Sage Publications, Inc

Why States Act through Formal International Organizations
Author(s): Kenneth W. Abbott and Duncan Snidal

Source: The Journal of Conflict Resolution, Vol. 42, No. 1 (Feb., 1998), pp. 3-32
Published by: Sage Publications, Inc.
Stable URL: http://www.jstor.org/stable/174551 .
Accessed: 17/11/2014 14:06
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.
Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve and extend access to The Journal of
Conflict Resolution.
http://www.jstor.org
This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

All use subject to JSTOR Terms and Conditions
Why States Act through Formal

International Organizations
KENNETH W. ABBOTT
Graduateand InternationalStudies
NorthwesternUniversitySchool of Law
DUNCAN SNIDAL
Departmentof Political Science
Universityof Chicago
Statesuseformalinternational
andmore
(IOs)tomanageboththeireverydayinteractions
organizations
international
conflicts.Yet,contemporary
international
dramatic
theorydoesnotexplain
episodes,including
thequestionof whystatesuseformalorganizations
theexistenceorformof IOs.Thisarticleaddresses
by
thatenablethemto performthosefunctions.
the functionsIOsperformandtheproperties
investigating
thatseesIOsas enablingstatesto achievetheirends,the
witha rational-institutionalist
perspective
Starting
authorsexaminepoweranddistributive
questionsandtheroleof IOsin creatingnormsandunderstanding.
of formalorganizations,
areidentifiedas the key properties
andtheir
andindependence
Centralization
witha widearrayof examples.IOsas community
furtherallow
is illustrated
representatives
importance
valuesandenforceinternational
commitments.
statesto createandimplement
community
? When the United States decided to reverse the Iraqiinvasion of Kuwait, it did not act
unilaterally (although it often does). It turned to the United Nations (UN) Security
Council.
e Whenthe SecurityCouncilsoughtto learnthe extentof chemical,biological, andnuclear
armsin Iraq,it did not rely on U.S. forces. It dispatchedinspectorsfromthe International
Atomic EnergyAgency (IAEA).
* When the internationalcommunity sought to maintainthe suspension of combat in
Bosnia, it did not rely only on nationalefforts. It sent in peacekeepingunits under the
aegis of the UN and NorthAtlanticTreatyOrganization(NATO).
? When statesliberalizedtradein servicesandstrengthenedintellectualpropertyprotection
in the UruguayRound,they were not contentto draftrules.They createdthe WorldTrade
Organization(WTO) and a highly institutionalizeddisputesettlementmechanism.
Formal international organizations (IOs) are prominent (if not always successful)
participants in many critical episodes in international politics. Examples in addition
comments
fromLeaBrilmayer,
AUTHORS'
NOTE:Wearegrateful
forvaluable
Judith
Charles
Goldstein,
andseminar
attheUniversity
Anne-Marie
JamesMorrow,
Slaughter,
participants
Lipson,AndrewMoravcsik,
LawSchool,New YorkUniversity
Law School,Princeton
at Berkeley,Harvard
of California
University
andtheProgram
on International
andSecurity(PIPES)at theUniversity
of
Politics,Economics
University,
Forfinancial
AbbottthankstheRussellBakerFund,theCharles
C.Linthicum
Fund,andthe
Chicago.
support,
Summer
Research
Grantprogram.
Northwestern
JOURNALOF CONFLICTRESOLUTION,Vol. 42 No. 1, February1998 3-32
? 1998 Sage Publications,Inc.

JOURNALOF CONFLICTRESOLUTION
to those above include the following: Security Council sanctions on Libya, IAEA
inspectorsin NorthKorea,UN peacekeepersin the MiddleEast, andso forth.The UN
secretary-general's1992 Agenda for Peace sets out an even broaderrange of current
and proposedUN functions in situationsof internationalconflict: fact finding, early
warning, and preventive deployment;mediation, adjudication,and other forms of
disputeresolution;peacekeeping;sanctionsandmilitaryforce;impartialhumanitarian
assistance;and postconflictrebuilding.But IO influence is not confined to dramatic
interventionslike these. On an ongoing basis, formalorganizationshelp managemany
significant areas of interstaterelations, from global health policy (the WHO) to
Europeansecurity(OSCE and NATO)to internationalmonetarypolicy (IMF). What
is more,participationin such organizationsappearsto reducethe likelihood of violent
conflict among memberstates (Russett,Oneal, and Davis in press).
IOs range from simple entities like the APEC secretariat,with an initial budget of
$2 million, to formidableorganizationslike the EuropeanUnion (EU)1and the World
Bank, which has thousandsof employees and multipleaffiliates and lends billions of
dollars each year. Specialized agencies like the ILO, ICAO, and FAO play key roles
in technical issue areas.New organizationslike UNEP, the EBRD, and the International Tribunalfor the formerYugoslaviaareregularlycreated.OlderIOs like NATO
and the Security Council are rethought and sometimes restructuredto meet new
circumstances.2As the examples illustrate,moreover,even the most powerful states
often act throughIOs. In short,"itis impossibleto imaginecontemporaryinternational
life" withoutformalorganizations(Schermersand Blokker 1995, 3).
Why do states so frequentlyuse IOs as vehicles of cooperation?What attributes
accountfor theiruse, and how do these characteristicsset formalorganizationsapart
from alternativearrangements,such as decentralizedcooperation,informalconsultation, and treaty rules? Surprisingly,contemporaryinternationalscholarshiphas no
clear theoreticalanswersto such questionsand thus offers limited practicaladvice to
policy makers.
We answerthese questionsby identifyingthe functionalattributesof IOs across a
range of issue areas. Although we are concerned with the concrete structureand
operationsof particularorganizations,we also see IOs as complex phenomenathat
implicate several lines of internationalrelations(IR) theory.Fromthis vantagepoint,
we identify two functional characteristicsthat lead states, in appropriatecircumstances,to preferIOs to alternateformsof institutionalization.These arecentralization
and independence.
IOs allow for the centralizationof collective activitiesthrougha concreteandstable
organizationalstructureanda supportiveadministrativeapparatus.These increasethe
1. Althoughwe discuss certainof its operations,we deliberatelyde-emphasizethe EU because some
would regardit as an exceptionalcase of institutionalization.
2. A discussion of IOs is an exercise in acronyms.The ones not identifiedin the text, in order,are the
World Health Organization(WHO), Organizationfor Security and Cooperation in Europe (OSCE),
InternationalMonetary Fund (IMF), Asia-Pacific Economic Cooperationforum (APEC), International
LaborOrganization(ILO), InternationalCivil AviationOrganization(ICAO),Food and AgricultureOrganization(FAO),UnitedNationsEnvironmentProgram(UNEP), andEuropeanBankfor Reconstructionand
Development(EBRD).

ORGANIZATIONS 5
Abbott,Snidal/ FORMALINTERNATIONAL
efficiency of collective activities and enhancethe organization'sability to affect the

understandings,environment,and interestsof states. Independencemeans the ability
to act with a degree of autonomywithin defined spheres.It often entails the capacity
to operateas a neutralin managinginterstatedisputesandconflicts. IO independence
is highly constrained:memberstates,especially the powerful,can limit the autonomy
of IOs, interferewith theiroperations,ignoretheirdictates,or restructureanddissolve
them. But as in many privatetransactions,participationby even a partiallyautonomous, neutralactorcan increaseefficiency and affect the legitimacyof individualand
collective actions. This provides even powerful states with incentives to grant IOs
substantialindependence.
The broad categories of centralizationand independenceencompass numerous
specific functions.Most IOs performmore thanone, thougheach has its own unique
combination.We do not enumerateevery such function or provide a comprehensive
typology. Instead,we highlightseveralof the most important.We focus especially on
the active functionsof IOs-facilitating the negotiationandimplementationof agreements, resolving disputes,managingconflicts, carryingout operationalactivities like
technical assistance, elaborating norms, shaping internationaldiscourse, and the
like-that IR theoryhas only sparinglyaddressed.Rationalstates will use or createa
formalIO when the value of these functionsoutweighsthe costs, notablythe resulting
limits on unilateralaction.
DistinguishingformalIOs fromalternativeformsof organizationis importantfrom
several perspectives.For IR scholars,who largelyabandonedthe study of formalIOs
in the move from the legal-descriptive tradition to more theoretical approaches,
developing such distinctionsshould "openup a large and importantresearchagenda"
with institutionalform and structureas centraldependentvariables(Young 1994, 4;
see also Koremenoset al. 1997). This will complementemergingworkon international
legalization, a closely related form of institutionalization(Burley and Mattli 1993;
Abbott and Snidal 1997; Keohane,Moravcsik, and Slaughter1997). Such research
will also benefit practitionersof conflict managementand regime design (Mitchell
1994). The policy implicationsof our analysis are significant as well. Many states,
notablythe United States,now resist the creationof IOs and hesitateto supportthose
alreadyin operation,citing the shortcomingsof internationalbureaucracy,the costs of
formal organization,and the irritationsof IO autonomy.This is an ideal time for
studentsof internationalgovernanceto focus on the otherside of the ledger.
The next section spells out ourtheoreticalapproach,drawinglessons fromthe ways
in which differentschools of theoryhave dealt with (or have failed to deal with) the
questionsposed above. It is followed by an analysisof the organizationalattributesof
centralizationand independenceand the functionsthey make possible-especially in
contexts of cooperation and nonviolent conflict. The final section explores two
composite functionsthatchallengeconventionalviews of 10 capabilitiesand demonstratethe complementarityof prevailingtheories:developing,expressing,andcarrying
out community norms and aspirationsand enforcing rules and commitments. We
conclude with the example of the Security Council in the Gulf War,which draws
togetherthese themes in the context of violent conflict.

PUTTING IOs INTO THEORY

AND THEORY INTO IOs
Ourprimaryapproachis rationalistand institutionalist.We assume, for simplicity,
thatstatesarethe principalactorsin worldpolitics andthatthey use IOs to createsocial
orderings appropriateto their pursuit of shared goals: producing collective goods,
collaboratingin prisoner'sdilemmasettings, solving coordinationproblems,and the
like. We startwith the pursuitof efficiency and employ the logic of transactioncosts
economics andrationalchoice (Snidal 1996), using analogieswith business firms and
medieval trading institutions.Decentralizedcooperationtheory and, especially, regime theoryprovide a strongdeductivebasis for this analysis.
Regime theory (Krasner 1983; Keohane 1984) represents a major advance in
understandinginternationalcooperation.It is self-consciouslytheoreticalandfocuses
directlyon the institutionalorganizationof internationalcooperation.But it has several
shortcomings.Most important,regime scholarsembracean earlierturnin IR, which
unnecessarilycoupled a move to theorywith a move away from considerationof IOs
themselves. This resultedin "thesteady disengagementof internationalorganization
scholars from the study of organizations,to the point that today one must question
whether such a field even exists any longer except in name only" (Rochester 1986,
783-84). Indeed, regime theory deals with institutionsat such a general level that it
has little to say aboutthe particularinstitutionalarrangementsthat organizeinternationalpolitics. Ourfocus on the concreteoperationsof formalIOs not only bringsthem
into regimetheorybutalso providesa broaderopportunityforIR theoryto differentiate
among institutionalforms andrecaptureinstitutionaldetails. We draw on the legaldescriptiveliteratureto accomplishthis.
Furthermore,regime theory has been rightly criticized for paying insufficient
attentionto issues of poweranddistributionin internationalpolitics.Wedrawon realist
considerations to supplement our institutionalistapproachin this regard. Finally,
althoughregime theoryhas paid increasingattentionto the role of ideas and normsin
internationalpolitics (Goldsteinand Keohane 1993), it has only begun to incorporate
these importantconsiderations.Here, we draw on constructivisttheory for guidance.
In sum, we enrich our primarilyrationalistapproachwith importantinsights from
several differenttraditions,which we see as complementaryratherthancompetitive.
Decentralizedcooperationtheorytakes as the problematicof internationalgovernance the existence of coordinationand collaborationproblemsrequiringcollective
action (Oye 1986; Stein 1983; Snidal 1985a). It assumes anarchy,often depicted in
game models, and analyzes how states cooperate in that spare context through
strategies of reciprocity and other forms of self-help. The dependent variable is
typicallycooperationin the abstract,andmuchof theresearchin this traditionhas been
directed to disproving the realist assertionthat cooperationin anarchyis unlikely.
There is no nuanced account of the forms of cooperation because the anarchy
assumptionmakes IOs and other institutionslargely irrelevant.However, the strong
assumptionsthat underliethe theory,such as the need for high-qualityinformation,
suggest that cooperation is unlikely without an adequate institutional contextalthoughthe theoryis only beginningto analyzethatcontext (Morrow1994). For our

