You are on page 1of 302



ISBN 0-321 - 43603-2

ATI:~NAL I T P: , ITI .__.. ._

Enduring Concepts and Contemporary Issues

Brandeis University

Columbia University



:\ew York San Fr.mcisco Boston London Toronto )-dne~ Tok:o Singapore ~ladrid ~ l cxi"'O Cttv ~I much Paris Cape T0\\11 Hong Kong ~l ontreal ,

Editor-in-Chief: Eric St:mo Senior ~larkt>ting ~lanagC'r: Elizabeth Fog~ Production ~lanager: Denise Phillip Pro_iect Coordination. Tt>.\1 Design. and Electronic Paue ~lakcup: tratford Publishing
Senices Con~r Designer/~lanager: John Callahan

Detailed Contents Preface PART 1


Con'r 1mages: Left image rottrtesy of Planet Art. Right image courtesy of PhotoDi c Senior Manufacturing Bu~-er: Dennis J. Para Printer and Binder: R. R. Donnelley & Sons Cover Printer: Phoenix Color Graphics




Power and Principle in Statecraft The Consequences of Anarchy The Mitigation of Anarchy


For pemliSl ion to use cop~Tighted material. grateful acknowledgment is made to the cop~Tight holders on the first page of each selection. which are be re b~ made part of this cop}Tight page.



ubrary of Congress Cataloging-in-Publication Data

Intemational politics: enduring concepts and conte i ues I [edited bv] , Robert J. :\ rt, Robert Jenis. -8tb ed. p. cm. locludes bibliographical references.
I SB~ 0-321~2 , alk. paper



The Political Uses of Force The Political Utility of Force Today The Spread of Nuclear Weapons



l. International relations. 2. World politics-1989- 3. Globalization. I. Art, Robert J.



Perspectives on Political Economy The Meaning of Globalization

309 347

11. Jenis, Robert. 1940JZ1242.15i4 200i 327.1-<k:22 20060 1543-S

The Pros and Cons of Globalization


~'Tight C 200"i ~- Pearson Education. Jnc.

So All rights reserved. part 0 f this publi may be reproduced. stored in a retrieval anon _ . . system . ar- banstuitted. m anv form or b-. anv _means, e1 ec.tromc, mechanical. photocop~mg. _. . recocd; 1 ~orotber.vise ithout the pnor written permission of the publisher. Printed in ~ the United Sbtes.


Conflict, War, and Terrorism The Uses of, and Reactions to, American Power Failed States, CiviJ Wars, and Nation-Building The Environment and Climate Change New Actors and New Forces

375 417 451 495 517


345178SHG--Doc-.oe 08






Anarchy and Its Consequences




Six Principles ofPolitical Realism


A Critique ofMorgenthau's Principles ofPolitical Realism




The Anarchic Structure of World Politics 29


Anarchy and the Strugglefor Power 50


Anarchy Is What States Make ofIt 61



The Conditionsfor Cooperation in World Politics 69


Kant Liberal Legacies, and Foreign Affairs 83


Alliances: Balancing and Bandwagoning 96


The Future ofDiplomacy



, iii



The uses and Umits ojlntemational Law





The International Political Economy


tnten?ational Institutions: can Interdependence Work?





The United Nations and International Security 12?


The Nature ofPolitical Economv 267 -'


The Uses of Force

Hegemony in lhe World Political Economy 283

141 149




The Great Divide in the Global Village 296



The Four Functions ofForce


Globalization ofthe Economy 309


The Diplomacy ofViolence


Will Globalization Swvive? 325

Coercive Diplomacy

163 177


Globalization and Governance 335


O.ffense, Defense, and the Security Dilemma



What Is Terrorism?


Tradjng in Illusions 34 7



Globalization's Missing Middle 355


The FUngibility ofForce 205


Why the GlobaJjzation Backlash is Stupid 361

The Strategic Logic ofSuicide TeJTorism


221 239


Contemporary World Politics




Nuclear Instability in South Asia 239


The Era ofLeading Power Peace 375


Nuclear StabiliOJ in South Asia 250

The Clash ofCMlizations? 391



Why Do They Hate Us? 406









Explaining the Bush Doctnne 417


The Rise ofChina: GetUng the Questions Right



Europe Without llfusions 526 432


can Democracy Stop Teaonsm.

Transnalional Activist Networks


Taming American Power 44 I



NGOs: Fighting Poverty, Hurting the Poor 539


Human Rights in World Politics


546 558

Failed States, Collapsed States, weak States: Causes and Indicators



The Five War.s ofGlobalization


Possible and Impossible Solutions to Ethnic Civil wars 459


The Global Governance ofthe Internet: Bringing the State Back In 567

IntemaUonal Lallv: The Trials ofGlobal Norms

58 I

Nation -Building: UN Surpasses u.s. on Learning Curve 481


Deconstructing Nation Building 490



The Iragedy ofthe Commons



Environmental Changes as causes ofAcute Conflict 50 I


Beyond Kyoto 508


T he Hrst ctli tion of Jnten wtional Politics appeared in 1973. Since then. the field of international relations hac; expe rienc.:cd a dramatic.: enric hme nt in the subje ct:, st11 dicd and the qmJi ty of works pu blisiJ ed. Poli tical economy c.:ame into its own as an itnportant subfi eld in the l 970s. New and itnportant works in the fiekl of security studi es appeared . The li terature on cooperation among states Aourished in the early 1980s, and impo rtant studi es about the environment began to appear in the mid-l 980s. Feminist, post-modernist, and constructivist critiques of the mainstrea m made their appearanc.:e also. w ith the end of the Cold War, these new issues came to the fore: hum an rights, the tension between state sovereignty and the obligations of the international communi ty, the global environm ent, chil wars, failed states, and nation-buildin g. The growing dive rsity of the field has closely mirrored the actual developm ents in international relations. As for the previous editions, in fashi oning the eigh th, we have kept in mind both the new developm ents in world politics and the }jterature that has accompanied them. Central to this ec.lition, though, as for the othe r seve n, is our belie f that the realm of internatjo nal politics diffe rs fundame ntall y from that of dom estic politks. Therefore, we have continuec.l to put both the developm ents anc.l the literature in the context of the patterns that stiJJ remain valid for unde rstanding the differences between politics in an anarchjc environment and politics that takes place under a government. As in the previous seven, the theme for thjs edition continues to revolve around enduring concepts and contemp orary issues in world politics. The eighth edition retai ns the four major subdivisi.ons of the seventh edition. We have leA: Part One as it appears in the seve nth edition. Part Two retai ns the three subsections of the seventh editi on, but has a new selection by Bruce Hoffman on teiTorism. Part Thre e has two new selections on globalization -on e by Martin Wolf and another by GeofTrey Garrett. Most of the changes in the eigh th edition come in Part Fou r. We have retained the four subdivisions of the seve nth edition but have added eleven new selec tions: articles by Fare ed Zakaria, Gregory Cause.' Stephen Walt, James Dobbins, James Payne, Thomas Hom er-Dixon, John Browne, Richard Betts and Thomas Christensen, Andrew Moravcsik, Sebastian Mallaby, and Daniel Drezner. The eighth edition of International Politics has fourt een new selections and is a little over 25 percen t new, but it continues to follow the four principles that have guided us thro ughout all previous editions: I . A selection of subjects that, while not exhaustively <.."'ve ring the field of intematiomJ politics, neverthe less encompasses most of the esse ntial topi<:s that we teach in our introductory courses.



