The European Trust Crisis

The European Trust Crisis
Yann Algan, Sergei Guriev, Elias Papaioannou, and Evgenia Passari

(Sciences Po and CERR; Sciences Po, EBRD, and CEPR; LBS and CEPR; Paris-Dauphine)
February 2017
Introduction
Motivation
Global rise of extremism and populism

Roots? [globalization, recession, inequality, immigration, etc.]
European crisis
Deep recession in the periphery (output contraction, rise in unemployment)
Animosity between core and periphery (historical roots?)
BREXIT
Trust and economic performance
Mechanisms [trust and institutions]
Drivers of trust
2
Introduction
This Paper.
Unemployment, Trust in Institutions and Extremism
Descriptive patterns (regional variation before-after the crisis)
Voting for extremist/nationalistic parties
Trust and beliefs; emphasis on institutions (EU, political parties, courts,
police) and general trust (focus of literature so far)
Unemployment and voting for extremist-populist parties
Simple (within region across time) correlations; ols.
Component of unemployment stemming from global aspects of the crisis
(industrial specialization, construction sector). 2sls.
European Trust Crisis. Unemployment and Beliefs-Trust
Trust (general, institutions); beliefs (towards immigration). ols
Heterogeneity (w.r.t. gender, age, demographics)
Industrial-specialization driven component of unemployment. 2sls.
3
Introduction
Structure
1. Related Literature
2. Data [voting, trust/beliefs, regional accounts]
3. Descriptive Before-After Analysis [voting, trust, beliefs]
4. Unemployment and Voting
OLS estimates [graphical illustration]
2sls estimates [first-stage, reduced-forms, 2sls]
5. Unemployment and the European Trust Crisis
OLS estimates [graphical illustrations]
Heterogeneity
6. Conclusion. Next Steps
4
Related Literature
Structure
Heterogeneity
5
Related Literature
Related Literature
1. Extremism/populism
2. European crisis
3. Trust and economic performance
6
Related Literature
Populism/Extremism
Empirics
Dornbusch and Edwards (1991). Latin America
Becker et al. (2016). BREXIT; correlates across regions
Autor et al. (2016, 2017). US; Global competition (from China)
Dippel et al. (2017). Germany; import competition
Colantone and Stanig (2017). European regions; Global competition
Theory
Acemoglu, Egorov, and Sonin (2013)
Guiso, Herrera, Morelli, and Sonno (2017).
Di Tella and Rotenberg (2016)
Theory and Empirics
Guiso, Herrera, Morelli, and Sonno (2017).
7
Related Literature
European Crisis. Aspects
Economics/finance
macro policies; banks; current account deficits; fiscal conditions;
coordination failures. [vast literature]
Politics of Currency Union [Alesina, Tabellini, and Trebbi (2017); Spolaore
(2013]
Institutional Divergence [Papaioannou (2016)]
Beliefs. [Guiso, Sapienza, and Zingales (2016); Guiso, Morelli, and Herrera
(2016)]
8
Related Literature
Trust and Economic Performance
Economic development/growth. [Tabellini, 2010; Algan and Cahuc, 2008;

Knack and Keefer, 1998]
Institutions and corruption
Financial development
Trade and foreign investment
Labour market institutions
Organizational structure
Determinants
Long-run factors; historical episodes [Nunn and Wantchekon 2011; Guiso,
Sapienza, and Zingales, 2016; Putnam, 1994]
Short-run; economic downturn [Ananiev and Guriev, 2016; Putnam, 2001 ]
Reviews. Algan and Cahuc (2013); Guiso, Sapienza and Zingales (2011); Durlauf
and Fafchamps (2005); Fernandez (2014)
9
Data
Structure
Heterogeneity
10
Data
Data. EU Regions (NUTS)
Regional output, employment, and industry composition. EUROSTAT

27 Countries
244 Regions
General trust, trust towards institutions and beliefs. European Social
Surveys (ESS).
24 countries
183 regions
Electoral Data
National elections; country-specific sources
Classification of political parties orientation
11
Data
Countries and Regions
List of Countries
Austria (6), Belgium (3), Bulgaria (6), Cyprus (1), Czech Republic (8), Denmark
(5), Estonia (1), Finland (5), France (8), Germany (16), Greece (8), Hungary (6),
Iceland (1), Ireland (2), Italy (18), Netherlands (12), Norway (7), Poland (16),
Portugal (5), Romania (8), Slovakia (4), Slovenia (2), Spain (17), Sweden (8),
Switzerland (5), UK (12).
Country Groups
Centre: Austria, Belgium, France, Germany, Luxembourg, Netherlands,
Switzerland.
South: Cyprus, Greece, Italy, Portugal, Spain.
North: Denmark. Ireland, Iceland, Norway, Sweden, UK, Finland.
Former Transition: Bulgaria, Czech Republic, Estonia, Hungary, Poland,
Slovakia, Slovenia, Romania.
12
Descriptive Analysis
Structure
Heterogeneity
13
Descriptive Patterns across European Regions
Unemployment and output