ORGANIZATIONS 7
purposes, however, it performs a useful service by emphasizing that institutional

capacities other than centralized enforcement are crucial in mediating interstate
relations.
Regime theory, in contrast, deals explicitly with institutionalfactors affecting
cooperation, and regime scholars frequentlymention IOs. But they downplay the
distinctiveinstitutionalrole(s) of IOs, perhapsin continuedreactionagainstthe earlier
preoccupationwith formal organizations.For example, Martin (1992) depicts the
EuropeanEconomic Community(EEC) and the CoordinatingCommitteefor Export
Controls (COCOM) as importantbut nevertheless quite rudimentaryforums for
intergovernmentalbargaining;Weber (1994) emphasizes the broad political and
symbolic goals of the EBRD. Neitherdiscusses the organizations'primaryoperational
roles. Keohane's(1984) AfterHegemonyalso emphasizesintergovernmental
bargaining, arguingthatregimeshelp statesreachspecific agreementsby reducingtransaction
costs, improving information,and raising the costs of violations. But this valuable
analysis also excludes many significant operationalactivities of IOs.3 In all these
works, furthermore,regime scholars treat internationalinstitutionsas passive. Regimes are seen, for example,as embodyingnormsandrules or clarifyingexpectations
(Keohane 1984; Yarbroughand Yarbrough1992; Garrettand Weingast 1993), functions also performedby treatiesand informalagreements.Regimes are also seen as
forums in which states can interact more efficiently: like Keohane and Martin,
Moravcsik's(1991) analysis of the Single EuropeanAct treatsIOs as sites of, but not
as agents in, cooperation.Indeed, the canonicaldefinition of regime (Krasner1983)
encompasses only norms and collective choice procedures,makingno provision for
the active andindependentIO functions-and the correspondinginstitutionalformsthat we emphasizebelow.
Legal scholarship continues to offer descriptive accounts of the history and
institutionalarchitectureof IOs, as well as doctrinalanalysis of norms and texts,
especially the normative output of organizationssuch as ILO treaties or General
Agreementon Tariffsand Trade(GATT)/WTOpanel decisions (Bowett 1982; Kirgis
1993). More importantfor present purposes, another strand of doctrinal theory
addresses the constitutionallaw of IOs, including membership and voting rules,
externalrelations,finance, and the authorityof specific organs (Amerasinghe1994;
Sohn 1950, 1967; Dupuy 1988; Shihata 1991, 1995). The best of this work is
comparative,examining how common problems of organizationand operationare
addressedin the constitutivedocumentsand practicesof variousIOs (Schermersand
Blokker 1995; Chayes andChayes 1995). Unfortunately,"inthe landof legal science,
there is no stronglyestablishedtraditionof developing theories on IOs" (Schermers
andBlokker 1995, 8; see also Brownlie 1990, 679). Nevertheless,legal scholarshiplike some earlierwork in IR, notablyCox andJacobson(1973)-carefully differentiates among institutionalforms and emphasizes institutionaldetails, an important
contributionthatwe use in our analysis.
3. Keohane (1984) does discuss monitoring,but Glaser (1995) argues that regime theorists do not
explain why monitoringmust be done centrally.

Realist theory finds both legal and regime scholarshipnaive in treating IOs as
serious political entities. Realists believe states would never cede to supranational
institutions the strong enforcementcapacities necessary to overcome international
anarchy.Consequently,IOs and similarinstitutionsare of little interest;they merely
reflect nationalinterestsandpoweranddo not constrainpowerfulstates(Mearsheimer
1995; Strange1983; for a more nuancedview, see Glaser 1995). We acceptthe realist
point that states arejealous of theirpower and deeply concernedwith the distributive
consequences of theirinteractions.Yet, realistsunderestimatethe utility of IOs, even
to the powerful.The United States, at the peak of its hegemony,sponsorednumerous
IOs, includingGATT,IMF,andNATO;these organizationshave provided"continuing
utility . . . as instruments ... for regime and rule creation" (Karns and Mingst 1990,
29). Even the Soviet Union, the very model of a modernrepressivehegemony, used
the Council for MutualEconomic Assistance to organizeeconomic relationswithin
the easternbloc. We arguethatpowerfulstates structuresuch organizationsto further
their own interestsbut must do so in a way thatinduces weaker states to participate.
This interplayis embeddedin 10 structureand operations.
Finally,KratochwilandRuggie (1986) arguethatonly constructivist(interpretivist)
theory-focusing on norms,beliefs, knowledge,and understandings-can satisfactorily explain formal organizations.We accept the insight that social constructionsare
fundamentalelements of internationalpolitics (Wendt1992, 1995; Barnett1993) and
agree that IOs are-in part-both reflections of and participantsin ongoing social
processes and prevailingideas (Finnemore1996; Kennedy 1987). But the role of IOs
is best understoodthrougha synthesis of rationalist(includingrealist) and constructivist approaches.States consciously use IOs both to reduce transactioncosts in the
narrowsense and,morebroadly,to createinformation,ideas, norms,andexpectations;
to carryout and encouragespecific activities;to legitimateor delegitimateparticular
ideas and practices; and to enhance their capacities and power. These functions
constituteIOs as agents,which, in turn,influencethe interests,intersubjectiveunderstandings,andenvironmentof states(McNeely 1995). Potentially,these roles give IOs
an influence well beyond their material power, which is trivial on conventional
measures. Indeed, IO activities may lead to unintendedconsequences for member
states, a fear often expressed by U.S. politicians. Yet, IO autonomyremains highly
constrainedby state interests,especially those of the powerful-a fact often demonstratedby U.S. politicians.
Although we adopt a predominantlyrationalisttheoreticalapproach,we are concernedwith highlightingthe importanceof formalIOs as empiricalphenomenarather
than with maintaininga particulartheoreticaldogma. None of the individual approaches mentionedadequatelyexplains why states use formal IOs; each holds key
insights. In identifyingformalIOs as an importantcategory of institutionalizationto
be explained,therefore,we proceedin a more interpretivemode, drawingon different
strandsof argumentationto highlight ways in which formal IOs function to manage
interstatecooperationand conflict.4
4. On the use of rationalchoice as an interpretivedevice, see Ferejohn(1991), Johnson(1991), and
Snidal (1985b).

ORGANIZATIONS 9
Abbott,Snidal/FORMALINTERNATIONAL
THE FUNCTIONS OF IOs:

CENTRALIZATION AND INDEPENDENCE
Two characteristicsdistinguishIOs from otherinternationalinstitutions:centralization (a concreteand stable organizationalstructureand an administrativeapparatus
managingcollective activities) and independence(the authorityto act with a degree
of autonomy,and often with neutrality,in defined spheres).5The very existence of a
centralizedsecretariatimplies some operationalautonomy,but this is often limited to
administrativeandtechnicalmattersandsubjectto close supervisionby governments.
In othersituations-sometimes involvingthe same organizations-substantive autonomy andneutralityareessential.The rangeandpotentialimportanceof these activities
lead us to treatindependenceas a separatecategory.
Centralizationand independenceenhanceefficiency. An analogy to privatebusiness firms is instructive.The firm replaces contractualrelations among suppliers,
workers,and managers;it substitutesa centralized,hierarchicalorganizationfor the
horizontal,negotiatedrelationsof contract.In Coase's (1937) theory,firmsareformed
when the transactioncosts of direct contractingare too high for efficient operation.
Similarly,the move from decentralizedcooperationto IOs occurs when the costs of
direct state interactionoutweigh the costs of internationalorganization,including
consequentconstraintson unilateralaction (Trachtman1996).
Centralizationand independence represent different forms of transactioncost
economizing. Small businesses draw mainly on the centralizationbenefits of formal
organization,interposinga legal entity with the ability to manageemployees hierarchically and the capacityto contract,sue, and be sued. The owners still manage the
business directly,though their interactionsare more highly structured.Investors in
largerfirms additionallybenefit by grantingautonomyand supervisoryauthorityto
professional managers;in Berle and Means's (1968, 5) famous phrase, there is a
"separationof ownership and control."The situationis similar in complex IOs, in
which memberstates grantsome authorityto IO organsand personnelbut supervise
them through structuresresembling the corporateshareholdersmeeting, board of
directors, and executive committee. Introducing these new actors changes the
relations amongstatesandallows themto achievegoals unattainablein a decentralized
setting.
Centralizationand independenceproducepolitical effects beyond mereefficiency.
In these respects, IOs resemble governments and private associations more than
business firms. Independence,in particular,enables IOs to shape understandings,
influence the terms of state interactions,elaboratenorms, and mediate or resolve
member states' disputes. The acts of independentIOs may be accorded special
legitimacy, and they affect the legitimacy of members'actions. Even centralization,
seemingly more mechanical, can alter states' perceptions and the context of their
interactions.
5. Centralizationand independenceare mattersof degree, not only among IOs but even between IOs
andrelatedinstitutions.Forexample,the Groupof Seven is not a formalIO but merelya negotiatingforum.
Its organizationalpractices(e.g., a rotatingchair)neverthelessprovide some centralizationbenefits, and it
partakesof some autonomy,as in legitimatingmembers'actions.

