The MIT Press Is Collaborating With JSTOR To Digitize, Preserve and Extend Access To International Security

Security Seeking under Anarchy: Defensive Realism Revisited
Author(s): Jeffrey W. Taliaferro

Source: International Security, Vol. 25, No. 3 (Winter, 2000-2001), pp. 128-161
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/2626708
Accessed: 26-08-2015 01:47 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to International Security.
http://www.jstor.org
This content downloaded from 141.161.91.14 on Wed, 26 Aug 2015 01:47:20 UTC
All use subject to JSTOR Terms and Conditions
SecuritySeeking under Jeffrey
W.Taliaferro
Anarchy
Defensive Realism Revisited
Does theinternational
systemprovide incentivesforexpansion? If so, should the United States seek
to guaranteeits long-termsecuritythrougha grand strategyof preponderance
(or primacy)and pursue opportunitiesto weaken potentialgreatpower com-
petitors,such as China? Alternatively,
does the internationalsystemprovide
moredisincentivesthanincentivesforaggression?Ifthisis thecase, should the
UnitedStatesseek to guaranteeits long-termsecuritythrougha grand strategy
of selectiveengagement?Two strandsof contemporaryrealismprovide differ-
ent answers to these questions.1
Offensiverealism holds thatanarchy-the absence of a worldwide govern-
ment or universal sovereign-provides strongincentivesfor expansion.2All
states striveto maximize theirpower relativeto otherstatesbecause only the
most powerfulstates can guarantee theirsurvival. They pursue expansionist
policies when and where thebenefitsof doing so outweighthecosts.Statesun-
der anarchyface the ever-presentthreatthatotherstateswill use forceto harm
or conquer them.This compels statesto improvetheirrelativepower positions
W Taliaferro
Jeffrey is AssistantProfessor
ofPoliticalScienceat TuftsUniversity.
I wishto thankDale Copeland,BernardFinel,BenjaminFrankel,

BenjaminMiller,Jennifer Ster-
ling-Folker,and the anonymousreviewersforInternational
Security forcommentson various
I am responsible
drafts. foranyremaining
errorsoromissions.
Earlierversionsofthisarticlewere
presented at the1999annualmeetings
oftheInternational
StudiesAssociationand theAmerican
PoliticalScienceAssociation.
1. For an overview of competingpost-Cold War U.S. grand strategiesderived fromoffensivereal-

ism and defensiverealism,see BarryR. Posen and Andrew L. Ross, "Competing Visions forU.S.
Grand Strategy,"International Vol. 21, No. 3 (Winter1996/97),pp. 5-53.
Secuirity,
2. The terms"aggressiverealism"(or offensiverealism)and "defensiverealism"originatedin Jack
Snyder,MythsofEmpire:DomesticPoliticsand International Ambition(Ithaca, N.Y.: Cornell Univer-
sityPress, 1991), pp. 11-12. Overviews of the offensiverealism-defensiverealism debate include
Sean M. Lynn-Jonesand Steven E. Miller,"Preface,"in Michael E. Brown,Owen M. Cote, Lynn-
Jones,and Miller,eds., ThePerilsofAnarchy:Contemporary Realismand InternationalSecurity(Cam-
bridge,Mass.: MIT Press,1995),pp. ix-xii;BenjaminFrankel,"RestatingtheRealistCase: An Intro-
duction," SecurityStudies,Vol. 5, No. 3 (Spring 1996), pp. xiv-xx; and Sean M. Lynn-Jones,
"Realism and America's Rise: A Review Essay," International Security,Vol. 23, No. 2 (Fall 1998),
pp. 157-182.
3. Examples of offensiverealism include JohnJ.Mearsheimer,"Back to the Future:Instabilityin
Europe after the Cold War," InternationalSecurity,Vol. 15, No. 1 (Summer 1990), pp. 5-56;
Mearsheimer,"The False Promiseof InternationalInstitutions,"International Vol. 19,No. 3
Secuirity,
(Winter1994/95),pp. 5-49, especiallypp. 10-15; Fareed Zakaria,FromWealthtoPozwer: TheUnulsual
Ititerniatioznal Vol. 25, No. 3 (Winter2000/01), pp. 128-161

Secturity,
? 2000 by the Presidentand Fellows of Harvard College and the Massachusetts Instituteof Technology.
128
SecuritySeekingunderAnarchy| 129
througharms buildups, unilateral diplomacy,mercantile(or even autarkic)

foreigneconomic policies, and opportunisticexpansion.3
Defensiverealismholds thattheinternationalsystemprovides incentivesfor
expansion only under certainconditions.Under anarchy,many of the means a
state uses to increase its securitydecrease the securityof otherstates.This se-
curitydilemma causes states to worryabout one another's futureintentions
and relativepower. Pairs of states may pursue purelysecurity-seeking strate-
gies,but inadvertently generatespirals of mutual hostilityor conflict.Statesof-
ten, although not always, pursue expansionistpolicies because theirleaders
mistakenlybelieve thataggressionis the only way to make theirstatessecure.
Defensive realism predictsgreatervariation in internationallydriven expan-
sion and suggests thatstatesought to generallypursue moderatestrategiesas
the best routeto security.Under most circumstances,the strongerstatesin the
internationalsystemshould pursue military, diplomatic,and foreigneconomic
policies that communicaterestraint.4
Defensiverealismhas recentlycome under attackfromcriticsof realismand
even fromfellow realists.Criticsof realism,such as Andrew Moravcsik and
Jeffrey Legro,faultvarious defensiverealisttheoriesforpositinga role fordo-
mesticpolitics,elitebeliefsystemsand misperceptions,and internationalinsti-
tutions. By including such variables in their theories, the critics argue,
defensiverealistseffectively repudiate the core assumptions of political real-
ism.5Offensiverealists,such as Fareed Zakaria and Randall Schweller,charge
Origins of America'sWorldRole (Princeton,N.J.: Princeton UniversityPress, 1998); Randall L.

Schweller,"Bandwagoning forProfit:BringingtheRevisionistStateBack In," International Security,
Vol. 19,No. 1 (Summer1994),pp. 72-107; Schweller,DeadlyImbalances:Tripolarity and Hitler'sStrat-
egyof WorldConquiest(New York:Columbia UniversityPress, 1997); Samuel P. Huntington,"Why
InternationalPrimacyMatters,"International Vol. 17, No. 4 (Spring 1993), pp. 68-83; and
Security,
EricJ.Labs, "Beyond Victory:OffensiveRealism and theExpansion ofWarAims," SecurityStudies,
Vol. 6, No. 4 (Summer 1997), pp. 1-49.
4. RobertJervis,"Cooperation under theSecurityDilemma," WorldPolitics,Vol. 30, No. 2 (January
1978), pp. 167-214,provides the theoreticalfoundationsfordefensiverealism.Examples of defen-
sive realism include Stephen M. Walt, The Originsof Alliances(Ithaca, N.Y: Cornell University
Press,1987); Walt,Revolution and War(Ithaca,N.Y: Cornell UniversityPress,1995); BarryR. Posen,
The SourcesofMilitaryDoctrine:France,Britain,and Germanybetweenthe WorldWars(Ithaca, N.Y:
Cornell UniversityPress,1984); Thomas J.Christensenand JackSnyder,"Chain Gangs and Passed
Bucks: PredictingAlliance Patternsin Multipolarity,"International Organization,Vol. 44, No. 2
(Spring 1990), pp. 137-168; Christensen,"Perceptionsand Alliances in Europe, 1860-1940,"Inter-
nationalOrganization,Vol. 51, No. 1 (Winter1997), pp. 65-98; Stephen Van Evera, "Offense,De-
fense,and theCauses of War,"International Vol. 22, No. 4 (Spring1998),pp. 5-43; and Van
Security,
Evera, Causes of War:Pozwer and theRootsofConflict(Ithaca, N.Y: Cornell UniversityPress, 1999).
5. JeffreyW. Legro and Andrew Moravcsik,"Is AnybodyStilla Realist?"International Security,Vol.
24, No. 2 (Fall 1999), pp. 5-55.
International
Security25:3 | 130
that defensiverealism cannot explain state expansion because it argues that

thereare neverinternationalincentivesforsuch behavior.6
I argue thatthedebate betweendefensiverealismand offensiverealismover
the implicationsof anarchyand the need to clarifydefensiverealism's auxil-
iaryassumptionsdeserve attentionforthreereasons. First,the outcome of this
theoreticaldebate has broad policy implications.Defensive realism suggests
that under certainconditions,pairs of nondemocraticstates can avoid war,
statescan engage in mutuallybeneficialcooperationwithoutthe assistance of
internationalinstitutions,and norms proscribingthe developmentand use of
weapons of mass destructionare largelyepiphenomenal.7In addition, offen-
sive realism and defensive realism generate radically differentprescriptions
formilitarydoctrine,foreigneconomic policy,militaryintervention, and crisis
management.8
Second, debates withinparticularresearchtraditions,not debates between
them,are more likelyto generatetheoreticalprogressin the study of interna-
tional politics.By developing and testingtheoriesderived fromthe same core
assumptions, researcherscan more easily identifycompeting hypotheses,
refinescope conditionsfortheories,and uncover new facts.Arguably,thisis a
more productivestrategyforthe accumulationof knowledge than the current
tendencyamong some scholarsto brand entireresearchprogramsas "degener-
ative."9 As Robert Jervisobserves: "Programs-and, even more, their first
6. Randall L. Schweller,"Neorealism's StatusQuo Bias: What SecurityDilemma?" SecurityStudies,

Vol. 5, No. 3 (Spring1996),pp. 90-121; and Fareed Zakaria, "Realism and Domestic Politics:A Re-
view Essay," International Security,Vol. 17, No. 1 (Summer 1992), pp. 177-198.
7. See Charles L. Glaser,"Realistsas Optimists:Cooperation as Self-Help,"SecurityStudies,Vol. 5,
No. 3 (Spring 1996),pp. 122-166; RobertJervis,"Realism,Neoliberalism,and Cooperation,"Inter-
nationalSecuirity,Vol. 24, No. 1 (Summer 1999),pp. 42-63; and KennethN. Waltz,"StructuralReal-
ism afterthe Cold War,"International Security,Vol. 25, No. 1 (Summer 2000), pp. 5-41.
8. For examples of competingforeignpolicy prescriptionsdrawn fromoffensiverealismand de-
fensiverealism,see Mearsheimer,"Back to the Future,"especially pp. 36-40, 54-56; Stephen M.
Walt,"The Case forFiniteContainment:AnalyzingU.S. Grand Strategy," International Vol.
Security,
14, No. 1 (Summer 1989), pp. 5-49; ChristopherLayne, "From Preponderance to OffshoreBal-
ancing: America's Future Grand Strategy,"International Security,Vol. 22, No. 1 (Summer 1997),
pp. 86-124; RobertJ.Art,"Geopolitics Updated: The Strategyof Selective Engagement,"Interna-
tionalSecurity,Vol. 23, No. 3 (Winter1998/99), pp. 79-113; Eugene Gholz, Daryl G. Press, and
Harvey M. Sapolsky,"Come Home, America:The Strategyof Restraintin the Face of Temptation,"
International Security,Vol. 21, No. 4 (Spring 1997), pp. 4-48; Thomas J.Christensen,"China, the
U.S.-JapanAlliance, and the SecurityDilemma in East Asia," International Vol. 23, No. 4
Security,
(Spring 1999), pp. 49-80; and Michael C. Desch, "Why Realists Disagree about the Third World
(and Why They Shouldn't)," Secuirity Studies,Vol. 5, No. 3 (Spring 1996), pp. 358-384.
9. See Legro and Moravcsik,"Is AnybodyStilla Realist?" especially pp. 9, 18-22; Paul Schroeder,
"Historical Realityvs. Neo-realistTheory,"International Security,Vol. 20, No. 1 (Summer 1995),
pp. 108-148; and JohnA. Vasquez, "The Realist Paradigm and Degenerative versus Progressive
ResearchPrograms,"AmericanPoliticalScienceReviezv, Vol. 91,No. 4 (December 1997),pp. 899-912.
cousins, paradigms-are notoriouslydifficultto confirmor disconfirm.Not

only do theyshape what countsas a factat all, but also thereare so many steps
between assumptionsand outlooks on the one hand and empiricalfindingson
the otherthatneitherin social nor in naturalsciences can the evidence ever be
unambiguous."10
Third,regardlessof whetherrealismis the dominanttheoreticalapproach in
internationalrelations,it remainsthebete noireof everynonrealistapproach.11
Proponentsof neoliberalinstitutionalism, various culturaltheories,democratic
peace theories,and constructivismall begin with the supposition thatrealism
is an extremelylimited,ifnot completelybankrupt,body of theory.12In thein-
terestof scholarlydialogue, it is importantto clarifythe predictionsof particu-
lar realisttheories.
I argue thatdefensiverealismproceeds fromfourauxiliaryassumptionsthat
specify how structuralvariables translateinto internationaloutcomes and
states' foreignpolicies. First,the securitydilemma is an intractablefeatureof
anarchy.Second, structuralmodifiers-such as the offense-defensebalance,
geographic proximity,and access to raw materials-influence the severityof
the securitydilemma between particularstates. Third,materialpower drives
states' foreignpolicies throughthe medium of leaders' calculations and per-
ceptions.Finally,domesticpoliticscan limitthe efficiency of a state's response
to the externalenvironment.
The firstsectionof thisarticlediscusses thedebates withincontemporaryre-
alism, drawing a distinction between theories of international politics
(neorealism) and theories of foreignpolicy (neoclassical realism), both of
10. RobertJervis,"Realism and theStudyofWorldPolitics,"International Organization,Vol. 52, No.

