moravcik, kritik

Jeffrey W Legro andAndrew Moravcsik

ealism the oldestand most prominent theoretical paradigm in international relations is introuble The problem is not lack of interest Realism remains the primary oralternative theory in virtually every major book and article addressing generaltheories of world politics particularly in security affairs Controversies be-tween neorealism and its critics continue to dominate international relationstheory debates Nor is the problem realismrsquos purported inability to make pointpredictions Many specic realist theories are testable and there remains muchglobal conict about which realism offers powerful insights Nor is the problemthe lack of empirical support for simple realist predictions such as recurrentbalancing or the absence of plausible realist explanations of certain salientphenomena such as the Cold War the ldquoend of historyrdquo1or systemic change ingeneral Research programs advance after all by the renement and improve-ment of previous theories to account for anomalies There can be little doubtthat realist theories rightfully retain a salient position in international relationstheory

International Security Vol 24 No 2 (Fall 1999) pp 5ndash55copy 1999 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology

5

Jeffrey W Legro is Associate Professor of Government and Foreign Affairs University of Virginia AndrewMoravcsik is Professor of Government Harvard University

We are grateful to Charles Glaser Joseph Grieco Gideon Rose Randall Schweller Jack SnyderStephen Van Evera Stephen Walt William Wohlforth and Fareed Zakaria for providing repeateddetailed corrections and rebuttals to our analysis of their respective work to Robert Art MichaelBarnett James Caporaso Thomas Christensen Dale Copeland Michael Desch David DesslerColin Elman Miriam Fendius Elman Daniel Epstein Martha Finnemore Stefano Guzzini GuntherHellmann Robert Jervis Peter Katzenstein Robert Keohane Stephen Krasner John MearsheimerJohn Owen Robert Paarlberg Stephen Rosen Anne-Marie Slaughter Nigel Thalakada AlexanderWendt and participants at colloquia at Brown University and Harvard Universityrsquos John M OlinInstitute for Strategic Studies for more general comments and to Duane Adamson and AronFischer for research assistance

1 We agree with much of the analysis in John Vasquez ldquoThe Realist Paradigm and Degenerativevs Progressive Research Programs An Appraisal of Neotraditional Research on Waltzrsquos BalancingPropositionrdquo American Political Science Review Vol 91 No 4 (December 1997) pp 899ndash912 But wedo not agree among other things that balancing behavior per se provides a strong test of realismor that realism is beyond redemption On various criticisms see also Francis Fukuyama The Endof History and the Last Man (New York Free Press 1992) Richard Ned Lebow and ThomasRisse-Kappen eds International Relations and the End of the Cold War (New York ColumbiaUniversity Press 1995) and Paul W Schroeder ldquoHistorical Reality vs Neorealist Theoryrdquo inMichael E Brown Sean M Lynn-Jones and Steven E Miller eds The Perils of Anarchy Contem-porary Realism and International Security (Cambridge Mass MIT Press 1995) pp 421ndash461 Peter JKatzenstein Robert O Keohane and Stephen D Krasner ldquoInternational Organization and the Studyof World Politicsrdquo International Organization Vol 52 No 4 (Autumn 1998) pp 670ndash674 and

The central problem is instead that the theoretical core of the realist approachhas been undermined by its own defendersmdashin particular so-called defensiveand neoclassical realistsmdashwho seek to address anomalies by recasting realismin forms that are theoretically less determinate less coherent and less distinc-tive to realism Realists like EH Carr Hans Morgenthau and Kenneth Waltzsought to highlight the manipulation accumulation and balancing of powerby sober unsentimental statesmen focusing above all on the limits imposedon states by the international distribution of material resources They viewedrealism as the bulwark against claims about the autonomous inuence ofdemocracy ideology economic integration law and institutions on worldpolitics Many recent realists by contrast seek to redress empirical anomaliesparticularly in Waltzrsquos neorealism by subsuming these traditional counterar-guments The result is that many realists now advance the very assumptionsand causal claims in opposition to which they traditionally and still claim todene themselves

This expansion would be unproblematic even praiseworthy if it took placeon the basis of the further elaboration of an unchanging set of core realistpremises It would be quite an intellectual coup for realists to demonstratemdashasrealists from Thucydides through Machiavelli and Hobbes to Morgenthausought to domdashthat the impact of ideas domestic institutions economic inter-dependence and international institutions actually reects the exogenousdistribution and manipulation of interstate power capabilities Some contem-porary realists do continue to cultivate such arguments yet such efforts appeartoday more like exceptions to the rule Many among the most prominent andthoughtful contemporary realists invoke instead variation in other exogenousinuences on state behaviormdashstate preferences beliefs and international insti-tutionsmdashto trump the direct and indirect effects of material power Such factorsare consistently treated as more important than power We term such an ap-proach ldquominimal realismrdquo because it retains only two core assumptionsmdashlittlemore than anarchy and rationalitymdashneither of which is distinctively realist By

Benjamin Frankel ed Realism Restatements and Renewal (London Frank Cass 1996) pp xindashxiiFor rejoinders see Kenneth N Waltz ldquoEvaluating Theoriesrdquo American Political Science Review Vol91 No 4 (December 1997) pp 913ndash918 Thomas J Christensen and Jack Snyder ldquoProgressiveResearch and Degenerative Alliancesrdquo American Political Science Review Vol 91 No 4 (December1997) pp 899ndash912 Colin Elman and Miriam Fendius Elman ldquoCorrespondence History vs Neo-realism A Second Lookrdquo International Security Vol 20 No 1 (Summer 1995) pp 182ndash193 Elmanand Elman ldquoLakatos and Neorealism A Reply to Vasquezrdquo American Political Science Review Vol91 No 4 (December 1997) pp 923ndash926 Randall L Schweller ldquoNew Realist Research on AlliancesRening not Refuting Waltzrsquos Balancing Propositionrdquo American Political Science Review Vol 91No 4 (December 1997) pp 927ndash930 and Stephen M Walt ldquoThe Progressive Power of RealismrdquoAmerican Political Science Review Vol 91 No 4 (December 1997) pp 931ndash935

International Security 242 6

reducing realist core assumptions to anarchy and rationality minimal realismbroadens realism so far that it is now consistent with any inuence on rationalstate behavior including those once uniformly disparaged by realists as ldquole-galistrdquo ldquoliberalrdquo ldquomoralistrdquo or ldquoidealistrdquo The concept of ldquorealismrdquo has thusbeen stretched to include assumptions and causal mechanisms within alterna-tive paradigms albeit with no effort to reconcile the resulting contradictions2

Contemporary realists lack an explicit nontrivial set of core assumptions Thosethey set forth either are not distinctive to realism or are overtly contradictedby their own midrange theorizing In sum the malleable realist rubric nowencompasses nearly the entire universe of international relations theory (in-cluding current liberal epistemic and institutionalist theories) and excludesonly a few intellectual scarecrows (such as outright irrationality widespreadself-abnegating altruism slavish commitment to ideology complete harmonyof state interests or a world state)

The practical result is that the use of the term ldquorealistrdquo misleads us as to theactual import of recent empirical research The mislabeling of realist claims hasobscured the majormdashand ironicmdashachievement of recent realist work namelyto deepen and broaden the proven explanatory power and scope of the estab-lished liberal epistemic and institutionalist paradigms The more precise themidrange theories and hypotheses contemporary realists advance the clearerit becomes that such claims are not realist Some subsume in a theoreticallyunconstrained way nearly all potential rationalist hypotheses about state be-havior except those based on irrational or incoherent behavior Others relyexplicitly on variation in exogenous factors like democratic governanceeconomic interdependence systematic misperception the transaction costndashreducing properties of international institutions organizational politics andaggressive ideology This is obscured because most realists test their favoredexplanations only against other variants of realismmdashnormally Waltzian neo-realismmdashrather than against alternative liberal epistemic and institutionalisttheories as they once did Recent realist scholarship unwittingly throws therealist baby out with the neorealist bathwater

Our criticism of recent realist theory is not a semantic quibble an invitationto yet another purely abstract debate about the labeling and relabeling of

2 Giovanni Sartori ldquoConcept Misinformation in Comparative Politicsrdquo American Political ScienceReview Vol 64 No 4 (December 1970) pp 1033ndash1053 This is another way in which our critiquediffers from that of Vasquez who has also charged that the realist paradigm is degeneratingVasquez argues that ldquothere is no falsication before the emergence of better theoryrdquo and thatalternative paradigms do not exist We demonstrate that they do Vasquez ldquoThe Realist Paradigmrdquop 910

Is Anybody Still a Realist 7

international relations ideal-types or a philosophical inquiry into the develop-ment of research paradigms It is a direct challenge to the theoretical distinct-iveness of contemporary realism one with immediate and signicant practicalimplications Recent realist theory has become a hindrance rather than a helpin structuring theoretical debates guiding empirical research and shapingboth pedagogy and public discussion It no longer helps to signal the analystrsquosadherence to specic deeper assumptions implicated in any empirical expla-nation of concrete events in world politics

If such complete confusion is possible some might be tempted to rejectrealismmdashand perhaps with it all ldquoismsrdquo in international relations theorymdashasinherently vague indeterminate contradictory or just plain wrong3 This is anunderstandable response but it is at the very least premature Althoughbattles among abstract ldquoismsrdquo can often be arid the specication of well-developed paradigms around sets of core assumptions remains central to thestudy of world politics By unambiguously linking specic claims to commoncore assumptions paradigms assist us in developing coherent explanationsstructuring social scientic debates considering a full range of explanatoryoptions dening the scope of particular claims understanding how differenttheories and hypotheses relate to one another and clarifying the implicationsof specic ndings While realism is not the only basic international relationstheory in need of clarication its long history and central position in the eldmake it an especially important focus for theory research pedagogy andpolicy analysis No other paradigm so succinctly captures the essence of anenduring mode of interstate interaction based on the manipulation of materialpowermdashone with a venerable history4 And it need not be incoherent Accord-ingly we shall propose not a rejection but a reformulation of realism in threeassumptionsmdasha reformulation that highlights the distinctive focus of realismon conict and material power

This article proceeds in three sections We begin by elaborating the desirablequalities of a theoretical paradigm in international relations and guided bythese criteria propose a formulation of realism that we believe captures itsenduring essence We then document the theoretical degeneration of recentldquominimal realistrdquo theory We conclude by highlighting the practical advantages

3 Vasquez ldquoThe Realist Paradigmrdquo and David A Lake and Robert Powell eds Strategic Choiceand International Relations (Princeton NJ Princeton University Press 1999)4 Jared Diamond Guns Germs and Steel The Fates of Human Societies (New York WW Norton1997)


for theoretical debate and empirical research of consistently adhering to anarrower and more rigorous reformulation of the realist paradigm

Realism as a Theoretical Paradigm

Realism many have observed is not a single theory but a family of theoriesmdashaldquoparadigmrdquo5 Nearly all scholars who have voiced an opinion on the subjectover the past quarter century agree that what makes it possible and useful tospeak about realism as a unied paradigm is the existence of a series of sharedcore assumptions In this section we rst discuss desirable attributes of a setof core assumptions then offer an appropriate reformulation of realism

Whether a paradigm is conceptually productive depends on at least tworelated criteria coherence and distinctiveness6 First and least controversial aparadigm must be logically coherent It must not contain internal logicalcontradictions that permit the unambiguous derivation of contradictory con-clusions To be sure given their breadth paradigms are likely to be incompleteThe use of differing auxiliary assumptions may thus generate multiple evencontradictory propositions But there must be a constraint on such deriva-tions7 When theoretical explanation of empirical ndings within a paradigmconsistently relies on auxiliary assumptions unconnected to core assumptionsto predict novel facts or clear up anomalies we learn little about the veracityof those assumptions When it relies on auxiliary assumptions contradictory tounderlying core assumptions our condence in those core assumptions shouldweaken8

5 Or a ldquobasic theoryrdquo ldquoresearch programrdquo ldquoschoolrdquo or ldquoapproachrdquo For similar usage see StephenVan Evera cited in Benjamin Frankel ldquoRestating the Realist Caserdquo in Frankel Realism p xiii andWalt ldquoThe Progressive Power of Realismrdquo We do not mean to imply more with the term ldquopara-digmrdquo than we state6 For a fuller account of the desirable criteria see Jeffrey Legro and Andrew Moravcsik ldquoIsAnybody Still a Realistrdquo Weatherhead Center for International Affairs Working Paper Series(Cambridge Mass Harvard University 1998) There we also employ these standards to rejectparadigmatic denitions of realism based on ideal-typical outcomes (eg ldquopessimismrdquo orldquoconictrdquo) vague concepts (eg ldquopower and interestrdquo) intellectual history or outcomes predictedby more than one theory (eg ldquobalancingrdquo)7 Our central criticism of recent realism is not that the realist paradigm is incoherent or indistinctsimply because it generates various even conicting theories and hypotheses We do not believethat disagreement among realists per se is a sign of degeneration See Walt ldquoThe Progressive Powerof Realismrdquo pp 932ndash9338 See Imre Lakatos ldquoFalsication and the Methodology of Scientic Research Programsrdquo inLakatos and Alan Musgrave eds Criticism and the Growth of Knowledge (Cambridge CambridgeUniversity Press 1970) pp 131ndash132


Second and more important for our purposes here a paradigm must bedistinct Its assumptions must clearly differentiate it from recognized theoreti-cal alternatives Paradigmatic formulations must make sense not only on theirown terms but also within the context of broader social scientic debates9

Only in this way can we speak meaningfully of testing theories and hypothesesdrawn from different paradigms against one another or about the empiricalprogress or degeneration of a paradigm over time The appropriate level ofgenerality number of assumptions and empirical scope of a paradigm are nottherefore qualities intrinsic to any single paradigm but depend on the schol-arly debate in which the paradigm is employed

Realism coexists in a theoretical world with at least three paradigmaticalternatives for which core assumptions can been elaborated The rst theinstitutionalist paradigm contains theories and explanations that stress the roleof international institutions norms and information Examples include thetransaction costndashbased analyses of functional regime theorists and perhapsthe sociological institutionalism espoused by some constructivists10

The second alternative the liberal paradigm contains theories and explana-tions that stress the role of exogenous variation in underlying state preferencesembedded in domestic and transnational state-society relations Paradigmaticliberal assumptions underlie most of what are referred to as ldquosecond-imagerdquo(and many ldquosecond-image reversedrdquo) theories Examples include claims aboutthe autonomous impact of economic interdependence domestic representativeinstitutions and social compromises concerning the proper provision of publicgoods such as ethnic identity regulatory protection socioeconomic redistribu-tion and political regime type11

9 Fundamental debates are always (at least) ldquothree-corneredrdquo pitting two (or more) theoriesagainst the data See ibid p 11510 For a statement of core assumptions see Robert O Keohane International Institutions and StatePower (Boulder Colo Westview 1989) Stephen D Krasner ed International Regimes (Ithaca NYCornell University Press 1983) and Keohane After Hegemony Cooperation and Discord in the WorldPolitical Economy (Princeton NJ Princeton University Press 1984)11 For a statement of core assumptions see Andrew Moravcsik ldquoTaking Preferences Seriously ALiberal Theory of International Politicsrdquo International Organization Vol 51 No 4 (Autumn 1997)pp 513ndash553 Helen V Milner ldquoRationalizing Politics The Emerging Synthesis of InternationalAmerican and Comparative Politicsrdquo International Organization Vol 52 No 4 (Autumn 1998)pp 759ndash786 Robert O Keohane and Joseph S Nye Power and Interdependence World Politics inTransition (Boston Little Brown 1977) Michael W Doyle ldquoLiberalism and World Politicsrdquo Ameri-can Political Science Review Vol 80 No 4 (December 1986) pp 1151ndash1169 Richard Cooper TheEconomics of Interdependence Economic Policy in the Atlantic Community (New York ColumbiaUniversity Press 1968) and Elman and Elman ldquoCorrespondencerdquo p 924 all concur that suchtheories are nonrealist


The third less well-articulated alternative the epistemic paradigm containstheories and explanations about the role of collective beliefs and ideas onwhich states rely in calculating how to realize their underlying goals12 Incontrast to liberal theories (which stress the way the ideas shared or manipu-lated by groups inuence state preferences and policy) and institutionalisttheories (which stress the role of formal norms and institutions in providinginformation to states) the epistemic paradigm stresses exogenous variation inthe shared beliefs that structure means-ends calculations and affect perceptionsof the strategic environment13 Examples include many arguments about cul-ture (strategic organizational economic and industrial) policy paradigms inparticular issue areas group misperception standard operating proceduresand some types of social learning14

A paradigm is only as powerful and useful as its ability to rule out plausiblecompeting assumptions and explanations about the world Enduring interna-tional relations paradigms have helped to focus our attention on particular coreassumptions and causal mechanisms Debates among realists liberalsepistemic theorists and institutionalists have traditionally centered around thescope power and interrelationship of variation in material capabilities (real-ism) national preferences (liberalism) beliefs (epistemic theory) and interna-tional institutions (institutionalism) on state behavior A formulation of realism

