have recently given birth, contained in Directive 92/85, and maintains that the measures
approved in each member state in implementation of that directive must be regarded as an indivisible
package of minimum rights, which include maternity leave, pay, protection against dismissal during
pregnancy and maintenance of employment rights. It follows that an employee cannot be allowed to
choose from among the rights in that package those which she sees as most advantageous, such as
maternity leave, and seek to disapply other provisions in that package, which seem less favourable to her,
such as the fact that the pay she receives in that period is less than full pay.
It concludes that the Community legislation should be interpreted as not precluding the application to
employees of the disputed provisions contained in the EOC maternity scheme.
19.
Ireland considers that Directive 92/85, adopted to protect the health and safety of pregnant workers
and to uphold their employment rights, contains the main provisions governing maternity in the case of
working women, except as regards circumstances not covered by the directive, in which case it is
necessary to rely on other provisions of Community law, in particular Directive 75/117 or Directive 76/207.
It considers, however, that in this case the answer to the preliminary questions is to be found in Directive
92/85, the provisions of which do not preclude the application of the disputed provisions to female
employees of the EOC.
20.
With respect to the first question, the Commission considers that the disputed stipulations, in so far as
they grant employees the right to receive full pay while on maternity leave of 14 weeks duration, conform
to Directive 92/85. However, under those same stipulations, if the employee does not return to work for at
least one month she must repay the difference between what she received and the amount of the
statutory maternity pay. To the extent to which that provision means that a woman who does not return to
work fails to receive, during maternity leave, income equivalent to what she would have 890 received if
she had been absent from work on account of illness, the Commission considers it incompatible with
Directive 92/85.
It suggests that the second and third questions be considered together, since both deal with situations
in which maternity leave partly coincides with unfitness for work on grounds of illness, and that they are
governed by the principle that an employee cannot be in both sets of circumstances at the same time.
That appears to the Commission to be a perfectly legitimate rule, provided that its application does not
mean that the employee is deprived of her physical and legal right to a continuous and protected period
of at least 14 weeks maternity leave. It would be contrary to the spirit of Directive 92/85 for an employee
to be compelled to give up her maternity leave in order to be able to take sick leave.
The Commission also suggests a combined answer to the fourth and fifth questions, since both deal
with the maintenance of rights under the employment contract during a womans absence on account of
maternity. The Commission considers that the right to annual leave and pension rights must continue to
accrue in the maternity leave period prescribed by art 8 of Directive 92/85. However, periods of leave to
care for a child, which exceed the length provided for by that article, do not enjoy protection and it would
not therefore be incompatible with that directive for the right to annual leave and pension rights not to
accrue during those periods.

VITHE JURISDICTION OF THE COURT OF JUSTICE TO ANSWER THE QUESTIONS SUBMITTED

BY THE INDUSTRIAL TRIBUNAL
21.
In the same way as the Commission in its written observations, I too have wondered whether the
Court of Justice has jurisdiction to answer the questions. At first sight, it would not appear particularly
useful to interpret Community provisions laying down the principle of equal pay and the principle of equal
treatment for men and women, and the provisions of Directive 92/85 relating to maternity leave and the
maintenance of employment rights during that period, when the disputes pending before the national
court, although certainly relating to the compatibility with Community law of certain clauses of the
applicants employment contracts, do not derive from the actual application of those clauses to them.
22.
In that connection, according to settled case law, by virtue of the cooperation between the Court of
Justice and the national courts provided for by art 177 of the Treaty it is solely for the national court before
which the dispute has been brought, and which must assume responsibility for the subsequent judicial
decision, to determine in the light of the particular circumstances of each case both the need for a
preliminary ruling in order to enable it to deliver judgment and the relevance of the questions which it
submits to the court. Consequently, where the questions submitted by the national court concern the
interpretation of a provision of Community law, the Court of Justice is, in principle, bound to give a ruling
(see the judgment in Aprile Srl (in liq) v Amministrazione delle Finanze dello Stato Case C-125/94 [1995]
ECR I-2919 (paras 1617)).
23.
But the court has also taken the view that in order to determine whether it has jurisdiction, it is a matter
for the Court of Justice to examine the conditions in which the case has been referred to it by the national
court. The spirit of cooperation which must prevail in the preliminary ruling procedure requires the national
court to have regard to the function entrusted to the 891 Court of Justice, which is to assist in the
administration of justice in the member states and not to deliver advisory opinions on general or
hypothetical questions (see the judgment in Meilicke v ADV/ORGA AG Case C-83/91 [1992] ECR I-4871
(para 25)).
24.
And it also said that a request for a preliminary ruling from a national court may be rejected only if it is
quite obvious that the interpretation of Community law sought by that court bears no relation to the actual
nature of the case or the subject-matter of the main action (see the judgment in Furlanis Costruzioni
Generali SpA v Azienda Nazionale Autonoma Strade (ANAS) Case C-143/93 [1995] ECR I-3633 (para
12)).
25.
However, it is beyond doubt that the disputes in which the Industrial Tribunal must adjudicate are not
hypothetical. They are real disputes, in which the applicants seek a declaration that certain clauses of
their employment contracts are contrary to Community law and therefore void, in order to preclude their
application to them in the future, should the case arise. Since national law allows them to bring such
actions, I am of the opinion that the questions submitted are objectively necessary in order to resolve the
disputes and that the Court of Justice must answer the questions (see the judgment in Union Royale
Belge des Socits de Football Association ASBL v Bosman Case C-415/93 [1996] All ER (EC) 97, [1995]
ECR I-4921 (para 65)).

VIICONSIDERATION OF THE PRELIMINARY QUESTIONS

26.
To date, the court has given several judgments on the application of the principle of equal pay and the
principle of equal treatment for men and women to cases involving the employment rights of workers who
are pregnant or who have recently given birth. Examples are Gillespie v Northern Health and Social
Services Board Case C-342/93 [1996] All ER (EC) 284, [1996] ECR I-475, dealing with pay during
maternity leave; Dekker v Stichting Vormingscentrum voor Jong Volwassenen (VJV-Centrum) Plus Case
C-177/88 [1990] ECR I-3941, concerning a refusal to appoint a pregnant woman; Handels- og
Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiverforening Case C-179/88 [1990] ECR I-
3979 (Hertz), concerning the dismissal of a woman on account of incapacity for work which commenced
after maternity leave and was attributable to an illness caused by the confinement; Habermann-
Beltermann v Arbeiterwohlfahrt, Bezirksverband Ndb/Opf-eV Case C-421/92 [1994] ECR I-1657,
concerning the possibility of declaring a contract void or capable of being avoided in consequence of the
legal prohibition of night work by pregnant women; Webb v EMO Air Cargo (UK) Ltd Case C-32/93 [1994]
4 All ER 115, [1994] ECR I-3567, concerning the non-comparability of the situation of a pregnant woman
who is unable to carry out the task for which she was recruited with that of a man suffering the same
incapacity for medical or other reasons; and Larsson v Ftex Supermarked A/S Case C-400/95 [1997]
ECR I-2757, concerning the possibility of taking into account, for the purposes of dismissal, of absence
due to a womans incapacity for work as a result of pregnancy, before the start of maternity leave. An
opinion has been delivered but judgment is still awaited in Caisse Nationale dAssurance Viellesse des
Travailleurs Salaris v Thibault Case C-136/9510 concerning the possibility of a woman not being given an
annual assessment report because she was absent from work on maternity leave; 892Pedersen v
Kvickly Skive Case C-66/96 (1997) ECJ Transcript (opinion) 10 July, concerning the right to equal pay
when a womans pregnancy renders her unfit for work; and Brown v Rentokil Ltd Case C-394/9611,
concerning the possibility of dismissal of a pregnant worker because her absence due to unfitness for
work attributable to pregnancy exceeded the period which, under the employment contract, rendered
workers liable to dismissal on grounds of illness.
10
Ed: for the judgment and opinion, now see Caisse Nationale dAssurance Viellesse des Travailleurs Salaris v Thibault
Case C-136/95 [1998] All ER (EC) 385, [1998] ECR I-2011.
11
Ed: for the judgment and opinion, now see Brown v Rentokil Ltd Case C-394/96 [1998] All ER (EC) 791.

A common denominator in all those cases is that it was or is necessary for them to be resolved by
reference to an interpretation of the Community legislation which was in force at the material time, namely
either art 119 of the Treaty and Directive 75/117, in the case of the principle of equal pay; or Directive
76/207 in the case of equal treatment regarding access to employment, training and promotion, and
working conditions.
27.
In order to deal with the present case, however, recourse may now be had to Directive 92/85, since
the wording of the clauses which the applicants in the main proceedings seek to have declared void is the
wording adopted in order to bring them into line with the text of that directive.
 Since it is an individual directive which, as indicated by its very title, purports to apply measures to
encourage improvements in the safety and health at work of pregnant workers and workers who have
recently given birth or are breastfeeding, I consider that, in order to answer the preliminary questions, it
will be necessary to rely primarily on the provisions of that directive and, on a subsidiary basis, on the
remainder of the Community legislation of more general scope.
28.
Before examining the various questions, I must point out that, in order to answer them, it is necessary
to reformulate them since, otherwise, in view of the complexity and originality of the national legislation on
which the disputed clauses are partly based, the terminological problems associated with the various
concepts would be insoluble, and the answers might not be helpful to the national court.
I shall therefore reformulate the questions, relying for that purpose, in view of the lack of detail in the
order for reference regarding the factual and legal background to the disputes, upon the documents
produced by the national court and by the parties which have submitted observations in these
proceedings, and on the replies to the written questions put by the court to the respondent and to the UK
government, in order to clarify the scope of some of the disputed clauses.

(1) The first question

29.
I infer that by this question the national court wishes to ascertain whether the Community law
provisions which it cites mean that an employer, who is prepared, for the benefit of his employees, to go
beyond the legal provisions applicable to pay during maternity leave, is precluded from imposing, as a
quid pro quo, by means of clauses like those of the EOC Maternity Scheme, the requirement that the
employees declare, before commencing maternity leave, that they intend returning to work and give an
undertaking to repay the difference between the full salary paid to them and the statutory maternity pay
that they would have received had they not given an 893 undertaking to return to work, in the event of
their failing to come back to work for at least one month.
30.
Among the rights connected with the employment contract which must be guaranteed to women,
under art 11(2)(b) of Directive 92/85, is the maintenance of a payment and/or entitlement to an adequate
allowance during the period of maternity leave, the minimum period of which is to be 14 continuous
weeks, of which two are compulsory. Under art 11(3), the allowance is deemed to be adequate if it
guarantees income at least equivalent to that which the worker concerned would receive in the event of a
break in her activities on grounds connected with her state of health, subject to any ceiling laid down
under national legislation.
31.
Under the EOC maternity scheme, employees who fulfil the stricter requirements regarding length of
service and type of contract are entitled to continue to receive full pay for three months and one week. If
their length of service is less, they will receive statutory maternity pay, which consists of nine-tenths of
normal pay over a given period, to be paid for the first six weeks, and a fixed amount which at present is
5455 per week for the remainder of the period. If she does not fulfil the conditions for entitlement to the
latter allowance, the worker may apply for the maternity allowance, which amounts to 5455 per week.
In addition, the EOC Staff Handbook provides, with respect to remuneration for an employee who is
unfit for work through illness, that he will be entitled to full pay for a period not exceeding the first six
months, within a period of 12 months. Thereafter, he will receive half-pay for a maximum of 12 months
within a period of four years. If incapacity continues, in principle nothing will be payable by the employer
and the employee will then be able to receive certain state benefits.
The issue is whether, as the applicants in the main proceedings and the Commission contend,
Directive 92/85 precludes the EOC Maternity Scheme from requiring employees to give an undertaking to
return to work after maternity leave in order to receive full pay and to refund the difference between that
amount and statutory maternity pay if they do not return, on the ground that such a requirement means
that a woman in those circumstances will not have received income equivalent to that to which she would
have been entitled if the break in her activities had been attributable to an illness. In the Commissions
view, art 11(2)(b) and (3) of Directive 92/85 must be construed as meaning that the term adequate, used
in the text in relation only to the allowance, must also apply to remuneration, so that it will be necessary to
consider the question of adequacy in each case, having regard to the legitimate expectations of the
particular employee, and drawing a comparison with the amount she would receive if on sick leave.
32.
I do not agree with that interpretation, for the following reasons. First, because the text of art 11(2)(b)
requires maintenance of a payment, not of pay. It is thus accepted at the outset that the income received
by a worker on maternity leave does not coincide with the earnings she receives while at work or in any
other circumstances. Indeed, it would be difficult to pay her a higher sum when she was not working than
she would receive if working. I must therefore conclude that the directive does not stand in the way of
remuneration being, during maternity leave, lower than the wages received by the worker whilst at work or
in any other circumstances.
894
33.
Secondly, because the distinction between pay and allowances is linked with the source of the income
payable to the woman. Thus, with regard to pay, reference must be made to art 119 of the Treaty which
defines it broadly as:
 the ordinary basic or minimum wage or salary and any other consideration, whether in cash
or in kind, which the worker receives, directly or indirectly, in respect of his employment from his
employer.
That definition has been completed by the case law of this court which, since 1971, has included within
that concept immediate or future consideration and, in 1990, it added that the benefits paid by an
employer to a worker in respect of his employment are to be classified as pay whether they are paid
under a contract of employment, by virtue of legislative provisions or on a voluntary basis (see the
judgments in Defrenne v Belgium Case 80/70 [1971] ECR 445 (para 6) and Barber v Guardian Royal
Exchange Assurance Group Case C-262/88 [1990] 2 All ER 660, [1990] ECR I-1889 (para 20),
respectively).
By a process of elimination, the allowance will, necessarily, be any income of a public or private nature
received by a woman whilst on maternity leave and not paid by the employer in respect of her
employment. That definition, it seems to me, will in most cases include amounts paid by social security
authorities, either directly or through management agencies.
In my opinion, the terminology used for the two concepts supports that interpretation. Where the
provision refers to a payment, it is accompanied by the word maintenance. This means that, either by
operation of law or by virtue of a collective agreement or individual contract, it will be the employer who
must ensure that the employee receives, whilst on maternity leave and in respect of the employment
relationship, a given level of income which, as I pointed out earlier, does not necessarily have to coincide
with full pay. On the other hand, where the provision refers to an . . allowance, it is accompanied by the
words entitlement to, which brings it closer to the sphere of social security protection and distances it
somewhat from the concept of pay.
34.
The distinction drawn by the provision between the maintenance of a payment and entitlement to an .
allowance is helpful in considering whether the word adequate refers only to the allowance or must be
deemed to extend to the payment, as the Commission submits. Given that the consideration paid by the
employer to the worker in respect of her employment is, as a general rule, greater than the amount of the
benefits paid by the social security authorities, for the simple reason that such benefits are usually based
on contributions and consists, to a greater or lesser extent, of a percentage of income, it must be
concluded that, where Directive 92/85 indicates that the allowance will be deemed adequate if it
guarantees income at least equivalent to that which the worker would receive in the event of a break in
her activities on grounds connected with her state of health, subject to any ceiling laid down by national
legislation, it is referring to allowances paid by national social security schemes and not to pay from the
employer in respect of employment. Clearly the latter must also be adequate, but I do not believe that, in
practice, an employer, whether public or private, who has to negotiate working conditions with his
employees, individually or collectively, will be able to pay less to his employees during maternity leave
than they would receive from the social security authorities by way of sickness benefits.
That interpretation is supported by the last recital in the preamble to Directive 92/85, in which it is
stated:
895
the concept of an adequate allowance in the case of maternity leave must be regarded as a
technical point of reference with a view to fixing the minimum level of protection and should in no
circumstances be interpreted as suggesting an analogy between pregnancy and illness.
If the purpose of an adequate allowance is to set the minimum level of protection, I do not see how it
could mean that, in each specific case, a woman will be entitled to receive the same income during
maternity leave as she receives when unfit for work through illness.
35.
Thirdly, I do not agree with the interpretation advocated by the Commission because art I 1(2)(b) of
Directive 92/85 contemplates the possibility of guaranteeing at the same time maintenance of a payment
and entitlement to an adequate allowance. If that meant that both of them had to be adequate within the
meaning of the directiveby guaranteeing income at least equivalent to that which the worker would
receive in the event of a break in her activities on grounds of healththe result would be that she could
be entitled to receive from her employer the remuneration which the latter is required to pay his
employees in the event of incapacity for work and, in addition, the allowance provided by the social
security scheme by way of sickness benefit. On the contrary, I believe that the only possibility of both
being payable arises where the social security benefit, which normally represents a minimum, may be
supplemented by the employer, either by operation of law or under an agreement with his employees.
36.
Finally, there is yet another reason for which the interpretation proposed by the Commission seems to
me unacceptable. On examining the system set up by the respondent to pay its employees whilst they are
unfit for work, I note that there are three possible situations: in the first, for six months within a 12-month
period, they receive full pay. Thereafter, they are paid only half pay for a maximum of twelve months
within a period of four years. Finally, payment from the employer to a worker who is unfit for work ceases.
The Commission asserts that the adequacy of the income received during maternity leave will depend on
the legitimate expectations of the employee concerned regarding level of pay when she is unfit for work.
But, what will be the legitimate expectations of an employee who is already in the second stage, that is to
say when she is receiving only half pay? And I need not speculate on the expectations of a person whose
incapacity for work has gone on to the point that she no longer receives any pay at all.
37.
 It is clear from the documents before the court, first, that the national legislation requires employers to
pay employees who fulfil certain conditions regarding length of service and level of earnings statutory
maternity pay, consisting of nine-tenths of their ordinary wages for the first six weeks of maternity leave
and a fixed amount which at present is 5455 per week for the remainder of the period. The latter
amount coincides with the maternity allowance and with the statutory sickness benefit paid by the
employer for a maximum of 28 weeks.
It has also been shown that the respondent improves on those conditions for the benefit of its
employees by paying them the difference between statutory maternity pay and full pay for 14 weeks
maternity leave, requiring them in return to declare, before starting leave, that they intend returning to
work and to actually return for at least one month.
896
38.
In the light of the interpretation which I propose for art 11(2)(b) of Directive 92/85, I do not believe that
its provisions mean that an employer, who is prepared, for the benefit of his employees, to go beyond the
legal provisions applicable to pay during maternity leave, is precluded from imposing, as a quid pro quo,
by means of clauses like those of the EOC Maternity Scheme, the requirement that the employees
declare, before commencing maternity leave, that they intend returning to work and give an undertaking
to repay the difference between the full salary paid to them and the statutory maternity pay that they
would have received had they not given an undertaking to return to work, in the event of their failing to
come back to work for at least one month.
39.
The applicants in the main proceedings also allege discrimination regarding equal pay for men and
women stemming from the fact that, to be entitled to receive full pay whilst on maternity leave, they must
undertake to return to work and, if they fail to do so, they are obliged to repay the difference between that
amount and the statutory maternity pay, since that condition is not imposed on workers in general as a
precondition for receiving full pay whilst they are unfit for work, nor are such workers required to repay the
difference if they do not resume work when declared fit to do 50.
40.
According to settled case law of this court, discrimination can arise only through the application of
different rules to comparable situations or the application of the same rule to different situations (see the
judgment in Finanzamt Kln-Altstadt v Schumacker Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-
225 (para 30)). Now, in Gillespie [1996] All ER (EC) 284, [1996] ECR I-475 (paras 17, 20), in which it was
sought to clarify whether the principle of equal pay contained in art 119 of the Treaty and developed in
Directive 75/117 made it compulsory to maintain full pay for workers on maternity leave, the Court of
Justice held that women taking maternity leave provided for by national legislation are in a special
position which requires them to be afforded special protection, but which is not comparable either with
that of a man or with that of a woman actually at work, reaching the conclusion that neither art 119 of the
Treaty nor Directive 75/117 impose the obligation that workers should receive full pay during maternity
leave.
In response to the question whether Community law laid down specific criteriaand, if so, whatfor
determining the amount of benefit to be paid to workers on maternity leave, the court answered ([1996] All
ER (EC) 284, [1996] ECR I-475 (para 20)):
Nor did those provisions [art 119 of the Treaty and art 1 of Directive 75/117] lay down any
specific criteria for determining the amount of benefit to be paid to them during that period. The
amount payable could not, however, be so low as to undermine the purpose of maternity leave,
namely the protection of women before and after giving birth. In order to assess the adequacy of
the amount payable from that point of view, the national court must take account, not only of the
length of maternity leave, but also of the other forms of social protection afforded by national law in
the case of justified absence from work. There is nothing, however, to suggest that in the main
proceedings the amount of the benefit granted was such as to undermine the objective of
protecting maternity leave.
In Gillespie [1996] All ER (EC) 284, [1996] ECR I-475, the workers had received the following benefits
during maternity leave, under their collective 897 agreement: their full weekly pay for the first four weeks,
nine-tenths of their full pay for the next two weeks and, finally, half their full pay for 12 weeks, such
conditions being more advantageous than those laid down by the general applicable legislation, which
provided for the payment of nine-tenths of full weekly pay for six weeks and then a flat-rate allowance of
4795 per week for the next 12 weeks.
41.
I would also add that the situation of a man rendered unfit for work by illness and that of a woman
taking maternity leave are certainly not comparable in any way. The man, had he not been ill, would be
working and, on being declared fit, would have to return to work, whereas maternity leave is granted to a
woman in order to safeguard her biological state during and after pregnancy and to protect the special
relationship between mother and child in the period following confinement (see the judgment in Hofmann
v Barmer Ersatzkasse Case 184/83 [1984] ECR 3047 (para 25)). Moreover, as I stated in the opinion
which I delivered in Pedersen v Kvickly Skive Case C-66/96 (1997) ECJ Transcript (opinion) 10 July,
during maternity leave a woman is not only released from work but also from any other obligation under
her employment contract, whereas a man or woman declared unfit for work is required to undergo the
therapeutical treatment prescribed by the doctor in order to assist recovery.
42.
 For the reasons which I have just set out I consider that art 119 of the Treaty and Directive 75/117
likewise do not mean that an employer, who is prepared, for the benefit of his employees, to go beyond
the legal provisions applicable to pay during maternity leave, is precluded from imposing, as a quid pro
quo, by means of clauses like those of the EOC Maternity Scheme, the requirement that the employees
declare, before commencing maternity leave, that they intend returning to work and give an undertaking
to repay the difference between the full salary paid to them and the statutory maternity pay that they
would have received had they not given an undertaking to return to work, in the event of their failing to
come back to work for at least one month.

