Beruflich Dokumente
Kultur Dokumente
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/2706525?seq=1&cid=pdf-reference#references_tab_contents
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.
International Organization Foundation, MIT Press, University of Wisconsin Press and Cambridge University
Press are collaborating with JSTOR to digitize, preserve and extend access to International Organization.
http://www.jstor.org
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for
internationalregimes
Robert 0. Keohane
325
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
326 International Organization
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 327
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
328 International Organization
functions, and attempts to explain their behavior on the basis of environ-
mentalfactors such as the competitiveness of markets. It is thereforea sys-
temic theory, unlike the so-called "behavioral theory of the firm," which
examines the actors for internalvariationsthat could account for behavior
not predictedby microeconomictheory.
A systemic focus permits a limitation of the number of variables that
need to be considered. In the initial steps of theory-building,this is a great
advantage:attemptingto take into account at the outset factors at the foreign
policy as well as the systemic level would lead quickly to descriptive com-
plexity and theoreticalanarchy.Beginningthe analysis at the systemic level
establishes a baseline for future work. By seeing how well a simple model
accounts for behavior, we understandbetter the value of introducingmore
variablesand greatercomplexity into the analysis. Withoutthe systemic mi-
croeconomic theory of the firm, for instance, it would not have been clear
what puzzles needed to be solved by an actor-orientedbehavioraltheory.
A systems-level examinationof changes in the strength and extent of
internationalregimesover time could proceed throughhistoricaldescription.
We could examine a large numberof cases, attemptingto extract generaliza-
tions about patterns from the data. Our analysis could be explicitly com-
parative, analyzing different regimes within a common analytical frame-
work, employing a methodology such as George's "focused compari-
son."5 Such a systematic comparativedescriptioncould be quite useful, but
it would not provide a theoretical frameworkfor posing questions of why,
and under what conditions, regimes should be expected to develop or be-
come stronger. Posing such fundamentalissues is greatly facilitated by a
priori reasoningthat makes specific predictionsto be comparedwith empiri-
cal findings. Such reasoning helps us to reinterpretpreviously observed
patternsof behavior as well as suggestingnew questions about behavior or
distinctionsthat have been ignored:it has the potentialof "discovering new
facts."6 This can be useful even in a subject such as internationalpolitics,
where the variety of relevantvariablesis likely to confound any comprehen-
sive effort to build deductive theory. Deductive analysis can thus be used in
interpretationas well as in a traditional strategy of theory-buildingand
hypothesis-testing.
This analysis follows the traditionof microeconomictheory by focusing
on constraints and incentives that affect the choices made by actors.7We
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 329
8 I am indebtedto AlexanderJ. Field for makingthe importanceof this pointclear to me. See
his paper, "The Problemwith Neoclassical InstitutionalEconomics: A Critiquewith Special
Referenceto the North/ThomasModel of Pre-1500 Europe,"Explorationsin EconomicHis-
tory 18 (April 1981).
9 Lance E. Davis and Douglass C. North adopt this strongform of rationalisticexplanation
when they arguethat "an institutionalarrangementwill be innovatedif the expected net gains
exceed the expected costs." See their volume, Institutional Change and American Economic
Growth(Cambridge:CambridgeUniversity Press, 1971).
10Two of the classic works are JamesMarchand HerbertSimon,Organizations(New York:
Wiley, 1958);and RichardCyert and James March,The Behavioral Theoryof the Firm (En-
glewood Cliffs, N.J.: Prentice-Hall,1963).
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
330 International Organization
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 331
14
WilliamFellner, Competitionamong the Few (New York: Knopf, 1949).
15
Donald J. Puchala, "Domestic Politics and Regional Harmonizationin the European
Communities," World Politics 27,4 (July 1975), p. 509.
16 There are exceptions to this generalization,such as Tiebout's "votingwith the feet" mod-
els of populationmovementsamongcommunities.Yet only one chapterof fourteenin a recent
survey of the public-choiceliteratureis devoted to such models, which do not focus on au-
thoritativedecision-makingprocesses. See Dennis C. Mueller, Public Choice (Cambridge:
CambridgeUniversityPress, 1980).For a brilliantlyinnovativework on "exit" versus "voice"
processes, see Albert0. Hirschman,Exit, Voice, and Loyalty(Cambridge:HarvardUniversity
Press, 1970).
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
332 International Organization
17
Anyone who has thought about Hobbes's tendentious discussion of "voluntary"
agreementsin Leviathan realizes the dangers of casuistry entailed in applying voluntaristic
analysisto politics, especially when obligationsare inferredfrom choices. This articlefollows
Hobbes's distinctionbetween the structureof constraintsin a situation,on the one hand, and
actor choices, on the other; but it does not adopt his view that even severely constrained
choices ("your freedomor your life") create moralor politicalobligations.