10
CENTRALIZATION
It is no great theoreticalinsight that an established organizationalstructureand
centralizedadministrativesupportcan rendercollective activitiesmoreefficient:even
studentsof internationalgovernanceare not contentto communicateby e-mail; they
formthe InternationalStudiesAssociationandthe InternationalLaw Association.This
simple insight goes farto explainthe proliferationof IOs in this centuryin a periodof
increasing issue complexity and a growing numberof states. The (inter)subjective
effects of centralizationare less apparent,thoughequally important.We considerthe
benefits of centralizationunder two headings-support for direct state interaction
(the principalfocus of regime theory)and operationalactivities (the traditionalfocus
of IO studies). Here, we emphasizeconcreteactivities in which governmentsremain
closely involved;the following section introducesbroaderfunctionsalso requiringIO
autonomy.
SUPPORT FOR STATEINTERACTIONS
The organizationalstructureof IOs enhanceseven the passive virtuesrecognized

by regime theory.An establishedorganizationprovides a stable negotiatingforum,
enhancingiterationandreputationaleffects. Such a stableforumalso allows for a fast
responseto suddendevelopments.The SecurityCouncil,for example, is organizedso
thatit can functionon shortnotice, with each memberrequiredto maintaincontinuous
representationat UN headquarters.A permanentorganizationalso reinforcesaccepted
norms:the most favorednation(MFN) principleinstantiatedin the WTO provides a
sounderbasis for state expectationsthanany informalarrangement.
In other ways too, centralizationshapes the political context of state interactions.
IOs provideneutral,depoliticized,or specializedforumsmore effectively thanalmost
any informalor decentralizedarrangement.This enablesa broaderrangeof behavior:
the superpowerscould discuss technicalnuclearissues within the IAEA without the
intrusionof high politics, even at the height of the cold war.IOs also serve as partisan
forumsfor political coalitions:the UnitedNationsConferenceon TradeandDevelopment (UNCTAD)for developing countries,the Organizationfor Economic Cooperation and Development(OECD) for industrializedstates.Finally,IOs strengthenissue
linkages by situatingthem within common organizationalstructures,as the WTO has
done for goods, services, and intellectualpropertyrights.
Formalorganizationsfurtherembody the precise termsof state interaction.Representationand voting rules "constitutionalize"balancesamong states having different
levels of power, interest,or knowledge. Stateswith advancednucleartechnology and
large supplies of nuclear raw materialare guaranteedseats on the IAEA Board of
Governors;states with majorshippingand carrierinterestshave equal representation
on the InternationalMaritimeOrganization(IMO) Council. Such decision structures
frequentlyguaranteedisproportionateinfluenceforpowerfulstates.Yet,they may also
constitutionalizeprotectionfor weaker states and hold the powerful accountableto
fixed rulesandprocedures.Forexample,boththe SecurityCouncilandtheEU Council

ORGANIZATIONS 11
are structuredso that the most powerful memberscan block affirmativeactions but,
even if united,cannotapproveactions withoutsupportfrom smallerpowers.
Such considerationsoften lead to elaborateorganizationalstructures.The substantive workof many IOs takesplace in specializedcommitteesstaffedby theirsecretariats.The OECDuses morethan200 committeesandworkinggroups;the IMOprepares
treatiesin substantivegroupingslike the maritimesafety and marineenvironmental
protectioncommittees.Such committeesare often formallyopen to all members,but
specializationoccurs naturallybecause of differences in interest,expertise, and resources.Delegation can also be encouragedinstitutionally:in the thirdUN law of the
sea conference (UNCLOS III), the chairs of open-ended committees sometimes
scheduledmeetings in rooms capableof holding only 30 people!6
Organizationalstructureinfluencesthe evolution of interstatecooperationas conditions change. For example, several environmentalagreementswere facilitatedby
appointing UNEP as secretariatand the World Bank as financial administrator,
obviating the need for new institutions.These institutionallinks are often contested
because of their distributionalimplications.The advancedcountriesfought to locate
new intellectual propertyrules in the WTO (ratherthan in the World Intellectual
PropertyOrganization[WIPO])so they could enforcetheirrightsmoreeffectively. In
other cases, organizationalstructurescreate vested interests that impede change or
politicize issues, as in the United Nations Education,Scientific, and CulturalOrganization (UNESCO) during the 1970s. More generally,because IOs are designed for
stability,they may not adaptsmoothlyto changingpowerconditions,as the continuing
makeupof the SecurityCouncilattests.Yet,the gradualreductionof U.S. voting power
in the IMF, mandatedby its declining share of capitalcontributions,illustrateshow
organizationalstructurecan facilitatesuch adaptation.
Most IOs include a secretariator similar administrativeapparatus.In simple
consultative organizations,the secretariatneed only assist with the mechanics of
decentralizedinteraction.The 1985 ViennaOzone Conventionassignedthe following
functionsto its secretariat:"(a)To arrangefor andservicemeetings...; (b) To prepare
and transmitreportsbased upon informationreceived ... ; (d) To preparereportson
its activities .. .; (e) To ensurethe necessarycoordinationwith otherrelevantinternational bodies . . .; (f) To perform such other functions as may be determined"
("Vienna Convention" 1985, 1532). The secretariatfor the Convention on LongRange TransboundaryAir Pollution(LRTAP)performedsimilarfunctionswith only
five professionals.Levy (1993, 84) notes thatthe staff had "littletime to do anything
else but keep the meetingsrunningsmoothly."
Even such modest activities can strengtheninternationalcooperation.Here, we
draw on the analogy to the medieval law merchantand the correspondingtheoretical
literature(Milgrom,North,andWeingast1990;Calvert1995;Morrow1994). Informal
consultationsproducedsufficientinformationon the identityof untrustworthytraders
to supporta substantialvolume of trade.Yet, modestefforts by centraladministrators
at commercialfairs to collect and relay additionalinformationcreateda new equilibrium at a higherlevel of exchange.
6. Personalcommunicationfrom BernardOxman,memberof the U.S. delegation,21 May 1997.

12
Most IOs performmore extensive supportivefunctions.Law-makingconferences

like UNCLOS III or the Rio conference on the environmentand developmentrely
heavily on their secretariats.IO personnelcoordinateand structureagendas,provide
backgroundresearch,andpromotesuccessful negotiations.They keep trackof agreements on particularissues, trade-offs,and areasof disagreement,periodicallyproducing texts thatconsolidatethe currentstate of play. They also transmitprivateoffers or
assurances,improvingthe flow of information.
IO staffs supportdecentralizedcooperationbetweenmajorconferences.The large,
expert OECD secretariatcollects, produces, and publishes informationrelevant to
national economic policy coordination.The WTO secretariatassists in numerous
negotiations, from the settlement of disputes to sectoral talks under the services
agreement.IO staffs also supportthe decentralizedimplementationof norms.UNEP,
secretariatfor the Basel convention on the transboundarymovement of hazardous
wastes, providesinformationstatesneed to manageactivitiesunderthe treaty;the ILO
receives, summarizes,and circulatesnationalreportson treatyimplementation.
Experience under the internationaltrade regime testifies to the importance of
organizationalstructureand administrativesupport.The originalGATTwas a normative and consultativearrangement;almostall organizationalfeatureswere removed at
the instance of the United States. Yet, member states soon needed more extensive
organizationalstructureand support.As membershipexpanded and complex new
issues appeared on the agenda, GATTbegan its metamorphosis into the WTO, a
true 10.
MANAGING SUBSTANTIVE OPERATIONS
IOs do more than supportintergovernmentalnegotiations;they manage a variety

of operationalactivities. A prototypicaloperationalorganizationis the WorldBank,
which finances massive development projects, borrows on world capital markets,
reviews state investment proposals, provides technical assistance and training in
many disciplines, generates extensive research and publications, and performs
other substantiveactivities.Operationalorganizationsnormallyhave sizable budgets
and bureaucracies,complex organizationalstructures,and substantial operational
autonomy.7
Memberstates of an 10 like the WorldBank use the institutionas an agent, taking
advantageof its centralizedorganizationandstaffto carryoutcollective activities.The
analogy of the large business corporation,with its dispersed owner-investorsand
professionalmanagers,is apt. Comparedwith a decentralizedapproachbased on ad
hoc contracting,a formal organizationprovides efficiency gains that outweigh the
accompanyingcosts in termsof money,humanresources,andconstraintson unilateral
action.Especiallywhen participatingstatesdifferin power,centralizedoperationswill
have significantdistributionalconsequences.
7. We reservefor the following sectiondiscussionof those functionsthatturndirectlyon independence

and neutrality.

ORGANIZATIONS 13
IO operations also significantly influence the capabilities, understandings,and

interestsof states. This is most apparentwith outputssuch as informationand rules.
But it is also true of more material activities like technical assistance and joint
production.Indeed, virtually all of the activities discussed below promote certain
norms and practicesamong states, often in unanticipatedways.
Pooling
Many IOs are vehicles for pooling activities,assets, or risks. Some pooling can be
accomplished on a decentralizedbasis, as in a business partnership,but a separate
entity with a stable organizationalstructureand specialized staff can greatlyreduce
transactioncosts while providingadditionaladvantages.
Consider the World Bank again. As in other internationalfinancial institutions
(IFIs), memberspool financialresourcesthroughcapital contributionsand commitments.Pooling providesa solid cushionof capitalthatenablestheWorldBankto make
credible financial commitmentsto borrowers,who rely on them for costly planning
and investmentdecisions, and to worldcapitalmarkets,in which the bankborrowsat
advantageousrates. In addition, this common effort promotes burden sharing in
providing a collective good and may limit the competitionfor influence that characterizes some bilateral assistance. Similarly,by combining development loans in a
common portfolio,bankmemberspool, and therebyreduce,theirindividualrisk.
Pooling enables the WorldBank to achieve economies of scale by carryingout a
largevolume of activities,establishinguniformproceduresandbuildingup a common
body of data. These economies allow it to develop greatertechnical expertise on
various aspects of country and project assessment than could most states and to
innovate in emerging areaslike "basicneeds."Finally,the bank's broadjurisdiction
creates a horizontaladvantageakin to economies of scope: by dealing with virtually
all needy countries,the bank can targetglobal prioritieswhile avoiding duplication
and gaps in coverage.8
The largeststates,especially the United States,could mobilize sufficientcapitalto
They areunwilling to
accomplishtheirinternationalfinancialobjectivesunilaterally.9
do so, however, for internationaland domestic political reasons and because of
competing priorities.Indeed,the United States is actively working to strengthenthe
IFIs, in part because their broad membershipand assessment structuresencourage
In the meantime,althoughthe G-7 countriesbearmost of the costs
wide cost sharing.10
of the IFIs, they also retain the greatest share of voting power and influence on
management.Duringthe cold war,they successfully excludedthe Soviet bloc andthe
People's Republic of China. Yet, the United States has been unable consistently to
dictateIFI decisions on specific transactions.
8. Of course, as Kratochwil(1996) notes, large-scalecentralizedoperationsmay not be necessary or
desirable in all cases. The MaastrichtTreaty'ssubsidiarityprinciple adopts this view, while authorizing
supranationalactivity when the scale of the problemmakes thatappropriate.
9. The desireto benefitfrompoolingis neverthelessreflectedin U.S. TreasurySecretaryRubin'slament
thatthe "UnitedStates cannotbe the lenderof last resortto the world"(quotedin Sanger 1995).
10. The G-7 countriesalso benefit from IFI independence,as discussedbelow.