(2) The second question

43.
By this question, the national court appears to be asking whether, under the provisions of Community
law which it cites, it is not permissible, by recourse to clauses like those forming part of the respondents
Maternity Scheme, to provide that, where a worker has indicated that she wishes to commence maternity
leave on any date in the six weeks prior to the expected week of confinement and is declared unfit for
work on account of pregnancy immediately before that date, then, if the birth occurs whilst she is in that
situation, the commencement of her maternity leave may be backdated to the later of the following two
dates: the beginning of the period of sick absence or the start of the sixth week prior to the expected week
of confinement.
44.
As regards the length of maternity leave, Directive 92/85 imposes two obligations on the member
states: the first is that workers must be allowed a continuous period of at least 14 weeks before and/or
after the confinement; the second is that that period must include compulsory maternity leave of at least
two weeks, likewise allocated before and/or after confinement. In accordance with those provisions, the
United Kingdom gave to all women a general right, which previously had not existed, to take periods of
maternity leave of that duration, the same provisions being contained in the EOC Maternity Scheme. The
compulsory maternity leave of two weeks commences in the United 898 Kingdom on the day of
confinement, and maternity leave may be extended for the period necessary to satisfy that obligation.
45.
The applicants consider it discriminatory for a woman who is unfit for work to be able to exercise her
right to remain on sick leave with full pay if her illness is not attributable to pregnancy and she gives birth
whilst in that situation.
46.
I do not agree with that interpretation. Before Directive 92/85 imposed on the member states the
obligations under review here, that is to say when the only Community provision dealing with pregnancy
and maternity was art 2(3) of Directive 76/207which did no more than authorise them to adopt
provisions to uphold womens specific rights for those two reasonsthe court held, in Hertz [1990] ECR I-
3979 (para 15) that it is for every member state to fix periods of maternity leave in such a way as to
enable female workers to absent themselves during the period in which the disorders inherent in
pregnancy and confinement occur.
47.
On the basis that the purpose of maternity leave is to permit a worker, without prejudice to her
employment rights, to be absent from work because of imminent or recent maternity, and provided that
the length of the periods laid down by art 8 of Directive 92/85 is complied with, I conclude that neither the
provisions of that directive nor those of Directive 76/207 preclude a provision like that contained in the
respondents Maternity Scheme under which, where an employee has indicated that she wishes to
commence maternity leave on any date in the six weeks prior to the expected week of confinement and is
declared unfit for work on account of pregnancy immediately before that date, then, if the birth occurs
while she is in that situation, the commencement of her maternity leave may be backdated to the later of
the following two dates: the beginning of the period of sick absence or the start of the sixth week prior to
the expected week of confinement.

(3) The third question

48.
By this question, I believe, the national court wishes to ascertain whether under Community law it is
not permissible, by recourse to provisions like those in the EOC Maternity Scheme, to prohibit a woman
who has begun her maternity leave or is on unpaid maternity leave from being accorded sick leave on full
paybeing entitled in the latter case to claim only statutory sick payand to impose the requirement that,
in order to be granted such leave, she has stated, three weeks in advance, her intention to return to work
on a specified day, thereby bringing to an end, if the birth has occurred, her special maternity leave
arrangements.
49.
In order to answer that question, I think a distinction must be drawn between, on the one hand,
maternity leave properly so called, as provided for by Directive 92/85, that is to say, the continuous period
of 14 weeks allocated before and/or after confinement, during which in any event workers rights under
their employment contracts are safeguarded and, if appropriate, a payment to them or entitlement to an
adequate allowance is maintained, that period being the one provided for both by national legislation and
by the provisions of the EOC Maternity Scheme, and, on the other hand, any other leave, paid or
otherwise, which the national legislature or the employer, by granting enhanced contractual terms, sees fit
to afford to a worker who has recently given birth. I shall consider first maternity leave properly so called.
899
50.
It seems to me to be entirely logical for a woman who is on maternity leave not to be able, at the same
time, to be declared unfit for work on account of illness. What sense would it make, legally, if she could be
regarded as being simultaneously in both situations? Would it be appropriate, nevertheless, to conclude
that she can interrupt her maternity leave, take sick leave and, on recovering, go back to the previous
situation?
51.
I see a number of reasons for rejecting that interpretation. In the first place, it must be borne in mind
that, although a period of 14 weeks maternity leave must without fail be provided for in the legislation of
the member states, the internal regulations of undertakings and in contracts of employment, as far as the
worker is concerned, apart from the two weeks compulsory maternity leave which, in the United Kingdom,
commence on the day of the birth, the entitlement may be waived. Secondly, Directive 92/85 clearly
provides, in art 8, that the maternity leave is to comprise a continuous period of at least 14 weeks. It is not
therefore possible to take part of that period and then take the remainder at a later stage.
Thirdly, under art 11(4) of Directive 92/85, the maintenance of a payment and/or entitlement to an
adequate allowance may be made conditional upon the worker concerned fulfilling the conditions of
eligibility for such benefits under national legislation, amongst which it is not appropriate to include earlier
periods of work exceeding 12 months immediately prior to the expected date of confinement. The
possibility of treating maternity leave as having ended and being given paid sick leave will, for a woman
who falls ill after giving birth and is not entitled to receive any income during maternity leave, offer
indubitable advantages.
52.
For those reasons, I consider that the provisions of Directive 92/85 do not preclude a clause like that
contained in the EOC Maternity Scheme under which a woman who is in the course of her 14 weeks
maternity leave must, provided that the two weeks of compulsory leave have been completed, accept that
her special maternity leave arrangements have definitively ceased so that she can be granted paid sick
leave.
53.
As regards unpaid maternity leave, the entitlement to and duration of which are determined by
reference to the conditions fulfilled by the workers and for which remuneration is limited to the payment by
the employer of statutory maternity pay for the first four weeks, I do not consider that it can be regarded
as maternity leave within the meaning of Directive 92/85. In the present case, the workers are entitled to
be absent from work for up to 52 weeks. That period includes the 14 weeks maternity leave and the
remainder will be unpaid maternity leave. The applicants consider that the clause in the Maternity
Scheme which makes receipt of full pay, if they fall ill during the period of unpaid leave, conditional upon
their giving up their special arrangements, discriminates against women who find themselves in that
situation.
That type of unpaid leave is a right which, being reserved in the United Kingdom exclusively for
women, falls within the exception provided for in art 2(3) of Directive 76/207 which enables the member
states to provide for special treatment for women, in particular as regards pregnancy and maternity, for
which there is no parallel in the treatment accorded to their male colleagues.
54.
I consider that Community law does not preclude the application of a clause of the kind at issue here.
Although it does provide for different treatment as between men and women, in that women, if they fall ill
whilst on 900 unpaid maternity leave are not automatically granted the right to sick leave on full pay (they
are entitled only to statutory sick pay), whereas male workers automatically enjoy that right for the first six
months of a period of 12 months, it must also be recognised that the situations are different, since men
are not granted any right to unpaid parental leave. But there is another important difference which
becomes apparent when both fall ill: to be able to claim that right, a male worker must be at work (and I
do not think that the fact that he may fall ill whilst on annual leave changes that assessment), whilst the
woman is on unpaid leave, during which she is relieved of the obligation to work.
55.
In those circumstances, and provided that the two weeks compulsory maternity leave have been
taken, I consider that none of the provisions cited by the national court precludes recourse to stipulations
of the kind contained in the EOC Maternity Scheme which prohibit a woman who has begun her maternity
leave or is on unpaid maternity leave from being accorded sick leave on full paybeing entitled in the
latter case to claim only statutory sick payand to impose the requirement that, in order to be granted
such leave, she has stated, three weeks in advance, her intention to return to work on a specified day,
thereby bringing to an end, if the birth has occurred, her special maternity leave arrangements.

(4) The fourth question

56.
By this question, the national court seeks a ruling as to whether, under the Community law provisions
which it cites, it is not permissible, by recourse to a stipulation of the kind contained in the EOC Maternity
Scheme, to limit the period for which annual leave entitlement accrues to the 14 weeks maternity leave
and to exclude such accrual whilst the woman concerned is on leave of another kind in order to care for a
newborn child.
57.
Article 11(2)(a) of Directive 92/85 requires the member states, whilst a worker is taking the maternity
leave provided for in art 8, to safeguard the rights connected with her employment contract. There is no
doubt that the accrual of annual leave is one of the rights connected with an employees employment
contract. Nor is there any doubt that, for the 14-week period of maternity leave in the United Kingdom,
regardless of any right to receive any particular income, employees are guaranteed the accrual of annual
leave.
The applicants nevertheless claim the right to the accrual of annual leave whilst they are on unpaid
maternity leave, which is granted to them if they fulfil certain requirements regarding length of service.
58.
As I stated earlier, Directive 92/85 merely lays down the obligation to safeguard rights connected with
the employment contract whilst the worker is taking the maternity leave provided for in art 8, that is to say
for 14 weeks. For the remaining time, up to the 52 weeks representing the maximum period for which an
employee of the EOC may be absent from work on account of pregnancy and maternity, she will be in the
special situation which I described when considering unpaid maternity leave in connection with the
answer to be given to the third question. For the reasons which I put forward at that time, I consider that
Directive 92/85 does not impose the obligation to safeguard an employees rights under her employment
contract, such as the accrual of annual leave, beyond the 14 weeks maternity leave provided for in art 8.
59.
The applicants maintain that that clause, which is ostensibly neutral, involves indirect discrimination,
since, during their working life, it is women 901 who most frequently take unpaid leave, since they are
allowed to take unpaid maternity leave.
60.
Quite apart from the fact that I consider that reserving solely to women the availability of unpaid leave
to look after a newborn child does not help promote equality of opportunity between the sexes, since what
it does in reality is to perpetuate in society the idea that it is women who as a matter of priority should take
care of the childrenwith all the concomitant adverse effects on their future careersI do not share the
view put forward by the applicants.
61.
The Court of Justice has repeatedly held that, to justify a finding of indirect discrimination, a provision
formulated in neutral terms, which in this case is contained in the Staff Handbook, whereby any unpaid
leave will reduce annual leave in proportion to the length of the leave of absence, must in fact be
detrimental to a much larger number of women than men 12.
12
See the judgments in Gerster v Freistaat Bayern Case C-1/95 [1997] ECR I-5253 (para 30), Kording v Senator fr
Finanzen Case C-100/95 [1997] ECR I-5289 (para 16), Megner v Innungskrankenkasse Vorderpfalz Case C-444/93
[1996] All ER (EC) 212, [1995] ECR I-4741 (para 24) and De Weerd, ne Roks v Bestuur van de
Bedrijfsvereniging voor de Gezondheid, Geestelijke en Maatschappelijke Belangen Case C-343/92 [1994] ECR I-571
(para 33).

However, the argument that that provision adversely affects a much larger number of women because
they more frequently take unpaid leave cannot succeed in the present case since the leave taken by the
women referred to by the applicants in support of their allegation of indirect discrimination reflects special
arrangements which have no connection with unpaid leave taken voluntarily for personal reasons, which
is available to both men and women. In the first place, in the United Kingdom only women may take
unpaid leave to look after a newborn child, of a predetermined duration and with the right to return to
work13. In the second place, unpaid leave of that kind is a protective measure covered by art 2(3) of
Directive 76/207, which is afforded to a woman because she has recently given birth and the grant of
such leave and the rules concerning it, where it exceeds the minimum requirements of Directive 92/85,
are matters reserved exclusively to the member states.

13
In Coloroll Pension Trustees Ltd v Russell Case C-200/91 [1995] All ER (EC) 23, [1994] ECR I-4389 (para 103),
the Court of Justice stated in reply to questions from the High Court as to whether, in the sphere of equal pay for men
and women, art 119 of the Treaty was applicable to company pension schemes which at all times have had members of
only one sex, thata worker cannot rely on art 119 in order to claim pay to which he could be entitled if he belonged to
the other sex in the absence, now or in the past, in the undertaking concerned of workers of the other sex who perform
or performed comparable work. In such a case, the essential criterion for ascertaining that equal treatment exists in the
matter of pay, namely the performance of the same work and receipt of the same pay, cannot be applied.

62.
For the foregoing reasons, I am of the opinion that Community law does not preclude the limitation, by
means of stipulations such as those of the EOC Maternity Scheme, of the time during which annual leave
accrues to the 14 weeks maternity leave and the exclusion of such accrual during unpaid maternity leave.

(5) The fifth question

63.
By this question the national court wishes to ascertain whether any of the provisions which it cites
precludes imposition, for the acquisition of pension rights during maternity leave, of the condition that the
woman must receive pay or statutory maternity pay from her employer, with the result that she is
prevented from acquiring pension rights if, during maternity leave, she 902 receives income from funds
unconnected with the employer or if she receives no income at all, and likewise if she takes unpaid
maternity leave.
64.
I must point out once again that art 11(2)(a) of Directive 92/85 provides that workers must, when on
maternity leave, which is to be of at least 14 weeks duration, have the rights connected with their
employment contracts maintained. There is no doubt at all that the accrual of pension rights, particularly
where, as in this case, they form part of an occupational scheme financed entirely by the employer, is one
of the rights connected with the workers employment contract.
However, Cl 8.1 of the EOC Maternity Scheme provides that where an employee is not entitled to
receive full pay during her maternity leave, the 14 weeks absence will be taken into account for the
purposes of the accrual of pension rights only if she is entitled to receive statutory maternity pay.
65.
Consequently, an EOC employee who does not meet the requirements of the Maternity Scheme for
the employer to pay her statutory maternity pay, both in the case of her receiving during maternity leave
any of the state benefits available in respect of maternity and in the case of her receiving no income at all,
will be prevented from accruing any pension rights under the occupational scheme of which she is a
member.
66.
It is true that such a possibility appears to be acceptable by virtue of art 6(g) of Directive 86/378, in so
far as it provides that provisions contrary to the principle of equal treatment are to include those based on
sex, which suspend the retention or acquisition of rights during periods of maternity leave or leave for
family reasons which are granted by law or agreement and are paid by the employer. In fact the rule in the
Maternity Scheme at issue is consistent with that provision: pension rights continue to accrue whilst pay,
in the form either of salary or of statutory maternity pay, is paid by the employer. On the other hand, the
accrual of pension rights is interrupted when the income received by the woman during that period is paid
out of public or private funds unconnected with the employer or when she receives no income at all.
67.
However, despite the fact that the wording which Council Directive (EC) 96/97 amending Directive
86/378 on the implementation of the principle of equal treatment for men and women in occupational
social security schemes (OJ 1997 L46 p 20) laid down as the new version of that article has not changed
that provision, I consider that art 11(2)(a) of Directive 92/85, under which workers are to be guaranteed
during maternity leave of a continuous period of at least 14 weeks the rights connected with their
employment contracts, must be interpreted as meaning that the accrual of pension rights under an
occupational scheme during the 14 weeks maternity leave may not be made conditional upon the
workers receiving, in the form of pay or statutory maternity pay, income paid by the employer.
In my opinion, it would be contrary to the aim pursued by Directive 92/85 if, during the period of
maternity leave for which it provides, the guarantee of one of the rights connected with the employment
contract, seen as an unconditional right, could be made subject, by virtue of a provision of the kind now
under consideration, to the condition that the woman must receive pay from her employer during that
period, when the directive itself provides that the member states are entitled to make the right to pay or to
an allowance conditional upon the worker concerned fulfilling the requirements laid down by national
legislation in order to qualify for those advantages.
903
68.
On the other hand, on the basis of the reasons set out in relation to the third and fourth questions
regarding unpaid maternity leave, I consider that Community law does not preclude a woman from being
denied the possibility of accruing pension rights under an occupational scheme whilst she is on unpaid
leave of that kind.
69.
In the light of the reasons which I have just expounded, I am of the opinion that art 1 1(2)(a) of
Directive 92/85, in conjunction with the provisions of art 8 thereof, precludes making the accrual of
pension rights under an occupational scheme during the 14 weeks maternity leave conditional upon the
worker receiving, in the form of wages or statutory maternity pay, income paid by her employer.

VIIICONCLUSION
70.
In view of the foregoing considerations, I propose that the following answers be given to the questions
submitted by the Industrial Tribunal, Manchester:
(1) In circumstances such as those of the present case, neither art 119 of the EC Treaty, nor
Council Directive (EEC) 75/117 on the approximation of the laws of the member states relating to
the application of the principle of equal pay for men and women, nor Council Directive (EEC)
76/207 on the implementation of the principle of equal treatment for men and women as regards
access to employment, vocational training and promotion, and working conditions, nor Council
Directive (EEC) 92/85 on the introduction of measures to encourage improvements in the safety
and health at work of pregnant workers and workers who have recently given birth or are
breastfeeding precludes the application of clauses of the kind examined in this opinion under
which: (1) an employer, who is prepared, for the benefit of his employees, to go beyond the legal
provisions applicable to pay during maternity leave, imposes, as a quid pro quo, by means of
clauses like those of the EOC Maternity Scheme, the requirement that the employees declare,
before commencing maternity leave, that they intend returning to work and give an undertaking to
repay the difference between the full salary paid to them and the statutory maternity pay that they
would have received had they not given an undertaking to return to work, in the event of their failing
to come back to work for at least one month; (2) where a worker has indicated that she wishes to
commence maternity leave on any date in the six weeks prior to the expected week of confinement
and is declared unfit for work on account of pregnancy immediately before that date, then, if the
birth occurs whilst she is in that situation, the commencement of her maternity leave may be
 backdated to the later of the following two dates: the beginning of the period of sick leave or the
start of the sixth week prior to the expected week of confinement; (3) a woman who has begun her
maternity leave or is on unpaid maternity leave is not allowed to be accorded sick leave on full pay
being entitled in the latter case to claim only statutory sick payand is required, in order to be
granted such leave, to have stated, three weeks in advance, her intention to return to work on a
specified day, thereby bringing to an end, if the birth has occurred, her special maternity leave
arrangements, provided that the two weeks of compulsory maternity leave have already been
taken; (4) the time for which annual leave accrues is limited to the 14 weeks duration of the 904
maternity leave, and such accrual is excluded during unpaid maternity leave.
(2) Article 11(2)(a) of Directive 92/85, in conjunction with art 8 thereof, precludes making the
accrual of pension rights under an occupation scheme during the 14 weeks maternity leave
conditional upon the employees receiving, by way of salary or statutory maternity pay, income from
her employer.

27 October 1998.

The COURT OF JUSTICE delivered the following judgment.

1.
By order of 15 October 1996, received at the Court of Justice of the European Communities on 23
December 1996, the Industrial Tribunal, Manchester, referred to the Court of Justice for a preliminary
ruling under art 177 of the EC Treaty five questions on the interpretation of art 119 of the Treaty, Council
Directive (EEC) 75/117 on the approximation of the laws of the member states relating to the application
of the principle of equal pay for men and women (OJ 1975 L44 p 19), Council Directive (EEC) 76/207 on
the implementation of the principle of equal treatment for men and women as regards access to
employment, vocational training and promotion, and working conditions (OJ 1976 L39 p 40) and Council
Directive (EEC) 92/85 on the introduction of measures to encourage improvements in the safety and
health at work of pregnant workers and workers who have recently given birth or are breastfeeding (OJ
1992 L348 p 1).
2.
Those questions were raised in proceedings between Mrs Boyle and others and their employer, the
Equal Opportunities Commission (the EOC), concerning the Maternity Scheme applied by the latter to its
staff. The national court states that the EOC is agreed to be an emanation of the state for the purposes,
as far as it is concerned, of the direct effect of the directive at issue.

The national legislation

3.
In the United Kingdom, the Employment Rights Act 1996 grants female employees a general right to
maternity leave. Section 72 provides that that leave commences either on the date which the employee
notifies to her employer as the date on which she intends her period of absence from work to commence,
or on the first day after the beginning of the sixth week before the expected week of childbirth on which
she is absent from work wholly or partly because of pregnancy, whichever is the earlier.
4.
Section 73(1) of the Act provides that the maternity leave period continues for the period of 14 weeks
from its commencement or until the birth of the child, if later. In any event, an employee may not work
during the period of two weeks commencing with the date of childbirth.
5.
In addition, ss 79 to 85 of the Act grant employees who have the general right to maternity leave and
who have been continuously employed for a period of not less than two years at the beginning of the
eleventh week before the expected week of childbirth, the right to return to work with their employer at
any time during the period of 29 weeks from the beginning of the week in which childbirth occurred.
6.
According to s 164 of the Social Security Contributions and Benefits Act 1992, pregnant employees
who have worked for a continuous period of at least 26 weeks ending with the week immediately
preceding the fourteenth week 905 before the expected week of confinement and whose earnings exceed
a certain level are to be entitled to receive from their employer payments known as statutory maternity
pay (SMP) if they have ceased working, wholly or partly because of pregnancy or confinement.
7.
Section 165(1) of that Act provides that SMP is to be payable for a maximum period of 18 weeks. For
women who have worked for a continuous period of at least two years preceding the fourteenth week
before the expected week of confinement for an employer required to pay them SMP, s 166 specifies that
SMP is to be equivalent to nine-tenths of the womans normal weekly earnings for the first six weeks, and
for the following 12 weeks is to be payable at a prescribed rate. At the material time that prescribed rate
was 5455. Women who do not satisfy the condition relating to length of service receive the fixed rate for
the whole 18 weeks.
8.
Furthermore, ss 151 to 163 of the Social Security Contributions and Benefits Act 1992 grant male and
female employees who are incapable of work the right to receive payments known as statutory sick pay
(SSP) from their employer for a period of up to 28 weeks, at the rate of 5455 per week.
 The dispute in the main proceedings
9.
The six applicants in the main proceedings are all employees of the EOC of childbearing age. They
have completed at least one years service with their employer and are not employed on a casual,
standby or short-notice appointment, nor are they employed on a fixed-term appointment of less than two
years. At least three of them have taken maternity leave in the recent past.
10.
The employment contract entered between EOC and its employees comprises, first, the Staff
Handbook, which applies to all workers and, second, the Maternity Scheme, which applies to female
workers.
11.
According to the Staff Handbook, staff who are unfit to work because of illness are entitled to their full
salary for a maximum of six months in any 12-month period. Thereafter, they receive half pay up to a
maximum of 12 months in any four-year period. Another clause provides that any leave taken without pay
reduces the annual leave entitlement by a proportion of the amount of unpaid leave taken.
12.
The dispute in the main proceedings centres on the Maternity Scheme. The persons concerned
applied to the Industrial Tribunal, Manchester, for a declaration that certain conditions of the scheme are
void or unenforceable in so far as they discriminate against female employees and are thus contrary to art
119 of the Treaty or Directives 75/117, 76/207 or 92/85.
13.
According to one of those clauses, any member of staff who has rendered at least one years paid
service with the EOC and is not employed on a casual, standby or short-notice appointment, or employed
on a fixed-term appointment of less than two years is entitled to three months and one weeks maternity
leave on full pay for the period of continuous absence before and after childbirth. However, in order to
benefit from that right, the employee must state that she intends to return to work in the EOC after
childbirth and agree to be liable to repay any payment made during that period (excluding SMP to which
she is entitled in any event), should she fail to return.
14.
According to another contested clause of the Maternity Scheme, a member of staff who is entitled to
paid maternity leave may also qualify for supplementary unpaid maternity leave provided, inter alia, that
the total of those two periods of leave do not exceed 52 weeks.
906
15.
Furthermore, the Maternity Scheme provides that if a member of staff specifies that she wishes to
begin her maternity leave during the six weeks before the expected week of childbirth, and is absent on
account of a pregnancy-related sickness immediately before the date on which she asked to begin her
maternity leave and childbirth occurs during that period of absence, the date on which paid maternity
leave commences can be brought forward to whichever is the later of the sixth week before the expected
week of childbirth and the beginning of the period of absence on account of sickness.
16.
Furthermore, according to the Maternity Scheme, paid sick leave is not granted once paid maternity
leave has begun or during a period of supplementary unpaid maternity leave. There may, however, be an
entitlement to SSP during unpaid maternity leave. Where a member of staff provides at least three weeks
notification of her intention to return to work on a specified date, she is entitled to paid sick leave from that
date. Paid sick leave following childbirth terminates the maternity leave and the supplementary unpaid
maternity leave arrangements.
17.
Finally, the Maternity Scheme states that members of staff who are not entitled to paid leave of
absence retain their contractual rights and benefits, except remuneration, during the first 14 weeks of
leave. In particular, annual leave continues to accrue. The period of absence only accrues for pension
purposes if the employee is in receipt of SMP.
18.
The national court points out that, pursuant to the foregoing provisions, employees on any form of paid
leave, apart from paid maternity leave, including sick leave and paid special leave, are not required to
agree to repay any part of their salary if they do not return to work after the period of leave. Furthermore,
it is agreed that substantially more women employees than men employees take periods of unpaid leave
in the course of their careers, largely because they take supplementary maternity leave.
19.
The Industrial Tribunal, Manchester, had doubts as to the compatibility of such provisions with
Community law and decided to stay proceedings in order to refer the following questions to the Court of
Justice for a preliminary ruling:
In circumstances such as those of the present case, do any of the following matters infringe the
prohibition of unfair and/or unfavourable treatment of women because of pregnancy, childbirth,
maternity and/or sickness in relation thereto under EC law (in particular Article 119 of the Treaty of
Rome and/or Council Directive 75/117/EEC and/or Council Directive 76/207/EEC and/or Council
Directive 92/85/EEC):
 (1) A condition that maternity pay, beyond the Statutory Maternity Pay, is paid only if the woman
states that she intends to return to work and agrees to be liable to repay such maternity pay if she
does not return to work for one month on the conclusion of maternity leave.
(2) A condition that where a woman, who is absent on paid sick leave with a pregnancy related
illness, gives birth during such absence, her maternity leave may be backdated to the later date of
either six weeks before the expected week of childbirth or when the sickness leave began.
(3) A prohibition on a woman, who is unfit for work for any reason whilst on maternity leave, from
taking paid sick leave, unless she elects to return to work and terminate her maternity leave.
907
(4) A condition limiting the time during which annual leave accrues to the statutory minimum
period of 14 weeks maternity leave and accordingly excluding any other period of maternity leave.
(5) A condition limiting the time in which pensionable service accrues during maternity leave to
when the woman is in receipt of contractual or statutory maternity pay and accordingly excluding
any period of unpaid maternity leave?