. gs tlnt are mainh- anahtical in ton lt'nt tl nt hke 1 ssuc \' h '" 2 1ndh;dua1reldm the tudent to tht fun dame a1 'lt . d th;lt t11ereb, introduce nt . 1 6 Id u?r. one anot 1 ,tn . debates and point ohiew in t 1' Je . . . . troductions to each [)art that summan'le tlw <:vntral <.:once t P s the , . . 1tors Lll 3. E<I ., "'ter that oraamze the centr .u the me.!> ol t'<lC:h })ut , .. nd t1 0 1a1 < St stucIen t mu. lll ...., . . relate the readings to one another. _ :\reader that ran be used eithe~ a the core around" luc:~1 to dc.'sign an 4 . t od ,.t ... . course or as the }JI1111aJY supplemc' nl to ennch .tn as icrned text ~ m r lh. 0 : Finally. in putting togetl1er thi. and predous editiOJ.lS. \\'C n'cei\'ed exeellent achice frol1l the tollo,,inl1 colleagues. whom we would like to thank for the ti llle I . . Ge01:get o"-n. U n~,er ~: r 11not 1y McKeown ,md cart> they took: AneIrew Ben nett. Unh-ersity of ~orth Carolina at Chap~ I H ~: Roshn Suno\\'117., V n hersity or Tex~ at Arlington: Robert J. GJiffiths. Um,ersity of North Carolina at Greensbor . Linda s. Adam~. Baylor Unhersity: Timoth~ M. C~le. Unive rsity of i\ laine: Robe~ c. Gray, Fmnkun & \lar hail College: j<UJleS A. M1tc.: hell, California State University. t\ortluidge: ~l argaret E. Scranton. UniYersity of Arkansas at Little Rock ecker. Dartmouth College: James A. Caporaso, University of \Vashin ~ Da,;d G. B ton: Ken \ \'ise. Creighton {;nj,ersity: Sonia Gardenas. Trinity College: Philf p Schrodt. lJnhersity of Kansas: and Jane Cramer. Unher i~ of Oregon.




U nlike dom estic politics, international politics takes place in an aren a that has no central gove rning body. From this central fact Row impo rtant cons eque nces for the behavior of states. In Part One, we explo re three of them: the role that principle and morality can and shou ld play in state craft; the effects that anarchy has on how state s view and relate to one anot her; and the ways th at the harsh er edge s of anarchy can be mitigated, even if not wholly re moved.


Citizens, students , and scholars ahke often take up the stud y of inter natio nal politics because they want their coun try to behave in as princ ipled a way as possible. But they soon discover that principle and powe r, morality and state craft do not easily mLx. \Vhy shou ld this be? Is it inevitable? Can and shouJd state s seek to do good in tl1e world? Will they e ndan ger them selves and ham 1 othe rs if tl1ey try? Thes e are timeless ques tions , having been asked by obse rvers of inte mati onal politics in nearly ever y previous era. They there fore mak e a good start ing poin t for thinking about the natu re of international politics and the choices states face in our era. Hans J. Mor gent hau, one of the leading prop onen ts of the appr oach know n as Real ism (also known as Powe r Politics), takes the classic Realist position: unhersal stand ards of morality cann ot be an invariable guid e to state craft because there is an 'ineluctable tension between the moral com man d and the requ irem ents of successful political actio n." Rath er than base state craft on morality, Mor gent hau argues that state acto rs must think and act in te m1s of powe r and must do what ever it takes to de fe nd the national inter ests of their state . J. Ann Tickne r, com men ting on the primacy of pow er in Morgent hau's writings. explains tl1at "vhat he cons ide rs to be a reali stic desc ripti on of inte rnational politics is only a pictu re of the past and there fore not a prediction abou t the futu re, and prop oses what she cons iders to be a fe minist alter native. A world in which state actor s think of pow er in te rms of collective e mpo werm ent, not in te rms of leverage over one anot her, coul d prod uoo more cooperatjve outc ome s and pose fewe r conflicts be tween the dicta tes of morality and the pow er of self-interest.



E,en those who argue that. 1:10r~ity sh~uJd play a. large role in statecraft a , edge that in ternational po!J hcs IS not like dom esttc politics. rn the latt, cknowl. ao~ernme nt: in the forme r, the re is none. As a <..:onsec1uenc;c no . er, there i.~ " 'th . ' agency e . abo"e the individual states "~ authonty ~n cl .PO'-'''er to make laws and settleXt~ts . dt~. [Jute . States can make comrmtments and tr eaties. but no sO\ereign po\ . . . s. T1 . compliance and pumshes deVIaoon 11s-t I1e a b sence of a supremever eusu res . . . . power--i wl1at is meant by tI1e anarc h1c env:u onmen t o f mte rnationaI politics A h ~ . . narc therefore said to constitute a state of wctr: \vhen all else lails, fo rce is the tl.)' ts ratio-th e final and legitimate arbiter of wsputes among states. l tuna The state of war does not mean that every nation is constantly at the bri k war or actuaUy at ,.var with otbe_r nations. Most countJies, though , do feel th~ea~: ened b~ some _stat~s at some .hme, and evet) ' state has exp e ri enced periotls of intense msecunty. 1\o two contiguo us states, moreove r, have had a history of dos friendly relations uninterru pted by severe te nsion if not outright war. Becausee~ nation cannot look to a supreme body to enforce laws, nor coun t on other nations for constant aid and support, it must rely on its own eff01t s, particularly for defense agail1St attack. Coexiste nce in an anarchic environm ent thus requires self-help. The psychological outlook that self-help breeds is best describe d by a sayi ng common among British statesme n since Palme rston: "Great Britain has no pe rm anent enemies or permane nt friends, she has only permane nt inte rests." Although states must provide the wherewi thal to achieve their own ends, they do not always reach their foreign policy goals. The goals may be grandiose; the means available , meage r. The goals may be attainable; the means selected, inappropriate . But even if the goals are realistic and the means both available and appropriate, a state can be frustrate d in pursuit of its e nds. The reason is simple but fundame ntal to an understanding of inte rnational politics: what one state does will ine..,itably impinge on some othe r states-o n some be neficially, but on others adve rsely. What one state desires, anothe r may covet. What one thinks its just due, anotl1e r may find threate ning. Steps that a state takes to achieve its goals may be rende red useless by the countersteps others take. No state, there fore, can afford to disregard the effects its actions will have on other nations' behavior. In this sense state behavior is continge nt: what one state does is dep ende nt in part upon what othe rs do. Mutual depende nce means that each must take the othe rs into account. Mutual depende nce affects nothing more powerfully than it does securitythe meast_res states take to protect their ter ritory. Like othe r foreign-policy goal_, I s ~e secunty of one state is contingent upon the behavior of other states. HereiJ) lies the se:unty dilemTYUJ to which eac:h state is subject: in its e fforts to preserve or enhance Its own security, one state can take measure s that decrease the secwity of other states and cause them to take counterm easures that ne utralize the actions of the first sta~ and that ~ay even menace it. The first state may feel impeJled to take ~rth~r actions, provoking additional counte nneasures . . . and so fortl1 . The secunty dilemma means that an action-reaction spiral can occur between two states or among several of the m ~ h orcmg eac to spend ever larger sums on arms to be no more secure than befo All will c re. run 1aster me rely to stay where they are.
r\l the hf:'<~rt of the