14
NonfinancialIOs providesimilaradvantages.The publichealthactivitiesof WHO,

like otherUN-specializedagencies, arebasedon the pooling of nationalcontributions
and cost sharing (though the industrializedcountries bear the bulk of the costs);
economies of scale provide operationalefficiencies. The WHO smallpox campaign
illustratesthe horizontalbenefitsof centralization:a single global campaignagainsta
contagious disease is more effective than decentralizedefforts because global scope
avoids gaps in coverage. (The IAEA nuclear safeguards system offers a similar
advantage.)In addition,the stable organizationalstructureof WHO and the reputation-stakingeffect of membershipencourageparticipation.Free-riderproblems remain, but the organizationcan alleviatethem by using its own resources.WHO also
provides effective technical assistance by pooling financial and technical resources
and accumulatingexpertise; its global scope diffuses new technologies and allows
rationalprioritizationof needs. By enhancingthe developmentand transmissionof
ideas, technical activities of specialized organizationshave significantly shaped the
interests and identities of states. At the same time, they have helped less developed
states acquire capacities essential to both nationalpolicy making and international
activity.
An example of the limits of pooling illustratesthese effects and the importanceof
realist and constructivistconsiderations.UNESCO's scientific arm was intendedto
promote the public goods aspects of scientific research by pooling international
scientific facilities andcreatinga centralclearinghouse.The organizationwas initially
oriented towardthe needs of scientists: executive boardmembersdid not represent
governments.With the cold war, however, state interests assertedthemselves. The
board was reorganizedto represent states, and UNESCO's orientation shifted to
national science. Finnemore(1996) documentshow UNESCO technical assistance
subsequentlypromotednationalscience programseven in stateswheretherewas little
need for them. Thus, UNESCO helped shape states'identities,interests,andcapabilities in the areaof science policy even thoughits initialglobal objectiveswerefrustrated
by interstaterivalries.
Joint Production
Alchian and Demsetz's (1972) theory of the firm suggests that a centralized
organizationis particularlyimportantwhen workers, managers,and other "inputs"
must work in teams, producinga joint output. In these situations, the hierarchical
organizationof the firm makes it easier for managers,themselves beholden to the
owners ("residualclaimants"),to monitor, reward, and discipline employees. 10
personnelengage in similarteamworkandthus aretypically organizedhierarchically,
with supervisionby and on behalf of memberstates.
Beyond this, statesthemselvessometimesformmultinational"teams"to engage in
productionactivities. Experts from several Europeanstates cooperate in subatomic
researchthroughthe EuropeanOrganizationfor NuclearResearch(CERN),an IO that
operatesa nuclearlaboratory;the Airbus projectis a similarexample. In additionto
holdingparticipantsresponsible,these organizationspool resourcesandrisks, achieve
economies of scale, avoid duplicationand unproductivecompetition,and ensurethat

ORGANIZATIONS 15
the outputs,includingtechnologicalexternalities,are shared.Projectslike CERN and

Airbus resemble business firms even more than the typical 10. Indeed, Airbus,
originallycreatedas a partnershipunderFrenchlaw,is being transformedinto a private
corporationto bettercoordinatethe participants.
Perhaps the best example of interstatejoint productionis the NATO military
alliance.Commonwarplans, specializationof militarytasks,joint exercises, common
equipmentandinterchangeableparts,and,of course,theconductof battleareexamples
of teamworkparexcellence. NATO'sintegratedcommand-operating hierarchically
on behalfof memberstatesas residualclaimants-organizes, monitors,anddisciplines
participantsin the joint activities of the alliance, probably the most successful in
history.1'
Norm Elaboration and Coordination
Statesarrangecooperativerelationshipsthroughagreements.As Williamson(1985,
1994) and others have pointed out, bounded rationalityand high transactionand
informationcosts make it difficult for states-like the parties to any contract-to
anticipateand provide for all possible contingencies.The longer and more complex
the relationship,the more significantthe contingencies;the greaterthe investmentin
specific assets, the greaterthe uncertaintyandriskof opportunism.The domesticlegal
system helps alleviatethese problemsby supplyingmissing termsanddecision rules,
but the internationalinstitutionalcontextis comparativelythin."First,in international
law, there is not a very complete body of law that can be appliedto supply missing
terms.... Second, ... thereis generallyno disputeresolutiontribunalwith mandatory
jurisdiction. . . . The alternative, of course, is to write comprehensive contracts"
(Trachtman1996, 51-54).
Thereis anotheralternative:to createproceduresfor theelaborationof normswithin
an 1O.Decentralizedproceduresdo not addressthe problemsof transactioncosts and
opportunism.Even with coordinationissues-in which equilibriacan sometimes be
reachedwithoutcommunication-these problemscan stymie cooperationwhen there
are many actors, complex problems,and distributiveconflicts. The stable organizational structureof IOs addressesboth issues. Establishedproceduresfor elaborating
rules, standards,and specifications enhance cooperationeven when member states
retainthe power to rejector opt out-as they do even in 1Oswith relativelyadvanced
legislative procedures,like the ILO. Nonbindingrecommendationscan become de
facto coordinationequilibria,relied on by states and other internationalactors. This
gives IOs some power to affect internationalnormsand statebehaviorandpotentially
much greaterpower with the backingof key states.
As always, powerful statesexert disproportionateinfluence over normelaboration
and structurelegislative processes to ensure their influence. Here, too, however,
protectionfor weakerstates may be the price of theirparticipation,and the effective11. The analogy is imperfect. NATO's organizationdiffers from that of a firm. Nevertheless, team
analysis suggests why a formal 10 is valuable, whereas the standardpublic goods analogy reduces the
problemsimply to one of individual(under)provision.See Olson and Zeckhauser(1966).

16
ness of an established rule-makingprocedurerequiresthat powerful states respect

those arrangements.For example, powerful states often limit IO jurisdiction to
technicalareaswith limiteddistributionalimpact;as a result,IO legislativeprocedures
may go forward-up to a point, at least-less influencedby narrownationalinterests
and differentialpower thandirectintergovernmentalbargaining.
ManyIOs engage in normelaboration,especially of a technicalkind.The EU, most
notably, has issued a huge numberof directives, regulations, and other legislative
acts-affecting everythingfromfranchiseagreementsto telecommunicationinterconnectivity standardsto tax policy-though manyimportantissues have been addressed
throughinterstateagreementsand mutualrecognition.The preparationof proposed
legislation is housed exclusively in the commission to facilitate a depoliticized and
expert approach.
Many other IOs carryout extensive legislative programs,frequentlyfocusing on
coordinationrules. The ICAO promulgatesinternational"rulesof the air";the InternationalTelecommunicationsUnion (ITU) coordinatesnationalbroadcastingstandards;the Customs CooperationCouncil implementscommon customs rules; and the
Codex AlimentariusCommission harmonizesfood standards.Although technical,
these standardshave importanteffects on (and within) states, as the concern over
privileging Codex standards under the North American Free Trade Agreement
(NAFTA) demonstrated.Althoughthe associatedIOs are quite weak, their influence
is strengthenedby the self-enforcingnatureof coordinationequilibria.
INDEPENDENCE
Althoughcentralizationoften requiressome operationalautonomy,many valuable
1Ofunctionsrequiremore substantiveindependence.The participationof an IO as an
independent,neutralactor can transformrelationsamong states, enhancingthe efficiency and legitimacy of collective and individualactions. These functionsrequirea
delicate balance among short- and long-termcollective and distributionalinterests.
Powerfulstates will not enteran organizationthey cannotinfluence,yet undermining
the independence of an organizationperformingthe functions discussed here will
simultaneouslyreduceits effectiveness and theirown ability to achieve valued ends.
Analogies from the business firm and the law merchant illustrate the point.
Shareholdersin a largecorporationmustmonitormanagersto limit agency costs. Yet,
if majorshareholderscause managersto favortheirinterestsunduly,othersmay refuse
to invest. If shareholdersgenerally assertexcessive control, moreover,they lose the
advantagesof professionalmanagement.The law merchantanalogy is even sharper.
Powerful princes granted monopoly privileges to independent guilds of foreign
merchants,enabling them to embargothe princes themselves if they took advantage
of the merchants (Grief, Milgrom, and Weingast 1994). By eliminating princes'
incentives to cheat, these arrangementsenabled them to make the binding commitments necessaryto induce mutuallybeneficial trade.The princescould withdrawthe
guilds' privileges, of course, but were constrainedfrom doing so by the resultingloss
of trade.

ORGANIZATIONS 17
SUPPORT FOR DIRECT STATEINTERACTION
IndependentIOs promoteintergovernmentalcooperationin more proactiveways

than those discussed earlier;they are initiating as well as supportiveorganizations.
The governing body is often authorizedto call togethermember states to consider
currentproblems.IO personnelalso influencenegotiationagendas.On a high political
plane, UNEP kept ozone protectionalive when interstatenegotiationsdeadlockedand
built supportfor the MontrealProtocol.The UN secretary-generalmay put before the
Security Council any matterthat, in his opinion, threatensinternationalpeace and
security.At the administrativelevel, the ILOgoverningbody sets the GeneralConference agendawith assistancefromtheInternationalLaborOffice. At the technicallevel,
IO and conference officials advance specific proposals and suggest linkages or
trade-offs:the presidentof UNCLOS III was authorizedto defer contentiousvotes to
forge a consensusduringdeferment;the negotiatingtext advancedby GATTDirectorGeneralDunkel duringthe UruguayRound catalyzed the falteringnegotiationsand
helped bridgesubstantivedifferences.
IO officials arealso prominentmembersof the epistemiccommunitiesthatdevelop
andtransmitnew ideasfor internationalgovernance(Haas 1992). DrakeandNicolaidis
(1992, 76) documentthe role of IOs in developingthe concepts behindthe liberalization of tradein services:a "comparativelysmall numberof expertsin the GNS [Group
on Negotiationin Services] and on the GATT,UNCTADand OECD staffs [were] the
main source of the specific kinds of new ideas neededto carrythe policy projectto a
conclusion."The UN Economic Commissionon LatinAmericais well known as the
source of many ideas regardingeconomic developmentthat rallied the Groupof 77.
Such autonomousefforts can modify the political, normative,and intellectualcontext
of interstateinteractions.These factors are not purely exogenous, as in structural
theories or constructivistapproachesthat locate them in general societal trends,but
are tied to the agency and interestsof IOs (Ness and Brechin 1988; Scott 1992).
Independence is equally importantin implementation.The ILO committee of
experts-a group of privateindividuals-comments on nationalreports.Some ILO
organs use these comments to highlight noncompliancewith ILO conventions and
recommendationsand to invite governmentsto submitadditionalinformation.Other
IOs reporton statecompliancein additionto, or in lieu of, nationalreports.IO officials
furthermonitor state conduct, in more or less intrusiveways, althoughenforcement
remainsdecentralized.Forexample,the WTOregularlyreviews the generaleffects of
nationaltradepolicies.
MANAGING SUBSTANTIVE OPERATIONS
In the above examples, IOs facilitateinterstatecollaborationby pushing negotiations forward.This role could be played by, say, a dominantstate, but suspicions of
bias might impede cooperation;an independent10 may be more acceptablebecause
it is neutral.For many substantiveIO operations,however,it is the existence of a truly
independentthirdparty,not the absence of bias per se, that enables states to achieve
theirends.

18
Laundering
Launderinghas a negativeconnotationfromits associationwith runningill-gotten
gains throughseemingly independentfinancialinstitutionsuntil they come out clean,
having lost their original characterand taint. Without necessarily adopting that
connotation,we use the termadvisedlybecausethe process at work in IOs is similar:
activities that might be unacceptablein their original state-to-stateform become
acceptable when run through an independent,or seemingly independent,IO. The
concept should be familiarto IR scholarswho are reluctantto accept CentralIntelligence Agency funds but eagerly acceptNationalScience Foundationgrantsoverseen
by independentacademicpanels.
Appropriatelyenough,theWorldBank,IMF,andotherIFIsprovideclearexamples.
States may prefer developmentassistancefrom an independentfinancial institution
over directaid from anotherstate, especially a formercolonial power or one seeking
political influence. IFI restrictionson nationalautonomy(e.g., on project design or
broadereconomic policies) may not carrythe same domesticpolitical implicationsof
dependenceand inferiorityas would conditionsimposed directlyby, say, the United
States or France.These considerationsmay make IFI conditionsa superiormeans of
promotingdomestic reforms.
IFIs equally serve a launderingfunctionfor donorstatesseeking to avoid domestic
and internationalcontroversies.The WorldBank's charterrequires,for example, that
developmentloansbe madewithoutregardforthe "politicalcharacter"of therecipient;
disregardof this factoris difficultwithinthe United States,wherefinancialassistance
budgets require congressional approval. The United States called on the IMF to
manage the 1980s debt crisis, keeping the issue less politicized and more technical.
Similarly,the Soviet Union launderedsubsidiesto subordinatestatesin EasternEurope
throughCouncil for MutualEconomicAssistance (CMEA)tradingpractices,muting
domestic opposition to these political and economic arrangementsboth at home and
in recipient states (Marreese1986). IFIs also inhibit domestic special interestsfrom
distortingpolicy for otherpurposes,as in the case of tied aid.
Althoughthe obligationto participatein IFIs may be strong,doing so helps donor
states curtailaid recipients'expectations,thus preservingflexibility. Although international intermediariesdiminish a donor state's leverage over recipient states, this
factor is offset by decreasesin otherstates' leverage and in competitionfor leverage
among donors.Donor states as a group,of course, retaincontrolover the IFIs. But it
is the fund, not the United States or Germany,thatimposes austerityon borrowers.
The autonomy needed for successful launderinggives IOs influence over the
substanceof their activities.For example,IFI staff have significantinputinto lending
criteriaandadjustmentpolicies and,increasingly,into social, environmental,andother
related policies. Robert McNamara was able to broaden development discourse
beyond economic growthto include social factorsand to reorientWorldBank policy
(Finnemore1996; Sanford 1988). The point should not be overstated.McNamara's
reforms were hardly radical, and Western countries were largely receptive. Subsequently,the Reagan administrationpushed the WorldBank partiallyback toward