The Community legislation

20.
Article 119 of the Treaty provides that member states are required to ensure and to maintain the
application of the principle that men and women should receive equal pay for equal work.
21.
According to art 1 of Directive 75/117, the principle of equal pay laid down in art 119 of the Treaty is
intended to eliminate, for the same work or for work to which equal value is attributed, all discrimination
on grounds of sex with regard to all aspects and conditions of remuneration.
22.
According to art 1(1) thereof, Directive 76/207 is intended
to put into effect in the Member States the principle of equal treatment for men and women as
regards access to employment, including promotion, and to vocational training and as regards
working conditions.
23.
Article 2(1) of that directive provides:
For the purposes of the following provisions, the principle of equal treatment shall mean that
there shall be no discrimination whatsoever on grounds of sex either directly or indirectly by
reference in particular to marital or family status.
24.
However, according to art 2(3), Directive 76/207 is without prejudice to provisions concerning the
protection of women, particularly as regards pregnancy and maternity.
25.
Article 5(1) of that directive provides:
Application of the principle of equal treatment with regard to working conditions, including the
conditions governing dismissal, means that men and women shall be guaranteed the same
conditions without discrimination on grounds of sex.
26.
As regards Directive 92/85, art 8 of that directive, concerning maternity leave, provides:
(1) Member States shall take the necessary measures to ensure that workers within the
meaning of Article 2 are entitled to a continuous period of maternity leave of at least 14 weeks
allocated before and/or after confinement in accordance with national legislation and/or practice. (2)
The maternity leave stipulated in paragraph 1 must include compulsory maternity leave of at least
two weeks allocated before and/or after confinement in accordance with national legislation and/or
practice.
27.
As regards employment rights, art 11 of Directive 92/85 states:
In order to guarantee workers within the meaning of Article 2 the exercise of their health and
safety protection rights as recognized in this Article, it shall be provided that (2) in the case
referred to in Article 8, 908the following must be ensured: (a) the rights connected with the
employment contract of workers within the meaning of Article 2, other than those referred to in point
(b) below; (b) maintenance of a payment to, and/or entitlement to an adequate allowance for,
workers within the meaning of Article 2; (3) the allowance referred to in point 2(b) shall be deemed
adequate if it guarantees income at least equivalent to that which the worker concerned would
receive in the event of a break in her activities on grounds connected with her state of health,
subject to any ceiling laid down under national legislation; (4) Member States may make
entitlement to pay or the allowance referred to in points 1 and 2(b) conditional upon the worker
concerned fulfilling the conditions of eligibility for such benefits laid down under national legislation.
These conditions may under no circumstances provide for periods of previous employment in
excess of 12 months immediately prior to the presumed date of confinement.
 The first question
28.
By its first question, the national court essentially asks whether art 119 of the Treaty, as given specific
expression by Directive 75/117 and Directives 76/207 or 92/85, precludes a clause in an employment
contract which makes the payment, during the period of maternity leave referred to by art 8 of Directive
92/85, of pay higher than the statutory payments in respect of maternity leave conditional on the womans
undertaking to return to work after the birth of the child for at least one month, failing which she is required
to repay the difference between the amount of the pay she will have received during the period of
maternity leave, on the one hand, and the amount of those payments, on the other.
29.
As regards, first, Directive 92/85, the Commission submits that art 11(2)(b) and (3) requires the
payment to a worker on maternity leave of an amount at least equivalent to that which that woman would
receive under her employment contract if she were on sick leave. Where, as in this case, the employer
has undertaken to pay workers on sick leave their full salary, women on maternity leave should, in
accordance with the aforementioned provisions of the directive, receive an equivalent income. In those
circumstances, it would be inconsistent with art 11 of Directive 92/85 if female workers were required to
repay the difference between the full salary they received from their employer during their maternity leave
and the payments to which they are entitled during maternity leave under national legislation, in the event
that they did not return to work after childbirth.
30.
In that respect, it should be noted that it was in view of the risk that the provisions relating to maternity
leave would be ineffective if rights connected with the employment contract were not maintained, that the
Community legislature provided, in art 11(2)(b) of Directive 92/85, that maintenance of a payment to,
and/or entitlement to an adequate allowance for workers to whom the directive applies must be ensured
in the case of the maternity leave referred to in art 8.
31.
The concept of pay used in art 11 of that directive, like the definition in the second paragraph of art 119
of the Treaty, encompasses the consideration paid directly or indirectly by the employer during the
workers maternity leave in respect of her employment (see Gillespie v Northern Health and Social
Services Board Case C-342/93 [1996] All ER (EC) 284, [1996] ECR I-475 (para 12)). By 909 contrast, the
concept of allowance to which that provision also refers includes all income received by the worker during
her maternity leave which is not paid to her by her employer pursuant to the employment relationship.
32.
According to art 11(3) of Directive 92/85, the allowance
shall be deemed adequate if it guarantees income at least equivalent to that which the worker
concerned would receive in the event of a break in her activities on grounds connected with her
state of health, subject to any ceiling laid down under national legislation.
This is intended to ensure that, during her maternity leave, the worker receives an income at least
equivalent to the sickness allowance provided for by national social security legislation in the event of a
break in her activities on health grounds.
33.
Female workers must be guaranteed an income of that level during their maternity leave, irrespective
of whether, in accordance with art 11(2)(b) of Directive 92/85, it is paid in the form of an allowance, pay or
a combination of the two.
34.
Although the wording of art 11 refers only to the adequate nature of the allowance, the income
guaranteed to female workers during maternity leave must none the less also be adequate within the
meaning of art 11(3) of Directive 92/85 if it is paid in the form of pay or in conjunction with an allowance,
as the case may be.
35.
However, although art 11(2)(b) and (3) requires the female worker to receive, during the period of
maternity leave referred to in art 8, income at least equivalent to the sickness allowance provided for
under national social security legislation in the event of a break in her activities on health grounds, it is not
intended to guarantee her any higher income which the employer may have undertaken to pay her, under
the employment contract, should she be on sick leave.
36.
It follows that a clause in an employment contract according to which a worker who does not return to
work after childbirth is required to repay the difference between the pay received by her during her
maternity leave and the statutory payments to which she was entitled in respect of maternity leave is
compatible with art 11(2)(b) and (3) of Directive 92/85 in so far as the level of those payments is not lower
than the income which the worker concerned would receive, under the relevant national social security
legislation, in the event of a break in her activities on grounds connected with her state of health.
37.
Next, as regards art 119 of the Treaty, as given specific expression by Directive 75/117 and Directive
76/207, the applicants in the main proceedings submit that the requirement that a woman must repay the
contractual pay received by her during maternity leave, in so far as it exceeds SMP, if she does not return
to work after childbirth constitutes discrimination against a woman for reasons of pregnancy, and is
therefore contrary to the principle of equal pay. For other forms of paid leave, such as sick leave, workers
in general are entitled to the agreed salary without having to undertake to return to work at the end of their
leave.
38.
Since the consideration paid by an employer under legislation or an employment contract to a woman
on maternity leave is based on the employment relationship, it constitutes pay within the meaning of art
119 of the Treaty and art 1 of Directive 75/117 (see the judgment in Gillespie [1996] All 910 ER (EC) 284,
[1996] ECR I-475 (para 14)). It therefore cannot also fall within the scope of Directive 76/207.
39.
Furthermore, it is settled case law that discrimination involves the application of different rules to
comparable situations or the application of the same rule to different situations (see the judgments in
Gillespie [1996] All ER (EC) 284, [1996] ECR I-475 (para 16), and Finanzamt Kln-Altstadt v Schumacker
Case C-279/93 [1995] All ER (EC) 319, [1995] ECR I-225 (para 30)).
40.
As the Community legislature acknowledged when adopting Directive 92/85, pregnant workers and
workers who have recently given birth or who are breastfeeding are in an especially vulnerable situation
which makes it necessary for the right to maternity leave to be granted to them but which, particularly
during that leave, cannot be compared to that of a man or a woman on sick leave.
41.
The maternity leave granted to a worker is intended, first, to protect a womans biological condition
during and after pregnancy and, second, to protect the special relationship between a woman and her
child over the period which follows pregnancy and childbirth (see Hofmann v Barmer Ersatzkasse Case
184/83 [1984] ECR 3047 (para 25) and Caisse Nationale dAssurance Viellesse des Travailleurs Salaris
v Thibault Case C-136/95 [1998] All ER (EC) 385, [1998] ECR I-2011 (para 25)).
42.
A clause in an employment contract which makes the application of a more favourable set of rules
than that prescribed by national legislation conditional on the pregnant woman, unlike any worker on sick
leave, returning to work after childbirth, failing which she must repay the contractual maternity pay in so
far as it exceeds the level of the statutory payments in respect of that leave, therefore does not constitute
discrimination on grounds of sex for the purposes of art 119 of the Treaty and art 1 of Directive 75/117.
43.
However, the level of those payments must satisfy the requirements laid down in art 11(2)(b) and (3) of
Directive 92/85.
44.
In view of the foregoing, the answer to the first question must be that art 119 of the Treaty, art 1 of
Directive 75/117 and art 11 of Directive 92/85 do not preclude a clause in an employment contract which
makes the payment, during the period of maternity leave referred to by art 8 of Directive 92/85, of pay
higher than the statutory payments in respect of maternity leave conditional on the workers undertaking
to return to work after the birth of the child for at least one month, failing which she is required to repay
the difference between the amount of the pay she will have received during the period of maternity leave,
on the one hand, and the amount of those payments, on the other.

The second question

45.
By its second question, the national court essentially asks whether art 119 of the Treaty, as given
specific expression by Directive 75/117 and Directives 76/207 or 92/85, precludes a clause in an
employment contract from requiring an employee who has expressed her intention to commence her
maternity leave during the six weeks preceding the expected week of childbirth, and is on sick leave with
a pregnancy-related illness immediately before that date and gives birth during the period of sick leave, to
bring forward the date on which her paid maternity leave commences either to the beginning of the sixth
week preceding the expected week of childbirth or to the beginning of the period of sick leave, whichever
is the later.
911
46.
The applicants in the main proceedings submit that such a clause constitutes discrimination against
women in so far as, unlike any other worker who is sick, a female worker who is unfit for work is not able
to exercise her contractual right to unconditional paid sick leave if her illness is pregnancy-related and she
gives birth while on sick leave. The female worker is thus required to take paid maternity leave on less
favourable terms and, in particular, to repay a part of the salary received during that period if she does not
return to work after the birth of the child.
47.
It should be noted at the outset that, in so far as it concerns the determination of the beginning of the
period of maternity leave, the question falls within the scope of Directive 76/207, in particular art 5(1)
thereof on working conditions, and not art 119 of the Treaty or Directive 75/117.
48.
Next, the contested clause applies to the case of a pregnant employee who has expressed her wish to
commence her maternity leave during the six weeks preceding the expected week of childbirth.
49.
 In that respect, although art 8 of Directive 92/85 provides for a continuous period of maternity leave of
at least 14 weeks, including compulsory maternity leave of at least two weeks, it none the less leaves it
open to the member states to determine the date on which maternity leave is to commence.
50.
Furthermore, pursuant to art 2(3) of Directive 76/207, it is for every member state, within the limits laid
down in art 8 of Directive 92/85, to fix periods of maternity leave so as to enable female workers to be
absent during the period in which the disorders inherent in pregnancy and confinement occur (see the
judgment in Handels- og Kontorfunktionaerernes Foribund i Danmark v Dansk Arbejdsgiverforening Case
C-179/88 [1990] ECR I-3979 (para 15)).
51.
National legislation may therefore, as here, provide that the period of maternity leave commences with
the date notified by the person concerned to her employer as the date on which she intends to commence
her period of absence, or the first day after the beginning of the sixth week preceding the expected week
of childbirth during which the employee is wholly or partly absent because of pregnancy, should that day
fall on an earlier date.
52.
The clause to which the second question relates merely reflects the choice made in such national
legislation.
53.
Furthermore, for the reasons described at paras 42 and 43 above, the requirement that a female
worker on maternity leave undertake to repay the pay received under her contract during her maternity
leave, in so far as it exceeds the level of the payments provided for by national legislation during that
leave, if she fails to return to work after childbirth cannot constitute unfavourable treatment of that worker.
54.
The answer to the second question must therefore be that art 8 of Directive 92/85 and art 5(1) of
Directive 76/207 do not preclude a clause in an employment contract from requiring an employee who
has expressed her intention to commence her maternity leave during the six weeks preceding the
expected week of childbirth, and is on sick leave with a pregnancy-related illness immediately before that
date and gives birth during the period of sick leave, to bring forward the date on which her paid maternity
leave commences either to the beginning of the sixth week preceding the expected week of childbirth or
to the beginning of the period of sick leave, whichever is the later.
912

The third question

55.
By its third question, the national court asks whether art 119 of the Treaty, as given specific expression
by Directive 75/117 and Directives 76/207 and 92/85, precludes a clause in an employment contract from
prohibiting a woman from taking sick leave during the minimum period of 14 weeks maternity leave to
which a female worker is entitled pursuant to art 8 of Directive 92/85 or any supplementary period of
maternity leave granted to her by the employer, unless she elects to return to work and thus terminate her
maternity leave.
56.
First, as regards Directive 92/85, a distinction must be drawn between, on the one hand, the period of
maternity leave of at least 14 weeks referred to by art 8 of that directive and, on the other, any
supplementary leave which, as in the present case, the employer is prepared to offer pregnant workers,
and workers who have recently given birth or are breastfeeding.
57.
In so far as the clause at issue prohibits a woman from taking sick leave during the period of maternity
leave referred to by art 8 of Directive 92/85which, in the United Kingdom is, in principle, 14 weeks
unless she terminates that leave, it must be examined in the light of that provision.
58.
In that respect, although the member states are required, pursuant to art 8 of the aforesaid directive, to
take the necessary measures to ensure that workers are entitled to a period of maternity leave of at least
14 weeks, those workers may waive that right, with the exception of the two weeks compulsory maternity
leave provided for in art 8(2), which, in the United Kingdom, commence on the day on which the child is
born.
59.
Furthermore, art 8 of Directive 92/85 provides that the period of maternity leave provided for therein
must be at least 14 continuous weeks, allocated before and/or after confinement. It follows from the
purpose of that provision that the woman cannot interrupt or be required to interrupt her maternity leave
and return to work, and complete the remaining period of maternity leave later.
60.
In contrast, if a woman becomes ill during the period of maternity leave referred to by art 8 of Directive
92/85 and places herself under the sick leave arrangements, and that sick leave ends before the expiry of
the period of maternity leave, she cannot be deprived of the right to continued enjoyment, after that date,
of the maternity leave provided for by the aforementioned provision until the expiry of the minimum period
of 14 weeks, that period being calculated from the date on which the maternity leave commenced.
61.
 Any other interpretation would compromise the purpose of maternity leave, in so far as that leave is
intended to protect not only the womans biological condition but also the special relationship between a
woman and her child over the period which follows pregnancy and childbirth. The continuous period of
maternity leave of at least 14 weeks, allocated before and/or after confinement, is intended in particular to
provide the woman with the guarantee that she can look after her newborn baby in the weeks following
childbirth. Except in exceptional circumstances, she cannot therefore be deprived of that guarantee for
reasons of health.
62.
In so far as it prohibits a woman from taking sick leave during any leave granted by the employer in
addition to the period of maternity leave provided for by art 8 of Directive 92/85, unless she terminates
that leave, such a clause does not fall within the scope of that provision.
913
63.
The third question also seeks to ascertain whether the clause at issue constitutes discrimination as
regards the right to sick leave and, therefore, falls within the scope of Directive 76/207, in particular art
5(1) thereof, concerning conditions of employment. Article 119 of the Treaty and Directive 75/117 are
therefore not in point. In view of the foregoing, the third question need be examined only in so far as the
clause of the employment contract referred to therein applies to the supplementary period of maternity
leave granted by the employer to female workers.
64.
In that respect, the principle of non-discrimination laid down in art 5 of Directive 76/207 does not
require a woman to be able to exercise simultaneously both the right to supplementary maternity leave
granted to her by the employer and the right to sick leave.
65.
Consequently, in order for a woman on maternity leave to qualify for sick leave, she may be required to
terminate the period of supplementary maternity leave granted to her by the employer.
66.
The answer must therefore be that a clause in an employment contract which prohibits a woman from
taking sick leave during the minimum period of 14 weeks maternity leave to which a female worker is
entitled pursuant to art 8(1) of Directive 92/85, unless she elects to return to work and thus terminate her
maternity leave, is not compatible with Directive 92/85. By contrast, a clause in an employment contract
which prohibits a woman from taking sick leave during a period of supplementary maternity leave granted
to her by the employer, unless she elects to return to work and thus terminate her maternity leave, is
compatible with Directives 76/207 and 92/85.

The fourth question

67.
By its fourth question, the national court essentially seeks to ascertain whether art 119 of the Treaty, as
given specific expression by Directive 75/117 and Directives 76/207 or 92/85, precludes a clause in an
employment contract from limiting the period during which annual leave accrues to the minimum period of
14 weeks maternity leave to which female workers are entitled under art 8 of Directive 92/85 and from
providing that annual leave ceases to accrue during any period of supplementary maternity leave granted
to them by their employer.
68.
First, the accrual of annual leave constitutes a right connected with the employment contract of
workers for the purposes of art 11(2)(a) of Directive 92/85.
69.
It follows from that provision that such a right need only be ensured during the period of maternity
leave of at least 14 weeks to which workers are entitled under art 8 of Directive 92/85.
70.
Here, the duration of that period of leave is, in principle, fixed at 14 weeks in the United Kingdom.
71.
Consequently, the directive does not preclude a clause, such as that to which the question referred to
the court relates, according to which annual leave ceases to accrue during any period of supplementary
maternity leave granted by employers to pregnant workers or workers who have recently given birth or
who are breastfeeding.
72.
Second, the particular rules concerning the accrual of annual leave constitute an integral part of
working conditions within the meaning of art 5(1) of Directive 76/207 and cannot therefore also fall within
the scope of art 119 of the Treaty or Directive 75/117.
914
73.
In that respect, the applicants point out that, according to the EOCs Staff Handbook, if unpaid leave is
taken (sick leave, special leave or supplementary maternity leave), the annual leave entitlement is
reduced by a proportion of the amount of unpaid leave taken. However, since a substantially greater
proportion of women than men take periods of unpaid leave because they take supplementary maternity
leave, that rule which is ostensibly gender-neutral constitutes indirect discrimination against women,
contrary to art 5(1) of Directive 76/207.
74.
First, as is clear from the documents before the court, all employees of EOC who take unpaid leave
cease to accrue annual leave during that period. According to the EOC Staff Handbook, unpaid leave
includes both sick leave and special leave, which are available to any worker, as well as supplementary
maternity leave granted by EOC in addition to the 14 weeks maternity leave provided for by the
Employment Rights Act 1996.
75.
Such a clause therefore does not constitute direct discrimination since the accrual of annual leave
during the period of unpaid leave is interrupted for both men and for women who take unpaid leave. It is
therefore necessary to consider whether such a clause can constitute indirect discrimination.
76.
The court has consistently held that indirect discrimination arises where a national measure, albeit
formulated in neutral terms, works to the disadvantage of far more women than men (see esp Gerster v
Freistaat Bayern Case C-1/95 [1997] ECR I-5253 (para 30) and Kording v Senator fr Finanzen Case C-
100/95 [1997] ECR I-5289 (para 16)).
77.
In that respect, it should be noted that, as the national court points out, substantially more women than
men take periods of unpaid leave during their career because they take supplementary maternity leave,
so that, in practice, the clause at issue applies to a greater percentage of women than men.
78.
However, the fact that such a clause applies more frequently to women results from the exercise of the
right to unpaid maternity leave granted to them by their employers in addition to the period of protection
guaranteed by art 8 of Directive 92/85.
79.
Female workers who exercise that right subject to the condition that annual leave ceases to accrue
during the period of unpaid leave cannot be regarded as at a disadvantage compared to male workers.
The supplementary unpaid maternity leave constitutes a special advantage, over and above the
protection provided for by Directive 92/85 and is available only to women, so that the fact that annual
leave ceases to accrue during that period of leave cannot amount to less favourable treatment of women.
80.
The answer must therefore be that Directives 92/85 and 76/207 do not preclude a clause in an
employment contract from limiting the period during which annual leave accrues to the minimum period of
14 weeks maternity leave to which female workers are entitled under art 8 of Directive 92/85 and from
providing that annual leave ceases to accrue during any period of supplementary maternity leave granted
to them by their employer.

The fifth question

81.
It appears from the documents before the court that, by its fifth question, the national court essentially
seeks to ascertain whether art 119 of the Treaty, as given specific expression by Directive 75/117 and
Directives 92/85 or 76/207, precludes a clause in an employment contract from limiting, in the context of
an occupational scheme wholly financed by the employer, the accrual of 915 pension rights during
maternity leave to the period during which the woman receives the pay provided for by that employment
contract or national legislation.
82.
The accrual of pension rights in the context of an occupational scheme wholly financed by the
employer constitutes one of the rights connected with the employment contracts of the workers for the
purposes of art 11(2)(a) of Directive 92/85.
83.
As stated at para 69, above, in accordance with that provision, such rights must be ensured during the
period of maternity leave of at least 14 weeks to which female workers are entitled under art 8 of Directive
92/85.
84.
Although, in accordance with art 11(4) of Directive 92/85, it is open to member states to make
entitlement to pay or the adequate allowance referred to in art 11(2)(b) conditional upon the worker
concerned fulfilling the conditions of eligibility for such benefits laid down under national legislation, no
such possibility exists in respect of rights connected with the employment contract within the meaning of
art 11(2)(a).
85.
The accrual of pension rights under an occupational scheme during the period of maternity leave
referred to by art 8 of Directive 92/85 cannot therefore be made conditional upon the womans receiving
the pay provided for by her employment contract or SMP during that period.
86.
Since the clause to which the fifth question relates is contrary to Directive 92/85, it is not necessary to
interpret art 119 of the Treaty, as given specific expression by Directive 75/117 and Directive 76/207.
87.
The answer to the fifth question must therefore be that Directive 92/85 precludes a clause in an
employment contract from limiting, in the context of an occupational scheme wholly financed by the
employer, the accrual of pension rights during the period of maternity leave referred to by art 8 of that
directive to the period during which the woman receives the pay provided for by that contract or national
legislation.