security dilem ma are these two constraints: the inherent dil fit ultv in dtstingub hing betwe<:n offensive and defe:nsi,e postures. and the 111 abil:t: >f one .;t<ltC: to believe or trust that another state's present pacific intentiom ..;;11 remaiu so . The capability to defend eau also provide the capability to attack In addiug to its arms. state A may know that its aim is defensi\'e . that its intf'ntions are pcac:c:J'ul. and therefore that it has no aggressive designs on state B. 1n a world whe re states must Jc,ok to themselves for protection , however, B will e:-xamine A's a<:tions c:arehdly ami snspi<:iously. B may th ink that A will attack him when 1\s arms beeorne powerful enough and that Ns protestations of friendship are designed to lull him into lowering his guard. But even if B believes A's actions are not directed against him, B cannot assume that Ns intentions \\ill remain peacefuL Anan;hy makes it ilnpossibl e for A to bind itself to eontinuing to respect B's interests in the F uture. B must allow for the possibility that what A can do to hi m, A sometime might do. The need to assess capabilities along with intentions , or, the equivalent, to allow for a change in intentions, makes state actors profoundly conservative. They prefer to err on the side of safety. to have too much rather than too Httle. Because security is the basis of existence and the prerequisite for the achievement of all oth~r goals, state actors must be acutely sensitive to the security adions of others. The security dilem ma th us means that state actors cannot risk not reacting to the security actions of other states, but that in so reacting they can produce circumstances that leave them worse off than before. The anarchic environment of international politics, then, allows every state to be the fi nal judge of its 0\\111 interests, but requires that each provide the means to attain them. Because the absence of a central authority permits wars to occur. secu rity considerations become paramount. Because of the effects of the secwity dilemma, efforts of state leaders to protect their peoples can lead to severe tension and war even when all parties sincerely desire peace. Two states. or two groups of states, each satisfied with the status quo and seeking only security, may not be able to achieve it. Conflicts and wars \vith no economic or ideological basis can occur. The outbreak of war, therefore, does not necessarily mean that some or all states seek expansion , or that humans have an innate drive for power. That states go to war when none of them wants to, however. does not imply that they never seek war. The security dile mma may explain some wars; it does not explain aU wars. States often do e>.perience conB of interest over trade, real estate. ideology, and preskts tige. For example. when someone asked Francis I what differences led to his constant wars \vitl1 Charles V, he replied: '' None whatever. We agree pe rfectly. We both wan t control ofltaly!'' (Cited in Frederick L. Schuman,l nternational Politics. 7th ed. , New York, 1953, p. 283. ) If states cannot obtain what they w;mt by blackmail, bribery, or threats, they may resort to war. Wars can occur when no one wcmts them; wars usually do occur when someone wants them. Reallits argue tl1at even under propitious circumstances, international cooperation is difficult to achieve because in anarchy, states are often more concerned \\ith relative advantages than with absolute gains. That is, because international politics is a self-help system in which each state must be prepared to rely on its own resources and strength to furth er its interests, national leaders often seek to become more powerful than their potential adversari es. Cooperation is then made difficult not only by

1"1\K I 1

/\N/\ 1'(\.n I /'\ I'< V


~VI'<Jl ~UL I'I\..C::>

tht lc.\tr l>llwr:> " ill dt(': tl ~111d luil to liw np lc1 tll<'ir a~e, nwnts . . . . rrl . . . 1 t also I llt l'~~-''''-'h {'d tll't'd to ~tin a sup eno r J~<>stllon . w rv;L, on 1" not that state Ut:tor . 1y th~: l't't1 h'<i \\it h tatu:-. hut that Lite~ kar tktt amm gcments whi th bcn dit ll s lre<:o11, ,idt ~~atl'r hcmfits to othe rs than to thcttl. will n nder their <:oun tr,, vll' btll pr 0 ' nnd coerdon in I r 11 ncr1bl e lo pn'. ~un. t w utttr<. . . 1\ennt'th ~ \\'al tz d<'H 'IO{)S the ubo\'t' pollltS Ill ore full) bv ' ll ll],r . . . . . . ' < ';:t. . le rt>twc between hicrarcl uc (dotnC' ttc:) and <tnarcht.c (tnternational) ptngth c dtf. .. 01 ~ I I. )UtJo t '.", 111 IJ e shows " 'In tl C cJstn1 n o rc::tpa b'l't 'tc>s ( tIlf' n.. Ialive [)Owct ttt<:al sYs. 11 . post :~'tates ) in anarc:hi<: systems is so itnpott<H and lays o ut tlw wa 'S in whic:h. ti of .tl <~~~~ bchado r difle rs in hi<nm:hi<: and anarc:luc s~st<ms. P lhcaJ Thtre is broad agre eme nt at nong Realists on the ~onsequcnc:es of anar<:h t tatt's behc:wior, but not tota l agre ement. On<.> bran t_ of H '<Uists , who are callccLr l "ofl(msive Rt~<ilists ... argu<.' I hat t~te con seque~1ces o~ anarch~ go rar beyo nd protlu1 ~ ing securi~ ilile mm as and m~ ng c:oopt rabo n h<trd to com e by _They asstt't th~t anarchr forces states. and esiJec:tallv the grea t po" ers. to bec ome 'power tn,. . . . . ..,,Jil\tzers" because the only w~y t~ assu re. the sta~e_s sccut~ty ts to he: tl:_e most powerful state in the syst em. Ofle nsJVe reah sm env1s1ons a dog -eat-dog world of internationa l politics in whi ch power and rear uo minate great powe r interactions and in which war, or the threat or war, amo ng the grea t pow ers or amo ng thei r proxies is a constant feat ure of inte mational rela tion s. Joh n J. Mea rs heirner lays out the tenets of tl1js brand or Realism . In an anarchic conditio n, howe,er. the que stio n to a k may not be, ''\Vhy does war occu r?'' but rather ''\Vhy uoes war not occu r mor e rreq ucn tly than it does?'' Instead of asking "\Vh y do state s not coo pera te more to ach ieve common interests? we should ask ''Given ana rchy and the secm i ty dile m ma, how is it that states are able to cooperate at aJJ?'' Ana rchy and tl1e security cWe mma do not produce thei r effects automatically, and it is not selr evid ent that stat es are power maxirnizers. Thus Alexande r \Ven dt argues that \Valtz and oth er Rea lists have missed the exte nt to which the unpleas ant patterns they describe are "soc ially w nstructed''i.e., stem from the actors' beliefs, pe rceptions, and inte rpre tations of others beba,ior. If national lead ers believe that ana rch~ requ ires an asse1tive stance that endangers others. conflict will be gen erated. But if they thin k they have more freedom of action and do not take the hostility of othe rs for gran ted, they may hr able to crea te more peaceful relationships. In this view, stru cture (ana rchy) does not dete rmi ne state actio n; agency (hum an decision ) doe s.



Even Realists note that conflict and warfare is not a constan t cha racteristic of ~temational politics. Most states remain at pea ce with most othe rs most of ~~e ~e . State actors have developed a num ber of ways of cop ing witl1 ana rchy, of g.un mg more than a mcxlicum of security, of regulating thei r com petition with other states, and of developing patt erns that contain , but do not elim inat e, the dangers of


~roup th,lll to~t'llt<'r, thcrc:h~ reproducing ~omc of the advantages of ~mall 1111111 hers. The conditions acto rs l'ac:e also in(hwnc:e their fates. The ban iers of anarchy are 111 orc likel v to be O\'Ncomc when actors have long time horizons, when even succcssrt~llv ~xploiti 11g othc rs produces an outcome that is onJy < little better than t mutmJ ' cooperation. wlwn being ex-ploited by others is only slightly worse than mutmtl noncoope ration. nnd vvhe n mutual cooperation is much better than unre stric:ted competition. Und er such circum tances. states are particularly Ukely to unde 1take wntingcnt strategies such a<; tit-for-tat. That is. they will cooperate with others if othe rs do li kewise and refuse to cooperate if' others have refused to cooperate witl1 them. Mos t strikingly. it appears that democracies may never have gone to war agai nst each othe r. This is not to say, as Wood row Wilson did, that democracies are inhe rently peace ful. They seem to fight as many wars as do dict atorships. But. as Michael \11/. Doyle shows, they do not fight ea<.:h othe r. If this is corr ect- and . of course. both d1e evidence and the reasons are ope n to dispute it implies that anarchy and th.e seculity dile mm a do not prevent peaceful and even harmonio us relati.o ns amo ng states that share certain common values and beli efs. Democracies are relatively recent developments . For a longer peri od of time . two specific dev ices -inte rnational law < dipl omacy- have proven useful in lml resolving conflicts among states. Although not enfo rced by a world gove rn ment. international law can provide norms for behavio r and mechanisms for settling disputes. The effective ness of inte rnational law deri ves from the willingness of state s to observe it. Its power exte nds no furth er than the disposition of states 'to agre e to agree."' Whe re less than vital inte rests are at stake, state actors may accept settleme nts that are not entirely satisfactory because the~ think the precede nts or principles jus ti~~ the compromises made. Much of inte rnational law reflects a consensus amo ng states on what is of equal benefit to aU. as, for example. the rules regulating inte mational t'Ommunications. Oiplomac), too, can faciHtate cooperation and resolve disputes . If diplomacy is skillful, and the legitimate inte rests of the parties in disp ute are take n into acco unt, understandings can ofte n be reached on issues that mig ht othe rwise lead to war. These points and othe rs are explored more fully by Stanley I Ioffm ann < H<ms J. Morgenthau . Uld National leaders use these two traditional tools within a balance-of-powe r syste m. Much muligned l~y Presiden t Wilson < his followers and misunde rstood br many md othe rs. balanc:e ol power refers to the way in which stability is achieved through the conflicting efl01ts of .individual states, \vhether or not anv or all of them deliberatelv ' I pursue that goal . Just as Adam Smith argued that if every indhidual pursued his or her