12 An ldquoepistemerdquo or ldquosystem of understandingrdquo implies a collective mentality and should bedistinguished from purely psychological approaches about individual perceptions and personalitytraits although these may share similarities Our use of the word seeks to situate the paradigmbetween deep constitutive connotations of ldquosocial epistemerdquo in John G Ruggie ldquoTerritoriality andBeyond Problematizing Modernity in International Relationsrdquo International Organization Vol 47No 4 (Winter 1993) p 157 and interest-group focus of ldquoepistemic communityrdquo in Peter M HaasldquoIntroduction Epistemic Communities and International Policy Coordinationrdquo International Or-ganization Vol 46 No 1 (Winter 1992) pp 1ndash3513 On the role of beliefs in rationalist theory see Jon Elster ldquoIntroductionrdquo in Elster ed RationalChoice (New York New York University Press 1986) pp 1ndash33 and Arthur Denzau and DouglassNorth ldquoShared Mental Models Ideologies and Institutionsrdquo Kyklos Vol 47 No 1 (Spring 1994)pp 3ndash3114 John Odell US International Monetary Policy Markets Power and Ideas as Sources of Change(Princeton NJ Princeton University Press 1982) Paul Egon Rohrlich ldquoEconomic Culture andForeign Policyrdquo International Organization Vol 41 No 1 (Winter 1987) pp 61ndash92 Kathryn SikkinkIdeas and Institutions Developmentalism in Brazil and Argentina (Ithaca NY Cornell University Press1991) Peter Hall ldquoPolicy Paradigms Social Learning and the Staterdquo Comparative Politics Vol 25No 3 (April 1993) pp 275ndash295 Judith Goldstein Ideas Interests and American Trade Policy (IthacaNY Cornell University Press 1993) Alastair Iain Johnston Cultural Realism Strategic Culture andGrand Strategy in Chinese History (Princeton NJ Princeton University Press 1995) Jeffrey WLegro Cooperation under Fire Anglo-German Restraint during World War II (Ithaca NY CornellUniversity Press 1995) Peter J Katzenstein ed The Culture of National Security Norms Identityand World Politics (New York Columbia University Press 1996) Dan Reiter Crucible of Beliefs(Ithaca NY Cornell University Press 1996) and Elizabeth Kier Imagining War French and BritishMilitary Doctrine between the Wars (Princeton NJ Princeton University Press 1997)


that subsumed all the core assumptions underlying these other theories wouldbe a misleading guide to theoretical debate or empirical research Perpetuallyunderspecied perhaps internally contradictory such a formulation wouldevade rather than encourage potentially falsifying theoretical counterclaimsthereby defeating the basic purpose of grouping theories under paradigms inthe rst place Surely realism with its enduring commitment to the statesman-like manipulation of conict and power is more than just a generic form ofrationalism Realism must therefore remain distinct from its liberal epistemicand institutionalist counterparts

realism as a paradigm three core assumptionsMany among the most prominent contemporary forms of realism lack bothcoherence and distinctiveness To see precisely why and how this is so how-ever we must rst demonstrate that a coherent distinct formulation of the coreassumptions underlying the realist paradigm is possible practical and pro-ductive Three ldquocorerdquo assumptions are necessary and sufcient for this pur-pose Our formulation comprises the essential elements of a social scientictheory namely assumptions about actors agency and structural constraint15

Though few if any formulations in the realist literature are identical to this onemany overlap16

assumption 1mdashthe nature of the actors rational unitary politicalunits in anarchy The rst and least controversial assumption of realismconcerns the nature of basic social actors Realism assumes the existence of aset of ldquoconict groupsrdquo each organized as a unitary political actor that ration-ally pursues distinctive goals within an anarchic setting Within each territorialjurisdiction each actor is a sovereign entity able to undertake unitary actionBetween jurisdictions anarchy (no sovereign power) persists Realists assumemoreover that these sovereign conict groups are rational in the conventionalsense that they select a strategy by choosing the most efcient available meansto achieve their ends subject to constraints imposed by environmental uncer-tainty and incomplete information17

15 James S Coleman Foundations of Social Theory (Cambridge Mass Harvard University Press1990)16 Randall L Schweller and David Priess suggest this denition although they neglect it in theirmidrange theorizing Schweller and Priess ldquoA Tale of Two Realisms Expanding the InstitutionsDebaterdquo International Studies Review Vol 41 No 1 (1997) pp 1ndash32 Walt comes close in Walt ldquoTheProgressive Power of Realismrdquo p 932 For an all-inclusive denition including many of theseelements see Frankel ldquoRestating the Realist Caserdquo17 Kenneth N Waltz Theory of International Politics (Reading Mass Addison-Wesley 1979) p 94Stephen D Krasner Structural Conict The Third World against Global Liberalism (Berkeley Univer-


What is essential to the logic of realist theory is not the particular scope ofthe actors but the ability to draw a sharp distinction between anarchy amongactors and hierarchy within them As Kenneth Waltz Robert Gilpin and manyothers have noted under other historical circumstances one might replacestates with tribes domains principalities city-states regional political unionsor whatever other conict group enjoys a monopoly of legitimate force withinterritorial jurisdictions In modern international relations the state is generallyaccepted as the dominant form of political order able to pursue a unitaryforeign policy18

assumption 2mdashthe nature of state preferences fixed and uniformlyconflictual goals The second realist assumption is that state preferencesare xed and uniformly conictual19 Interstate politics is thus a perpetualinterstate bargaining game over the distribution and redistribution of scarceresources Much of the power of realist theory leading realists like CarrMorgenthau and Waltz consistently maintained comes from the assumptionthat state preferences are xed It is this assumption they argue that releasesus from the ldquoreductionistrdquo temptation to seek the causes of state behavior inthe messy process of domestic preference formation from the ldquomoralistrdquo temp-tation to expect that ideas inuence the material structure of world politicsfrom the ldquoutopianrdquo temptation to believe that any given group of states havenaturally harmonious interests and from the ldquolegalistrdquo temptation to believethat states can overcome power politics by submitting disputes to commonrules and institutions20

sity of California Press 1985) p 28 Hans J Morgenthau Politics among Nations The Struggle forPower and Peace (New York Alfred A Knopf 1973) pp 7ndash8 Robert Gilpin ldquoNo One Loves aPolitical Realistrdquo in Frankel Realism p 7 and Robert O Keohane ldquoRealism Neorealism and theStudy of World Politicsrdquo in Keohane ed Neorealism and Its Critics (New York Columbia UniversityPress 1986) pp 1ndash26 This rationality can be bounded the precise level of calculating ability isinessential to our purposes here as long as miscalculations are random if they are not then othertheories may take over18 Gilpin ldquoNo One Loves a Political Realistrdquo and Kenneth N Waltz ldquoRealist Thought andNeorealist Theoryrdquo in Robert L Rothstein ed The Evolution of Theory in International Relations(Columbia University of South Carolina Press 1991) p 3719 Preferences should remain clearly distinct from strategies State preferences are dened overstates of the social world and are therefore ldquoprestrategicrdquo that is they remain uninuenced byshifts in the strategic environment such as the distribution of power Preferences are akin toldquotastesrdquo that states bring to the international bargaining table although they themselves may ofcourse result from forms of international interaction other than those being studied as do nationalpreferences resulting from economic interdependence See Robert Powell ldquoAnarchy in Interna-tional Relations Theory The Neorealist-Neoliberal Debaterdquo International Organization Vol 48 No2 (Spring 1994) pp 313ndash344 and Moravcsik ldquoTaking Preferences Seriouslyrdquo20 Morgenthau Politics among Nations pp 2ndash12 Waltz Theory of International Politics pp 18ndash37and Waltz ldquoRealist Thought and Neorealist Theoryldquo pp 21ndash37


Despite their general agreement on the assumption of xed preferencesrealists display far less agreement about the precise nature of such preferencesMost assume only that in Waltzrsquos oft-cited phrase states ldquoat a minimum seektheir own preservation and at a maximum drive for universal dominationrdquomdashan elastic assumption much criticized for its vagueness Such an impreciseassumption negates the explanatory value of assuming xed preferences21

From game theorists like Robert Powell to constructivists like AlexanderWendt there is broad agreement that this does not constitute a sharp enoughassumption about the nature of the statemdashthat is of its state-society relationsand resulting state preferencesmdashon which to build explanatory theory In aworld of status quo states and positive-sum interactions for example tradi-tional realist behaviors may well not emerge at all Lest we permit the entirerange of liberal epistemic and institutional sources of varying state prefer-ences to enter into realist calculations a narrower assumption is required22

We submit that a distinctive realist theory is therefore possible only if weassume the existence of high conict among underlying state preferencesmdashwhat John Mearsheimer labels a ldquofundamentally competitiverdquo world andJoseph Grieco sees as one dominated by relative gains seeking (a high valueof k)23 Only then does a rational government have a consistent incentive toemploy costly means to compel others to heed its will Only then thereforeshould we expect to observe recurrent power balancing the overriding im-perative to exploit relative power and (in extreme cases) concern about sur-vival and security as well as other realist pathologies24 In short realists view

21 Waltz Theory of International Politics p 11822 Powell ldquoAnarchy in International Relations Theoryrdquo p 315 Alexander Wendt ldquoSocial Theoryof International Politicsrdquo unpublished manuscript Dartmouth College 1998 p 309 Randall LSchweller ldquoNeorealismrsquos Status-Quo Bias What Security Dilemmardquo in Frankel Realism Moravc-sik ldquoTaking Preferences Seriouslyrdquo Jeffrey W Legro ldquoCulture and Preferences in the InternationalCooperation Two-steprdquo American Political Science Review Vol 90 No 1 (March 1996) pp 118ndash137Barry Buzan Charles Jones and Richard Little The Logic of Anarchy Neorealism to Structural Realism(New York Columbia University Press 1993) and John Gerard Ruggie ldquoInternational RegimesTransactions and Change Embedded Liberalism in the Postwar Economic Orderrdquo in KrasnerInternational Regimes23 John J Mearsheimer ldquoBack to the Future Instability in Europe after the Cold Warrdquo InternationalSecurity Vol 15 No 1 (Summer 1990) pp 5ndash56 and Joseph M Grieco ldquoAnarchy and the Limitsof Cooperation A Realist Critique of the Newest Liberal Institutionalismrdquo International Organiza-tion Vol 42 No 3 (Summer 1988) pp 485ndash507 Grieco maintains that states seek both absoluteand relative gains The relative importance of relative gains is given by the coefcient k The higherthe value of k Grieco maintains the stronger the incentives for relative-gains seeking and the morepronounced the tendency to engage in ldquodefensive positionalistrdquo realist behavior For a moredetailed analysis see pp 25ndash27 below24 Schweller puts this well ldquoIf states are assumed to seek nothing more than their own survivalwhy would they feel threatened Anarchy and self-preservation alone are not sufcient Predatory states motivated by expansion and absolute gains not security and the fear of relative


the world as one of constant competition for control over scarce goods Thisexplicit assumption of xed and uniformly conictual preferences is the mostgeneral assumption consistent with the core of traditional realist theory Gov-ernments may conict over any scarce and valuable good including agricul-tural land trading rights and allied tribute as in the time of Thucydidesimperial dominion as observed by historians from Ancient Rome through theRenaissance religious identity dynastic prerogatives and mercantilist controlas in early modern Europe national and political ideology as in most of thelate nineteenth and twentieth centuries or purely economic interests for asWaltz himself observes ldquoeconomic and technological competition is often askeen as military competitionrdquo25

Note that in addition to its generality this assumption is more permissivethan it might appear at rst glance for three reasons First it does not denythat in world politics zero-sum conict nearly always coexists with positive-sum conicts (or tractable collective action problems) This is in fact impliedby our proposed realist assumption that in world politics states face bargainingproblems because conventional bargaining theory commonly disaggregatesnegotiations into distributional and integrative elements26 The assumptioninsists only that the explanatory power of realism is limited largely to thedistributive aspect of such mixed-motive interstate bargaining Explainingintegrative aspects requires a nonrealist theory

losses are the prime movers of neo-realist theory Without some possibility for their existence thesecurity dilemma melts away as do most concepts associated with contemporary realismrdquo Schwel-ler ldquoNeorealismrsquos Status-Quo Biasrdquo pp 91 119 Somewhat perversely for a realist he citesFukuyama The End of History pp 254ndash255 See also Moravcsik ldquoTaking Preferences SeriouslyrdquoCharles L Glaser ldquoRealists as Optimists Cooperation as Self-Helprdquo in Brown Lynn-Jones andMiller The Perils of Anarchy and Andrew Kydd ldquoSheep in Sheeprsquos Clothing Why Security SeekersDo Not Fight Each Otherrdquo Security Studies Vol 7 No 1 (Autumn 1997) pp 153ndash15425 Kenneth N Waltz rdquoThe Emerging Structure of International Politicsldquo International SecurityVol 18 No 2 (Fall 1993) p 57 Waltz Theory of International Politics Robert Gilpin War and Changein World Politics (Cambridge Cambridge University Press 1981) and Michael Mastanduno ldquoDoRelative Gains Matter Americarsquos Response to Japanese Industrial Policyrdquo in David A Baldwined Neorealism and Neoliberalism The Contemporary Debate (New York Columbia University Press1993)26 Disaggregating the interactions between two may be empirically and theoretically challengingbut the conceptual distinction between the two dimensions of preferences remains unavoidableHoward Raiffa The Art and Science of Negotiation (Cambridge Mass Harvard University Press1982) James D Morrow ldquoSocial Choice and System Structure in World Politicsrdquo World PoliticsVol 41 No 1 (October 1988) pp 75ndash97 and Stephen D Krasner ldquoGlobal Communications andNational Power Life on the Pareto Frontierrdquo in Baldwin Neorealism and Neoliberalism pp 234ndash249These theorists do not of course concede to a theory based on material resources the sole abilityto explain the outcome of conict-prevailing beliefs asymmetrical interdependence or preferenceintensity institutional context and various process-level theories may also play a role


Second this assumption does not exclude most variants of so-called defen-sive realismmdashin which states are assumed to have a preference for securityThis is because the assumption of xed uniformly conictual preferences neednot mean that every set of state preferences actually are conictual It is consis-tent also with the view thatmdashas even Mearsheimer and others commonlythought of as ldquooffensive realistsrdquo contendmdashstate preferences are on averageconictual In the latter case governments must make worst-case assumptionsacting ldquoas ifrdquo preferences were xed uniform and conictual if high uncer-tainty prevents governments from distinguishing true threats27 Either way wemay assume for the purposes of analysis that preferences are conictual

Third we assume only that underlying preferences are xed and conictualnot that the resulting state policies and strategies or systemic outcomes (thedependent variables of any theory of world politics) are necessarily conictualObserved political conict may be deterred or dissuaded by domination brib-ery threats or balancing For most realists the fundamental problem of state-craft is to manage conict in a world where state interests are fundamentallyopposed Indeed even if underlying preference functions generate zero-sumconicts among substantive ends (or are randomly distributed behind a veilof uncertainty) it might reasonably be assumed that all states have a xeduniform preference to minimize the political costs of bargaining itselfmdashtheblood and treasure squandered in warfare sanctions and other forms ofcoercion Under such circumstances we maintain states have a strong incen-tive to bargain efciently and to avoid futile endeavors This is the basis of theconsistent realist concern from Thucydides to Morgenthau for moderation instatecraft

assumption 3mdashinternational structure the primacy of material ca-pabilities The rst two assumptionsmdashnamely that states (or other hierarchi-cal conict groups) are unitary rational actors in international politics and thatthey hold conicting preferencesmdashimply that realism is concerned primarilywith the determinants of distributive bargaining among states These assump-tions however remain insufcient to distinguish realist theory for two relatedreasons First they characterize only agents but not the structure of theirinteraction We still know nothing even in principle about how the outcomesof interstate bargaining in anarchy are determined Second the two assump-

27 John J Mearsheimer ldquoThe False Promise of International Institutionsrdquo in Brown Lynn-Jonesand Miller The Perils of Anarchy p 337 Eric Labs ldquoBeyond Victory Offensive Realism and theExpansion of War Aimsrdquo Security Studies Vol 6 No 4 (Summer 1997) pp 1ndash49 and Robert GilpinldquoThe Richness of the Tradition of Political Realismrdquo in Keohane Neorealism and Its Critics


tions describe a world of constant background conditions What permits us toexplain variation in world politics

