Putnamdiplomacy and Doestic Politicspdf

Diplomacy and Domestic Politics: The Logic of Two-Level Games
Author(s): Robert D. Putnam

Source: International Organization, Vol. 42, No. 3 (Summer, 1988), pp. 427-460
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/2706785
Accessed: 10-06-2015 17:31 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to International Organization.
http://www.jstor.org
This content downloaded from 130.64.11.153 on Wed, 10 Jun 2015 17:31:18 UTC
All use subject to JSTOR Terms and Conditions
Diplomacyand domesticpolitics:
thelogicof two-levelgames
RobertD. Putnam
of domesticand
Introduction:the entanglements
internationalpolitics
Domesticpoliticsand international relationsare oftensomehowentangled,

butour theories have not yet sorted out thepuzzlingtangle.It is fruitless
to debatewhether domestic politicsreallydetermine internationalrelations,
or thereverse. The answer to that question is clearly "Both, sometimes."
Themoreinteresting questionsare"When?"and"How?" Thisarticleoffers
a theoreticalapproach to thisissue,butI beginwitha storythatillustrates
thepuzzle.
One illuminating exampleof how diplomacyand domesticpoliticscan
becomeentangled culminated at theBonn summitconference of 1978.1In
the mid-1970s, a coordinatedprogramof globalreflation, led by the "lo-
comotive"economiesoftheUnitedStates,Germany, and Japan,had been
proposed to fosterWestern recovery from the first
oil shock.2Thisproposal
An earlierversionofthisarticlewas deliveredat the1986annualmeeting oftheAmerican

PoliticalScienceAssociation.Forcriticisms
andsuggestions,I amindebtedtoRobertAxelrod,
NicholasBayne,HenryBrady,JamesA. Caporaso,BarbaraCrane,ErnstB. Haas, Stephan
Haggard,C. RandallHenning, PeterB. Kenen,Robert0. Keohane,StephenD. Krasner, Jacek
Kugler,Lisa Martin, JohnOdell,RobertPowell,KennethA. Shepsle,StevenStedman,Peter
Yu, members of researchseminarsat theUniversities
of Iowa, Michigan,and Harvard,and
two anonymous reviewers.I am gratefulto theRockefeller Foundationforenablingme to
completethisresearch.
1. The following accountis drawnfromRobertD. Putnamand C. RandallHenning,"The
BonnSummit of1978:HowDoes International EconomicPolicyCoordination Actually
Work?"
BrookingsDiscussion Papers inInternationalEconomics, no. 53 (Washington,D.C.: Brookings
Institution,
October1986),and RobertD. Putnamand NicholasBayne,HangingTogether:
Cooperation and Conflictin the Seven-Power Summits,rev. ed. (Cambridge,Mass.: Harvard
UniversityPress,1987),pp. 62-94.
2. Amonginterdependent economies,mosteconomists
believe,policiescan oftenbe more
iftheyareinternationally
effective coordinated.
Forrelevant see PutnamandBayne,
citations,
Hanging Together,p. 24.
InternationalOrganization42, 3, Summer 1988

andtheMassachusetts
?D1988bytheWorldPeace Foundation ofTechnology
Institute
428 International
Organization
had receiveda powerful boostfromtheincoming Carteradministration and

was warmlysupported bytheweakercountries, as wellas theOrganization
forEconomicCo-operationand Development(OECD) and manyprivate
economists, whoarguedthatitwouldovercomeinternational payments im-
balancesand speedgrowth all around.On theotherhand,theGermansand
theJapaneseprotested thatprudent andsuccessful economicmanagers should
notbe askedto bail outspendthrifts. Meanwhile,Jimmy Carter'sambitious
NationalEnergyProgramremaineddeadlockedin Congress,whileHelmut
Schmidtled a chorusofcomplaints abouttheAmericans'uncontrolled ap-
petiteforimported oil andtheirapparentunconcern aboutthefalling dollar.
All sides concededthatthe worldeconomywas in serioustrouble,butit
was notclearwhichwas moreto blame,tight-fisted Germanand Japanese
fiscalpoliciesor slack-jawedU.S. energyand monetary policies.
At the Bonn summit,however,a comprehensive packagedeal was ap-
proved,theclearestcase yetof a summitthatleftall participants happier
thanwhentheyarrived.HelmutSchmidt agreedtoadditional fiscalstimulus,
amounting to 1 percentofGNP, Jimmy Cartercommitted himself to decon-
troldomesticoil pricesbytheendof 1980,andTakeo Fukudapledgednew
effortsto reacha 7 percentgrowthrate.Secondaryelementsin theBonn
accordincludedFrenchand BritishacquiescenceintheTokyoRoundtrade
negotiations; Japaneseundertakings to fosterimportgrowthand restrain
exports;and a genericAmericanpromiseto fightinflation. All in all, the
Bonn summitproduceda balancedagreement of unparalleled breadthand
specificity.Moreremarkably, virtually all partsofthepackagewereactually
implemented.
Most observersat the timewelcomedthe policiesagreedto at Bonn,
althoughin retrospecttherehas been muchdebate about the economic
wisdomofthispackagedeal. However,myconcernhereis notwhether the
deal was wise economically, but how it became possiblepolitically.My
researchsuggests,first, thatthekeygovernments at Bonnadoptedpolicies
differentfromthosethattheywouldhave pursuedin theabsenceof inter-
nationalnegotiations, butsecond,thatagreement was possibleonlybecause
a powerful minority withineach government actuallyfavoredon domestic
groundsthepolicybeingdemandedinternationally.
WithinGermany,a politicalprocesscatalyzedby foreign pressureswas
surreptitiously orchestrated by expansionists insidethe Schmidtgovern-
ment.Contrary to thepublicmythology, theBonndeal was notforcedon
a reluctant or "altruistic"Germany.In fact,officials in the Chancellor's
OfficeandtheEconomicsMinistry, as wellas intheSocialDemocraticparty
andthetradeunions,hadarguedprivately inearly1978thatfurther stimulus
was domestically desirable,particularly in view of the approaching1980
elections.However,theyhad littlehope of overcoming theoppositionof
theFinanceMinistry, the Free Democraticparty(partof thegovernment
coalition),and thebusinessand bankingcommunity, especiallytheleader-
Diplomacyand domesticpolitics 429
shipoftheBundesbank.Publicly,HelmutSchmidtposedas reluctant to the

end. Onlyhisclosestadvisorssuspectedthetruth:thatthechancellor"let
himself be pushed"intoa policythathe privately favored,butwouldhave
foundcostlyandperhapsimpossibleto enactwithout thesummit's package
deal.
Analogously,in Japana coalitionof businessinterests, the Ministry of
Tradeand Industry (MITI), theEconomicPlanningAgency,and someex-
pansion-minded politicianswithintheLiberalDemocraticPartypushedfor
additionaldomesticstimulus,usingU.S. pressureas one of theirprime
arguments againstthestubborn resistance oftheMinistry ofFinance(MOF).
Without internaldivisionsin Tokyo,it is unlikelythattheforeign demands
wouldhave been met,but withoutthe externalpressure,it is even more
unlikelythattheexpansionists could have overridden thepowerful MOF.
"Seventypercentforeignpressure,30 percentinternalpolitics,"was the
disgruntled judgmentofone MOF insider."Fifty-fifty," guessedan official
fromMITI.3
In theAmericancase, too, internal politicking reinforced, and was rein-
forcedby,theinternational pressure.Duringthesummit preparations Amer-
icannegotiators occasionallyinvitedtheirforeign counterparts to putmore
pressureon the Americansto reduceoil imports.Key economicofficials
withinthe administration favoreda tougherenergypolicy,buttheywere
opposedby the president'sclosestpoliticalaides, even afterthe summit.
Moreover,congressional opponentscontinued to stymie oil pricedecontrol,
as theyhadunderbothNixonandFord.Finally,inApril1979,thepresident
decidedongradualadministrative decontrol, bringingU.S. pricesuptoworld
levelsby October1981.His domesticadvisorsthuswona postponement of
thispoliticallycostlymoveuntilafterthe 1980presidential election,butin
theend,virtually everyone ofthepledgesmadeat Bonnwas fulfilled. Both
proponents and opponentsofdecontrolagreethatthesummit commitment
was at the centerof the administration's heatedintramural debateduring
thewinterof 1978-79and instrumental in thefinaldecision.4
In short,the Bonn accord represented genuineinternational policyco-
ordination.Significantpolicychangeswerepledgedandimplemented bythe
key participants. Moreover-althoughthiscounterfactual claimis neces-
sarilyharderto establish-thosepolicychangeswouldveryprobablynot
have been pursued(certainly notthesame scale and withinthesame time
frame)in theabsenceoftheinternational agreement. Withineach country,
one factionsupported thepolicyshiftbeingdemandedofitscountry inter-
3. Fora comprehensive
accountoftheJapanesestory,see I. M. DestlerandHisao Mitsuyu,
"Locomotiveson Different
Tracks:Macroeconomic Diplomacy,1977-1979,"inI. M. Destler
and Hideo Sato, eds., Coping with U.S.-Japanese Economic Conflicts(Lexington, Mass.:
Heath,1982).
4. Foran excellentaccountofU.S. energypolicyduring
thisperiod,see G. JohnIkenberry,
"MarketSolutionsforStateProblems:The Internationaland DomesticPoliticsofAmerican
Oil Decontrol," InternationalOrganization42 (Winter1988).
430 International
Organization
nationally,but thatfactionwas initiallyoutnumbered. Thus,international

pressurewas a necessaryconditionforthesepolicyshifts.On the other
hand,without forceswouldnothave suf-
domesticresonance,international
ficedto producetheaccord,no matter howbalancedand intellectuallyper-
suasivetheoverallpackage.In theend,each leaderbelievedthatwhathe
was doingwas in his nation'sinterest-andprobablyin his own political
interest,too,eventhoughnotall hisaides agreed.5Yet without thesummit
accordhe probablywouldnot(orcouldnot)havechangedpoliciesso easily.
In thatsense,theBonndeal successfully mesheddomesticandinternational
pressures.
Neithera purelydomesticnora purelyinternationalanalysiscouldaccount
forthisepisode.Interpretationscast intermseitherofdomesticcauses and
internationaleffects("Second Image"6)or of internationalcauses and do-
mesticeffects ("Second ImageReversed"7)wouldrepresent merely"partial
equilibrium" analysesandwouldmissan important partofthestory,namely,
how the domesticpoliticsof severalcountriesbecame entangledvia an
internationalnegotiation.The eventsof 1978illustrate thatwe mustaim
insteadfor"generalequilibrium" theoriesthataccountsimultaneously for
theinteraction ofdomesticand internationalfactors.Thisarticlesuggestsa
conceptualframework for understanding how diplomacyand domestic
politicsinteract.
Domestic-international
entanglements:
thestateoftheart
Muchof theexistingliterature on relationsbetweendomesticand interna-
tionalaffairsconsistseitherofad hoclistsofcountless"domesticinfluences"
on foreign policyor ofgenericobservations thatnationaland international
affairsare somehow"linked.'8 JamesRosenauwas one ofthefirst scholars
tocallattentiontothisarea,buthiselaboratetaxonomy of"linkagepolitics"
generatedlittlecumulative research,exceptfora flurry ofworkcorrelating
domesticand international "conflictbehavior."9
A second streamof relevanttheorizing began withthe workby Karl
5. It is notclearwhetherJimmyCarterfullyunderstood
thedomesticimplications
of his
Bonn pledgeat the time.See Putnamand Henning,"The Bonn Summit,"and Ikenberry,
"MarketSolutionsforStateProblems."
6. KennethN. Waltz, Man, theState, and War:A TheoreticalAnalysis(New York: Columbia
University Press,1959).
7. PeterGourevitch, "The SecondImageReversed:The International
SourcesofDomestic
Politics,"International 32 (Autumn1978),pp. 881-911.
Organization
8. I am indebtedto StephanHaggardforenlightening
discussionsaboutdomesticinfluences
on international
relations.
9. JamesRosenau,"TowardtheStudyofNational-International Linkages,"in his-
Linkage
Politics: Essays on the Convergenceof National and InternationalSystems(New York: Free
Press,1969),as wellas his "Theorizing
AcrossSystems:LinkagePoliticsRevisited,"inJon-
athanWilkenfeld,ed., ConflictBehavior and Linkage Politics (New York: David McKay, 1973),
especiallyp. 49.
Deutschand ErnstHaas on regionalintegration. 10Haas, in particular, em-

phasizedtheimpactofpartiesandinterest groupsontheprocessofEuropean
integration,and hisnotionof "spillover"recognizedthefeedbackbetween
domesticand international developments. However,thecentraldependent
variablein thisworkwas thehypothesized evolutionofnew supranational
institutions,ratherthanspecificpolicydevelopments, and whenEuropean
integration stalled,so did thisliterature. The intellectual heirsof thistra-
dition,suchas JosephNye and RobertKeohane,emphasizedinterdepend-
enceandtransnationalism, buttheroleofdomesticfactorsslippedmoreand
moreoutoffocus,particularly as theconceptofinternational regimescame
to dominatethesubfield.11
The "bureaucraticpolitics" school of foreignpolicyanalysisinitiated
anotherpromising attackon the problemof domestic-international inter-
action.As GrahamAllisonnoted,"Appliedto relationsbetweennations,
thebureaucratic politicsmodeldirectsattention to intra-nationalgames,the
overlapof whichconstitutes international "12
relations. Nevertheless, the
natureof this"overlap" remained unclarified, and the theoreticalcontri-
butionof thisliterature did notevolvemuchbeyondtheprinciple thatbu-
reaucratic interests in
matter foreign policymaking.
Morerecently, themostsophisticated workon thedomesticdeterminants
of foreignpolicyhas focusedon "structural"factors,particularly "state
strength." The landmark worksof Peter Katzenstein and Stephen Krasner,
forexample,showedtheimportance ofdomesticfactors inforeign economic
policy.Katzenstein the
capturedtheessenceof problem: "The main purpose
of all strategiesof foreigneconomicpolicyis to make domesticpolicies
compatible withtheinternational politicaleconomy." 13 Bothauthorsstressed
the crucialpoint that central decision-makers ("the state") mustbe con-
cerned simultaneously with domestic and internationalpressures.
10. Karl W. Deutsch et al., Political Communityin the NorthAtlanticArea: International