4 (Autumn 1998), pp. 971-991, at p. 975.
11. See Mike Winnerstig,"Dancing theMaster's Waltz:The Hidden Influenceof 20 Years of Theory
ofInternational Politics,"paper presentedat theannual meetingof theInternationalStudies Associ-
ation,Washington,D.C., February16-20, 1999. Based on an investigationof thereferencesmade to
sixteenprominentscholars in the Social Science CitationIndex and a quantitativeand qualitative
StudiesQuarterlyand International
analysis of all articlesin International Securitybetween 1990 and
1997,Winnerstigconcludes thatcontemporaryrealismis farfromthedominantapproach in thein-
ternationalrelationsfield.
12. See, for example, Richard Rosecrance and ArthurA. Stein, "Beyond Realism: The Study of
Grand Strategy,"in Rosecrance and Stein,eds., The DomesticBases ofGrandStrategy(Ithaca, N.Y.:
Cornell University Press, 1993); Robert 0. Keohane and Lisa L. Martin, "The Promise of
Institutionalist Theory,"International Vol. 20, No. 1 (Summer 1995), pp. 39-51; Alexander
Security,
Wendt,"AnarchyIs What States Make of It: The Social Constructionof Power Politics,"Interna-
tionalOrganization, Vol. 46, No. 2 (Spring 1992), pp. 391-425; Jeffrey
W. Legro, Cooperation uinder
Fire:Anglo-German RestraintduringWorldWarII (Ithaca,N.Y.: Cornell UniversityPress,1995); Peter
J.Katzenstein,ed., The CultureofNationalSecurity(New York:Columbia UniversityPress, 1996);
and Alastair lain Johnston,CulturalRealism:StrategicCuilture and GrandStrategyin ChineseHistory
(Princeton,N.J.:PrincetonUniversityPress, 1995).
International
Security25:3 | 132
whichhave defensiveand offensivevariants.I divide realismalong theselines

because doing so allows us to distinguishbetween debates over the implica-
tions of anarchy and the empirical range of particulartheories.The second
section examines the four assumptions underlying the defensive variants
of neorealismand neoclassical realism. The thirdsection responds to several
criticismsthatrealistsand nonrealistsraise about those assumptions and the
explanatorypower of defensiverealism.The conclusiondiscusses the implica-
tions of the debate forU.S. grand strategyand offerssome suggestionsforfu-
tureresearch.
Intrarealist
Debates
Realist theoriesshare certaincore assumptions,but thereare two crosscutting

divisions withincontemporaryrealism.13 First,neorealismseeks to explain in-
ternationaloutcomes,such as the likelihoodof major war,the prospectsforin-
ternational cooperation, and aggregate alliance patterns among states.
Neoclassical realism, on the other hand, seeks to explain the foreignpolicy
strategiesof individual states.Second, realistsdisagree about thelogical impli-
cations of anarchy.This is the crux of the debate between offensiverealism
and defensive realism. Below I discuss the four categories of realist theory
and how theoffensive-defensive dichotomytranscendsthe distinctionbetween
neorealismand neoclassical realism.
NEOREALISM AND NEOCLASSICAL REALISM

Neorealismand neoclassical realismdifferbased on the phenomenaeach seeks
to explain, or the dependent variable. In this sense, neorealismand neoclassi-
cal realism are complementary;each purportsto explain phenomena thatthe
otherdoes not.14
13. For statementsof realism's core assumptions,see Frankel,"Restatingthe Realist Case," pp.
xiv-xviii;RobertGilpin, "No One Loves a Political Realist,"Secuirity Studies,Vol. 5, No. 3 (Spring
1996),pp. 3-26, at pp. 6-8; Colin Elman, "Horses forCourses: Why NotNeorealistTheoriesof For-
eign Policy?" SecurityStuidies,Vol. 6, No. 1 (Autumn 1996), pp. 7-53, at pp. 18-21; and Randall L.
Schweller,"New RealistResearchon Alliances: Refining,Not Refuting,Waltz's BalancingProposi-
tion,"AmericanPoliticalScienceReview,Vol. 91, No. 4 (December 1997), pp. 927-930.
14. Schweller,"New RealistResearchon Alliances," p. 329, and Charles L. Glaser,"The Necessary
and Natural Evolution of StructuralRealism," unpublished manuscript,Universityof Chicago,
May 1999, make the same point. Gideon Rose, who originatedthe term"neoclassical realism,"
however,sees neoclassical realism,offensiverealism,and defensiverealismas competingtheories
of foreignpolicy.See Rose, "Neoclassical Realism and Theories of ForeignPolicy," WorldPolitics,
Vol. 51, No. 1 (October 1998), pp. 144-172.
Neorealismis a body of internationalrelationstheorythatbuilds upon a few

assumptions about the internationalsystemand the units thatit comprises.15
Neorealist theoriesseek to explain internationaloutcomes-phenomena that
resultfromthe interactionof two or more actors in the internationalsystem.
For example, explainingthe likelihoodof major or hegemonicwar fallswithin
the purview of neorealism.16Other examples of internationaloutcomes in-
clude internationalcooperation,arms races, crisisbargaining,aggregatealign-
mentpatterns,and thewar pronenessof theinternationalsystem.In short,one
cannot attributethese phenomena to the behavior of any one state.
Neorealism cannotmake predictionsabout theforeignpolicybehaviorof in-
dividual states. It cannot,for example, answer the followingquestion: What
will a particularstate faced with these circumstanceslikely do? As Kenneth
Waltz observes, a strictlysystemictheory"can tell us what pressuresare ex-
ertedand what possibilitiesare posed by systemsof different structure,but it
cannottell us just how, and how effectively,the unitsof a systemwill respond
to those pressuresand possibilities.""7Waltz's balance-of-powertheoryis the
most prominentexample of the neorealistapproach. Neorealism also encom-
passes hegemonictheoriesof war and change,power transitionand long cycle
theories,and systemstheory.18
Neoclassical realismseeks to explain why different states or even the same
state at differenttimes pursues particular strategies in the international
arena.19It generatesprobabilisticpredictionsabout how individual states re-
15. For a contrastingview, see Elman, "Horses for Courses," especially pp. 21-47; and Colin
Elman, "Cause, Effect,and Consistency:A Response to KennethWaltz," Secuirity Studies,Vol. 6,
No. 1 (Autumn 1996), pp. 58-61. AlthoughI agree with Elman's argument-namely,thatthereis
no epistemologicalor methodologicalreason why one cannot derive testablehypothesesabout
states' foreignpolicies fromWaltz's balance-of-powertheory-I nonetheless reserve the term
"neorealism" fortheoriesof internationalpolitics.
16. I draw upon the definitionsofferedby Dale C. Copeland, TheOriginsofMajor War:Hegemonic
Rivalryand theFear of Decline (Ithaca, N.Y.: Cornell UniversityPress, forthcoming), p. 3; Robert
Gilpin, Warand Changein WorldPolitics(New York: Cambridge UniversityPress, 1981), pp. 15,
197-198; and JackS. Levy,"The Causes of War: A Review of Theories and Evidence," in Philip E.
Tetlock,JoL. Husbands, RobertJervis,Paul C. Stern,and Charles Tilly,eds., Behavior, Society,and
NuclearWar,Vol. 1 (New York:OxfordUniversityPress, 1989), p. 307, n. 73.
17. KennethN. Waltz,TheoryofInternational Politics(Reading,Mass.: Addison-Wesley,1979),p. 73.
See also Waltz, "InternationalPolitics Is Not Foreign Policy,"SecurityStudies,Vol. 6, No. 1 (Au-
tumn 1996), pp. 54-57.
18. See, forexample,Gilpin, Warand Changein WorldPolitics;A.F.K. Organski,WorldPolitics(New
York:Knopf,1968);JacekKugler and A.F.K. Organski,"The Power Transition:A Retrospectiveand
ProspectiveEvaluation," in Manus I. Midlarsky,ed., HandbookofWarStudies(Ann Arbor:Univer-
sityof Michigan Press, 1996), pp. 1.71-194;George Modelski, LongCyclesin WorldPolitics(Seattle:
Universityof WashingtonPress,1987); and RobertJervis,SystemsEffects: Complexity in Politicaland
Social Life(Princeton,N.J.:PrincetonUniversityPress, 1997).
19. Elman, "Horses forCourses," p. 12. See also Zakaria, FromWealthto Pozwer, pp. 14-18.
International
Security25:3 | 134
spond to systemicimperatives.Phenomena such as individual states' grand

strategies,militarydoctrines,foreigneconomic policy, alliance preferences,
and crisisbehavior fall withinneoclassical realism's purview.Neoclassical re-
alism cannot predict the aggregate internationalconsequences of individual
states' strategies.
While building on Waltz's assumptions about anarchy,neoclassical realists
explicitlyrejectthe injunctionthat theoriesought not to include explanatory
variables at different
levels of analysis.20Gideon Rose notes thata state's rela-
tive materialcapabilitiesset the parametersof its foreignpolicy.He observes,
however,"thatthe impact of such power capabilitieson foreignpolicy is indi-
rectand complex,because systemicpressuresmust be translatedthroughin-
terveningvariables at the unit level."'21
OFFENSIVE REALISM VERSUS DEFENSIVE REALISM

Whereas neorealism and neoclassical realism seek to explain differentphe-
nomena, the divide between offensiverealism and defensive realism repre-
sentsa fundamentaldivergenceon the implicationsof anarchy.Thus offensive
realismand defensiverealismare theoreticalcompetitorsbecause theygener-
ate differentpredictionsand policyprescriptions.22 This division subsumes the
neorealist-neoclassicaldichotomy.23 Table 1 illustrateshow the offensivereal-
ism-defensiverealism debate cuts across the divide between neorealismand
neoclassical realism. This two-partclassificationscheme refersto particular
theories,not particulartheorists.Specifictheoriesfallintothese categories,but
scholars may work in more than one category.
AuxiliaryAssumptions
ofDefensiveRealism
Four auxiliaryassumptionsdefinedefensiverealism.The firsttwo specifythe

incentivesforinterstateconflictor cooperation.The lattertwo specifythe links
20. Waltz, TheoryofInternational Politics,p. 75.

21. Rose, "Neoclassical Realism and Theories of ForeignPolicy,"p. 146.
22. Other scholars make thisdistinction,but employ idiosyncraticterms.Stephen G. Brooks sub-
stitutesthelabels "neorealism"foroffensiverealismand "postclassicalrealism"fordefensivereal-
ism; Robert G. Kaufmann substitutesthe term "pessimistic structuralrealism" for offensive
realism and "optimisticstructuralrealism" fordefensiverealism;and Charles L. Glaser uses the
term"contingentrealism" instead of defensiverealism.See Brooks,"Dueling Realisms," Interna-
tionalOrganization,Vol. 51, No. 3 (Summer 1997), pp. 445-477; Kaufmann,"A Two-Level Interac-
tion:Structure,Stable Liberal Democracy,and U.S. Grand Strategy,"SecurityStudies,Vol. 3, No. 4
(Summer 1994), p. 683ff.;and Glaser,"Realists as Optimists,"pp. 52-54.
23. Dale C. Copeland, "Neorealism and the Myth of Bipolar Stability:Toward a New Dynamic
Theoryof Major War," SecurityStudies,Vol. 5, No. 3 (Spring 1996), pp. 29-89.
SecuritySeekingunderAnarchy 135
Table 1. Categoriesof Contemporary

Realism.
PhenomenaTo Be
Explaineda Assumptionsabout Anarchy
Defensiverealism Offensiverealism
The international
system The international
system
providesincentivesfor always providesincentives
expansiononlyunder forexpansion.
certainconditions.
Neorealism Balance-of-power theory Hegemonic theory of war
(KennethWaltz) (RobertGilpin)
Theories that seek to Dynamic differentials Power transitiontheory
explaininternational theory (A.F.K.Organskiand
outcomes-forexample, (Dale Copeland) JacekKugler)
the likelihoodof great
power war, the durability Great power cooperation Balance-of-intereststheoryb
of alliances,or the theories (RandallSchweller)
likelihoodof international (RobertJervis,Charles
cooperation Glaser,and Benjamin Theoryof greatpower
Miller) politics
(JohnMearsheimer)
Neoclassical realism Balance-of-threattheory State-centered realism
(StephenWalt) (FareedZakaria)
Theories that seek to Domestic mobilization Theory of war aims
explainthe external theory (EricLabs)
behaviorof individual (Thomas Christensen)
states-for example,
militarydoctrineforce Offense-defense theories Hegemonic theory of
posture,alliance (StephenVan Evera, foreignpolicy
preferences, foreign Thomas Christensenand (WilliamWohlforth)
economicpolicy,or the JackSnyder,and Charles
pursuitof accommodative Glaserand Chaim
or belligerentdiplomacy Kaufmann)
aThe distinction betweenneorealismand neoclassicalrealismis best understoodas a con-

tinuum,nota concretedivision.Severaltheoriesstaddlethe linebetweenthetwo because
theyseek to explainbothsystemicoutcomesand theforeignpolicybehaviorsof particu-
lar states. For example,RandallSchweller'sbalance-of-interests theory,Dale Copeland's
dynamicdifferentials theory,and JohnMearsheimer's theoryof greatpowerpoliticsgen-
eratetestablehypotheseson the likelihoodofmajorwarand thelikelydiplomaticand mil-
itarystrategiesof greatpowers.
bUnlikemostoffensiverealisttheories,Schweller'sbalance-of-interests theorydoes notas-
sume thatrelativepower maximization and aggressionare the logicalconsequences of
anarchy.His theorydraws a sharpdistinction betweenrevisionist and statusquo states.
He does notattribute or statusquo interestto anarchy,however.Status
states' revisionist
quo and revisionist interestsare unit-levelvariables.See RandallL. Schweller,DeadlyIm-
balances: Tripolarityand Hitler'sStrategy of World Conquest (New York:Columbia Univer-
sityPress, 1997),pp. 22-26.
International
Security25:3 | 136
between externalforcesand the actual foreignpolicy behavior of individual

states.
ASSUMPTION 1: THE INTRACTABILITY OF THE SECURITY DILEMMA

The securitydilemma is an intractablefeatureof the internationalsystem.24
Jervisdefines the securitydilemma as a situation "in which the means by
which a state triesto increase its securitydecreases the securityof others."25
Anarchyproduces uncertainty, and statescan never be certainof others'pres-
ent or futureintentionsor the relativedistributionof capabilities over time.
Anarchyinduces states to engage in self-helpbehavior.States react to reduc-
tionsin theirsecurityby takingsteps to increasetheirown security,thus miti-
gating the securitypolicies of others.
Why should the efforts of one stateto make itselfsecure cause otherstatesto
feelless so? Charles Glaser posits threeways throughwhich makingone's ad-
versaries insecure can prove self-defeating.First,even security-seekingpoli-
cies can set in motion a process that reduces the state's own military
capabilities-the abilityto performparticularmilitarymissions. Second, self-
help strategiesmay increase the value an adversaryplaces on expansion as a
means of self-defense,whichin turnmakes deterrenceharder.Third,bothmili-
tarybuildups and alliances can change the adversary'sbeliefsabout the state's
motives,thus convincingthe adversarythatthe state is inherentlymore dan-
gerous than previouslythought.An adversarymay conclude thata state har-
bors "greedy" motives-that is, a desire to expand for reasons other than
security.Armsbuildups may simplybe a waste of a state's finiteresources,be-
cause othersmay be able to meet or exceed its level of armament.In short,a
statethatinitiatesa militarybuildup to increaseits securitymay inadvertently
set in motion a chain of events thatleaves it less secure.26
ASSUMPTION 2: STRUCTURAL MODIFIERS AND THE SECURITY DILEMMA