We thus require a third and pivotal assumption namely that interstatebargaining outcomes reect the relative cost of threats and inducements whichis directly proportional to the distribution of material resources In contrast totheories that emphasize the role of issue-specic coordination persuasiveappeals to shared cultural norms or identities relative preference intensityinternational institutions or collective norms in shaping bargaining outcomesrealism stresses the ability of states absent a common international sovereignto coerce or bribe their counterparts This is consistent with the assumptionsoutlined above If underlying state preferences are assumed to be zero-sumthere is generally no opportunity (absent a third party at whose expense bothbenet) for mutually protable compromise or contracting to a common insti-tution in order to realize positive-sum gains Nor can states engage in mutuallybenecial political exchange through issue linkage The primary means ofredistributing resources therefore is to threaten punishment or offer a sidepayment It follows that the less costly threats or inducements are to the senderand the more costly or valuable they are to the target the more credible andeffective they will be Each state employs such means up to the point wheremaking threats and promises are less costly to them than the (uniform) benetsthereby gained28

The ability of a state to do this successfullymdashits inuencemdashis proportionalto its underlying power which is dened in terms of its access to exogenouslyvarying material resources For realists such variation does not reduce tovariation in preferences beliefs or institutional position States faced with asimilar strategic situation will extract a similar proportion of domestic re-sources With xed uniform preferences a large state will thus expend moreresources and is therefore more likely to prevail The obvious example ismilitary force but there is no reason to exclude from the realist domain theuse of commercial or nancial sanctions boycotts and inducements to achieveeconomic endsmdashcommonly termed ldquomercantilismrdquomdashregardless of whether theoutcome is connected with security or the means are military Realists needonly assume that efcacy is proportional to total material capabilities It followsthat the strong do what they can and the weak suffer what they must

28 Coleman argues that coercionmdashldquowhere the superordinate agrees to withhold an action thatwould make the subordinate worse off in exchange for the subordinatersquos obeying the superordi-naterdquomdashis a ldquosomewhat specialrdquo case of exchange Coleman Foundations of Social Theory p 29 andKenneth A Oye Economic Discrimination and Political Exchange World Political Economy in the 1930sand 1980s (Princeton NJ Princeton University Press 1992)


Realists have long insisted that control over material resources in worldpolitics lies at the core of realism When Morgenthau Waltz and Gilpinproclaim that the central premise of realism is the ldquoautonomy of the politicalrdquothey mean that by treating material capability as an objective universal andunalienable political instrument independent of national preferences institu-tions and perceptions realists isolate the essence of world politics This simplenotion gives force to Morgenthaursquos and Waltzrsquos consistent dismissal of idealsdomestic institutions economic interests psychology and other sources ofvaried state preferencesmdasha position inherited (almost verbatim) from NiccolograveMachiavelli Friedrich Meinecke and Max Weber29 For all these realists ma-terial resources constitute a fundamental ldquorealityrdquo that exercises an exogenousinuence on state behavior no matter what states seek believe or construct30

This is the wellspring of the label ldquorealismrdquo Realism we maintain is only asparsimonious and distinctive as its willingness to adhere rmly to this assump-tion This assertion above all else distinguishes realism from liberal epistemicand institutionalist explanations which predict that domestic extraction ofresources and interstate interaction will vary not with control over materialresources but with state preferences beliefs and information

The Degeneration of Contemporary Realist Theory

So far we have argued that a distinct realist paradigm must rest on three coreassumptions The power of these premises can be seen in contemporary realisttheories that adhere rmly to them Despite his curious reluctance to makeexplicit assumptions of conictual preferences and rationality Kenneth Waltzrsquosinuential neorealist theory which stresses the polarity of the internationalsystem is broadly consistent with these premises John Mearsheimer rsquos gloomypredictions about the future of Europe derived from consideration of theconsequences of shifts in polarity on national military policy are as well31

Joanne Gowa adheres to core realist assumptions in her provocative argumentthat both the democratic peace and postndashWorld War II international liberaliza-tion were designed in large part to generate ldquosecurity externalitiesrdquo within abipolar structure of power32 Stephen Krasner Robert Gilpin and David Lake

29 The language in Morgenthau Politics among Nations p 5 is echoed almost verbatim in WaltzTheory of International Politics On Weber see Michael Joseph Smith Realist Thought from Weber toKissinger (Baton Rouge Louisiana State University Press 1986)30 Frankel ldquoRestating the Realist Caserdquo pp xiindashxiv31 Mearsheimer ldquoBack to the Futurerdquo32 Joanne Gowa Allies Adversaries and International Trade (Princeton NJ Princeton UniversityPress 1994)


have argued that the level of overall openness in the world economy is afunction of the concentration of control over economic capabilities33 RobertKeohane while in other senses not a realist applies a similar logic to the roleof hegemons in international economic institutions34 Gilpin and Paul Kennedyaddress the historical succession of security orders35 On a recognizably realistbasis Dale Copeland explains major war and Christopher Layne criticizes thedemocratic peace thesis36 Robert Powellrsquos game-theoretical reformulation ofrealism in terms of increasing returns to material capabilities like closelyrelated theories of offense and defense dominance ts within the three coreassumptions as does Barry Posenrsquos analysis of variation in military doctrine37

Among those who claim to be realists today however adherence to thesecore realist premises is the exception rather than the rule Most recent realistscholarshipmdashnotably that of ldquodefensiverdquo and ldquoneoclassicalrdquo realistsmdashatlyviolates the second and third premises To illustrate this tendency we rst turnbriey to recent developments in abstract realist theory focusing particularlyon explicit denitions of realism then trace three trends in recent empiricaltheory and research that highlight the slide of realism into liberal epistemicand institutionalist theory respectively

minimal realism in theoryMost recent formulations of the realist paradigm are inconsistent with ourtripartite formulation Most important among these for our purposes here iswhat we term ldquominimal realismrdquo Minimal realists seek to dene a distinct andcoherent realist paradigm with reference to a set of assumptions less restrictivethan the three we outline above

33 Stephen D Krasner ldquoState Power and the Structure of Foreign Traderdquo World Politics Vol 28No 3 (April 1976) pp 317ndash347 Gilpin War and Change in World Politics David A Lake PowerProtection and Free Trade International Sources of US Commercial Policy (Ithaca NY CornellUniversity Press 1988) and Lake ldquoLeadership Hegemony and the International Economy NakedEmperor or Tattered Monarch with Potentialrdquo International Studies Quarterly Vol 37 No 4(December 1993) pp 459ndash48934 Keohane After Hegemony35 Gilpin War and Change in World Politics and Paul Kennedy The Rise and Fall of the Great Powers(New York Random House 1987)36 Dale Copeland Anticipating Power Dynamic Realism and the Origins of Major War (Ithaca NYCornell University Press forthcoming) and Christopher Layne ldquoKant or Cant The Myth of theDemocratic Peacerdquo in Brown Lynn-Jones and Miller The Perils of Anarchy pp 287ndash33137 Robert Powell ldquoAbsolute and Relative Gains in International Relations Theoryrdquo AmericanPolitical Science Review Vol 85 No 4 (December 1991) pp 701ndash726 and Barry R Posen The Sourcesof Military Doctrine France Britain and Germany between the Wars (Ithaca NY Cornell UniversityPress 1984) pp 69 229


The most extreme among minimal realists maintain that realismrsquos distinct-iveness vis-agrave-vis other international relations paradigms lies solely in our rstassumptionmdashthe existence of rational actors in an anarchic setting JosephGrieco for example maintains that realists need only assume rationality andanarchymdashin other words the pursuit of rational ldquoself-helprdquo strategiesmdashtoderive a concern about security and autonomy a measure of underlying stra-tegic conict strategies of relative-gains seeking and balancing of materialpower and other elements of realist theory38 Outside of a small group of suchrealists however a variety of scholars agree that the assumption of hierarchicalactors interacting rationally in an anarchic world is insufcient to distinguishrealism As we discuss below this assumption is shared by almost all otherschools39 Because anarchy and rationality are constant moreover assumingthem tells us little about the distinctive realist variables and causal mechanismsfor explaining variation in state behavior

Other recent denitions of a realist paradigm therefore include additionalassumptions which seek to serve the same functions of social theory as oursecond and third assumptions namely to specify agency and structure andthe interaction between them Two assumptions are particularly common Firststates seek to realize a xed set of underlying preferences ranging from de-fending their territorial integrity and political independence to expanding theirinuence over their international environment (often referred to somewhatmisleadingly as ldquosecurityrdquo and ldquopowerrdquo respectively) Second among thepolitical means states employ to resolve the resulting conicts force and thethreat of force are preeminent Nearly all the authors considered in this articlebase their discussion of realism on such a denition even when some fail tomake this explicit40

38 Joseph M Grieco ldquoRealist International Theory and the Study of World Politicsrdquo in MichaelW Doyle and G John Ikenberry eds New Thinking in International Relations Theory (Boulder ColoWestview 1997) pp 166ndash168 is most explicit39 The transmethodological consensus on this point is near universal In addition to WendtPowell Moravcsik Legro and Schweller cited above in n 22 see Helen V Milner ldquoThe Assump-tion of Anarchy in International Relations Theoryrdquo Review of International Studies Vol 17 No 1(January 1991) pp 67ndash8540 This is true also of some more unwieldy denitions Elman and Elman ldquoLakatos and Neore-alismrdquo p 923 dene the realist hard core as rational strategic states in anarchy seeking survivalwith limited resources Ashley Tellis ldquoReconstructing Political Realism The Long March to Scien-tic Theoryrdquo Security Studies Vol 5 No 2 (Winter 1995ndash1996) p 3 describes ldquopolitical actionsaimed at enhancing securityrdquo as the ldquominimum realist programrdquo Sean M Lynn-Jones and StevenE Miller ldquoPrefacerdquo in Brown Lynn-Jones and Miller The Perils of Anarchy pp ixndashx focus onrationality anarchy and power but make no assumption that underlying goals conict and limittheir denition to the use of military force We see a similar move in Buzan Jones and Little TheLogic of Anarchy which seeks to integrate interdependence preferences information and institu-tions into a ldquorealistrdquo theory tied together only by the fact that it is systemic


Yet even this more elaborate form of minimal realism fails to distinguishrealism from its alternative paradigms because nearly all variants of liberalepistemic and institutionalist theories share the same three assumptions41

Consider for example functional regime theory democratic peace theorytheories of ldquoaggressorrdquo states ldquoendogenousrdquo theories of international tradepolicy and strategic culture theory Surely none is realist yet each concurs thatin an anarchic world system no superordinate institution can establish amonopoly of legitimate force rational unitary states are the major actors42

(Although it is true that liberals and epistemic theorists focus on contestationamong subnational actors in the process of preference or belief formation theygenerally hold that they act rationally thereafter) Nearly all agree moreoverthat states are self-interested and their preferences at least in security matterslie somewhere between security and power Indeed nearly all go much furtherassuming that a perfect underlying harmony of interest is so rare as to bealmost irrelevant a measure of conict over underlying values and interestsall modern theories agree is endemic to world politics Nearly all concurfurthermore that governments generally place a high perhaps superordinatevalue on national security territorial integrity and political independenceThey also agree that a central and often decisive instrument available tostatesmdashthe ultima ratio at least in the abstractmdashis coercive force In sum amongmodern international relations theories the claims that ldquopower and interestsmatterrdquo that states seek to ldquoinuencerdquo one another in pursuit of often conict-ing ldquoself-interestsrdquo and that ldquoself-helprdquo through military force is an importantperhaps the most important instrument of statecraft are trivial

Most clearly missing from minimal realism as compared to the tripartitedenition with which we began are any distinctive assumptions about thesource and resolution of conict Yet its adherents continue to employ realistrhetoric and claim consistency with traditional realist theory This lack of

41 Some sociological theories take the somewhat different view that actors behave according toa noninstrumental ldquologic of appropriatenessrdquo whereby actors conform to internalized rules im-posed by society See Martha Finnemore National Interests in International Society (Ithaca NYCornell University Press 1996) pp 28ndash31 and James March and Johan Olsen Rediscovering Insti-tutions The Organizational Basis of Politics (New York Free Press 1989)42 Stefano Guzzinirsquos assessment goes to the heart of the matter ldquoThe closest we can get to asingle assumption that would demarcate realism is the idea of anarchy [But] traditionaldefenders of collective security [as well as rsquodemocratic peacersquo liberals] have the same starting pointRather than setting Realism apart from other international theories the assumption of anarchy setsInternational Relations apart from other disciplinesrdquo Guzzini Realism in International Relations andInternational Political Economy The Continuing Story of a Death Foretold (New York Routledge 1998)pp viiindashix See also Helen V Milner ldquoThe Assumption of Anarchy in International Relationsrdquo andKeohane ldquoIntroductionrdquo After Hegemony


distinctiveness is not simply a matter of abstract denition It is we argue themost striking common characteristic of contemporary midrange ldquorealistrdquo theo-ries Increasingly realist research invokes factors extraneous even contradic-tory to the three core realist assumptions but consistent with core assumptionsof existing nonrealist paradigms This degeneration takes three distinct formsdepending on whether realists invoke exogenous variation in preferencesbeliefs or international institutions These correspond respectively to realistdegeneration into liberal epistemic and institutionalist theories Below weconsider each in turn

from realism to liberalism power is what states want it to beThe traditional realist viewmdashabout which there was until recently little dis-agreementmdashassumes that state preferences are xed and uniform Morgenthauand Waltz we have seen believed that this assumption accounts for realismrsquospower and parsimony43 Still there has been heated debate among modernrealists over precisely which xed uniform preferences should be ascribed tostates Morgenthau emphasizes power itself as a goal by which he may havemeant a generalized desire to expand44 Waltz speaks of survival as the ulti-mate goal of states but allows that states may seek anything between minimalsurvival and world domination As we have seen this assumption imposesalmost no constraint on state behavior because it subsumes the entire spectrumof possible motivations of states from pure harmony to zero-sum conictundened and untheorized Only outright self-abnegation is excluded45 Thishas given rise to a variety of formulations of the precise specication of statepreferences For our purposes we need note only that throughout there hasbeen agreement in principle that realism must assume xed and uniform pref-erences without which it loses its distinctiveness and power

Yet many intellectual descendents of Morgenthau and Waltz reject even thisThey neither simply disagree about the specic nature of xed assumptions to

43 Morgenthau speaks for nearly all realists in arguing that realism must ldquoguard against twopopular fallacies the concern with motives and the concern with ideological preferences History shows no exact and necessary correlation between the quality of motives and the qualityof foreign policyrdquo Morgenthau Politics among Nations pp 5ndash7 Waltz Theory of International Politicsp 29 see also pp 65ndash66 79 90 108ndash112 196ndash198 271 and Grieco ldquoRealist International Theoryrdquop 16544 Morgenthaursquos use of the term ldquopowerrdquo can be ill-dened and overly expansive See Inis LClaude Power and International Relations (New York Random House 1962) pp 25ndash3745 Waltz Theory of International Politics pp 118 126 Fareed Zakaria speaks for most contemporaryrealists when he terms Waltzrsquos writings on such questions ldquoconfused and contradictoryrdquo ZakariaFrom Wealth to Power The Unusual Origins of Americarsquos World Role (Princeton NJ PrincetonUniversity Press 1998) pp 26ndash28


be assumed nor even challenge the notion that they are conictual They rejectthe underlying notion of xed preferences itself Nearly all argue that statebehavior is inuenced not just by power calculations but by the varying pointson the spectrum between motivations of security and power (expansion) onwhich different states nd themselves Such explanations inevitably importconsideration of exogenous variation in the societal and cultural sources ofstate preferences thereby sacricing both the coherence of realism and appro-priating midrange theories of interstate conict based on liberal assumptionsSuch theories include those that stress the nature of domestic representativeinstitutions (eg the democratic peace) the nature of economic interests (egliberal interdependence theories) and collective values concerning nationalidentity socioeconomic redistribution and political institutions

Our skeptical judgment is hardly new A generation ago Arnold Wolfersdrew the consequences of such ad hoc extensions of realist theory ldquoOneconsequence of distinctions such as these [between hostile and status quostates] is worth mentioning They rob [realist] theory of the determinate andpredictive character that seemed to give the pure power hypothesis its peculiarvalue It can no longer be said of the actual world for example that a powervacuum cannot exist for any length of timerdquo46 This tendency is evident in thework of self-styled realists like Jack Snyder Joseph Grieco Fareed ZakariaRandall Schweller and Stephen Van Evera

jack snyder on imperialism We begin with Jack Snyder rsquos analysis of im-perialism to which we owe the label ldquodefensive realismrdquo Snyder sets out toexplain ldquooverexpansionrdquomdashsituations in which great powers expand beyondthe point where they trigger overwhelming countercoalitions and disastrouscounterpressures Unlike some of the theorists we examine below Snyderprovides a detailed theory to back his claims about the importance of domesticpolitics For Snyder the taproot of overexpansion lies in the misrepresentationof domestic interests such that small rent-seeking groups can prot at theexpense of diffuse constituenciesmdasha general tendency exacerbated by deliber-ate manipulation of ideology and logrolling among ldquocartelizedrdquo interestgroups The extent to which states are prone to such pathologies is a functionSnyder argues of the timing of industrialization47