Organization in the Light of Historical Experience (Princeton: PrincetonUniversityPress,
1957) and Ernst B. Haas, The Unitingof Europe: Political, Social, and Economic Forces,
1950-1957(Stanford,
Calif.:Stanford
University
Press,1958).
11. Robert0. Keohaneand JosephS. Nye, Powerand Interdependence (Boston:Little,
Brown,1977).On theregimeliterature,
including
itsneglectofdomesticfactors,
see Stephan
HaggardandBethSimmons, "TheoriesofInternational
Regimes,"International
Organization
41 (Summer1987),pp. 491-517.
12. Graham T. Allison, Essence of Decision: Explaining the Cuban Missile Crisis (Boston:
Little,Brown,1971),p. 149.
13. Peter J. Katzenstein, ed., Between Power and Plenty: Foreign Economic Policies of
AdvancedIndustrialStates (Madison:University
of WisconsinPress, 1978),p. 4. See also
Katzenstein,
"International
RelationsandDomesticStructures:
ForeignEconomicPoliciesof
AdvancedIndustrial
States,"International
Organization
30 (Winter1976),pp. 1-45; Stephen
D. Krasner,"UnitedStatesCommercial and MonetaryPolicy:Unravellingthe Paradoxof
ExternalStrength
and InternalWeakness,"in Katzenstein,BetweenPowerand Plenty,pp.
51-87; and Krasner, Defending the National Interest:Raw Materials Investmentsand U.S.
ForeignPolicy(Princeton:
Princeton
University
Press,1978).
432 International
Organization
Moredebatable,however,is theiridentification of"statestrength" as the

keyvariableofinterest. Giventhedifficulties ofmeasuring "statestrength,"
thisapproachcourtstautology,14and efforts to locateindividualcountries
on thisambiguouscontinuum haveprovedproblematic.15"State strength,"
ifreinterpreted as merelytheoppositeofgovernmental fragmentation,is no
doubtofsomeinterest inthecomparative studyofforeign policy.However,
Gourevitch is quitecorrectto complainthat"the strongstate-weakstate
argument suggeststhat. .. t-heidentityofthegoverning coalitiondoes not
matter.This is a veryapoliticalargument."16 Moreover,because "state
structures" (as conceivedin thisliterature)
varylittlefromissueto issueor
fromyearto year,suchexplanations areill-suitedforexplaining differences
acrossissuesor acrosstime(unless"time" is measuredin decadesor cen-
turies).A moreadequateaccountof thedomesticdeterminants offoreign
policyandinternational relationsmuststresspolitics: parties,socialclasses,
interestgroups(both economicand noneconomic),legislators,and even
publicopinionand elections,notsimplyexecutiveofficials and institutional
arrangements. 17
Some workin the"state-centric" genrerepresents a unitary-actor
model
runamok."The centralproposition ofthispaper,"notesone recentstudy,
"is thatthestatederivesitsinterests fromandadvocatespoliciesconsistent
withtheinternational systemat all timesandunderall circumstances."' 18 In
fact,on nearlyallimportant issues"centraldecision-makers" disagreeabout
whatthenationalinterest andtheinternational contextdemand.Even ifwe
arbitrarilyexcludethelegislaturefrom"thestate"(as muchofthisliterature
does),itis wrongtoassumethattheexecutiveis unified initsviews.Certainly
thiswas truein none of thestatesinvolvedin the 1978negotiations. What
was "the" positionof the Germanor Japanesestateon macroeconomic
policyin 1978,oroftheAmericanstateonenergy policy?Iftheterm"state"
is to be used to mean "centraldecision-makers," we shouldtreatit as a
plural noun: not "the state, it . . ." but "the state, they . . ." Centralex-
ecutiveshavea specialroleinmediating
domesticandinternational
pressures
preciselybecause theyare directlyexposedto bothspheres,notbecause
14.Forexample,see Krasner,"UnitedStatesCommercial andMoneyPolicy,"p. 55: "The

centralanalyticcharacteristic thatdetermines the abilityof a stateto overcomedomestic
resistanceis itsstrengthin relationto itsownsociety."
15.HelenMilner,"Resisting theProtectionist
Temptation: IndustryandtheMakingofTrade
Policyin France and the UnitedStatesduringthe 1970s,"International 41
Organization
(Autumn1987),pp. 639-65.
16. Gourevitch, "The SecondImageReversed,"p. 903.
17.In theirmoredescriptive work,"state-centric" scholarsareoftensensitivetotheimpact
ofsocialandpoliticalconflicts,suchas thosebetweenindustry andfinance,laborandbusiness,
and export-oriented versusimport-competing sectors.See Katzenstein, BetweenPqwerand
Plenty,pp. 333-36,forexample.
18. David A. Lake, "The Stateas Conduit:The International SourcesofNationalPolitical
Action,"presented at the1984annualmeeting oftheAmerican PoliticalScienceAssociation,
p. 13.
theyare unitedon all issues norbecause theyare insulatedfromdomestic

politics.
Thus,thestate-centric literature
is an uncertain foundationfortheorizing
abouthowdomesticand international politicsinteract.
Moreinterestingare
recentworksabout the impactof the international economyon domestic
politicsand domesticeconomicpolicy,suchas thoseby Alt,Evans,Gour-
evitch,and Katzenstein.19 These case studies,representing diversemeth-
odologicalapproaches,displaya theoretical sophistication
on theinterna-
tional-to-domestic
causal connectionfargreaterthanis characteristic of
comparablestudieson thedomestic-to-international halfoftheloop. Never-
theless,theseworksdo notpurportto accountforinstancesof reciprocal
causation,nor do theyexaminecases in whichthe domesticpoliticsof
severalcountriesbecameentangled internationally.
In short,we need to move beyondthemereobservation thatdomestic
factorsinfluenceinternational affairsand vice versa,and beyondsimple
catalogsofinstancesof suchinfluence, to seek theoriesthatintegrate
both
spheres,accountingfortheareas ofentanglement betweenthem.
Two-levelgames: a metaphor
fordomestic-international
interactions
Overtwodecadesago RichardE. Waltonand RobertB. McKersieoffered

a "behavioraltheory"of social negotiations
thatis strikinglyapplicableto
international andcooperation.20
conflict Theypointedout,as all experienced
negotiatorsknow,thattheunitary-actor assumption is oftenradicallymis-
leading.As RobertStrausssaid of the Tokyo Roundtradenegotiations:
"Duringmytenureas Special TradeRepresentative, I spentas muchtime
negotiating
withdomesticconstituents (bothindustry and labor)and mem-
bers of the U.S. Congressas I did negotiatingwithour foreigntrading
partners."21
19.JamesE. Alt,"CrudePolitics:OilandthePoliticalEconomyofUnemployment inBritain

andNorway,1970-1985,"BritishJournal ofPoliticalScience17(April1987),pp. 149-99;Peter
B. Evans, Dependent Development: The Alliance of Multinational,State, and Local Capital
in Brazil(Princeton:Princeton
University
Press, 1979);PeterGourevitch,
Politicsin Hard
Times: ComparativeResponses to InternationalEconomic Crises (Ithaca, N.Y.: Cornell Uni-
versityPress, 1986); Peter J. Katzenstein,Small States in WorldMarkets: IndustrialPolicy in
Europe(Ithaca,N.Y.: CornellUniversity
Press,1985).
20. RichardE. Waltonand RobertB. McKersie, A Behavioral TheoryofLabor Negotiations:
An Analysis of a Social InteractionSystem(New York: McGraw-Hill, 1965).
21. RobertS. Strauss,"Foreword,"in JoanE. Twiggs,The TokyoRoundofMultilateral
Trade Negotiations: A Case Studyin BuildingDomestic Supportfor Diplomacy (Washington,
D.C.: Georgetown forthe Studyof Diplomacy,1987),p. vii. Former
Institute
University
SecretaryofLaborJohnDunlopis said to haveremarkedthat"bilateralnegotiations
usually
requirethreeagreements-oneacrossthetableandone on each sideofthetable,"as citedin
Howard Raiffa,The Art and Science of Negotiation (Cambridge,Mass.: Harvard University
Press,1982),p. 166.
Organization
434 International
The politicsofmanyinternational negotiations be conceived

can usefully
as a two-levelgame. At the nationallevel, domesticgroupspursuetheir
interestsbypressuring thegovernment to adoptfavorablepolicies,andpol-
iticiansseek powerby constructing coalitionsamongthosegroups.At the
internationallevel,nationalgovernments seekto maximizetheirownability
to satisfydomesticpressures,whileminimizing theadverseconsequences
offoreign developments. Neitherofthetwogamescanbe ignoredbycentral
decision-makers, so longas theircountriesremaininterdependent, yetsov-
ereign.
Each nationalpoliticalleaderappearsat bothgameboards.Acrossthe
internationaltablesithisforeigncounterparts,andathiselbowssitdiplomats
and otherinternational advisors.Aroundthedomestictablebehindhimsit
partyand parliamentary figures,spokespersons fordomesticagencies,rep-
resentatives ofkeyinterest groups,and theleader'sownpoliticaladvisors.
Theunusualcomplexity ofthistwo-level gameis thatmovesthatarerational
fora playerat one board(suchas raisingenergyprices,concedingterritory,
or limitingautoimports)maybe impolitic forthatsameplayerat theother
board.Nevertheless, therearepowerful incentivesforconsistencybetween
the two games. Players(and kibitzers)will toleratesome differences in
rhetoricbetweenthetwogames,butin theendeitherenergypricesriseor
theydon't.
The politicalcomplexities fortheplayersin thistwo-levelgameare stag-
gering.Anykeyplayerat theinternational tablewhois dissatisfied
withthe
outcomemayupsetthegameboard,and conversely, anyleaderwho fails
to satisfyhis fellowplayersat thedomestictablerisksbeingevictedfrom
hisseat.On occasion,however,cleverplayerswillspota moveononeboard
thatwill trigger realignments on otherboards,enablingthemto achieve
otherwise unattainable objectives.This "two-table"metaphor capturesthe
dynamicsof the 1978negotiations betterthananymodelbased on unitary
nationalactors.
Otherscholarshave notedthe multiple-game re-
natureof international
lations.Like Waltonand McKersie,Daniel Druckmanhas observedthata
negotiator "attemptsto builda packagethatwillbe acceptablebothto the
otherside and to his bureaucracy."However,Druckmanmodelsthe do-
mesticand international processesseparatelyand concludesthat"the in-
teractionbetween the processes . . . remains a topic for investigation."
22
RobertAxelrodhas proposeda "Gamma paradigm,"in whichthe U.S.

pursuespoliciesvis-a-vistheSovietUnionwithan eye towards
president
maximizing athome.However,thismodeldisregards
hispopularity domestic
22. DanielDruckman,
"BoundaryRole Conflict:
Negotiation
as Dual Responsiveness,"
in
I. WilliamZartman,ed., The Negotiation Process: Theories and Applications(Beverly Hills:
Sage, 1978),pp. 100-101,109.Fora reviewofthesocial-psychological
literature
on bargainers
as representatives,
see Dean G. Pruitt,Negotiation
Behavior(New York:AcademicPress,
1981),pp. 41-43.
cleavages,and itpostulatesthatone oftheinternational actors-theSoviet

leadership-caresonlyaboutinternational gainsandfacesno domesticcon-
straintwhiletheother-theU.S. president-cares onlyaboutdomestic gains,
exceptinsofaras hispublicevaluatestheinternational Prob-
competition.23
ablythemostinteresting empirically based theorizingabouttheconnection
betweendomesticand international bargainingis thatofGlennSnyderand
Paul Diesing.Thoughworkingin theneo-realist traditionwithits conven-
tionalassumptionofunitary actors,theyfoundthat,infullyhalfofthecrises
theystudied,top decision-makers were not unified.They concludedthat
prediction outcomesis significantly
ofinternational improvedbyunderstand-
inginternal especiallywithrespecttominimally
bargaining, acceptablecom-
24
promises.
Metaphorsare nottheories,butI am comforted by Max Black's obser-
vationthat"perhapseverysciencemuststartwithmetaphor and end with
algebra;and perhapswithoutthe metaphortherewouldneverhave been
any algebra."25Formalanalysisof any game requireswell-defined rules,
choices,payoffs, andeventhen,manysimpletwo-
players,andinformation,
person,mixed-motive gameshaveno determinate solution.Derivinganalytic
challenge.In whatfollowsI
solutionsfortwo-levelgameswillbe a difficult
hopeto motivatefurther workon thatproblem.
Towards a theoryof ratification:

the importanceof "win-sets"
Considerthefollowing stylizedscenariothatmightapplyto anytwo-level