Voting for extremist/populist parties
General Trust [focus of literature]
Trust towards Institutions [EU, local political parties, courts, police]
Role of Immigrants [economy, cultural life]
14
Unemployment. [Source. Eurostat]
15
Regional Unemployment and Output. 2000-2014.
16
Recession Intensity.
Regional Unemployment and Output Before and after the Crisis
17
Voting for Extreme Parties. [Source. Country-Specific Electoral Datasets]
18
Trust towards the European Union. [Source. ESS]
19
Trust in National Parliament. [Source. ESS]
20
Trust in the Legal System. [Source. ESS]
21
Trust towards the Police. [Source. ESS]
22
General Trust. [Source. ESS]
23
Role of Immigrants in the Economy. [Source. ESS]
24
Role of Immigrants in Countrys Cultural Life. [Source. ESS]
25
Unemployment and Voting
Structure
Heterogeneity
26
Graphical analysis
OLS difference-in-difference specifications
Further Evidence. BREXIT analysis (out-of-sample prediction)
Towards causation. Identification idea
First-stage relationship
Reduced-form relationship
2sls estimates
27
Unemployment and Extremist Voting

within European Regions [2000-2015]
28
Unemployment and Extremist Voting.

Before-After the Crisis
29
BREXIT Digression.
Changes in Unemployment (2014-2007) and Leave Vote
30
BREXIT Digression, cont.

Unemployment (2014) and Vote to Leave the EU
31
Towards Causation. Issues
Reverse causation [unlikely, but cannot be ruled out]

Measurement error [present esp. with regional output]
Omitted variables
Time-invariant [region fixed-effects]
General time-varying [year/period, groupXyear fixed-effects]
Region time-varying [cannot be ruled out]
32
Identification Idea
Isolate global, unrelated to European regions cycle, component of

economic downturn during the crisis (2008-2015)
Industrial composition and differential industry global shocks [Ananyev and

Guriev, 2016]
Role of construction [e.g., United States, United Kingdom, Spain, Ireland,
Greece, Portugal]
Semi-exogenous variation (instrument)

Share of construction
33
Identifying Assumptions
First-stage fit. Regional employment (and output) shock related to

Share of construction before the crisis
Exclusion Restriction. Industrial specialisation, share of construction, before the

crisis affects voting (and trust/beliefs) only via affecting the intensity of the crisis
(rise in unemployment, fall of output)
Reasonable (?)
Other channels (??)
34
First-Stage. Construction Share before the Crisis and

Changes in Unemployment (2015-2008)
35
First-Stage, cont. Construction Share before the Crisis

and Changes in Unemployment, cond. for sectoral shares
36
Reduced-Form Estimates. Construction Share before

the Crisis and Changes in Extreme Voting
37
Reduced-Form Estimates, cont. Construction and

Changes in Extreme Voting, cond. for sectoral shares
38
2sls. Initial Construction Share, Changes in

Unemployment and Changes in Extremist Voting
Initial (before Crisis) Construction Share, Changes in Unemployment, and Changes in
Extremist Voting
(1) (2) (3)
Panel A. 2SLS Estimates
Change in Unemployment [Before-After] 1.689*** 2.4475*** 2.1844***

[standard error] (0.2135) (0.2005) (0 .3351)
Panel B. "Reduced-Form" Estimates
Before Crisis Construction Share (2005-2007) 3.1956*** 3.5445*** 3.5445***

[standard error] (0.4968) (0.7841) (0.7841)
overall R-squared 0.150 0.361 0.361

Panel C. "First-Stage" Estimates
Before Crisis Construction Share (2005-2007) 1.8925*** 1.7279*** 1.6226***

[standard error] (0.1549) (0.1963) (0.2032)
First Stage F-score 149.35 42.30 33.57

p-value 0.00 0.00 0.00
Panel D. OLS Estimates
Change in Unemployment [Before-After] 1.5024*** 1.8969*** 1.5715***

[standard error] (0.1146) (0.1899) (0.1717)
First Stage F-score 149.35 42.30 33.57
Observations 162 162 162