tb \ ( ''' .,J,ow' l k .lt, w11 il anardt) .111d tlw ~<.!tnrily dile mm a inhi~)il lOOP r,1t 1011 ltt, t 1 no' pn v 1rl 11 lllllf llll'r of < \ ns an<~ na~ional ~lral~W~!> can m.1kc> it 1 ,1:>it1 Jo1 ..., 11, a< h w-. c com m on tnt k Coo pera tton .'s usual!) easH :r il' tiH'n ,m a :; 111 111lutunhcr of actor~. i\ol onlyc:an (acl1 tnore card sLily ubsen:e th~. otlu r' hut. 111 t< tol' kno '' that tlwir irnpa t:l on tlw s~ste tn is great enough so that t1 tltt') fail to coopcralt with otlwrs. jom l cn tc:rp rir;cs arc likely .to fail. Fu.rth~ml~>rC .. wltt' ll tlw numl>tr or a<:tor!. is IHr~e. there may be mechamsm!. ClOd ltiSh tutto ns




sl'l f:.intt.-rel'\l, the interaction of i ndiviclual C:'gois m~ would t'nhar t<.:c n . \\t'l~lh, S() illtC?nleltiOilal reJatl011S theOriSts have <ll'~(~ lhUl l'V(-'11 if <.: Vel)' Stat;lhOI)aj }X)wer at the expense of the others, no one stat~ '"rdl. lt k?l~ ~~minate. ln hotJ1 C:eck~ sa ueneral good can be tl1e unintendeJ product of selfl"h llldtVtd ual a<:tions M . I Or(!OV(! . tJ \ . t} t h. I. . k - I . . ~ (;'e,en if mo t states desm" otuy to ~ep Wtlat le) ;we...1< 11 own llttc rcsts dictate tl r, the, b~:md togetJ1er in order to resist any state or coaution of states that thre t )at a ens to . dominate them. The balance-of-power system is likely to prevent any one state's a<..:qnitin h mom. It will not, however, benefit all states equally nor maintain the peace~ egc. I Wa 1 nently. Rewards will be unequaI because of mequal'ti es tn power an cl c~:pe ttise.erma. "'rill occur because they are one of ilie means by whkh states can presetve what th rs have or acquire what they covet. ~mall states m.ay even be eliminated by their rno~~ powerfUJ neighbors. The intemational syst~m "'ilJ be ~nstable, however, only if states hat ock A to what they tl1ink is the strongest s1de. \i\T IS called handwagoning or th domino theory argues tl1at the international syste1~ is precarious because successfu~ aggression will attract many followers, either out of fear or out of a desire to share the spoils of victory. Stephen M. Wait disagrees, dra,ving on b<~lance-of-power theory and histOJical evidence to argue that, ratl1er tl1an bandwagonmg, under most conditions states balance against emerging threats. They do not throw in th eir lot "'~th the stronger side. Instead they join witl1 otl1ers to prevent any state from becorniJ)g 50 strong that it could dominate the system. Power balancing is a strategy followed by individual states acting on their own. Oilier ways of coping \vith anarchy, which may supplement or exist alongside this impulse, are more explicitly collective. Regimes and institutions can help overcome anarchy and facilitate cooperation. When states agree on the plinciples, ruJes, and norms that should govern behavior, iliey can often ameliorate the security dilemma and increase the scope for cooperation. Institutions may not only embody common understandings but, as Robert 0 . Keohane argues, they can also help states work toward mutually desired outcomes by providing a framework for long-run agreements, making it easier for each state to see whether oiliers are living up to their promises, and increasing the costs the state will pay if it cheats. . In the .sec~ri~ area, the United Nations has the potential to be an especially t~portant.lOsti~~on. Adam Roberts assesses the United Nations' role in coping with confltct wttlun states through the mechanism of peacekeeping missjons. He finds that the demand for such missions since the end of the Cold War has expanded dramatically, but notes the formidable obstacles that must be overcome if ~e UN is to fulfil! the hopes that so many state leaders and citizens have for it in this area.



Six Principles ofPolitical Realism


1. Politic:al realism believes that politics, like society in general , is governed hy

objective laws that have their roots in human nature. [n order to improve society it is first necessary to understand the laws by which society lives. The operation of these laws being impervious to our preferences, men will challenge them only at the risk of failure. Realism, believing as it does in the objectivity of the laws of politic , must also believe in the possibility of developing a rational theory that reflects. however imperfectly and onc-sidedly, these objective laws. 1t believes also, then, in the possibility of distinguishing in politics between truth and opinion-between what is true objectively and rationally, supported by evidence and illuminated by reason. and what is only a subjective judgment, divorced from the facts as they are and in formed by prejudice and wishful thinking. Hu man nature, in which the laws of politics have their roots, has not changed since the classic:al philosophies of China, [ndja, and Greece endeavored to discover these laws. I Ience, novelty is not necessarily a virtue in political theory, nor is old age a defect. The fat:t that a theory of politics, if there be such a theory. has never been heard of before tends to <:reate a presumption against, rather than in favor of, its soundness. Conversely, the fact that a theory of politics was developed huudreds or even thousands of years ago-as was tJ1e theory of ilie balanc:e of power--does not create a presumption Ulat it must be outmoded and obsolete .... For realism, theory consists in ascertaining facts and giving U1em meaning through reason. It assumes that the character of a foreign policy can be ascertained

fro111 .PoliliCII Among Nations, 5th edition hy llans

l9i2 bv Alfred A. Knopf, .a division of Randc,m llou!.e, Inc. Reprinted by pennission of Alfred A. Knopf. a di\ision of Hundom Hou~c. Inc. Portions of the text and some footnotes have been omitted.