game.Negotiators representingtwo organizations meetto reachan agree-
mentbetweenthem,subjectto theconstraint thatanytentative agreement
mustbe ratifiedby theirrespectiveorganizations.The negotiatorsmightbe
heads ofgovernment representingnations,forexample,or laborand man-
agementrepresentatives, or partyleadersin a multiparty coalition,or a
financeministernegotiatingwithan IMF team,orleadersofa House-Senate
conference committee, or ethnic-group leadersin a consociational
democ-
racy.For themoment, we shallpresumethateach sideis represented by a
singleleaderor "chiefnegotiator," has no indepen-
and thatthisindividual
23. RobertAxelrod,"The GammaParadigm theDomesticInfluence

forStudying on Foreign
StudiesAsso-
Policy,"preparedfordeliveryat the1987AnnualMeetingoftheInternational
ciation.
24. Glenn H. Snyderand Paul Diesing, ConflictAmong Nations: Bargaining,Decision Mak-
ing,and SystemStructureinInternationalCrises (Princeton:PrincetonUniversityPress, 1977),
pp. 510-25.
Press,1962),p.
25. Max Black,Modelsand Metaphors(Ithaca,N.Y.: CornellUniversity
Politics,"WorldPolitics
242,as citedin DuncanSnidal,"The Game TheoryofInternational
38 (October1985),p. 36n.
436 International
Organization
dentpolicypreferences, butseekssimplyto achievean agreement

thatwill
be attractiveto his constituents.26
It is convenientanalytically
to decomposetheprocessintotwostages:
1. bargaining betweenthenegotiators,
leadingto a tentative
agreement;
call thatLevel I.
2. separatediscussionswithineach groupofconstituents
aboutwhether
theagreement;
to ratify call thatLevel II.
This sequentialdecomposition intoa negotiationphase and a ratification
phase is usefulforpurposesof exposition,althoughit is notdescriptively
accurate.In practice,expectational effectswillbe quiteimportant. There
are likelyto be priorconsultations and bargaining at Level II to hammer
outan initialpositionfortheLevel I negotiations. Conversely, theneedfor
Level II ratification is certainto affecttheLevel I bargaining. In fact,ex-
pectationsofrejectionat Level II mayabortnegotiations at Level I without
anyformalactionat Level II. For example,eventhoughboththeAmerican
and Iraniangovernments seem to have favoredan arms-for-hostages deal,
negotiations collapsedas soon as theybecamepublicand thusliableto de
facto"ratification." In manynegotiations, the two-levelprocessmaybe
iterative,as the negotiators tryout possibleagreements and probetheir
constituents' views. In morecomplicatedcases, as we shallsee later,the
constituents' viewsmaythemselves evolveinthecourseofthenegotiations.
Nevertheless, therequirement thatanyLevel I agreement must,intheend,
be ratified at Level II imposesa crucialtheoretical linkbetweenthetwo
levels.
"Ratification" mayentaila formalvotingprocedureat Level II, suchas
theconstitutionally requiredtwo-thirds voteoftheU.S. Senateforratifying
treaties,but I use the termgenerically to referto any decision-process at
Level II thatis requiredto endorseor implement a Level I agreement,
whether formally or informally.It is sometimes convenient to thinkofrat-
ificationas a parliamentary function, butthatis notessential.The actorsat
Level II mayrepresent bureaucratic agencies,interestgroups,socialclasses,
or even "publicopinion."For example,iflaborunionsin a debtorcountry
withhold necessarycooperation froman austerity program thatthegovern-
menthas negotiatedwiththe IMF, Level II ratification of theagreement
maybe said to have failed;ex ante expectations aboutthatprospectwill
surelyinfluence theLevel I negotiations betweenthegovernment and the
IMF.
Domesticratification ofinternationalagreements mightseempeculiarto
democracies. As theGermanFinanceMinister recentlyobserved,"The limit
of expandedcooperationlies in thefactthatwe are democracies,and we
26. To avoidunnecessary complexity,
myargument is phrasedintermsofa single
throughout
chiefnegotiator,
although inmanycases someofhisresponsibilities
maybe delegatedtoaides.
LaterinthisarticleI relaxtheassumptionthatthenegotiator
has no independentpreferences.
need to secureelectoralmajorities at home."27However,ratification need

notbe "democratic"in anynormalsense. For example,in 1930theMeiji
Constitution was interpreted as givinga specialroleto theJapanesemilitary
in theratificationoftheLondonNaval Treaty;28 and duringtheratification
of any agreement betweenCatholicsand Protestants in Northern Ireland,
presumably theIRA wouldthrowitspowerontothescales. We needonly
stipulatethat,forpurposesofcounting"votes" in theratification process,
differentformsofpoliticalpowercan be reducedto somecommondenom-
inator.
The onlyformalconstraint on the ratificationprocessis thatsince the
identicalagreement mustbe ratified by bothsides, a preliminary Level I
agreement cannotbe amendedat Level II withoutreopeningthe Level I
negotiations.In otherwords,finalratification mustbe simply"voted" up
or down;anymodification to theLevel I agreement countsas a rejection,
unlessthatmodification is approvedby all otherpartiesto theagreement.29
Congresswoman LynnMartincapturedthe logic of ratification whenex-
plainingher supportforthe 1986tax reformbill as it emergedfromthe
conference committee: "As worriedas I am aboutwhatthisbilldoes, I am
evenmoreworriedaboutthecurrent code. The choicetodayis notbetween
thisbilland a perfectbill; thechoiceis betweenthisbilland thedeathof
tax reform."30
Giventhisset ofarrangements, we maydefinethe"win-set"fora given
Level II constituencyas thesetofall possibleLevel I agreements thatwould
"win"-that is, gainthenecessarymajority amongtheconstituents-when
simplyvotedup or down.31For twoquitedifferent reasons,thecontoursof
theLevel II win-setsare veryimportant forunderstanding Level I agree-
ments.
First,largerwin-setsmake Level I agreementmorelikely,ceterisparibus.32
mustfallwithintheLevel II win-
any successfulagreement
By definition,
27. Gerhardt
Stoltenberg,
WallStreetJournal
Europe,2 October1986,as citedinC. Randall
Henning,Macroeconomic Diplomacy in the1980s: Domestic Politics and InternationalConflict
Among the UnitedStates, Japan, and Europe, AtlanticPaper No. 65 (New York: Croom Helm,
fortheAtlanticInstitute
forInternational 1987),p. 1.
Affairs,
28. Ito Takashi,"Conflicts
and Coalitionin Japan,1930:PoliticalGroupsand theLondon
Naval Disarmament Conference,"in Sven Groennings et al., eds, The Studyof Coalition
Behavior(New York: Holt, Rinehart, & Winston,1970);KobayashiTatsuo,"The London
Naval Treaty, 1930," in JamesW. Morley, ed., Japan Erupts: The London Naval Conference
and the Manchurian Incident, 1928-1932 (New York: Columbia UniversityPress, 1984), pp.
11-117.I am indebtedto WilliamJaroszforthisexample.
29. Thisstipulation
is, infact,characteristic
ofmostreal-world
ratification
procedures,
such
as House andSenateactionon conference committeereports,
althoughitis somewhatviolated
by theoccasionalpracticeofappending "reservations"
to theratification
oftreaties.
30. New YorkTimes, 26 September 1986.
31. For theconceptionof win-set,see KennethA. Shepsleand BarryR. Weingast,"The
Institutional
FoundationsofCommittee Power,"American PoliticalScienceReview81 (March
1987),pp. 85-104.I am indebtedto ProfessorShepsleformuchhelpon thistopic.
32. To avoidtedium,I do notrepeatthe"otherthingsbeingequal" provisoin each ofthe
propositionsthatfollow.Undersomecircumstances an expandedwin-set mightactuallymake
practicablesomeoutcomethatcouldtrigger a dilemmaof collectiveaction.See VincentP.
Crawford, "A TheoryofDisagreement in Bargaining,"
Econometrica 50 (May 1982),pp. 607-
37.
438 International
Organization
setsof each of thepartiesto theaccord.Thus,agreement is possibleonly

ifthosewin-setsoverlap,and thelargereach win-set,themorelikelythey
areto overlap.Conversely, thesmallerthewin-sets, thegreatertheriskthat
thenegotiations willbreakdown.For example,duringtheprolongedpre-
warAnglo-Argentine negotiations overtheFalklands/Malvinas, severalten-
tativeagreements were rejectedin one capitalor the otherfordomestic
politicalreasons;whenitbecameclearthattheinitialBritishandArgentine
win-setsdid notoverlapat all, warbecamevirtually inevitable.33
A brief,but important digression:The possibilityof failedratification
suggests thatgametheoretical analysesshoulddistinguish betweenvoluntary
and involuntary defection.Voluntary defectionrefersto reneging by a ra-
tionalegoistin the absence of enforceable contracts-themuch-analyzed
problemposed,forexample,in theprisoner'sdilemmaand otherdilemmas
of collectiveaction.Involuntary defection insteadreflectsthebehaviorof
of
an agentwhois unableto deliveron a promisebecause failedratification.
Even thoughthesetwotypesofbehaviormaybe difficult to disentanglein
someinstances,theunderlying logicis quitedifferent.
The prospectsforinternational cooperationin an anarchic,"self-help"
worldare oftensaid to be poorbecause "unfortunately, policymakersgen-
erallyhave an incentiveto cheat."34However,as Axelrod,Keohane,and
othershavepointedout,thetemptation todefectcanbe dramatically reduced
If
amongplayerswho expectto meetagain.35 policymakers in an anarchic
worldwerein factconstantly temptedto cheat,certainfeaturesofthe1978
storywouldbe veryanomalous.For example,eventhoughtheBonnagree-
mentwas negotiatedwithexquisitecare, it containedno provisionsfor
temporal balance,sequencing, thatmight
or partialconditionality havepro-
tectedthepartiesfromunexpecteddefection. Moreover,theGermansand
theJapaneseirretrievably enactedtheirparts thebargainmorethansix
of
monthsbeforethepresident'sactionon oil pricedecontroland nearlytwo
yearsbeforethatdecisionwas implemented. Once theyhad done so, the
temptation to the president to renege should have been overpowering, but
in factvirtually no one on either side of the decontroldebate within the
administration dismissedthe Bonn pledge as In
irrelevant. short,the Bonn
"promise"had politicalweight,becausereneging wouldhave had highpo-
liticaland diplomatic costs.
33. The SundayTimes InsightTeam, The FalklandsWar(London: Sphere,1982);Max
HastingsandSimonJenkins,
TheBattlefortheFalklands(NewYork:Norton,1984);Alejandro
Dabat and Luis Lorenzano, Argentina: The Malvinas and the End of MilitaryRule (London:
Verso,1984).I am indebtedto Louise Richardson
forthesecitations.
34. MatthewE. Canzoneriand Jo Anna Gray,"Two Essays on MonetaryPolicyin an
InterdependentWorld,"InternationalFinanceDiscussionPaper219 (BoardofGovernorsof
theFederalReserveSystem,February1983).
35. RobertAxelrod,TheEvolutionofCooperation (New York:Basic Books,1984);Robert
0. Keohane, AfterHegemony: Cooperation and Discord in the World Political Economy
(Princeton:PrincetonUniversity
Press, 1984),esp. p. 116; and the specialissue of World
Politics,"CooperationUnderAnarchy,"KennethA. Oye,ed., vol. 38 (October1985).
On the otherhand,in any two-levelgame,the credibility of an official