Costs
88.
The costs incurred by the UK government, the Irish government and the Austrian government, and by
the European Commission, which have submitted observations to the Court of Justice, are not
recoverable. Since these proceedings are, for the parties to the main proceedings, a step in the
proceedings pending before the national court, the decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the questions referred to it by the Industrial
Tribunal, Manchester, by order of 15 October 1996, hereby rules: (1) Article 119 of the EC Treaty, art 1 of
Council Directive (EEC) 75/117 on the approximation of the laws of the member states relating to the
application of the principle of equal pay for men and women and art 11 of Council Directive (EEC) 92/85
on the introduction of measures to encourage improvements in the safety and health at work of pregnant
workers and workers who have recently given birth or are breastfeeding do not preclude a clause in an
employment contract which makes the payment, during the period of maternity leave referred to by art 8
of Directive 92/85, of pay higher than the statutory payments in respect of maternity leave conditional on
the workers undertaking to return to work after the birth of the child for at least one month, failing which
she is required to repay the difference between the amount of the 916 pay she will have received during
the period of maternity leave, on the one hand, and the amount of those payments, on the other. (2)
Article 8 of Directive 92/85 and art 5(1) of Council Directive (EEC) 76/207 on the implementation of the
principle of equal treatment for men and women as regards access to employment, vocational training
and promotion, and working conditions do not preclude a clause in an employment contract from requiring
an employee who has expressed her intention to commence her maternity leave during the six weeks
preceding the expected week of childbirth, and is on sick leave with a pregnancy-related illness
immediately before that date and gives birth during the period of sick leave, to bring forward the date on
which her paid maternity leave commences either to the beginning of the sixth week preceding the
expected week of childbirth or to the beginning of the period of sick leave, whichever is the later. (3) A
clause in an employment contract which prohibits a woman from taking sick leave during the minimum
period of 14 weeks maternity leave to which a female worker is entitled pursuant to art 8(1) of Directive
92/85, unless she elects to return to work and thus terminate her maternity leave, is not compatible with
Directive 92/85. By contrast, a clause in an employment contract which prohibits a woman from taking
sick leave during a period of supplementary maternity leave granted to her by the employer, unless she
elects to return to work and thus terminate her maternity leave, is compatible with Directives 76/207 and
92/85. (4) Directives 92/85 and 76/207 do not preclude a clause in an employment contract from limiting
the period during which annual leave accrues to the minimum period of 14 weeks maternity leave to
which female workers are entitled under art 8 of Directive 92/85 and from providing that annual leave
ceases to accrue during any period of supplementary maternity leave granted to them by their employer.
(5) Directive 92/85 precludes a clause in an employment contract from limiting, in the context of an
occupational scheme wholly financed by the employer, the accrual of pension rights during the period of
maternity leave referred to by art 8 of that directive to the period during which the woman receives the pay
provided for by that contract or national legislation.

917

[1998] All ER (EC) 918

Gestevisin Telecinco SA v European Commission (France

intervening)
(Case T-95/96)
EUROPEAN COMMUNITY: Other European Community
COURT OF FIRST INSTANCE OF THE EUROPEAN COMMUNITIES (THIRD CHAMBER, EXTENDED
COMPOSITION)
JUDGES TIILI (PRESIDENT), BRIT, LENAERTS, POTOCKI AND COOKE
10 MARCH, 15 SEPTEMBER 1998
European Community Admissibility Action for failure to act Private sector television company
complaining to European Commission that grants to public television companies contravening state aid
rules Commission failing to adopt decision concerning complaints Applicant bringing action for
declaration of failure to act Whether private undertaking having standing EC Treaty, art 175.

The applicant undertaking was one of three private commercial television companies in Spain. In March
1992, it lodged complaints with the European Commission seeking to have the grants which the public
television companies received from the Spanish state declared incompatible with the common market
within the meaning of art 921 of the EC Treaty. In February 1996 the Commission, which still had not ruled
on the complaints, informed the applicant that following completion of a study commissioned in December
1993 on public television in other countries, it had requested the Spanish authorities to provide a number
of further details and explanations necessary for investigating the complaints. No decision was adopted
subsequently and thereafter the applicant applied to the Court of First Instance of the European
Communities for a declaration under art 1752 of the Treaty that, by failing to adopt a decision in relation to
the applicants complaints and by failing to initiate the procedure provided for under art 93(2) 3, the
Commission had failed to fulfil its obligations under the Treaty. Before the Court of First Instance, the
Commission contended that only the person to whom an act was potentially addressed had the capacity
to bring an action under art 175, and that since any decision taken under an art 93(2) procedure would be
addressed to a member state, the action brought by the applicant was inadmissible. Furthermore, it
contended that such an interpretation did not deprive individuals of the right to legal protection, since they
could bring an action before the national courts in order to determine the illegality of state measures.
1
Article 92, so far as material, provides: (1) Save as otherwise provided in this Treaty, any aid granted by a Member
State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring
certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be
incompatible with the common market.
2
Article 175, so far as material, provides: Any natural or legal person may complain to the Court of Justice that an
institution of the Community has failed to address to that person any act other than a recommendation or an opinion.
3
Article 93, so far as material, provides: (2) If the Commission finds that aid granted by a State or through State
resources is not compatible with the common market having regard to Article 92, or that such aid is being misused, it
shall decide that the State concerned shall abolish or alter such aid

918

Held (1) On its proper construction, the third paragraph of art 175 entitled individuals to bring an action
for failure to act against a Community institution which they claimed had failed to adopt a measure where
that measure would have been of direct and individual concern to them. The persons directly and
individually concerned for the purposes of art 93(2) were those persons, undertakings or associations
whose interests might be affected by the grant of the aid in question, in particular competing undertakings
and trade associations. In the instant case, the applicants undisputed position as a party concerned
within the meaning of art 93(2) ensued from its status as one of the private television channels in
competition with the public television channels to which the contested grants were made and from the fact
that both complaints lodged by it prompted the preliminary investigation carried out by the Commission in
relation to those grants (see p 928 f and p 929 f h, post); T Port GmbH & Co KG v Bundesanstalt fr
Landwirtschaft und Ernhrung Case C-68/95 [1996] ECR I-6065 applied.
(2) Moreover, the possible existence of a remedy at domestic level, whereby the applicant could
challenge the grant of the disputed allowances to the public channels, could not affect the admissibility of
the applicants claim. Accordingly, the failure by the Commission to take any decision following its
initiation of the preliminary investigation procedure in relation to the grants made by the various Spanish
state bodies to the public television companies was of direct and individual concern to the applicant. It
followed that the applicants claim for a declaration that the Commission had failed to act was admissible
and would, in the circumstances, be granted (see p 929 j to p 930 c and p 933 c, post); Kahn
Scheepvaart BV v European Commission Case T-398/94 [1996] ECR II-477 applied.

Notes
For locus standi in relation to actions for failure to act, see 51 Halsburys Laws (4th edn) para 256. For a
relevant case see, 21(1) Digest (2nd reissue) 64, 143.
For the EC Treaty, art 175 (as amended by art G.54 of the Treaty on European Union), see 50
Halsburys Statutes (4th edn), Current Service 97.

Cases cited
Association of Sorbitol Producers within the EC (ASPEC) v European Commission Case T-435/93 [1995]
ECR II-1281.
Associazone Italiana Tecnico Economica del Cemento (AITEC) v European Commission Case T-277/94
[1996] ECR II-351.
Automec Srl v EC Commission Case T-24/90 [1992] ECR II-2223.
Buckl (Joseph) und Shne OHG v EC Commission Joined cases C-15/91 and C-108/91 [1992] ECR I-
6061.
Chambre Syndicale Nationale des Entreprises de Transport de Fonds et Valeurs (Sytraval) v European
Commission Case T-95/94 [1995] ECR II-2651.
Chevalley v EC Commission Case 15/70 [1970] ECR 975.
Compagnie franaise de lazote (Cofaz) SA v EC Commission Case 169/84 [1986] ECR 391.
Dumez v European Commission Case T-126/95 [1995] ECR II-2863.
European Commission v Chambre Syndicale Nationale des Entreprises de Transport de Fonds et Valeurs
(Sytraval) Case C-367/95 P (1998) Transcript, 2 April, ECJ.
European Parliament v EC Council Case 13/83 [1985] ECR 1513.
Fdration Franaise des Socits dAssurance (FFSA) v European Commission Case C-174/97 P
[1998] ECR I-1303.
919
Fdration Franaise des Socits dAssurance (FFSA) v European Commission Case T-106/95 [1997]
ECR II-229.
Fdration Nationale du Commerce Extrieur des Produits Alimentaires v France Case C-354/90 [1991]
ECR I-5505.
Forening af Jernskibs- og Maskinbyggerier i Danmark, Skibsvrftsforeningen v European Commission
Case T-266/94 [1996] ECR II-1399.
Germany v EC Commission Case 84/82 [1984] ECR I-1451.
Goldstein v European Commission Case T-286/97 (1998) Transcript, 15 July, CFI.
Iannelli & Volpi SpA v Meroni Case 74/76 [1977] ECR 557.
Intermills (SA) v EC Commission Case 323/82 [1984] ECR 3809.
Kahn Scheepvaart BV v European Commission Case T-398/94 [1996] ECR II-477.
Lorenz (Gebr) GmbH v Germany Case 120/73 [1973] ECR 1471.
Matra SA v EC Commission Case C-225/91 [1993] ECR I-3203.
Oliveira (Estabelecimentos Isidoro M) SA v European Commission Case T-73/95 [1997] ECR II-381.
Pantochim SA v European Commission Case T-107/96 [1998] ECR II-311.
Plaumann & Co v EEC Commission Case 25/62 [1963] ECR 95.
Port (T) GmbH & Co KG v Bundesanstalt fr Landwirtschaft und Ernhrung Case C-68/95 [1996] ECR I-
6065.
Rijn-Schelde-Verholme (RSV) Maschinefabrieken en Scheepswerven NV v EC Commission Case 223/85
[1987] ECR 4617.
Socit internationale diffusion et de ldition (SIDE) v European Commission Case T-49/93 [1995] ECR
II-2501.
Socit Nouvelle des Usines de PonlieueAciries du Temple v ECSC High Authority Joined cases
42/59 and 49/59 [1961] ECR 53.
Spain v EC Commission Case C-312/90 [1992] ECR I-4117.
Steinike and Weinlig (Firma) v Germany Case 78/76 [1977] ECR 595.
Stichting Certificatie Kraanverhuurbedrijf (SCK) v European Commission Joined cases T-213/95 and T-
18/96 [1997] ECR II-1739.
Syndicat Franais de lExpress International (SFEI) v La Poste Case C-39/94 [1996] All ER (EC) 685,
[1996] ECR I-3547, ECJ.
Vereniging van Exporteurs in Levende Varkens v EC Commission Joined cases T-481/93 and T-484/93
[1995] ECR II-2941.
William Cook plc v EC Commission Case C-198/91 [1993] ECR I-2487.
William v EC Commission Case 246/81 [1982] ECR 2277.

Application
By an application lodged on 17 June with the registry of the Court of First Instance of the European
Communities 1996 Gestevisin Telecinco SA, a company governed by Spanish law, sought, inter alia, a
declaration under art 175 of the EC Treaty that the European Commission had failed to fulfil its obligations
under the Treaty, first, by failing to adopt a decision in relation to the complaints made by the applicant
against the Kingdom of Spain for breach of art 92 of the Treaty and secondly by failing to initiate the
procedure provided for under art 93(2) of the Treaty. Gestevisin Telecinco SA was represented by S
Muoz Machado, of the Madrid Bar, with an address for service in Luxembourg at the Chambers of
Carlos Amo Quiones, 2 Rue Gabriel Lippmann. The Commission was represented initially by G Rozet,
Legal Adviser, and F Castillo de la Torre, of its Legal Service, and then by Mr Rozet and J Guerra
Fernndez, of its Legal Service, acting as agents, with an address for service in Luxembourg at the office
of Carlos Gmez de la Cruz, of its Legal Service, Wagner Centre, Kirchberg. The 920 French Republic
intervened in support of the Commission and was represented by C de Salins, Deputy Director of the
Department of Legal Affairs at the Ministry of Foreign Affairs, and G Mignot, Secretary for Foreign Affairs,
acting as agents, with an address for service in Luxembourg at the French Embassy, 8B Boulevard
Joseph II. The language of the case was Spanish. The facts are set out in the judgment of the court.

15 September 1998.

The COURT OF FIRST INSTANCE (Third Chamber, Extended Composition)

delivered the following judgment.

BACKGROUND TO THE DISPUTE

1.
There are ten television companies in Spain, of which three are in the private sector and seven are
public service broadcasters.
2.
The main source of finance for the private television companies is revenue generated by advertising.
The public television companies, on the other hand, are only partially funded by advertising. They are
managed either directly by the state through the intermediary of the public body RTVE, or governed by an
indirect system of management which has branches in various regional channels created for that purpose
in the various autonomous Spanish communities.
3.
From the start of their activities, all the public television companies have received, in varying
proportions, funds from the authorities within whose jurisdiction they fall. So, they receive funds from two
sources; advertising revenue and state grants.
4.
The applicant, Gestevisin Telecinco SA, a company incorporated under Spanish law in Madrid
(Spain), is one of the three private commercial companies. On 2 March 1992, it lodged with the European
Commission a complaint (the first complaint) seeking to have the grants which the regional television
companies receive from their respective autonomous communities declared incompatible with the
common market within the meaning of art 92 of the EC Treaty.
5.
By letter dated 30 April 1992 the Commission acknowledged receipt of that complaint and informed the
applicant that it had
decided to ask the Spanish authorities for specific information in order to determine whether
or not the practices complained of were compatible with the Community provisions relating to state
aid.
A request for information in that form was sent to the Spanish authorities on the same day.
6.
On 25 November 1992, the applicant sent a letter to the Commission with a view to obtaining
information on progress in relation to its complaint. In a letter dated 3 December 1992, the Commission
informed it that, by letter dated 28 October 1992, it had reminded the Spanish authorities of their duty to
reply to the request for information which had been sent to them.
7.
On 12 November 1993, the applicant lodged another complaint seeking to have the subsidies granted
by the central Spanish state to the public body RTVE declared incompatible with the common market
under art 92 of the Treaty (the second complaint).
8.
On 24 November 1993, the applicant sent a letter to Mr Van Miert, the member of the Commission with
responsibility for competition matters, informing him of the existence of the two above-mentioned
complaints, of the fact that the aid being challenged in those complaints had not been notified, and 921 of
the irreversible consequences of the Commissions slowness in dealing with those complaints.
9.
In December 1993, the Commission instructed a firm of outside consultants to carry out a study of the
funding of public television companies in the Community as a whole.
10.
In February 1994, it responded to a telephone request for information from the applicant by saying that
it had decided to await completion of that study before continuing its investigation into the complaints
concerned and so before taking any decision to initiate a procedure under art 93(2) of the Treaty.
11.
On 12 May 1995, it divulged, in response to a further telephone request for information, that the report
from the outside firm of consultants, which had been corrected following various delays in its drafting, was
to be sent to it before the end of the month. It received the final report in question during October 1995 at
the latest.
12.
However, by the beginning of February 1996, it had still not ruled on the applicants complaints.
Consequently, in a registered letter dated 6 February 1996, received on 8 February 1996, the applicant
formally called upon the Commission to rule on the two complaints in accordance with art 175 of the
Treaty and to initiate the procedure under art 93(2) of the Treaty.
13.
In a letter dated 20 February 1996, the Commission replied as follows:
Having considered your complaint in the light of Articles 92 et seq of the Treaty and following
completion of a study commissioned in December 1993 on the funding of public television in other
Member States, the Directorate-General for Competition, by letters dated 18 October 1995 and 14
February 1996, requested the Spanish authorities to provide a number of further details and
explanations necessary for investigating this case.
14.
After that letter, the Commission did not adopt a decision on the two complaints filed by the applicant.

PROCEDURE
15.
The applicant brought this action by an application lodged with the registry of the Court of First
Instance of the European Communities on 17 June 1996.
16.
By application lodged with the registry of the Court of First Instance on 8 November 1996, the French
Republic applied to intervene in support of the defendant. That application was granted by an order of the
President of the Third Chamber (Extended Composition) dated 4 February 1997.
17.
Upon hearing the report of the Judge Rapporteur, the Court of First Instance (Third Chamber,
Extended Composition) decided to open the oral procedure without preparatory inquiry. However, as a
measure of organisation of procedure, provided for in art 64 of the Rules of Procedure, the parties were
requested to respond to certain questions at the hearing.
18.
The parties made their submissions and replied to the questions posed by the Court of First Instance
at the hearing on 10 March 1998.

FORMS OF ORDER SOUGHT

19.
The applicant requests that the Court of First Instance should: declare that the Commission failed to
fulfil its obligations under the Treaty by not adopting a decision on the two complaints lodged by it and by
not initiating the procedure laid down in art 93(2) of the Treaty; alternatively, annul the Commissions 922
decision contained in its letter of 20 February 1996; order the defendant to pay the costs; order the
intervener to pay its own costs and the costs incurred by the applicant as a result of its intervention.
20.
The Commission requests that the Court of First Instance should: declare the claim for a declaration of
failure to act inadmissible, alternatively, dismiss it as unfounded; declare the claim for annulment
inadmissible; order the applicant to pay the costs.
21.
The French Republic supports the form of order sought by the Commission.

FAILURE TO ACT

Arguments of the parties

Admissibility
22.
The Commission states first of all that the decision it will adopt at the end of the administrative
procedure, pursuant to arts 92ff of the Treaty, will be sent to the Kingdom of Spain. The procedure for
supervising state aid is based on dialogue between the Commission and the member state concerned,
unlike the procedure for applying arts 85 and 86, which follows different rules under which a complainant
has a decisive role (see the judgment in Associazone Italiana Tecnico Economica del Cemento (AITEC) v
European Commission Case T-277/94 [1996] ECR II-351 (para 71)). Since the complainant does not
have any status in this context, it is inconceivable that a decision should be addressed to it directly (see
the opinion of Advocate General Tesauro in William Cook plc v EC Commission Case C-198/91 [1993]
ECR I-2487).
23.
Furthermore, the provisions of the third paragraph of art 175 of the Treaty cannot be interpreted so
broadly as to allow interested third parties the possibility of bringing an action. In this regard, the
defendant observes that the capacity to bring an action under art 175 of the Treaty is more limited than
the capacity to bring an action under art 173 of the Treaty. Only the person to whom an act is potentially
addressed has the capacity to bring an action under art 175 and that is not the position here (see the
judgments in William v EC Commission Case 246/81 [1982] ECR 2277 (para 16) and AITEC [1996] ECR
II-351 (para 62)).
24.
Secondly, the Commission considers that the fact of this action being inadmissible does not
necessarily mean that the applicant is deprived of the right to legal protection. It observes that it does not
have exclusive jurisdiction to adjudge a state measure to be state aid. National courts can also rule on
this question in order to determine the consequences of the illegality of measures in question under
national law (see the judgments in Firma Steinike and Weinlig v Germany Case 78/76 [1977] ECR 595
(para 14), Fdration Nationale du Commerce Extrieur des Produits Alimentaires v France Case C-
354/90 [1991] ECR I-5505 and Syndicat Franais de lExpress International (SFEI) v La Poste Case C-
39/94 [1996] All ER (EC) 685, [1996] ECR I-3547 (paras 3153)). It also challenges the applicants claim
that it has no remedies under Spanish law.
25.
Finally, the Commission states that the legal protection afforded by the Court of First Instance cannot
in any event serve to cure the deficiencies in legal protection at national level (see the opinion of
Advocate General Gulmann in Joseph Buckl und Shne OHG v EC Commission Joined cases C-15/91
and C-108/91 [1992] ECR I-6061 (para 27) and the judgment in Kahn Scheepvaart BV v European
Commission Case T-398/94 [1996] ECR II-477 (para 50)).
923
26.
The applicant observes that, more than four years after the lodging of the first complaint and more
than two and a half years after the second, the Commissions failure to define its position in relation to the
two complaints and to initiate a procedure under art 93(2) of the Treaty persists.
27.
It points out that, in a letter dated 6 February 1996, received on 8 February 1996, it formally called
upon the Commission to take action under the second paragraph of art 175 of the Treaty. It considers
that, in view of the excessive length of time which elapsed from the time when the two complaints were
lodged, the Commission wrongfully failed to act and was under a duty to define its position within two
months. That period expired without the Commission having defined its position.
28.
In its letter of 20 February 1996, the Commission did not adopt a position. On the contrary, it avoided
taking any position on the basis that it had requested supplementary information from the Spanish
government and that the complaints were still being examined. The court has ruled in this regard that
where an institution is called upon to define its position, a letter from that institution stating that the
questions raised are in the process of being considered does not amount to the definition of a position
releasing the institution from its wrongful failure to act (see the judgment in Socit Nouvelle des Usines
de PonlieueAciries du Temple v ECSC High Authority Joined cases 42/59 and 49/59 [1961] ECR 53).
29.
The applicant further observes that the Commission seeks to justify its inaction by the unacceptable
argument that the preliminary investigation into the state measures to which the complaints relate is not
yet complete. Such a method of proceeding infringes the fundamental right to effective legal protection.
30.
The applicant also observes that the Commission was under a duty in this case to initiate inter partes
proceedings under art 93(2) of the Treaty and then to rule on the compatibility of the aid. Such steps and
the resulting failure to adopt any such decisions affected it directly and individually in its capacity as
complainant and as a competitor of the companies benefiting from the aid (see the judgments in
Compagnie franaise de lazote (Cofaz) SA v EC Commission Case 169/84 [1986] ECR 391, Socit
internationale diffusion et de ldition (SIDE) v European Commission Case T-49/93 [1995] ECR II-2501
and Chambre Syndicale Nationale des Entreprises de Transport de Fonds et Valeurs (Sytraval) v
European Commission Case T-95/94 [1995] ECR II-2651 (Sytraval No 1)). The coherence of the
Community system of legal protection requires that its locus standi in this case should also be recognised.
31.
The applicant also observes that the conditions for admissibility under art 175 of the Treaty are
comparable to those imposed by art 173 of the Treaty, as the court stated in its judgment in Chevalley v
EC Commission Case 15/70 [1970] ECR 975.
32.
It also takes the view that there is no possibility of bringing an action before the national court in this
case since the aid of which it complains is granted pursuant to budgetary laws against which an individual
cannot bring an action under Spanish law. Furthermore, the fact that the beneficiaries of the aid are public
companies means that the instruments implementing those laws are internal and unpublished and also
incapable of challenge. Even if that were not the case, no national court would be bold enough to hold
that the grants made to the public television companies amounted to state aid, knowing that the matter
had been before the Commission for four years without it having initiated an inter partes procedure under
art 93(2) of the Treaty. Finally, because of the 924 attitude of the Commission in this case, no national
court could require the grants concerned to be repaid if they were found to be incompatible (see the
judgment in Rijn-Schelde-Verholme (RSV) Maschinefabrieken en Scheepswerven NV v EC Commission
Case 223/85 [1987] ECR 4617).
33.
The French Republic, as intervener, referring to the operative part of the judgment in SFEI [1996] All
ER (EC) 685, [1996] ECR I-3547, challenges the applicants argument that no national court would be
minded to classify a measure which has been the subject of investigation by the Commission for several
years as state aid. Under the operative part of that judgment, a national court may rule on such a question
even if it is pending before the Commission at the same time. Moreover, the national court may request
clarification from the Commission or refer a question to the Court of Justice for a preliminary ruling under
art 177 of the Treaty.
Substance
34.
The applicant emphasises that it is settled case law that the procedure under art 93(2) of the Treaty is
indispensable where the Commission experiences serious difficulties in assessing whether aid is
compatible with the common market. The Commission cannot confine itself to the preliminary stage
provided for by art 93(3) so as to take a favourable decision in relation to aid unless it is in a position to
reach the firm view, following a preliminary investigation, that the aid is compatible with the Treaty (see
the judgments in William Cook plc [1993] ECR I-2487, Germany v EC Commission Case 84/82 [1984]
ECR I-1451 and SIDE [1995] ECR II-2501).
35.
In this case, the length of time which elapsed from the time when the complaints were lodged in itself
shows that the Commission is having serious difficulties in assessing the compatibility of the aid
concerned with the common market. The fact that it requested an external report on the funding methods
for public television companies only confirms that hypothesis. Even once that report was produced, the
Commission continued to experience difficulties in assessing the aid concerned given that, several
months later, it had still not adopted a position in relation to the facts complained of and was still
requesting supplementary information from the Spanish authorities.
36.
In its judgment in Gebr Lorenz GmbH v Germany Case 120/73 [1973] ECR 1471 the Court of Justice
furthermore recognised that the Commission has a reasonable period of two months to make an initial
assessment of any aid notified to it. The Commission also has a duty to carry out a preliminary
investigation within a reasonable period where a member state has not only failed to notify the aid but
also implemented it in breach of its Community obligations.
37.
In adopting the attitude which it has assumed in this case, the Commission is also failing to observe its
procedural rights under the Treaty in the context of the procedure under art 93(2) of the Treaty. The
applicants rights can only be respected if it is able to challenge decisions made by the Commission
without initiating the procedure under art 93(2) (see the judgments in William Cook plc [1993] ECR I-2487
and Matra SA v EC Commission Case C-225/91 [1993] ECR I-3203). Those procedural rights are also
meaningless if the Commission is allowed to prolong its preliminary investigation into state measures
indefinitely.
38.
The applicant further disputes that the obligation to initiate an administrative procedure under art 93(2)
is subject to a prior finding of aid under art 92(1) of the Treaty. The Commissions administrative practice
shows that it has in the past initiated such procedures where it was in doubt as to whether the 925 state
measures in question could be regarded as aid (see the judgment in Sytraval No 1 [1995] ECR II-2651
(para 79)). In any event, in its judgment in Fdration Franaise des Socits dAssurance (FFSA) v
European Commission Case T-106/95 [1997] ECR II-229 (FFSA No 1), the Court of First Instance stated
that granting public funds to a company constituted state aid even if art 92 might subsequently prove to
be inapplicable by virtue of art 90(2) of the Treaty.
39.
Finally, the applicant considers that account must be taken of the fact that, since no procedure has
been initiated under art 93(2), the Spanish authorities are still disbursing the contested grants to the
Spanish public television companies even though those grants were challenged several years ago. The
applicant submits that, in these circumstances, the Commission was under a duty to act which it has
failed to fulfil contrary to the Treaty.
40.
The Commission argues that, whilst it is true that it has not made any decision as to the existence of
state aid or the initiation of the procedure under art 93(2) of the Treaty, it has none the less taken a
number of actions to enable it to analyse all facets of a particularly complex problem, common as it is to a
large number of member states.
41.
It points out that it entered into correspondence with the Spanish authorities between 30 April 1992
and 8 February 1993 and that it subsequently commissioned a study on the exploitation and operation of
public television channels in the Community in December 1993. Following receipt of that study, it again
entered into correspondence with the Spanish authorities between 18 October 1995 and 5 July 1996.
During the period when the study was being drafted, it temporarily refrained from taking other initiatives
which might overlap with the study.
42.
In these circumstances, the procedure relating to the measures concerned cannot be considered as
having been suspended. In fact, most of the period of two and a half years which passed between the
time when the second complaint was lodged and the time when the applicant called upon the
Commission to act was spent on producing the external study referred to above.
43.
The Commission points out that neither the Treaty nor secondary legislation provides for a time limit
within which it is obliged to react to a complaint relating to non-notified state aid.
44.
In this case, account must also be taken of the complexity, both legal and political, of the matter in
question. The manner in which this case was dealt with called for a particularly cautious approach
because of the recent opening-up of televisual activity to competition. The first complaint was the first
ever of its kind and related to seven different regional grants. The complaints lodged by the applicant
furthermore raised delicate problems relating to effects on intra-Community trade, offsetting the cost of
public service obligations and classification in terms of aid, particularly because of the lack of accounting
transparency prevailing at times in the public companies in question.
45.
The time taken in dealing with this case cannot therefore be considered as constituting a wrongful
failure to act, contrary to the rules of the Treaty and in particular to the obligation to initiate the procedure
under art 93(2) of the Treaty.
46.
The Commission also draws attention to the serious repercussions which a decision to initiate the
procedure under art 93(2) of the Treaty would have on public television companies throughout the
Community. Indeed, in such a case, the grant of such aid would have to be suspended, a step which
would be contrary to the principle of sound administration (see the judgment in Spain v EC Commission
Case C-312/90 [1992] ECR I-4117).
926
47.
Finally, it states that it must first of all rule on the question whether the contested grants can be
classified as aid within the meaning of art 92(1) of the Treaty before being able to rule on their
compatibility with the common market. In that connection, it disputes that it has developed a practice
whereby it initiates the procedure under art 93(2) in order to determine whether the state measures can
be classified as aid within the meaning of art 92(1) of the Treaty.
48.
It concludes from those considerations that it was not able to define its position or to take the decisions
requested by the applicant at the time when it was formally called upon to do so. In this regard, it refers to
the opinion of Advocate General Edward in Automec Srl v EC Commission Case T-24/90 [1992] ECR II-
2223.