Countries
Robust standard errors in parentheses, adjusted for 4 clusters in group of countries
*** p<0.01, ** p<0.05, * p<0.1
39
Unemployment and Trust/Beliefs
Structure
Heterogeneity
40
Unemployment and the European Trust Crisis
Structure
Graphical analysis. Before-After Crisis
OLS difference-in-difference specifications. All regions; all years
Difference specifications. Before-after Crisis
2sls difference specifications. Before-after Crisis
Aspects
Trust in the EU [comparison with trust in the UN]
Trust in National Political Parties
Trust in Courts [comparison with trust in Police]
General Trust
Immigrants Economic Role of [comparison with cultural life role]
41
Unemployment and Distrust in the European Parliament
42
Unemployment and Distrust in National Institutions
43
Unemployment and Distrust in National Courts
44
Unemployment and Distrust in the Police
45
Unemployment and General Trust
46
Unemployment and Role of Immigration in the Economy
47
Unemployment and Role of Immigration in Cultural Life
48
Trust, Immigration. OLS Estimates.

Full Sample of European Regions (2000-2015)
Trust Trust Nat. Trust EU Trust United Trust Nat. Trust Legal Immigration
People Parliament Parliament Nations Police System Economy Cult. Life
(1) (2) (3) (4) (5) (6) (7) (8)
Unemployment -0.15 -0.74*** -0.39*** -0.08 -0.12 -0.48*** -0.63*** -0.10

[standard error] (0.098) (0.157) (01279) (0.1039) (0.0984) (0.0874) (0.0876) (0.0758)
Within R2 0.015 0.15 0.052 0.003 0.008 0.083 0.19 0.006

Overall R2 0.872 0.812 0.547 0.726 0.833 0.844 0.772 0.839
Obsrevations 1039 1039 1039 1039 1039 1039

Region (NUTS) 187 187 187 187 187 187 187 187
Countries 24 24 24 24 24 24 24 24
Region Fived-Effects Yes Yes Yes Yes Yes Yes Yes Yes
ESS Round Fixed-Effects Yes Yes Yes Yes Yes Yes Yes Yes
49
Trust. 2sls Estimates. Before-After the Crisis

Change (before-after Trust Trust Nat. Trust EU Trust United Trust Nat. Trust Legal
crisis) in People Parliament Parliament Nations Police System
(1) (2) (3) (4) (5) (6)
Change Unemployment -0.1234 -1.1022*** -0.4778*** 0.0879 0.2099** -0.6294***

[standard error] (0.1137) (0.0831) (0.0903) (0.0906) (0.0832) (0.1022)
Share of construction -0.2158 -2.0522** -0.8896*** 0.1637 0.3908 -1.1718**

[standard error] (0.1823) (0.2798) (0.0500) (0.2326) (0.1717) (0.3502)
Panel C. OLS Estimates
Change Unemployment -0.1240** -0.8031** -0.3104 0.0240 -0.0226 -0.4630*

[standard error] (0.0131) (0.1855) (0.2329) (0.1212) (0.0244) (0.1946)
First-Stage F-score 38.79 38.79 38.79 38.79 38.79 38.79
Initial Industry Shares Yes Yes Yes Yes Yes Yes

Obsrevations/Regions 122 122 122 122 122 122
Countries 24 24 24 24 24 24
50

the Crisis and Changes in Trust in European Parliament
51

the Crisis and Changes in Trust in Nat. Parliament
52

the Crisis and Changes in Trust in Legal System
53
Immigration. 2sls Estimates. Before-After the Crisis

Initial (before Crisis) Construction Share, Changes in Unemployment,
and Changes in Attitudes towards Immigration
(1) (2)
Immigration Immigration
Economy Cult. Life Economy Cult. Life
(1) (2) (1) (2)
Unemployment -0.5077** 0.0229 -0.8824** -0.1992

[standard error] (0.2149) (0.2178) (0.4204) (0.3039)
Share of construction -0.9453** 0 .0426 -1.5519** -0.3503

[standard error] (0.2769) (0.4867) (0.4333) (0.5123)
Panel C. OLS Estimates
Unemployment -0.3842* 0.0868 -0.5198* 0.014

[standard error] (0.1266) (0.1360) (0.1808) (0.1064)
First-Stage F-score 38.79 38.79 93.67 93.67
Obsrevations/Regions 122 122 122 122

Countries 24 24 24 24
54
Conclusion
Structure
Heterogeneity
55
Conclusion
Summary
Link between (changes in) unemployment and (rise in) extremist voting
Correlation across European regions
Component of rise in unemployment explained by pre-crisis share of
construction
Unemployment and Distrust

EU (but not the UN); national politics/parliament; legal system and courts
(but not the police)
No impact in general trust (perhaps due to opposing effects)
Role of immigrants in the economy (as compared to social life)
OLS; 2SLS (share of construction before the crisis as an instrument for
changes in unemployment)
56
Conclusion
Next Steps
Conduct the analysis at finer geographic unit (NUTS3).