The securitydilemma is inescapable, but it does not always generateintense
competitionand war. In addition to the gross distributionof power in the in-
24. Securitydilemmas existin any anarchicenvironment,not just in the internationalsystem.See

Barry R. Posen, "The SecurityDilemma in Ethnic Conflict,"in Michael E. Brown, ed., Ethnic
Conflictand International Security(Princeton,N.J.:PrincetonUniversityPress, 1993), pp. 103-124;
and Chaim Kaufmann,"Possible and Impossible Solutions to EthnicCivil Wars,"International Se-
curity,Vol. 20, No. 4 (Spring 1996), pp. 136-175.
25. Jervis,"Cooperation under the SecurityDilemma," p. 178.
26. Charles L. Glaser, "The SecurityDilemma Revisited," WorldPolitics,Vol. 50, No. 1 (October
1997), pp. 171-201.
ternationalsystem, other material factors,which I referto as "structural

modifiers,"may increaseor decrease the likelihoodof conflict.27 These include
the offense-defense balance in militarytechnology,geographic proximity,ac-
cess to raw materials,internationaleconomicpressure,regionalor dyadic mili-
tary balances, and the ease with which states can extractresources from
conquered territory.28
Defensive realistsassume thatstructuralmodifiershave a greaterinfluence
on the likelihood of internationalconflictor cooperationthan does the gross
distributionof power. The gross distributionof power refersto the relative
share of theinternationalsystem'smaterialcapabilitiesthateach statecontrols.
Polarity,or thenumberof greatpowers in the internationalsystem,is the most
common measure of the gross distributionof power. Structuralmodifiers,on
the otherhand, referto the relativedistributionof capabilitiesthatenable indi-
vidual statesto carryout particulardiplomaticand militarystrategies.This in
turninfluencesthe severityof the securitydilemma between particularstates
or in regional subsystems.Thus one may thinkof the structuralmodifiersas
mediatingthe effectsof systemicimperativeson the behavior of states.29
27. For a discussion of structuralmodifiers,see Glenn D. Snyder,"Process Variablesin Neorealist

Theory,"SecurityStudies,Vol. 5, No. 3 (Spring 1996),pp. 167-192,at pp. 168-171. Snyder's concept
of structuralmodifiersis similarto Stephen Van Evera's concept of the "fine-grainedstructureof
power" and BarryBuzan's concept of "interactioncapacity."See Van Evera, Causes of War:Pozver
and theRootsofConflict(Ithaca, N.Y.: Cornell UniversityPress, 1999), pp. 7-8; and Buzan, Charles
Jones,and RichardLittle,TheLogicofAnarchy:Neorealismto Structural Realism(New York:Colum-
bia UniversityPress, 1993), pp. 69-83.
28. For debates on the definitionand measurementof the offense-defense balance, see Jack S.
Levy,"The Offense/DefenseBalance in MilitaryTechnology:A Theoreticaland HistoricalAnaly-
sis," InternationalStudiesQuarterly,Vol. 28, No. 2 (June1984), pp. 219-238; Scott D. Sagan, "1914
Revisited:Allies, Offense,and Instability,"InterniationalSecurity,Vol. 11,No. 2 (Fall 1986),pp. 151-
176; Sean M. Lynn-Jones, "Offense-DefenseTheoryand Its Critics,"SecurityStudies,Vol. 4, No. 1
(Summer 1995),pp. 660-691; Van Evera, Causes of War,pp. 160-166; Charles L. Glaser and Chaim
Kaufmann,"What Is the Offense-DefenseBalance and Can We Measure It?" International Security,
Vol. 22, No. 4 (Spring1998),pp. 44-82; JamesW. Davis, Jr.,BernardI. Finel,Stacie E. Goddard, Ste-
phen Van Evera,and Charles L. Glaser and Chaim Kaufmann,"Correspondence:TakingOffenseat
Offense-DefenseTheory,"International Security,Vol. 23, No. 3 (Winter1998/99),pp. 179-206; Rich-
ard K. Betts,"Must War Find a Way? A Review Essay," International Security,Vol. 24, No. 2 (Fall
1999),pp. 166-198,at pp. 178-179; and Kier A. Lieber,"Grasping the TechnologicalPeace: The Of-
fense-DefenseBalance and InternationalSecurity,"Internationcal Security,Vol. 25, No. 1 (Summer
2000), pp. 71-104, at pp. 74-77.
29. On thispoint,my treatmentof structuralmodifiersdiffersfromGlenn Snyder's. Snyder con-
tends that structuralmodifiers"are roughly analogous macroeconomicinfluences,like interest
rates or governmentalregulation,on microeconomicrelationsbetween firms;theyaffectthe be-
havior of all actorsmore or less evenly,but theyare different in kind fromfactorslike thenumber
of actors(firms)or the distributionof power among them-variables which clearlydeterminethe
structureof the system(market)."Snyder,"Process Variables in Neorealist Theory,"p. 169.
International
Security25:3 | 138
Consider,forexample, offense-defense theoryand balance-of-threat theory.

It makes littlesense to speak of a systemwideoffense-defense balance in mili-
tarytechnology.The possession of particularmilitarytechnologiesand weap-
ons systemsinfluencesthe relativeease with which a state can attackor hold
territory.The objectiveoffense-defense balance affectsthe strategiesof individ-
ual states and the interactionbetween pairs of states; it does not change the
gross distributionof power in the internationalsystem.30 Similarly,balance-of-
threattheorydoes not posit that states always balance against the greatest
threatin the internationalsystem.Rathertheygenerallybalance against states
thatpose an immediatethreatto theirsurvival.31
Defensive realism, in both its neorealist and neoclassical realist variants,
challengesnotionsthatthesecuritydilemma always generatesintenseconflict.
In thisrespect,defensiverealismcorrectsdeductive flawsboth in Waltz's core
model and in offensiverealism.Waltz holds thatanarchyand the need forsur-
vival oftenforcestates to forgomutually beneficialcooperation. At a mini-
mum,cooperationis difficult because statesare sensitiveto how it affectstheir
currentand futurerelativecapabilities.32Cooperation oftenproves to be im-
possible, particularlyin the securityarena,because stateshave everyincentive
to maintainan advantage over theircompetitors.33 Some offensiverealistsgo
furtherin arguingthatcooperationcan put a state's survivalin jeopardy.John
Mearsheimerargues thatanarchyleaves littleroom fortrustbecause "a state
may be unable to recoverif its trustis betrayed."34
Defensiverealismfaultsthese argumentsforbeing incomplete.Cooperation
is risky,but so is competition.States cannot be certainof the outcome of an
arms race or war beforehand,and losing such a competitioncan jeopardize a
state's security.Waltz's balance-of-powertheoryand Mearsheimer's offensive
realism require that states evaluate the risks of cooperationand competition,
but theydo not explain variationin competitiveor cooperativebehavior.35 This
has implicationsforboth foreignpolicy and internationaloutcomes.
30. Glaser and Kaufmann,"What Is the Offense-DefenseBalance?" p. 57. For a different view, see
Van Evera, CausesofWar,chap. 6; and Emily0. Goldman and RichardB. Andres,"SystemicEffects
of MilitaryInnovationand Diffusion,"SecurityStudies,Vol. 8, No. 4 (Summer 1999), pp. 79-125.
31. Walt,OriginsofAlliances,pp. 21-34, 262-285. Walt definesthreatas a compositeof a state's ag-
gregate power, offensivemilitarycapabilities,geographic proximity,and perceived aggressive
intentions.
32. Waltz, TheoryofInternationalPolitics,p. 105.
33. Ibid.
34. Mearsheimer,"Back to theFuture,"p. 12; and Mearsheimer,"False Promiseof InternationalIn-
stitutions,"p. 12.
35. Glaser,"Realists as Optimists,"pp. 130-133.
The defensive variants of neorealism and neoclassical realism specifythe

conditionsunder which cooperativeinternationaloutcomes and less competi-
tive state behavior, respectively,become more likely.According to offense-
defensetheoryproponents,at the operationaland tacticallevel, improvements
in firepower(e.g., machine guns, infantryantitankweapons, surface-to-air
missiles,and tacticalnuclear weapons) should favor the defense because at-
tackersare usually more vulnerableand detectablethan are well-preparedde-
fenders.At the strategiclevel, the anticipatedhigh costs and risksof conquests
should deter even greedy leaders.36
The nuclear revolution-specifically the development of secure second-
strikecapabilities by the declared nuclear states-provides strongdisincen-
tives forintendedwar.37This does not mean thatpairs of nuclear-armedstates
will not engage in political-military competitionin thirdregions or limited
conventionalconflictshortof all-outwar.38Ratherit suggeststhatintended(or
premeditated)wars-wars thatbreak out as the resultof a calculated decision
by at least one partyto resortto the massive use of forcein the pursuitof its
objectives-become highly unlikely.39Conversely,if the offensedominates,
then states have an incentiveto adopt aggressive strategies.Similarly,states'
abilitiesto extractresourcesfromconquered territory influencethe likelihood
of internationalconflict.Whereindustrialcapacity,strategicdepth,or raw ma-
terialsare cumulative,defensiverealistswould expect statesto pursue expan-
sionistpolicies.40
36. Glaser and Kaufmann,"What Is the Offense-DefenseBalance?" p. 64.

37. Van Evera, "Correspondence:TakingOffenseat Offense-DefenseTheory,"p. 195; Miller,When
OpponentsCooperate, pp. 64-66; and RobertJervis,TheMeaningoftheNuclearRevolution:Statecraft
and theProspectofArmageddon (Ithaca,N.Y.: CornellUniversityPress,1989),pp. 4-5, 19-21,29-35.
38. Lieber,"Grasping the TechnologicalPeace," pp. 100-102. Lieber writes,"A finalpredictionof-
fense-defensetheorymakes about behavior under nuclear defense is thatstates should not com-
pete or fighttoo intenselyover territory beyond thehomeland or the homeland of close allies." He
claims that frequentSoviet and U.S. interventionsaround the world during the Cold War
disconfirmthis proposition.This is not reallya testableproposition,but rathera policy prescrip-
tion that flows fromoffense-defense theory(and defensiveneoclassical realism in general). See
BarryR. Posen and Stephen Van Evera, "Defense Policy and the Reagan Administration:Depar-
turefromContainment,"International Security,Vol. 8, No. 1 (Summer1983),pp. 3-45, at p. 33; Walt,
"The Case forFiniteContainment,"pp. 22-30; StephenVan Evera, "Why Europe Matters,Why the
ThirdWorldDoesn't: AmericanGrand StrategyaftertheCold War,"Journal ofStrategicStudies,Vol.
13, No. 2 (June 1990), pp. 1-51, especially pp. 4-5; Van Evera, Causes of War,pp. 245-246; and
Gholz, Press, and Sapolsky,"Come Home, America,"pp. 14-15.
39. See BenjaminMiller,WhenOpponentsCooperate:GreatPowerConflict and Collaboration in World
Politics(Ann Arbor:Universityof Michigan Press,1995),pp. 27-28. The term"intendedwar" says
nothingabout the objectivesof the attackingstate. Intended wars encompass both conflictsiniti-
ated forself-aggrandizement (i.e., greed) and preventivewars (i.e., conflictsinitiatedto block or re-
tard the furtherrise of an adversary).
40. Van Evera, Causes of War,pp. 108-112. See also Walt, "The Case for Finite Containment,"
pp. 19-22; Peter Liberman, Does ConquestPay? The Exploitationof OccupiedIndustrialSocieties
International
Security25:3 | 140
According to Mearsheimer,states must constantlyworry about theirsur-

vival because potentialcompetitorsmay tryto eliminatethemat any time.He
argues, "States operate in both an internationalpolitical environmentand an
internationaleconomic environment,and the formerdominates the latterin
cases where the two come into conflict."-41
This implies thatstateswill heavily
discountthe futureby favoringshort-term militarypreparednessover longer-
term objectives, such as economic prosperity,when and if the two goals
conflict.42
Again, defensiverealismfindsthis argumentlacking and specifiesthe con-
ditions under which states are more likelyto heavily discount the futureand
prefershort-term militarypreparednessto long-termeconomic prosperity.For
example, where geography provides defense frominvasion or blockade, de-
fensiveneoclassical realismwould expecta stateto favorlong-termobjectives.
Similarly,a state with relativelyweak neighborscan affordto take a longer-
termperspectiveand devote a greaterportionof its national resourcesto do-
mesticprograms.A relativelybenign threatenvironmentremoves the incen-
tives for the development of strongcentralinstitutionswithinthe state. For
example, geographic separation fromEurope and the relative weakness of
Canada and Mexico allowed the United States to survive the first150 years of
its independence without developing strong state institutions(i.e., a large
standingarmy,an efficienttax system,and a large centralbureaucracy).43
There are several circumstances,however,where defensiverealism expects
states to favor short-termmilitarypreparedness over long-termeconomic
prosperity.States that lack defensibleborders or have strongneighborswill
have a powerfulincentiveto build strongcentralinstitutions,maintainlarge
standing armed forces,and adopt offensivemilitarydoctrines.The rise of
Prussia is the classic example of how a precarious threat environment
influencesboth a state's grand strategyand the developmentof its domestic
political institutions.The original Hohenzollern territorieswere noncontigu-
ous and lacked defensibleborders.44Externalvulnerabilityprovided strongin-
(Princeton,N.J.:PrincetonUniversityPress, 1996); and Carl Kaysen, "Is War Obsolete? A Review