46 Arnold Wolfers ldquoThe Pole of Power and the Pole of Indifferencerdquo in Wolfers ed Discord andCollaboration Essays in International Politics (Baltimore Md Johns Hopkins University Press 1962)p 42 see also pp 86 156 160 Even in Waltz Theory of International Politics we encounter onlythe assertion rather than the derivation of the primacy of systemic concerns47 Jack Snyder Myths of Empire Domestic Politics and International Ambition (Ithaca NY CornellUniversity Press 1991)


Snyder presents this argument as an improvement of realism by integratingdomestic factors consistent with it ldquoMy arguments stressing the domesticdeterminants of grand strategyrdquo he argues ldquoare fully consistent with thedefensive version of realismrdquomdashan ascription he defends with reference toMorgenthau48 Yet while many treat Snyder rsquos argument as a denitive state-ment of defensive realism his position has been criticized for its heavy relianceon domestic factors As Zakaria observes ldquoWhile neorealism is loosely de-picted as leaving domestic politics out many defensive realists in fact havedisplayed the opposite tendency using domestic politics to do all the work intheir theories In the end we are left not with a novel combination ofsystemic and domestic determinants but with a restatement of the traditionalInnenpolitik caserdquo49 At the very least Snyder rsquos effort to redene realism asincluding assumptions and causal mechanisms not traditionally connectedwith it has led realists into conceptual confusion about whether realism meansanything at all To employ more traditional terminology if as Zakaria assertsrealism subsumes both what Waltz terms ldquostructural factorsrdquo at the systemlevel and classic diplomatic historians the Primat der Aussenpolitik and domes-tic and societal factors that alter state preferences which diplomatic historiansterm the Primat der Innenpolitik what is excluded Are any concrete assump-tions of this theory still distinctly realist

Yet the problem is even more fundamental What is innovative in Snyder rsquosexplanation draws almost exclusively on an existing nonrealist internationalrelations paradigm Snyder rsquos is a classically liberal analysis of the impact onforeign policy of shifting domestic state-society relations in modernizing socie-ties As a matter of intellectual history Snyder rsquos theory is drawn from JohnHobson and as Zakaria notes the left-liberal and social democratic GermanInnenpolitik school As a matter of social science theory its core assumptionsare almost identical to contemporary theories of the democratic peace and ofthe role of domestic institutions in trade policy both of which rest on specicimplications of domestic misrepresentation and rent seeking for foreign policyIn sum there is a disjuncture between label and reality Snyder rsquos midrangetheory does not conrm realist assumptions it demonstrates the power andgenerality of fundamental liberal assumptions beyond the simple case of thedemocratic peace His theoretical language which terms all of this ldquorealistrdquo

48 Ibid p 12 see also pp 19ndash20 6449 Fareed Zakaria ldquoRealism and Domestic Politics A Review Essayrdquo in Brown Lynn-Jones andMiller The Perils of Anarchy quote on p 463 and Zakaria From Wealth to Power pp 32ndash33 181ndash183We do not endorse all of Zakariarsquos criticisms of Snyder


simply obscures if not misstates outright the signicance of his importantempirical result

joseph grieco on relative gains Joseph Griecorsquos proposal to dene real-ism in terms of statesrsquo concerns about relative gains provides another examplethis one from political economy of how the line between power and prefer-ences can become blurred when realism is not rigorously dened Grieco positsthat states are ldquodefensive positionalistsrdquo in search of securitymdasha desire thatmakes them sensitive to relative rather than absolute gains States cooperatelessmdashor more precisely they cooperate under different circumstancesmdashthanthe mere presence of mutual benets might lead us to expect because theymust ldquopay close attention to how cooperation might affect relative capabilitiesin the futurerdquo50 Despite much criticism of this formulation and disagreementabout whether the gains in question are actually ldquorelativerdquo Grieco clearlycaptures an essential quality of realism namely its assumption of underlyingconictmdasha quality we highlight in our statement of core assumptions51

Grieco is aware that states do not always forgo ldquoabsoluterdquo economic benetsfor ldquorelativerdquo geopolitical gains so that any theory must state the antecedentconditions under which relative-gains seeking occurs Given that not all statesin all situations are equally sensitive to gaps in payoffs he argues we shouldemploy a factor (termed k) that measures sensitivity to gaps between payoffs(relative gains) alongside absolute gains We can thus restate Griecorsquos causalclaim as follows When k is high states are more motivated to seek relativegains (or limit losses) This simply displaces the causal question howeverfor we are now impelled to ask What determines the value of k Whatmotivates states to worry about relative gains Is this motivation distinctivelyrealist

In answering these questions Grieco is driven to tinker with the assumptionof xed preferences thus revealing that his relative gainsndashseeking denitionof realism lacks theoretical coherence and distinctiveness How does Griecoseek to establish the ldquorealistrdquo nature of his argument He does so by assumingthat the issue area in question explains variation in k Specically k is alwayshigh in security affairs an assumption endorsed by Mearsheimer and others52

Yet this assumed correlation between security policy and relative-gains seeking(even if it were clearly realist) is unsustainable On the one hand there are

50 Joseph M Grieco ldquoAnarchy and the Limits of Cooperation A Realist Critique of the NewestLiberal Institutionalismrdquo in Baldwin Neorealism and Neoliberalism p 138 and Powell ldquoAbsoluteand Relative Gains in International Relations Theoryrdquo51 For the subsequent debate see Baldwin Neorealism and Neoliberalism52 Mearsheimer ldquoThe False Promise of International Institutionsrdquo p 342


numerous security issuesmdashsay interactions among democracies the construc-tion of security regimes or power politics without increasing returnsmdashaboutwhich it is difcult to conclude that there is any incentive to pursue a relativegainsndashseeking strategy53 Even more striking economic conict alone can giverise to realist and mercantilist dynamics without the involvement of anysecurity interestmdashas scholars such as Stephen Krasner Michael MastandunoJames Fearon and David Lake have demonstrated54 As many critics havenoted neither Griecorsquos analysis of postndashTokyo round trade policy nor his otherwork reveals convincing evidence that ldquorelative gainsrdquo in those areas could beexploited to threaten national security55

Cut loose from the claim that all security conicts necessarily generateintense underlying conict (a high value of k) however the ldquorelative-gainsseekingrdquo account of realism no longer imposes any a priori theoretical con-straint on variation in state preferences (variation in k) The argument becomesinstead When state interests clash for whatever reason conict is more likelyYet because other theoriesmdashrealist liberal epistemic and institutionalistmdashalsopredict that conict may result from opposed interests and offer explanationsof that variation in interests there is nothing distinctly realist about relative-gains seeking per se56 In seeking to specify the determinants of variation in kGrieco himself invokes variation in the nature of individual statesmdashincludingldquoprevious experiencesrdquo ldquoreputation for exploitationrdquo and whether they areldquolong-term ally or adversaryrdquomdashas well as more traditionally realist factorsconnected with relative power57 Indeed nonrealist studies of trade policy nd

53 Bruce M Russett Grasping the Democratic Peace Principles for a PostndashCold War World (PrincetonNJ Princeton University Press 1993) Robert Jervis ldquoSecurity Regimesrdquo International OrganizationVol 36 No 2 (Spring 1982) pp 357ndash378 Jack Snyder ldquoAverting Anarchy in the New EuroperdquoInternational Security Vol 14 No 4 (Spring 1990) pp 5ndash41 and George Downs and Keisuke IidaldquoAssessing the Theoretical Case against Collective Securityrdquo in Downs ed Collective Securitybeyond the Cold War (Ann Arbor University of Michigan Press 1994) pp 17ndash3954 Krasner Structural Conict Mastanduno ldquoDo Relative Gains Matterrdquo James D Fearon ldquoRa-tionalist Explanations for Warrdquo International Organization Vol 49 No 3 (Summer 1995) pp 379ndash414 and Lake ldquoLeadership Hegemony and the International Economyrdquo55 Peter Liberman ldquoTrading with the Enemy Security and Relative Economic Gainsrdquo InternationalSecurity Vol 21 No 1 (Summer 1996) pp 155ndash158 Moravcsik ldquoTaking Preferences Seriouslyrdquopp 538ndash540 and Robert O Keohane ldquoInstitutionalist Theory and the Realist Challenge after theCold Warrdquo in Baldwin Neorealism and Neoliberalism p 280ff56 Because k cannot be observed directly and it is difcult to differentiate security from powerseekingmdashhence the security dilemmamdashit is difcult to know how this theory could be testedabsent a theory of the determinants of k There has been to our knowledge no attempt to measurek independently of state behavior Cf Mearsheimer ldquoThe False Promise of International Institu-tionsrdquo pp 347ndash34857 See Joseph M Grieco ldquoRealist Theory and the Problem of International Cooperationrdquo Journalof Politics Vol 50 No 3 (August 1988) pp 610ndash612 and Grieco Cooperation among Nations EuropeAmerica and Non-Tariff Barriers to Trade (Ithaca NY Cornell University Press 1990) pp 45ndash47


that particularly strong pressure from economic interest groupsmdashthe classicliberal explanation for protectionmdashis concentrated in precisely those areas(government procurement and industrial standardization) in which Griecorsquosstudy of the Tokyo round nds unexplained relative-gains seeking58

Absent a tighter paradigmatic denition of realism and more detailed spe-cication of its causal mechanisms this fundamental indeterminacy and lackof theoretical distinctiveness cannot be surmounted The central problem forGrieco is quite simply that relative-gains concerns conict inefcient bargain-ing and suboptimal cooperation are predicted by all major rationalist (andsome nonrationalist) theories of international relations The key differencesamong paradigms lie not in whether they predict interstate conictmdashall domdashbutin when why and under what circumstances they predict conict Bargainingfailures such as those Grieco observes in the General Agreement on Tariffs andTrade may result from inefcient bargaining under uncertainty as institution-alists and negotiation analysts maintain from particularly conictual societalpreferences as liberals argue or from a lack of shared language or culturalcapital as some epistemic theorists assertmdashas well as concerns about futurepower as realists contend Without a more precise specication of realismGrieco cannot distinguish these empirically or theoretically59

neoclassical realism Whereas Snyder and Grieco stress the preference ofstates for security a new generation of realists recently heralded by GideonRose as ldquoneoclassical realistsrdquo (NCRs) stresses the other pole of Waltzrsquos loosespecication of state preferencesmdashthe natural desire of all states to wieldexternal inuence60 States the NCRs argue do not simply respond defensivelyto threats they exploit power differentials to expand their inuence over theirexternal environmentmdasha view of international politics quite different from thatbased on the simple assumption that states seek security Some of these real-istsmdashnotably Zakaria as we have seenmdashare harsh critics of Snyder and othersfor their purported ad hoc reliance on domestic factors to explain conictamong states assumed only to seek security

58 For a review of this literature James E Alt Jeffry Frieden Michael J Gilligan Dani Rodrikand Ronald Rogowski ldquoThe Political Economy of International Trade Enduring Puzzles and anAgenda for Inquiryrdquo Comparative Political Studies Vol 29 No 6 (December 1996) pp 689ndash71759 Krasner ldquoGlobal Communications and National Powerrdquo and Keohane ldquoInstitutionalist Theoryand the Realist Challengerdquo If such bargaining failure cannot be attributed to concerns about expost cheating Grieco argues it conrms realist claims Yet Grieco concedes the existence of acompeting liberal explanation in a long footnote but then drops the point Grieco ldquoAnarchy andthe Limits of Cooperationrdquo pp 486ndash48860 Gideon Rose ldquoNeoclassical Realism and Theories of Foreign Policyrdquo World Politics Vol 51 No1 (October 1998) pp 144ndash172


Yet ironically neoclassical realism (NCR) suffers from precisely the sameweaknesses as defensive realism namely theoretical indeterminacy and a reli-ance on exogenous variation in state preferences Most NCRs seek to incorpo-rate in one form or another variation between states with underlying statusquo and revisionist preferences The incorporation of variation in underlyingdomestic preferences we argue undermines (if not eliminates) the theoreticaldistinctiveness of NCR as a form of realism by rendering it indistinguishablefrom nonrealist theories about domestic institutions ideas and interests Forrealists however these domestic preference shifts moreover remain ad hoc61

As with defensive realists this inclination toward indeterminacy and in-distinctness is not a purely abstract concern but adversely inuences theempirical work of some of realismrsquos latest and brightest defenders Considerthe work of Zakaria and Schweller

fareed zakaria on nineteenth-century america Fareed Zakaria offersan insightful analysis of the reasons why the US government moved towardexpansion in the late nineteenth century more slowly and less thoroughly thanshifts in relative power predict To explain this neorealist anomaly Zakariarejects the traditional realist assumption of a unitary state in favor of a distinc-tion between domestic state apparatus (state) and society (nation) State powerhe argues depends not just on control over resources but on the ability ofstates to extract those resources from society62 The tendency of states to expandis thus a function of the international and domestic power of the state Bothhe contends were necessary for late-nineteenth-century US expansion Inso-far as states are inuenced by relative power and can muster societal supportfor their policies they exploit opportunities to wield inuence

Zakariarsquos argument is a noteworthy effort to bridge the gap between domes-tic and international politics Yet it rests decisively on treating a statersquos abilityto extract societal resources not simply as an exogenous factor predictablyrelated to geographical control over material resources but also as a functionof particular domestic political circumstances Zakaria compounds the inherentindeterminacy of an unweighted combination of material and domestic politi-cal sources of power by offering no general theory (or even consistent inter-

61 Ibid Rose seeks to make a virtue of this citing Aristotle for the proposition that domesticpolitics is simply too complex a subject about which to generalize This claim must come as asurprise not only to scholars of comparative and US politics but to those who study thedemocratic peace economic interdependence aggressive ideologies and other domestic determi-nants of security policy In any case no more recent support for the assumption is provided 62 For a similar argument see Thomas J Christensen Useful Adversaries Grand Strategy DomesticMobilization and Sino-American Conict 1947ndash1958 (Princeton NJ Princeton University Press1996)


pretation) of shifts in domestic state power Absent a theory of domesticpolitics any argument about why a particular state can extract more or fewerresources from domestic society (even an argument implying irrational statebehavior) becomes consistent with what Zakaria terms ldquostate-centered real-ismrdquo This is reected in the exceptionally wide range of considerations thathe admits affected ldquothe degree to which national power can be converted intostate powerrdquomdashincluding technological ideological institutional partisan cul-tural and racial inuences63

Although Zakaria employs an indeterminate assemblage of causal factorshe draws disproportionately on precisely those liberal factors cited by contem-porary liberal democratic peace or endogenous tariff theoristsmdashas well as earlytwentieth-century ldquoidealistsrdquo (This is particularly ironic given his widely citedcriticism of Snyder for adhering to just this Primat der Innenpolitik) Zakariareturns repeatedly to a core claim of democratic peace theory namely thatlegislative or judicial control over the executive undermines its ability todeploy force aggressively except where expected costs are low64 He frequentlyinvokes mutual recognition among liberal republics economic modernizationpublic unwillingness to increase taxes for overseas adventures or militaryprocurement popular opinion on questions like race and partisan politicsmdashallwell-developed liberal causal mechanisms Surely Morgenthau Carr andGeorge Kennan would be hard pressed to recognize in such a view a renewalof classical realism65

randall schweller on interwar foreign policy Randall Schweller rsquosbook on the security policy of the great powers between the world wars DeadlyImbalances offers another instructive example of how recent realists have cometo rely on ad hoc variation in state preferences in lieu of variation in (evenbroad measures of) capabilities66 Schweller argues that the decisive cause ofchanges in state behavior during the 1930s was a perceived shift in the powerstructure from multipolarity to tripolarity This appears at rst glance to be atraditional realist argument yet the shift in polarity in the 1930s was not in

63 Zakaria From Wealth to Power p 3864 Ibid pp 60ndash67 77ndash78 90ndash12765 Zakaria cites Otto Hintze and Morgenthau in his defense Yet Hintze viewed state structureas a product of international circumstances not the reverse Morgenthau we shall see did notbelieve that his claims about moral restraint were realist Vague though Morgenthaursquos notion ofldquopowerrdquo may be he rmly rejects appeals to public opinion ldquoThe governmentrdquo he wrote ldquomustrealize that it is the leader and not the slave of public opinion [Public opinion is] continuouslycreated and recreated by informed and responsible leadershiprdquo Morgenthau Politics among Na-tions pp 133ndash135 205 chap 966 Randall L Schweller Deadly Imbalances Tripolarity and Hitlerrsquos Strategy of World Conquest (NewYork Columbia University Press 1998)


Schweller rsquos view the result of exogenous shifts in the distribution of materialresources Instead it reected the decisions of particular revisionist nationsnotably Germany to build up their military forces beyond what was toleratedby others or required for security Germanyrsquos leap from a lesser power to majorpower ldquopolerdquo for example occurs suddenly as a result of Adolf Hitlerrsquos riseto power in 1933 and his particular idiosyncratic conception of German na-tional interest67