commitment may be low, even ifthe reputational costs of reneging arehigh,
forthe negotiator maybe unableto guaranteeratification. The failureof
Congressto ratify abolitionofthe "AmericanSellingPrice" as previously
agreedduringtheKennedyRoundtradenegotiations is oneclassicinstance;
anotheris the inabilityof JapanesePrimeMinisterSato to deliveron a
promisemade to PresidentNixonduringthe "TextileWrangle.' '36 A key
obstacleto Westerneconomiccoordination in 1985-87was theGermans'
fearthattheReaganadministration wouldbe politically unableto carryout
any commitment it might make to cut the U.S. budget deficit,no matter
howwell-intentioned thepresident.
Unlikeconcernsaboutvoluntary defection, concernabout"deliver-abil-
ity"was a prominent elementin theBonnnegotiations. In thepost-summit
pressconference, President Carterstressedthat"each ofus hasbeencareful
not to promisemorethanhe can deliver."A majorissue throughout the
negotiations was Carter'sownabilityto deliveron hisenergycommitments.
The Americansworkedhardto convincetheothers,first, thatthepresident
was underseveredomesticpoliticalconstraints on energyissues, which
limitedwhathe could promise,but second,thathe could deliverwhathe
was preparedto promise.The negotiators in 1978seemedto followthis
presumption aboutone another: "He willdo whathe has promised,so long
as whathe has promised is clear and within his power."
Involuntary defection, and the fear of it,can be just as fatalto prospects
as
forcooperation voluntary defection. Moreover, in somecases, itmaybe
difficult,both for the other side and for outside analysts,to distinguish
voluntary and involuntary defection, particularly since a strategicnegotiator
to
mightseek misrepresentvoluntarya defection as involuntary. Such be-
havioris itselfpresumably subject to some reputational constraints, although
it is an important empirical question how far reputations generalize from
to
collectivities negotiators and vice versa. Credibility (and thusthe ability
tostrikedeals)atLevel I is enhancedbya negotiator' s (demonstrated) ability
to "deliver" at Level II; this was a major strength of Robert Strauss in the
Tokyo Round negotiations.37
Involuntary defection can onlybe understood withintheframework ofa
two-level game. Thus, to return to the issue of win-sets, the smaller the win-
sets,thegreatertheriskof involuntary defection, and hencethemoreap-
plicable the literatureabout dilemmas of collective action.38
36. I. M. Destler,HaruhiroFukui,and Hideo Sato, The TextileWrangle:ConflictinJapanese-
American Relations, 1969-1971 (Ithaca, N.Y.: Cornell UniversityPress, 1979), pp. 121-57.
37. GilbertR. Winham,"RobertStrauss,theMTN, and theControlof Faction,"Journal
of World Trade Law 14 (September-October 1980), pp. 377-97, and his InternationalTrade
and theTokyoRound(Princeton: PrincetonUniversity
Press,1986).
38. Thisdiscussionimplicitly
assumesuncertaintyaboutthecontoursofthewin-sets
on the
partoftheLevel I negotiators,
forifthewin-setswereknownwithcertainty,thenegotiators
wouldneverproposeforratification an agreement
thatwouldbe rejected.
440 International
Organization
is thattherelativesize
The secondreasonwhywin-setsize is important
of thejoint gains
of therespectiveLevel II win-setswillaffectthedistribution
fromthe internationalbargain. The largerthe perceived win-setof a nego-
themorehe canbe "pushedaround"bytheotherLevel I negotiators.
tiator,
Conversely, a smalldomesticwin-setcan be a bargaining advantage:"I'd
liketo acceptyourproposal,but I could nevergetit acceptedat home."
Lamenting thedomesticconstraintsunderwhichone mustoperateis (inthe
wordsofone experienced Britishdiplomat)"the naturalthingto say at the
beginning ofa toughnegotiation."
39
This generalprinciplewas, of course,firstnotedby ThomasSchelling
nearlythirtyyearsago:
The powerofa negotiator oftenrestson a manifestinabilityto make
concessions and meet demands.... When the United States Govern-
mentnegotiateswithothergoverments. . . if the executive branchne-
gotiatesunderlegislativeauthority,withitspositionconstrainedby
law, . . . thentheexecutivebranchhas a firmpositionthatis visibleto
itsnegotiating suchas this]runthe
partners.... [Ofcourse,strategies
riskofestablishing an immovablepositionthatgoes beyondtheability
oftheotherto concede,and thereby provokethelikelihoodof stale-
mateor breakdown.40
Writing pointofview,Schellingstressedwaysinwhich
froma strategist's
win-setsmaybe manipulated, but even whenthewin-setitselfis beyond
thenegotiator's control,he mayexploititsleverage.A ThirdWorldleader
whosedomesticpositionis relatively weak (Argentina'sAlfonsin?)should
be able to drivea betterbargainwithhisinternationalcreditors,otherthings
beingequal, thanone whosedomesticstanding is moresolid(Mexico's de
ofwinning
la Madrid?).41The difficulties congressionalratification
are often
exploitedby Americannegotiators. Duringthenegotiation of thePanama
CanalTreaty,forexample,"theSecretary ofStatewarnedthePanamanians
severaltimes.. . thatthenewtreatywouldhavetobe acceptabletoat least
sixty-seven senators,"and "Carter,ina personalletterto Torrijos,warned
thatfurther concessionsbytheUnitedStateswouldseriously threaten
chances
forSenateratification.'"42Preciselyto forestall
suchtactics,opponentsmay
demandthata negotiator ensurehimself"negotiating room" at Level II
beforeopeningtheLevel I negotiations.
The "sweet-and-sour"implications of win-setsize are summarized in
Figure1, representing a simplezero-sumgamebetweenX and Y. XMand
39. GeoffreyW. Harrison, in JohnC. Campbell, ed., Successful Negotiation: Trieste1954

(Princeton:PrincetonUniversity
Press,1976),p. 62.
40. ThomasC. Schelling,TheStrategyof Conflict Mass.: HarvardUniversity
(Cambridge,
Press,1960),pp. 19-28.
41. I am grateful
to Lara Putnamforthisexample.For supportingevidence,see RobertR.
Kaufman,"Democraticand Authoritarian Responsesto theDebt Issue: Argentina,Brazil,
Mexico," InternationalOrganization39 (Summer 1985), pp. 473-503.
42. W. MarkHabeebandI. WilliamZartman,ThePanamaCanalNegotiations
(Washington,
D.C.: JohnsHopkinsForeignPolicyInstitute,
1986),pp. 40, 42.
xm r r 'I r IY
YX Y2 XI Y3
FIGURE 1. Effectsof reducingwin-setsize

YM represent themaximum outcomesforX and Y, respectively,whileX1
and Y1 represent theminimal At thisstage
outcomesthatcouldbe ratified.
any agreement by both
in the rangebetweenX1 and Y1 could be ratified
parties.Ifthewin-setofY werecontracted to,say,Y2 (perhapsbyrequiring
a largermajority outcomesbetweenY1 and Y2 wouldno
forratification),
longerbe feasible,and therangeoffeasibleagreements wouldthusbe trun-
cated in Y's favor.However,if Y, emboldenedby thissuccess,wereto
reduceits win-setstillfurtherto Y3 (perhapsby requiringunanimity for
would suddenlyfindthemselvesdeadlocked,
the negotiators
ratification),
forthewin-setsno longeroverlapat all.43
Determinantsof the win-set
to understand
It is important affectwin-setsize. Three
whatcircumstances
setsoffactorsare especiallyimportant:
proposedmodelsoflinkedgamesakin
in otherfieldshaverecently
43. Severalinvestigators
to this"two-level"game. KennethA. Shepsleand his colleagueshave used the notionof
"interconnectedgames" to analyze,forexample,thestrategy simultaneously
of a legislator
embeddedin twogames,one in thelegislative arenaand theotherin theelectoralarena.In
thismodel,a givenactionis simultaneously a movein twodifferent games,and one player
maximizes thesumofhispayoffs fromthetwogames.See Arthur Denzau,WilliamRiker,and
KennethShepsle,"Farquharson andFenno:SophisticatedVotingandHomeStyle,"American
Political Science Review 79 (December 1985),pp. 1117-34; and KennethShepsle, "Cooperation
Arrangements,"
and Institutional unpublished manuscript, February1986.This approachis
similarto modelsrecentlydevelopedby economists working in the "rationalexpectations"
genre.In thesemodels,a government contendssimultaneously againstothergovernmentsand
againstdomestictradeunionsovermonetary policy.See, forexample,KennethRogoff, "Can
Monetary
International PolicyCooperation be Counterproductive," JournalofInternational
Economics18 (May 1985),pp. 199-217,and RolandVaubel,"A PublicChoiceApproachto
Organization,"
International Public Choice51 (1986),pp. 39-57. GeorgeTsebelis("Nested
Games: The Cohesion of French Coalitions," BritishJournal of Political Science 18 [April
1988],pp. 145-70)has developeda theoryof "nestedgames,"in whichtwoalliancesplaya
competitivegameto determine whiletheindividual
totalpayoffs, playerswithineach alliance
contendovertheirshares.FritzSharpf("A Game-Theoretical InterpretationofInflationand
Unemployment inWestern Europe,"Journal ofPublicPolicy7 [19881, pp. 227-257)interprets
macroeconomic policyas thejointoutcomeoftwosimultaneous games;inone,thegovernment
playsagainsttheunions,whilein theother,it respondsto theanticipated reactionsof the
electorate.JamesE. Altand BarryEichengreen ("Paralleland Overlapping Games:Theory
and an Application to theEuropeanGas Trade," unpublished manuscript, November1987)
offera broadertypology oflinkedgames,distinguishingbetween"parallel"games,in which
"thesameopponents playagainstone anotherat thesametimein morethanone arena,"and
"overlapping"games,whicharise"whena particular playeris engagedat thesametimein
gamesagainstdistinct opponents, and whenthestrategy pursuedin one gamelimitsthestrat-
egiesavailablein theother."Detailedcomparison of thesevariouslinked-game modelsis a
taskforthefuture.
Organization
442 International
* Level II preferences
and coalitions
* Level IL institutions
* Level I negotiators'
strategies
Let us considereach in turn.
1. The size of the win-setdepends on the distributionof

power, preferences,and possible coalitions among Level II
constituents.
Anytestabletwo-leveltheoryofinternational negotiationmustbe rooted
in a theoryof domesticpolitics,thatis, a theoryabout the powerand
preferences ofthemajoractorsat Level II. Thisis nottheoccasionforeven
a cursoryevaluationoftherelevant exceptto notethatthetwo-
alternatives,
level conceptualframework could in principlebe marriedto such diverse
perspectives as Marxism,interest
grouppluralism, bureaucratic politics,and
neo-corporatism. For example,armsnegotiations mightbe interpreted in
termsof a bureaucratic politicsmodelof Level II politicking, whileclass
analysisorneo-corporatism might be appropriateforanalyzing international
macroeconomic coordination.
Abstracting fromthedetailsof Level II politics,however,it is possible
thatgovernthesize ofthewin-sets.
to sketchcertainprinciples Forexample,
thelowerthecost of "no-agreement" to constituents,thesmallerthewin-
set.44Recall thatratification
pitsthe proposedagreement, not againstan
arrayofother(possiblyattractive)alternatives,butonlyagainst"no-agree-
ment."145 No-agreement oftenrepresents the statusquo, althoughin some
cases no-agreement mayin factlead to a worsening situation;thatmightbe
a reasonabledescriptionofthefailedratificationoftheVersaillesTreaty.
Someconstituents mayfacelowcostsfromno-agreement, andothershigh
costs,and theformer willbe moreskepticalofLevel I agreements thanthe
latter.Membersoftwo-wage-earner familiesshouldbe readierto strike,for
example,thansole breadwinners, and small-town barbersshouldbe more
isolationist bankers.In thissense,someconstituents
thaninternational may
offereithergenericoppositionto,orgenericsupport for,Level I agreements,
moreorless independently ofthespecific contentoftheagreement, although
naturally decisionsaboutratification
otherconstituents' willbe closelycon-
ditionedon thespecifics.The size of thewin-set(and thusthenegotiating
44. ThomasRomerandHowardRosenthal, "PoliticalResourceAllocation,

ControlledAgen-
das, and theStatusQuo," PublicChoice33 (no. 4, 1978),pp. 27-44.
45. In moreformaltreatments, theno-agreement outcomeis calledthe"reversionpoint."
A givenconstituent'sevaluationofno-agreement corresponds to whatRaiffatermsa seller's
"walk-away price,"thatis, thepricebelowwhichhe wouldprefer "no-deal."(Raiffa,
Artand
ScienceofNegotiation.) No-agreement is equivalenttowhatSnyderandDiesingterm"break-
down,"or theexpectedcostofwar.(Snyderand Diesing,Conflict AmongNations.)
roomof theLevel I negotiator) dependson therelativesize of the "isola-

tionist"forces(who oppose international cooperationin general)and the
"internationalists" (whooffer"all-purpose"support).All-purpose support
forinternational agreements is probablygreaterin smaller,moredependent
countrieswithmoreopen economies,as comparedto moreself-sufficient
countries,liketheUnitedStates,formostofwhosecitizensthecostsofno-
agreement are generallylower.Ceterisparibus,moreself-sufficient states
withsmallerwin-setsshouldmakefewerinternational agreements anddrive
harderbargainsin thosethattheydo make.
In some cases, evaluationof no-agreement maybe the onlysignificant
disagreement amongthe Level II constituents, because theirinterests are
relativelyhomogeneous. For example,ifoil importsare to be limitedby an
agreement amongtheconsuming nations-thesortof accordsoughtat the
Tokyo summitof 1979,forexample-thenpresumably everyconstituent
wouldprefer tomaximizehisnation'sshareoftheavailablesupply,although
someconstituents maybe morereluctant thanothersto pushtoo hard,for
fearof losingtheagreement Similarly,
entirely. in mostwage negotiations,
theinterests of constituents (eitherworkersor shareholders) are relatively
homogeneous, and themostsignificant cleavagewithintheLevel II consti-
tuenciesis likelyto be between"hawks" and "doves," depending on their
willingnessto riska strike.(WaltonandMcKersiereferto theseas "bound-
ary" conflicts,in whichthenegotiator is caughtbetweenhis constituency
and the externalorganization.) Otherinternationalexamplesin whichdo-
mesticinterestsare relativelyhomogeneousexceptforthe evaluationof
no-agreement mightincludetheSALT talks,thePanamaCanal Treatyne-
gotiations,and the Arab-Israeliconflict.A negotiator is unlikelyto face
criticismat homethata proposedagreement reducestheopponents'arms
toomuch,offers too littlecompensation forforeignconcessions,orcontains
toofewsecurity guarantees fortheotherside,although ineachcase opinions
maydiffer on how muchto riska negotiating deadlockin orderto achieve
theseobjectives.
The distinctive natureofsuch"homogeneous"issuesis thrown intosharp
reliefby contrasting themto cases in whichconstituents' preferences are
moreheterogeneous, so thatanyLevel I agreement bearsunevenly onthem.
Thus,an internationally coordinatedreflationmayencounter domesticop-
positionbothfromthosewho thinkit goes too far(bankers,forexample)
and fromthosewho thinkit does notgo farenough(unions,forexample).
In 1919,someAmericansopposedtheVersaillesTreatybecauseitwas too
harshon thedefeatedpowersand othersbecauseitwas too lenient.46 Such
patternsare even morecommon,as we shallshortly see, wherethenego-
tiationinvolvesmultipleissues, such as an armsagreement thatinvolves
tradeoffsbetweenseaborneandairborne weapons,ora laboragreement that
46. Thomas A. Bailey, Woodrow Wilson and the Great Betrayal (New York: Macmillan,
1945), pp. 16-37.
444 International
Organization
involves tradeoffsbetween take-home payandpensions. (WaltonandMcKersie