Findings of the court

Preliminary observations
49.
Article 93 of the Treaty provides for a special procedure of constant review and supervision of state aid
by the Commission. In relation to new aid which member states may intend to institute, there is a
procedure without which no aid can be considered properly granted and the Commission must be
informed of any plans to grant or alter aid prior to such plans being put into effect.
50.
The Commission then proceeds to carry out a preliminary investigation into the planned aid. If, at the
end of that investigation, it appears that a plan is not compatible with the common market, it must initiate
the procedure provided for in the first sub-paragraph of art 93(2) of the Treaty forthwith.
51.
In the context of that procedure, a distinction must therefore be drawn between, on the one hand, the
preliminary stage of investigating aid, instituted under art 93(3) of the Treaty, whose aim is solely to
enable the Commission to form an initial view on the compatibility, in part or in whole, of the aid in
question and, on the other hand, the analysis stage referred to in art 93(2) of the Treaty, which is intended
to give the Commission full information on all the details of the case (see the judgments in William Cook
plc [1993] ECR I-2487 (para 22) and Matra SA [1993] ECR I-3203 (para 16)).
52.
The procedure under art 93(2) is indispensable if the Commission experiences serious difficulties in
assessing whether aid is compatible with the common market. The Commission cannot therefore limit
itself to the preliminary phase under art 93(3) and take a favourable decision on a state measure which
has not been notified unless it is in a position to reach the firm view, following an initial investigation, that
the measure cannot be classified as aid within the meaning of art 92(1) or that the measure, whilst
constituting aid, is compatible with the common market. On the other hand, if the initial analysis has
resulted in the Commission taking the contrary view or has not even enabled all the difficulties raised by
the assessment of the measure in question to be overcome, the institution has a duty to gather all
necessary views and to that end to initiate the procedure under art 93(2) (see, on this point, the
judgments in Germany v EC Commission Case 84/82 [1984] ECR I-1451 (para 13), William Cook plc
[1993] ECR I-2487 (para 29), Matra SA [1993] ECR I-3203 (para 33) and European Commission v
Chambre Syndicale Nationale des Entreprises de Transport de Fonds et Valeurs (Sytraval) Case C-
367/95 P (1998) ECJ Transcript, 2 April (para 39) (Sytraval No 2)).
53.
Where interested third parties submit complaints to the Commission relating to state measures which
have not been notified under art 93(3), the 927 Commission is bound, in the context of the preliminary
stage referred to above, to conduct a diligent and impartial examination of the complaints in the interests
of sound administration of the fundamental rules of the Treaty relating to state aid, which may make it
necessary for it to examine matters not expressly raised by the complainants (see Sytraval No 2 (1998)
Transcript, 2 April (para 62)).
54.
Finally, it must be remembered that the Commission has exclusive jurisdiction to determine whether
aid is incompatible with the common market (see the judgments in Firma Steinike and Weinlig [1977]
ECR 595 (paras 910) and Fdration Nationale du Commerce Extrieur des Produits Alimentaires v
France Case C-354/90 [1991] ECR I-5505 (para 14)).
55.
It follows from all those rules that, at the end of the preliminary stage of the investigation into a state
measure, the Commission has a duty to adopt one of the following three decisions vis--vis the member
state concerned: it may decide that the state measure at issue does not constitute aid within the
meaning of art 92(1) of the Treaty; or it may decide that the measure, although constituting aid within the
meaning of art 92(1), is compatible with the common market under art 92(2) or (3); or it may decide to
initiate the procedure under art 93(2).
56.
As the law so stands, it is appropriate first of all to consider whether the claim for a declaration that the
Commission has failed to act is admissible and then, if appropriate, whether it is well founded.
Admissibility
57.
Under the third paragraph of art 175 of the Treaty, any natural or legal person may complain to the
Court of Justice that an institution of the Community has failed to address to that person any act other
than a recommendation or an opinion.
58.
 In its judgment in T Port GmbH & Co KG v Bundesanstalt fr Landwirtschaft und Ernhrung Case C-
68/95 [1996] ECR I-6065 (para 59), the court stated that, just as the fourth paragraph of art 173 allows
individuals to bring an action for annulment against a measure of an institution not addressed to them
provided that the measure is of direct and individual concern to them, the third paragraph of art 175 must
be interpreted as also entitling them to bring an action for failure to act against an institution which they
claim has failed to adopt a measure which would have concerned them in the same way.
59.
The Commission is therefore wrong to consider that the claim for a declaration of failure to act is
inadmissible on the sole ground that the applicant is not the potential addressee of any measures it might
adopt in this case (see para 55, above).
60.
In this case, it is appropriate to examine to what extent the measures which the Commission allegedly
failed to adopt can be considered to be of direct and individual concern to the applicant.
61.
In that connection, it follows from the judgment of the Court of First Instance in Association of Sorbitol
Producers within the EC (ASPEC) v European Commission Case T-435/93 [1995] ECR II-1281 (para 60),
that an undertaking must be considered to be directly affected by a decision of the Commission relating to
state aid where there is no doubt that the national authorities intend to implement their plan to grant aid.
In the present case, it is common ground that the various grants at issue have already been granted by
the Spanish authorities concerned and continue to be granted. In those circumstances, it must be
considered as established that the applicant is directly affected.
928
62.
As to whether it is individually affected, it is settled case law that natural or legal persons are
individually concerned by a decision where that decision affects them by reason of the attributes peculiar
to them or by reason of factual circumstances differentiating them from all other persons (see the
judgments in Plaumann & Co v EEC Commission Case 25/62 [1963] ECR 95 at 197, 226, Vereniging van
Exporteurs in Levende Varkens v EC Commission Joined cases T-481/93 and T-484/93 [1995] ECR II-
2941 (para 51) and Forening af Jernskibs- og Maskinbyggerier i Danmark, Skibsvrftsforeningen v
European Commission Case T-266/94 [1996] ECR II-1399 (para 44)).
63.
It is therefore appropriate to consider in this case whether the applicant would be individually affected
by a decision which the Commission might adopt in relation to the member state concerned at the end of
the preliminary stage of investigation, to the effect that the state measure at issue does not constitute aid,
or to the effect that it does constitute aid but is compatible with the common market, or to the effect that it
necessitates the procedure under art 93(2) of the Treaty to be initiated.
64.
It is settled case law that where, without initiating the procedure under art 93(2), the Commission finds,
on the basis of art 93(3), that a state measure does not constitute aid, or that such a measure, although
constituting aid, is compatible with the common market, the persons concerned, beneficiaries of the
procedural guarantees laid down in art 93(2), may secure compliance therewith only if they are able to
challenge such a Commission decision before the Community judicature (see, most recently, the
judgment in Sytraval No 2 (1998) ECJ Transcript, 2 April (para 37), and, previously, the judgments in
William Cook plc [1993] ECR I-2487 (para 23) and Matra SA v EC Commission Case C-225/91 [1993]
ECR I-3203 (para 17)). The same would apply, in this case, in the event that the Commission took the
view that the grants made to the Spanish public television companies amounted to aid but that they did
not fall within the prohibition laid down in art 92 of the Treaty by virtue of art 90(2) thereof (see the
judgment in FFSA No 1 [1997] ECR II-229 (paras 172, 178) confirmed on appeal by order in Fdration
Franaise des Socits dAssurance (FFSA) v European Commission Case C-174/97 P [1998] ECR I-
1303 (FFSA No 2)).
65.
The persons concerned for the purposes of art 93(2) of the Treaty, who are thus to be considered as
directly and individually concerned, are those persons, undertakings or associations whose interests
might be affected by the grant of the aid, in particular competing undertakings and trade associations (see
the judgment in SA Intermills v EC Commission Case 323/82 [1984] ECR 3809 (para 16)).
66.
In the present case, the Commission has not disputed that the applicant is a party concerned within
the meaning of art 93(2), this position ensuing from its status as manager of one of the three private
television channels in competition with the public television channels to which the contested grants were
made and from the fact that both complaints lodged by it prompted the preliminary investigation carried
out by the Commission in relation to those grants.
67.
Furthermore, the applicant properly brought an action before the Community judicature which has sole
jurisdiction, to the exclusion of any national court, to determine whether the Commission failed, in breach
of the Treaty, to initiate the procedure under art 93(2) of the Treaty, which is a necessary precondition for
the adoption of a final decision which would be of direct and individual concern to the applicant, such as a
decision declaring 929 compatible with the common market, an aid whose classification has until that
point raised serious difficulties.
68.
 In that connection, the possible existence of a remedy at domestic level, whereby the applicant could
challenge the grant of the disputed allowances to the public channels, cannot affect the admissibility of
the applicants claim for a declaration of failure to act (see the judgment in Kahn Scheepvaart BV [1996]
ECR II-477 (para 50)).
69.
Accordingly, the failure by the Commission to take any decision following its initiation of the preliminary
investigation procedure in relation to the grants made by the various Spanish state bodies to the public
television companies must be considered to be of direct and individual concern to the applicant.
70.
It follows that the claim for a declaration that the Commission failed to act is admissible.
Substance
71.
In order to rule on the substance of the claim for a declaration that the Commission has failed to act, it
is necessary to determine whether, at the time when the Commission was formally called upon to define
its position within the meaning of art 175 of the Treaty, it was under a duty to act (see the orders in
Dumez v European Commission Case T-126/95 [1995] ECR II-2863 (para 44) and Goldstein v European
Commission Case T-286/97 (1998) CFI Transcript, 15 July (para 24)).
72.
Since it has exclusive jurisdiction to assess the compatibility of state aid with the common market, the
Commission must, in the interests of sound administration and the fundamental rules of the Treaty
relating to state aid, conduct a diligent and impartial examination of a complaint alleging aid to be
incompatible with the common market (see the judgment in Sytraval No 2 (1998) ECJ Transcript, 2 April
(para 62)).
73.
As regards the period within which the Commission must rule on such a complaint, it is appropriate to
note that, in relation to art 85 of the Treaty, the Court of First Instance has already ruled that the
Commission cannot postpone indefinitely defining its position in relation to an application for clearance
under art 85(3), an area in which it has exclusive jurisdiction (see the judgment in Stichting Certificatie
Kraanverhuurbedrijf (SCK) v European Commission Joined cases T-213/95 and T-18/96 [1997] ECR II-
1739 (para 55)). In that case, the Court of First Instance held it to be a general principle of Community law
that the Commission must act within a reasonable time in adopting decisions following administrative
procedures relating to competition policy (see [1997] ECR II-1739 (para 56) and the cases cited therein).
74.
It follows that the Commission cannot prolong indefinitely its preliminary investigation into state
measures in relation to which there has been a complaint under art 92(1) of the Treaty where it has, as in
this case, agreed to initiate such an investigation.
75.
Whether or not the duration of an administrative procedure is reasonable must be determined in
relation to the particular circumstances of each case and, especially, its context, the various procedural
stages to be followed by the Commission, the complexity of the case and its importance for the various
parties involved (see the judgments in Estabelecimentos Isidoro M Oliveira SA v European Commission
Case T-73/95 [1997] ECR II-381 (para 45) and SCK [1997] ECR II-1739 (para 57)).
930
76.
In this case, it is appropriate first of all to examine whether the Commission should, as the applicant
claims, have undertaken a preliminary investigation of the disputed grants made to the public television
companies within a reasonable time of two months, such as the period specified in the judgment in Gebr
Lorenz GmbH [1973] ECR 1471 (para 4).
77.
In referring to a period of two months, that judgment was based on the need to take into account the
legitimate interests of the member state concerned in quickly ascertaining for certain the position as to the
legality of measures which have been notified to the Commission.
78.
That consideration cannot be entertained where the member state concerned has implemented aid
without having notified the Commission beforehand. If the state had doubts as to whether the aid planned
was state aid, it would be at liberty to safeguard its interests by informing the Commission of the planned
aid which would place the Commission under an obligation to define its position within a period of two
months (see the judgment in SFEI [1996] All ER (EC) 685, [1996] ECR I-3547 (para 48)).
79.
So, the two-month period referred to in Gebr Lorenz GmbH cannot apply as it stands in a case such
as this, where the aid in dispute has not been notified to the Commission.
80.
Next, it is appropriate to note that the applicants first complaint was lodged on 2 March 1992 and the
second complaint on 12 November 1993. It follows that, at the time when the Commission was formally
called upon to act pursuant to art 175 of the Treaty, that is to say on 8 February 1996, the date when the
applicants letter of 6 February 1996 inviting the Commission to act was received, the Commissions
preliminary investigation had so far taken 47 months for the first complaint and 26 months for the second.
81.
Those periods are so long that they should have been sufficient to enable the Commission to close the
preliminary stage of investigation into the aid in question and thus be in a position to adopt a decision
thereon (see para 55, above), unless it could show exceptional circumstances justifying such periods.
82.
In that connection, the Commission points out that the first complaint was the first of its kind it had ever
received, that member states may legitimately pursue non-commercial aims in the televisual sector and
that delicate problems were raised in relation to the effect on intra-Community trade and the offsetting of
costs of performing public service obligations under art 90(2) of the Treaty. At the hearing, it referred to
the existence of the Protocol on the System of Public Broadcasting in the Member States annexed to the
EC Treaty by the Treaty of Amsterdam (2 October 1997; OJ 1997 C340 p 109).
83.
However, it is apparent from the submissions and pleadings of the parties that the only real difficulty
facing the Commission in this case relates to the extent to which the disputed grants made to the Spanish
public television companies are intended to compensate them in respect of specific public service
obligations imposed upon them by national legislation. When assessing that issue, the Protocol cited
above cannot be taken into consideration as it was adopted almost 19 months after the applicant called
upon the Commission to act, which was even before the opening, on 29 March 1996, of the inter-
governmental conference which led to the Treaty of Amsterdam being concluded.
84.
The Commission furthermore attempts to justify the length of the delays in question by reference to the
steps it took after the applicant filed its complaints.
85.
On this point, it is appropriate to note that before the applicant called upon the Commission to act, it
twice, on 30 April 1992 and 18 October 1995, formally 931 requested information from the Spanish
authorities in relation to the grants in question. In December 1993, it also commissioned a firm of
consultants to produce an in-depth study on the funding of public television companies in the Community
as a whole.
86.
However, those steps in no way justify the Commission having prolonged to the extent that it did its
preliminary investigation into the aid in question, thereby considerably exceeding the period of
deliberation reasonably necessary to assess the aid under art 90(2) of the Treaty. Accordingly, and even if
it is accepted that the above Protocol annexed to the EC Treaty by the Treaty of Amsterdam shows the
politically sensitive nature of the subject matter in question for the member states, the Commission
should, at the time when it was called upon to act, have been in a position to adopt a decision declaring
that the grants in question did not constitute aid, or that they did constitute aid but were compatible with
the common market or that serious difficulties obliged it to initiate the procedure under art 93(2), thus
allowing all parties concerned, and in particular the member states, to submit their observations.
Furthermore, it could equally have adopted within the periods concerned, a hybrid decision combining,
according to the circumstances, one or more of the decisions of principle set out above in respect of
different aspects of the state measures in question (see, on this point, the judgments in Iannelli & Volpi
SpA v Meroni Case 74/76 [1977] ECR 557 (paras 1417) and Pantochim SA v European Commission
Case T-107/96 [1998] ECR II-311 (para 51)).
87.
At this stage, it is now appropriate to consider further the extent to which the Commission defined its
position upon the applicants request to act contained in its letter of 20 February 1996.
88.
The applicant has rightly pointed out that that letter does not in any way define the Commissions
position in relation to the complaints in question because the Commission simply states that, having
examined the complaints and following completion of an outside study, it asked the Spanish authorities for
supplementary information. A letter from an institution called upon to act under art 175 of the Treaty
stating that the questions raised are being examined does not in fact amount to the defining of a position
such as to release it from its duty to act (see the judgments in Socit Nouvelle des Usines de Ponlieue
Aciries du Temple v ECSC High Authority Joined cases 42/59 and 49/59 [1961] ECR 53 and
European Parliament v EC Council Case 13/83 [1985] ECR 1513 (para 25)).
89.
It is, furthermore, common ground that the Commission had still not adopted any of the decisions
mentioned above at the time when this action was being considered.
90.
It follows from the foregoing that the Commission had, by 8 April 1996, failed to act, this date being two
months from receipt by it, on 8 February 1996, of the letter calling upon it to act on the ground that it
refrained from adopting a decision declaring that the state measures in question did not amount to aid
within the meaning of art 92(1) of the Treaty, or that those measures were to be classified as aid within
the meaning of art 92(1) but were compatible with the common market under art 92(2) and (3), or that the
procedure until art 93(2) of the Treaty had to be initiated, or from adopting a combination of these various
possible decisions according to the circumstances.
91.
Therefore, the claim for a declaration of failure to act must be considered to be well founded.
92.
 Accordingly, as it was only made in the alternative, there is no reason to rule on the claim for
annulment.
932

COSTS
93.
Under art 87(2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs if
they have been applied for in the successful partys pleadings.
94.
Since the Commission is the unsuccessful party, it will be ordered to pay the costs incurred by the
applicant, as sought by the applicant, other than the costs incurred as a result of the intervention by the
French Republic.
95.
Pursuant to art 87(4) of the Rules of Procedure, the French Republic shall bear its own costs. In
addition, it shall bear the costs incurred by the applicant as a result of its intervention.
On those grounds, the Court of First Instance (Third Chamber, Extended Composition) hereby: (1)
Declares that the European Commission failed to fulfil its obligations under the EC Treaty by failing to
adopt a decision following the two complaints lodged by the applicant on 2 March 1992 and 12 November
1993; (2) Orders the Commission to pay the applicants costs other than those incurred as a result of the
intervention by the French Republic; (3) Orders the French Republic to bear its own costs together with
the costs incurred by the applicant by reason of its intervention.