Essay," InternationalSecurity,
Vol. 14, No. 4 (Spring 1990), pp. 42-64, especially pp. 54-57.
41. JohnJ.Mearsheimer,"Disorder Restored,"in Graham Allison and GregoryF. Treverton,eds.,
Rethinking America'sSecurity:BeyondCold WortoNezvWorldOrder(New York:W.W.Norton,1992),
pp. 213-237, at p. 222.
42. Brooks,"Dueling Realisms," p. 452.
43. See Eliot A. Cohen, "The Strategyof Innocence?The United States,1920-1945,"in Williamson
Murray,MacGregorKnox, and Alvin Bernstein,eds., TheMakingofStrategy: Rutlers,States,and War
(New York:Cambridge UniversityPress, 1994), pp. 428-465.
44. On thereciprocalrelationshipamong externalthreat,statebuilding,and militarystrategy(and
foreignpolicy in general), see Charles Tilly, "Reflectionson the History of European State-
centivesforthe developmentof efficientstate institutionsto extractresources

fromdomesticsociety,a standingarmy,and a preferenceforoffensivemilitary
doctrines.Successive Prussiankings,mostnotablyFredericktheGreat,triedto
acquire, throughopportunisticexpansion and shrewd alliances,additional ter-
ritoriesneeded to consolidate and round offthe kingdom's borders.45
Likewise, when the offense-defense balance favors the offenseor if a state
lacks defensibleborders,one should expect that state to adopt a very short-
termperspectivewhen faced with a risingexternalthreat.This in turnmay
cause states to engage in truncatedor hasty diplomacy,conceal grievances,
adopt offensivemilitaryforcepostures,and seize first-moveadvantages.46
ASSUMPTION 3: THE INFLUENCE OF MATERIAL CAPABILITIES ON FOREIGN POLICY

Defensive neoclassical realistsassume thatin the shortrun,the relativedistri-
bution of power is oftenuncertainand leaders oftenface ambiguous and con-
tradictoryinformation.Thereforesuch foreignpolicy theoriesposit an explicit
role forleaders' preexistingbeliefsystems,images of adversaries,and cogni-
tive biases in the process of intelligencegathering,net assessment,military
planning,and foreignpolicy decisionmaking.Much of what defensiveneoclas-
sical realistsseek to explain would be simplyinexplicablewithoutreferenceto
the perceptionsof centraldecisionmakers.47
The role of such perceptionalvariables becomes particularlyimportantdur-
ing periods of rapid power fluctuation.48They also play an importantrole dur-
ing noncrisis periods and periods when the distributionof power remains
relativelystable. Benjamin Miller finds that benign images of the opponent,
balancingbeliefs,and ideological similarity,along withmultipolarityand com-
mon fears of revolution,are necessaryconditionsforthe emergenceof great
power concerts.49
Making," in Tilly,ed., TheFormation ofNationalStatesin WesternEurope(Princeton,N.J.:Princeton

UniversityPress, 1975); Bruce D. Porter,Warand theRise of theState:The MilitaryFoundationsof
ModernPolitics(New York: Free Press, 1994); and BarryR. Posen, "Nationalism,the Mass Army,
and MilitaryPower," International Vol. 18, No. 2 (Fall 1993), pp. 80-124.
Security,
45. See Gordon A. Craig, The PoliticsofthePrussianArmy,1640-1945 (Oxford:Clarendon, 1955);
and H.W. Koch, A HistoryofPrussia (New York: Dorsett,1978), pp. 102-139.
46. Van Evera, Causes of War,pp. 45-53.
47. See Posen, Sourcesof MilitaryDoctrine,pp. 67-69; and Christensen,"Perceptions and Alli-
ances," pp. 68-70. For a discussion of why leaders oftendraw the wrong historicallessons, see
RobertJervis,Perception andMisperception Politics(Princeton,N.J.:PrincetonUniver-
in International
sityPress, 1976), chap. 6; and Yuen Foong Khong, Analogiesat War:Munich,Korea,Dien BienPhu,
and theVietnamDecisionsof1965 (Princeton,N.J.:PrincetonUniversityPress, 1992), chap. 2.
48. Christensen,"Perceptionsand Alliances," p. 92; and Walt,Revolutionand War,chap. 2.
49. Miller,WhenOpponentsCooperate, pp. 110-119.
International
Security25:3 | 142
Finally,leaders' perceptionsplay a criticaland at times pernicious role in

shaping how states respond to the structuralmodifiers.Oftenthe "objective"
offense-defense balance is sharplyat odds with civilian and militaryleaders'
perceptionsof it.50Officialsoftendraw upon the "lessons of history"in formu-
latingmilitarydoctrineor allow organizationalprioritiesto overridelegitimate
securityrequirements.51 The most oft-citedinstancein which thishappened is
the "cult of the offensive"among the European great powers beforeWorld
War 1.52
ASSUMPTION 4: DOMESTIC POLITICS AND SYSTEMIC IMPERATIVES

The defensivevariantof neoclassical realismposits a role fordomesticpolitics
in shaping states' foreignpolicies. Furthermore, defensiveneoclassical realism
specifiesthe conditionsunder which domesticpoliticsmattersin foreignpoli-
cy.For example,duringperiods of imminentexternalthreat,thecalculationsof
centraldecisionmakersare paramount.Over the longertermor in the absence
of an immediateexternalthreat,national leaders will have more difficulty in
mobilizingdomesticresourcesforforeignpolicy.Furthermore, leaders' mobili-
zation effortsmay laterrestricttheirabilityto readjusttheirforeignpolicies in
response to changes in the externalenvironment.
Thomas Christensen'sdomesticmobilizationtheoryaddresses the problem
of how domesticpoliticsconstrainsstates' abilitiesto adjust theirforeignpol
icies.53In the late 1940s and 1950s,U.S. and Chinese leaders soughtto mobilize
domestic resourcesto balance against the Soviet Union, but lacked sufficient
"national politicalpower" to do as theypleased. PresidentHarryTrumanand
Chairman Mao Zedong used domesticallypopular but unnecessary foreign
policies in secondaryareas as a diversionfornecessarybut unpopular policies
in primaryareas. These secondarypolicies set in motiona chain of events cul-
minatingin the U.S. and subsequent Chinese interventionsin the Korean War
and the 1958 Quemoy-Matsu crisis.54
50. Offense-defense theoryhas both an objectivecomponentand a perceptionalcomponent.The

objectiveoffense-defense balance is a structuralmodifier.Elite perceptionsof the offense-defense
balance are unit-levelphenomena.See the exchangebetween Davis and Van Evera in "Correspon-
dence: Taking Offenseat Offense-DefenseTheory,"pp. 179-182, 195-200.
51. Posen, SourcesofMilitaryDoctrine,pp. 67-69.
52. StephenVan Evera, "The Cult oftheOffensiveand theOriginsof theFirstWorldWar,"Interna-
tionalSecurity, Vol. 9, No. 1 (Summer 1984),pp. 58-107; Van Evera, Causes ofWar,pp. 193-239; and
Christensen,"Perceptionsand Alliances," pp. 82-83.
53. Thomas J.Christensen,Usefuil Adversaries:GraiidStrategy,DomesticMobilization, and Sino-Amner-
ican Conflict,1947-1958 (Princeton,N.J.:PrincetonUniversityPress, 1996), p. 256.
54. Ibid., pp. 32-76, 194-241.
Defensive neoclassical realism relies on a top-down conceptionof political

behaviorand rejectsliberalism'sassumptionthatthefundamentalactorsin in-
ternationalpoliticsare risk-averse,rationalindividuals or groups withinsoci-
ety.55Leaders weigh options and make decisions based primarilyon their
strategicsituationand an assessmentof relativepower. State autonomyvis-a-
vis civil society,organizationalpolitics,and civil-militaryrelations,however,
can constrainthe efficiencyof leaders' responses to systemicimperatives.For
example, statestrength(i.e., the extractivecapacityof a state's centralpolitical
institutions)influencesboth the amount of militarypower a state can project
abroad and the scope of its grand strategy.56
Consider the grand strategiesof the superpowers during the Cold War.
Aaron Friedbergargues thatwhile the gross distributionof power and struc-
turalmodifierspushed theUnited Statesand theSovietUnion towardconfron-
tation,internalfactorsshaped the typesof strategieseach side pursued. In the
U.S. case, a combinationof weak state institutions,the material interestsof
various societal actors,and an embedded antistatistideology eventuallyled
to the adoption of a flexibleresponse strategyand a limited program of
power creation.57 The Soviet Union, on the otherhand, lacked all of the coun-
tervailingdomestic influences.As a result,during most of the Cold War,the
Soviet Union pursued a more ambitious militarydoctrine(i.e., full war fight-
ing) than did the United States and undertooka farmore expansive program
of power creation.58
CriticismsofDefensiveRealism
The fourauxiliaryassumptions discussed above definedefensiveneorealism

and defensiveneoclassical realism.Recently,however,both nonrealistsand of-
55. Andrew Moravcsik,"TakingPreferencesSeriously:A Liberal Theoryof InternationalPolitics,"

International Organization,
Vol. 51, No. 4 (Autumn 1997), pp. 513-553.
56. Both realistand nonrealistworks use variationin state strengthto explain national security
and foreigneconomic policies. See Zakaria, FromWealthto Power,pp. 95-113; Stephen D. Krasner,
DefendingtheNationalInterests:Raw MaterialsInvestment and U.S. ForeignPolicy(Princeton,N.J.:
PrincetonUniversityPress, 1978); Aaron L. Friedberg,"Why Didn't the United States Become a
GarrisonState?" International Security,Vol. 16, No. 4 (Spring 1992), pp. 109-142; Friedberg,In the
ShadowoftheGarrisonState:America'sAnti-Statism and Its Cold WarGrandStrategy(Princeton,N.J.:
PrincetonUniversityPress,2000); PeterJ.Katzenstein,ed., BetweenPowerand Plenty:ForeignEco-
nomicPoliciesofAdvancedIndustrialSocieties(Madison: Universityof WisconsinPress, 1978); and
Michael C. Desch, "War and StrongStates,Peace and Weak States?"International Organization, Vol.
50, No. 2 (Spring 1996), pp. 237-268.
57. Friedberg,In theShadowoftheGarrisonState,p. 66.
58. Ibid., p. 75.
Security25:3 | 144
International
fensiverealistshave raised several theoreticaland empiricalcritiquesof each.

The followingsection examines the most prominentcritiquesand concludes
thatmost proceed fromincompleteor flawed analysis.
CRITICISM 1: THERE IS NO SECURITY DILEMMA

Randall Schwellerchallenges the existenceof the securitydilemma and faults
Waltz forallegedlyrelyingon uncertainty, insteadof structure,to explain inter-
national conflictand balancing. He writes,"Predatorystatesmotivatedby ex-
pansion and absolute gains, not securityand the fearof relativelosses, are the
prime movers of neorealisttheory.Withoutsome possibilityof . . . [the] exist-
ence [of predatorystates],the securitydilemma melts away, as do most con-
cepts associated with contemporaryrealism."59
Schweller contendsthatbalancing would not occur in an internationalsys-
tem comprisedentirelyof security-seeking states,unless stateswere uncertain
of one another'smotives.He argues thatifall statesseek security,"the security
dilemma is always apparent,not real." If aggressors (i.e., states that arm for
nonsecurityreasons) do exist,then therewould be no securitydilemma "but
ratheran example of a state or coalitionmobilizingforthe purpose of expan-
sion and targetsof thataggressionrespondingby acquiringarms and forming
alliances to defend themselves."If,on the otherhand, real aggressorsdo not
exist,this reliance on uncertaintyto explain war and balancing behavior vio-
lates "realism's most basic tenetthatconflictsof interestamong statesare gen-
uine ratherthan the resultof misunderstandingand misperception."60
Schweller's critique of the securitydilemma misses the mark in threere-
spects. First,he fails to appreciate the role that uncertaintyplays in both
Waltz's balance-of-powertheoryand defensiveneorealism.Glaser notes that
neorealismassumes that states are black boxes "that provide no information
about internaldifferences, except forthe observable outputs of theirinterna-
tional policy choices."61States do not rely on the internalcharacteristicsof
otherstates-for example,theirpoliticalor economicsystems-to discerntheir
motives.Instead,theymust draw inferencesfromobservable internationalbe-
havior.Nevertheless,when a statefacesa securitydilemma,manypolicies that
would improve the state's securityalso send ambiguous informationabout
motives.
59. Schweller,"Neorealism's Status Quo Bias," p. 119.