termthese"factional"conflicts, because the negotiator is caughtbetween
contending factionswithinhis ownorganization.)
The problemsfacingLevel I negotiators dealingwitha homogeneous(or
"boundary")conflict arequitedifferent fromthosefacingnegotiators dealing
witha heterogeneous (or "factional")conflict. In theformer case, themore
thenegotiator can winat Level I-the higherhisnationaloil allocation,the
deeperthecutsin Sovietthrow-weight, thelowertherenthe promisesfor
theCanal, and so one-the better his odds ofwinning ratification.In such
cases, the negotiator may use the implicit threat from his own hawks to
maximizehisgains(or minimize hislosses) at Level I, as Carter and Vance
did in dealingwiththePanamanians.Glancingoverhis shoulderat Level
II, thenegotiator's mainproblemin a homogeneous conflict is to manage
thediscrepancy betweenhis constituents' expectations and thenegotiable
outcome.Neithernegotiator is likelytofindmuchsympathy fortheenemy's
demandsamonghisownconstituents, normuchsupport forhisconstituents'
positionsin theenemycamp.The effectofdomesticdivision,embodiedin
hard-line oppositionfromhawks,is to raisetheriskofinvoluntary defection
and thusto impedeagreement at Level I. The commonbeliefthatdomestic
politicsis inimicalto international cooperationno doubtderivesfromsuch
cases.
The taskof a negotiator grappling insteadwitha heterogeneous conflict
is morecomplicated, butpotentially moreinteresting. Seekingto maximize
thechancesofratification, he cannotfollowa simple"themore,thebetter"
ruleof thumb;imposingmoreseverereparations on theGermansin 1919
wouldhavegainedsomevotesat Level II butlostothers,as wouldhastening
thedecontrolof domesticoil pricesin 1978.In some cases, theselinesof
cleavagewithintheLevel II constituencies willcut across theLevel I di-
vision,and the Level I negotiator mayfindsilentallies at his opponent's
domestictable. Germanlabor unionsmightwelcomeforeignpressureon
theirown government to adopta moreexpansivefiscalpolicy,and Italian
bankersmightwelcomeinternational demandsfora moreaustereItalian
monetary policy.Thus transnational alignments mayemerge,tacitor ex-
plicit,in whichdomesticinterests pressuretheirrespectivegovernments to
adoptmutually supportive policies.Thisis, ofcourse,myinterpretation of
the 1978Bonnsummit accord.
In suchcases, domesticdivisionsmayactuallyimprovetheprospectsfor
international cooperation.For example,considertwodifferent distributions
of constituents' preferences as betweenthreealternatives: A, B, and no-
agreement. If45 percentoftheconstituents ranktheseA > no-agreement >
B, 45 percentrankthemB > no-agreement > A, and 10percentrankthem
B > A > no-agreement, thenbothA and B are in thewin-set,eventhough
B wouldwinina simpleLevel-Il-only game.On theotherhand,if90 percent
rankthe alternatives A > no-agreement > B, while10 percentstillrank
themB > A > no-agreement, thenonlyA is in thewin-set.In thissense,
a government dividedis morelikelyto be able to strikea

thatis internally
thanone thatis firmly
deal internationally committed to a singlepolicy.47
Conversely,to imposebindingex ante instructions on the negotiators in
sucha case mightexcludesome Level I outcomesthatwould,in fact,be
in bothnations.48
ratifiable
Thus farwe have implicitly assumedthatall eligibleconstituents will
participate process.In fact,however,participation
in theratification rates
varyacrossgroupsand acrossissues,and thisvariationoftenhas implica-
tionsforthesize ofthewin-set.For example,whenthecostsand/or benefits
of a proposedagreementare relativelyconcentrated, it is reasonableto
expectthatthoseconstituents whoseinterests are mostaffected willexert
specialinfluence process.49
ontheratification OnereasonwhyLevelII games
are moreimportant fortradenegotiations thanin monetary mattersis that
the "abstentionrate" is higheron internationalmonetary issues thanon
trade issues.50
The composition oftheactiveLevel II constituency (andhencethechar-
acterof thewin-set)also varieswiththepoliticization of theissue. Politi-
cizationoftenactivatesgroupswhoare less worriedaboutthecostsofno-
agreement, thusreducingtheeffective win-set.For example,politicization
ofthePanamaCanal issue seemsto havereducedthenegotiating flexibility
on bothsidesofthediplomatic Thisis one reasonwhymostprofes-
table.51
sionaldiplomatsemphasizethevalueofsecrecyto successfulnegotiations.
However,WoodrowWilson's transcontinental tourin 1919reflected the
oppositecalculation,namely,thatby expanding theactiveconstituency he
of theVersaillesTreaty,althoughin theend this
could ensureratification
strategyprovedfruitless.52
Anotherimportant of our discussionthusfarhas been the
restriction
47. Raiffanotesthat"themorediffusethepositions each side,theeasieritmight

are within
be to achieveexternalagreement."(Raiffa,Artand ScienceofNegotiation, p. 12.) For the
conventional thatdomestic
view,bycontrast, forinternational
a precondition
unityis generally
agreement,see Michael Artisand Sylvia Ostry,InternationalEconomic Policy Coordination,
ChathamHouse Papers:30 (London:Routledge& KeganPaul, 1986),pp. 75-76.
48. "Meaningful withothernationsbecomesverydifficult
consultation whentheinternal
processofdecision-makingalreadyhas someofthecharacteristicsofcompactsbetweenquasi-
sovereign Thereis an increasing
entities. to hazarda hard-won
reluctance domesticconsensus
forum."HenryA. Kissinger,
inan international "DomesticStructure andForeignPolicy,"in
James N. Rosenau, ed., InternationalPolitics and Foreign Policy (New York: Free Press,
1969),p. 266.
(New York:Basic Books,1975)on howthe
49. See JamesQ. Wilson,PoliticalOrganization
by whetherthecosts and thebenefits
politicsof an issue are affected or
are concentrated
diffuse.
50. Anotherfactorfostering is thegreatercomplexity
abstention and opacityof monetary
issues;as GilbertR. Winham ("Complexity inDanielDruckman,
Negotiation,"
inInternational
ed., Negotiations: A Social-Psychological Perspective [Beverly Hills: Sage, 1977], p. 363)
thehandof a negotiator
can strengthen
observes,"complexity he
vis-a-vistheorganization
represents."
51. Habeeb and Zartman,Panama Canal Negotiations.
52. Bailey, Wilsonand the Great Betrayal.
446 International
Organization
assumption thatthenegotiations involveonlyone issue. Relaxingthisas-

sumption has powerful consequencesfortheplayat bothlevels.53Various
groupsat Level II arelikelytohavequitedifferent preferences ontheseveral
issues involvedin a multi-issue negotiation. As a generalrule,the group
withthegreatestinterestin a specificissue is also likelyto holdthemost
extremepositionon thatissue. In the Law of the Sea negotiations, for
example,the DefenseDepartment feltmoststrongly aboutsea-lanes,the
Department oftheInterior aboutsea-bedmining rights, and so on.54Ifeach
groupis allowedto fixtheLevel I negotiating positionfor"its" issue,the
resulting packageis almostsureto be "non-negotiable" (thatis, non-rati-
fiablein opposingcapitals).55
Thus,thechiefnegotiator is facedwithtradeoffs acrossdifferent issues:
howmuchto yieldon mining rightsin orderto getsea-laneprotection, how
muchto yieldon citrusexportsto geta betterdeal on beef,and so on. The
implication ofthesetradeoffs fortherespectivewin-sets can be analyzedin
termsofiso-voteor "politicalindifference" curves.Thistechniqueis anal-
ogous to conventional indifference curveanalysis,exceptthatthe opera-
tionalmeasureis voteloss, notutility loss. Figure2 providesan illustrative
Edgeworth box analysis.56The most-preferred outcomeforA (theoutcome
whichwinsunanimousapprovalfromboththebeefindustry and thecitrus
industry) is theupperright-hand corner(AM),and each curveconcave to
pointAMrepresents thelocusofall possibletradeoffs betweentheinterests
ofranchers andfarmers, suchthatthenetvoteinfavorofratification at A's
Level II is constant.The bold contourA1-A2represents theminimalvote
necessaryforratification by A, and the wedge-shapedarea northeastof
A1-A2represents A's win-set.Similarly, B1-B2represents theoutcomesthat
are minimally ratifiable
by B, and thelens-shaped area betweenA1-A2and
B1-B2represents the set of feasibleagreements. Althoughadditionalsub-
tleties(suchas thenatureofthe"contractcurve")might be extracted from
thissortof analysis,thecentralpointis simple:thepossibility of package
deals opensup a richarrayofstrategic alternatives
fornegotiators in a two-
levelgame.
One kindofissue linkageis absolutelycrucialto understanding how do-
mesticand international politicscan becomeentangled.57 Suppose thata
majorityof constituents at Level II oppose a givenpolicy(say, oil price
53. I am grateful
to ErnstB. Haas and Robert0. Keohaneforhelpful
adviceon thispoint.
54. Ann L. Hollick, U.S. Foreign Policy and the Law of the Sea (Princeton: Princeton
University
Press,1981),especiallypp. 208-37,andJamesK. Sebenius,Negotiating theLaw
oftheSea (Cambridge,
Mass.: HarvardUniversity Press,1984),especiallypp. 74-78.
55. Raiffa,Art and Science of Negotiation, p. 175.
56. I amindebtedto Lisa MartinandKennethShepsleforsuggestingthisapproach,although
theyarenotresponsible formyapplication ofit.Notethatthisconstruction
assumesthateach
issue,takenindividually,
is a "homogeneous" type,nota "heterogeneous"type.Constructing
iso-votecurvesforheterogeneous-type issuesis morecomplicated.
57. I am grateful
to HenryBradyforclarifying thispointforme.
A1 Am
B1
A2
BM B2
curvesfor two-issuenegotiation
FIGURE 2. Political indifference
decontrol),butthatsomemembers ofthatmajoritywouldbe willing toswitch

theirvote on thatissue in returnformorejobs (say, in exportindustries).
If bargainingis limitedto Level II, thattradeoff
is nottechnically feasible,
butifthechiefnegotiator can brokeran internationaldealthatdeliversmore
jobs (say, via fastergrowthabroad),he can, in effect,overturn theinitial
outcomeat the domestictable. Such a transnational issue linkagewas a
crucialelementin the 1978Bonnaccord.
Note thatthis strategyworksnot by changingthe preferences of any
domesticconstituents, butratherbycreating a policyoption(suchas faster
exportgrowth) thatwas previously beyonddomesticcontrol.Hence,I refer
to thistypeof issue linkageat Level I thataltersthefeasibleoutcomesat
Level II as synergisticlinkage. For example, "in the Tokyo Round . ..
nationsused negotiation to achieveinternal reform in situationswherecon-
stituency pressureswould otherwisepreventactionwithoutthe pressure
448 International
Organization
benefits)
(and tradeoff thatan externalpartner 58 Economic
couldprovide.'"
interdependencemultipliestheopportunities domesticcoalitions
foraltering
(and thuspolicyoutcomes)by expandingtheset offeasiblealternatives in
thisway-in effect, politicalentanglements
creating acrossnationalbound-
aries. Thus,we shouldexpectsynergistic linkage(whichis, by definition,
explicableonlyin termsoftwo-levelanalysis)to becomemorefrequent as
interdependencegrows.
2. The size of the win-setdepends on the Level II

political institutions.
Ratificationprocedures clearlyaffectthesize ofthewin-set.Forexample,
ifa two-thirds voteis requiredforratification, thewin-setwillalmostcer-
tainlybe smallerthanifonlya simplemajority is required.As one experi-
encedobserverhas written: "UndertheConstitution, oftheone
thirty-four
hundredsenatorscan blockratification of any treaty.This is an unhappy
and uniquefeatureofourdemocracy.Because oftheeffective vetopower
of a smallgroup,manyworthyagreements have been rejected,and many
treatiesare neverconsideredforratification."59 As notedearlier,theU.S.
separationofpowersimposesa tighter constraint on theAmericanwin-set
thanis truein manyothercountries.This increasesthebargaining power
of Americannegotiators, but it also reducesthe scope forinternational
cooperation. It raisestheoddsforinvoluntary defection andmakespotential
partners warieraboutdealingwiththeAmericans.
The TradeExpansionAct of 1974modified U.S. ratificationprocedures
inan efforttoreducethelikelihood ofcongressional tampering withthefinal
deal and henceto reassureAmerica'snegotiating partners. AftertheAmer-
ican SellingPricefiasco,it was widelyrecognizedthatpiecemealcongres-
sionalratificationof anynewagreement wouldinhibit internationalnegoti-
ation.Hence,the1974Actguaranteed a straightup-or-down voteinCongress.
However,to satisfycongressional an elaboratesystemofpri-
sensitivities,
vate-sectorcommittees was established toimprovecommunication between
the Level I negotiators and theirLevel II constituents, in effectcoopting
the interestgroupsby exposingthemdirectlyto theimplications of their
demands.60 Preciselythistacticis describedin thelabor-management case
byWaltonand McKersie:"Insteadoftakingresponsibility fordirectlyper-
II
suadingtheprincipals[Level constituents] to reducetheirexpectations,
[theLevel I negotiator] structures thesituationso thatthey(or theirmore
immediate representatives) willpersuadethemselves."61
58. GilbertR. Winham,"The Relevanceof Clausewitzto a Theoryof International
Nego-
tiation,"preparedfordeliveryat the 1987annualmeeting
of theAmericanPoliticalScience
Association.
59. JimmyCarter,Keeping Faith: Memoirsof a President(New York: Bantam Books, 1982),
p. 225.
60. Winham(see note37); Twiggs,TheTokyoRound.
61. Walton and McKersie, Behavioral Theoryof Labor Organizations,p. 321.
Not all significant ratificationpracticesare formalized; forexample,the