This shift in emphasis from variation in material capabilities to variation instate preferences is the essence of Schweller rsquos theoretical contribution Schwel-ler criticizes Stephen Walt for his unwillingness to integrate fully variation instate preferences He proposes Waltrsquos evolution from ldquobalance of powerrdquo toldquobalance of threatrdquo be extended one step further to ldquobalance-of-interestrdquo theoryState behavior and international outcomes he argues vary with the distribu-tion of both state power and state preferences but primarily the latter ldquoThemost important determinant of alignment decisionsrdquo Schweller asserts ldquois thecompatibility of political goals not imbalances of power or threatrdquo68 Onesimply cannot assume that states uniformly seek any particular goal Somestates (ldquowolvesrdquo and ldquojackalsrdquo he terms them) have an intrinsic desire forrevision or risky gain while other states (ldquolambsrdquo and ldquolionsrdquo) seek only thestatus quo Schweller rsquos predictions are driven by this distinction Revisionistshe argues ldquobandwagon for protrdquo and thus seek ldquominimum winning coali-tionsrdquo while status quo states seek to balance only against threats and arecomfortable with overwhelming power Many other similar predictions fol-low69

With this analysis Schweller reverses the causal arrow of realism Ratherthan arguing as have realists for centuries that the distribution of powerinuences state behavior despite varying preferences he offers a compellingand creative account of how governments adjust their power to their prefer-ences Coherence and distinctiveness are thereby sacriced The coherence ofrealism is undermined because it is unclear what set of common nontrivialassumptions would permit us to explain state behavior as a function of bothvariation in power and variation in underlying state preferences Schweller

67 Ibid pp 26ndash29 93ndash12068 Ibid p 22 and Randall L Schweller ldquoBandwagoning for Prot Bringing the Revisionist StateBack Inrdquo in Brown Lynn-Jones and Miller The Perils of Anarchy pp 255ndash257 Andrew Kydd termsthis view ldquomotivational realismrdquo Kydd ldquoSheep in Sheeprsquos Clothingrdquo This it should be notedviolates Rosersquos explicit denition of NCR which assumes that systemic factors remain empiricallymore important69 See Schweller Deadly Imbalances pp 84ndash89 and Schweller ldquoNew Realist Research on Alli-ancesrdquo pp 928ndash929


moreover neither consistently invokes nor develops any particular theory ofdomestic preferencesmdashlet alone a distinctively realist one70 To the contraryinsofar as we can read a theory into Schweller rsquos empirical interpretations hetends like Zakaria to invoke concrete causal mechanisms central to liberal andepistemic theories In the few cases in which he speculates on the underlyingsources of state preferences Schweller cites the democratic peace and theaggressiveness of totalitarian dictators71 What is excluded from such an analy-sis And if preexisting liberal international relations theories offer more de-tailed compelling and empirically robust causal mechanisms what is gainedby subsuming a thinner version under a loose conception of ldquorealismrdquo72

To defend the realist label Schweller neither links his argument to a coherentset of realist assumptions nor explicitly distinguishes it from nonrealist theoryInstead he invokes intellectual history73 Morgenthau he maintains employedthe distinction between status quo and revisionist states as well as writingabout the role of domestic politics ideas and institutions Schweller rsquos appealto the intellectual history of classical realism deserves our closer attentionbecause itmdashand the characteristic error in social science theorizing it repre-sentsmdashis also found in the scholarship of Zakaria Snyder and others

Efforts to dene realism by reference to intellectual history in general andclassical realism in particular are deeply awed The coherence of theories isnot dened by their intellectual history but by their underlying assumptionsand causal mechanisms Resort to intellectual history offers a circular deni-tion restating rather than resolving the question of what realism is It is whatrealists believe and realists are those who believe it Moreover intellectualtraditions and even individual statements contain unresolved often contra-dictory tensions Thus scholars have long debated whether the arguments ofrealists from Thucydides to Kennan are in fact coherently realist74 It is note-worthymdashalthough thoroughly unnoted in current debatesmdashthat Carr and Mor-genthau themselves denied that any argument they advanced was ipso factorealist Both were generally careful to distinguish the realist parts of their

70 See Schweller Deadly Imbalances pp 31ndash3871 See for example ibid pp 200ndash20172 This confusing situation can arise because Schweller like other contemporary realists tests histheory against neorealism but ignores nonrealist alternatives73 Ibid p 2074 Those with concerns other than social scientic explanation may legitimately see indetermi-nacy and richness as a virtue Michael W Doyle concludes a recent study of realism by warningthat if we ldquowant to retain the range of insight embodied in the works of Thucydides MachiavelliHobbes and Rousseau we need to reject a monolithic conception of a Realist modelrdquo DoyleWays of War and Peace Realism Liberalism and Socialism (New York WW Norton 1997) p 195 seealso pp 137ndash160


analysis from nonrealist partsmdashoften by chapter or section75 For all thesereasons rigorous assumptions not received authority should determine thecoherence of social scientic theory

On the specic issue of status quo states moreover Schweller (like Snyderand Zakaria) simply misreads Morgenthau In this regard a true realist Mor-genthau introduces ldquostatus quordquo and ldquorevisionistrdquo policies (he terms them ldquothepolicy of imperialismrdquo and ldquothe policy of the status quordquo and adds also theldquopolicy of prestigerdquo) as strategies not preferences That is to say he seeks toshow that policies that appear to be the result of varying ideologies andintentions are in fact tactics in a common ldquostruggle for powerrdquo In the threechapters devoted to these policies in Politics among Nations such policies areexplained as responses to shifts in relative power owing to factors such as ldquolostwarsrdquo and ldquoweaknessrdquo76 For Schweller and other contemporary realists bycontrast the status quorevisionist distinction refers to exogenous variation instate preferences independent of power which in turn reects varied domesticcircumstances and state-society relations This is precisely the sort of theoreticalappeal that Morgenthau in this sense a true realist rejects explicitly Wolfersagain summarizes the matter succinctly ldquo[In the claim that] countries that seekself-extension tend to be the initiators of power competition and the resort toviolence lies the signicant kernel of truth in the idealist theory of aggres-sionrdquo77 Schweller has transformed realism into idealism

stephen van evera on the causes of war A nal example of the slidefrom power to preferences is visible in the work of Stephen Van Everamdashargu-ably the most inuential scholar among a generation that has revitalizedtheoretical debates in security studies Van Everarsquos magisterial study of thecauses of war aims explicitly to improve realism by highlighting the weak-nesses of objective material measures of aggregate power and redirecting usto consider a broader range of factors78

75 For example Morgenthau distinguishes consistently and explicitly between realist and nonre-alist elements The realist elements rest on ldquothe concept of interest dened in terms of powerrdquowhich ldquosets politics as [a] sphere of action and understandingrdquo independent of law moralityor economics Morgenthau Politics among Nations pp 5 7ndash8 12 He explicitly sets off sections onnorms and institutions with an introduction presenting them as alternatives to the initial realisttheory Ibid p 227 EH Carrrsquos classic realist statement contains similarly self-conscious dichoto-mies Carr The Twenty-Yearsrsquo Crisis 1919ndash1939 (New York Harper and Row 1964) pp 93ndash9476 See especially Morgenthau Politics among Nations pp 50ndash51 About this there is no exegeticalambiguity in the relevant chapters Still we agree with Robert Keohane that there is considerablecontradiction and paradigmatic ambiguity in these theorists77 Wolfers Discord and Collaboration p 9678 Stephen Van Evera Causes of War Power and the Roots of Conict (Ithaca NY Cornell Univer-sity Press 1999) citations are from a manuscript copy


Although Van Evera does employ some progressive extensions of the realistresearch program notably hypotheses on the consequences of variation in theoffense-defense balance his major focus is on preferences and perceptions notobjective measures of power79 He is admirably clear moreover in explicitlyrejecting both a Waltzian interpretation of biased preferences (which treatsthem as an indirect consequence of underlying power structures) and a ran-dom psychological or irrational explanation of them Instead Van Evera verymuch like Snyder in his work on imperialism attributes biased perceptions ofpower to the generalizable impact of four factors manipulation by elitesself-serving bureaucracies militarism and nationalist ideology Where thesefactors are present aggression and war are more likely As in Snyder rsquos workthe preferences and relative power of social groups are the underlying inde-pendent variables while perceptions and ideas often serve as an interveningprocess that widens and deepens the domestic inuence of those groups80

An obvious objection to such a broad denition of realism is simply that itlacks any analytical coherence What common assumptions can it claim withrealism

The problem here however is not simply the breadth and questionablecoherence of Van Everarsquos brand of realism A deeper aw is that the concretecausal mechanisms Van Evera citesmdashhis willingness to make these very explicitis among the most admirable qualities of his workmdashstems from existing non-realist international relations paradigms Most relate state behavior to inequali-ties and biases in the preferences and power of particularistic domestic interestgroups who mislead or coerce the less powerful to pursue policies to theirnarrow advantage Such arguments lie at the very core of the liberal interna-tional relations paradigm in which domestic misrepresentation is an importantsource of interstate conict in issues ranging from war to tariff policy

What could be more classically liberal for example than Van Everarsquos well-reasoned conjecture that misperceptions ldquooriginate with the worldrsquos propagan-dists spin doctors and professional obfuscators whose self-serving falsehoodsbecome national misperceptionsrdquo and that ldquopublics misperceive because theyare misled by national leaders state bureaucracies or propagandistsrdquo81 In hisinuential article ldquoPrimed for Peacerdquo Van Evera explains postndashCold War peace

79 Ibid p 880 Ibid pp 9ndash10 Van Evera is quite explicit that he considers this move realist ldquoThe theoriesdiscussed here address the effects of the structure of power or of perceptions of the structure ofpower As such they fall into the Realist camp Their explanatory power therefore adds to theoverall explanatory power of Realism and bolsters Realist arguments that power factors stronglyshape international politicsrdquo81 Ibid


in Western Europe with reference to a bold series of classical liberal generali-zations economically modern politically democratic nationalistically satisedgovernments with civilian control over the military liberal education complextechnological networks and generous social welfare systems do not provokewars82 Although more modern and sophisticated Van Everarsquos core thesis isan intellectual descendent of arguments advanced by early twentieth-centuryliberal ldquoidealistsrdquo like Normal Angell John Hobson Lionel Robbins andLeonard Wolff Properly understood in terms of its general assumptions andcausal processes Van Everarsquos scholarship is a conrmation of the unexpectedlyrobust predictive power of the assumptions underlying liberal or epistemicparadigmsmdasheven in the area of pure security studies83 The dilution of recentrealism obscures this essential commonality

Whereas the modern liberal international relations paradigm explicitly linksdomestic misrepresentation to general causal mechanisms and core assump-tions underlying phenomena from the democratic peace to tariff policy VanEverarsquos hypothesesmdashdespite the brilliance with which they are elaboratedmdashremain theoretically ad hoc They are related to no explicit set of paradigmaticassumptionsmdashthough we have seen their true provenance Little is gained andmuch lost by disconnecting such arguments from the liberal assumptions thatunderlie them and presenting them instead as realist Van Evera is doing morehere than simply challenging a narrow neorealist formulation of realism In allbut name Van Evera like Snyder Grieco Zakaria and Schweller has trans-formed realism into its opposite

from realism to epistemic theory power is what states believe it to beRealismrsquos central analytical leverage parsimony and distinctiveness derivefrom its ability to explain social life simply through variation in the distributionof objective material power capabilities rather than preferences perceptionsor norms As Benjamin Frankel succinctly puts it realism assumes ldquothat thereare things out there that exist independently of our thoughts and experienceWhen we admonish an individual to be realistic we urge that individual togive up beliefs or notions that y in the face of realityldquo84 Yet while contempo-rary realists continue to speak of international ldquopowerrdquo their midrange expla-

82 Stephen Van Evera ldquoCauses of War Volume 2 Misperception and Its Rootsrdquo unpublishedbook manuscript pp 9ndash1083 For a study demonstrating why it is necessary to treat Van Evera as something other than arealist in order to engage in any sort of orderly empirical testing see Peter Liberman ldquoThe Spoilsof Conquestrdquo International Security Vol 18 No 2 (Fall 1993) pp 125ndash15384 Frankel ldquoRestating the Realist Caserdquo p xiii


nations of state behavior have subtly shifted the core emphasis from variationin objective power to variation in beliefs and perceptions of power

This poses a fundamental problem If the perceptions and beliefs abouteffective means-ends calculations of states given adequate information con-sistently fail to correspond to material power relationships then power is atbest one of a number of important factors and perhaps a secondary one Theparsimony and coherence of realist theory is eroded85 When recent realiststheorize this relationship explicitly moreover they are forced to borrow propo-sitions more fully elaborated in existing epistemic theories which theorize theinuence of societal beliefs that structure means-ends calculations and affectperceptions of the environment If realism subsumes alongside traditionalmaterial capabilities factors such as national ideology organizational biasesand perceptions what remains theoretically distinctive If any governmentacting on the basis of geopolitical national interest or the aims of a particularis-tic interest group or ideationally induced strategies or misperceptions is inaccord with ldquorealistrdquo theory what plausible constraints on state behavior areexcluded

We have already glimpsed this tendency in the work discussed in thepreceding section As well as relying on exogenous variation in preferencesthese works accord causal signicance to exogenous shifts in collective beliefsabout means-ends relations Snyder and Van Evera dip into epistemic theorywhen they highlight ldquoblowbackrdquo whereby elites and states become trapped intheir own myths The resulting policies no longer serve either elite interestsas liberals predict or the maintenance of the balance of power as realistspredict86 Van Evera points in particular to the cultural factors independentfrom actual technology and military feasibility that shape how states view theoffense-defense balance87 Schweller invokes epistemes when he asserts that

85 Robert Jervis ldquoRealism Game Theory and Cooperationrdquo World Politics Vol 40 No 3 (April1988) pp 317ndash349 It is of course consistent with realism to trace the nature of perception andcalculation back to the distribution of material power as does Waltz in Theory of InternationalPolitics pp 168ndash17286 Snyder Myths of Empire pp 41ndash42 49 Van Evera recognizes the problem and resolves it byassertion ldquoThe Realist family includes causes lying in the structure of international power and inthe misperceptions of that structure although rather limited room is allowed for misperceptionsrdquoVan Evera Causes of War p 9 n 1287 Van Evera Causes of War chap 6 Building on Van Everarsquos earlier work Christensen andSnyder emphasize perception and misperception of the offense-defense balance to explain alliancepatterns in nineteenth- and twentieth-century Europe In contrast to Waltz and traditional realistshowever Christensen and Snyder do not view perceptions as endogenous to power (or as arandom product of uncertainty) but as the result of the lessons of past wars and the relativedomestic power of civilians and the military Their aspiration to synthesize different theories andlevels of analysis is a progressive step more generallymdasha point to which we return in the nal


contemporary Germany and Japan are not polar powers because they choosenot to have military power This he argues is not solely because they are statusquo powers but because of specic path-dependent beliefs about the efcacyof force In this Schweller would appear to join ranks with culturalists likeThomas Berger and Peter Katzenstein in arguing that German and Japaneseantimilitarism results from the socially embedded lessons of World War II notcurrent strategic opportunities and constraints88 Zakariarsquos study suggests thecentrality of a ldquocultural paradigm shiftrdquo in the ideas that underlay how Ameri-cans thought about foreign policymdasha shift he leaves unexplained89 In turningto culture Zakaria joins the long tradition of realists who nd US foreignpolicy anomalous because ldquorealism is largely alien to American culturerdquo90 Wecan more closely observe the shift from realism to epistemic theory in consid-ering the work of Stephen Walt and William Wohlforth

stephen walt on alliances Stephen Walt is an effective critic of neoreal-ism which he considers too spare to explain balancing behavior accurately Hetherefore seeks to supplant ldquobalance-of-powerrdquo theory with ldquobalance-of-threatrdquo theory Alliances are triggered by imbalances of ldquothreatrdquo not imbal-ances of ldquopowerrdquo Unlike the concept of power the concept of external ldquothreatrdquoincludes ldquoperceived state intentionsrdquo alongside more clearly realist variableslike economic resources military technology and geography Walt goes on todemonstrate convincingly that the primary purpose of alliances is to balancenot to bandwagon against threatsmdashan important contribution