60. Ibid., pp. 117-118.
61. Glaser,"The SecurityDilemma Revisited,"p. 195.
Second, statesfacereal uncertaintynot onlyabout one another'spresentmo-

tives but also about theirfuturemotives and relativecapabilities. States can
neverbe certainof one another's futureintentions,regardlessof whetherthey
harbor "greedy" or "security-seeking"motives in the present. Uncertainty
about thefutureand anticipatedshiftsin therelativedistributionof power cre-
ate incentivesand disincentivesforcooperativeor competitivepolicies in the
present.Windows of opportunitytemptdeclining states to contemplatewar
beforethe power shiftis complete. For example, in the late 1940s and early
1950s,the loss of its nuclear monopoly and the developmentof thermonuclear
weapons created a window of opportunityfor the United States. Preventive
war argumentswere common withinthe JointChiefsof Staffand the Truman
administration.62 Few scholars,however,would classifythe United States as
harboringgreedy motives.63
Third,Schwellerignorestheincidenceof security-driven expansion in world
history.The historicalrecordabounds with cases of states thatpursued secu-
rity-driven expansion or preventivewar. This is not to say thatdefensivereal-
ists deny the existenceof predatorystates;in fact,theydo not. Few defensive
realists would classifythe expansionist behavior of Napoleonic France and
Nazi Germanyas securitydriven.64
CRITICISM 2: GREEDY STATES AND INTERNATIONAL CONFLICT

Andrew Kydd argues thatgreedystates,not the securitydilemma,are theper-
missive cause of internationalconflictand balancing behavior.65He constructs
a formalmodel in which all states are securityseekers. The model assumes
62. Marc Trachtenberg,"A WastingAsset: American Strategyand the ShiftingNuclear Balance,

1949-1954,"in Trachtenberg, ed., Historyand Strategy(Princeton,N.J.:PrincetonUniversityPress,
1991),pp. 100-152,at pp. 107-111.See also Van Evera, CausesofWar,chap. 4; and JackS. Levy,"De-
cliningPower and the PreventiveMotivationforWar," WorldPolitics,Vol. 40, No. 1 (October1987),
pp. 82-107.
63. Of course, states can address relativedecline by seeking additional allies or making conces-
sions to adversaries,instead of launching preventivewars. See Randall L. Schweller,"Domestic
Structureand PreventiveWar: Are Democracies More Peaceful?" WorldPolitics,Vol. 44, No. 2 (Jan-
uary 1992),pp. 235-269; and WilliamC. Wohlforth, "Realism and the End of the Cold War,"Inter-
nationalSecurity,Vol. 19, No. 3 (Winter1994/95),pp. 91-129.
64. BothVan Evera and Dale C. Copeland argue thatwindows of opportunityand preventivewar
calculationsdrove Adolf Hitler's decision forwar against theSoviet Union in June1941. Ideology
and self-aggrandizement propelled Hitler's overall grand strategy.See Van Evera, Causes of War,
pp. 94-99; Copeland, "Economic Interdependenceand War: A Theoryof Trade Expectations,"In-
ternational Vol. 20, No. 4 (Spring 1996), pp. 5-41, at pp. 33-39; and Copeland, Originsof
Security,
Major War,pp. 139-145.
65. Andrew Kydd, "Sheep in Sheep's Clothing:Why SecuritySeekers Do Not FightEach Other,"
SecurityStudies,Vol. 7, No. 1 (Autumn 1997), pp. 114-155.
International
Security25:3 | 146
thatall stateshave completeinformationabout one another'spreferences.The

result would be a no-war equilibrium: States would have no incentive to
arm against or attackanother.This equilibriumwould exist regardlessof the
relative distributionof power, the offense-defense balance, trends in power
growth,or the cumulativityof resources.Kydd thencomplicatesthe model by
adding uncertaintyabout futureintentions.Uncertaintymay underminethe
no-warequilibrium,because greedystatescannotachieve theirmaximumpay-
offby not attacking.
Althoughstatescan neverbe completelycertainof one another's presentor
futureintentions,theydo have two means to reduce uncertaintyto manage-
able levels. First,formoderndemocraticstates,the policymakingprocess is so
open thattheycannothelp but reveal theirtruepreferences.This transparency
gives otherstatesample informationabout a democracy'sgreedyor security-
seeking motivations.66Second, all states, both democracies and nondemoc-
racies, can employ costlysignals to reveal theirbenign intentions.Costly sig-
nals are strategiesthat one type of actor in a game can take that othertypes
mightfindtoo costly For example, in a crisisbargaininggame, a costlysignal
would be some act thatraises the cost ofbackingdown. Actorswho have a low
expectedvalue forwar or forthe objectat stake are more likelyto send such a
signal.
Kydd citesfourexamples of costlysignals.67The firstis ideological modera-
tionon thepartofa stateconsideredaggressiveby others.For a leadershipthat
trulybelieves in an aggressiveideology,ideological moderationhas real costs.
For moderate leaders, on the otherhand, the costs of ideological moderation
are considerablyless. Second, tolerationof domestic minoritiescan signal be-
nign intentions,whereas leaders who oppress minoritiesmightfeel littlecon-
straintagainst dominating other states and peoples. Third, benign policies
toward weaker neighboringstates can signal security-seeking intentions."By
allowing substantiallatitude and freedomwithin the geographical range in
66. Ibid., p. 138. Kydd does not considerhimselfan offensiverealistand is broadlysympatheticto

defensiverealism.Indeed, his formalmodel illustratesthatthesearchforsecuritydoes not directly
translateintorelativepower maximization.My criticismof his articlemainlyconcernstwo points:
(1) the conclusionsdrawn about the transparencyof democraticpoliticalsystems;and (2) the his-
torical incidence of war or internationalcrises short of war between pairs of security-seeking
states.Andrew Kydd, correspondencewith the author,August 18 and 21, 2000.
67. Kydd's conceptionof costlysignalingand reassuranceturnson transparency, not on audience
costs. For examples of the latter,see JamesD. Fearon,"Domestic PoliticalAudiences and the Esca-
lation of InternationalDisputes," AmericanPoliticalScienceReview,Vol. 88, No. 3 (September1994),
pp. 577-592; and KurtTaylorGaubatz, "DemocraticStatesand Commitmentin InternationalRela-
tions,"International Organization, Vol. 50, No. 1 (Winter1996), pp. 109-139.
which it could enforcestrictobservanceof its wishes ifit chose to,the security-

seekinggreatpower signals thatit is uninterestedin conquest and domination
forits own sake."68Finally,true security-seeking states will pursue unilateral
arms controland moderate militarypolicies. In short,greedy states are un-
likelyto undertakesuch costlysignals.
Defensiverealistswould agree withsome aspects of Kydd's model, but they
would disagree with the broad conclusions he draws fromit. For example, in
relationswith theirweaker neighbors,statescan and do employ arms control,
defensivemilitarydoctrines,forcepostures,and moderate policies. Such ac-
tions can only mitigate,but not eliminate, the security dilemma.69Like
Schweller,Kydd underestimatesthe difficulty of discerningstates' futurein-
tentionsand power trends: "With respect to fears about futuremotivations,
while theyare conceivablyable to motivateconflictand war,theyhave histori-
cally rarelydone so. It is difficult
to explain manyhistoricalarms races or wars
as a resultof fearsabout the possible futurepreferencesof a currentbenignre-
gime."70On the contrary, the Peloponnesian War,the War of the Spanish Suc-
cession, the Seven Years' War, the French Revolutionary Wars, Russia's
mobilization in the July1914 crisis,the 1962 Sino-Indian War,the 1965 and
1971 Indian-PakistaniWars,and the 1990 Kashmircrisisare all cases where se-
curity-seeking statesprecipitatedinternationalcrisesor wars in partbecause of
uncertaintyabout adversaries' intentionsand impendingpower shifts.71
Japan's expansion in the 1930s and early 1940s and China's interventionin
the Korean War are classic cases of how security-driven policies and fearsof
adversaries' futureintentionscan provoke conflict.The notion that Japan
could best provide forits securitythroughempire and autarkyoriginatedin
the lessons that its militaryplanners drew fromGermany's defeat in World
War I. If futureconflictsresembledthatwar, a state's abilityto win would de-
pend largelyon its abilityto mobilize economic resources.The Japanesehome
islands, however, lacked the natural resources needed to fighta prolonged
war, which in turnmade Japan vulnerable to exploitationor attackfromthe
68. Kydd, "Sheep in Sheep's Clothing,"pp. 143.

69. Glaser,"Realists as Optimists,"pp. 143-146; and Walt,OriginsofAlliances,chap. 6.
70. Kydd, "Sheep in Sheep's Clothing,"p. 117.
71. See PatriceLouis-Rene Higonet, "The Origins of the Seven Years' War,"JournalofModernHis-
tory,Vol. 40, No. 1 (March 1968), pp. 57-90; Luigi Albertini,The OriginsoftheWarof1914, Vol. 2,
trans.and ed. Isabella M. Massey (New York:OxfordUniversityPress, 1953), p. 559; Van Evera,
Causes of War,pp. 77, 79, 84; Copeland, OriginsofMajor War,pp. 79-117; Walt,Revolutionand War,
pp. 46-128; Miller,WhenOpponentsCooperate, pp. 125-174; and Devin T. Hagerty,TheConsequences
ofNuclearProliferation:LessonsfromSouthAsia (Cambridge,Mass.: MIT Press, 1998), pp. 117-176.
Security25:3 | 148
International
United States, the Soviet Union, the Netherlands,or Great Britain.72To that
end, elementsof theJapaneseImperial Armypursued a measured expansion-
ist strategyin resource-richManchuria and northernChina from 1931 to
1937.73Japaneseperceptionsof vulnerabilityand inevitablehostilityfromthe
West and the Soviet Union dominated decisionmakingwithinJapan's army
and naval generalstaffsand thecabinetsof successive primeministers-Prince
Konoe Fumimaro,Baron Hiranuma Kiichiro,Gen. Abe Nobuyuki,Adm. Yonai
Mitsumasa, and finallyGen. Tojo Hideki.74
The July1937 clash between Japanese and Kuomintang (KMT or Chinese
Nationalist)troopsnear Beijing's Marco Polo Bridge escalated into a full-scale
war betweenJapanand the forcesof Chiang Kai-shek.The so-called China In-
cidentimposed tremendousstrainson Japan's economy;increasedJapan's de-
pendence on the United States,Great Britain,and the Netherlandsforoil and
raw materials;and heightenedtensionbetween Tokyo and Washington.Fur-
thermore,theJapaneseImperialArmy'soperationsnear theSino-Sovietboard-
er in Manchuria broughtit into directconfrontationwith the Soviet army.75
Perceptionsof vulnerability,considerationsof sunk costs, and a fading win-
dow of opportunitydrove the militarychiefsand the Konoe and Tojo cabinets
to undertake various high-riskstrategies.These included Japan's expansion
into the Dutch East Indies to acquire oil and raw materials;the use of diplo-
matic and later militarymeans to stop the flow of arms to the KMT through
FrenchIndochina;the conclusionof theTripartite Pact withNazi Germanyand
72. On the perceived lessons of WorldWar I and the originsof totalwar planningin theJapanese
ImperialArmy,see Michael A. Barnhart,JapanPreparesforTotalWar:The SearchforEconomicSecu-
rity,1919-1941 (Ithaca, N.Y: Cornell UniversityPress, 1987), pp. 22-23; and Mark R. Peattie,
IshiwaraKanji and Japan'sConfrontation withthe West(Princeton,N.J.:PrincetonUniversityPress,
1975), chaps. 1-3.
73. See Seki Hiroharu, "The Manchurian Incident, 1931," in James William Morely,ed., Japan
Erupts:The LonidonNaval Conference and theManchurianIncident,1928-1932 (New York:Columbia
UniversityPress, 1984), pp. 143-170; and Shimada Toshikiko,"Designs on North China," in
Morely,ed., The China Quagmire:Japan'sExpansionon theAsian Continent, 1933-1941 (New York:
Columbia UniversityPress, 1983), pp. 135-156.
74. Konoe served as prime ministerfromJune4, 1937 to January5, 1939; July22, 1940 to July18,
1941; and July18 to October 18, 1941. Hiranuma,Abe, and Yonai held the premiershipfromJanu-
ary 5 to August 30, 1939; August 30, 1939 to January16, 1940; and January16 to July22, 1940,re-
spectively.Tojo succeeded Konoe on October18, 1941,and held thepremiershipuntilJuly18, 1944.
75. The KMT and theChinese CommunistParty(CCP) suspended theirbittercivil war and forged
a tacticalalliance against theJapanese.In theirpursuitof the CCP in Manchuria,Japanesetroops
repeatedlyviolated the Sino-Soviet border in 1938-39. For a detailed examinationof the unde-
clared war betweenJapaneseand Soviet forces,see Hata Ikuhiko,"The Japanese-SovietConfronta-
tion, 1935-1939," in James William Morely, ed., DeterrentDiplomacy:Japan,Germany,and the
U.S.S.R., 1935-1940(New York:Columbia UniversityPress,1976),pp. 129-178;and Alvin D. Coox,
Nomohan:JapanagainstRussia,1939, 2 vols. (Stanford,Calif.: StanfordUniversityPress, 1985).
Italyas a means to deterpotentialadversaries;and ultimatelythe decision for

war with the United States.76
Similarlythe pursuitof security,uncertaintyabout the otherside's capabili-
ties and intentions,miscommunication,and misperceptionsset the stage for
conflictbetween the United States and China in the autumn of 1950. Percep-
tions of vulnerability,
not greed, drove the Trumanadministration'sdecisions
to send ground troops to the defense of South Korea followingthe NorthKo-
rean invasion on June25, 1950. InitiallyPresidentTruman,Secretaryof State
Dean Acheson, and otherU.S. officialssought to restorethe statusquo ante: a
divided Korean peninsula with two sovereignstateson eitherside of the 38th
parallel. The administrationescalated its war aims fromcontainmentto roll-
back in September,in large part to remove the perceived long-termthreatto
East Asian security-namely,the NorthKorean regime.77Because Soviet inter-
vention appeared unlikelyand Chinese intentionsand capabilities remained
unclear,the Trumanadministrationsaw an opportunityto reunifyKorea. At
the same time,the administrationdid not intendto help Chiang and the KMT
retakethe mainland or extend the war beyond the Korean peninsula. Truman
and otherofficialspubliclymade statementsto thateffect.78 Nonetheless,Gen.
Douglas MacArthur'sadvance across the 38thparallel,along withthe Truman
administration'searlierdecisions to send the U.S. SeventhFleet to the Taiwan
Strait,deny diplomatic recognitionto the new communistregime in Beijing,
and continueaid to Chiang and the KMT on Taiwan, convincedMao thatany
U.S. militarypresence on the Korean peninsula posed a threat.In short,Chi-
nese forcesenteredthe Korean War because Mao fearedthe Trumanadminis-
tration'sfutureintentions.79
76. Jeffrey W. Taliaferro,"Quagmires in the Periphery:ForeignWars and EscalatingCommitment