J. Morgeuthau. Copyright C


HY AND ITS CUN:":It.\,lUt:l'l\...t:..:>


of the politic<tl ads JWrfont <d and of Ll tt' ~< >resPe b! rl IlJ'()tl\! 1 tltc. find oul what ~l. l<sm< 11 hav . a c . t Thus we can tmh t < e U<.:tu ll 1 e ,le !\. . I . . , ''s o1 t tc t'Otl~t qm ne< . I (I eseeable consequences of l lC't r acts '' <' can su rn1 . ~ a y llie what Jom'. and rmm ( le or . . . . . ob~.-cth cs mi"ht ha"e been. . . . t> f the facts is not enough. To gJVc ' neanmg to the fac:t . I thetr. ~ . . ) et e,amul.lhOll o 1 . I rea1. '"''tI1 a kind of na. ra,v th Jc . we must approach po JtJca . . ratJonal . (:f. . . I nrttcrial of foregn po ' ) t . ggests to us the poss1b e meamngs o oretgn poli<:y. 1 . 1 n Ol1er Hl outiJJle a map t 1 su 1 must meet a c . . 10 elves in the position of a state man " cl ' <;rtam . word we put ours . an W<' ask ourselv"s what r pro bJem of .oreJgn policv under certain Circumstance st alten'ltti'ves are from which a statesman may chome who mu 111 eet . . t he rationa1 ' this roblem under these circumstances (presurn~g alv.~ays tl~al he acts in a . p . . . ratJona1 man ner), a11d which of these rational altemahvcs this p<HtJC:u lar statesrnan s acting under these circumstances. is Hkely t~ choose. It 1 the test m~ of this ration~ hypothesis against the. actual [~cts and .U:er consequences that gvcs theoretical . . meaning to the facts of international poh~~s. 2. The main signpost that helps pohtical reaiJsrn. to find ls way through the landscape of intemational politics .is the concept of mte~est cle An ed in terms of power. This concept provides the lmk between reason tJ)'ll~~ to understand international politics and the facts to be understood. It sets pol1tcs as an autonomous sphere of action and understanding apart from other spheres, such as economics (understood in terms of interest defined as wealth), ethics, aesthetics, or religion. Without such a concept a theory of politics, international or domesti<.:, would be altogether impossible, for without it we could not distinguish between political and non political facts, nor could we bring at least a measure of systentati c order to the political sphere. We assume that statesmen think and act in terms of interest defined as power, and the evidence of history bears that assumption out. That assumption allows us to retrace and anticipate, as it were, the steps a statesman-past, present, or future-has taken or will take on the political scene. \Ve look over his shoulder when he writes his ruspatches; we listen in on his conversation with other statesmen; we read and anticipate his very thoughts. Thinking in te rms of interest defined as power, we think as he does, and as disinterested observers we under stand his thoughts and actions perhaps better than he, the actor on the political scene, does himself. Tile conc~pt of inter~st defined as power imposes intellectual discipline upon the observer, mfu~es rational order into the subject matter of politics, and thus mak~s the theoretical understanding of politics possible. On the side of the actor, it fro~des fo.r rational discipline in action and creates that astounding continuity in roreign policy which makes A mencan, Bntis h or Russian fo re ign p olic); appear as . ' an mtelligible rati a} on c:ontinuum, by and large consistent within itself regardless ' cr f the diuerent ti ' ~ o mo ves, prererences, and intellectual and moral qualities of sue . . . . f. cessiVe statesmen A reali t th 0 mtemational politics then will guard agrunst s eory 1 r::..Iac. tw0 I ' ' . es: popu ar cu 1 the eoncem WJ'th motives and the concem with . tdeolo~c:a referen('eS. P li To search for the clue to 1 . or~ gn ~ cy exclusively in the motives of statesmen is both futile and d eceptive. It 15 futile because motives are the most illusive of
._. t':~.:untll<h t < 111
J "

.tnd oh"..: I''. ';!f .Jike. Do we real.!~ lo10\\ \\hat our own 111 oti\c~ aH '!Am ' i,at do Wl' kno" oftl lf"' ruoti\es of others~ \<:t t'H'n il ',\1' }J,,rj ,,< ccs" to tlw real moti' es of state~ men, that kno" lcch;e "vcJilld IH'lp us little in understauding fo reigrl policies, and migh t ~veil lead us astra).

p~vrhdu r ,,] ] ,11 rL urtl'cl

intc rt'~tc; 'nd


L 1 Jc~- ar~. frequc:rttl; he~ond re<.:o~nition, h~ tlw


on' <!


It is

tnw that the kuflwl<dge of' the statesman's mothes may g1ve us one <~rnong man\' clu(S as trJ what tlw direction of his fo reign policy might be. It cannot (Q''c us, how~vcr. the one cl11c hy which to predict ltis foreign policies. History show no t:::xacl and necessary condation he::tween the quality of motives and the quality of . . foreign policy. This is true in both moral and political tenns. We cannot <:onclude from th e good intentions of a statesman that h1s forc1gn policies will be eith er morally praiseworthy or politically successful. Judging his motives, we can say that he will not intentionally pursue policies that are morally wrong, but we can say 11othing about th e probability of their success. If we want t~ know the moral and political qualities of his actions, we must know them, not h1s motives. How often have statesmen been motivated by the desire to impro,e the world, and ended by making it worse? And how often have they sought one goal, and ended by achieving something they neither ex-pected nor desired? ... A realist the01y of international politics will also avoid the other popular fallacy of equating the foreign policies of a statesman with his philosophic or political sympathies, and of deducing the former from the latter. Statesmen, especially under contemporruy conditions, may well make a habit of presenting their foreign policies in terms of their philosophic and political sympathies in order to gain popular et support for them. Y they will distinguish with Lincoln between their "official duty," which is to think and act in tem1s of the national interest, and their "personal .vish," which is to see their own moral values and political principles realized throughout the world. Political realism does not require, nor does it condone, indifference to political ideals and moral principles, but it requires indeed a sharp distinction between the desirable and the possible-between what is desirable everywhere and at all times and what is possible under the concrete circumstances of time and place. It stands to reason that not all foreign policies have always followed so rational, objective, and unemotional a course. The contingent elements of personality, prejudice, and subjective preference, and of all the weaknesses of intellect and will which Hesh is heir to, are bound to deflect foreign policies from their rational course. Especially where foreign policy is conducted under the conditions of democratic control, the need to marshal popular emotions to the support of foreign policy cannot fail to impair thE' rationality of foreign policy itself. Yet a theory of foreign policy which aims at rationality must for the time being, as it were, abs.tract from these irrational elements and seek to paint a picture of foreign policy whtch presents the rational essence to be found in experience. \\ithout the contingent deviations from rationality which are also found in experience .... The difference between international politics as it actually is and a rational ived from it is like the difference between a photog~aph and a painted theOJy. de1 po~ratt. The photograph shows everything that can be seen by the naked eye: the pamted portrait does not show everything that Catl be seen by the naked eye, but jt




Poltttca reaJJsm . . . ement . a1 realitv is replete w1th contingenctcs and svstemic irrati It know. t1 polittc tat , I. . . onali. . d . ts to tJle hpical inAuences they exert upon ore1gn pohc:v. Y it sh et bes an pom , ' ares . a1 sOCJ tl1e00. tlte need for the sake o f.tl1eoretica1unc.1 J a] erstanding' to s ress t ,nth . . . . . . al emen ts of political realih; for 1t 1s t 1 1ese rat1onaJ elements that makt tJte mbon e1 . . . lli bJ f'o. tlleon Political realism presents the theorettcal constru t reallh mte gt e 1 ' . . . c: of : r cy a rationaJ ,oretgn poJ 1 . which exnenence can. ne' er completelv achte,e. r . . . . polic)' to b A t tJ1e sa me time political realism constders a rahonal foretgn k r li . . . e r . poJ1 , , for onl) a rational roreign po c~r mtrumtZes ris s and m.,,. c\'' good wretgn . . aJ . ...lQ. mizes benefits and, hence, compltes both wtth tl1~ mor precept of prudence and tJ political requ irement of success. Political realism wc.mts the photographk pie~~ of the political world to resemble as much ~ pos_sible its pa~nted po1tJait. Aware of tJ1e inevitable gap between good-that IS, rahonaJ-fore1 policy and gn foreign policy as it actually is, political re~sm mai~tains not only that th~ory must focus upon the rational elements of political reality. _but also that foregn policy ought to be rational in 'iew of its O\Vll moral and practical purposes. Hence. it is no argument against the theory here presented that actual foreign policy does not or cannot li,e up to it. That ~~m_ent misun~er~tands th~ i~tention of tllis book, which is to present not an indiscnmmate descnption of polit1cal realitv, but a rational theot:' of international politics. Far from being invalidated by the f~ct tllat, for instance. a perfect balance of power policy will scarcely be found in reality, it assumes tllat reality, being deficient in this respect, must be understood and e,aluated as an approximation to an ideal S)-stem of balance of power. 3. Realism assumes that its key c-oncept of interest defined as power is an objective category which is universally valid, but it does not endow that concept "'ith a meaning that is fixed once and for all. The idea of interest is indeed of the essence of politics and is unaffected by the circumstances of time and place. Thu<:ydides statement, born of the experiences of ancient Greece, that "identity of interests is the surest of bonds whether between states or individuals" was taken up in tile nineteenth century by Lord Salisbuxy's remark that "the only bond of union that endures" among nations is "the absence of an clashing interests." It was erected into a general principle of government by George \IVashington:
A small knowledge of human nature will convinc-e us. that, ,vitJl far the greatest part of mankind, interest is the governing principle; and that almost every man is more or less.