Walt explicitly labels this move as a progressive and parsimonious revisionof realist balance-of-power theory91 Yet ldquobalance-of-threatrdquo theory in fact sac-

section Yet there remains considerable ambiguity whether Christensen and Snyder believe this isa ldquoprogressive problem shiftrdquo within the realist paradigm or a form of theory synthesis Certainlythey are often cited as realists (eg Grieco ldquoRealist International Theoryrdquo p 181) Either wayChristensen and Snyder clearly demonstrate the fundamental limits not just of neorealist theorybut of the realism paradigm more broadly Thomas J Christensen and Jack Snyder ldquoChain Gangsand Passed Bucks Predicting Alliance Patterns in Multipolarityrdquo International Organization Vol 44No 2 (Spring 1990) pp 144 166 Christensen ldquoPerceptions and Alliances in Europerdquo InternationalOrganization Vol 51 No 1 (Winter 1997) p 65 and Christensen and Snyder ldquoProgressive Researchand Degenerate Alliancesrdquo pp 920ndash92188 Schweller Deadly Imbalances pp 19 88 164ndash168 20089 In an otherwise highly complimentary review historian Walter McDougall calls our attentionto precisely this empirical indeterminacy McDougall ldquoAmerican Empire Review of FareedZakaria From Wealth to Power The Unusual Origins of Americarsquos World Rolerdquo New York Times BookReview May 3 1998 p 2590 Mearsheimer ldquoThe False Promise of International Institutionsrdquo p 37691 For specic claims of a progressive shift see Stephen M Walt The Origins of Alliances (IthacaNY Cornell University Press 1987) pp viii 5 21 263ndash265 The underlying problem is that aprediction of balancing is not unique to realism Nearly all international relations paradigms andtheories predict that states align and balance against threats to the realization of onersquos interestswhether the latter are status quo or revisionist Why else would a rational government form a


rices the theoretical coherence and distinctiveness not just of neorealism butof realist theory more broadly By combining exogenous changes in power andstate perceptions of the intentions of others into a single variable Waltrsquosldquobalance-of-threatrdquo approach excludes virtually no potential cause of rationalbalancing short of irrational altruistic or incoherent state action92 Combiningintentions and power into dimensions of a single variable without an a prioriweighting numerous critics have observed is a source of fundamental inde-terminacy Walt himself concedes that ldquoone cannot determine a priori whichsources of threat will be most important in any given caserdquo93 An example ishis analysis of the Cold War bipolar balance in Europemdasharguably the mostimportant single set of alliances in the twentieth century and a ldquocriticalrdquo realistcase Perceived state intentions (and geography) reverse the predictions de-rived from pure power balancing thereby leading country after country to sidewith the overwhelming US-led coalition Absent a clear weighting of factorshow are we to judge (even in principle) whether this conrms or disconrmsWaltrsquos basic theory

Having cast the theoretical net so widely Walt necessarily encountersdifculty clearly dening plausible alternative theories (neorealism aside)against which to test his own At rst glance he appears to treat ideology asan alternative explanation Yet in fact Walt rejects only very primitive forms ofideological motivation almost absent from international relations theory nota-bly that governments ally with those who espouse similar formal ideologicaldoctrines94 ldquoBalance-of-threatrdquo theory subsumes most other ideological argu-

military alliance Theories differ in their predictions about conditions under which states balanceLiberal theories predict balancing against ldquoaggressorrdquo states institutionalist theories predict bal-ancing within institutions epistemic and some constructivist theories predict balancing where itis perceived as efcacious and realist theories predict balancing against power92 For a striking statement see ibid p 149 ldquoIn the Arab world the most important source ofpower has been the ability to manipulate onersquos own image and the images of onersquos rivals in theminds of other Arab elites We are therefore dealing with two broad types of balancingbalancing conducted by military means [and] balancing conducted by political means directed atan opponentrsquos image and legitimacyrdquo There remains ambiguity about whether perceptions ofintentions involve basic variations in preferences or beliefs about such intentionsmdashthe heart ofwhich lies in the lack of theoretical constraint Walt is able to impose on ldquostate intentionsrdquo93 Ibid p 22 Our criticism is not simply this indeterminacy but that his concept of ldquothreatrdquosubsumes all but the most implausible of prevailing rationalist explanations Ibid p 26 Waltztakes a similar view arguing that Walt (like Schweller and others) should not be seen as ldquoincreasingthe explanatory power of defective theory and making it more preciserdquo but bringing in extra-theo-retical variables ldquoWaltrdquo he writes ldquo[has] unfortunately taken the imaginative application of thetheory to be a new onerdquo Waltz ldquoEvaluating Theoriesrdquo p 916 See also Gunther Hellmann andReinhard Wolf ldquoNeorealism Neoliberal Institutionalism and the Future of NATOrdquo Security Stud-ies Vol 3 No 1 (Autumn 1993) pp 3ndash4394 Douglas J MacDonaldrsquos review in Journal of Politics Vol 51 No 2 (August 1989) p 796accuses Walt of employing a ldquorigidrdquo denition Walt acknowledges for example that the Soviet


ments (as a primary source of ldquostate intentionsrdquo) in particular the claimmdashfarmore common in the international relations literaturemdashthat states ally againstideologies perceived as ldquoaggressiverdquo and with those perceived as ldquounthreaten-ingrdquo Factors such as Pan-Arabism in the Middle East Hitler rsquos expansionistview of security before and during World War II and ldquothe divisive character ofSoviet Marxist-Leninism an ideology calling for the authoritative leader-ship of the Socialist system by Moscowrdquo play central causal roles in hisempirical explanations95

Of more concern than the indeterminacy of Waltrsquos theory is its appropriationof nonrealist causal mechanisms The ldquoaggressive intentionsrdquo underlyingldquothreatsrdquo include precisely those predicted by nearly all nonrealist explana-tions of alliance formation As seen in the examples just cited interpretationsof the intentions of others play a central role in the alliances that occur andthese are shaped in turn by the compatibility of strategic beliefs and percep-tions held by different countries as epistemic theory predicts The allianceamong postwar West European democracies (along with a few noncommunistauthoritarian states) reects in large part the perception that they posed lessof a threat to one another than did the Soviet Union and its alliesmdashwhich helpsreverse the impact of material variables Walt does not clearly specify whetherthe Western perception of aggressive Soviet intentions is the product of under-lying preferences as liberal theories of peace and war (notably democraticpeace theory) predict or of strategic beliefs and perceptions as epistemictheory predicts We do know however that these intentions are exogenous toeconomic capabilities military technology and geographymdashdistinctive vari-ables in traditional realist theory What is gained by terming this unwieldysynthesis a progressive extension of ldquorealismrdquo thereby impeding any possibleempirical challenge from more plausible nonrealist explanations

william wohlforth on the end of the cold war The centrality of per-ceptions and beliefsmdashand thus epistemic theorymdashis even clearer in WilliamWohlforthrsquos analysis of Soviet (and US) policy during the Cold War96 LikeZakaria Wohlforth argues that state behavior is shaped most fundamentally

Union allied with leftist regimes in the Middle East and the United States did not but he doesnot treat such actions as ideologically motivated because neither superpower demanded that itsallies alter their domestic policies95 Walt The Origins of Alliances pp 149 168 266 and Stephen M Walt ldquoAlliance Formation andthe Balance of World Politicsrdquo in Brown Lynn-Jones and Miller The Perils of Anarchy pp 214 243(emphasis in original)96 We set aside another problem namely Wohlforthrsquos evident reliance on a distinction betweenstatus quo and revisionist states as a ldquocontextualrdquo factor explaining Soviet preferences and inparticular the absence of a ldquoWorld War IIIrdquo As we have already discussed degeneration into liberaltheory we focus here on Wohlforthrsquos overt perceptual challenge to objective power analysis in


not by exogenous variation in objective power but in varying perceptions ofpower Unlike Zakaria however who employs perceptions of ruling elitesprimarily as a method to observe the working of more fundamental variablesWohlforth asserts that perceptions are exogenous variables systematicallyskewed and of great causal importance This becomes the centerpiece of histheoretical innovation97 Objective power shifts he argues ldquocan account nei-ther for the Cold War nor its sudden endrdquo Either objective measures of powerare ldquonot even roughly accurate indicatorsrdquo of true power or ldquopower does notmatterrdquo98 The end of the Cold War Wohlforth contends is instead a storyabout reactions to (often questionable) perceptions of power

If perceptions and power diverge however power no longer necessarilyserves as the primary independent variable driving state behavior If bothpower and beliefs about power matter it becomes unclear in principle whenone or the other predominates The coherence of the realist core is erodedInsofar as Wohlforth seeks to render this problematic mix determinatemdashandcareful historical reconstruction is an unambiguous strength of his workmdashhedoes so by replacing realist variables and causal mechanisms with those drawnfrom existing epistemic and liberal theories Wohlforth generally views statesas being guided by embedded beliefs about foreign policy that are relativelyresistant to change even when experience with the material environmentclearly signals the need for it

Consider for example Wohlforthrsquos explanation of the timing of the suddenSoviet perception of decline in the late 1980s that in his view brought the ColdWar to an end He relies on four concrete causal mechanisms at least three ofwhich (and perhaps all four) are more consistent with international relationsparadigms other than realism The rst and ambiguous factor is the ldquoscientic-technical revolutionrdquo which sparked a desire to reform the socialist economiesIt is unclear whether Wohlforth views this as a straightforward source ofmaterial weakness as realist theory would have it or as a shift in the dominantmodels and standards for economic growth as epistemic theory would predictor as a qualitative change in domestic views about the need to link economicmodernization with an opening to the West as liberal theory would suggest99

William C Wohlforth ldquoRealism and the End of the Cold Warrdquo in Brown Lynn-Jones and MillerThe Perils of Anarchy pp 32 36ndash3797 William C Wohlforth The Elusive Balance Power and Perceptions during the Cold War (IthacaNY Cornell University Press 1993) pp 1ndash1798 Wohlforth ldquoRealism and the End of the Cold Warrdquo p 4199 Wohlforth The Elusive Balance pp 233ndash234 242ndash243 251 and Wohlforth ldquoRealism and the Endof the Cold Warrdquo pp 19ndash22 37


Whereas it remains unclear whether Wohlforthrsquos rst factor is realist theremaining three clearly are not100 One was that the Soviet Union from JosephStalin through Leonid Brezhnev was guided by a ldquocorrelation-of-forcesrdquo theorythat not only saw capitalism as a threat but also held that states bandwagonto power especially military power This led successive leaders to discountevidence of Soviet decline Perceptions and ideas suddenly shifted in the late1980s because of the endogenous dynamics of epistemic structures The corre-lation of forces model he argues became so entrenched and formalized thatit generated excessive expectations making it more vulnerable to suddenchange from failed predictions101 Still another factor was the new role of theUnited States in the 1980s which no longer ldquobuttressedrdquo the Soviet perceptionof well-being by treating it as a rising power whose interests had to beaccommodated102 Does this reasoning not reduce relative power to whateveronersquos enemy acknowledges it to bemdasha theme more constructivist than realist

The nal factor was the formation of an overwhelming balance of poweragainst the Soviet Union in which revolutions in East Central Europe consti-tuted the nal decisive steps103 East European revolutions Wohlforth argueshad a symbolic effect on Soviet power perceptions because they ldquobegan to callsocialismrsquos vitality into questionrdquo Such an explanation faces precisely thedifculties that plague Waltrsquos theory of alliances Like Walt Wohlforth offersneither a distinctively realist explanation for why an ever-expanding anti-So-viet coalition should form nor more fundamentally an explanation for whychanges in regime type should inuence state calculations Why do East Euro-pean governments not move toward the Soviet Union as it declines andmellows Moreover Wohlforth treats the Soviet policy choicemdashits withdrawalfrom Eastern Europe and opening to the Westmdashas one designed to inducechanges in Western perceptions of the Soviet threat rather than to alter thebalance of power104 But what is realist about this world in which imbalancesimages and internal politics override sober calculations of relative power

100 For a presentation of a clearer realist argument see Stephen G Brooks and William CWohlforth ldquoHow Identities Change Material Forces Identity Transformation and the End of theCold Warrdquo unpublished manuscript March 1999101 Wohlforth The Elusive Balance pp 52ndash53 220ndash229 250102 Wohlforth ldquoRealism and the End of the Cold Warrdquo pp 21ndash22 32ndash35103 Ibid pp 23 34 38ndash39104 Ibid p 23 It remains unclear whether Wohlforth means to argue that regime type actuallyshifted East European policies or shifted Soviet perception of its relative power Either way theconnection to relative power capabilitiesmdasheven understood in a more ne-grained sensemdashremainsunclear Wohlforth also stresses the unintended consequences of Soviet policy shifts particularlyin Eastern Europe


from realism to regime theory power is what states agree it to beThe realist assumption that the distribution of material resources is the criticalexogenous variable determining state behavior implies not only that the ac-tions of states remain uninuenced by variation in state preferences or beliefsbut that such actions remain essentially uninuenced by international institu-tions In the realist view governments cannot induce changes in their relativeinuence by entering into multilateral commitments From Bismarckian criti-cisms of the nineteenth-century Concert of Europe to Mearsheimerrsquos criticismof contemporary regime theory realists have delighted in demonstrating thatinternational institutions reect and ratify but do not transform existingpower relations Those who believe otherwise are ldquolegalistsrdquo105

Yet after rejecting the independent inuence of international institutions forcenturies realists are suddenly embracing ldquolegalismrdquo Recent realist theory notonly treats international institutions as autonomous forces in world politicsbut at times views their impact as far stronger if also far more ad hoc thandoes conventional regime theory The tendency of recent realists to reverse thiscausal mechanism not only undermines realismrsquos coherence and distinctive-ness but is ultimately parasitic on existing regime theorymdashnot least becauserealists have yet to offer a distinctive theory of why international institutionsinuence state behavior We illustrate this tendency with examples drawn fromthe prominent work of Joseph Grieco and Charles Glaser

joseph grieco on european monetary integration A striking exampleof the slide from realist to institutionalist assumptions is found in JosephGriecorsquos attempt to employ realist theory to explain European integration andmore generally the formation of international economic regimes For realiststhe agreement among European Union members at Maastricht in 1991 to moveto a single currencymdashEconomic and Monetary Union (EMU)mdashappears anoma-lous106 Absent coercion how could states primarily concerned with ldquorelativegainsrdquo ever agree to surrender basic elements of state sovereignty to an ambi-tious international institution of this kind

105 On Bismarck see Henry Kissinger Diplomacy (New York Simon and Schuster 1994) pp120ndash131 On the realist view of international law and institutions see Mearsheimer ldquoThe FalsePromise of International Institutionsrdquo Recall that institutionalists adhere to nearly all the sameassumptions as realistsmdashan underlying state of anarchy states as rational egotistical actors sub-stantial conict of interestmdashbut argue that governments faced with collective action problems cancontract among themselves to mitigate the major disadvantages of anarchy106 Grieco ldquoRealist International Theoryrdquo pp 184ndash186 Liberals and institutionalists have rela-tively little trouble explaining this outcome See Andrew Moravcsik The Choice for Europe SocialPurpose and State Power from Messina to Maastricht (Ithaca NY Cornell University Press 1998)


In a bold move Grieco seeks to extend realist theory to encompass andexplain such behaviormdashthereby generating a realist theory of internationalinstitutions He proposes the ldquobinding hypothesisrdquo whereby weak statesrather than forming a balancing coalition against or submitting to the will ofa larger state propose legal commitments that allocate voting rights withininternational institutions so as to redistribute power from the powerful tothemselves EMU he argues was a Franco-German bargain in which Germanysurrendered its power to satisfy French and Italian fears that cooperationwould undermine their power In sum international institutions are a meansof alienating and transferring state power

By introducing an autonomous role for international institutionsmdashone evenmore powerful than that institutions play in most conventional studies ofregimesmdashGriecorsquos reformulation sacrices realismrsquos coherence and distinctive-ness It sacrices coherence because the analysis rests on contradictory (ifunstated) sets of assumptions about the constraints on state behavior most ofwhich cannot be traced back to the exogenous impact of relative powerAccording to Griecorsquos reformulation institutional commitments might be afunction of underlying state power as realists have traditionally argued orthey might be a factor alongside state power as Griecorsquos case study seems tosuggest or they might even be a determinant of state power as the bindinghypothesis suggests107 Because these contradictory directions of causality co-exist Griecorsquos formulation of realism subsumes the entire universe of interna-tional relations theories about international institutions Thus a governmentfacing a powerful country may balance against it submit to its demands orcontract with it Bargaining outcomes and institutional commitments mayfavor strong states or weak states International commitments may be credibleor not No rational state calculations strategies or outcomes are privileged orexcluded Grieco offers no assumptions specifying even in principle where tolook for causal mechanisms antecedent conditions or weighting of competingconsiderations that would render these predictions more determinate in anyspecic case As Grieco himself concedes realist predictions about ongoingnegotiations over EMU are therefore fundamentally indeterminate108

107 Joseph M Grieco ldquoState Interests and Institutional Rule Trajectories A Neorealist Interpreta-tion of the Maastricht Treaty and European Economic and Monetary Unionrdquo in Frankel Realismpp 287ndash290108 Grieco with admirable honesty concedes the indeterminacy ldquoIt will be of intense interest tostudents of international politicsrdquo he concludes ldquoto observe whether institutions [ie Griecorsquosbinding hypothesis] or underlying differentials of power [ie the conventional realist argument]will have a greater impact on the future course of European monetary affairsrdquo Ibid p 304 Yet