18 Kenneth N. Waltz, Theory of International Politics (Reading, Mass.: Addison-Wesley,
1979).
19Externalitiesexist wheneveran actingunitdoes not bearall of the costs, or fails to reapall
of the benefits, that result from its behavior. See Davis and North, InstitutionalChange and
American Economic Growth, p. 16.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 333
20
Olson, The Logic of Collection Action; BruceM. RussettandJohnD. Sullivan,"Collective
Goods and InternationalOrganization,"with a comment by MancurOlson Jr., International
Organization 25,4 (Autumn1971);JohnGerardRuggie,"CollectiveGoods and FutureInterna-
tional Collaboration,"American Political Science Review 66,3 (September 1972); Duncan
Snidal, "Public Goods, PropertyRights, and Political Organization,"International Studies
Quarterly 23,4 (December 1979),p. 544.
21 Keohane, "The Theory of HegemonicStability";CharlesP. Kindleberger,The World in
Depression, 1929-1939 (Berkeley: University of CaliforniaPress, 1974);MancurOlson and
RichardZeckhauser,"An EconomicTheoryof Alliances,"Review of Economics and Statistics
48,3 (August 1966), reprintedin Bruce M. Russett, ed., Economic Theories of International
Politics (Chicago:Markham,1968).For a criticalappraisalof work placingemphasison public
goods as a rationalefor forminginternationalorganizations,see JohnA. C. Conybeare,"Inter-
nationalOrganizationsand the Theory of PropertyRights," International Organization 34,3
(Summer1980),especially pp. 329-32.
22 My use of the word "functions" here is meant to designate consequences of a certain
patternof activity, particularlyin termsof the utilityof the activity;it is not to be interpretedas
an explanationof the behaviorin question,since there is no teleologicalpremise,or assumption
that necessity is involved. Understandingthe functionof internationalregimeshelps, however,
to explainwhy actors have an incentiveto create them, and may thereforehelp to makebehav-
ior intelligiblewithin a rational-choicemode of analysisthat emphasizesthe role of incentives
and constraints.For useful distinctionson functionalism,see Ernest Nagel, The Structure of
Scientific Explanation (New York: Harcourt,Brace, 1961),especially "Functionalismand So-
cial Science," pp. 520-35. I am gratefulto RobertPackenhamfor this referenceanddiscussions
of this point.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
334 International Organization
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 335
markets, and that the history of the developmentof credit marketscould be informativefor
studentsof internationalregimes. The analogyseems to hold. RichardEhrenbergreportsthat
the developmentof credit arrangementsin medieval EuropeanBourses reduced transaction
costs (since money did not need to be transportedin the form of specie) and providedhigh-
qualityinformationin the formof merchants'newslettersandexchangesof informationat fairs:
"duringthe Middle Ages the best informationas to the course of events in the world was
regularlyto be obtainedin the fairsandthe Bourses" (p. 317).The Boursesalso providedcredit
ratings,which providedinformationbut also served as a crude substitutefor effective systems
of legal liability.Althoughthe descriptionsof credit marketdevelopmentin works such as that
by Ehrenbergare fascinating,I have not been able to find a historically-groundedtheory of
these events. See RichardEhrenberg,Capital and Finance in the Age of the Renaissance: A
Study of the Fuggers and Their Connections, translatedfromthe Germanby H. M. Lucas (New
York: Harcourt,Brace, no date), especially chap. 3 (pp. 307-333).
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
336 International Organization
at any given price will vary directly with the desirabilityof agreementsto
states and with the ability of internationalregimes actually to facilitate the
makingof such agreements.The conditionfor the theory's operation(thatis,
for regimesto be formed)is that sufficientcomplementaryor common inter-
ests exist so that agreementsbenefitingall essential regime memberscan be
made.
The value of theories of marketfailurefor this analysis rests on the fact
that they allow us to identify more precisely barriersto agreements. They
therefore suggest insights into how internationalregimes help to reduce
those barriers, and they provide richer interpretationsof previously ob-
served, but unexplained, phenomenaassociated with internationalregimes
and internationalpolicy coordination.In addition,concepts of marketfailure
help to explainthe strengthand extent of internationalregimesby identifying
characteristicsof internationalsystems, or of internationalregimes them-
selves, that affect the demandfor such regimesand therefore,given a supply
schedule, their quantity.Insightsfromthe market-failureliteraturetherefore
take us beyond the trivial cost-benefit or supply-demandpropositionswith
which we began, to hypotheses about relationshipsthat are less familiar.