. . . Ie:.tSt ~eek to sho" one thing the nakt d eve cannot see: the ho" . or at ~ d 1 ortra)e . IHunan tsst>nce of t le person p < a1 L tl . . . . . 1 . 1 COtltains not onJv a theoretic r>ut. so a not mative cl



Jt \\a'i

ho d n d t nb.r14~d upou in our century by Ma.x \\'e ber'

0 b!iC rvatJon:

Intt.r< t' ,lf ri:ll .twl ideal no~ ic!eas rlominate direct!} the acbonc; of r:'e>~- Yd ~-tehrt: c t . "imJ't s of th~ ''odd' c.:rcat(-d by t I1e:.e 1deas have very IJ fteT 1 served as .S\\1tces .: mini~e th~:- trach on ,,}, 1(:1, th< dyn:uni~m of inte rests kept action:. m0'-1ng.

Y the kind of interest determining political action in a particular _period.of ct historv depends upon lhc political and cultural context ~'~thi~ whi~h fore~gn pol~c~ is for~ulated. The goals that 1night be pursued by nations m thtlr fore_gn polic! can run the whole gamut of objectives any nation has ever pursued or tmght poSSIbly pursue. . t . d ilie The same observations apply to the concept of power. Its conten an manner of its use are determined by the political and cultural environment. Power may comprise anything that establishes and mai~tains the control of man o~~r end, from phystcal an - hus j)Ower covers all social relationships whiCh serve that d nlc . T I th violence to the most subtle psychological ties by which one min contro s ana er. Power covers the domination of man by man, boili when it is disciplined b~ moral ends and controlle<..l by constitutional safeguards, as in vVestern democrac1es, and when it is that untamed and barbaric force which finds its laws in nothing but its , 0 .. 11 strength and its sole justification in its aggrandizement. Political realism does not assume that the contemporary conditions under which foreign policy operates, with their extreme instability and the ever present threat of large-scale violence, cannot be changed. The balance of power, for instance. is indeed a perennial element of all pluralistic societies, as the authors of The Federalist papers well knew; yet it is capable of operating, as it does in the United States, under the conditions of relative stability and peaceful conflict. If the factors that have given rise to these conditions can be duplicated on the international scene, similar conditions of stability and peace will then prevail there, as they have over long stretches of history among certain nations. What is true of the general character of international relations is also true of the nation state as the ultimate point of reference of contemporary foreign policy. While tJ1e realist indeed believes that interest is the perennial standard by which political action must be judged and directed, the contemporary connection between interest and the nation state is a product ofhistory, and is therefore bound to disappear in the course of history. Nothing in the realist position militates against the assumption that the present division of the political world into nation states 'vill be replaced by larger units of a quite different character, more in keeping with the technical potentialities and the moral requirements of the contemporan world. .. The realist parts company 'vith other schools of thought before the all-important question of how tile contemporary world is to be transformed. The realist is persuaded tllat this transformation can be achieved only through the \\-'Orkmanlike manipulation of tile perennial forces tllat have shaped the past as they will the future. The realist cannot be persuaded that we can bring about that transformation by confronting a political reality that has its own laws with an abstrnct ideal that refuses to take those L into account. aws

under its ioHuence. ~otives of public "i.rtue may for a time, or in particular instances. actuate men to the observance of a conduct purely disinterested: but they are not of ~h-es sufficient to produce persevering conformity to the refined dictates and ~gatiOns of social duty. Few men are capable of making a continual sacrifice of~ ~ of private interest, or advantage, to the common good. It is vain to exclaun agamst the depravity of human nature on this account; the fact is so, the e>:perience ~f ~age and nation has proved it and we must in a great measure, change the c.'Onsti tution .of man, before we can make it othernise. ~o institution, not built on the pre sumptiVe truth ofthese maxims can succeed. I




l\llitical rC'alis 111 is :.~ware of thC:' moral sigllificattC<-' of political acn 11 is . '\ ' ,~\r. of the ineluctable . tension . between . the moraJ. eo m111 1nc1 an<.) lth . . . t ,1 t e ,..., 111 nts of succcs ful pohtical action. And rt 1s umVJ lhng to aloss . over e, b I I rt'qur.. l ohlitl'r.ttl' 1hat tension and thus to obfuscate ot 1 t _1~ moral and the polHic:al is~d h' making it appear as though the stark facts of pohhc_s we re morally more satisr Y ir~e than the~ actually are, and the moral law less_ex~chng than it actually is. lh>;ilism maintains that universal moral pnncrples can not be applied to t actions of states in their abst~act universal fon~ulation , but that th~y must be~~ tered through the concrete crrcumstances of tim ~ <ll~cl place. The rndividuai ma . say for himself: .. Fiat justitia, ~ereat mundu~s (Let J~Stlcc ~c done, even if the Worl~ perish):' but the state has no n~ht to say~~ m the _name of l_hose \vho are in its care. Both individual and state must JUdg~ p~l~hcal achon by um~'e rsal moral principles, such as that ofliberty. Yet while the mdtVJdual has a moral nght to sacJilke himself in de fen se of such a moral principle, the state has no righ t to let its moral tlisappro. bati.on of the infringemen t of libe1ty get in the way of successful political action itself inspired by the moral principle of national survival. There can be no poLtical morality without prudence; that is, without conside ration of the political consequences of seemingly moral action. Realism, then , considers prudence-the weighing of the consequences of alternative poutical actions-to be the supreme virtue in politics. Ethics in the abstract judges action by its conformity with the moral law; political ethics judges action by its political consequences. Classical and medieval philosophy knew this, and so did Lincoln when he said:

)' a1 t T -~ purltuing tlw1r r~'>lKdJH' loc)'K tt allllntmnr,, <Jur 0\ ' 11 iu<:ludcu, as po ttJc <'J I .JL . . 11 fth"Jn \ne\ "'<' ' . l''' . )e . Jt<Jcsts rl fwcu in ltrrns of powcr. \v< are a 1 1 t o do JUStrce to a11 o t ~ c <>lhl'r 11 . \ Vc arc a ) c <J JU ...... then 1 a1 '< trc I . t I . ale able to do Jll'-tic:c l o all of them Jn a c ll' sense. cl ~d them in t 11s as 11on . ._.. .I h ther nations. whilt: pronation s as \\( ' .Jlld~t' our own anu. a\1ng Jl' ~e r t t reflect urstring polidc:" that respect tlJe mterests of o (;lp"l>lc ofJ> m o ' , tecting and promoting those of our own. ~1od eratlon po licv cannot at <>

I do the very best I know how, the vel)' best r can. and I mean to keep doing so until the

end. If the end brings me out alJ right, what is said against me won't amount to anything. If the end brings me out wrong, ten angels swearing I was right would make no difference. 5. Political realism refuses to identify the moral aspirations of a particular nation with the moral laws that govern the universe. As it distinguishes between truth and opinion, so it distinguishes between truth and idolatry. All nations are tempted-and few have been able to resist the temptation for long-to clothe tlteir own particular aspirations and actions in the moral purposes o f the universe. To know that nations are subje(..'t to the moral Law is one thing, while to pretend to know with certainty what is good and evil in the relations among nations is quite another. There is a world of difference between the belief that all nations stand under the judgment of God, inscrutable to the human mind, and the blasphemous conviction that God is always on one's side and that what one ,vilJs oneself cannot fail to be willed by God also. The lighthearted equation between a particular nationalism and the counsels of Providence is ~orally indefensible, for it is that very sin of pride against which the Greek tragedians and the Biblica] prophets have warned rulers and ruJed. That ~uation is al~ poli~cally pernicious, for it is liable to engender the distortion in J~em~nt w~ch, m the blindness of crusading frenzy, destroys nations and CJvi.hzations-m the name of moral principle, ideal, or God himself. On the other hand, it is exactly the concept of interest defined in terms of power that saves us from both that moral excess and that political folly. For if we