ORGANIZATIONS 19
marketpolicies. Thus, IO autonomyremainsboundedby stateinterestsandpower,as

reflected in institutionalarrangements.
Such interventionscan cause IOs to be perceivedas politicized,respondingto the
interestsof certainstates or to issues beyond theirregularpurview.This occurredin
the 1960s and 1970s, when the WorldBank withheldloans from states thatexpropriated foreign property without compensation (Lipson 1985, 138-39); recently, the
United States linked support for World Bank lending to human rights in cases,
includingChinaandMalawi (Kirgis 1993, 572-75). Whatevertheirjustification,such
measures reflect a partialfailure we label dirty laundering.Powerful states face a
tension between the immediateadvantagesof dirty launderingversus the long-run
costs of jeopardizingIO independence.
Launderingis not limited to financial organizations.UN peacekeeping allows
powerful states to support conflict reduction without being drawn into regional
conflicts and discouragesotherpowers from takingadvantageof their inaction.This
simultaneouslyreassuressmall countriesthat the conflict will not be enlarged.The
IAEA performstwo differentlaunderingfunctions.First,recipientsmay prefertechnical assistancefroman independentagencyratherthana particularnuclearstate,even
though nuclearstates as a group dominatethe agency. Direct assistance may create
dependence,reducepolicy flexibility, and be domesticallycontroversial.IAEA technical assistance programsalso distance providerstates from recipient nuclear programs and inhibit the commercialrivalryamong suppliersthat otherwise facilitates
proliferation.Second, states subject to nuclear safeguardsmay be more willing to
admitindependentinternationalmonitorsintosensitivenuclearfacilitiesthanto permit
whentheUnitedStatestransferred
of anotherstate.Interestingly,
entryby representatives
bilateralsafeguardresponsibilitiesto the IAEA in 1962, some recipientsresisted the
new arrangement,fearing that nationalsof various states on the IAEA staff would
conduct covert intelligence missions. This suggests, however, not that the logic of
launderingis false but thatit turnson the perceivedindependenceof the organization.
Launderingthus has significant implications for the constitutive rules of IOs.
Although memberstates retainultimatecontrol, organizationsmust be structuredfrom theirorgansof governancedown to theirpersonnelpolicies-to createsufficient
independencefor launderingto succeed. A failing of the UN secretariatis that its
personnelare viewed as retainingtheirnationalidentities;by contrast,the "Eurocrat"
is seen as having loyalties beyond his or her individualstate.
Neutrality
Neutrality adds impartialityto independence.It enables IOs to mediate among
states in contested interactions, including disputes and allocation decisions. UN
neutralityunderliesmost of the functionsdiscussedin the secretary-general'sAgenda
for Peace, from fact-findingandotherformsof preventivediplomacythroughdispute
resolution and peacekeepingto postconflictconsolidationof peace. Even more than
laundering,neutralitydemandsthat institutionsbe bufferedfrom direct pressuresof
states.

20
IO as neutral information provider. Regime theory recognizes the importance of
informationbut does not emphasizedifferencesin its quality.Informationcreatedor

verified by an independent,neutralIO is more reliable than that providedby states
because it is free of nationalbiases. Considerthe air pollution monitoringstations
established in Europe under LRTAP.Data supplied by Sweden or Russia could be
perceived as biased, but a neutralsource of informationwas more credibleand could
supportgreatercooperation.The conventionprotectingAntarcticseals incorporated
an existing institution,the Scientific Commissionon AntarcticResearch,as a neutral
source and verifierof informationon the statusof seals and state activities. Based on
this information,the parties attaineda ratherhigh degree of cooperation. Similar
conventionswithoutneutralsourcesof information,such as thatconcerningAntarctic
marineliving resources,havebeenless successful.Finally,the 1991 GeneralAssembly
declarationon fact finding strengthensthe UN secretary-general'srole as a neutral
informationsource in politically chargedsituations;the GeneralAssembly has similarly encouragedthe secretary-generalto develop early-warningsystems for international disputes and humanitariancrises.
Internationalmonitoringorganizations,notablythose operatingundermultilateral
armscontroltreaties,provideoutstandingexamplesof neutralinformationproduction.
Fromthe perspectiveof manyparticipants,the neutralityof these organizationsis their
most importantfeature.Impartialinformationnot only deterscheatingby others but
also helps states assure othersof their own compliance(Abbott 1993). Although the
literatureon informalcooperationandthe U.S.-Soviet armscontrolexperiencesuggest
thatstatescan performthese functionson theirown (Glaser1995), the widespreaduse
of IOs testifies to the advantagesof third-partyneutrals.
IO as trustee. In private commercialdealings, neutralparties often hold assets
belonging to persons who cannot be trustedwith possession until a transactionis
completed. The escrow agent, for example, protects assets until all elements of the
transactionare ready for closing, while the trusteeholds assets on behalf of owners
who cannottake title immediately.
Such arrangementsarenot commonin IR, but notableexamplesexist. The Security
Council held Iraqresponsiblefor losses causedby its invasion of Kuwait.It required
Iraqto contributea percentageof its oil exportrevenuesto a UN compensationfund
fromwhich paymentswould be made.A compensationcommission(whose governing
council includes representativesof SecurityCouncil members)administersthe fund
as trustee for claimants.Subsequently,concernedabout humanitarianneeds in Iraq,
the council authorizedstates to importlimited amountsof Iraqioil with paymentsto
be made directly into a special escrow accountfor purchasesof food and medicine.
Similarly,an internationaloil pollutioncompensationfund is partof the IMO regime
governingoil spills in territorialwaters.
Building on the League of Nations mandatesystem, the UN charterestablishedan
internationaltrusteeshipsystem.Individualstatesweretypicallydesignatedas trustees
for various territories,with mixed results. But the charterdid establish standardsfor
trusteesand a trusteeshipcouncil to monitorthem. It even contemplatedthat the UN

ORGANIZATIONS 21
itself would performthe trusteefunctiondirectly,an extraordinaryexample of the IO

as a neutralparty.
TraditionalUN peacekeepingalso illustratesthe trusteefunction:UN forces patrol
or even control territoryto separate combatants,prevent conflict, and supervise
negotiatedcease-fires.UN neutralityalso allows majorpowers to supportpeacekeeping withoutchoosing sides amongfriendlystates,as in Cyprus.Blue-helmetneutrality
is crucial and guaranteedin multiple ways: operations are voluntary and require
continuingconsent of all parties,peacekeepersarefromcountrieswith no stakein the
conflict and under UN command, operationsare financed throughgeneral assessments, andtroopsareunarmed(observers)or lightly armedfor self-defense to prevent
uses of force inconsistentwith neutrality.But these restrictionscan limit the effectiveness of peacekeeping operations in some conflictual environments-as has been
evident in Bosnia. To deal with these limitations,the secretary-general'sAgenda for
Peace proposes a preventivetrusteefunction:UN-administereddemilitarizedzones,
established in advance of actual conflict to separatecontendingparties and remove
any pretextfor attack.
Neutral activities must be keenly attunedto the realities of internationalpower.
U Thant's quick withdrawalof the United Nations Emergency Force (UNEF) at
Egypt's request in 1967 was based on the legal principlerequiringconsent for UN
operationsbut equally reflectedthe realitythattwo contributingcountrieshad threatened to withdrawtroopsif Egyptianwishes were not respected.Nevertheless,like an
escrow agent, peacekeeping is effective when it furthersstate interests in limiting
conflict.
The Acheson-Lilienthal(Baruch)Plan would have createdan internationalagency
to managefissile material,contributedby the United States and the United Kingdom,
the existing nuclearpowers. This institutionalarrangement(which was not, of course,
adopted)resembleda trusteeshipwith the worldcommunityas beneficiary.It reflected
the vital interestsof donorstatesin preventingdestabilizingproliferation,but the plan
required a neutral trustee. The sponsors would not have been trusted to hold the
materialthemselves.
Similarly,underthe "commonheritage"principleof UNCLOS III, the convention
declaresthatrights to seabedresourcesare "vestedin mankindas a whole, on whose
behalf the Authorityshall act."The powers of the Seabed Authoritywere limited to
accordbetterwith marketprinciplesand U.S. interests,but it retainsits basic institutional structure,includingimportanttrusteecharacteristicsthatmay evolve over time.
IO as allocator. A neutralpartyoften allocates scarceresourcesamong claimants
to avoid paralyzingnegotiatingstandoffsand lingeringresentment:the parent,not the
children,slices the birthdaycake. IOs also serve this function.
The IAEA, for example, assists peaceful nationalnuclearprograms.It necessarily
evaluates proposedprojectsand allocates financial and personnelresources.Only a
neutralbody could be entrustedwith such responsibilityin a sensitive area.IFIs also
allocate scarce resourcesaccordingto projectworthiness.The WorldBank's charter
tries to guaranteeits neutralityby requiringthat it ignore the political characterof

22
potentialborrowers.The perceptionthatthe WorldBankpromotespromarketpolicies

on behalf of the Westernpowers and punishes governmentsthat pursue other goals
such as equity reduces its effectiveness. The WorldBank defends its neutralityby
presentingits policies as drivenby technicalanalysesratherthanvaluejudgments.It
has retaineda sufficientauraof neutralityto be entrustedwith allocatingfunds under
the Global EnvironmentFacility, the Ozone Trust Fund, and the climate change
convention.
IO as arbiter. According to Morgenthau(1967, 272), "despite ... deficiencies
[in] ... the legislative function [in internationalpolitics], a legal system might still be
capableof holdingin checkthepoweraspirationsof its subjectsif thereexistedjudicial
agencies thatcould speak with authoritywhenevera dissension occurredwith regard
to the existence or the importof a legal rule."Few internationalinstitutionsare truly
designed to restrainstate power, yet many help states resolve legal (and political)
disputes. Neutralityis essential for such institutions,just as for a judge in the law
merchantsystem (Milgrom, North, and Weingast 1990), the EuropeanCourt, or a
domestic court.
Infacilitative intervention,an IO operatesas "honestbroker"to reducetransaction
costs, improve informationaboutpreferences,transmitprivateoffers, and overcome
bargainingdeadlocks.ChapterVI of the UN charterrequiresstates to use traditional
measures-including good offices, mediation,conciliation and fact finding-to resolve disputes that threateninternationalpeace and security.The secretary-general
frequentlyprovides these services. The HumanRights Committeeprovides its good
offices in interstatedisputes and may appoint ad hoc conciliation commissions to
proposepossible settlements.Numerousinternationalconventions,fromthe Antarctic
to the NATOtreaties,providefor similarmeasuresif directnegotiationsfail. Even the
highly legalizedWTOunderstandingon disputesettlementallows membersto request
mediationor conciliationby the director-general.
In binding intervention,internationalinstitutionsissue legally binding decisions
with the consent of all parties.The mere possibility of binding externalintervention
may bring recalcitrantstates to the bargainingtable and make negotiatingpositions
more reasonable.The most common dispute resolution mechanism of this kind is
arbitration.Participatingstates agree on arbitrators,procedures,andjurisdictionand
decision.Whenagreementon these matterscannot
agreeto be boundby the arbitrators'
be reached,other neutralIOs sometimes fill the gap-as when the PermanentCourt
of Arbitrationselected the presidentof the U.S.-Iranclaims tribunal.
Arbitraltribunalsresolve disputeson an ad hoc basis, as in the 1941 U.S.-Canada
TrailSmelterarbitration,a leadingprecedentin internationalenvironmentallaw, or in
the secretary-general's"Rainbow Warrior"arbitrationbetween France and New
Zealand.They also handle classes of disputes such as the famous Alabama Claims
arbitrationfollowing the Civil War,the special claims commission for allied property
claims following World War II, and the Iran-U.S. claims tribunal.The following
commenton the RainbowWarriordisputeappliesto most of these cases: "Thissolution
is not withoutcritics in both countries.... However, ... the settlementprovedmuch