in InternationalConflict,"SecurityStudies,Vol. 7, No. 3 (Spring 1998), pp. 94-144; and Peter
Liberman,"The Offense-DefenseBalance, Interdependence,and War,"SecurityStudies,Vol. 9, Nos.
1/2 (Autumn 1999-Winter2000), pp. 59-92, especiallypp. 82-88. See also Van Evera, CausesofWar,
pp. 89-94.
77. I concede thatthe expansion of the Truman administration'saims in the Korean War-from
containmentto rollback-appears to provide strongsupport forone of offensiverealism's main
hypotheses.U.S. officialsescalated theiraims in response to perceived battlefieldopportunities
(i.e., Gen. Douglas MacArthur'ssuccessfullanding at Inchon) and systemicopportunities(i.e., the
Trumanadministration'sperceptionsof China's weakness followingthe civil war and the dimin-
ished probabilityofSoviet intervention).For thisargument,see Labs, "Beyond Victory,"pp. 34-39.
78. WilliamStueck,TheKoreanWar:An International History(Princeton,N.J.:PrincetonUniversity
Press),pp. 66-70; and RosemaryFoot, The WrongWar:AmericanPolicyand theDimensionsoftheKo-
reanConflict, 1950-1953 (Ithaca, N.Y.: Cornell UniversityPress, 1985), pp. 82-83.
79. See Christensen,UsefulAdversaries, pp. 138-193;Chen Jian,China'sRoad totheKoreanWar(New
York:Columbia UniversityPress, 1994); and Sergei N. Goncharov,JohnW. Lewis, and Xue Litai,
International
Security25:3 | 150
States may not always interpretan adversary's ideological moderationas a

even ifconcretediplomaticmoves accompany its rhetoric.80
signal of restraint,
Consider,forexample, the Eisenhower administration'sreactionto the "ideo-
logical thaw" in Soviet foreignpolicy followingthe death of JosefStalin in
March 1953. Premier Gregori Malenkov signaled a willingness to pursue
detentewith the United States. On March 15, he announced therewas no su-
perpowerdispute that"could not be settledby peaceful means, on thebasis of
mutual agreement."'81 Concrete actions followed this peace initiative,includ-
ing: Soviet restorationof diplomaticrelationswith Yugoslavia, Greece,and Is-
rael; the withdrawal of Soviet territorialdemands on Turkey; and Soviet
pressureon Beijingto end the Korean War.Nevertheless,PresidentDwight Ei-
senhower and Secretaryof State JohnFosterDulles saw the Soviet "peace of-
fensive" as a tactical ploy designed to undermine popular support in the
United States and among NATO allies forhigh defense expenditures,greater
burden sharing,the developmentof the hydrogenbomb, and the forwardde-
ploymentof U.S. troopsin WesternEurope.82The administrationresponded to
the Soviet peace offensiveby demanding politicalliberalizationin EasternEu-
rope, an Austrianneutralitytreaty,and the repatriationof all German prison-
ers of war. Eisenhower made two nuclear disarmamentproposals, the Atoms
forPeace Programand the Open Skies initiative,largelyto bolsterthe United
States' image abroad and make the Soviets appear intransigent.83 Similarly,in
1986-87 the Reagan administrationremainedwary of Moscow's intentionsde-
spite General SecretaryMikhail Gorbachev's willingnessto make deep (and in
some cases) unilateral cuts in intermediate-rangemissiles and conventional
forcesin Europe.84
UncertainPartners:Stalin,Mao, and the Korean War (Stanford,Calif.: StanfordUniversityPress,

1993).
80. Dale C. Copeland, "Do ReputationsMatter?" SecurityStudies,Vol. 7, No. 1 (Autumn 1997),
pp. 33-71.
81. Charge in the Soviet Union to the State Department,March 18, 1953, ForeignRelationsof the
UnitedStates:1952-1954,Vol. 8 (Washington,D.C.: U.S. GovernmentPrintingOffice,1984),p. 1131.
82. Kenneth A. Osgood, "Form before Substance: Eisenhower's Commitmentto Psychological
Warfareand Negotiations with the Enemy," DiplomaticHistory,Vol. 24, No. 3 (Summer 2000),
pp. 405-433, especially pp. 419-421.
83. Ibid., pp. 420-421. See also Vladislav Zubok and ConstantinePleshakov,Inside theKremlin's
Cold War:FromStalinto Khrushchev (Cambridge,Mass.: Harvard UniversityPress, 1996), pp. 154-
157; and Deborah Welch Larson, Anatomyof Mistrust:U.S.-SovietRelationsduringthe Cold War
(Ithaca, N.Y: Cornell UniversityPress, 1997), pp. 47-49.
84. Larson, AnatomyofMistrust,pp. 204-212; and Wohlforth, "Realism and the End of the Cold
War,"pp. 113-114.
Other parts of Kydd's model rely on essentiallyInnenpolitik argumentsto

challenge the existenceof the securitydilemma. He places considerable em-
phasis on domesticcostlysignals and the transparencyof democraticsystems,
but claims thathis model differsfromthe normativeand institutionalvariants
of the democraticpeace thesis.In actuality,it does not. Like Kydd's genericse-
curityseekers,democraticstatessupposedly projectinternalnormsof peaceful
conflictresolutionoutwardly.This signals theirbenignintentionstowardother
democracies. This mutual recognitionof shared norms for domestic conflict
resolutionsupposedly produces a shared, but exclusive,peace among demo-
craticstates.85
Defensive realismlargelyrejectsthe notionthatdemocraticstatesare better
able to signal intentionsthan are autocraticstates,because the multiplicityof
parties,actors,and interestgroups withina democraticstate can send mixed
messages about its intentions.Again, thereare manyhistoricalcases where po-
tentialadversaries misjudged the supposedly security-seeking preferencesof
democracies or where democracies failed to communicatetheirintentions.86
During theJuly1914 crisis,Britaindid not clearlysignal its resolve to support
France and Russia and protectBelgian neutralityuntil July28-29, by which
time the crisis was out of control.87The Japanese governmentand military
chiefsof staffin 1940-41 failed to discern the Roosevelt administration'strue
security-seekingpreferences,despite the supposed transparencyof the U.S.
political system.Similarly,throughoutthe Cold War, Soviet leaders viewed
their counterpartsin Washingtonas expansionist and irredeemablyhostile
(and vice versa).88Nor are these miscalculations limited to nondemocratic
states.The 1898 Fashoda crisis,the 1923 Ruhr crisis,the 1861 HMS Trentaffair,
and threeIndian-Pakistaniwars are all cases where pairs of democraticstates
85. JohnM. Owen, IV, "How LiberalismProduces DemocraticPeace," International Vol.

Security,
19, No. 2 (Fall 1994), pp. 87-125.
86. See BernardI. Finel and KristinM. Lord, "The SurprisingLogic of Transparency,"International
StudiesQuarterly, Vol. 48, No. 2 (June1999), pp. 325-339.
87. See Albertini,OriginsoftheWarof1812,pp. 514-527; and Van Evera, CausesofWar,pp. 222-224.
Copeland, on the otherhand, argues thatChancellorTheobald von BethmannHollweg and other
Germanleaders did not count on BritishneutralityduringtheJuly1914 crisis.See Copeland, Ori-
gins ofMajor War,pp. 60-64, 84-85, 92-93, 111-112.
88. Osgood, "Form beforeSubstance," pp. 412-413; GregoryMitrovich,Undermining theKremlin:
America'sStrategy to SubverttheSovietBloc,1947-1956(Ithaca,N.Y: CornellUniversityPress,2000),
especially pp. 122-177; Andrew Bennett,Condemnedto Repetition? TheRise,Fall, and RepriseofSo-
viet-RussianMilitaryInterventionism,1973-1996 (Cambridge,Mass.: MIT Press, 1999), pp. 128-131,
169-173,234-235; and Dale C. Copeland, "Trade Expectationsand the Outbreakof Peace: Detente,
1970-74,and the End of the Cold War,1985-91," SecurityStudies,Vol. 9, Nos. 1/2 (Autumn 1999-
Winter2000), pp. 15-59.
International
Security25:3 | 152
(or quasi-democraticstates) miscalculated each other's security-seekingmo-

tives despite the supposed openness of the domestic political systemsof the
statesinvolved.89
CRITICISM 3: UNIT-LEVEL FACTORS, AGGRESSIVE BEHAVIOR, AND WAR

Fareed Zakaria claims that defensive realism cannot explain much of world
politicsbecause its adherentsassert thatonly "pathological" states engage in
expansionistbehavior.Defensiverealism,he argues,assumes that"security[is]
plentiful"in the internationalsystemand that most states understandthis.90
Because states can easily obtain securitythroughthe pursuitof moderatefor-
eign policies,only stateswith defectivedomesticpoliticalsystemswill pursue
expansionist policies. In effect,defensive realism denies the existence of a
"real" securitydilemma and relies instead on unit-levelvariables to explain
greatpower expansion,conflict,and war. Zakaria incorrectly claims thatWaltz,
Jervis,and JackSnyderassume thatall statespursue (or oughtto pursue) mini-
mal security.He writes,"For thesescholars,thenatureof theinternationalsys-
tem dictates that a state possess limited externalinterests,maintain a small
military,and pursue a restrainedforeignpolicy."91Because the international
systempushes statestowardadoptingmoderategrand strategies,defensivere-
alism must place the causes of aggressionand conflictat the unit level.
Contraryto Zakaria's claims, defensiverealism assumes (as does offensive
realism) that at minimum all states seek to survive under anarchy.This as-
sumptiondoes not preclude expansive definitionsof states' securityrequire-
mentsor states' pursuitof nonsecuritygoals. Again, structuralmodifiers,such
as the offense-defense balance in militarytechnologyand states' geographic
proximityto one another,influencethe severityof the securitydilemma. As
noted above, under certaincircumstances,defensiveneorealismexpects states
to pursue expansioniststrategiesas a means to achieve security.92
CRITICISM 4: THE SECURITY DILEMMA AND POWER MAXIMIZATION

Several offensiverealists take power-maximizingbehavior to be the logical
consequence of anarchyand the securitydilemma. The internationalsystem
provides strongincentivesforthe pursuitof expansionistforeignpolicies be-
89. ChristopherLayne, "Kant or Cant? The Mythof the DemocraticPeace," Interniational

Secuirity,
Vol. 19, No. 2 (Fall 1994), pp. 5-49.
90. Zakaria, FromWealthto Pozver,p. 23.
91. Ibid., pp. 26-27, 30-31.
92. On this point,see Jervis,"Cooperation under the SecurityDilemma," pp. 189-190.
cause onlythestrongeststatesstand thebest chance of survival.Statesseek op-

portunitiesto weaken potentialadversaries and improve theirrelativepower
positions.Expansion is oftenthebest way to accumulatemorepower at theex-
pense of rivals.93Weak statesare unlikelyto pursue expansioniststrategiesbe-
cause the risks of doing so are high. Strongerstates rationallyadapt to the
internationalenvironmentand quickly learn to expand when and where the
costs and risks seem manageable. Eric Labs observes, "Successful expanders
learn frompast mistakesand theytryto go about expanding in a mannerthat
draws the least attentionof the othergreatpowers.... When Russia was de-
featedin the Crimean War,the lesson it took away was not thatit should not
tryto expand, but thatit should tryto expand elsewhere."94
Thereare two responses to thisargument-one deductive and the otherem-
pirical. Glaser identifiesthreedeductive reasons why power maximizationis
self-defeating. First,a statethatincreasesits relativepower mightnevertheless
decrease its own security.By makingan adversaryless secure,a statemightin-
advertentlyincreasethe value an adversaryplaces on expansion. Second, rela-
tive power maximization increases the probabilityof losing an arms race.
"Even a countrythat would preferto win an arms race-that is, that would
prefersuperiorityto parity-might choose cooperation over arms racing to
avoid the riskof losing the race." Third,by failingto distinguishbetween of-
fensiveand defensivepotential,the claim thatstatesmaximize relativepower
ignoresthe factthatdoing so may not maximize themilitarycapabilitiesthata
state needs fordeterrenceor defense.95
A relatedtheoreticalcritiqueconcernstheprevalenceof hegemony-a situa-
tionwhereone greatpower enjoysa preponderanceof thematerialcapabilities
in the internationalsystem.96Relative power maximizationposes no problem
93. Mearsheimer,"Back to the Future,"p. 14. See also JohnJ.Mearsheimer,GreatPowerPolitics