To explain outcomes so clearly at variance with traditional realist premisesGrieco further sacrices realismrsquos distinctiveness by invoking basic assump-tions and causal processes central to nonrealist paradigms To explain the mainpuzzle of EMU from a realist perspectivemdashwhy in a world of relative-gainsseekers powerful states (in this case Germany) would agree uncoerced toalienate sovereigntymdashGrieco is forced to reintroduce absolute gains and mis-perceptions Having rejected the possibility that Germany was ldquobalancingrdquoagainst the United States he concludes that there must have been commongains or that either Germany or its partners (or both) misperceived the truecosts of EMU To explain why the commitments of Germany and others arecredible moreover Grieco relies implicitly on the notion that institutionsstrengthen the credibility of commitmentsmdashthe core prediction of functionalregime theory Yet he neither acknowledges the transaction-cost logic of func-tional regime theory nor provides an explicit alternative to it Behind therhetoric realism has been transformed into its nemesis109

charles glaser on signaling and arms control Charles Glaser has ad-vanced a sophisticated synthetic view termed ldquocontingent realismrdquo Part of hisargument is designed to show that a stable world in which states signalpeaceful intent and engage in tacit or formal arms control is consistent withrealist theory Here Glaser like Grieco shifts the analytical focus from causalreliance on exogenous variation in the distribution of capabilities to exogenousvariation in the international informational and institutional environmentGlaser aims to show that even if we adopt structural realist assumptionscooperation is much more likely than realists commonly assume and can besubstantially assisted by international regimes Glaserrsquos argument is overtlyfunctional International institutions provide information to states that helpsthem to realize common interests and joint gains

in a realist theory of European integration in the 1990s shouldnrsquot the outcome of EMU be a decisivetheoretical prediction not a matter of empirical happenstance109 Grieco ldquoRealist International Theoryrdquo pp 185ndash186 and Grieco ldquoState Interests and Institu-tional Rule Trajectoriesrdquo p 286 Grieco rightly observes that the historical record does not conrmthat initial and ongoing support for the agreement by the most powerful government that ofGermany (rather than imposing an institutional solution on weaker countries) can be explained asan effort to balance against US monetary power Grieco invokes at various points the claim thatEMU generates absolute gains (as liberals maintain) and that Germany or others may havemisperceived the likely economic outcome (as epistemic theory might suggest) or because inter-national institutions enhance the credibility of national commitments (as institutionalists maintain)He also argues that Germany was forced to grant a quid pro quo in exchange for Germanunication but this ies in the face of a growing consensus that the German commitment to moveto EMU began months even years before reunication and did not weaken when reunicationwas complete For a review of the evidence see Moravcsik The Choice for Europe chap 6


To explain how and predict when this will occur Glaser adds an exogenousvariable to realismrsquos concern with relative power transaction costs110 To mini-mize conict Glaser assumes states must establish their own defensive inten-tions and reduce uncertainty about whether other states are ldquogreedyrdquomdashthat iswhether they prefer more than just security While unilateral policies maysometimes achieve the same end formal international institutions (ie armscontrol regimes) may help states achieve this efciently by signaling or enhanc-ing the credibility of commitments through monitoring111 According to GlaserldquoInstitutions that provide information and reduce transaction costs donot pose a problem for structural realism Nothing about the roles performedby this type of institution conicts with structural realismrsquos basic assump-tionsrdquo Glaserrsquos argument is that tacit coordination or perhaps formal interna-tional institutions can be employed to generate joint gains where thetransaction costs of decentralized signaling coordination and monitoring arehigh

Yet if Glaserrsquos reformulation of realism encompasses not only the distribu-tion of military power but also exogenous variation in costsmdashand implicitlyadmits albeit as a nonrealist factor the role of ldquogreedyrdquo statesmdashdoes it notencompass the assumptions of the institutionalist paradigm Has it become ageneric commitment not to a distinct realist theory but simply to a lowest-common-denominator rationalism Acknowledging that contingent realismmight appear overly broadmdasha dilution rather than a deepening of realistpremisesmdashGlaser explicitly seeks to establish his realist credentials by demon-strating that ldquocontingent realismrdquo does not change ldquostatesrsquo motives toaltruismrdquo or grant ldquotremendous control to an international authorityrdquo Yet thisonly serves to demonstrate the difculty contemporary realists face oncehaving appropriated (but not theoretically subsumed) nearly all rationalistalternatives in locating plausible competing theories Glaserrsquos alternatives arestraw men Both altruism and a world state have been utterly absent fromscholarly debates for nearly half a century112 Neither is advocated or analyzed

110 Glaser is ambiguous on the role of motives In ldquoRealists as Optimistsrdquo pp 394ndash397 he arguesthat ldquocontingent realism suggests the importance of motivesrdquo and renders standard power vari-ables ldquoless importantrdquo In Charles L Glaser ldquoThe Security Dilemma Revisitedrdquo World Politics Vol50 No 1 (October 1997) p 191 he argues ldquocontrary to the standard [realist] argument countriesshould not focus solely on capabilities but also on motivesrdquo Kydd ldquoSheep in Sheeprsquos Clothingrdquop 17 brings the distinction between democracy and nondemocracy as well as variation inideology to explain such behavior111 Glaser ldquoRealists as Optimistsrdquo p 410ff112 Ibid p 411 Another example is Mearsheimerrsquos detailed refutation of collective securitytheory yet Mearsheimer cites very few if any clear advocates of collective security (as opposed


in any signicant contemporary institutionalist or liberal scholarshipmdasha hand-ful of recent writings on altruistic motives for human rights or environmentalactivism aside More specically is not the effort to show that largely realistassumptions generate predictions of cooperation even institutionalization thecentral premise of Keohanersquos seminal statement of institutionalism in AfterHegemony And is Keohanersquos theoretical solution drawn from the transaction-cost economics of Oliver Williamson not identical to Glaserrsquos113 What isgained by terming all these competing rationalist claims ldquorealistrdquo

Practical Advantages and Broader Implications Why Reformulate Realism

The works considered above make innovative and valuable contributions toscholarly understanding of world politics particularly at the level of midrangepropositions There is much to be said in defense of their empirical insightsand midlevel theorizing which we have necessarily slighted here They belongamong the most fruitful advances in recent international relations scholarshipYet the fact that scholars working under a particular label produce interestingempirical insights is not the only criterion by which to judge a theoreticalparadigm The question is not simply whether such authors provide interestingexplanations but what their ndings tell us more generally about worldpolitics

A causal reading of recent realist research would lead one to believe thatrealists have successfully found innovative ways to build on core realist as-sumptions to explain new aspects of world politics Yet if the true assumptionsand causal mechanisms underlying much recent realist research by self-styledrealists was made explicit we have argued realismrsquos afnity with existingliberal epistemic and institutionalist theories of world politics would becomeclear This systematic mislabeling of ndings has tended to isolate realists fromimportant trends in international relations theory In lieu of fully theorizingfactors like domestic preferences collective epistemes and international insti-tutions many realists are tempted to deny that any true theories (eg ofdomestic politics) are possible Rose goes so far in his prominent review essayto proclaim this unwillingness to theorize domestic politics fully as a deningvirtue of contemporary realismmdasha claim for which he cites Aristotle114 Yet

to concerts or regimes) after the 1950s Robert Keohane explicitly rejects such an ldquoidealistrdquo accountKeohane After Hegemony pp 59ndash60 n 2113 On Williamsonian theory see Keohane After Hegemony114 See n 60


little is gained by distinguishing realism as a theory only to reintroduce liberalepistemic and institutionalist arguments in a vaguer form by loosening thespecication of core concepts At best this is insular at worst quite misleadingScholars would benet if the role of these factors and the assumptions aboutworld politics on which they rest were made explicit and theories weregrouped accordingly More rigorous and consistent links between hypothesesand underlying assumptions would clarify the landscape of international rela-tions theory Our general theoretical understanding would for example bequite different if the work of Snyder on imperialism Grieco on relative gainsVan Evera on war and Zakaria or Schweller on national expansion weretreatedmdashas their assumptions demandmdashas part of the same liberal researchprogram that has given rise to theories of the democratic peace and commercialpolicy

Yet the issue here is not just the provision of a simpler and more accurateguide to the actual assumptions of major international relations theories im-portant though that goal may be Like the authors of the articles we examineabovemdashall of whom took great care to underscore the realist nature of theirclaimsmdashwe believe that proper denition of basic theories has practical impli-cations for theoretical debates empirical research and pedagogy Specicallywe believe that adherence to our reformulation would facilitate more decisivetests among existing theories dene more sharply the empirical domain ofrealist theory and provide a superior foundation for multicausal synthesisbetween realism and other theories

refocusing empirical testsOne implication of our proposed reformulation of realism is that its conceptuallanguage permits scholars to represent the theoretical implications of ongoingempirical research Testing theories is a way of evaluating the assumptions thatunderlie them The proper identication of those assumptions is the mostimportant reason why the semantics of paradigmatic debates matter Webelieve that a central issue in international relations todaymdashas it has been for200 if not 2000 yearsmdashis to assess the relative inuence on world politics ofand the interactions among four factors the distribution of material resourcesthe distribution of preferences the distribution of beliefs and the distributionof information These factors are critical we submit regardless of the languageone uses to describe theoriesmdashformal or informal traditional or modernThese four categoriesmdashpower preferences beliefs and informationmdashroughlycorrespond to the four major categories of modern rationalist international


relations theory namely realist liberal epistemic and institutionalist theo-ries115 These theories correspond also to the four generic determinants of actorbehavior in fundamental rationalist social theory resources tastes beliefs andinstitutions116

Were empirical research consistently structured around the relative impor-tance of (or as we shall see below the interaction among) these factors thediscourses of major traditions in international relations theory would surelybecome more consistent with one another and with fundamental social theoryStudents of world politics could better judge what is at stake in empiricalresearch and theoretical debate Acceptance of a reformulated realism wouldclarify what is in fact a realist argument and what is not and thereby refocustheoretical debate and empirical research on the enduring issues of worldpolitics raised by realismrsquos traditional skepticism of intentions ideology andinstitutions Curiously insular contemporary realists are strikingly reluctant totest their hypotheses against nonrealist theories Loose formulations of realismdiscourage decisive empirical testing against fundamentally competing ration-alist views Proper paradigmatic denition reveals that the theoretical innova-tions in recent defensive and neoclassical realist research in fact conrmassumptions and causal mechanisms underlying the liberal epistemic andinstitutionalist paradigms more than those underlying the realist paradigm

This would open a number of new and compelling areas for empiricalconfrontations among theories that are currently blocked by contempor-ary minimal realist formulations We believe that more ne-grained empiri-cal debates would become theoretically inescapable Consider the followingpossibilities

imperialism By combining power and preferences in his explanation ofimperialism Snydermdashas Zakaria observesmdashblurs the relative importance ofthe two Subsequent realist studies of expansion including those by Zakariaand Schweller similarly fail to distinguish the role of power vacuums on theone hand and ldquostrongrdquo domestic states or ldquorevisionistrdquo aggressors on theother If scholars explicitly separated developed or tested nonrealist theoriesit would become possible to discern the relative inuence of each

alliances Walt by structuring his analysis of alliance formation as a di-chotomous contest between the ldquobalance of threatrdquo on the one hand and

115 For example Lake and Powell in Strategic Choice employ nearly identical categories butdifferent labels116 Coleman Foundations of Social Theory and Elster ldquoIntroductionrdquo


irrational ideology on the other subtly discourages empirical tests of realismand alternative explanations How much of alliance behavior can be explainedby capabilities geography and technology and how much by state ldquointen-tionsrdquo In assessing threat to what extent should scholars and policymakersbe concerned about military might and to what extent the management ofimages and the accumulation of cultural capital

cooperation Grieco by structuring discussions of cooperation around thedichotomy of ldquoabsolute-gains seekingrdquo and ldquorelative-gains seekingrdquo discour-ages investigation of competing sources of conictual (ldquorelative-gains seek-ingrdquo) behavior Who is correctmdashliberals who attribute conict to deadlockedpreferences epistemic theorists who point to conicting embedded beliefsrealists who invoke security externalities or institutionalists who highlightcoordination (bargaining) failure Current realist theory which combines allfour into ldquorelative-gains seekingrdquo evades this question More ne-grainedstudies would provide more insight

war and peace In their studies of hot and cold wars Van Evera andWohlforth focus on power and perceptions of power In doing so they eithersubsume or ignore a series of narrower explanations for the beliefs that theyconclude are at the heart of world politics Only recently however have webegun to see focused tests between variants of realist liberal epistemic andinstitutionalist theory117 More would be welcome

hegemony Scholars have isolated four different variants of hegemonicstability theory each grounded in a separate aspect of international leadershipA liberal variant stresses variation in differential competitiveness an epistemicvariant looks to shared ideas and beliefs between leader and followers aninstitutionalist variant emphasizes the provision of institutional infrastructureand a realist variant stresses the hegemonic provision of resources that permitsldquofollowerrdquo governments to defray the short-term costs of adjustment inexchange for which the hegemon gains inuence over the terms of futurecooperation or benets from security externalities As David Lake has ob-served more attention could be paid to the relative power of these fourexplanations118

the virtue of limits specifying realismrsquos proper explanatory domainAssumptions dene the empirical scope of a paradigm A more precise anddistinct paradigm based on more than a minimal commitment to rational state

117 See for example Russett Grasping the Democratic Peace118 Lake ldquoLeadership Hegemony and the International Economyrdquo


behavior in anarchy should thus offer a more precise specication of thepredicted empirical domain of realist theories Most realists today we haveseen assert that realist hypotheses should always enjoy analytical prioritymdashat least wherever security issues are involved119 Yet no single theory canor should claim to explain all of world politics or to be empirically preeminentunder all circumstances Assertions of blanket preeminence underminethe credibility of modern realism whereas acceptance of assumptions thatimpose explicit constraints on empirical domain would be a sign of theoreticalmaturity

The three assumptions we propose in the rst section of this article suggesta more sharply dened domain than that claimed by contemporary realistsyet one that eliminates many apparent realist anomalies Realist theory doesnot apply across the board to security affairs It is appropriate instead only tothose cases marked by severe underlying conict of interest (economic ideo-logical or political) relative to the cost of overt coercion or inducement If theunderlying preferences at stake are weak or the relative cost of exercisingpower is high states will have little incentive to threaten or provoke overtcoercion or inducement and the outcome of conict is more likely to reectan institutionalized focal point (as institutionalists argue) concordantdiscor-dant causal beliefs (as epistemic analysts argue) or the relative preferenceintensity (as liberals argue)120

For the realist link between total resources and bargaining outcomes to holdboth parties must consider that the issues at stake are of paramount impor-tance Realist claims should therefore be limited to circumstances in whichstates are motivated by strong and symmetrical underlying conicts in prefer-encesmdashoverlapping territorial economic or ideological claimsmdashor situationswhere the cost of coercion is so low (at least to one party) that its cost-effectiveuse is feasible This explains why security disputes among advanced industrialdemocracies tend to be resolved nonmilitarilymdasha liberal prediction consistentwith the near total suppression of realist politics among them observed bySchweller Snyder Grieco Van Evera and others121 In such cases realist theory

119 Waltz Theory of International Politics Mearsheimer ldquoFalse Promise of International Institu-tionsrdquo p 351 and Charles Lipson ldquoInternational Cooperation in Economic and Security Affairsrdquoin Baldwin Neorealism and Neoliberalism pp 60ndash84120 We should expect war and realist security dynamics for example only involving at least onestate sufciently ldquoaggressiverdquo to raise ldquovitalrdquo interests for all involved Here we nd support fromSchweller ldquoNew Realist Research on Alliancesrdquo pp 928ndash929 For a classic statement of thisposition see also Stanley Hoffmann Duties beyond Borders On the Limits and Possibilities of an EthicalInternational Politics (Syracuse NY Syracuse University Press 1981) pp 14ndash16121 We have noted these examples above Such scholars voice criticisms of what they take to beliberal views but they tend to take the form of either skepticism that democracies are stable or


is not disconrmed but is simply inappropriate because its assumptions arenot met Similarly in cases where the stakes are asymmetricalmdashfor examplethe Boer War Hitlerrsquos remilitarization of the Rhineland and more recentperipheral conicts in Vietnam Afghanistan and Chechnyamdashan intense pref-erence or ldquopolitical willrdquo concerning the particular issue at stake can compen-sate for a deciency in capabilities122 Specifying the proper domain of realisttheory is thus a means not only to constrain realism but also to strengthen itby limiting claims to domains where it should be expected to apply A leanerrealism may be meaner

beyond monocausal mania moving toward theory synthesisSome readers may object that debates about the accuracy and scope of uni-causal explanations of world politics are unnecessarily limited Is it realistic tomaintain that patterns of important complex events in world politics are theresult of a single factor Is not theory synthesis the real goal of the defensiveand neoclassical realists we critique however they label their empirical claimsDoesnrsquot excessive attention to ldquoismsrdquo encourage sectarian and semantic battlesamong schools that would be better treated as elements within broader inte-grated explanations Isnrsquot this the implication even of our own specication ofrealism in terms of interstate bargaining