ORGANIZATIONS 23
more acceptable-precisely because of its unimpeachablesource-than would have

been the same, or anyother,solutionarrivedat solely by thepartiesthemselves.Neither
government... could be accusedby its internalcritics of having yielded to the other"
(Franckand Nolte 1993, 166).
Many internationalagreements,frombilateralcommercialtreatiesto the law of the
sea convention,rely on arbitrationthroughad hoc panels or more permanentinstitutions. The GATT-WTOdispute resolution process is similar to arbitration.In the
interest of neutrality,the director-generalmaintainsa roster of qualified panelists,
suggests panelists to disputants,and names the panel if the parties cannot agree.
NAFTA incorporatesseveral arbitrationprocedures, including an innovative one
wherebyarbitratorsreview nationalantidumpingandcountervailingduty decisions to
ensurethatnationallaw was followed. The InternationalCentrefor the Settlementof
InvestmentDisputes (ICSID),affiliatedwith the WorldBank,providesneutralfacilities for arbitrationsbetweenprivateinvestorsand host governments.
The principalinternationaljudicial authorityis the InternationalCourtof Justice
(ICJ). Unlike domestic courts, it must be grantedjurisdictionby partiesto a dispute.
Most cases have arisenundertreatiesthatinclude submissionto ICJjurisdiction.The
ICJ also issues advisory opinions to UN organsand specialized agencies. The court
has issued a numberof decisions of significance but has not been heavily used by
states; GATI panels, for instance,have issued many more decisions than the ICJ. A
relatively small numberof states have acceptedcompulsoryjurisdiction,and efforts
to use the court during high-profile disputes led France and the United States to
terminatetheir acceptance,althoughnot withoutcost. The EuropeanCourtof Justice
and the EuropeanCourtof HumanRights (which also requiresacceptanceofjurisdiction) have been more successful.Indeed,the former-whose judges arechosen "from
persons whose independence is beyond doubt"-approaches the authority of the
judicial institutions Morgenthauhad in mind. Its judges have played a leading
(independent)role in promotingEuropeanlegal integration(Burley andMattli 1993).
Otherinternationalinstitutions,includingthe WTO appellatebody,may also develop
into successfuljudicial agencies.
10 AS COMMUNITY
REPRESENTATIVEAND ENFORCER
In this section, we considerbroaderandmorecontroversialfunctionsof formalIOs,
some of which go beyond a simple state-centricapproach.We examine how states
structureanduse formalorganizationsto createandimplementcommunityvalues and
normsand to assist in the enforcementof internationalcommitments.This discussion
demonstratesfurtherhow the study of IOs forces different theoretical schools to
engage one another.We discuss these two functionsseparately,thentogetherin a brief
examinationof the role of the UN in the Gulf War-an example that also illustrates
the significance of IOs in situationsof violent conflict.

24
THE IO AS COMMUNITY REPRESENTATIVE
States establish IOs to act as a representativeor embodimentof a communityof

states. This was a centralaspirationin the postwarorganizationalboom and remains
an important,if only partiallyfulfilled, aspect of IO operationstoday.
Communityinstitutionstake severalforms. They may be inclusive bodies such as
the GeneralAssembly, the town squareof internationalpolitics, createdas a forum in
which common issues can be addressed.Withinsuch institutions,stateswork out and
express theircommon interestsandvalues. The process may be largelyconsensual,as
when statesconsidersome problemof commonconcernsuch as environmentalchange
or the behaviorof a rogue state,or it may entail one set of statespressuringanotherto
accept new principlessuch as humanrights,the oceans as a commons, or democracy.
Othercommunityinstitutions,such as the SecurityCouncil,arerepresentativebodies.
These incorporatethe major actors (as realism would predict) as well as states
representingotherinterests.These smallerbodies instantiatepoliticalbargainsin their
representationrules while providing a more efficient forum in which to deal with
issues, especially those requiringoperationalresponses. Finally,communityinstitutions such as the ICJ are structuredto promote independenceand neutrality,their
actions constrainedby a charge to act in the common interest.All three types can
advancecommunityinterestswith special legitimacy.
The UN, established by the Allies when they had unchecked dominance, was
undoubtedlyintendedto serve theirown purposes.It was also based on a conception
of sharedinterestsandvalues thatwent well beyondlaunderingor even neutrality.The
charter'sbroadgoals presupposeda directrelationbetween nationalwelfare, conditions aroundthe globe, and the peaceful workingof the internationalcommunityas a
whole. The principalgoal was to maintaininternationalpeace and security,and UN
organs were authorizedto intervene-not just mediate-in interstatedisputes that
threatenedpeace. Othergoals were to develop friendlyrelationsamong states based
on the principles of equal rights and self-determination,to promote fundamental
freedoms, and to promote cooperationon a wide range of global problems. Shared
interests in many of these areas-human rights, democracy,and liberal economic
relations-are still developing.
Perhapsthe most importantfunctionof communityorganizationsis to develop and
express communitynormsand aspirations.Althoughthe GeneralAssembly lacks the
Security Council's power of action, it can have substantialimpact on international
politics by expressingsharedvalues on issues like humanrights,apartheid,decolonization, and environmentalprotection in ways that legitimate or delegitimate state
conduct.The UniversalDeclarationof HumanRights is a strikingexample.Although
the declarationcannotbe enforced,its explicit and sweeping formulationof standards
has significantly affected state behavior.Its norms have been included in binding
treaties,and the declarationitself has been incorporatedinto some nationalconstitutions, thereby influencing the characterand preferencesof states and, thus, of the
internationalsystem itself. Althoughsmallerstateshave been disproportionatelyheld
to account on this issue, even large states like the formerSoviet Union and reputed
nuclearstates like South Africahave been affected.

ORGANIZATIONS 25
Similarly,althoughGATT(unlikethe WTO) was intentionallycreatedwith as few

attributesof an independentIO as possible, its contractingparties and council have
formulatedimportantpolicies for the tradingcommunity,including"differentialand
morefavorabletreatment"for developingcountries.Althoughcontested,thisprinciple
has been reflected in subsequenttrade negotiations and the generalized system of
preferences.
Courts as independentinstitutionsalso formulateand express communitypolicy.
By enunciating,elaborating,and applyingrulespublicly,they educatethe community
and strengthenunderlyingnorms(Abbott1992). A highly unusualIO, the UN tribunal
dealing with warcrimesin the formerYugoslavia,combinesthese publicjudicial roles
with the closely relatedpublic role of prosecutor.But states have not fully embraced
the community functions of courts. Even the ICJ is structuredto minimize its
communityrole:itsjurisdictionrestson partyconsent,andits decisions haveno formal
status as precedents.Yet, ICJ decisions are regularlyrelied on, and the court has on
importantoccasions acted as expositor of fundamentalcommunityvalues, as in the
Iranianhostages case and,manywould say,Nicaragua'ssuit againstthe UnitedStates.
These decisions have importantmoral authorityeven when they cannot be enforced
in the traditionalsense. Similarfunctions are performedby the Europeanand InterAmerican Commissions and Courts of Human Rights, and even by quasi-judicial
bodies like the ILO governingbody.
The most controversial example of community representationis the Security
of international
Council's"primary
responsibilityforthemaintenance
peaceandsecurity."
The council is empoweredto investigateany situationthatmight lead to international
friction and recommend means of resolving the conflict, including terms of
settlement. It is furtherempoweredunderChapterVII to "takeaction"against any
threatto peace. When using armedforce, however, the council has proceededmuch
as with economic sanctions, calling on members to give effect to measures it has
approved.
An 10 with these powers could overcome free-riderproblemshamperingdecentralizedeffortsto maintainpeace. But the SecurityCouncilhas the deeperrationaleof
representingthe community.Because local disputesmight spill over and disruptthe
largercommunity,they affect the generalwelfare. Such disputes should not be dealt
with exclusively by the partiesthemselves,or by thirdstatesinterveningfor theirown
privateinterests,but by collective bodies thatconsiderthe effects of the dispute and
of externalinterventionon thegeneralwelfare.ChapterVIIIof thecharterevenauthorizes
regional organizationslike the Organizationof AmericanStates (OAS) to deal with
local disputes, althoughthey only take "enforcementaction"with council approval,
lest such action itself threatenthe peace of the largercommunity.Finally, situating
privatedisputes in termsof communityinterestsand institutionsbrings a heightened
level of political and moralpressureto bearon disputantsand potentialintervenors.
The creationanddevelopmentof IOs often representdeliberatedecisions by states
to changetheirmutuallyconstitutedenvironmentand,thus,themselves.IOs can affect
the interests and values of states in ways that cannot be fully anticipated.Yet, it is
importantto stressthattheseprocessesareinitiatedandshapedby states.Furthermore,
IOs areconstrainedby institutionalprocedures-including financialcontributionsand

26
leadershipappointments-that are controlledby states and, ultimately,by the ability

of (some) states to withdraw,albeit at some cost. These possibilities and limitations
make IOs an importantwindow into the relationbetweenrationalistandconstructivist
analysis.
IOs AS MANAGERSOFENFORCEMENT
The role of IOs in ensuringcompliancewith internationalcommitmentscan best

be understood by integratingmanagerialand enforcement views of the process.
Observing high levels of compliance with internationalagreements, even though
strong enforcement provisions are rarely included or used, the managerialschool
concludes that IR has focused too heavily on coercive enforcement.In this view,
noncompliancetypically results not from deliberatecheating but from ambiguityin
agreements,insufficientstatecapacity,orchanginginternationalanddomesticcircumstances (Chayes and Chayes 1995; see also Mitchell 1994; Young 1994). Resolution
of such problems lies not in stronger enforcement but in better management of
compliance.Downs, Rocke, and Barsoom (1996) counterthat, withoutenforcement,
states will cheat on agreementsand that observed compliance levels largely reflect
shallow agreementsthatrequirelittle change in statebehavior.
An overly sharp distinction between managerialand enforcement functions is
misleading. For many significant day-to-day activities-especially ones involving
coordination-incentives to defect arerelativelysmall comparedwith the benefits of
cooperation;here,the managerialapproachis sufficient.In othercases, some enforcement may be necessary,at least potentially.IOs supportboth kinds of activities.More
important,the strictlydecentralizedmodels thatunderpinthe enforcementview do not
apply strictlyto the richerenvironmentof internationalpolitics, especially when states
are numerousand face significant informationalproblems. In these more complex
settings, IOs can manageenforcementactivities to make them more effective and to
limit their adverseside effects.
Many IO functionsidentifiedearlierare valuablein implementingthe managerial
approach.Ambiguitycan be resolvedthroughdisputeresolutionandotherthird-party
procedures,includingfact finding,good offices, interpretationof internationalagreements, andmediation.Stateincapacityis addresseddirectlyby financialandtechnical
assistance. Emerging compliance problems due to changing circumstancescan be
managed by IO political and judicial organs with authorityto interpretand adapt
agreementsand elaboratenorms.
When enforcementis needed,IOs can facilitatedecentralizedaction.They increase
the prospectof continuedinteraction,often across issues, and generalizereputational
effects of reneging across membersof the organization.Some IOs directly monitor
statebehavior,producingcredibleneutralinformationnecessaryfor effective enforcement. IOs furtherprovide forums in which suspicious actions can be explained,
lowering the risk that misperceptionswill upset cooperation,and in which pressure
can be broughton transgressorstates. In these ways, internationallegal discussions
about "mobilizationof shame"can be understoodnot in the moral sense of creating