933

[1998] All ER (EC) 934

Canon Kabushiki Kaisha v Metro-Goldwyn-Mayer Inc (formerly Pathe

Communications Corp)
(Case C-39/97)

EUROPEAN COMMUNITY; Other European Community: INTELLECTUAL PROPERTY; Trade Marks

COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES
JUDGES RODRGUEZ IGLESIAS (PRESIDENT), GULMANN (RAPPORTEUR), RAGNEMALM,
WATHELET AND SCHINTGEN (PRESIDENTS OF CHAMBERS), KAPTEYN, MURRAY, EDWARD,
HIRSCH, JANN AND SEVN
ADVOCATE GENERAL JACOBS
20 JANUARY, 2 APRIL, 29 SEPTEMBER 1998
European Community Trade marks Likelihood of confusion Trade mark application opposed on
grounds of similarity with earlier mark Whether reputation of earlier mark to be taken into account when
determining likelihood of confusion Whether public perception that goods in question having different
places of origin sufficient to counter any likelihood of confusion Council Directive (EEC) 89/104, art 4(1)
(b).

MGM applied for the registration in Germany of the word trade mark CANNON to be used in respect of
films recorded on video tape cassettes production, distribution and projection of films for cinemas and
television organisations. CKK opposed that application on the ground that it infringed its own earlier word
trade mark Canon, which was registered in Germany in respect of still and motion picture cameras and
projectors; television filming and recording devices, television retransmission devices, television receiving
and reproduction devices, including devices for television recording and reproduction. The German
Patent Office dismissed the opposition for lack of similarity and CKKs appeal to the Federal Patent Court
was also dismissed. CKK subsequently appealed to the Federal Court of Justice, which stayed the
proceedings and referred to the Court of Justice of the European Communities for a preliminary ruling,
questions whether, on a proper construction of Council Directive (EEC) 89/104 to approximate the laws of
the member states relating to trade marks: (i) the distinctive character and reputation of the earlier trade
mark had to be taken into account when determining whether the similarity between the goods or services
covered by the two trade marks was sufficient to give rise to the likelihood of confusion, contrary to art
4(1)(b)1 of the directive; and (ii) whether there could be a likelihood of confusion where the public
perception was that the goods or services had different places of origin.
1
Article 4(1)(b), so far as material, is set out at p 947 c, post

Held (1) The more distinctive the earlier trade mark, the greater the risk of confusion: the protection of a
trade mark depended, in accordance with art 4(1)(b) of the directive, on there being a likelihood of
confusion and, as such, marks with a highly distinctive character, either per se or because of the
reputation they possessed on the market, enjoyed broader protection under the directive than marks with
a less distinctive character. It followed that art 4(1)(b) 934of the directive required, in addition to evidence
of similarity between the goods or services in question, that the distinctive character and in particular the
reputation of the earlier trade mark be taken into account when determining whether the similarity
between the goods or services covered by two trade marks was sufficient to give rise to the likelihood of
confusion (see p 948 f g, p 949 b j and p 950 h, post); Sabel BV v PUMA AG, Rudolph Dassler Sport
Case C-251/95 [1997] ECRI-6191 applied.
(2) The essential function of a trade mark was to guarantee the identity of the origin of the marked
product to the consumer or end user by enabling him to distinguish the product or service from others
which had another origin. Thus, a risk that the public might believe that the goods or services in question
came from the same or economically linked undertakings constituted a likelihood of confusion within the
meaning of art 4(1)(b) of the directive. It followed that, for the purposes of art 4(1)(b), there could be a
likelihood of confusion even where the public perception was that the goods were produced in different
places. Conversely, there could be no such likelihood where it did not appear that the public could believe
that goods or services came from the same or economically linked undertakings (see p 950 c to f j, post);
Sabel BV v PUMA AG, Rudolph Dassler Sport Case C-251/95 [1997] ECRI-6191 applied.

Cases cited
Bayerische Motorenwerke AG v Deenik Case C-63/97 (1998) Transcript (opinion) 2 April, ECJ.
British Sugar Plc v James Robertson & Sons Ltd [1996] RPC 281.
CNL-SUCAL (SA) NV v HAG GF AG Case C-10/89 [1990] ECR I-3711.
Sabel BV v PUMA AG, Rudolph Dassler Sport Case C-251/95 [1997] ECR I-6191.

Reference
By order of 12 December 1996, the Bundesgerichtshof (the Federal Court of Justice) referred to the Court
of Justice of the European Communities for a preliminary ruling under art 177 of the EC Treaty a question
(set out at p 948 b, post) on the interpretation of art 4(1)(b) of Council Directive (EEC) 89/104 to
approximate the laws of the member states relating to trade marks (OJ 1989 L40 p 1). That question
arose in proceedings between the Japanese company Canon Kabushiki Kaisha (CKK) and the American
corporation Metro-Goldwyn-Mayer Inc, formerly Pathe Communications Corp (MGM), following MGMs
application in Germany in 1986 for registration of the word trade mark CANNON to be used in respect of
certain goods and services. Written observations were submitted on behalf of: CKK, by G Jordan,
Rechtsanwalt, Karlsruhe; MGM, by W-W Wodrich, Rechtsanwalt, Essen; the French government, by K
Rispal-Bellanger, Assistant Director in the Legal Affairs Directorate of the Ministry of Foreign Affairs, and P
Martinet, Secretary of Foreign Affairs in the same directorate, acting as agents; the Italian government, by
Prof U Leanza, Head of the Legal Service in the Ministry of Foreign Affairs, acting as agent, and O
Fiumara, Avvocato dello Stato; the United Kingdom government, by L Nicoll, of the Treasury Solicitors
Department, acting as agent, and D Alexander, Barrister; and the European Commission, by J Grunwald,
Legal Adviser, and B J Drijber, of its Legal Service, acting as agents. Oral observations were made by:
CKK, represented by A Rinkler, Rechtsanwalt, Karlsruhe; MGM, represented by W-W Wodrich and J K
Zenz, Patentanwalt, Essen; the French government, represented by A de Bourgoing, Charg de 935
Mission in the Legal Affairs Directorate of the Ministry of Foreign Affairs, acting as agent; the Italian
government, represented by O Fiumara; the United Kingdom government, represented by D Alexander;
and the Commission, represented by J Grunwald. The language of the case was German. The facts are
set out in the opinion of the Advocate General.

2 April 1998.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
Article 4(1)(b) of Council Directive (EEC) 89/104 to approximate the laws of the member states relating
to trade marks (OJ 1989 L40 p 1) (the trade marks directive) prohibits the registration of a trade mark if
because of its identity with, or similarity to, [an] earlier trade mark and the identity or similarity of
the goods or services covered by the trade marks, there exists a likelihood of confusion on the part
of the public, which includes the likelihood of association with the earlier trade mark.
2.
In order to oppose the registration of a mark on the basis of that provision it is accordingly necessary
to show both that the mark is identical or similar to an earlier mark and that the goods or services covered
by both marks are identical or similar.
3.
The Bundesgerichtshof (the Federal Court of Justice) wishes to know whether, in assessing whether
goods or services should be considered to be similar within the meaning of that provision, the degree of
distinctiveness of an earlier mark, in particular its reputation, may be taken into account.

The trade marks directive

4.
The trade marks directive harmonises the provisions of national trade mark law which most directly
affect the functioning of the internal market (third recital of the preamble to the directive). Thus it
harmonises, inter alia, the grounds for refusing to register or invalidating a trade mark (arts 3 and 4), and
the rights conferred by a trade mark (arts 5ff).
5.
Under art 16(1) of the directive, member states were required to implement its provisions by 28
December 1991. However, by Council Decision (EEC) 92/10 (OJ 1992 L6 p 35) the EC Council made use
of the power conferred on it by art 16(2) and postponed the deadline for implementing the directive until
31 December 1992.
6.
Article 4(1) of the directive, which concerns the ability to register a mark, provides that:
A trade mark shall not be registered or, if registered, shall be liable to be declared invalid: (a) if
it is identical with an earlier trade mark, and the goods or services for which the trade mark is
applied for or is registered are identical with the goods or services for which the earlier trade mark
is protected; (b) if because of its identity with, or similarity to, the earlier trade mark and the identity
or similarity of the goods or services covered by the trade marks, there exists a likelihood of
confusion on the part of the public, which includes the likelihood of association with the earlier trade
mark.
7.
Similarly, art 5(1), which specifies the rights conferred by a trade mark, provides that:
936
The registered trade mark shall confer on the proprietor exclusive rights therein. The proprietor
shall be entitled to prevent all third parties not having his consent from using in the course of trade:
(a) any sign which is identical with the trade mark in relation to goods or services which are
identical with those for which the trade mark is registered; (b) any sign where, because of its
identity with, or similarity to, the trade mark and the identity or similarity of the goods or services
covered by the trade mark and the sign, there exists a likelihood of confusion on the part of the
public, which includes the likelihood of association between the sign and the trade mark.
8.
Marks with a reputation can benefit from yet further protection. Article 4(4)(a) gives member states the
option of refusing the registration of a mark in certain circumstances if the mark is similar or identical to an
earlier national mark which has a reputation, even though the goods or services in respect of which the
later marks application is made are not similar to the goods or services in respect of which the earlier
mark is registered:
Any Member State may furthermore provide that a trade mark shall not be registered or, if
registered, shall be liable to be declared invalid where, and to the extent that: (a) the trade mark is
identical with, or similar to, an earlier national trade mark within the meaning of paragraph 2 and is
to be, or has been, registered for goods or services which are not similar to those for which the
earlier trade mark is registered, where the earlier trade mark has a reputation in the Member State
concerned and where the use of the later trade mark without due cause would take unfair
advantage of, or be detrimental to, the distinctive character or the repute of the earlier trade mark

9.
Where the earlier mark is a Community trade mark provided for by Council Regulation (EC) 40/94 on
the Community trade mark (OJ 1994 L11 p 1) (the trade mark regulation) art 4(3) of the directive allows
for the same type of objection to registration to be made by the owner of an earlier Community trade mark
which has a reputation in the Community. In contrast to art 4(4)(a), art 4(3) requires, rather than merely
empowers, member states to afford such protection.
10.
Furthermore, art 5(2) (which concerns the use, as opposed to the registration, of a later mark) gives
member states a similar option to that provided in art 4(4)(a):
Any Member State may also provide that the proprietor shall be entitled to prevent all third
parties not having his consent from using in the course of trade any sign which is identical with, or
similar to, the trade mark in relation to goods or services which are not similar to those for which
the trade mark is registered, where the latter has a reputation in the Member State and where use
of that sign without due cause takes unfair advantage of, or is detrimental to, the distinctive
character or the repute of the trade mark.
11.
It should, however, be noted that, although the question refers to marks with a reputation and arts 4(4)
(a) and 5(2) mention such marks specifically, the Bundesgerichtshof has made it clear that the provision
in question in the present case is art 4(1)(b) for the reasons explained below (see para 26, below).
937

The facts
12.
On 29 July 1986, Pathe Communications Corp (Pathe), which is based in the United States of
America, submitted an application to register the word trade mark CANNON in respect of the following
goods and services: films recorded on video tape cassettes (video film cassettes); production, distribution
and projection of films for cinemas and television institutions.
13.
That application was opposed by Canon Kabushiki Kaisha (CKK) on the grounds that it infringed its
own word trade mark Canon. That mark had already been registered, inter alia, in respect of
still and motion picture cameras and projectors; television filming and recording devices,
television transmission devices, television receiving and reproduction devices, including tape and
disc devices for television recording and reproduction.
14.
At the time the opposition by CKK was lodged the trade marks directive had not been adopted and the
national German law on trade marks accordingly applied. That law is known as the Warenzeichengesetz
(the WZG). The directive, adopted on 21 December 1988 and due to be implemented by 31 December
1992 (see para 5, above), was implemented late into German law by a law adopted on 25 October 1994.
The principal provisions of that law came into force on 1 January 1995. However, the Bundesgerichtshof
explains that the present case must be judged on the basis of the law as it now stands, which gives effect
to the directive. The new German trade mark law is called the Markengesetz and the Bundesgerichtshof
explains that art 9(1)(2) of that law corresponds to art 4(1)(b) of the directive.
15.
According to the Bundesgerichtshof, it must be assumed for the purposes of legal assessment that the
two marks CANNON and Canon sound identical. They are not, however, applied in respect of identical
goods and services. The question with which the German courts have been confronted is whether the
respective goods and services can nevertheless be regarded as similar.
16.
When Pathes application was examined by the German authorities the first examiner considered that
the goods and services of the opposing parties were indeed similar and accordingly refused to register
the mark CANNON. The second examiner set aside that decision and dismissed the opposition on the
ground that there was no similarity. CKK appealed to the Bundespatentgericht (the Federal Patent Court)
but its appeal was dismissed by an order dated 6 April 1994. CKK then appealed to the
Bundesgerichtshof and it is in the context of those proceedings that the present reference has been
made.
17.
The Bundespatentgericht dismissed CKKs appeal because it agreed with the second examiner that
there was no similarity, within the meaning of para 5(4)(1) of the WZG, between the parties goods and
services. In its view, there could be similarity only if the goods or services, in accordance with their
economic significance and use, and with respect in particular to their usual places of manufacture and
sale, had such close points of contact that the average purchaser might form the opinion that they came
from the same business operation. CKK claims that 766% of the population knew its mark in November
1985 and the Bundesgerichtshof states that that must be taken as meaning that the mark Canon was a
recognised mark. The Bundespatentgericht, however, considered that the reputation of CKKs mark was
of no significance in assessing the similarity of the goods and services in question.
938
18.
The Bundespatentgericht observed that the goods video film cassettes specified in Pathes
application were closest to the goods television filming and recording devices, television transmission
devices, television receiving and reproduction devices, including tape and disc devices for television
recording and reproduction covered by CKKs mark. However, it considered that the two sets of goods
were not similar. Disagreeing with the view taken by the Thirtieth Senate of the Bundespatentgericht in a
similar case, it stated that it could not be assumed that video film cassettes were similar to the television
devices covered by CKKs mark or to the video cameras distributed by CKK.
19.
It explained that the Bundespatentgericht had already found, in 1989, that there was not a single
manufacturer of leisure electronic devices to be found among the video tape producers in the 1988
SEIBT Directory of German Industries; no significant changes had taken place in the meantime in that
respect, at least in relation to recorded video cassettes; and inquiries in relevant specialist shops had
shown that no name of a manufacturer of television devices or video recorders could be found in the
range of recorded video cassettes. The Bundespatentgericht accordingly considered that it could not be
assumed that the relevant average purchaser thought that recorded video tapes and the corresponding
recording and reproduction devices came from the same business operation. Even members of the public
were sufficiently aware of the different conditions for the manufacture of recorded cassettes and
understood that video cassettes and video recorders do not come from the same manufacturer.
20.
The Bundespatentgericht also rejected the possibility of similarity between the services specified in
Pathes application relating to production, distribution and projection of films for cinemas and television
institutions and the television filming devices etc. protected by CKKs mark. The Bundespatentgericht
considered that the fact that cameras and projectors were used to produce and project films did not
mislead persons, to an extent relevant for trade mark law purposes, to conclude that the producers of
such devices regularly also produced, distributed or projected films.
21.
In its appeal against the order of the Bundespatentgericht, CKK argues that, since the implementation
of the trade marks directive into German law, the approach of the Bundespatentgericht to the assessment
of the similarity of goods or services is no longer appropriate. It submits that its mark Canon is a famous
or well-known mark and that that fact, coupled with the fact that video film cassettes and video recording
and reproduction devices are offered through the same points of sale, should lead to the conclusion that
the goods covered by the two marks are similar and that there is consequently a likelihood of the public
being confused within the meaning of para 9(1)(2) of the Markengesetz (which corresponds to art 4(1)(b)
of the directive).

The question
22.
The Bundesgerichtshof has accordingly referred the following question to this Court of Justice for a
preliminary ruling:
May account be taken, when assessing the similarity of the goods or services covered by the
two marks, of the distinctive character, in particular the reputation of the mark with earlier priority
(on the date which determines the seniority of the later mark), in particular in such a way that
likelihood of confusion within the meaning of Article 4(1)(b) of the Directive 939 must be taken to
exist even if the public attributes the goods and/or services to different origins?
23.
The Bundesgerichtshof explains that the essential question is whether the adoption of the trade marks
directive requires the German courts to change their approach in assessing the similarity of goods or
services. It accordingly seeks to ascertain which criteria should be applied in assessing whether goods or
services are similar within the meaning of art 4(1)(b) of the directive.
24.
The order for reference contains the following information as to the implementation of the directive.
When implementing the directive, the German legislature started from the assumption that the concept of
the similarity of goods or services could not be understood in the same way as that concept had been
understood under the previous German law. In the explanatory memorandum to the draft
Markenrechtsreformgesetz (the Trade Mark Reform Law), it was stated that in future it would not be
possible to refer back to the static concept of similarity developed in the previous law.
25.
Under the previous law, there had to be objective similarity between the goods or services: there was
thus no protection under trade mark law where there was no objective similarity of goods and services,
however similar the marks and whatever the reputation of the earlier mark. Commentators argue that,
since implementation of the directive, that is no longer the case: there is now an inverse correlation
between, on the one hand, the similarity of the goods and services and, on the other, the similarity of the
marks and the distinctive character of the earlier mark. Thus the closer the marks and the more distinctive
the earlier mark, the less similarity of goods or services needs to be shown. According to the
Bundesgerichtshof, such an interpretation would mean that it would be considerably easier than under the
previous German law to demonstrate a likelihood of confusion.
26.
The Bundesgerichtshof recognises that, in certain circumstances, where the earlier mark has a
reputation, it can be protected even in relation to dissimilar goods and services by virtue of art 4(4)(a) of
the directive. Although that provision is optional, the Bundesgerichtshof states that it has been
implemented into German law by para 9(1)(3) of the Markengesetz. However, the Bundesgerichtshof
stresses that it is important to distinguish between the application of art 4(1)(b) and art 4(4)(a) because,
under national law, the initial registration of a mark in relation to dissimilar goods cannot be opposed per
se under the national provisions implementing art 4(4)(a): the person objecting can only commence an
action for cancellation of the mark once it has been registered or bring proceedings for infringement of his
own mark, the idea being that the registration procedure should be carried out in an abstract, systematic
way. Article 4(1)(b), on the other hand, is a ground for opposing the registration of a mark. The question
whether a particular use of a mark falls within art 4(1)(b) or art 4(4)(a) is accordingly of considerable
practical importance.

The meaning of confusion

27.
The question asks in part whether a likelihood2 of confusion must be taken to exist even if the public
attributes the goods or services to different origins. The meaning of confusion in art 4(1)(b) of the
directive has already been considered 940 by this court in its judgment in the case of Sabel BV v PUMA
AG, Rudolph Dassler Sport Case C-251/95 [1997] ECR I-6191.
2
The German version of the directive speaks of a risk of confusion, while the English version speaks of a likelihood of
confusion.

28.
That case concerned the interpretation of art 4(1)(b) of the directive in so far as it refers to a likelihood
of confusion on the part of the public, which includes the likelihood of association with the earlier trade
mark. The court explained that it had been submitted that
the likelihood of association may arise in three sets of circumstances: (1) where the public
confuses the sign and the mark in question (likelihood of direct confusion): (2) where the public
makes a connection between the proprietors of the sign and those of the mark and confuses them
(likelihood of indirect confusion or association); (3) where the public considers the sign to be similar
to the mark and perception of the sign calls to mind the memory of the mark, although the two are
not confused (likelihood of association in the strict sense). (See [1997] ECR I-6191 (para 16).)
29.
The court stated that it was therefore necessary to determine whether Article 4(1)(b) can apply
where there is no likelihood of direct or indirect confusion, but only a likelihood of association in the strict
sense (see [1997] ECRI-6191 (para 17)). It concluded: The terms of the provision itself exclude its
application where there is no likelihood of confusion on the part of the public (see para 18). Thus the
court held that the mere association which the public might make between two trade marks as a result of
their analogous semantic content is not in itself a sufficient ground for concluding that there is a likelihood
of confusion (see operative part) within the meaning of art 4(1)(b).
30.
It follows that if, in the present case, there is no likelihood of the public assuming that there is any sort
of trade connection between the marks Canon and CANNON, there is no likelihood of confusion within
the meaning of art 4(1)(b) of the directive. The European Commission suggests however that the question
refers to the attribution of goods or services to different places of origin; and that concept may reflect the
importance attached by the previous German trade mark law to the place of manufacture of the goods in
question. In that respect it should be noted that it is not sufficient to show simply that there is no likelihood
of the public being confused as to the place in which the goods are manufactured or the services
performed: if, despite recognising that the goods or services have different places of origin, the public is
likely to believe that there is a link between the two concerns, there will be a likelihood of confusion within
the meaning of the directive.