(New York:W.W. Norton,forthcoming), chap. 2.
94. Labs, "Beyond Victory,"p. 13.
95. Glaser, "The SecurityDilemma Revisited,"p. 145.
96. JackS. Levy observes,"Hegemony is a conceptthatis widely used, but is rarelydefinedwith
any degree of precision."See Levy,"Theories of General War," WorldPolitics,Vol. 37, No. 3 (April
1985), pp. 344-375, at pp. 348-349, n. 20. It is unclear whetherhegemony is synonymouswith
unipolarityor whethera hegemon can exist withina multipolaror bipolar internationalsystem.
Examples ofthe different definitionsofhegemonyemployed by realistsand nonrealistsin security
studies include Gilpin, Warand Changein WorldPolitics,pp. 29-30; Daniel Deudney,"The Philadel-
phia System:Sovereignty, ArmsControl,and Balance ofPower in theAmericanStates-Union,circa
1787-1861,"International Organization, Vol. 49, No. 2 (Spring 1995), pp. 191-228; BenjaminMiller,
"Competing Realist Perspectiveson Great Power Crisis Behavior,"Secuirity Studies,Vol. 5, No. 3
(Spring 1996), pp. 309-357, at p. 324; Charles A. Kupchan, "AfterPax Americana: Benign Power,
Regional Integration,and the Sources of a Stable Multipolarity," International Vol. 23, No.
Secutrity,
International
Security25:3 | 154
forhegemonic,long cycle,and power transitiontheories,which see world pol-

itics as a succession of hegemonicsystems.97Because states seek to maximize
relativepower and theconcentrationofpower promotesinternationalstability,
hegemonicperiods will be farmore common and of longerdurationthanbal-
ance-of-powertheoristsbelieve.98The power maximization argument does
presenta theoreticalproblemforotheroffensiverealists,such as Mearsheimer
and Labs, who argue thatstatesgenerallybalance againstincreasesin power.99
One way out of this problem is to make a distinctionbetween manual and
automatic expansion analogous to Inis Claude's distinctionbetween manual
and automaticbalancing.Manual expansion occurs when a statemakes a con-
scious bid to maximizerelativepower withthe aim ofbecomingthe dominant
statein the internationalsystem.Automaticexpansion,on the otherhand, oc-
curs when statesmake localized, incrementaleffortsto expand withthe aim of
Ifanarchyprovides incentivesforrel-
exploitinginternationalopportunities.100
ative power maximizationbut balancing is the norm,thenit only makes sense
forstates to engage in automaticexpansion.
CRITICISM 5: DEFENSIVE REALISM AND MATERIAL POWER

In their critique of contemporary realism, JeffreyLegro and Andrew
Moravcsikfaultdefensiverealism'sattentionto elitebeliefsand perceptionsof
materialpower. Legro and Moravcsik contend thatrealismis a rationalistre-
search program.The explicitinclusion of elite perceptionsand belief systems
effectivelyremoves defensive realism fromthe realist paradigm altogether.
They argue,"If the perceptionsand beliefsabout effectivemeans-endscalcula-
tions,given adequate information,consistentlyfail to correspondto material
power relationships,thenpower is at best one ofa numberof importantfactors
and perhaps a secondaryone. The parsimonyand coherenceofrealisttheoryis
eroded. When recentrealiststheorizethis relationshipmore explicitly,more-
2 (Fall 1998), pp. 40-80, at pp. 45-49; Michael Mastanduno and Ethan B. Kapstein, "Realism and
State Strategiesafterthe Cold War,"and Daniel Deudney and G. JohnIkenberry, "Realism, Struc-
turalLiberalism,and theWesternOrder,"bothin Kapstein and Mastanduno,eds., UnipolarPolitics:
Realismand StateStrategies aftertheCold War(New York:Columbia UniversityPress, 1999), pp. 3-
27, 103-137,respectively;and William C. Wohlforth, "The Stabilityof a Unipolar World,"Interna-
tionalSecurity, Vol. 24, No. 1 (Summer 1999), pp. 5-41, especially pp. 9-18.
97. Elman, "Horses forCourses," pp. 28-29.
98. GeoffreyBlainey,The Causes of War,3d ed. (New York:Free Press, 1988), chap. 8; and Miller,
"Competing Realist Perspectiveson Great Power Crisis Behavior,"pp. 323-325.
99. Labs, "Beyond Victory,"pp. 18-20; Mearsheimer,"Back to the Future." See also Eric J.Labs,
"Do Weak States Bandwagon?" SecurityStudies,Vol. 1, No. 3 (Spring 1992), pp. 383-416.
Re-
100. Elman, "Horses forCourses," pp. 28-29. See also Inis L. Claude, Jr.,Powerand International
lations(New York: Random House, 1962).
over,theyare forcedto borrow propositionsmore fullyelaborated in existing

epistemictheories,which theorizethe influenceof societal beliefs that struc-
ture means-ends calculationsand affectperceptionsof the environment."101
Legro and Moravcsik presenta flawed critiquein threerespects.First,most
defensiverealistsdo not claim thatstates' foreignpolicies or internationalout-
comes consistentlyfail to correspond to material power relationships.102 On
thecontrary, all variantsof contemporaryrealismassume thattheinternational
systemis mostly,although not exclusively,responsibleforstates' externalbe-
havior.Material capabilitiesshape the broad parametersof what can and will
happen in the internationalarena.
Over the long run,internationaloutcomes correspondto the relativedistri-
bution of materialcapabilities.In the shortrun,however,defensiveneoclassi-
cal realism expects an indirectand problematiccausal path between material
capabilities(boththegross distributionof power and structuralmodifiers)and
a state's foreignpolicy.As Aaron Friedbergnotes: "Structuralconsiderations
provide a useful point fromwhich to begin the analysis of internationalpoli-
ticsratherthana place at which to end it. Even ifone acknowledges thatstruc-
turesexistand are important,thereis stillthe question of how statesmengrasp
theircountersfromtheinside,so to speak."'103It is sensible,therefore, thatneo-
classical realisttheoriesspecifythe mechanism throughwhich policy inputs
translateintopolicy outputs-namely, the various diplomatic,military, foreign
economic, and national securitystrategies states actually pursue.
Materialcapabilitiescan influencestates' externalbehavioronly throughthe
medium of centraldecisionmakers'perceptions,calculations,and estimates.
Purely quantitativeindicatorsof capabilities simply cannot capture decision-
makers' assessments.Moreover,as William Wohlforthnotes, "All policies are
future-oriented. All decisions are bets on the future.A decision to reform,re-
trench,or go to war reflectsexpectationsabout futuretrendsand assessments
of the likely effectof today's policies on tomorrow'sdistributionof relative
power."104
Second, Legro and Moravcsik overstatethe extentto which one can classify
realismas a "rationalist"program.Classical realism,neorealism,and neoclas-
101. Legro and Moravcsik,"Is Anybody Still a Realist?" p. 35.

102. Van Evera contendsthatmost wars resultfromnational leaders' misperceptionsof the fine-
grained structureof power-in exaggerationof the power of the offense,the size of first-move
ad-
vantages,the size and frequencyof power fluctuations,and the cumulativityof resources.See Van
Evera, Causes of War,pp. 9-11.
103. Aaron L. Friedberg,The WearyTitan:Britainand theExperienceofRelativeDecline,1895-1905
(Princeton,N.J.:PrincetonUniversityPress, 1988), p. 8, n. 24.
104. Wohlforth, "Realism and the End of the Cold War,"p. 98.
International
Security25:3 | 156
sical realism have an ambiguous and tenuous relationshipto rationalmodels

of social behavior.Twentieth-centuryclassical realistsheld ambivalent(and of-
ten inconsistent) views on rationality.105 Consider the writings of Hans
Morgenthau.His six principlesof political realism adopt rationalreconstruc-
tionfromtheviewpointof politicalleaders as a way ofcomprehendingforeign
policy.He definespoliticalpower as a "psychologicalrelation"between weak
and strongactorsflowingfrom"the expectationof benefits,the fearof disad-
vantages, [and] the respector love formen or institutions."106 In a previous
book, Morgenthaucriticizesliberalismas a "repudiation of politics" and de-
clares,"Our civilizationassumes thatthe social world is susceptibleto rational
controlconceived afterthe model of the natural sciences, while the experi-
ences, domestic and international,of the age contradictthis assumption." 107
Similarly,JohnHerz notes an underlyingirrationality in human beings' simul-
taneous interdependenceand "necessityfordistrustingand possibly destroy-
ing" one another.108
Likewise, the microfoundationsof neorealismare unclear.109 Consider Rob-
ertKeohane's widely accepted claim thatrationalityis a hard-coreassumption
forboth classical realism and neorealism.110 Waltz, however,explicitlystates
thathis neorealistbalance-of-powertheory"requires no assumption of ratio-
nality,"and thatover time,the internationalsystemconditionsstatebehavior
through the processes of socialization and competition.111 In response to
Keohane's assertion,Waltz places "the notionof 'selection' in a positionof cen-
105. For the differencesbetween contemporaryrealism and the classical realism of Morgenthau,
Henry Kissinger,E.H. Carr,and Arnold Wolfers,see Richard Ashley,"The Povertyof Neo-real-
ism," International Organization,Vol. 38, No. 2 (Spring 1984), pp. 225-261; KennethN. Waltz,"The
Originsof War in NeorealistTheory,"in RobertI. Rotbergand Theodore K. Rabb, eds., TheOrigins
and Prevention ofMajor Wars(New York:Cambridge UniversityPress,1989),pp. 39-44; and Ashley
J.Tellis,"ReconstructingPolitical Realism: The Long March to ScientificTheory,"SecurityStudies,
Vol. 5, No. 3 (Spring 1996), pp. 3-104, at pp. 67-89.
106. Hans J. Morgenthau,PoliticsamongNations:The StruggleforPozverand Peace, 3d ed. (New
York:AlfredA. Knopf, 1966), pp. 1-6, 29.
107. Hans J. Morgenthau,Scientific Man versutsPozverPolitics(Chicago: Universityof Chicago
Press, 1946), p. 71.
108. JohnHerz, PoliticalRealismand PoliticalIdealism(Chicago: Universityof Chicago Press, 1951),
p. 16.
109. See Miles Kahler,"Rationalityin InternationalRelations,"International Vol. 52,
Organization,
No. 4 (Autumn 1998), pp. 919-942, at pp. 924-925; Brooks,"Dueling Realisms," pp. 453-455; and
Randall L. Schwellerand William C. Wohlforth, "Power Test:Evaluating Realism in Response to
the End of the Cold War,"SecurityStudies,Vol. 9, No. 3 (Spring 2000), pp. 60-108, at pp. 70-71.
110. Robert 0. Keohane, "Theory of World Politics," in Keohane, ed., Neorealismand Its Critics
(New York:Columbia UniversityPress, 1986), pp. 164-165, 173.
111. Waltz, TheoryofInternational Politics,pp. 74-77, 127-128.
tralimportance,"notingthat"one cannotexpect of politicalleaders the nicely

calculated decisions thatthe word 'rationality'suggests."112
Third,Legro and Moravcsik downplay the methodologicalreasons forex-
amining elite decisionmaking.For any foreignpolicy theoryto explain state
behavior,it must specifythe mechanismthroughwhich the independentvari-
able translatesinto policies.113Wohlforth'sresponse to criticsof realism's abil-
ity to explain the Soviet Union's peaceful decline is equally applicable here:
"Criticsof realismcontrasta simplisticview of the relationshipbetween [rela-
tive] decline and policy change against a nuanced and complex view of the re-
lationshipbetween theirfavored explanatoryvariable and policy change."114
By rulingan examinationof actual decisionmakingoutside the realistrubric,
Legro and Moravcsik effectively privilegenonrealisttheories.
CRITICISM 6: DEFENSIVE REALISM AND UNITARY RATIONAL ACTORS

Legro and Moravcsik argue thatthe inclusion of domesticvariables in defen-
sive neoclassical realism violates a core tenet of realism-the assumption of
unitary,rational actors existingin an anarchic environment.By rejectingthe
notion thatall states have fixedand uniformlyconflictualpreferences,defen-
sive neoclassical realistsmust relyon unit-levelfactorsto explain variationin
states' motivations.Legro and Moravcsikwrite,"Such explanationsinevitably
import consideration of exogenous variation in the societal and cultural
sources of state preferences,therebysacrificingboth the coherenceof realism
and approaching midrange theoriesof interstateconflictbased on liberal as-
sumptions."11'5To support this point, they cite Arnold Wolfers'sadmonition
against ad hoc extensionsof realism:"One consequence of distinctionssuch as
these [between status quo and revisioniststates] is worth mentioning.They
112. Waltz,"Reflectionson TheoryofInternational Politics:A Response to My Critics,"in Keohane,