cl thcr schools or thought t . I a]' the moderation of 11JOraJ judgment. I sm an o c. TJ1c cliffcrcnce then b~twccn po Jtica re 1 .. } J sm may la\'C ' ' \ I. s re-1 and it is J>rofo11 nd. Il owever much the theory o f poUtica re. 1 t I i1.l , 1 I been mi understood and misinterpreted, there rs no grunsaymg ts drstrnc:t1ve m ~ olitic:al . . . lectual and moral attitude to matters politic~. . P. I ntellectu ally the IJolitical realist mamtmns the autonom} of tl 1e h k . ' 1St . 1 e as the ec:onom ist, the lawyer, the mora1 mrun t am tl1errs. He t . rn s m :~r~~ ~f interest ueflned as power, as the econom~t th~nks i~ terms of rn~~rest defined as wealth; the lawyer, of the conform ity of ac:tion Wlth legal ru_les. t~~ . , t of the conformity of action with moral principles. Th e econom1st as s. f ':>" Tl " lawver moraus , le . . "How does this policv affect the wealth of society, or a segment o rt. . aw. L1e . k . "Is this poUcy in/ accord with the ru les o f 1 ?" Tt mora}j s t asks "Is tins po 1 ~~ ~~ accord with. moral principles?" And the political real ist ac;ks: "How does th1s p;'licy affect the power of the nation?" (Or of the federal government. of Congress, of the party, of agriculture, as the case may be.) The political realist is not unaware of the existence and relevanee of stan~ards of thought other than political ones. As political realist. he cannot _but subordinat~ these other standards to those of politics. And he parts company wrth other schools when they impose standards of thought appropriate to other spheres upon the . . . .. poiltical sphere. . . . This realist defense of the autonomy of the poilocal sphere against rts subven.1on by other modes of thought does not imply disregard for the existence and im~rtan~ tc; gned 1 of these other modes of thought. It rath.e r implies that each should be ass1 proper sphere and function. Political realism is based_ upon a,p!,w~~tie conc~p-~on of human nature. Real man is a composite of "econom1c man, political man, moral man," "religious man," etc. A man who was nothing but "political man would ~ a beast, for he wouJd be completely lacking in moral restraints. A man who was nothrng but 'moral man" would be a fool, for be wouJd be completely lacking in prudence. A man who was nothing but "religious man" wouJd be a saint, for he would be completely lacking in worldly desires. Recognjzing that these different facets of hwnan nature exist, politi~al ~ealis.m also recognizes that in order to understand one of them one has to deal w1th 1t on 1ts own tenns. That is to say, ifl want to understand ''reilgious man,.. I must for the time being abstract from the other aspects of human nature and deal with its religious aspect as if it were the only one. Furthermore, I must apply to the religious sphere the standards of thought appropriate to it, always remaining aware of the existcnt-e of other standards and their actual influence upon the religious qualities of man. What is true of this facet of human nature is true of all the others. ~o modem economist, for instance, would conceive of his science and its relations to other scienet'S of man in any other way. It is exactly through such a process of emancipation from

P ttics is indtt-d the pwvose of poltttcal re<ilism. It i in the nature of things that a theory of politics which is b~s d e upo _ J . . . principles will not meet \VJth unamn~ous ~pprova1-nor does, for that matt n such . forei!lll )olic,. For theory and poUcy ahke run counter to two trend . er, such s Jn our 1. t1: o 1 ,1 cui. ture "'hich are not able to reconcile Jemse ves to the assumptions and results rational. obJective t I1eoJ)' of' politics. 0 ne o f t11ese tren cls disparages the of.a ower in societv on grounds that stem from tl1e experience and phjJos0 . 1 role of P P1Yof the "' nineteenth century; we shall address ourselves to this tendenc)' later . In alist theory and practice of politic .greater detail The other trend, opposed to the re from the vety relationship that exists, and must exist, between the humas, st~ms and the political sphere .. .' . The hum an mind in its day-by-day operations :~~~d bear to look the tmth of politics strrugbt in the face. It must disguise di ot ' stort h belittle, and embelhsh the truth-t .e more so, the more th e individual is acti 1 involved in the processes of politics, and particularly in those of in ternational tics. For only by deceiving himself about the nature of politics and the role he ~~ on the political scene is man able to live contentecUy as a poutical animal with \u~~ self and his fellow men. . T~us it is in~vitable ~1at a theory ~hie~ tries .to ~nd_erstand international politics as It actually IS and as It ought to be m v1ew of Its mtnnsic nature, rather than as ke people would U to see it, must overcome a psychological resistance that most otl1er branches of learning need not face.

)prOj)J' . t ..J. rll" of thoucrht. and tlw development of one a1 tl tl't ~ ,111{1. 1 e to its . ' ~ tl ,.,,., . tiJtl ec.:ono 1111 cs has de' eloped as an autonomous theo1) ' of t]1e e .sub1 "'\;' m:t ltt't. m To contribute to a s1milcu develop men t in the (!eld o (' cononr1<:. , .. .L . oJ . n . . .. .tct '',1"-. ol'm .

A Criuque ofMorgenthau's

Prjnciples of Political Realism


/t is 11ol in (l.ivinglife hut ilt risking life that man i.s raised above the animal: tltat is why s11periority lws been accorded in humanity not to t/,e sex that brings fort/, b111 to tltat which kills.


l. The Writings ofGeorge Washington, edited by John C. Fitzpatrick (Washington: United States Printing Office, 1931-44), Vol. X, p. 363. 2. Marianne Weber, Max Weber (Tuebingen: J.C. B. Mohr, 1926), pp. 347-8. See also Max . s Weber, Gesammelte Auf t:itze zur Religionssociology (Tuebingen: J C. B. Mohr, 1920),

p. 252.

International politics Le; a man's world, a world of power and conflict in wltich warfare is a plivileged adivity. Traditionally, diplomacy, military service and the science of international politics have been largely male domains. In tl1e past women have rarely been included in the ranks of professional diplomats or the military; of the relatively few women who specialize in the academic discipline of international relations, few are security specialists. \Vomen political scientists who do study international relations tend to focus on areas such as international political economy. North-South relations and matters of distributive justice. Today, in the United States. where women are entering the military and the foreign service in greater nu mbers than ever befo re, they are rarely to be found in positions of military leadership or at the top of tl1e foreign policy estabhshment.2 One notable exception, Jeane Kirkpatrick, who was U.S. ambassador to t11e United Nations in the early 1980s, has described herself as "a mouse in a man's world" for il1 sp.ite of ~er au~horitative and fo rceful public style and strong conservative 'ere~ denttals, K1rkpatnck maintains that she failed to win the respect or attention of her male colleagues on matters of foreign policy.3 Kirkpatrick's story could serve to illustrate the discrimination that women often encounter \ovhen they rise to ltigh political office. However. the doubts as to wl_1ether a woman ':ould be strong enough to press the nuclear button (an issue ra1sed when a tearful Pahicia Schroeder was pictured sobbina on he r husband's shoulder as she bowed out of the 1988 U.S. presidential race)~suggest that iliere may be ~~~ even more fundamental barrier to womens entry into the highest ranks of the mrlitruy or of foreign policy making. Nuclear strategy, with its vocabularv of power, threat, force and deterrence, has a distinctly masculine ring;"' moreo~'er.
F~m J Ann Tid:ner, "A Critique of .\llorgenthau 's Principles of Political Realism" in Cende d 1 natw~wl Rrlativlls, eds. Hebecca Grant and Kathleen ~t'\vland. Published b} lndiana u . r ~ pnter-

nJVerstv Hetmnted b pcrn11sson o f Kat hi een Newland. Portions of the text and some rootn"'t ha ' Lress --Y 11 ve ~l v es omitted.