Assessment of the similarity of goods and services

31.
The main argument in this case has focused on the question whether the degree of distinctiveness of
a mark, in particular its reputation, can be taken into account when assessing whether goods or services
should be regarded as similar within the meaning of art 4(1)(b). In other words, is it permissible to
consider goods or services to be similar in relation to particularly distinctive marks when such goods or
services would not be considered to be similar in relation to other, less distinctive marks? Or should the
test for assessing the similarity of goods or services be objective (ie unrelated to the nature of the marks
in question)?
32.
All trade marks, if they are to perform their function, should be distinctive; a trade mark which is devoid
of any distinctive character is, pursuant to art 3(1)(b) of the directive, not to be registered and, if
registered, is liable to be declared invalid. But distinctiveness is a matter of degree. A trade mark might be
941 particularly distinctive either because it is well known or because it is of an unusual character. The
more well known or unusual a trade mark, the more likely it is that consumers might be confused into
believing there to be a trade connection between goods or services bearing the same or a similar mark.
As the court observed in its judgment in Sabel BV [1997] ECR I-6191 (para 24), the more distinctive the
earlier mark, the greater will be the likelihood of confusion. It should be noted, however, that in that case,
in contrast to the present case, it was not disputed that at least some of the goods to which the marks in
question related were the same; the question was whether the marks (as opposed to the goods) in
question were sufficiently similar to give rise to a likelihood of confusion.
33.
CKK, the French government and the Commission are all of the view that the degree of distinctiveness
of a mark is relevant to the test of the similarity of products or services. At the hearing the Italian
government stated that the notion of similarity is a very vague concept which cannot be based on
objective factors alone.
34.
They refer to the tenth recital of the preamble of the directive, which is in the following terms:
whereas it is indispensable to give an interpretation of the concept of similarity in relation to
the likelihood of confusion; whereas the likelihood of confusion, the appreciation of which depends
on numerous elements and, in particular, on the recognition of the trade mark on the market, on 3
the association which can be made with the used or registered sign, on the degree of similarity
between the trade mark and the sign and between the goods or services identified, constitutes the
specific condition for such protection; whereas the ways in which likelihood of confusion may be
established, and in particular the onus of proof, are a matter for national procedural rules which are
not prejudiced by the Directive

3
The word of, rather than on, appears in the English text by mistake.

35.
CKK and the French government argue that that recital, in particular the statement that it is
indispensable to give an interpretation of the concept of similarity in relation to the likelihood of confusion,
shows that the test of the similarity of goods or services is not to be regarded as an objective test.
36.
CKK also argues that it is important to be able to oppose the initial registration of a mark under art 4(1)
(b), rather than having to accept the initial registration and attack its use under other provisions. It
considers that parties to opposition proceedings are subject to lower costs and can present their rights
more effectively and more efficiently than in other proceedings.
37.
 Pathe and the United Kingdom, however, advocate an objective, independent assessment of the
similarity of the goods or services (ie an assessment made without regard to the nature or reputation of
the earlier mark). The United Kingdom maintains that to require, at the stage of registering a mark, that
the reputation of an earlier mark be taken into account when assessing the similarity of the goods or
services in question would place an undue burden on examiners and considerably lengthen the
registration process. Pathe also argues that large companies would deliberately delay registration
processes.
38.
Moreover, Pathe argues that flexible boundaries to the definition of similar goods or services would
cause legal uncertainty. One final argument made by the United Kingdom is that, if the question of the
likelihood of confusion had to be 942 addressed in order to decide whether goods or services were
similar, there would be no purpose in requiring such similarity: the only question would be whether or not
there was a likelihood of confusion; if that had been the intention, the directive would have had a different
structure.
39.
In my view, the decisive consideration in resolving the issue is the statement in the tenth recital of the
preamble to the directive that the appreciation of the likelihood of confusion depends in particular on the
recognition of the mark. That statement set in its context reads as follows:
Whereas the protection afforded by the registered trade mark, the function of which is in
particular to guarantee the trade mark as an indication of origin, is absolute in the case of identity
between the mark and the sign and goods or services; whereas the protection applies also in case
of similarity between the mark and the sign and the goods or services; whereas it is indispensable
to give an interpretation of the concept of similarity in relation to the likelihood of confusion;
whereas the likelihood of confusion, the appreciation of which depends on numerous elements and,
in particular, on the recognition of the trade mark on the market, [on] the association which can be
made with the used or registered sign, [on] the degree of similarity between the trade mark and the
sign and between the goods or services identified, constitutes the specific condition for such
protection
That statement makes it clear that the recognition of the mark, although not specifically mentioned in
art 4(1)(b) of the directive, is relevant in deciding whether there is sufficient similarity to give rise to a
likelihood of confusion.
40.
That view is also confirmed by the judgment of the court in Sabel BV [1997] ECR I-6191 (para 22), in
which it held that the likelihood of confusion must be appreciated globally, taking into account all
factors relevant to the circumstances of the case. It is true that that statement was made in a different
context: the court was there considering the question whether conceptual similarity of the marks alone
could give rise to confusion within the meaning of art 4(1)(b), in a situation in which the goods in question
were clearly the same. However the statement is one of general application.
41.
The UK government seeks to refute the argument that the tenth recital of the preamble to the directive
supports a global approach. It maintains that that recital means simply that in assessing similarity regard
should be had to the question whether the goods or services are such that the public might be confused
into thinking that they have the same trade origin, and that in making that assessment it is not permissible
to have regard to the reputation of the earlier mark.
42.
That explanation, however, requires the recital to be read as indicating that the question of confusion
is to be taken into account in assessing the similarity of goods or services, but that one element of the
confusion test, namely that of the recognition of the earlier mark on the market (which is mentioned
expressly in the recital), cannot be taken into account in such an assessment. I have difficulty in reading
the recital in that way. (The phrase recognition of the trade mark on the market to my mind refers to the
degree of distinctiveness of the mark: ie whether it is readily recognised by the public, either because it
has an inherently unusual nature or because of its reputation.)
43.
In addition, the dangers of lengthening the registration process by requiring consideration of an earlier
trade marks reputation do not appear to me to be as serious as Pathe and the UK government suggest.
The French 943 government stated at the hearing that, in its experience, such consideration did not
unduly lengthen or complicate the procedure. Indeed, it may be in the interest of legal certainty to ensure
that marks whose use may be challenged successfully are not registered in the first place. In any event, it
seems to me that the tenth recital of the preamble to the directive indicates that the reputation of a trade
mark should be taken into account in assessing the likelihood of confusion between two marks even if it
cannot be taken into account in assessing the similarity of goods and services. Moreover, the Community
Trade Mark Registry will be obliged to consider the question of the reputation of a mark in many cases
since the Community trade mark regulation contains a provision similar to art 4(4)(a) of the directive.
Under art 8(5) of the regulation the proprietor of an earlier mark which has a reputation can oppose,
subject to certain conditions, the registration of an identical or similar mark in relation to dissimilar goods
or services. That suggests that the practical problems of requiring registrars to consider the reputation of
a mark are not as great as has been argued.
44.
I would emphasise that although in my view the degree of recognition of the mark must be taken into
account in deciding whether there is sufficient similarity to give rise to confusion, the requirement of
similarity must be given full weight, both in assessing the similarity of the marks and in assessing the
similarity of the goods or services in question. It is therefore incorrect to suggest that, in consequence of
the implementation into national law of art 4(1)(b) of the directive, it may no longer be necessary in the
case of a particularly distinctive mark to establish the similarity of the goods or services in question. In
assessing the similarity of the goods or services it will be helpful to have regard to the factors suggested
by the UK and French governments.
45.
According to the UK government, the following type of factors should be taken into account in
assessing the similarity of goods or services: (a) the uses of the respective goods or services; (b) the
users of the respective goods or services; (c) the physical nature of the goods or acts of service; (d) the
trade channels through which the goods or services reach the market; (e) in the case of self-serve
consumer items, where in practice they are respectively found or likely to be found in supermarkets and in
particular whether they are, or are likely to be, found on the same or different shelves; (f) the extent to
which the respective goods or services are in competition with each other: that inquiry may take into
account how those in trade classify goods, for instance whether market research companies, who of
course act for industry, put the goods or services in the same or different sectors (taken from the
judgment of the English High Court in British Sugar Plc v James Robertson & Sons Ltd [1996] RPC 281).
46.
Whilst recognising that that list of factors is not exhaustive, the UK government observed at the
hearing that it nevertheless indicates a common denominator which should be present in all factors taken
into account in assessing the similarity of goods or services: namely that the factors are related to the
goods or services themselves.
47.
The French government likewise considers that, in assessing the similarity of goods and services, the
factors to be taken into account should include the nature of the goods or services, their intended
destination and clientele, their normal use and the usual manner of their distribution.
48.
The use of those objective factors to assess similarity does not however in my view preclude account
being taken of the degree of recognition of the mark 944 in deciding whether there is sufficient similarity to
give rise to a likelihood of confusion.
49.
Against that view it might be argued that the simpler and more objective the test of the similarity of
goods and services under art 4(1)(b), the less likely national trade mark registries or courts in different
member states would be to adopt different assessments as to whether a particular mark is confusing.
That would be consistent with the directives aim of harmonising member states trade mark laws.
50.
I accept that a flexible test of the similarity of goods or services might lead to different interpretations of
such similarity in different member states. It is indeed possible that, contrary to the view in certain
member states, a new mark might not be caught by art 4(1)(b) of the directive in one member state simply
because it is considered in that state that, despite the reputation of the earlier mark and a likelihood of
confusion, the goods or services are not sufficiently similar. In such a case, however, the fact that the
earlier mark has a reputation may well mean that in that member state art 4(4)(a) or art 5(2) of the
directive (concerning the protection of a mark in relation to dissimilar goods or services) would apply
instead. According to the understanding of the Commission, all member states have availed themselves
of the option provided by art 4(4)(a) (statement made by the Commission at the hearing in Bayerische
Motorenwerke AG v Deenik Case C-63/97 on 13 January 19984). Thus the ultimate result in all member
states (namely the prohibition, or cancellation, of the registration of a mark or prohibition of its use) would
often be the same.
4
Ed: the judgment was still pending at date of going to press. However, for the opinion of Advocate General Jacobs see
Bayerische Motorenwerke AG v Deenik Case C-63/97 (1998) ECJ Transcript (opinion) 2 April.

51.
By way of a final observation I would add that I do not consider it unjust for a trade mark owner to
benefit from protection in relation to a wider range of goods than those in relation to which the mark is
registered. It is not reasonable to require a trade mark owner to register his mark in relation to all types of
goods in relation to which use of his mark may give rise to a risk of confusion, because he may not be
using his mark in relation to such goods; indeed marks which are registered in respect of goods or
services in relation to which they are not used are liable to be struck off the register after five years on the
grounds of non-use (see arts 10 to 12 of the directive). Moreover, the criterion of confusion ensures that
when registering a mark in relation to a certain class of goods or services, the trade mark owner is not
thereby protected in relation to too broad a range of goods and services. The concept of confusion should
not be extended too far since, as I observed in my opinion in Sabel BV [1997] ECR I-6191 (paras 5051)
a broad interpretation would be contrary to the directives aim of assisting the free movement of goods.
However, where there is a genuine and properly substantiated likelihood of confusion, it is in my view not
only justifiable but necessary to protect both the consumer and the trade mark owner by disallowing the
registration of a later mark even in relation to similar goods and services in respect of which the earlier
mark is not registered.

Conclusion
52.
 Accordingly the question referred by the Bundesgerichtshof should in my opinion be answered as
follows:
945
In the assessment of the similarity of goods or services covered by two marks within the
meaning of art 4(1)(b) of Council Directive (EEC) 89/104 to approximate the laws of the member
states relating to trade marks, account may be taken of the distinctive character, in particular the
reputation, of the earlier mark in deciding whether there is sufficient similarity to give rise to a
likelihood of confusion. However, there will only be a likelihood of confusion within the meaning of
that provision if it is likely that the public will be confused into thinking that there is some sort of
trade connection between the suppliers of the goods or services in question.

29 September 1998.

The COURT OF JUSTICE

delivered the following judgment.
1.
By order of 12 December 1996, received at the Court of Justice of the European Communities on 28
January 1997, the Bundesgerichtshof (the Federal Court of Justice) referred to the Court of Justice for a
preliminary ruling under art 177 of the EC Treaty a question on the interpretation of art 4(1)(b) of Council
Directive (EEC) 89/104 to approximate the laws of the member states relating to trade marks (OJ 1989
L40 p 1).
2.
That question was raised in proceedings between the Japanese company Canon Kabushiki Kaisha
(CKK) and the American corporation Metro-Goldwyn- Mayer Inc, formerly Pathe Communications Corp
(MGM), following MGMs application in Germany in 1986 for registration of the word trade mark
CANNON to be used in respect of the following goods and services: films recorded on video tape
cassettes (video film cassettes); production, distribution and projection of films for cinemas and television
organisations.
3.
Referring to para 5(4)(1) of the Warenzeichengesetz (the WZG) (the former Trade Mark Law), CKK
opposed that application before the Deutsches Patentamt (the German Patent Office) on the ground that
it infringed its earlier word trade mark Canon, registered in Germany in respect of, inter alia:
still and motion picture cameras and projectors; television filming and recording devices,
television retransmission devices, television receiving and reproduction devices, including tape and
disc devices for television recording and reproduction.
4.
The first examiner of the German Patent Office considered that the two marks were analogous and
therefore refused registration on the ground that the respective goods and services were similar within the
meaning of para 5(4)(1) of the WZG. The second examiner set aside that decision and dismissed the
opposition for lack of similarity.
5.
The Bundespatentgericht (the Federal Patent Court) dismissed CKKs appeal against the latter
decision, holding that there was no similarity within the meaning of para 5(4)(1) of the WZG. It stated that
such similarity could be taken to exist only where the goods or services, having regard to their economic
significance and method of use and, in particular, their usual place of manufacture and sale, were so
similar that the average purchaser might form the opinion that they were manufactured by the same
enterprise in so far as similar or supposedly similar distinguishing signs were used. The court considered
that in the circumstances of the case that condition was not satisfied.
6.
CKK brought an appeal against the order of the Bundespatentgericht before the Bundesgerichtshof.
946
7.
In its order for reference, the Bundesgerichtshof makes the preliminary point that the case pending
before it must be decided on the basis of the Markengesetz (the new German Law on Trade Marks),
which entered into force on 1 January 1995 and which transposed the directive into German law and para
9(1)(2) of which corresponds to art 4(1)(b) of the directive.
8.
Article 4(1)(b) provides:
(1) A trade mark shall not be registered or, if registered, shall be liable to be declared invalid
(b) if because of its identity with, or similarity to, the earlier trade mark and the identity or similarity
of the goods or services covered by the trade marks, there exists a likelihood of confusion on the
part of the public, which includes the likelihood of association with the earlier trade mark.
9.
In order to illustrate the context and significance of the question referred the Bundesgerichthof gave
the following information: in this instance the two signs, CANNON and Canon, are pronounced in the
same way and the mark Canon has a reputation; in addition, as the Bundespatentgericht noted, the
public perception is that films recorded on video tape cassettes (video film cassettes) and recording and
reproduction devices for video tapes (video recorders) do not come from the same manufacturer; in
conformity with the principles laid down in the WZG, the Bundespatentgericht attached no importance in
its decision to the identical nature of the signs or to the reputation of the opposing partys trade mark;
since the Markengesetz must henceforth be applied, it is necessary to establish the criteria to be applied
in interpreting the concept of similarity of the goods or services covered by the trade marks for the
purposes of art 4(1)(b) of the directive; if no account is taken in this case of the reputation of the earlier
mark when assessing the likelihood of confusion, because there is no similarity between the goods or
services covered by the two marks, then on the basis of the findings made by the Bundespatentgericht
the appeal brought by the opposing party cannot succeed; it is possible, however, to interpret art 4(1)(b)
of the directive as meaning that the reputation of the earlier mark may not only reinforce the
distinctiveness of the mark as such, but may also be sufficient to exclude the view held by the public
concerning the place of origin (Herkunftssttte) of those goods or services from the assessment of the
similarity of the goods or services covered; according to academic opinion, when the likelihood of
confusion within the meaning of the Markengesetz is assessed, it may be necessary to establish a
correlation between the similarity of the goods, on the one hand, and the degree of similarity of the
respective signs and the distinctive character of the mark to be protected, on the other, in such a way that
the closer the marks resemble one another and the more distinctive the mark for which protection is
sought, the less the similarity of the goods need be.
10.
Finally, the Bundesgerichtshof points out that the interpretation of art 4(1)(b) of the directive is of
particular importance in practice in view of the fact that the ground for refusing registration provided for by
para 9(1)(3) of the Markengesetz cannot be relied on in a national opposition procedure but only before
the ordinary courts in the course of an action for cancellation of a trade mark or for its infringement (this
paragraph transposes art 4(4)(a) of the directive, according to which the member states may provide for
broader protection in the case of trade marks with a reputation, derogating from the requirement that the
goods or services should be similar).
947
11.
In the light of those considerations the Bundesgerichtshof stayed the proceedings and referred the
following question to the Court of Justice for a preliminary ruling:
May account be taken, when assessing the similarity of the goods or services covered by the
two marks, of the distinctive character, in particular the reputation, of the mark with earlier priority
(on the date which determines the seniority of the later mark), so that, in particular, likelihood of
confusion within the meaning of Article 4(1)(b) of Directive 89/104/EEC must be taken to exist even
if the public attributes the goods and/or services to different places of origin (Herkunftssttten)?
12.
In the first part of the question, the Bundesgerichtshof asks in substance whether, on a proper
construction of art 4(1)(b) of the directive, the distinctive character of the earlier trade mark, and in
particular its reputation, must be taken into account when determining whether the similarity between the
goods or services covered by the two trade marks is sufficient to give rise to the likelihood of confusion.
13.
CKK, the French and Italian governments and the European Commission are essentially in agreement
in proposing that the question be answered in the affirmative.
14.
On the other hand, MGM and the UK government consider that similarity between goods and services
must be assessed objectively and independently, and no account should be taken of the distinctive
character of the earlier mark or in particular of its reputation.
15.
It is to be noted, first, that the tenth recital of the preamble to the directive states:
the protection afforded by the registered trade mark, the function of which is in particular to
guarantee the trade mark as an indication of origin, is absolute in the case of identity between the
mark and the sign and goods or services the protection applies also in case of similarity between
the mark and the sign and the goods or services it is indispensable to give an interpretation of
the concept of similarity in relation to the likelihood of confusion the likelihood of confusion, the
appreciation of which depends on numerous elements and, in particular, on the recognition of the
trade mark on the market, [on] the association which can be made with the used or registered sign,
[and on] the degree of similarity between the trade mark and the sign and between the goods or
services identified, constitutes the specific condition for such protection.
16.
Second, the court has held that the likelihood of confusion on the part of the public, in the absence of
which art 4(1)(b) of the directive does not apply, must be appreciated globally, taking into account all
factors relevant to the circumstances of the case (see the judgment in Sabel BV v PUMA AG, Rudolph
Dassler Sport Case C-251/95 [1997] ECR I-6191 (para 22)).
17.
A global assessment of the likelihood of confusion implies some interdependence between the
relevant factors, and in particular a similarity between the trade marks and between these goods or
services. Accordingly, a lesser degree of similarity between these goods or services may be offset by a
greater degree of similarity between the marks, and vice versa. The interdependence of these factors is
expressly mentioned in the tenth recital of the preamble to the directive, which states that it is
indispensable to give an 948 interpretation of the concept of similarity in relation to the likelihood of
confusion, the appreciation of which depends, in particular, on the recognition of the trade mark on the
market and the degree of similarity between the mark and the sign and between the goods or services
identified.
18.
Furthermore, according to the case law of the court, the more distinctive the earlier mark, the greater
the risk of confusion (see [1997] ECR I-6191 (para 24)). Since protection of a trade mark depends, in
accordance with art 4(1)(b) of the directive, on there being a likelihood of confusion, marks with a highly
distinctive character, either per se or because of the reputation they possess on the market, enjoy broader
protection than marks with a less distinctive character.
19.
It follows that, for the purposes of art 4(1)(b) of the directive, registration of a trade mark may have to
be refused, despite a lesser degree of similarity between the goods or services covered, where the marks
are very similar and the earlier mark, in particular its reputation, is highly distinctive.
20.
Against that interpretation, MGM and the UK government have argued that to take into account the
distinctiveness of the earlier mark when assessing the similarity of the goods or services involves the
danger of prolonging the registration procedure. However, the French government has stated that in its
experience consideration of that factor when assessing the similarity of the goods or services covered did
not have the effect of unduly lengthening or complicating the registration procedure.
21.
In this context, it is important to note that even if the suggested interpretation makes the registration
procedure much lengthier, that cannot be decisive for the interpretation of art 4(1)(b) of the directive. In
any event, for reasons of legal certainty and proper administration, it is necessary to ensure that trade
marks whose use could successfully be challenged before the courts are not registered.
22.
It is, however, important to stress that, for the purposes of applying art 4(1)(b), even where a mark is
identical to another with a highly distinctive character, it is still necessary to adduce evidence of similarity
between the goods or services covered. In contrast to art 4(4)(a), which expressly refers to the situation in
which the goods or services are not similar, art 4(1)(b) provides that the likelihood of confusion
presupposes that the goods or services covered are identical or similar.
23.
In assessing the similarity of the goods or services concerned, as the French and UK governments
and the Commission have pointed out, all the relevant factors relating to those goods or services
themselves should be taken into account. Those factors include, inter alia, their nature, their end users
and their method of use and whether they are in competition with each other or are complementary.
24.
In the light of the foregoing, the answer to be given to the first part of the question must be that, on a
proper construction of art 4(1)(b) of the directive, the distinctive character of the earlier trade mark, and in
particular its reputation, must be taken into account when determining whether the similarity between the
goods or services covered by the two trade marks is sufficient to give rise to the likelihood of confusion.
25.
In the second part of the question the Bundesgerichtshof asks in substance whether there can be a
likelihood of confusion within the meaning of art 4(1)(b) of the directive where the public perception is that
the goods or services have different places of origin (Herkunftssttten).
949
26.
There is a likelihood of confusion within the meaning of art 4(1)(b) of the directive where the public can
be mistaken as to the origin of the goods or services in question.
27.
Indeed, art 2 of the directive provides that a trade mark must be capable of distinguishing the goods or
services of one undertaking from those of other undertakings, while the tenth recital in the preamble to the
directive states that the function of the protection conferred by the mark is primarily to guarantee the
indication of origin.
28.
Moreover, according to the settled case law of the court, the essential function of the trade mark is to
guarantee the identity of the origin of the marked product to the consumer or end user by enabling him,
without any possibility of confusion, to distinguish the product or service from others which have another
origin. For the trade mark to be able to fulfil its essential role in the system of undistorted competition
which the Treaty seeks to establish, it must offer a guarantee that all the goods or services bearing it have
originated under the control of a single undertaking which is responsible for their quality (see esp SA
CNL-SUCAL NV v HAG GF AG Case C-10/89 [1990] ECR I-3711 (paras 1413)).
29.
Accordingly, the risk that the public might believe that the goods or services in question come from the
same undertaking or, as the case may be, from economically linked undertakings, constitutes a likelihood
of confusion within the meaning of art 4(1)(b) of the directive (see Sabel BV [1997] ECR I-6191 (paras
1618)). Consequently, as the Advocate General states at para 30 of his opinion (above), in order to
demonstrate that there is no likelihood of confusion, it is not sufficient to show simply that there is no
likelihood of the public being confused as to the place of production of the goods or services.
30.
The answer to be given to the second part of the question must therefore be that there may be a
likelihood of confusion within the meaning of art 4(1)(b) of the directive even where the public perception
is that the goods or services have different places of production. By contrast, there can be no such
likelihood where it does not appear that the public could believe that the goods or services come from the
same undertaking or, as the case may be, from economically linked undertakings.

Costs
31.
The costs incurred by the French, Italian and UK governments and by the European Commission,
which have submitted observations to the Court of Justice, are not recoverable. Since these proceedings
are, for the parties to the main proceedings, a step in the action pending before the national court, the
decision on costs is a matter for that court.
On those grounds, the Court of Justice, in answer to the question referred to it by the
Bundesgerichtshof by order of 12 December 1996, hereby rules: On a proper construction of art 4(1)(b) of
Council Directive (EEC) 89/104 to approximate the laws of the member states relating to trade marks, the
distinctive character of the earlier trade mark, and in particular its reputation, must be taken into account
when determining whether the similarity between the goods or services covered by the two trade marks is
sufficient to give rise to the likelihood of confusion. There may be a likelihood of confusion within the
meaning of art 4(1)(b) of the directive even where the public perception is that the goods or services have
different places of production. By contrast, there can be no such likelihood where it does not appear that
the public could believe that the goods or services come from the same undertaking or, as the case may
be, from economically linked undertakings.

950

[1998] All ER (EC) 951

European Commission v France (United Kingdom intervening)

(Case C-43/96)
EUROPEAN COMMUNITY; Taxation: TAXATION: VAT and Customs and Excise
COURT OF JUSTICE OF THE EUROPEAN COMMUNITIES (SIXTH CHAMBER)
JUDGES SCHINTGEN (PRESIDENT OF THE SECOND CHAMBER, ACTING FOR THE PRESIDENT
OF THE SIXTH CHAMBER), MANCINI, KAPTEYN, MURRAY AND HIRSCH (RAPPORTEUR)
ADVOCATE GENERAL JACOBS
25 SEPTEMBER, 13 NOVEMBER 1997, 18 JUNE 1998
European Community Value added tax Right to deduct French rules providing tax not deductible on
certain forms of transport capable of being tool of trade Commission bringing action for failure by
member state to fulfil obligations under relevant Community legislation Whether Community provision
allowing national exclusions which pre-dated Community legislation applying to non-business expenditure
only Council Directive (EEC) 77/388, art 17(2) and (6).