The emphasis on efficiency in the market-failureliteratureis consistent
with our constraint-choiceanalysis of the decision-makingprocesses leading
to the formationand maintenanceof internationalregimes. Each actor must
be as well or better off with the regime than without it-given the prior
structureof constraints.This does not imply, of course, that the whole pro-
cess leadingto the formationof a new internationalregimewill yield overall
welfare benefits. Outsidersmay suffer; indeed, some internationalregimes
(such as alliancesor cartel-typeregimes) are specificallydesignedto impose
costs on them. These costs to outsiders may well outweigh the benefits to
members. In addition, powerful actors may manipulateconstraintspriorto
the formationof a new regime. In that case, althoughthe regimeper se may
achieve overall welfare improvements compared to the immediately pre-
ceding situation,the results of the joint process may be inferiorto those that
existed before the constraintswere imposed.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 337
problems, transactions cost and informational imperfections. Questions
of information, involving uncertainty and risk, will receive particular
attention, since their exploration has rich implications for interpretation
and future research.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
338 International Organization
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demandfor internationalregimes 339
ture does not prevent the development of bits and pieces of law.29Regimes
are much more importantin providingestablished negotiatingframeworks
(reducingtransactionscosts) and in helpingto coordinateactor expectations
(improvingthe quality and quantity of informationavailable to states). An
explanationof these two functions of internationalxegimes,with the help of
microeconomicanalysis, will lead to hypotheses about how the demandfor
internationalregimes should be expected to vary with changes in the nature
of the internationalsystem (in the case of transactionscosts) and about ef-
fects of characteristicsof the internationalregime itself (in the case of infor-
mation).
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
340 InternationalOrganization
them. Where issue density is low, ad hoc agreementsare quite likely to be
adequate:differentagreementswill not impingeon one anothersignificantly,
and there will be few economies of scale associated with establishinginter-
national regimes (each of which would encompass only one or a few
agreements).Whereissue density is high, on the other hand, one substantive
objective may well impingeon anotherand regimes will achieve economies
of scale, for instance in establishingnegotiatingproceduresthat are applica-
ble to a variety of potential agreementswithin similar substantive areas of
activity.3'
Furthermore,in dense policy spaces, complex linkages will develop
among substantive issues. Reducing industrial tariffs without damaging
one's own economy may depend on agriculturaltariff reductions from
others; obtaining passage through straits for one's own warships may de-
pend on wider decisions taken about territorialwaters; the sale of food to
one country may be more or less advantageousdependingon other food-
supply contracts being made at the same time. As linkages such as these
develop, the organizationalcosts involved in reconcilingdistinct objectives
will rise anddemandsfor overallframeworksof rules, norms, principles,and
procedures to cover certain clusters of issues-that is, for international
regimes-will increase.
Internationalregimes therefore seem often to facilitate side-payments
among actors within issue-areas covered by comprehensive regimes, since
they bringtogethernegotiatorsto consider a whole complex of issues. Side-
paymentsin general are difficultin world politics and raise serious issues of
transaction costs: in the absence .of a price system for the exchange of
favors, these institutionalimperfectionswill hinder cooperation.32Interna-
tional regimes may provide a partialcorrective.33The well-knownliterature
on "spillover" in bargaining,relatingto the EuropeanCommunityand other
integrationschemes, can also be interpretedas being concerned with side-
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 341
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
342 International Organization
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 343
39The classic discussion is in Thomas C. Schelling, The Strategy of Conflict (1960; Cam-
bridge:HarvardUniversityPress, 1980),chap. 4, "Towarda Theoryof InterdependentDeci-
sion." See also Schelling,Micromotivesand Macrobehavior(New York: Norton, 1978).
40 For an interestingdiscussion of regimes in these terms, see the paper in this volume by
Oran R. Young. On conventions, see David K. Lewis, Convention:A PhilosophicalStudy
(Cambridge:CambridgeUniversityPress, 1969).
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
344 International Organization
cheat without the other being allowed to do so). As Stein points out in this
volume, these situations are characterizedtypically by unstable equilibria.
Withoutenforcement, actors have incentives to deviate from the agreement
point:
[Each] actor requires assurances that the other will also eschew its
rationalchoice [and will not cheat, and] such collaborationrequires a
degree of formalization. The regime must specify what constitutes
cooperation and what constitutes cheating.41
In such situationsof strategicinteraction,as in oligopolisticcompetition
and world politics, systemic constraint-choicetheory yields no determinate
results or stable equilibria.Indeed, discussions of "blackmailing"or games
such as "prisoners' dilemma" indicate that, under certain conditions, sub-
optimal equilibria are quite likely to appear. Game theory, as Simon has
commented, only illustrates the severity of the problem; it does not solve
it. 42
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 345
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
346 International Organization
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 347
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
348 International Organization
52
Williamson,"A DynamicTheory of InterfirmBehavior," p. 592, originalitalics.