Japanesepropensity forseekingthebroadestpossibledomesticconsensus
beforeactingconstricts theJapanesewin-set, as contrastedwithmajoritarian
politicalcultures.Otherdomesticpoliticalpractices,too,can affect thesize
of thewin-set.Strongdisciplinewithinthegoverning party,forexample,
increasesthe win-setby wideningtherangeof agreements forwhichthe
Level I negotiator can expectto receivebacking.For example,in the 1986
House-Senateconference committee on taxreform, thefinalbillwas closer
to theSenateversion,despite(or rather, becauseof) Congressman Rosten-
kowski'sgreatercontrolofhis delegation, whichincreasedtheHouse win-
set. Conversely, a weakeningofpartydisciplineacrossthemajorWestern
nationswould,ceterisparibus,reducethe scope forinternational cooper-
ation.
The recentdiscussionof "state strength" and "stateautonomy"is rel-
evanthere.The greatertheautonomy ofcentraldecision-makers fromtheir
Level II constituents, thelargertheirwin-setand thusthegreaterthelike-
lihoodof achievinginternational agreement. For example,centralbankin-
sulationfromdomesticpoliticalpressuresineffect increasesthewin-setand
thustheodds forinternational monetary cooperation;recentproposalsfor
an enhancedrole forcentralbankersin international policycoordination
reston thispoint.62 However,two-levelanalysisalso impliesthat,ceteris
paribus,thestronger a stateis in termsof autonomyfromdomesticpres-
sures,the weakerits relativebargaining positioninternationally. For ex-
ample,diplomatsrepresenting an entrenched dictatorshipareless able than
representatives of a democracyto claimcrediblythatdomesticpressures
precludesomedisadvantageous deal.63Thisis yetanotherfacetof thedis-
concerting ambiguity ofthenotionof "statestrength."
For simplicity ofexposition, myargument is phrasedthroughout in terms
ofonlytwolevels.However,manyinstitutional arrangements requireseveral
levelsofratification, thusmultiplying thecomplexity (butperhapsalso the
importance) ofwin-set analysis.Consider,forexample,negotiations between
the UnitedStates and the EuropeanCommunity over agriculturaltrade.
According totheTreatyofRome,modifications oftheCommonAgricultural
Policyrequireunanimousratification by the Councilof Ministers,repre-
senting each ofthememberstates.In turn,eachofthosegovernments must,
in effect,winratification foritsdecisionwithinitsownnationalarena,and
in coalitiongovernments, thatprocessmightalso requireratification within
each of theparties.Similarly, on theAmericanside,ratification would(in-
formally, at least) necessitatesupportfrommost,if not all, of the major
agricultural organizations, andwithin thoseorganizations,furtherratification
by key interestsand regionsmightbe required.At each stage,cleavage
patterns, issue linkages,ratification procedures,side-payments, negotiator
62. Artis and Ostry,InternationalEconomic Policy Coordination. Of course, whetherthis
is desirablein termsofdemocratic
valuesis quiteanothermatter.
63. Schelling,Strategyof Conflict,p. 28.
450 International
Organization
and so on wouldneed to be considered.At some pointin this

strategies,
of further
analyticregressthe complexity decomposition wouldoutweigh
theadvantages,buttheexampleillustrates
theneedforcarefulthoughtabout
thelogicofmultiple-level
games.
3. The size of the win-setdepends on the strategiesof the

Level I negotiators.
Each Level I negotiator has an unequivocalinterestin maximizing the
otherside's win-set,butwithrespectto his own win-set,his motivesare
mixed.The largerhiswin-set, themoreeasilyhe canconcludean agreement,
but also theweakerhis bargaining positionvis-a-vistheothernegotiator.
Thisfactoftenposes a tacticaldilemma.For example,one effective wayto
demonstrate commitment to a givenpositioninLevel I bargaining is to rally
supportfromone's constituents (forexample,holdinga strikevote,talking
abouta "missilegap," or denouncing "unfairtrading practices"abroad).
On theotherhand,suchtacticsmayhaveirreversible effects
on constituents'
attitudes,hampering subsequentratification of a compromise agreement.f4
Conversely,preliminary consultations at home,aimedat "softening up"
one's constituentsin anticipation of a ratification
struggle,can undercuta
negotiator'sabilityto projectan implacableimageabroad.
Nevertheless, disregarding thesedilemmasforthemomentand assuming
thata negotiator wishes expandhis win-setin orderto encourageratifi-
to
cationof an agreement, he may exploitbothconventional side-payments
andgeneric"good will." The use ofside-payments to attractmarginal sup-
portersis, of course,quitefamiliar in gametheory,as well as in practical
politics.For example,theCarterWhiteHouse offered manyinducements
(suchas publicworksprojects)to helppersuadewavering Senatorsto ratify
the Panama Canal Treaty.65 In a two-levelgame the side-payments may
come fromunrelateddomesticsources,as in thiscase, or theymay be
receivedas partoftheinternational negotiation.
The role of side-payments in internationalnegotiations is well known.
However,thetwo-levelapproachemphasizesthatthevalue of an interna-
tionalside-payment shouldbe calculatedintermsofitsmarginal contribution
to thelikelihoodof ratification,ratherthanin termsof its overallvalue to
therecipient nation.Whatcountsat Level II is nottotalnationalcostsand
benefits,but their incidence, relative to existing coalitions and proto-
tradeconcession(orstillworse,a concession
coalitions.Anacross-the-board
on a productof interestto a committed free-tradecongressman) is less
thana concession(evenone oflesserintrinsic
effective value)thattipsthe
balancewitha swingvoter.Conversely,traderetaliationshouldbe targeted,
64. Walton and McKersie, Behavioral Theoryof Labor Organizations,p. 345.

65. Carter,Keeping Faith, p. 172. See also Raiffa,Art and Science of Negotiation, p. 183.
neitherat free-tradersnor at confirmed protectionists, but at the uncom-

mitted.
An experiencednegotiator familiarwiththe respectivedomestictables
shouldbe able to maximizethecost-effectiveness (to himand his constit-
uents)of theconcessionsthathe mustmaketo ensureratification abroad,
as wellas thecost-effectivenessofhisowndemandsandthreats, bytargeting
withan eye to theirLevel II incidence,bothat homeand
his initiatives
abroad.In thisendeavorLevel I negotiators are oftenin collusion,since
each has an interestin helpingthe otherto get the finaldeal ratified. In
effect,theyare movingjointlytowardspointsof tangencybetweentheir
respectivepoliticalindifference curves.The empiricalfrequency of such
targetingintradenegotiationsandtradewars,as wellas inotherinternational
negotiations,wouldbe a crucialtestof therelativemeritsof conventional
unitary-actoranalysisand thetwo-levelapproachproposedhere.66
In additionto theuse of specificside-payments, a chiefnegotiator whose
politicalstandingat homeis highcan moreeasilywin ratification of his
foreigninitiatives.
Although genericgoodwillcannotguarantee ratification,
as WoodrowWilsondiscovered,it is usefulin expandingthewin-setand
thusfostering Level I agreement, forit constitutesa kindof "all-purpose
glue" forhis supporting coalition.Waltonand McKersiecite membersof
the UnitedAuto Workerswho, speakingof theirreveredleader,Walter
Reuther,said,"I don'tunderstand oragreewiththisprofit-sharing idea,but
iftheRed Head wantsit,I willgo along."67 The Yugoslavnegotiator in the
Triestedisputelaterdiscounted thedifficultyofpersuading irredentistSlov-
enesto accepttheagreement, since"thegovernment [i.e., Tito]can always
influence publicopinionifit wantsto."68
Note thateach Level I negotiator has a stronginterest in thepopularity
ofhis oppositenumber,sincePartyA's popularity increasesthesize ofhis
win-set,and thusincreasesboththeodds of success and therelativebar-
gainingleverageofPartyB. Thus,negotiators shouldnormally be expected
to tryto reinforce
one another'sstanding withtheirrespective constituents.
significance
66. The strategic oftargeting inJohnConybeare,
at LevelII is illustrated "Trade
Wars:A Comparative StudyofAnglo-Hanse, Franco-Italian, and Hawley-Smoot Conflicts,"
WorldPolitics38 (October1985),p. 157:Retaliation in theAnglo-Hanse tradewarsdid not
havetheintended deterrent becauseitwasnot(andperhapscouldnothavebeen)targeted
effect,
atthecrucialmembers oftheopposingLevelII coalition.CompareSnyderandDiesing,Conflict
AmongNations,p. 552:"Ifonefacesa coerciveopponent, buttheopponent'smajority coalition
includesa fewwavering members inclinedto compromise, a compromise proposalthatsuits
theirviewsmaycause theirdefection andtheformation ofa differentmajoritycoalition.Or if
theopponent'sstrategy is accommodative, basedon a tenuoussoft-line coalition,one knows
thatcare is requiredin implementing one's owncoercivestretegy to avoidtheoppositekind
ofshiftin theotherstate."
67. Walton and McKersie, Behavioral Theoryof Labor Negotiations, p. 319.
68. VladimirVelebit,in Campbell,Trieste1954,p. 97. As notedearlier,ourdiscussionhere
agreement;
wishesto reacha ratifiable
assumesthattheLevel I negotiator in cases (alluded
ownpreferences
to later)whenthenegotiator's are morehard-linethanhis constituents,his
mightallowhimto resistLevel I agreements.
domesticpopularity
Organization
452 International
Partlyforthisreasonandpartlybecauseofmediaattention, participation
on theworldstagenormally givesa headofgovernment a specialadvantage
vis-a-vishisor herdomesticopposition.Thus,although internationalpolicy
coordination is hamperedby hightransaction costs,heads of government
mayalso reapwhatwe might term"transaction benefits."Indeed,therecent
evolutionofWesternsummitry, whichhas placedgreateremphasison pub-
licitythanon substance,seemsdesignedto appropriate these"transaction
benefits"withoutactuallyseekingthesortof agreements thatmightentail
transactioncosts.69
are likelyto disposeofmoreside-payments
Higherstatusnegotiators and
more"good will" at home,and henceforeigners preferto negotiatewitha
head ofgovernment thanwitha lowerofficial. In purelydistributive
terms,
a nationmighthave a bargaining advantageif its chiefnegotiator were a
mereclerk.Diplomatsare actingrationally, notmerelysymbolically, when
theyrefuseto negotiatewitha counterpart rank.America'sne-
of inferior
gotiatingpartnershavereasonforconcernwhenever theAmerican president
is domesticallyweakened.
tactics
and bargaining
Uncertainty
Level I negotiatorsare oftenbadlymisinformed about Level II politics,
on theopposing
particularly side. In 1978,the Bonn negotiatorswereusually
wrongin theirassessmentsof domesticpoliticsabroad; example,most
for
AmericanofficialsdidnotappreciatethecomplexdomesticgamethatChan-
cellorSchmidtwas playingover theissue of Germanreflation. Similarly,
Snyderand Diesingreportthat"decisionmakersin our cases onlyocca-
sionallyattempted such assessments,and whentheytriedtheydid pretty
miserably.... Governments generallydo not do well in analyzingeach
other'sinternalpoliticsin crises[and,I wouldadd, in normaltimes],and
indeeditis inherentlydifficult.
"70 Relaxing theassumption ofperfect infor-
mationto allowforuncertainty has manyimplications forourunderstanding
oftwo-levelgames.Let me illustrate a fewoftheseimplications.
Uncertaintyaboutthesize of a win-setcan be botha bargaining device
and a stumblingblockin two-level negotiation. In purelydistributive Level
I bargaining,
negotiatorshavean incentive to understatetheirownwin-sets.
Since each negotiatoris likelyto knowmoreabouthis own Level II than
his opponentdoes, theclaimhas someplausibility. This is akinto a tactic
69. Transaction
benefitsmaybe enhancedifa substantiveagreement is reached,although
sometimes leaderscan benefit
domestically
byloudlyrejecting
a profferedinternational
deal.
70. SnyderandDiesing,Conflict
AmongNations,pp. 516,522-23.Analogousmi*perceptions
in Anglo-American diplomacyare thefocusof RichardE. Neustadt,AlliancePolitics(New
York:ColumbiaUniversity Press,1970).
thatSnyderandDiesingdescribe,whennegotiators seektoexploitdivisions
withintheirown government by saying,in effect,"You'd bettermakea
deal withme,because thealternative to me is evenworse.""71
On the otherhand,uncertainty aboutthe opponent'swin-setincreases
one's concernabout the riskof involuntary defection.Deals can onlybe
struckifeach negotiator is convincedthattheproposeddeal lies withinhis
oppositenumber'swin-setandthuswillbe ratified. Uncertainty aboutparty
A's ratification lowerstheexpectedvalueoftheagreement to partyB, and
thuspartyB willdemandmoregenerousside-payments frompartyA than
would be needed underconditionsof certainty. In fact,partyB has an
incentive to feigndoubtaboutpartyA's abilityto deliver,preciselyinorder
to extracta moregenerousoffer.72
Thus,a utility-maximizing negotiator mustseek to convincehis opposite
numberthathis own win-setis "kinky,"thatis, thattheproposeddeal is
certainto be ratified, butthata deal slightlymorefavorableto theopponent
is unlikelytobe ratified. Forexample,ontheenergy issuein 1978,bysending
SenatorByrdon a personalmissionto Bonnbeforethesummit andthenby
in
discussinghispoliticalproblems length a with
tete-'a-tete the chancellor,
Cartersoughtsuccessfully to convince Schmidt that immediate decontrol
was politicallyimpossible,but thatdecontrolby 1981 was politicallydo-
if
able. Kinkywin-setsmaybe morecredible theypivot on what Schelling
calls a "prominent"solution,suchas a 50-50 split,forsuchoutcomesmay
be distinctly more"saleable" at home.Anotherrelevanttacticis forthe
negotiator actuallyto submita trialagreement forratification, in orderto
demonstrate that it is notin his win-set.
Uncertainty aboutthe contoursof therespective"politicalindifference
curves" thushas strategic uses. On theotherhand,whenthenegotiators
are seekingnovelpackagesthatmightimprovebothsides' positions,mis-
representation ofone's win-setcanbe counterproductive. Creativesolutions
thatexpandthescopeforjointgainandimprovetheodds ofratification are
likelyto requirefairlyaccurateinformation aboutconstituents' preferences
andpointsofspecialneuralgia.The analysisoftwo-level gamesoffers many
illustrations
ofZartman'sobservation thatall negotiationinvolves"thecon-
trolledexchangeofpartialinformation. 73
71. Synderand Diesing, ConflictAmong Nations, p. 517.