Under art 237 of Annex II to the French Code Gnral des Impts a, value added tax was not deductible
on bicycles, motorcycles, private motor cars, boats, aeroplanes and helicopters; but it was deductible on
commercial vehicles such as vans, lorries and tractors. The European Commission informed the French
government that, since it did not confer the right to deduct value added tax on vehicles used for driving
instruction, the French rule was contrary to art 17(2)b of the subsequent Council Directive (EEC) 77/388
on the harmonisation of the laws of the member states relating to turnover taxescommon system of
value added tax: uniform basis of assessment, which provided that tax was deductible in respect of goods
and services used for the purposes of a persons taxable transactions. Although the French government
amended its legislation in respect of vehicles used for driving instruction, the Commission considered that
the legislation still affected taxable persons whose work necessarily involved certain measures and forms
of transport. The French government contended that such legislation was authorised by art 17(6) c of
Directive 77/388, which provided that, until the Council had adopted specific legislation as to the types of
expenditure which were not tax deductible, member states could retain all the exclusions provided for
under national legislation which pre-dated the directive. That provision also specified that value added tax
should in no circumstances be deductible on expenditure which was not strictly business expenditure.
The Commission subsequently brought an action under art 169 of the EC Treaty for a declaration that, by
maintaining in force legislation which denied taxable persons the right to deduct value added tax on
means of transport which constituted the tool of their trade, France had failed to fulfil its obligations under
Directive 77/388. The Commission contended, inter alia, that the national exclusions authorised by art
17(6) only related to non-business expenditure.

a
Article 237, so far as material is set out at p 258 f, post
b
Article 17(2), so far as material is set out at p 957 j to p 958 a, post
c
Article 17(6), so far as material is set out at p 958 b c, post
 951

Held On a proper construction of art 17(6) of Directive 77/388, the rules which the Council was called
on to adopt were not automatically limited to expenditure which was not strictly business expenditure; and
in those circumstances the expression all the exclusions clearly comprised exclusions for expenditure
which was strictly business expenditure. That provision accordingly authorised the member states to
retain national rules which denied taxable persons the right to deduct VAT on means of transport which
constituted the very tool of their trade. Furthermore, that interpretation was confirmed by the legislative
history of the directive which made it clear that the member states had been unable to agree on the
arrangements applicable specifically to expenditure on passenger transport. It followed that, by
maintaining in force the legislation at issue, the French government had not failed to fulfil it obligations
under art 17(2) of Directive 77/388. The application would therefore be dismissed (see p 959 h to p 960 e,
post).

Notes
For the right to deduct, see 52 Halsburys Laws (4th edn) para 2033.

Case cited
Lennartz v Finanzamt Mnchen III Case C-97/90 [1991] ECR I-3795.

Application
By application lodged at the registry of the Court of Justice of the European Communities on 14 February
1996, the European Commission brought an action under art 169 of the EC Treaty for a declaration that,
by maintaining in force legislation which denied taxable persons the right to deduct value added tax on
means of transport which constituted the very tool of their trade, France had failed to fulfil its obligations
under Council Directive (EEC) 77/388 on the harmonisation of the laws of the member states relating to
turnover taxescommon system of value added tax: uniform basis of assessment, and in particular art
17(2) thereof. The Commission was represented by H Michard and E Traversa, of its Legal Service,
acting as agents, with an address for service in Luxembourg at the office of C Gmez de la Cruz, also of
its Legal Service, Wagner Centre, Kirchberg. The French Republic was represented by C de Salins,
Deputy Director in the Legal Directorate, Ministry of Foreign Affairs, assisted by G Mignot, Foreign Affairs
Secretary in the same ministry, acting as agents, with an address for service in Luxembourg at the French
Embassy, 8B Boulevard Joseph II. The United Kingdom, which intervened in support of France, was
represented by S Ridley, of the Treasury Solicitors Department, acting as agent, with an address for
service in Luxembourg at the British Embassy, 14 Boulevard Roosevelt. The language of the case was
French. The facts are set out in the opinion of the Advocate General.

13 November 1997.

The Advocate General (F G Jacobs)

delivered the following opinion.
1.
In this case the European Commission has brought an action against France under art 169 of the EC
Treaty for a declaration that France is in breach of its obligations under the Council Directive (EEC)
77/388 on the harmonisation of the laws of the member states relating to turnover taxescommon
system of value added tax: uniform basis of assessment (OJ 1977 L145 p 1) (the Sixth Directive) in
particular art 17(2), by not allowing taxable persons under certain circumstances to deduct VAT on means
of transport put to business use.
952

The relevant Community rules

2.
Article 17(2) of the Sixth Directive provides:
In so far as the goods and services are used for the purposes of his taxable transactions, the
taxable person shall be entitled to deduct from the tax which he is liable to pay: (a) value added tax
due or paid in respect of goods or services supplied or to be supplied to him by another taxable
person
3.
However, art 17(6) of the directive provides:
Before a period of four years at the latest has elapsed from the date of entry into force of this
Directive, the Council, acting unanimously on a proposal from the Commission, shall decide what
expenditure shall not be eligible for a deduction of value added tax. Value added tax shall in no
circumstances be deductible on expenditure which is not strictly business expenditure, such as that
on luxuries, amusements or entertainment.
 Until the above rules come into force, Member States may retain all the exclusions provided for
under their national laws when this Directive comes into force.
4.
The EU Council has not as yet taken the decision envisaged by the first sentence of that provision.

The contested French rules

5.
Article 237 of Annex II of the French Code Gnral des Impts (the CGI) provides:
VAT shall not be deductible on vehicles or machines, whatever their nature, designed for the
transport of persons or for mixed use which constitute fixed assets or, if not, are not intended for
resale in a new state.
6.
Subject to certain exceptions the rule covers bicycles, motorcycles, private motor cars, boats,
aeroplanes and helicopters. It does not apply to commercial vehicles such as vans, lorries and tractors.
Helicopters do not qualify for deduction even where they are used for aerial photography, publicity, pilot
training or topographical or geodesic surveysd.
d
See Documentation administrative de base (DB) of the French tax authority, srie 3 CA, division D, feuillets 15321533
(version of 1 May 1990).

7.
The Commission claims that art 237 of Annex II of the CGI, as interpreted and applied by the French
tax authorities, is contrary to art 17(2) of the Sixth Directive in so far as it denies taxable persons the right
to deduct VAT on goods which constitute the very tool or object (loutil ou objet mme) of their trade. By
that term the Commission confines its application to cases where the goods in question are a necessity
to such an extent that they condition in an absolute manner the exercise of the trade itself, as distinct
from cases in which the goods contribute substantially to facilitating the exercise of the trade.
8.
By way of example of goods used as the actual tool of a taxable persons trade the Commission refers
in its application to the case of a firm which uses helicopters for aerial work other than the transport of
passengers. In its letter of formal notice of 6 September 1991 the Commission also referred to the case of
a driving school whose business depended on the use of motor cars for the purpose of giving driving
instruction. However, following an amendment to the French legislation e, permitting deduction of VAT on
motor cars used exclusively for the 953 purposes of driving instruction, the Commission no longer
pursues its complaint in that regard.
e
See art 237 septies A of the CGI, introduced by art 13 of the Law of 26 July 1991.

9.
By contrast the Commissions application does not extend to, for example, cars used by sales
representatives or by veterinary surgeons. Although considerably facilitating the exercise of a trade, such
goods cannot, according to the Commission, be regarded as tools of a trade.
10.
France and the United Kingdom, which has intervened in support of France, contend that the second
sub-paragraph of art 17(6) of the directive expressly permits the maintenance of all the French provisions,
whichas the Commission concedespredate the entry into force of the Sixth Directive. I share that
view.
11.
The Commission contends that France reads the second sub-paragraph of art 17(6) out of context. In
its view the term exclusions in the second sub-paragraph of art 17(6) is a contraction of the expression
expenditure which is not strictly business expenditure in the second sentence of the first sub-paragraph
of that provision. Expenditure on goods which are the tools of a taxable persons trade cannot be
regarded as expenditure which is not strictly business expenditure and therefore cannot be the subject of
exclusions retained under the second sub-paragraph.
12.
However, that analysis is incorrect. The first sentence of the first sub-paragraph of art 17(6) provides
for the adoption by the Council of rules specifying expenditure that shall not be eligible for a deduction of
value added tax. The second sentence of the sub-paragraph adds that VAT shall in no circumstances be
deductible on expenditure which is not strictly business expenditure, such as that on luxuries,
amusements or entertainmentf. It is apparent from the wording and structure of the sub-paragraph that
what was envisaged was the adoption by the Council of comprehensive rules on all the categories of
expenditure which were not eligible for deduction, includingbut not limited toexpenditure that was not
strictly business expenditure.
f
In the light of other language versions it would appear that the term in no circumstances should be understood as
meaning not in any event: see, for example, the Dutch (in elk geval), French (en tout tat de cause), German (auf
jeden Fall) and Italian (comunque) language versions.

13.
 Under the second sub-paragraph of art 17(6) member states are permitted until the above rules come
into force to retain all the exclusions provided for under their national laws when the directive came into
force. Thus the second sub-paragraph is linked not to the second sentence of the first sub-paragraph but
to the first: member states are entitled under the second sub-paragraph to retain all the exclusions in the
areas which are to be the subject of common rules adopted by the Council under the first sentence of art
17(6).
14.
The above analysis is consistent with the legislative history of art 17(6).
In its explanatory memorandum accompanying its proposal for the Sixth Directive g the Commission
explained that:
g
See Bulletin of the European Communities, Supplement 11/73, note 5, p 18.

certain expenditure, even though incurred in the ordinary course of business, is also incurred
for private purposes, and apportionment of such expenditure between business and private
purposes could not be adequately supervised.
15.
Accordingly, art 17(6) of the proposal provided:
954
Value added tax on the following shall not be deductible: (a) expenditure on accommodation,
lodging, restaurants, food, drink, entertainment and passenger transport, unless incurred by an
undertaking whose principal or subsidiary business is the pursuit of such activities; (b) expenditure
on luxuries; (c) entertainment expenditure.
16.
Thus it is clear first that, in drafting the proposal, the Commissions concern was not merely that
certain items of expenditure incurred by taxable persons were not strictly business expenditure but also
that certain expenditure, although incurred in connection with the normal operation of the business, was
difficult to apportion between business and private use h. Secondly, it is apparent from a comparison of the
proposed and adopted texts of art 17(6) that, although, at the moment of the adoption of the Sixth
Directive, member states were in substantial agreement with regard to certain categories of expenditure,
in particular luxuries, amusements and entertainment, no agreement could be reached on the treatment
of passenger transport.
h
See also to the same effect the second recital in the preamble to the proposal for a Twelfth Council Directive on the
harmonisation of the laws of the member states relating to turnover taxes Common system of value added tax:
expenditure not eligible for deduction of value added tax (OJ 1983 C37 p 8).

17.
Further guidance as to what the Commission itself considered to be the scope of the matters upon
which a decision was deferred at the moment of the adoption of the Sixth Directive is provided by the
proposals submitted by the Commission under the first sentence of art 17(6). The Commissions proposal
of 25 January 1983 for a Twelfth Council Directive on the harmonisation of the laws of the member states
relating to turnover taxes Common system of value added tax: expenditure not eligible for deduction of
value added tax (OJ 1983 C37 p 8), in addition to laying down rules disallowing deduction of tax on
transport expenses (art 2), accommodation, food and drink (art 3), entertainment expenditure (art 4) and
amusements and luxuries (art 5), contained detailed provisions on means of transport. Article 1(1) of the
proposal provided that VAT was not to be deductible on the purchase, manufacture, importation, leasing
or hire, use, modification, repair or maintenance of passenger cars, pleasure boats, private aircraft or
motor cycles. Passenger car was defined as
any road vehicle (including any trailer) other than one which, by its design and equipment, is
intended solely for the transport of goods or is intended for industrial or agricultural use or has a
seating capacity of more than nine persons including the driver.
Article 1(2) provided exceptions for vehicles or craft which were: (a) used for carriage for hire or reward;
(b) used for driving training or instruction; (c) hired out; (d) part of the stock in trade of a business.
18.
On 20 February 1984 the Commission presented an amended proposal in which it altered the
treatment of passenger cars and motor cycles (see OJ 1984 C56 p 7). Under a new paragraph 1a,
inserted into art 1, member states were to restrict the right of deduction to a proportion of the VAT on such
goods. In addition it proposed the insertion of a new art 3a providing as follows:
A taxable person may request application of Article 17(2) of Directive 77/388/EEC in respect of
the items of expenditure listed in Articles 1, 2 and 3 955above if he can furnish proof that such
expenditure has been made exclusively for business purposes.
Member States shall maintain in force or introduce arrangements for verifying ex post facto that
such expenditure was indeed made exclusively for business purposes.
19. Thus the Commissions proposals submitted pursuant to the first sentence of art 17(6) (now
withdrawn following continued disagreement in the Council) sought the adoption of common rules on the
restrictions to be imposed on input tax deductions on means of transport in general and on the scope of
the exceptions to be made for certain types of business such as driving schools or car-hire firms or, more
generally, for taxable persons able to show exclusive business use. In other words, they were directed
precisely at the matters which are the subject of the present case.
20.
During the proceedings the Commission has stressed the severe disruption to the VAT system caused
by rules disallowing deduction of tax on passenger vehicles and craft. It argues that France has not
shown why motor cars and other craft are singled out for special treatment and contends that the French
rules are disproportionate to the aim of preventing tax evasion. Private use of such goods could be dealt
with in the normal way, that is to say, either by means of an adjustment of a taxable persons input tax
deduction under arts 17 and 19 or by means of an output tax charge under arts 5(6) and 6(2).
21.
I do not think it is necessary for France in these proceedings to explain the merits of its rules. As I
have explained, it is clear from the text and legislative history of art 17(6), and from the proposals
submitted by the Commission pursuant to the first sentence of that provision, that as yet there are no
common rules on deduction of VAT on passenger vehicles and craft and that, pending the adoption of
such rules, member states are entitled to retain all the exclusions in that area which they applied at the
moment when the Sixth Directive came into force. Article 17(6) does not, as the Commission seems to
suggest, merely confer on member states the power to retain anti-evasion measures whose legality falls
to be tested judicially against the principle of proportionality.
22.
I do not in any event find it altogether surprising that some member states consider that passenger
vehicles and craft merit special treatment. As the Commission itself suggested in the explanatory
memorandum to its proposal for a Sixth Directive, the private use of such goods is particularly difficult to
monitor; moreover, in the case of luxury vehicles the final consumption and business elements may be
indistinguishable. In addition the extremely high value of such goods provides considerable incentive for
undeclared private use. For similar reasons some member states also restrict the deductibility of
expenditure on motor vehicles for the purposes of income and corporation taxes (see eg as regards
France: Fiscal 1996 pp 162163, 309; and, as regards the United Kingdom: Butterworths UK Tax Guide
(199596) pp 479480).
23.
That is not to say that the Councils failure to reach agreement is not regrettable. As the Commission
explained with commendable clarity at the hearing, rules preventing taxable persons from deducting VAT
on such important categories of expenditure severely disrupt the functioning and neutrality of the VAT
system. I doubt moreover whether the risk of tax evasion can justify total exclusion of such goods from
the deduction mechanism.
24.
It is however equally clear that the problem calls for a legislative solution. That point is amply
demonstrated by the weaknesses of the criterion suggested 956 by the Commission in these
proceedings. The mere fact that a passenger vehicle or craft constitutes a tool of the trade does not
remove the possibility of tax evasion or non-taxation of final consumption. Conversely, a vehicle or craft
which is not a tool of the trade may nevertheless be essential for the running of a business, as is shown
by the examples which the Commission itself gives as cases outside the scope of its challenge, namely
motor vehicles used by travelling sales representatives or veterinary surgeons in rural areas. The
perhaps necessarilysimplistic criterion proposed to the Court of Justice of the European Communities
by the Commission in these proceedings may be contrasted with the rather more sophisticated set of
rules which it put forward in its proposal for a Twelfth Directive. It is in any event clear that, as France and
the United Kingdom point out, the criterion proposed here has no basis in the Sixth Directive.
25.
Contrary to the Commissions assertion, I do not think the comments which I made in my opinion in
Lennartz v Finanzamt Mnchen III Case C-97/90 [1991] ECR I-3795 (paras 7879) of are of assistance to
it. As the United Kingdom points out, in that opinion I merely suggested that art 17(6) did not authorise a
general exclusion such as a rule treating goods as being used wholly for private purposes where the
element of business use was very small. Drawing support from the proposal for a Sixth Directive and the
proposal for a Twelfth Directive I concluded that the measures falling within the contemplation of the last
sentence of art 17(6) were those concerning specific categories of goods, such as motor vehicles, whose
use for business purposes was difficult to verify. The French rules in issue here clearly fall within that
category.

Conclusion
26.
Accordingly, I am of the opinion that the court should: (1) dismiss the Commissions application; (2)
order the Commission to pay the costs.

18 June 1998.

The COURT OF JUSTICE (Sixth Chamber)

delivered the following judgment.
1.
 By application lodged at the registry of the Court of Justice of the European Communities on 14
February 1996, the European Commission brought an action under art 169 of the EC Treaty for a
declaration that, by maintaining in force legislation which denies taxable persons the right to deduct value
added tax (VAT) on means of transport which constitute the very tool of their trade, the French Republic
has failed to fulfil its obligations under Council Directive (EEC) 77/388 on the harmonisation of the laws of
the member states relating to turnover taxesCommon system of value added tax: uniform basis of
assessment (OJ 1977 L145 p 1) (the Sixth Directive), and in particular art 17(2) thereof.

The Sixth Directive

2.
Article 17(2) of the Sixth Directive provides as follows:
In so far as the goods and services are used for the purposes of his taxable transactions, the
taxable person shall be entitled to deduct from the tax which he is liable to pay: (a) value added tax
due or paid in respect of goods or services supplied or to be supplied to him by another taxable
person
3.
Article 17(6) provides:
957
Before a period of four years at the latest has elapsed from the date of entry into force of this
Directive, the Council, acting unanimously on a proposal from the Commission, shall decide what
expenditure shall not be eligible for a deduction of value added tax. Value added tax shall in no
circumstances be deductible on expenditure which is not strictly business expenditure, such as that
on luxuries, amusements or entertainment.
Until the above rules come into force, Member States may retain all the exclusions provided for
under their national laws when this Directive comes into force.
4.
On 25 January 1983 the Commission submitted to the EC Council a Proposal for a Twelfth Directive
on the harmonisation of the laws of the member states relating to turnover taxescommon system of
value added tax: expenditure not eligible for deduction of value added tax (OJ 1983 C37 p 8), which was
amended by another proposal submitted by the Commission to the Council on 20 February 1984 (OJ
1984 C56 p 7). That proposal was not adopted by the Council.

The national legislation

5.
The French rule at issue is art 237 of Annex II to the French Code Gnral des Impts (the CGI),
which entered into force on 27 July 1967 and which provides as follows:
Value added tax shall not be deductible on vehicles or machines, whatever their nature,
designed for the transport of persons or for mixed use which constitute fixed assets or, if not, are
not intended for resale in a new state.
6.
The basic documentation of the French tax authority (Srie 3 C A, Division D, feuillets 1532 and 1533,
updated on 1 May 1990) states that the vehicles covered by that provision are bicycles, motorcycles,
private motor cars, boats, aeroplanes and helicopters. However, the aforesaid rule does not apply to
commercial vehicles such as vans, lorries, tractors and other highly specialised vehicles. Furthermore,
helicopters are not eligible for deduction even where they are used for aerial photography, publicity, pilot
training, or topographical or geodesic surveys.

The pre-litigation procedure

7.
By letter of 6 September 1991 the Commission informed the French Republic that it regarded art 237
of Annex II to the CGI as incompatible with the Sixth Directive and, in particular, art 17(2) thereof, in so far
as it does not confer the right to deduct VAT on vehicles used for the purposes of driving instruction.
8.
By letter of 6 September 1991 the French government informed the Commission that the contested
provision had been amended, with effect from 1 January 1993, in such a way as to render vehicles or
equipment used exclusively for the purposes of driving instruction eligible for deduction.
9.
By letter of 12 July 1993 the Commission informed the French government that both the condition that
a vehicle must be used exclusively for the purposes of driving instruction and the exclusion of the right to
deduct which continued to affect taxable persons whose work by its very nature involved the use of
certain means and forms of transport (such as helicopters used for lifting by an undertaking engaged in
aerial work) were in breach of art 17(2) of the Sixth Directive.
958
10.
By letter of 4 October 1993 the French government replied that the condition that a driving-school
vehicle must be used exclusively for the purposes of driving instruction had been relaxed by
administrative order of 4 February 1993. It also pointed out that the exclusion of the right to deduct VAT
on means of transport which constitute the very object of a taxable persons trade was authorised by art
17(6) of the Sixth Directive.
11.
In the case of vehicles used for the purposes of driving instruction, the Commission decided to
discontinue the procedure. However, taking the view that the principle of the right to deduct VAT on a
means of transport which constitutes the very object of a taxable persons trade was fundamental, the
Commission issued a reasoned opinion to the French Republic on 8 November 1994, requesting it to take
the requisite measures within a period of two months.
12.
In its reply of 9 January 1995, the French government expressed its total disagreement with the
Commissions analysis and set out in more detail the observations it had formulated in its reply to the
letter of formal notice.

The application
13.
In support of its application, the Commission claims that the exclusion, provided for by art 237 of
Annex II to the CGI, of the right to deduct VAT on goods which constitute the very tool or object of a
taxable persons trade is contrary to art 17(2) of the Sixth Directive.
14.
Admittedly, the second sub-paragraph of art 17(6) of the directive expressly authorises member states
to retain provisions excluding the right to deduct which, like art 237 of Annex II to the CGI, predate the
entry into force of the Sixth Directive.
15.
According to the Commission, however, the exclusion of the right to deduct provided for by art 17(6) of
the Sixth Directive relates only to expenditure which is not strictly business expenditure. Thus, the only
expenditure liable to be excluded from the right to deduct is that incurred by a taxable person on goods
and services which are not absolutely essential for the operation of his business. That possibility is
designed to prevent a taxable person from being able to obtain for his own final use goods and services
which have not been taxed.
16.
That interpretation cannot be accepted, as it is not consistent with the wording of art 17(6) of the Sixth
Directive.
17.
The first sub-paragraph of art 17(6) of the Sixth Directive provides that the Council is to decide what
expenditure is not eligible for a deduction of VAT. The next sentence states that value added tax shall in
no circumstances be deductible on expenditure which is not strictly business expenditure. It follows, in
particular, from that second sentence that the rules which the Council is called upon to adopt are not
automatically limited to expenditure which is not strictly business expenditure.
18.
In those circumstances, the expression all the exclusions, used in the second sub-paragraph of art
17(6), clearly comprises expenditure which is strictly business expenditure. That provision accordingly
authorises the member states to retain national rules which deny taxable persons the right to deduct VAT
on means of transport which constitute the very tool of their trade.
19.
As the Advocate General has pointed out in paras 14 to 16 of his opinion, that interpretation is
confirmed by the origin of art 17(6) of the Sixth Directive. In the first place, in the explanatory
memorandum accompanying its proposal for the Sixth Directive (see Bulletin of the European
Communities Supplement 11/73 p 1), the Commission stated that certain expenditure, even though
incurred in 959 the ordinary course of the undertakings business, would be difficult to apportion between
business use and private use. Secondly, it is clear from a comparison of the wording of art 17(6) proposed
by the Commission and that adopted by the Council that, when the Sixth Directive was adopted, the
member states were unable to agree on the arrangements applicable specifically to expenditure on
passenger transport.
20.
In the light of those considerations, it is apparent that, by maintaining in force legislation which denies
taxable persons the right to deduct VAT on means of transport which constitute the very tool of their trade,
the French Republic has not failed to fulfil its obligations under the Sixth Directive, and in particular art
17(2) thereof. The application for a declaration that it has failed to fulfil its obligations must therefore be
dismissed as unfounded.

Costs
21.
Under art 69(2) of the Rules of Procedure, the unsuccessful party is to be ordered to pay the costs.
Since the Commission has been unsuccessful, it must be ordered to pay the costs. Under art 69(4) of the
Rules of Procedure, member states and institutions which intervene in the proceedings are to bear their
own costs.
On those grounds, the Court of Justice (Sixth Chamber) hereby: (1) Dismisses the application; (2)
 Order the European Commission to pay the costs; (3) Orders the United Kingdom to bear its own
costs.

960