53Powerand Interdependence,especiallypp. 54-58 and 146-53. LindaCahnalso foundlags,
particularlyin the wheat regime;see "National Power and InternationalRegimes."
54 Power and Interdependence, p. 55.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 349
basis for hypotheses that could guide researchon fluctuationsin the strength
and extent of internationalregimes.
Understandingthe value of governmentalopenness for makingmutually
beneficial agreements helps to account for the often-observed fact that ef-
fective internationalregimes-such as the GATT in its heyday, or the Bret-
ton Woods internationalmonetary regime55-are often associated with a
great deal of informalcontact and communicationamong officials. Govern-
ments no longer act within such regimes as unitary, self-containedactors.
"Transgovernmental"networks of acquaintance and friendship develop,
with the consequences that supposedly confidentialinternaldocuments of
one governmentmay be seen by officials of another;informalcoalitions of
like-mindedofficials develop to achieve common purposes; and critical dis-
cussions by professionals probe the assumptions and assertions of state
policies.56These transgovernmentalrelationshipsincrease opportunitiesfor
cooperation in world politics by providingpolicy makers with high-quality
informationabout what their counterpartsare likely to do. Insofar as they
are valued by policy makers, they help to generate demandfor international
regimes.
The information-producing"technology" that becomes embedded in a
particularinternationalregime also helps us to understandwhy the erosion
of American hegemony duringthe 1970s has not been accompaniedby an
immediatecollapse of internationalregimes, as a theory based entirely on
supply-sidepublic goods analysis would have predicted. Since the level of
institutionalizationof postwarregimeswas exceptionallyhigh, with intricate
and extensive networksof communicationamongworking-levelofficials, we
should expect the lag between the decline of American hegemony and the
disruptionof internationalregimes to be quite long and the "inertia" of the
existing regimes relatively great.
The majorhypothesis to be derived from this discussion of information
is that demandfor internationalregimes should be in part a function of the
effectiveness of the regimes themselves in providinghigh-qualityinforma-
tion to policy makers. The success of the institutions associated with a
regime in providing such information will itself be a source of regime
persistence.
Three inferences can be made from this hypothesis. First, regimes ac-
companiedby highly regularizedproceduresand rules will provide more in-
formationto participantsthan less regularizedregimesand will therefore,on
55 On the GATT, see GardnerPatterson,Discriminationin InternationalTrade:The Policy
Issues (Princeton:PrincetonUniversityPress, 1966);on the internationalmonetaryregime,see
Robert W. Russell, "TransgovernmentalInteractionin the InternationalMonetarySystem,
1960-1972," International Organization 27,4 (Autumn 1973) and Fred Hirsch, Money Interna-
tional, rev. ed. (Harmondsworth,England:PelicanBooks, 1969),especiallychap. 11, "Central
BankersInternational."
56 Robert0. Keohane and Joseph S. Nye, "Transgovernmental Relationsand International
Organizations,"WorldPolitics 27,1 (October 1974):39-62.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
350 International Organization
57 These first three inferencesfocus only on the demand side. To understandthe degree to
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 351
help to define the dimensionsof the problemand provide some guidancefor
thinkingabout the future consequences of present actions.
5. Copingwith uncertainties:insuranceregimes
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
352 International Organization
Calleo, ed., Money and the ComingWorldOrder(New York: New York UniversityPress for
the Lehrman Institute, 1978); Ronald McKinnon, Money in International Exchange: The Con-
vertibleCurrencySystem (New York: OxfordUniversity Press, 1979).
60
Arrow, Essays in the Theory of Risk-Bearing, pp; 134-43.
61
In personalcorrespondence,RobertJervis has suggestedan interestingqualificationto this
argument.He writes:"If we look at relationsthat involve at least the potentialfor highconflict,
then schemes that tie the fates of all the actors togethermay have utilityeven if the actors are
concernedabout catastrophicevents which will affect them all. They can worry that if some
states are not affected, the latterwill be muchstrongerthanthe ones who have been injured.So
it wouldmakesense for themto workout a schemewhichwould insurethata disasterwouldnot
affect theirrelativepositions, even thoughthis would not meanthatthey would all not be worse
off in absolute terms." The point is certainlywell taken, althoughone may wonder whether
such an agreementwould in fact be implementedby the states that would make large relative
gains in the absence of insurancepayments.
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 353
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
354 International Organization
6. Conclusions
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions
The demand for international regimes 355
This content downloaded from 150.203.154.121 on Wed, 24 Feb 2016 01:23:53 UTC
All use subject to JSTOR Terms and Conditions