Account better for unobservable features and regional heterogeneity.
Fine-tune parties political orientation
Look at EU parliament elections
Examine jointly the role of the recession and immigration
Decompose opposing effects of unemployment on trust
57

The European Trust Crisis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The European Trust Crisis

Uploaded by

Copyright:

Available Formats

The European Trust Crisis

Yann Algan, Sergei Guriev, Elias Papaioannou, and Evgenia Passari

Global rise of extremism and populism

European Crisis. Aspects

Trust and Economic Performance

Economic development/growth. [Tabellini, 2010; Algan and Cahuc, 2008;

Data. EU Regions (NUTS)

Regional output, employment, and industry composition. EUROSTAT

Countries and Regions

Descriptive Patterns across European Regions

Unemployment and output

Unemployment. [Source. Eurostat]

Regional Unemployment and Output. 2000-2014.

Voting for Extreme Parties. [Source. Country-Specific Electoral Datasets]

Trust towards the European Union. [Source. ESS]

Trust in National Parliament. [Source. ESS]

Trust in the Legal System. [Source. ESS]

Trust towards the Police. [Source. ESS]

General Trust. [Source. ESS]

Role of Immigrants in the Economy. [Source. ESS]

Role of Immigrants in Countrys Cultural Life. [Source. ESS]

Unemployment and Voting

Unemployment and Extremist Voting

Unemployment and Extremist Voting.

BREXIT Digression, cont.

Towards Causation. Issues

Reverse causation [unlikely, but cannot be ruled out]

Isolate global, unrelated to European regions cycle, component of

Industrial composition and differential industry global shocks [Ananyev and

Semi-exogenous variation (instrument)

First-stage fit. Regional employment (and output) shock related to

Exclusion Restriction. Industrial specialisation, share of construction, before the

First-Stage. Construction Share before the Crisis and

First-Stage, cont. Construction Share before the Crisis

Reduced-Form Estimates. Construction Share before

Reduced-Form Estimates, cont. Construction and

2sls. Initial Construction Share, Changes in

(1) (2) (3)

Panel A. 2SLS Estimates

Change in Unemployment [Before-After] 1.689*** 2.4475*** 2.1844***

Panel B. "Reduced-Form" Estimates

Before Crisis Construction Share (2005-2007) 3.1956*** 3.5445*** 3.5445***

overall R-squared 0.150 0.361 0.361

Before Crisis Construction Share (2005-2007) 1.8925*** 1.7279*** 1.6226***

First Stage F-score 149.35 42.30 33.57

Change in Unemployment [Before-After] 1.5024*** 1.8969*** 1.5715***

First Stage F-score 149.35 42.30 33.57

Observations 162 162 162

Unemployment and the European Trust Crisis

Unemployment and Distrust in the European Parliament

Unemployment and Distrust in National Institutions

Unemployment and Distrust in National Courts

Unemployment and Distrust in the Police

Unemployment and General Trust

Unemployment and Role of Immigration in the Economy

Unemployment and Role of Immigration in Cultural Life

Trust, Immigration. OLS Estimates.

Unemployment -0.15 -0.74*** -0.39*** -0.08 -0.12 -0.48*** -0.63*** -0.10

Within R2 0.015 0.15 0.052 0.003 0.008 0.083 0.19 0.006

Obsrevations 1039 1039 1039 1039 1039 1039

Trust. 2sls Estimates. Before-After the Crisis

Panel A. 2SLS Estimates

Change Unemployment -0.1234 -1.1022*** -0.4778*** 0.0879 0.2099** -0.6294***

Panel B. "Reduced-Form" Estimates

Change in Unemployment [Before-After] 1.689* 2.4475* 2.1844***

Before Crisis Construction Share (2005-2007) 3.1956* 3.5445* 3.5445***

Before Crisis Construction Share (2005-2007) 1.8925* 1.7279* 1.6226***

Change in Unemployment [Before-After] 1.5024* 1.8969* 1.5715***

Unemployment -0.15 -0.74* -0.39* -0.08 -0.12 -0.48* -0.63* -0.10

Change Unemployment -0.1234 -1.1022* -0.4778* 0.0879 0.2099 -0.6294*

Share of construction -0.2158 -2.0522 -0.8896* 0.1637 0.3908 -1.1718**

Change Unemployment -0.1240 -0.8031 -0.3104 0.0240 -0.0226 -0.4630*

Unemployment -0.5077 0.0229 -0.8824 -0.1992

Share of construction -0.9453 0 .0426 -1.5519 -0.3503