ORGANIZATIONS 27
guilt among states but in an instrumentalsense of enhancingreputationaland other

incentives to abide by commitments.
IOs also have some directavenues of enforcement.These includerequirementsof
nationalreporting-wherein failureto reportitself indicatesimproperbehavior-and
the issuance of findings by the IO itself. The ILO has issued such reportswith respect
to labor practices,even in the case of powerful states such as the Soviet Union and
Britain.A less frequentsanctionoccursthroughresolutionscriticizingstatebehavior.
Such practices pressure, states to change their behavior both by impairing their
internationalstandingandby empoweringprivategroupsto pressurenationalgovernments, thus increasing"audiencecosts" (Fearon 1994). Currently,the G-7 states are
workingto empowerthe IMF to make findings on nationaleconomic policies and to
issue public criticismwith precisely these goals in mind.
A second means of direct enforcementis withholdingIO benefits, as the IAEA
suspendedtechnicalassistanceafterIsraelbombedan Iraqinuclearreactor.The IMF's
"conditionality"requirementsand the WorldBank's requirementson development
loans have expandedover the postwarperiod,and these agencies have frequentlyhad
strongeffects on the policies of memberstates.
Finally, IOs play an importantrole as managersof enforcement,authorizingand
giving meaningto retaliation,thus ensuringthatenforcementactivities arenot excessively disruptiveto the largerinternationalcommunity.Thispossibilityis differentially
developed. The GATTonly once authorizedretaliation,whereasWTOpracticeis still
emerging;the Security Council, by contrast,has authorizedeconomic sanctions on
numerousoccasions. Martin(1992, 245) finds IOs importantin managingeconomic
sanctionsbecausetheyprovidea frameworkfor side paymentsamongretaliatingstates
and increase incentives to cooperatein sanctions so as not to jeopardize the "broad
functional benefits these organizationsprovided."12Furthermore,such validationis
akin to laundering:when an IO legitimates retaliation,states are not vigilantes but
upholdersof communitynorms,values, and institutions.The IO imprimaturclarifies
retaliatorybehavior so that it will be seen by the target state for what it is, not as
noncooperationby the retaliatingstate, while reassuringthirdpartiesthatthe retaliating state is acting appropriately.(Again, influentialstates might seek IO approvalto
disguise their noncooperativeacts as retaliation,a form of dirty laundering,but this
practiceis limited by its self-defeatingcharacterand IO independence.)IO approval
frequentlylimits the severity and durationof state retaliation,as the WTO does by
limiting the amountof retaliationand the economic sectors targeted.Indeed, the IO
may negotiatea responsewith the retaliatingstateto maximizethird-partysupportfor
the action.Such managerialactivitiescounteract"echoeffects"andareimprovements
over strictlydecentralizedenforcement.
CHAPTER VII: THE USES AND LIMITS OF DIRECT ENFORCEMENT
The Security Council's experience with ChapterVII illuminates the role of the
community representativein constructinginterests,the possibility of more forcible
12. Martin(1992, 245) also finds it importantthatthe leading "sender"be willing to bear extra costs,
suggesting a possible limitationto IO enforcementcapacityin the absence of "leadership."

28
methods of direct enforcement,and, equally important,their limitations. As noted

above, the original conception of ChapterVII involved independentaction by the
SecurityCouncil on behalf of the communityof states,using militaryunits provided
"onits call"by memberstatesandguidedby a militarystaffcommittee.This was direct
enforcementexcept thatthe unitsto be deployed,even the membersof the committee,
were to be provided by states. This distinguishes ChapterVII from, say, the independent ability of the IMF to cut off funds to a country that violates its financial
commitments.Moreover,ChapterVII has neveroperatedas originallyintended.In the
two principalepisodes in which militaryforce has been used-Korea and the Gulf
War-the council insteadauthorizednationalmilitaryactions,led in both cases by the
United States. How are these episodes to be understood?
In the more cynical view, both are examples of dirty laundering.By obtaining
SecurityCouncil approval,the United Statescast essentiallyunilateralaction as more
legitimate collective action. The same interpretationcan be appliedto various OAS
enforcement actions against Castro's Cuba. Arguably,the organizationswere not
sufficiently independentof U.S. influenceto convertthe measurestakeninto genuine
community action. In the Gulf War,these measureswere transparentlynational:the
council simply called on other states to cooperate with the United States, which
was already operating in the Gulf theater, and coalition forces were visibly
dominated by the United States, whose troopseven retainedtheir own uniformsand
commanders.
Yet, these episodes can also be seen in a more affirmativelight. The institutional
underpinningsessential to the original vision of ChapterVII had never been put in
place: there were no agreementsfor the provision of nationalforces, no emergency
units standing by, no military staff committee. Lacking appropriateinstitutional
arrangements,the council carried out its community responsibilities in the only
practicableway, by shifting from direct to indirectenforcement,lending its institutional authorityto legitimateaction by willing nations.Its membershipstructureand
voting rules made the council sufficientlyindependentand representativeto perform
a genuine launderingfunction.13The United States, after all, assiduously courted
council approval(partlyby moving morecautiously)for reasonsof both domestic and
internationalpolitics. The imprimaturof the councilwas essentialto otherparticipants:
Middle Easternstates,for example,neededit to justify cooperationwith the coalition.
In this episode,just as Claude(1966, 74) putit morethan30 yearsago, "proclamations
of approvalor disapprovalby organsof the UnitedNations,deficient as they typically
are in ...
effective supportive power, are really important .... [S]tatesmen, by so
obviously attachingimportanceto them, have made them important."14

The affirmativeview sees the council, especially duringthe Gulf War,as representing the community of states. This representativestatus, not simply the formal
proceduresof ChapterVII, led the UnitedStatesandotherstatesto seek council action:
13. The currentdebateover the compositionof the councilreflectsthe idea thatsuch an institutionshould
be more representativeof the communityon behalf of which it acts.
14. See also Haas (1958) and, for a more skepticalview, see Slater(1969).

ORGANIZATIONS 29
SecurityCouncilresolutionson Iraqcarrieduniquepoliticalweightbecausethey came

from the established communityinstitutionwith primaryresponsibilityfor international peace and security. Resolutions condemning the Iraqi invasion of Kuwait
as unlawful,declaringvoid the incorporationof Kuwaititerritoryinto Iraq,denouncing human rights and environmental abuses by Iraqi forces, authorizing member
states to cooperate with U.S. forces, forcing the destructionof Iraqiweapons, and
holding Iraq financially responsible for its actions are clear expressions of the
shared moral and legal sense of organizedinternationalsociety. The 1Owas the locus
for giving meaning to state action. The United States, even as the clearly dominant
power in coercive activity,had good reasons to act not simply from might but from
persuasion.
Thus, realist, constructivist,and rational-regimeargumentscome togetherin considerationof the role of IOs in the Gulf crisis. Although some might preferto find a
singular"winner"amongthe threeexplanations,we believe each explainsa significant
partof the episode and thatany unidimensionalexplanationwould be incomplete.In
any event, IOs provide an importantlaboratoryin which to observe the operationof
these differentaspects of internationalpolitics.
CONCLUSION
For several decades, states have taken IOs more seriously than have scholars.
WhereasformalIOs have been seriouslyneglectedin the theoreticalstudy of international regimes, they have played a major role in many, if not most, instances of
interstatecollaboration.By taking advantageof the centralizationand independence
of IOs, states are able to achieve goals thatthey cannotaccomplishon a decentralized
basis. In some circumstances,the role of IOs extends even furtherto include the
development of common norms and practices that help define, or refine, states
themselves. At the same time, becauseissues of power and distributionarepervasive,
states are wary of allowing IOs too much autonomy.Thus, we do not claim that IOs
are supplantingthe states system. We do claim thatIOs providean importantsupplement to decentralizedcooperation that affects the nature and performanceof the
internationalsystem. Scholarsmust take IOs more seriously if they are to understand
interstaterelations.
Although we have presentedthe case for the importanceof formal institutionsin
internationalcooperation,the shortcomingsof many actualorganizationsgo without
saying. In addition, in emphasizing the possibilities for formal organizations,we
should not ignore the difficulty and even impossibilityof some of the tasks that are
presentedto them. Despite these severe limitations,the fact that IOs have not been
abandonedby states is testimonyto both their actualvalue and theirperhapsgreater
potential. A better theoreticaland empiricalunderstandingof formal organizations
should help improvetheirperformance.