72. I am gratefulto Robert0. Keohaneforpointing out theimpactof uncertainty on the
expectedvalueofproposals.
73. I. WilliamZartman,The50% Solution(GardenCity,N.J.:AnchorBooks,1976),p. 14.
The presentanalysisassumesthatconstituentsare myopicabouttheotherside'sLevel II, an
assumption thatis notunrealistic
empirically.
However,a fullyinformed constituentwould
considerthepreferences ofkeyplayerson theotherside,forifthecurrent proposallies well
within theotherside's win-set,
thenitwouldbe rationalfortheconstituent to voteagainstit,
hopingfora second-round proposalthatwas morefavorableto himand stillratifiable
abroad;
thismightbe a reasonableinterpretation
of SenatorLodge's positionin 1919(Bailey,Wilson
and theGreatBetraya[). Consideration
ofsuchstrategic
votingat LevelII is beyondthescope
ofthisarticle.
Organization
454 International
and reverberation
Restructuring
Formallyspeaking,game-theoretic analysisrequiresthatthe structure of
issues and payoffsbe specifiedin advance.In reality,however,muchof
whathappensin anybargaining situationinvolvesattempts by theplayers
to restructure thegameand to alterone another'sperceptions ofthecosts
ofno-agreement and thebenefits ofproposedagreements. Such tacticsare
moredifficult intwo-levelgamesthaninconventional negotiations,because
itis harderto reachconstituentson theothersidewithpersuasivemessages.
Nevertheless, governments do seektoexpandoneanother'swin-sets.Much
ambassadorialactivity-wooingopinionleaders,establishing contactwith
oppositionparties,offering foreignaid to a friendly,butunstablegovern-
ment,and so on-has preciselythisfunction. WhenJapaneseofficials visit
CapitolHill, or BritishdiplomatslobbyIrish-American leaders,theyare
seekingtorelaxdomesticconstraints thatmight otherwise prevent theadmin-
istrationfromcooperating withtheirgovernments.
Anotherilluminating exampleof actionsby a negotiator at theopposing
Level II to improvethe odds of ratification occurredduringthe 1977ne-
gotiationsbetweentheInternational Monetary FundandtheItaliangovern-
ment.InitialIMF demandsforausterity triggered strongoppositionfrom
theunionsand left-wing parties.Although theIMF's bargaining positionat
Level I appearedstrong, theFund's negotiator soughtto achievea broader
consensuswithinItalyin supportof an agreement, in orderto forestall
involuntary defection.Accordingly, afterdirectconsultations withtheunions
leaders,theIMF restructured
and leftist itsproposalto focuson long-term
investment and economicrecovery(incidentally, an interestingexampleof
without
targeting), backingofffromits short-term demands.Ironically, the
initialCommunist supportforthisrevisedagreement subsequently collapsed
because of conflictsbetweenmoderateand doctrinaire factionswithinthe
party,illustratingtheimportance ofmultilevel analysis.74
In someinstances,perhapseven unintentionally, internationalpressures
"reverberate"withindomesticpolitics,tippingthe domesticbalance and
thusinfluencing the internationalnegotiations. Exactlythiskindof rever-
berationcharacterized the 1978summit negotiations.DieterHiss, theGer-
mansherpaand one ofthosewhobelievedthata stimulus program was in
Germany'sowninterest, laterwrotethatsummits changenationalpolicy
onlyinsofaras theymobilizeand/orchangepublicopinionand theatti-
tude of political groups.... Oftenthatis enough, if the balance of
74.JohnR. Hillman,"The MutualInfluence ofItalianDomesticPoliticsandtheInternational

MonetaryFund,"TheFletcher Forum4 (Winter 1980),pp. 1-22.LuigiSpaventa,"Two Letters
Policy,Italy,1973-77,"in JohnWilliAmson,
ofIntent:ExternalCrisesand Stabilization ed.,
(Washington,
IMF Conditionality forInternational
D.C.: Institute Economics,1983),pp. 441-73,
arguesthattheunionsandtheCommunists actuallyfavoredtheausterity measures,butfound
theIMF demandshelpful in dealingwiththeirowninternal Level II constituents.
opinionis shifted, a baremajority

providing stymied
forthepreviously
actions of a strongminority.... No countryviolates its own interests,
ofitsinterests
thedefinition
butcertainly a summit
can changethrough
and give-and-take.75
withitspossibletradeoffs
Fromthepointofviewoforthodoxsocial-choicetheory, reverberation is
problematic, forit impliesa certaininterconnectedness amongthe utility
functions of independent actors,albeitacrossdifferent levelsof thegame.
Two rationalesmay be offeredto explainreverberation amongutility-
maximizing egoists.First,ina complex,interdependent, butoftenunfriendly
world,offending foreignersmaybe costlyin thelongrun."To getalong,
go along" may be a rationalmaxim.This rationaleis likelyto be more
commonthemoredependent(or interdependent) a nation,and it is likely
to be morepersuasiveto Level II actorswho are moreexposedinterna-
tionally,suchas multinational corporationsand international banks.
A secondrationaletakesintoaccountcognitive factorsand uncertainty.
It would be a mistakeforpoliticalscientiststo mimicmosteconomists'
disregardforthe suasive elementin negotiations.76Given the pervasive
uncertainty thatsurrounds manyinternationalissues,messagesfromabroad
can changeminds,movetheundecided,and heartenthosein thedomestic
minority. As one reluctant Germanlatecomerto the"locomotive"cause in
1978explainedhisconversion, "In theend,eventheBankforInternational
Settlements [thecautiousBasle organization of centralbankers]supported
theidea ofcoordinated an enthusiastic
relation."Similarly, advocateofthe
programwelcomedthe international pressureas providing a useful"tail-
wind"in Germandomesticpolitics.
Suasivereverberation is morelikelyamongcountries withclose relations
and is probablymorefrequent in economicthanin political-military nego-
tiations.Communiques fromthe Westernsummitsare oftencitedby par-
ticipantsto domesticaudiencesas a wayoflegitimizing theirpolicies.After
one suchstatement by ChancellorSchmidt,one ofhisaidesprivately char-
acterizedtheargument valid,butpolitically
as "not intellectually useful."
Conversely, it is widelybelievedby summit participants thata declaration
contrary to a government's currentpolicycould be used profitably by its
opponents.Recentcongressional proposalsto ensuregreaterdomesticpub-
licityforinternational commentary on nationaleconomicpolicies(including
hithertoconfidential IMF recommendations) turnon theidea thatreverber-
ationmightincreaseinternational cooperation.77
75.DieterHiss,"Weltwirtschaftsgipfel:BetrachtungeneinesInsiders[WorldEconomicSum-
mit:Observations of an Insider],"in JoachimFrohnand ReinerStaeglin,eds., Empirische
Wirtschaftsforschung (Berlin:Dunckerand Humblot,1980),pp. 286-87.
76. On cognitiveand communications of international
explanations see, for
cooperation,
example,ErnstB. Haas, "WhyCollaborate? andInternational
Issue-Linkage Regimes,"World
Politics32 (April1980),pp 357-405;RichardN. Cooper,"International inPublic
Cooperation
Healthas a Prologueto Macroeconomic Cooperation,"BrookingsDiscussionPapers in In-
ternational Economics44 (Washington, D.C.: Brookings 1986);andZartman,
Institution, 50%
Solution,especiallyPart4.
77. Henning,Macroeconomic Diplomacy in the 1980s, pp. 62-63.
456 International
Organization
Reverberation as discussedthusfarimpliesthatinternational pressure

expandsthe domesticwin-setand facilitates agreement. However,rever-
berationcan also be negative,in thesensethatforeign pressuremaycreate
a domesticbacklash.Negativereverberation is probablyless commonem-
piricallythanpositivereverberation, simplybecauseforeigners are likelyto
forgopublicpressureifitis recognizedto be counterproductive. Cognitive
balancetheorysuggeststhatinternational pressureis morelikelyto rever-
beratenegatively ifitssourceis generally viewedbydomesticaudiencesas
an adversaryratherthanan ally.Nevertheless, thepreciseeffect
predicing
offoreign pressureis admittedly although
difficult, reverberation
empirically,
seemsto occurfrequently in two-levelgames.
The phenomenon ofreverberation (alongwithsynergistic issuelinkageof
the sortdescribedearlier)precludesone attractive short-cut to modeling
two-levelgames.If nationalpreferences wereexogenousfromthepointof
view of international relations,thenthe domesticpoliticalgamecould be
moldedseparately, and the"outputs"fromthatgamecouldbe used as the
"inputs"to theinternational game.78The divisionof laborbetweencom-
parativepoliticsand international relationscould continue,thougha few
curiousobserversmightwishto keeptrackoftheplayon bothtables.But
ifinternationalpressuresreverberate withindomesticpolitics,or ifissues
can be linkedsynergistically, thendomesticoutcomesare notexogenous,
and thetwolevelscannotbe modeledindependently.
The role of the chiefnegotiator
In thestylizedmodelof two-levelnegotiationsoutlinedhere,thechiefne-
gotiatoris theonlyformallinkbetweenLevel I and Level II. Thus far,I
haveassumedthatthechiefnegotiator has no independentpolicyviews,but
acts merelyas an honestbroker,or ratheras an agenton behalfof his
constituents.Thatassumption simplifies
powerfully theanalysisoftwo-level
games.However,as principal-agent theoryremindsus, thisassumption is
thepreferences
Empirically,
unrealistic.79 ofthechiefnegotiator maywell
divergefromthoseofhisconstituents.Two-levelnegotiationsarecostlyand
78. Thisis theapproachused to analyzetheAnglo-Chinese overHongKong

negotiations
in Bruce Bueno de Mesquita,David Newman,and AlvinRabushka,ForecastingPolitical
Events: The Future of Hong Kong (New Haven: Yale UniversityPress, 1985).
see TerryM. Moe, "The New Economicsof Organi-
79. For overviewsof thisliterature,
zation," AmericanJournalof Political Science 28 (November 1984),pp. 739-77; JohnW. Pratt
and Richard J. Zeckhauser, eds., Principals and Agents: The Structureof Business (Boston,
Mass.: HarvardBusinessSchoolPress,1985);andBarryM. Mitnick, "The TheoryofAgency
andOrganizationalAnalysis,"prepared fordeliveryatthe1986annualmeeting American
ofthte
is onlyindirectly
PoliticalScienceAssociation.Thisliterature to ourconcernshere,
relevant
forithas notyetadequatelyaddressedtheproblems posedbymultiple (or constit-
principals
uents,inourterms).Forone highly formalapproachto theproblem ofmultiple
principals,see
R. DouglasBernheim and MichaelD. Whinston, "CommonAgency,"Econometrica 54 (July
1986),pp. 923-42.
riskyforthechiefnegotiator, and theyofteninterfere