Neorealismand Its Critics,p. 118. See also Waltz,TheoryofInternational Politics,pp. 74-77, 127-128.
For differentviews on whether,and ifso to what extent,Waltz's neorealistbalance-of-powerthe-
ory relies on rational choice as opposed to an evolutionary mechanism (or some other
decisionmakingmodel), see Elman, "Horses forCourses," pp. 42-44; Brooks,"Dueling Realisms,"
pp. 453-454; Joao Resende-Santos,"Anarchy and the Emulation of MilitarySystems," Security
Studies,Vol. 5, No. 3 (Spring 1996), pp. 193-260, at p. 209, n. 56; Scott D. Sagan, "More Will Be
Worse,"in Sagan and KennethN. Waltz,TheSpreadofNuclearWeapons:A Debate(New York:W.W.
Norton,1995),pp. 50-55, 86-87; and JackS. Levy,"Learningand ForeignPolicy: Sweeping a Con-
ceptual Minefield,"International Organization,Vol. 48, No. 2 (Spring1994),pp. 279-312,at pp. 296-
298.
113. AlexanderL. George and TimothyJ.McKeown, "Case Studies and Theoriesof Organizational
Decision Making," in RobertCoulam and RichardSmith,eds., Advancesin Information Processingin
Vol. 2 (Greenwich,Conn.: JAIPress, 1985), pp. 21-58, at p. 35.
Organizationis,
114. Wohlforth, "Realism and the End of the Cold War,"pp. 108-109.
115. Legro and Moravcsik,"Is Anybody Still a Realist?" p. 23.
International
Security25:3 | 158
rob [realist]theoryof the determinateand predictivecharacterthat seems to

give the pure power hypothesisits peculiar value. It can no longerbe said of
theactual world,forexample,thata power vacuum cannotexistforany length
of time."116
Legro and Moravcsik base theirsweeping critiqueof contemporaryrealism
on Imre Lakatos's methodology of scientificresearch programs (MSRP), al-
though they evade specifyingtheir philosophy of science.117 They write,
"When theoreticalexplanationof empiricalfindingswithina paradigm consis-
tentlyrelies on auxiliaryassumptions unconnected to core assumptionsto pre-
dictnovel factsor clear up anomalies,we learnlittleabout theveracityof those
assumptions.When it relieson auxiliaryassumptionscontradictory to underly-
ing core assumptions, our confidence in those core assumptions should
weaken."118
Debates over whether Lakatos's MSRP is even an appropriate standard
against which to judge internationalrelations and foreignpolicy theories,
Legro and Moravcsikfailto address the moreimportantquestion:Why should
one care? What are the empirical consequences of not adhering to Lakatos's
standard for judging scientificresearch programs?119 Legro and Moravcsik
faultthe inclusion of domestic variables in JackSnyder's logrollingtheoryof
imperialismand Joseph Grieco's application of defensive neorealism to the
studyof internationaltrade disputes,but theydo not show how such theoreti-
cal "degeneration"hinders our understandingof real-worldphenomena.120
Conclusions
Offensive realism predicts frequent internationallydriven expansion and

holds thatall statesstriveto maximize relativepower. Anarchycompels states
116. Arnold Wolfers,"The Pole of Power and thePole of Indifference," in Wolfers,Discordand Col-
laboration:Essays in International
Politics(Baltimore,Md.: JohnsHopkins UniversityPress, 1992),
p. 42.
117. On this point, see the lettersby GuntherHellmann and Jeffrey W. Taliaferro,"Correspon-
dence: Brother,Can You Spare a Paradigm? (Or Was AnybodyEver a Realist?),"International Secu-
rity,Vol. 25, No. 1 (Summer 2000), pp. 171-172, 179.
118. Legro and Moravcsik, "Is Anybody Still a Realist?" p. 9 (emphasis in original). In note 8,
Legro and Moravcsik cite Imre Lakatos, "Falsificationand the Methodologyof ScientificResearch
Programs,"in Lakatos and Alan Musgrave,eds., Criticism and theGrowthofKnowledge(Cambridge:
Cambridge UniversityPress, 1970), pp. 131-132.
119. For criticismsof the applicabilityof Lakatos's MSRP, see Stephen M. Walt,"The Progressive
Power of Realism," AmericanPoliticalScienceReviezv, Vol. 91, No. 4 (December 1997), pp. 931-935;
and Randall L. Schweller,"Correspondence:Brother,Can You Spare a Paradigm? (Or Was Any-
body Ever a Realist?),"International Vol. 25, No. 1 (Summer 2000), pp. 174-178.
Secuirity,
120. Legro and Moravcsik,"Is Anybody Still a Realist?" pp. 23-27.
to seek opportunitiesto weaken potentialadversaries and improve theirrela-

tive power positions. Defensive realism presents a slightlymore optimistic
view of internationalpolitics. States striveto maximize relative security,not
relativepower. The internationalsystemitselfprovides incentivesforexpan-
sion and aggressivestrategiesonly under verylimitedconditions.Statesoften
can achieve securityby pursuing moderate foreignpolicies.
This articlehas soughtto advance theintrarealistdebate in threeways. First,
it drew a distinctionbetween neorealism and neoclassical realism, both of
which have offensiveand defensive variants. Dividing realism along these
lines allows us to distinguishbetween different assumptions about the impli-
cations of anarchyand the empiricalrange of particulartheories.Second, the
articleexamined fourauxiliaryassumptions underlyingdefensiveneorealism
and defensiveneoclassical realism: (1) The securitydilemma is an intractable
featureof anarchy;(2) structuralmodifiersinfluencethe severityof the secu-
ritydilemma in particularregions or between particularstates; (3) material
power drives states' foreignpolicies throughthe medium of leaders' calcula-
tions and perceptions;and (4) domestic politics limitsthe efficiency of states'
responsesto systemicimperatives.Third,the articleresponded to several criti-
cisms raised by offensiverealistsand nonrealists.
As noted at the outset, the intrarealistdebate has implications for the
conductof foreignpolicy.Since 1991,theUnited Stateshas enjoyedpreponder-
ance in all underlyingcomponents of power: militarycapabilities, techno-
logy, geography,population, and economic resources. It currentlyfaces no
single great power competitoror hostile great power coalition. Unipolarity
may well last forseveral decades.121Over the long run,however,the relative
distributionof power will change, and new great power competitorswill
arise.
Defensive realism highlightsthe long-termperils for the United States in
pursuing short-term, unilateral,and potentiallyprovocative policies such as
the developmentof a national missile defense (NMD) systemand the pursuit
of "humanitarian"militaryinterventionsand "cruise missile diplomacy" in
othergreatpowers' spheresof influence.At firstglance,the costs and risksas-
sociated with these and other policies appear minimal. Leaving aside ques-
tions of technicalfeasibilityor even the existenceof a ballistic missile threat
fromNorthKorea, Iran,or Iraq, only the United Stateshas the resourcesto de-
velop an NMD. Only U.S. leadership and militarycapabilities (exercised
121. Wohlforth,
"Stabilityof a Unipolar World,"pp. 29-41.
International
Security25:3 | 160
throughNATO) could bring an end to war and ethnic cleansing in Bosnia-

Herzegovina and Kosovo. Last year,theUnited Stateswaged an air war in Ser-
bia forseventy-eightdays withoutsufferinga single casualty.
Althoughthesecurityrationalebehind thesepolicies may seem perfectlyob-
vious to officialsin Washington,othergreatpowers will worryabout a future
malign turnin U.S. intentions.The securitydilemma continuesto operate un-
der unipolarity:Steps the United States takes to enhance its securitywill de-
crease the securityof otherstates.The developmentand possible deployment
of NMD may shiftthe offense-defense balance (or at least leaders' perceptions
of the balance), thus negatingthe pacifyingeffectsof mutual assured destruc-
tion.This in turnwill forceRussia to place even greaterrelianceon its strategic
nuclear weapons, thusjeopardizingthe prospectsfora START III treaty.China
will likelyrespond to NMD developmentby increasingits numberofintercon-
tinentalballisticmissiles,which mightspark a nuclear arms race with India.
The U.S. victoryin the Kosovo conflictexposed the inadequacies of the West-
ern European states' militarycapabilities.The conductof the war,and the dip-
lomatictalksthatproceeded it,exacerbatedtensionsbetweentheUnited States
and its NATO allies and between the West and Russia. Another round of
NATO expansion would only increase Russia's sense of vulnerabilityand en-
circlement.AlthoughChina, Russia, Japan,and the European Union will not
be able to counterbalancethe lone superpower in the near future,the United
Statescan ill affordto alienate themin theend. Defensiverealismsuggeststhat
the next administrationcan best ensure U.S. securityin the twenty-first cen-
turythrougha strategyof selectiveengagement.
Beyond its policy relevance,the currentintrarealistdebate suggests at least
threeavenues forfutureresearch.First,both offensiveneorealistsand defen-
sive neorealistsshould be more explicitabout the assumptions that underlie
theirtheories.Second, both offensiverealismand defensiverealismshould de-
vote more attentionto the motivationsforexpansionistbehavior.122Defensive
realistsnote that states will sometimesengage in security-driven expansion,
but that over the long run, self-aggrandizementwill prove self-defeating.
Offensiveneorealists pay considerable attentionto expansionist states and
see them as the prime movers in internationalpolitics. The motivationfor
expansionistbehavior-greed or security-remains underdeveloped in both
camps.123 Third,both the offensiveand defensivevariantsof neoclassical real-
122. Rose, "Neoclassical Realism and Theories of ForeignPolicy,"p. 165.

123. Glaser,"The SecurityDilemma Revisited,"pp. 200-201.
ism hold thatperceptionsand misperceptionsamong top leaders may inhibita

state's abilityto respond to changes in relativecapabilities.The link between
objectivechanges in relativepower and leaders' perceptionsof relativepower
is underdeveloped. Futureoffensiveand defensiveneoclassical realisttheories
should be more explicit in incorporatinginsightsfromcognitive and social
psychology.124The debate between the two branchesof contemporaryrealism
over the implicationsof anarchyneed not lead to a permanentbifurcation.By
combiningdefensiverealism'sassumptionsabout structuralmodifierswithof-
fensive realism's assumptions about expansionist states, scholars might de-
velop more powerfulinternationalrelationsand foreignpolicy theories.
124. See JamesM. Goldgeier,"Psychologyand Security,"SecurityStudies,Vol. 6, No. 4 (Summer

1997), pp. 137-166.

The MIT Press Is Collaborating With JSTOR To Digitize, Preserve and Extend Access To International Security

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The MIT Press Is Collaborating With JSTOR To Digitize, Preserve and Extend Access To International Security

Uploaded by

Copyright:

Available Formats

Security Seeking under Anarchy: Defensive Realism Revisited

Author(s): Jeffrey W. Taliaferro

I wishto thankDale Copeland,BernardFinel,BenjaminFrankel,

1. For an overview of competingpost-Cold War U.S. grand strategiesderived fromoffensivereal-

Ititerniatioznal Vol. 25, No. 3 (Winter2000/01), pp. 128-161

througharms buildups, unilateral diplomacy,mercantile(or even autarkic)

Origins of America'sWorldRole (Princeton,N.J.: Princeton UniversityPress, 1998); Randall L.

that defensiverealism cannot explain state expansion because it argues that

6. Randall L. Schweller,"Neorealism's StatusQuo Bias: What SecurityDilemma?" SecurityStudies,

cousins, paradigms-are notoriouslydifficultto confirmor disconfirm.Not

10. RobertJervis,"Realism and theStudyofWorldPolitics,"International Organization,Vol. 52, No.

whichhave defensiveand offensivevariants.I divide realismalong theselines

Realist theoriesshare certaincore assumptions,but thereare two crosscutting

NEOREALISM AND NEOCLASSICAL REALISM

Neorealismis a body of internationalrelationstheorythatbuilds upon a few

spond to systemicimperatives.Phenomena such as individual states' grand

OFFENSIVE REALISM VERSUS DEFENSIVE REALISM

Four auxiliaryassumptionsdefinedefensiverealism.The firsttwo specifythe

20. Waltz, TheoryofInternational Politics,p. 75.

Table 1. Categoriesof Contemporary

aThe distinction betweenneorealismand neoclassicalrealismis best understoodas a con-

between externalforcesand the actual foreignpolicy behavior of individual

ASSUMPTION 1: THE INTRACTABILITY OF THE SECURITY DILEMMA

ASSUMPTION 2: STRUCTURAL MODIFIERS AND THE SECURITY DILEMMA

24. Securitydilemmas existin any anarchicenvironment,not just in the internationalsystem.See

ternationalsystem, other material factors,which I referto as "structural

27. For a discussion of structuralmodifiers,see Glenn D. Snyder,"Process Variablesin Neorealist

Consider,forexample, offense-defense theoryand balance-of-threat theory.

The defensive variants of neorealism and neoclassical realism specifythe

36. Glaser and Kaufmann,"What Is the Offense-DefenseBalance?" p. 64.

According to Mearsheimer,states must constantlyworry about theirsur-

(Princeton,N.J.:PrincetonUniversityPress, 1996); and Carl Kaysen, "Is War Obsolete? A Review

centivesforthe developmentof efficientstate institutionsto extractresources

ASSUMPTION 3: THE INFLUENCE OF MATERIAL CAPABILITIES ON FOREIGN POLICY

Making," in Tilly,ed., TheFormation ofNationalStatesin WesternEurope(Princeton,N.J.:Princeton

Finally,leaders' perceptionsplay a criticaland at times pernicious role in

ASSUMPTION 4: DOMESTIC POLITICS AND SYSTEMIC IMPERATIVES

50. Offense-defense theoryhas both an objectivecomponentand a perceptionalcomponent.The

Defensive neoclassical realism relies on a top-down conceptionof political

The fourauxiliaryassumptions discussed above definedefensiveneorealism

55. Andrew Moravcsik,"TakingPreferencesSeriously:A Liberal Theoryof InternationalPolitics,"

fensiverealistshave raised several theoreticaland empiricalcritiquesof each.

CRITICISM 1: THERE IS NO SECURITY DILEMMA

59. Schweller,"Neorealism's Status Quo Bias," p. 119.

Second, statesfacereal uncertaintynot onlyabout one another'spresentmo-

CRITICISM 2: GREEDY STATES AND INTERNATIONAL CONFLICT

62. Marc Trachtenberg,"A WastingAsset: American Strategyand the ShiftingNuclear Balance,

thatall stateshave completeinformationabout one another'spreferences.The

66. Ibid., p. 138. Kydd does not considerhimselfan offensiverealistand is broadlysympatheticto

which it could enforcestrictobservanceof its wishes ifit chose to,the security-

68. Kydd, "Sheep in Sheep's Clothing,"pp. 143.

Italyas a means to deterpotentialadversaries;and ultimatelythe decision for

76. Jeffrey W. Taliaferro,"Quagmires in the Periphery:ForeignWars and EscalatingCommitment

States may not always interpretan adversary's ideological moderationas a

UncertainPartners:Stalin,Mao, and the Korean War (Stanford,Calif.: StanfordUniversityPress,

Other parts of Kydd's model rely on essentiallyInnenpolitik argumentsto

85. JohnM. Owen, IV, "How LiberalismProduces DemocraticPeace," International Vol.

(or quasi-democraticstates) miscalculated each other's security-seekingmo-

CRITICISM 3: UNIT-LEVEL FACTORS, AGGRESSIVE BEHAVIOR, AND WAR

CRITICISM 4: THE SECURITY DILEMMA AND POWER MAXIMIZATION

89. ChristopherLayne, "Kant or Cant? The Mythof the DemocraticPeace," Interniational

cause onlythestrongeststatesstand thebest chance of survival.Statesseek op-