30
REFERENCES
Sanger,David E. 1995. Big powers plan a worldeconomic bailoutfund. TheNew YorkTimes,8 June,D1.
Abbott,KennethW. 1992. GATTas a publicinstitution:The UruguayRoundandbeyond. BrooklynJournal
of InternationalLaw 31:31-85.
. 1993. Trustbutverify:Theproductionof informationin armscontroltreatiesandotherinternational
agreements.Cornell InternationalLaw Journal26:1-58.
Abbott, KennethW., and Duncan Snidal. 1997. The many faces of internationallegalization.Draftpaper
presentedat the Conferenceon Domestic Politics and InternationalLaw, June,Napa Valley,CA.
Alchian, Armen, and HaroldDemsetz. 1972. Production,informationcosts and economic organization.
AmericanEconomicReview62:777-95.
Amerasinghe,C. F. 1994. The law of the internationalcivil service. 2d rev. ed. Oxford,UK: Clarendon.
Barnett,Michael. 1993. Institutions,roles and disorder:The case of the Arab states system. International
Studies Quarterly37:271-96.
Bowett, D. W. 1982. The law of internationalinstitutions.4th ed. London:Stevens.
Berle, Adolf A., andGardinerC. Means. 1968. Themoderncorporationandprivateproperty.Rev. ed. New
York:Harcourt,Brace & World.
Brownlie, Ian. 1990. Principles of public internationallaw. 4th ed. New York:OxfordUniversityPress.
Burley,Anne-Marie,andWalterMattli.1993. Europebeforethe court:Apoliticaltheoryof legal integration.
InternationalOrganization47:41-76.
Calvert,RandallL. 1995. Rationalactors,equilibriumand social institutions.In Explainingsocial institutions, edited by Jack KnightandItai Sened, 57-94. Ann Arbor:Universityof MichiganPress.
Chayes,Abraham,andAntoniaHandlerChayes.1995. Thenew sovereignty:Compliancewith international
regulatoryagreements.Cambridge,MA: HarvardUniversityPress.
Claude, Inis Jr. 1966. Collective legitimizationas a political functionof the United Nations. International
Organization20:367-79.
Coase, R. H. 1937. The natureof the firm. Economica4:386-405.
Cox, Robert W., and Harold K. Jacobson,eds. 1973. The anatomy of influence. New Haven, CT: Yale
UniversityPress.
Downs, GeorgeW.,David M. Rocke, andPeterN. Barsoom.1996. Is the good news aboutcompliancegood
news aboutcooperation?InternationalOrganization50:379-406.
Drake, William, and Kalypso Nicolaidis. 1992. Ideas, interests and institutionalization.International
Dupuy,Rene Jean,ed. 1988.A handbookon internationalorganisations.Hingham,MA: KluwerAcademic.
Fearon,James. 1994. Domestic political audiencesand the escalationof internationaldisputes.American
Political Science Review88:577-92.
elections in earlyStuartEngland.In The
Parliamentary
Ferejohn,John. 1991. Rationalityandinterpretation:
economic approachto politics, editedby KristenR. Monroe,279-305. New York:HarperCollins.
Finnemore,Martha.1996. Nationalinterestsin internationalsociety. Ithaca,NY: CornellUniversityPress.
Franck,Thomas M., and Georg Nolte. 1993. The good offices function of the UN secretary-general.In
UnitedNations,dividedworld: The UN'srole in internationalrelations,2d ed., editedby AdamRoberts
and Benedict Kingsbury,143-82. Oxford,UK: Clarendon.
Garrett,Geoffrey,and BarryWeingast.1993. Ideas, interestsand institutions:Constructingthe European
Community'sinternalmarket.In Ideas and foreign policy: Beliefs, institutionsand political change,
edited by JudithGoldsteinand RobertO. Keohane,173-206. Ithaca,NY: CornellUniversityPress.
Glaser,Charles. 1995. Realists as optimists:Cooperationas self-help. InternationalSecurity19:50-93.
Goldstein, Judith,and Robert0. Keohane, eds. 1993. Ideas and foreign policy: Beliefs, institutionsand
political change. Ithaca,NY: CornellUniversityPress.
Grief, Avner,Paul Milgrom, and BarryR. Weingast.1994. Merchantgilds. Journal of Political Economy
102:745-76.
Haas, Ernst. 1958. Beyondthe nation-state.Stanford,CA: StanfordUniversityPress.
Haas, Peter M., ed. 1992. Knowledge,power and internationalpolicy coordination.InternationalOrganization 46 (Special issue): 1-390.

ORGANIZATIONS 31
Johnson,JamesD. 1991. Rationalchoice as a reconstructivetheory.In The economicapproachto politics,
edited by KristenR. Monroe, 113-42. New York:HarperCollins.
Kars, Margaret,and KarenMingst, eds. 1990. The United States and multilateralinstitutions.Boston:
Unwin Hyman.
Kennedy,David. 1987. The move to institutions.CardozoLaw Review 8:841-988.
Keohane,Robert0. 1984. Afterhegemony.Princeton,NJ: PrincetonUniversityPress.
Keohane,RobertO., AndrewMoravcsik,and Anne-MarieSlaughter.1997. A theoryof legalization.Draft
paperpresentedat the Conferenceon Domestic Politics andInternationalLaw,June,NapaValley, CA.
Kirgis, FredericL. Jr. 1993. Internationalorganizationsin their legal setting. 2d ed. St. Paul, MN: West.
Koremenos, Barbara,Charles Lipson, Brian Portnoy,and Duncan Snidal. 1997. Rational international
institutions.Paperpresentedat the AmericanPolitical Science Association Meetings, 28-31 August,
Washington,DC.
Krasner,StephenD., ed. 1983. Internationalregimes.Ithaca,NY: Corell UniversityPress.
Kratochwil,Friedrich.1996. Internationalorganization(s):Globalizationand the disappearanceof "publics." Unpublishedmanuscript.
Kratochwil,Friedrich,andJohnGerardRuggie. 1986. The stateof the arton the artof thestate.International
Levy, MarkA. 1993. Europeanacid rain:The power of tote-boarddiplomacy.In Institutionsforthe Earth:
editedby PeterM. Haas,RobertO. Keohane,
Sourcesof effectiveinternationalenvironmentalprotection,
and MarcA. Levy, 75-132. Cambridge,MA: MIT Press.
Lipson,Charles.1985. Standingguard:Protectingforeigncapital in the nineteenthand twentiethcenturies.
Berkeley:Universityof CaliforniaPress.
Marreese,Michael. 1986. CMEA:Effectivebutcumbersomepoliticaleconomy.InternationalOrganization
40:287-327.
Martin,Lisa L. 1992. Coercivecooperation.Princeton,NJ: PrincetonUniversityPress.
McNeely,ConnieL. 1995. Constructingthenation-state:Internationalorganizationandprescriptiveaction.
Westport,CT:Greenwood.
Mearsheimer,John. 1995. The false promiseof internationalinstitutions.InternationalSecurity19:5-49.
Milgrom,Paul R., Douglass C. North, and BarryR. Weingast.1990. The role of institutionsin the revival
of trade:The law merchant,privatejudges, andthe champagnefairs.Economicsand Politics 2 (1): 1-23.
Mitchell, Ronald. 1994. Regime design matters:Intentionaloil pollution and treatycompliance.International Organization48:425-58.
Moravcsik, Andrew. 1991. Negotiating the Single EuropeanAct: National interests and conventional
statecraftin the EuropeanCommunity.InternationalOrganization45:19-56.
Morgenthau,H. 1967. Politics amongnations. 4th ed. New York:Knopf.
Morrow,JamesD. 1994. The forms of internationalcooperation.InternationalOrganization48:387-424.
Ness, GaryD., and Steven R. Brechin. 1988. IOs as organizations.InternationalOrganization42:245-74.
Olson, Mancur,andRichardZeckhauser.1966. An economic theoryof alliances.Review of Economicsand
Statistics48:266-79.
Oye, KennethA., ed. 1986. Cooperationunderanarchy.Princeton,NJ: PrincetonUniversityPress.
Rochester,J. Martin.1986. The rise and fall of internationalorganizationas a field of study.International
Russett,Bruce, JohnR. Oneal, and David R. Davis. In press. The thirdleg of the Kantiantripodfor peace:
Internationalorganizationsand militarizeddisputes, 1950-85. InternationalOrganization42.
Sanford,Jonathon.1988. The WorldBank and poverty:The plight of the world's impoverishedis still a
majorconcernof the internationalagency.AmericanJournalof Economicsand Sociology 47:257-765.
Schermers,Henry,and Niels Blokker.1995. Internationalinstitutionallaw: Unitywithindiversity.3d rev.
ed. Cambridge,MA: KluwerLaw International.
Scott, W. Richard.1992. Organizations:Rational,naturaland open systems.3d ed. Englewood Cliffs, NJ:
PrenticeHall.
Shihata, IbrahimF. I., ed. 1991. The WorldBank in a changing world. Vol. 1. Norwell, MA: Kluwer
Academic Publishers.
, ed. 1995. The WorldBankin a changingworld.Vol.2. Cambridge,MA: KluwerLaw International.

32
Slater,Jerome.1969. Thelimitsof legitimationin internationalorganizations:The Organizationof American

States and the Dominicancrisis. InternationalOrganization23:48-72.
Snidal,Duncan. 1985a. Coordinationversusprisoners'dilemma:Implicationsfor internationalcooperation
and regimes. AmericanPolitical Science Review79:923-42.
.1985b. The game theoryof internationalpolitics. WorldPolitics 38:25-57.
. 1996. Politicaleconomy andinternationalinstitutions.InternationalReviewof Lawand Economics
16:121-37.
Sohn, Louis. 1950. Cases and other materialson world law. Brooklyn,NY: Foundation.
. 1967. Cases on UnitedNations law. 2d rev. ed. Brooklyn,NY: Foundation.
Stein,Arthur.1983. Coordinationandcollaboration:Regimesin an anarchicworld.In Internationalregimes,
edited by StephenD. Krasner,115-40. Ithaca,NY: CornellUniversityPress.
Strange,Susan. 1983. Cave! Hic dragones:A critiqueof regime analysis. In Internationalregimes,edited
by StephenD. Krasner,337-54. Ithaca,NY: CornellUniversityPress.
Trachtman,Joel P. 1996. The theory of the firm and the theory of internationaleconomic organization:
Towardcomparativeinstitutionalanalysis.Unpublishedmanuscript.
Vienna conventionfor the protectionof the ozone layer. 1985. InternationalLegal Materials 26:1529-40.
Weber,Steven. 1994. The EuropeanBank for reconstructionand development.InternationalOrganization
48 (1): 1-38.
Wendt,Alex. 1992. Anarchyis whatstatesmakeof it:The social constructionof powerpolitics. International
. 1995. Constructinginternationalpolitics. InternationalSecurity20:71-81.
Williamson,Oliver. 1985. The economic institutionsof capitalism.New York:Free Press.
. 1994. The mechanismsof government.Oxford,UK: OxfordUniversityPress.
Yarbrough,Beth V., and RobertM. Yarbrough.1992. Cooperationand governance in internationaltrade.
Princeton,NJ: PrincetonUniversityPress.
Young, OranR. 1994. Internationalgovernance.Ithaca,NY: CornellUniversityPress.


Sage Publications, Inc

Uploaded by

Copyright:

Available Formats

Sage Publications, Inc

Uploaded by

Document Information

Original Description:

Original Title

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

Sage Publications, Inc

Uploaded by

Copyright:

Available Formats

Why States Act through Formal International Organizations

Author(s): Kenneth W. Abbott and Duncan Snidal

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

Why States Act through Formal

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

efficiency of collective activities and enhancethe organization'sability to affect the

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

PUTTING IOs INTO THEORY

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

purposes, however, it performs a useful service by emphasizing that institutional

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

THE FUNCTIONS OF IOs:

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

The organizationalstructureof IOs enhanceseven the passive virtuesrecognized

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

Most IOs performmore extensive supportivefunctions.Law-makingconferences

IOs do more than supportintergovernmentalnegotiations;they manage a variety

7. We reservefor the following sectiondiscussionof those functionsthatturndirectlyon independence

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

IO operations also significantly influence the capabilities, understandings,and

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

NonfinancialIOs providesimilaradvantages.The publichealthactivitiesof WHO,

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

the outputs,includingtechnologicalexternalities,are shared.Projectslike CERN and

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

ness of an established rule-makingprocedurerequiresthat powerful states respect

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

IndependentIOs promoteintergovernmentalcooperationin more proactiveways

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

marketpolicies. Thus, IO autonomyremainsboundedby stateinterestsandpower,as

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

IO as neutral information provider. Regime theory recognizes the importance of

informationbut does not emphasizedifferencesin its quality.Informationcreatedor

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

itself would performthe trusteefunctiondirectly,an extraordinaryexample of the IO

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

potentialborrowers.The perceptionthatthe WorldBankpromotespromarketpolicies

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

more acceptable-precisely because of its unimpeachablesource-than would have

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

THE IO AS COMMUNITY REPRESENTATIVE

States establish IOs to act as a representativeor embodimentof a communityof

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

Similarly,althoughGATT(unlikethe WTO) was intentionallycreatedwith as few

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

leadershipappointments-that are controlledby states and, ultimately,by the ability

The role of IOs in ensuringcompliancewith internationalcommitmentscan best

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

guilt among states but in an instrumentalsense of enhancingreputationaland other

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM

methods of direct enforcement,and, equally important,their limitations. As noted

effective supportive power, are really important .... [S]tatesmen, by so

obviously attachingimportanceto them, have made them important."14

This content downloaded from 210.212.199.162 on Mon, 17 Nov 2014 14:06:11 PM