We agree Our purpose in this article is not to freeze the lines betweenunicausal paradigms The replacement of what John Ruggie has termedldquomonocausal maniardquo with such multicausal even multiparadigmatic synthe-ses we believe is desirable even imperative It is the future of internationalrelations theory The unavoidable rst step however is to develop a set ofwell-constructed rst-order theories Multicausality without a rigorous under-lying structure only muddies the waters encouraging ad hoc argumentationand obscuring the results of empirical tests123

We submit moreover that a major advantage perhaps the most importantone of our proposed reformulation of the denition of realism is that itsuggests an easily operationalizable and internally coherent mechanism for

concerns about the transition to democracy neither of which supports realist claims against liberalones122 Moravcsik ldquoTaking Preferences Seriouslyrdquo pp 523ndash524 See also Posen The Sources of MilitaryDoctrine pp 60ndash61 Morrow ldquoSocial Choice and System Structurerdquo pp 83ndash84 and Andrew MackldquoWhy Big Nations Lose Small Wars The Politics of Asymmetrical Conictrdquo World Politics Vol 27No 2 (January 1975) pp 175ndash200123 John Ruggie personal communication For a view that one theory must dominate see EthanB Kapstein ldquoIs Realism Dead The Domestic Sources of International Politicsrdquo InternationalOrganization Vol 49 No 4 (Autumn 1995) pp 751ndash774


synthesizing realism with other theories This mechanism we have termed theldquotwo-stagerdquo or ldquotwo-steprdquo method124 The two-stage or two-step method as-sumes as any rationalist (or boundedly rational) theory of state behavior mustthat in world politics in which states are the relevant actors the domestic andtransnational state-society relations of preference and belief formation can beanalytically separated from the strategic logic of interstate interaction whetherexplained by realism or institutionalism If we understand international rela-tions as a bargaining problem as realists do theories that account for thedistribution and intensity of national preferences (in Krasner rsquos much-citedapplication of bargaining theory the shape and location of the Pareto frontier)are distinct from theories of bargaining and collective action (which concernhow to ldquoget tordquo or ldquomove alongrdquo the Pareto frontier)125

Two implications follow from this dichotomy First each major internationalrelations theory paradigm enjoys a comparative advantage in explaining adifferent input into the bargaining game Liberal and liberal constructivisttheories focus on exogenous variation in underlying state preferences (notpolicies or strategies) hence the analyst concerned with the causes and conse-quences of variation in state preferences will nd liberal theory most usefulEpistemic theories highlight exogenous variation in collective beliefs that guideactors in their pursuit of goals hence the analyst concerned with the causesand consequences of changing conceptions of means-ends relations or theboundedness of rationality will nd such theories most useful Both realist andinstitutionalist theories by contrast take specic congurations of state pref-erences and beliefs as given and focus on the impact of exogenous variationin external systemic constraintsmdashresources in the case of realists and informa-tion in the case of institutionalists For the analyst interested in explainingvarying outcomes where preferences and beliefs are xed the interaction logicof realist or institutionalist theory may be more useful

The second implication of the two-stage or two-step method is that it sug-gests a more defensible and internally consistent approach to theory synthesisthan that commonly employed today Most leading contemporary scholarsmdashincluding Waltz Keohane and those whose work we analyze in this articlemdashrecommend that we synthesize theories by automatically considering realismrst (with preferences assumed to be invariant) and then introducing compet-ing theories of preference or belief change as needed to explain residualvariance ldquoLiberalism makes sense within the explanatory constraints

124 Legro ldquoInternational Cooperation Two-steprdquo and Moravcsik ldquoTaking Preferences Seriouslyrdquo125 Krasner ldquoGlobal Communications and National Powerrdquo


imposed by realismrdquo126 or ldquoWhen realist theories are found wanting we shouldsupplement them with new culturalist theoriesrdquo127 This conventional proce-dure we have argued in detail elsewhere lacks any coherent methodologicalor theoretical justication Methodologically it overtly introduces omitted vari-able bias by arbitrarily privileging realist explanations of any phenomenaunicausally explained by realist liberal and epistemic theories without evertesting the latter two explanations Theoretically the conventional approachcontradicts its own assumption of state rationality and xed preferences whichimplies precisely the opposite If preferences and beliefs vary across states andissues we must rst explain how they vary128

It is important to recognize of course that as one moves away from staticdecisions toward long-term change the explanatory domains suggested by thetwo-step methodmdashliberalism explains preferences epistemic theories explainbeliefs whereas realism or institutionalism help explain strategic interactionbased on resources and informationmdashbecome increasingly dependent on spe-cic empirical attributes of the situation A dynamic view opens up morecomplicated linkages among the various elements of rationalist theory FromOtto Hintze to Charles Tilly realists have made a case for preference andidentity formation via a particular subset of ldquosecond-image reversedrdquo argu-ments They maintain that conict and war have denitively shaped states andtheir desires such that the very identity and preferences of states adapt overtime129 Similarly Barry Posen and John Mearsheimer have argued that collec-

126 Robert O Keohane ldquoTheory of International Politics Structural Realism and Beyondrdquo inKeohane ed Neorealism and Its Critics (New York Columbia University Press 1990) p 192 Seealso Stephen Brooks ldquoDueling Realismsrdquo International Organization Vol 51 No 3 (Summer 1997)pp 471ndash472127 Michael C Desch ldquoCulture Clash Assessing the Importance of Ideas in Security StudiesrdquoInternational Security Vol 23 No 1 (Summer 1998) p 170128 The ldquotwo-steprdquo model to explain static bargaining outcomes follows in our view directlyfrom the rationality assumption shared by realism and most of its basic competitors As classicstudies of power by Robert Dahl and others taught us a generation ago it is impossible to modelstrategic interaction without rst determining preferences or beliefs (or both) independently of thestrategic circumstances Only where the pattern of preferences is consistent with the realist as-sumptions abovemdashpreferences are intense symmetrical and zero-summdashis it proper even to con-sider realist theory In any other casemdashsay a situation where preferences are compatible or wherethe collective action problem is informationalmdashrealism is not simply incorrect it is completelyinappropriate Thus in classical bargaining theory the locations of ideal points and outside options(preferences) are almost always relevant whereas linkage to threats and inducements are onlyrelevant under specic conditions This mechanism for theory synthesis is the most powerful basictool that rationalist social science theory has developed for this sort theory synthesis129 Otto Hintze The Historical Essays of Otto Hintze (New York Oxford University Press 1975)Charles Tilly ed The Formation of National States in Western Europe (Princeton NJ PrincetonUniversity Press 1975) Grieco ldquoRealist International Theoryrdquo and Layne ldquoKant or Cantrdquo pp 326ndash327


tive beliefsmdashfor example nationalismmdashcan be the product of internationalsecurity competition not simply of mass culture societal development ordomestic political manipulation130 Institutionalists both of a regime-theoreticaland constructivist variety argue that over time institutions can also shapepreferences and ideasmdashthough as yet there is no theory of this phenomenon131

This dynamic view of preference and belief construction may allow for muchmore complex claims about the relationship between realism and its competi-tors when studying long-term phenomenamdashanother implication of our refor-mulation concerning the empirical scope of different paradigms

Still by clearly specifying the assumptions about state preferences involvedour reformulation of realism encourages acceptance of the two-stage or two-step synthesis as a rst-cut explanation of discrete episodes of state behaviorThis would we believe permit realists who seek to incorporate domesticfactors to draw more explicitly on vibrant bodies of relevant nonrealist theorysuch as the literature on the democratic peace economic interdependenceideas in foreign policy and credible commitments Conversely a clearlydened realist theory about the role of material resources in shaping theoutcome of interstate conict offers a salutary correction to those liberalepistemic and institutionalist theories that ignore or attempt to implicitlysmuggle power into their analysis

Conclusion

Perhaps the most useful way to judge the power of a social scientic paradigmis by examining what it is able to exclude By this standard the realist para-digm is degenerating Its conceptual foundations are being ldquostretchedrdquo beyondall recognition or utility132 There exists no set of shared nontrivial assumptionsthat can distinguish the arguments shared by realists today Instead of chal-lenging competing liberal epistemic and institutional theories realists nowregularly seek to subsume their causal mechanisms Realism has become littlemore than a generic commitment to the assumption of rational state behaviorOne result is ad hoc appeals to exogenous variation in national preferencesbeliefs and international institutions Others to be sure elaborate more de-

130 Barry R Posen ldquoNationalism the Mass Army and Military Powerrdquo International Security Vol18 No 2 (Fall 1993) pp 80ndash124 and Mearsheimer ldquoBack to the Futurerdquo131 Alexander Wendt ldquoAnarchy Is What You Make of It The Social Construction of PowerPoliticsrdquo International Organization Vol 46 No 2 (Spring 1992) pp 391ndash425132 On ldquoconceptual stretchingrdquo see Sartori ldquoConcept Misinformation in Comparative Politicsrdquop 970


tailed midrange causal propositions about the causes and consequences of suchvariation but the explicitness of these arguments serves only to highlight theirliberal institutional or epistemic provenance From the perspective of therealist paradigm with which we began this article we ask ldquoIs anybody still arealistrdquo From the perspective of minimal realism the question becomes ldquoIseverybody now a realistrdquo Either way realism is in need of reformulation

The tendency to label nearly all rationalist explanations of state behaviorldquorealistrdquo misstates the broader signicance of the empirical research that self-styled realists have recently conducted Its real signicance lies not in therevitalization of core realist premises to which its connection is tenuous at bestIt lies instead in the empirical validation of assumptions about world politicsthat realists traditionally reject The mislabeling of realism has obscured themajor achievement of this research in the 1990s namely to demonstrate inimportant areas of security studies the explanatory power of liberal epistemicand institutionalist theories Here many of the realists considered above aswell as critics like Vasquezmdashall of whom explicitly defend adherence to real-ism despite anomalies because there appears to exist no alternative para-digmmdashunderstate the problem133 The real problem is not simply the use ofad hoc arguments to patch anomalies but the systematic use of argumentsfrom existing alternative paradigms

Instead of acknowledging this trend recent realist writings defend it byinviting us to return to the early 1940smdasha period in which realists such as EHCarr convinced scholars that the central debate in international relations theoryshould be between ldquorealistsrdquo who believe in rationality prudence and theimportance of national self-interest and ldquoidealistsrdquo who believe in the uniformharmony of state interests the power of altruistic motivations or the possibil-ity of world government Whether this dichotomy was a useful guide ftyyears ago remains an open question Its unsuitability today should be obviousto all These two categories are too vague too broad too open-ended toonormative and too dismissive of contemporary nonrealist theory to be of muchuse as a guide to social scientic theory and research134 The major develop-ment in international relations theory over the past three decades is insteadthe emergence and rm establishment of more subtly differentiated rationalist

133 Vasquez ldquoThe Realist Paradigmrdquo pp 909ndash911134 Some constructivists seem also to encourage the use of this dichotomy We have not howeverconsidered a constructivist ldquotheoryrdquo here because we take seriously those who warn that ldquocon-structivismrdquomdashlike ldquomaterialismrdquo ldquorationalismrdquo and other such broad categories of social theoryndashdoes not dene a discrete international relations paradigm or theory It should not therefore be


theoriesmdashvariants of liberal epistemic and institutionalist theories These arepotent competitors to realist claims and should be recognized as such Anycategorization of international relations theories that fails to accord these acentral and distinct place is profoundly misleading

One corrective to the degeneration of contemporary realism would be ofcourse simply to jettison the term altogether We believe it is too soon tocontemplate such a radical solution It would be preferable for realists and theirinterlocutors to observe greater precision in stating and applying its premisesA commitment to ldquorealismrdquo should signal far more than a belief in staterationality and international anarchy It should mark a commitment to a par-ticular rationalist theory of state behavior in anarchy one stressing the resolu-tion of international conict through the application of material powercapabilities The true role of such capabilities can be appreciated only throughconceptual clarity not conceptual stretching Acceptance of our tripartite refor-mulation of realism would provide theoretical foundations clearly distinctfrom other rationalist theories generate crisper empirical predictions andcontribute to more rigorous multicausal syntheses Such a coherent and distinctrealist paradigm would be t to assume its rightful role in the study of worldpolitics

employed as a counterpart to realism liberalism institutionalism or epistemic theory Construc-tivist arguments might be found in any of these categories A realist versus idealistconstructionistdichotomy would thus be unhelpful See Wendt ldquoSocial Theory and International Politicsrdquo andMartha Finnemore and Kathryn Sikkink ldquoInternational Norm Dynamics and Political ChangerdquoInternational Organization Vol 52 No 2 (Spring 1998) pp 890 909ndash912


The central problem is instead that the theoretical core of the realist approachhas been undermined by its own defendersmdashin particular so-called defensiveand neoclassical realistsmdashwho seek to address anomalies by recasting realismin forms that are theoretically less determinate less coherent and less distinc-tive to realism Realists like EH Carr Hans Morgenthau and Kenneth Waltzsought to highlight the manipulation accumulation and balancing of powerby sober unsentimental statesmen focusing above all on the limits imposedon states by the international distribution of material resources They viewedrealism as the bulwark against claims about the autonomous inuence ofdemocracy ideology economic integration law and institutions on worldpolitics Many recent realists by contrast seek to redress empirical anomaliesparticularly in Waltzrsquos neorealism by subsuming these traditional counterar-guments The result is that many realists now advance the very assumptionsand causal claims in opposition to which they traditionally and still claim todene themselves

This expansion would be unproblematic even praiseworthy if it took placeon the basis of the further elaboration of an unchanging set of core realistpremises It would be quite an intellectual coup for realists to demonstratemdashasrealists from Thucydides through Machiavelli and Hobbes to Morgenthausought to domdashthat the impact of ideas domestic institutions economic inter-dependence and international institutions actually reects the exogenousdistribution and manipulation of interstate power capabilities Some contem-porary realists do continue to cultivate such arguments yet such efforts appeartoday more like exceptions to the rule Many among the most prominent andthoughtful contemporary realists invoke instead variation in other exogenousinuences on state behaviormdashstate preferences beliefs and international insti-tutionsmdashto trump the direct and indirect effects of material power Such factorsare consistently treated as more important than power We term such an ap-proach ldquominimal realismrdquo because it retains only two core assumptionsmdashlittlemore than anarchy and rationalitymdashneither of which is distinctively realist By

Benjamin Frankel ed Realism Restatements and Renewal (London Frank Cass 1996) pp xindashxiiFor rejoinders see Kenneth N Waltz ldquoEvaluating Theoriesrdquo American Political Science Review Vol91 No 4 (December 1997) pp 913ndash918 Thomas J Christensen and Jack Snyder ldquoProgressiveResearch and Degenerative Alliancesrdquo American Political Science Review Vol 91 No 4 (December1997) pp 899ndash912 Colin Elman and Miriam Fendius Elman ldquoCorrespondence History vs Neo-realism A Second Lookrdquo International Security Vol 20 No 1 (Summer 1995) pp 182ndash193 Elmanand Elman ldquoLakatos and Neorealism A Reply to Vasquezrdquo American Political Science Review Vol91 No 4 (December 1997) pp 923ndash926 Randall L Schweller ldquoNew Realist Research on AlliancesRening not Refuting Waltzrsquos Balancing Propositionrdquo American Political Science Review Vol 91No 4 (December 1997) pp 927ndash930 and Stephen M Walt ldquoThe Progressive Power of RealismrdquoAmerican Political Science Review Vol 91 No 4 (December 1997) pp 931ndash935















































































































































































































































Conclusion


























































































































































































































































Conclusion




















































































































































































































































Conclusion















































































































































































































































Conclusion









































































































































































































































Conclusion



































































































































































































































Conclusion































































































































































































































Conclusion

























































































































































































































Conclusion





















































































































































































































Conclusion
















































































































































































































Conclusion












































































































































































































Conclusion







































































































































































































Conclusion


































































































































































































Conclusion



























































































































































































Conclusion






















































































































































































Conclusion


















































































































































































Conclusion












































































































































































Conclusion







































































































































































Conclusion


































































































































































Conclusion






























































































































































Conclusion























































































































































Conclusion



















































































































































Conclusion














































































































































Conclusion








































































































































Conclusion



































































































































Conclusion






























































































































Conclusion

























































































































Conclusion




















































































































Conclusion














































































































Conclusion









































































































Conclusion




































































































Conclusion































































































Conclusion



























































































Conclusion






















































































Conclusion

















































































Conclusion













































































Conclusion








































































Conclusion




































































Conclusion
































































Conclusion




























































Conclusion






















































Conclusion


















































Conclusion











































Conclusion




































Conclusion































Conclusion

























Conclusion




















Conclusion
















Conclusion


























Documents

moravcik, kritik