withhis otherprior-
ities,so it is reasonableto ask whatis in itforhim.
The motivesofthechiefnegotiator include:
1. Enhancinghis standing in theLevel II gameby increasing his politi-
cal resourcesor by minimizingpotentiallosses. For example,a head
ofgovernment mayseek thepopularity thathe expectsto accrueto
himifhe concludesa successfulinternational agreement, or he may
anticipatethattheresultsof theagreement (forexample,faster
growthor lowerdefensespending)willbe politically rewarding.
2. Shiftingthebalanceofpowerat Level II in favorofdomesticpoli-
cies thathe prefersforexogenousreasons.International negotiations
sometimes enablegovernment leadersto do whattheyprivately
wish
to do, butare powerlessto do domestically. Beyondthenow-familiar
1978case, thispatterncharacterizes manystabilizationprogramsthat
are (misleadingly)
said to be "imposed"by theIMF. For example,in
the 1974and 1977negotiations betweenItalyand theIMF, domestic
conservative forcesexploitedIMF pressureto facilitate
policymoves
thatwereotherwiseinfeasible internally.80
3. To pursuehis ownconceptionofthenationalinterest
in theinterna-
tionalcontext.This seemsthebestexplanation
ofJimmy Carter's
prodigious on behalfofthePanamaCanal Treaty,as wellas
efforts
of WoodrowWilson'sultimately fatalcommitment
to theVersailles
Treaty.
It is reasonableto presume,at leastin theinternationalcase oftwo-level
bargaining, thatthechiefnegotiator willnormally giveprimacyto his do-
mesticcalculus,ifa choicemustbe made,notleastbecausehisownincum-
bencyoftendependson his standingat Level II. Hence, he is morelikely
to presentan international agreement forratification,
the less of his own
politicalcapitalhe expectsto havetoinvesttowinapproval,andthegreater
thelikelypoliticalreturns froma ratifiedagreement.
Thisexpandedconceptionof theroleof thechiefnegotiator impliesthat
he has, in effect, a vetooverpossibleagreements. Even ifa proposeddeal
lieswithin hisLevel II win-set,thatdealis unlikely
tobe struckifhe opposes
it.81Since thisprovisoapplieson bothsidesoftheLevel I table,theactual
international bargaining set maybe narrower-perhaps muchnarrower-
thantheoverlapbetweentheLevel II win-sets.Empirically, thisadditional
constraint is oftencrucialto theoutcomeoftwo-levelgames.One momen-
tousexampleis thefateoftheVersaillesTreaty.Thebestevidencesuggests,
first, thatperhaps80 percentof theAmericanpublicand of theSenatein
1919favoredratification ofthetreaty, ifcertainreservationswereattached,
and second,thatthosereservations wereacceptableto theotherkey sig-
80. Hillman,"MutualInfluence,"and Spaventa,"Two LettersofIntent."
81. Thispowerofthechiefnegotiator is analogousto whatShepsleandWeingasttermthe
"penultimate"or "ex postveto" powerofthemembers ofa Senate-Houseconference
com-
mittee.(Shepsleand Weingast,"Institutional
FoundationsofCommittee Power.")
Organization
458 International
natories,especiallyBritainandFrance.In effect, itwas Wilsonhimself who

vetoed this otherwise ratifiable
package, tellingthe dismayed French Am-
bassador,"I shallconsentto nothing."82
Yet anotherconstraint on successfultwo-levelnegotiation derivesfrom
the leader's existingdomesticcoalition.Anypoliticalentrepreneur has a
fixedinvestment in a particularpatternofpolicypositionsand a particular
supporting coalition.If a proposedinternational deal threatens thatinvest-
ment,or ifratification wouldrequirehimto construct a differentcoalition,
thechiefnegotiator willbe reluctant to endorseit,evenif(judgedabstractly)
it couldbe ratified. Politiciansmaybe willingto riska fewoftheirnormal
supporters inthecause ofratifying an international agreement, butthegreater
thepotentialloss, thegreatertheirreluctance.
In effect,thefixedcosts of coalition-building thusimplythisconstraint
on thewin-set:How greata realignment ofprevailing coalitionsat Level II
wouldbe requiredto ratify a particular proposal?For example,a tradedeal
mayexpandexportopportunities forSiliconValley,but harmAliquippa.
Thisis finefora chiefnegotiator (forexample,Reagan?)whocan easilyadd
Northern California yuppiesto his supportcoalitionand who has no hope
ofwinning Aliquippasteelworkers anyhow.Buta different chiefnegotiator
witha different supportcoalition(forexample,Mondale?)might finditcostly
or even impossibleto convertthegainsfromthesame agreement intopo-
liticallyusable form.Similarly, in the 1978"neutronbomb" negotiations
betweenBonnand Washington, "askingtheUnitedStatesto deploy[these
weapons]in WestGermanymighthave beenpossiblefora Christian Dem-
ocraticGovernment; fora Social Democraticgovernment, it was nearly
impossible.'"83 Undersuch circumstances, simple"median-voter" models
ofdomesticinfluences on foreign policymaybe quitemisleading.
Relaxingthe assumptionthatthe chiefnegotiator is merelyan honest
broker,negotiating on behalfofhis constituents, opensthepossibility that
theconstituents maybe moreeagerforan agreement (ormoreworriedabout
"no-agreement") thanhe is. Empiricalinstancesare not hardto find:in
early1987,Europeanpublicswerereadierto acceptGorbachev's"double-
zero" armscontrolproposalthanEuropeanleaders,justas intheearly1970s
theAmericanpublic(or at leastthepolitically activepublic)was moreeager
a
for negotiated end to theVietnam War than was theNixonadministration.
As a rule,thenegotiator retainsa vetooveranyproposedagreement in such
cases. However, ifthe negotiator's own domestic standing (or indeed, his
incumbency) would be threatened ifhe were to rejectan agreement that falls
withinhisLevel II win-set,and ifthisis knownto all parties,thentheother
sideat Level I gainsconsiderable leverage.DomesticU.S. discontent about
82. Bailey, Wilsonand the Great Betrayal, quotationat p. 15.
83. RobertA. StrongandMarshalZeringue,
"The NeutronBombandtheAtlantic Alliance,"
presentedat the1986annualmeetingoftheAmerican PoliticalScienceAssociation,
p. 9.
theVietnamWarclearlyaffected theagreement reachedat theParistalks.84

Conversely, iftheconstituents are(believedtobe) hard-line, thena leader's
domesticweaknessbecomesa diplomatic asset. In 1977,forexample,the
Americanscalculatedthat"a delayin negotiating a treaty. .. endangered
[Panamanian PresidentOmar]Torrijos'position;and PanamawithoutTor-
rijosmostlikelywouldhavebeenan impossible negotiating partner."85 Sim-
ilarly,inthe1954Triestenegotiations, theweakItaliangovernment claimed
that"'Unless something is done in our favorin Trieste,we can lose the
election.'That card was playedtwo or threetimes[reportedthe British
negotiatorlater],and it almostalwaystooka trick."86
My emphasison thespecialresponsibility ofcentralexecutivesis a point
ofaffinitybetweenthetwo-levelgamemodeland the"state-centric" liter-
ature,even thoughtheunderlying logicis different.In this"Janus"model
of domestic-international interactions,transnational politicsare less prom-
inentthanin some theoriesof interdependence.87 However,to disregard
"cross-table"alliancesat Level II is a considerablesimplification, and itis
moremisleading, thelowerthepoliticalvisibility oftheissue,andthemore
frequent thenegotiations betweenthegovernments involved.88Empirically,
forexample,two-levelgamesin the EuropeanCommunity are influenced
bymanydirecttiesamongLevel II participants, suchas nationalagricultural
spokesmen. In somecases,thesamemultinational actormayactuallyappear
at morethanone Level II table.In negotiations overmining concessionsin
some less-developedcountries,forexample,the same multinational cor-
poration maybe consulted privatelybyboththehomeandhostgovernments.
In subsequentworkon the two-levelmodel,the strategicimplications of
directcommunication betweenLevel II playersshouldbe explored.
Conclusion
The mostportentous development in thefieldsofcomparative politicsand

relationsin recentyearsis the dawningrecognition
international among
in each fieldof the need to takeintoaccountentanglements
practitioners
betweenthe two. Empiricalillustrations of reciprocalinfluencebetween
domesticand international
affairsabound.Whatwe neednoware concepts
1969-1973,"
84. I. WilliamZartman,"Reality,Image,and Detail:The ParisNegotiations,
in Zartman, 50% Solution,pp. 372-98.
85. Zbigniew PowerandPrinciple
Brzezinski, (NewYork:Farrar,StrausandGiroux,1983),
p. 136,as quotedin Habeeb and Zartman,Panama Canal Negotiations,pp. 39-40.
86. Harrisonin Campbell,Trieste1954,p. 67.
87. SamuelP. Huntington, "Transnational in WorldPolitics,"WorldPolitics
Organizations
25 (April 1973),pp. 333-68; Keohane and Nye, Power and Interdependence;Neustadt,Alliance
Politics.
byMajorWesternPowersin Bargaining
88. BarbaraCrane,"PolicyCoordination withthe
Organization
ThirdWorld:Debt Reliefand theCommonFund," International 38 (Summer
1984),pp. 399-428.
Organization
460 International
andtheories thatwillhelpus organizeandextendourempirical observations.

Analysisin termsoftwo-levelgames offersa promisingresponseto this
challenge.Unlike state-centric
theories,the two-level approachrecognizes
the inevitabilityof domesticconflictabout whatthe "nationalinterest"
requires.Unlikethe "Second Image" or the "Second ImageReversed,"
the two-levelapproachrecognizesthatcentraldecision-makers striveto
reconciledomestic andinternational
imperatives As
simultaneously. wehave
seen, statesmen in thispredicamentface distinctivestrategic
opportunities
and strategicdilemmas.
Thistheoretical approachhighlightsseveralsignificant ofthelinks
features
betweendiplomacy and domestic including:
politics,
* theimportant betweenvoluntary
distinction and involuntarydefection
frominternationalagreements;
* thecontrastbetweenissueson whichdomesticinterests are homogene-
ous, simplypittinghawksagainstdoves,and issueson whichdomestic
are moreheterogeneous,
interests so thatdomesticcleavagemayac-
tuallyfosterinternational
cooperation;
* thepossibilityof synergistic movesat
issue linkage,in whichstrategic
one game-table unexpectedcoalitionsat thesecondtable;
facilitate
* theparadoxicalfactthatinstitutional arrangementswhichstrengthen
decision-makers at homemayweakentheirinternational bargainingpo-
sition,and vice versa;
* theimportance oftargetinginternational
threats,offers,
and side-
payments withan eye towardstheirdomesticincidenceat homeand
abroad;
* thestrategic aboutdomesticpolitics,and thespecial
uses ofuncertainty
of "kinkywin-sets";
utility
* thepotentialreverberation pressureswithinthedomes-
ofinternational
ticarena;
* thedivergences betweena nationalleaderand thoseon
ofinterest
and in particular,
whosebehalfhe is negotiating, theinternational
impli-
in domesticpolitics.
cationsofhis fixedinvestments
Two-levelgamesseem a ubiquitousfeatureof social life,fromWestern
economicsummitry to diplomacyin theBalkansandfromcoalitionpolitics
inSriLankatolegislativemaneuveringonCapitolHill.Far-rangingempirical
researchis needednow to testand deepenourunderstanding of how such
gamesare played.

Putnamdiplomacy and Doestic Politicspdf

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

Putnamdiplomacy and Doestic Politicspdf

Uploaded by

Copyright:

Available Formats

Diplomacy and Domestic Politics: The Logic of Two-Level Games

Author(s): Robert D. Putnam

Domesticpoliticsand international relationsare oftensomehowentangled,

An earlierversionofthisarticlewas deliveredat the1986annualmeeting oftheAmerican

InternationalOrganization42, 3, Summer 1988

had receiveda powerful boostfromtheincoming Carteradministration and

shipoftheBundesbank.Publicly,HelmutSchmidtposedas reluctant to the

nationally,but thatfactionwas initiallyoutnumbered. Thus,international

Deutschand ErnstHaas on regionalintegration. 10Haas, in particular, em-

10. Karl W. Deutsch et al., Political Communityin the NorthAtlanticArea: International

Moredebatable,however,is theiridentification of"statestrength" as the

14.Forexample,see Krasner,"UnitedStatesCommercial andMoneyPolicy,"p. 55: "The

theyare unitedon all issues norbecause theyare insulatedfromdomestic

Overtwodecadesago RichardE. Waltonand RobertB. McKersieoffered

19.JamesE. Alt,"CrudePolitics:OilandthePoliticalEconomyofUnemployment inBritain

The politicsofmanyinternational negotiations be conceived

RobertAxelrodhas proposeda "Gamma paradigm,"in whichthe U.S.

cleavages,and itpostulatesthatone oftheinternational actors-theSoviet

Towards a theoryof ratification:

Considerthefollowing stylizedscenariothatmightapplyto anytwo-level

23. RobertAxelrod,"The GammaParadigm theDomesticInfluence

dentpolicypreferences, butseekssimplyto achievean agreement

need to secureelectoralmajorities at home."27However,ratification need

setsof each of thepartiesto theaccord.Thus,agreement is possibleonly

On the otherhand,in any two-levelgame,the credibility of an official

39. GeoffreyW. Harrison, in JohnC. Campbell, ed., Successful Negotiation: Trieste1954

FIGURE 1. Effectsof reducingwin-setsize

Determinantsof the win-set

1. The size of the win-setdepends on the distributionof

44. ThomasRomerandHowardRosenthal, "PoliticalResourceAllocation,

roomof theLevel I negotiator) dependson therelativesize of the "isola-

involves tradeoffsbetween take-home payandpensions. (WaltonandMcKersie

a government dividedis morelikelyto be able to strikea

47. Raiffanotesthat"themorediffusethepositions each side,theeasieritmight

assumption thatthenegotiations involveonlyone issue. Relaxingthisas-

decontrol),butthatsomemembers ofthatmajoritywouldbe willing toswitch

2. The size of the win-setdepends on the Level II

Not all significant ratificationpracticesare formalized; forexample,the

and so on wouldneed to be considered.At some pointin this

3. The size of the win-setdepends on the strategiesof the

64. Walton and McKersie, Behavioral Theoryof Labor Organizations,p. 345.

neitherat free-tradersnor at confirmed protectionists, but at the uncom-

71. Synderand Diesing, ConflictAmong Nations, p. 517.

74.JohnR. Hillman,"The MutualInfluence ofItalianDomesticPoliticsandtheInternational

opinionis shifted, a baremajority

Reverberation as discussedthusfarimpliesthatinternational pressure

The role of the chiefnegotiator

78. Thisis theapproachused to analyzetheAnglo-Chinese overHongKong

riskyforthechiefnegotiator, and theyofteninterfere

natories,especiallyBritainandFrance.In effect, itwas Wilsonhimself who

theVietnamWarclearlyaffected theagreement reachedat theParistalks.84

The mostportentous development in thefieldsofcomparative politicsand

andtheories thatwillhelpus organizeandextendourempirical observations.

You might also like