:\lt.'t\'\lf;Vic,tll~ jud~t>d to he lackiHg in qualit ies which the . ( '\{ 1k, lt IMs .tlso be tn suK_!!PSl<>d th at , alth ough 111 0r c w omen are en s~ tcrrn~ ,, ,rJd of public polic~ . the)' are more comfo1tabli.' dc uling with <.lor tc.rrng thl ! . ~u ch ;t..; :-tx.'htl "~ Ila rL' t I t ar e m t-:bl .1 1 . l<~ ore co ~npau e \\' ll l t 1e rr nurturing skill c i. nestr s. , ssuc5 l:u,:!t' nurnbl'r ol women m th e ra nk of the pe ace m o, e m en t sugges ts that Y the et . n. not uninterestf>d in i sues of w ar an d pe ace, al though th e ir freq uent ~?rnen frollliH\tiouul secUJit; po lic~ has often br an de d th em a. naive, uninform ed tssen t unpahiot ic. 0 r eve n In thi ch ap te r I propose to expl ore th e qu es tion of why in te rnat ional .. pe rce1ved as a m 1 an s worId an d w l1y women 1e mrun so , unc} re presen po1 s er ted . lhc higher eche lo ns of th e fo re ign policy es tablis hm en t, th e military and th e aca~n t~e d~cipline of inte rnational re lations. Sinc e I be ueYe th at th er e is something a~m 1c this field that re nd ers it partic ularly in hosp ita ble an d un at tr ac tive to wome:ut in tend to focus on th e natul'e of th e disc iplin e itsel f ra th er th an on possible stra~e ~ gies to remo,e barrie rs to women's ac cess to hi gh policy positions. As 1 have already suggested, th e issues th at are gi ven p1iorit;: in foreign policy are is ues wi th which men have had a special affinity. Mor eove r. if it is p1i marily men who are describi ng thes e issues and constructin g theori es to explain th e worki ngs of th e international system , might we not e>.'P ect to fin d a masculine pe rspective in the academ ic discipline also? If this w ere so th en it could be argued that the exclusio n of wom en has operated not only at the level of discrimination bu t al so through a pr ocess of self-se lection which begins w ith th e way in which \V C ar e taught abou t in te rn ational relations.
,,,HU<.'Il .n v



aboul in teruational politics sinc:fc 19li4.'"'~ Y1 orgcnthau 's principles n po lt<.: reaI'JS .. n be su mmarized as follows: m <.:a .

. f SC'hohrs :-lJIUt pra<.\JtiOI\I.;r:> uf 1nttrrcalislll f:-otn "h id t se ,c ral gcnC'raLJ. On<, . tl 1 l 1St al e'> ha\t:ob~ellltu,un.!> I d \lthollah \l orgcnth au Iat Hm " m te Jf ll et na tional n~ 1c ~-> . l'f1 . . rourt and ambiguou!> U !>C:: o f IJ 't~ l'rcc1ut'lllk be.en t:riti.dz1c:.dhlor his lack of SCJ~n I <: ln~l . . . ,,h ic h the maJor . . .. Jangttc:tg<:, lh<:-'>(; SIX pnn<.:tp (;S et\ .C, Sl~l1l fiCc tlv Iramec t 1e '"a ) m I . the \\ 'es t ha \e th ou gh t il) intemational relations scholars_~nd lO ad . . pr tll on er s 111


In or de r to investigate this claim th at th e disciplin e of in te rn atio nal relations, as it has traditionally be en defined by re alism , is ba se d on a masculin e world vie w, I propose to examine the six ptin ciples of political realis m formul at ed by Hail S J. M orge nt hau in his cla'isic work Politics A mong Nations . l shall use so m e ideas fro m feminist theory to show that the way in which Morge nt ba u desc rib es and expla ins in te rn ational politics, an d th e prescriptio ns th at en sue are em bedded in a mascu lin e perspective. Th en I shall suggest so me ways in which fe minis t th eory migh t he lp us begin to conceptualize a world view from a feminine pe rspective and to formulate a fe minist episte mology of in te rnational re la tions. Drawin g on thes e observations I shall conclude with a refo rmulation of Mor gent ha u 's six principles. Male critics of contemporary realism ha ve al re ady raised m any of the same ques tio ns about realism that I shall address. However, in un de rtaking this exercise. I ho~ to make a link between a grow ing critical pe rspe ct ive on intemati?~al relations .th eor~ an ~ fem inist writers inte rested in global issues. Adding a femm!St ~rspective to Its discourse could also he lp to make th e fie ld of in te m ationaJ relations more ac-cessible to women scholars an d practitione rs.

er deveJop t rational theory that reHe<.:ls thes e obJective laws. l fi d . 2 TJ 1e mai~ signpost of poli tical reaHsm is the concept of interest c e nfe 1 ~ . term s of power which in fuses ration al orcl . to the subject matter o po I . ... tics, and thus makes th e theoretical u_ er m 1 ~s possl'blc nd~rstandmg of P? tti Political realism stresses the rational, obJe ctive and unemobon . 3 Real ism assumes that in terest de fi ne d as power an obJe cb ve categorv IS r .1 which is universally valid but not w1t1 a m th t fixed once an d 10-r eanmg a lS all. Power is the control of man over man .. . . 4. Politjcal realism is aware of th e moral . significance of political actlon. I_t IS al so awe:lre of the te nsion between th e m oral command and the re qwre ments of successful political ac tion. . 5. Political realism refuses to identify t) ratio~s f a partic:u lar al 1e more aspt o nation wi th th e moral laws that gove rn th e universe . It IS th e concep t of inte rest defined in te rms of power that saves us from moral excess an d political folly. .. . 6. T he poH tical realist maintains th e auto , nomy of the_roli.?cal ~~~.ere, h~ as ~s "How does this policy affect the power of the nation?' Political realis n~ LS base d on a plur ali stic conception of human nature. A man who was no_tht~g but "political man" would be a beast, for he would be co mpletely lacki~ ~ m moral restraints. But , in order to develop an autonomous theory of political behavi our, "political man " must be abstra cted from othe r aspe cts of hu m an nature .6

. 1. Polilicc;, like society in general , ts gove 0 b b' f v Jaws that have their rl roots in !tun 1a11 natu re, which is unc ne _'f o Jtehc 'e~o re it is possible to 1angm~: _


= h o s e n to focus on Hans J. Morge nthau's six principles of politic al realisrn they re present one of the most importa nt st ate men ts of co ntemporaiY

I am not going to argue that Morge ntha u is incorrect. in his P?lt_Tayal ~f the inte mational system. I do be lieve, howev er, that it is a partial desc np tio n of m te rnati.onal politics because it is based on as sumptions about human nature_tl:at are pa1tial and that privilege masculinity. Firs t, it is necessary to de~e masc:ulim~.~tl fe mininity. Accordin g to almost all fe min ist theorists, masculiruty an d fenunuuty refe r to a set of sociall y constructed catego ries, which vary in time and place, rather than to biological de te rminants. In the V\T est. conceptual dichot omies such as obje ctivity vs. subjectivity, reason vs . emotion, mind vs. body. cu lture vs. na tu re. self vs . othe r or autonom y vs. relatedness, know ing vs . being and public vs. private have typically been used to describe mal e/fe male differences by feminists and nonfeminists alike. 7 In the Unite d States, psyc hological tests conduc te d across diffe rent socioeconomic grou ps confirm that indivi duals perceive these dich otomies as mas culin e and fe minine and also that the ch ar acte ris tics ass<xiated with masculi.nitv ar . e




mon' luvhh ,~tlued l>' m en and women alikP ." ,. ., . . . . .. . It IS ilJipo rtant t<) st res~ 1 thttt th l'~"e eharnc te m tJc s are stereolyptcal; th e~ do no t nec:cssarily de . ' 10Wevet - Jt 1a ' 11 .tJ m en or women, w I10 can c~11t') c I ntcl ensl JC.:S an<. I rnod es of thouscnbc ind ' ht . 'Vid. "i th th e oppo ite sex. g US!!<x.:iat~;:d a ,ocabu l<:lf) that contains many of th e w on]s as sociate d wi th ih as I h;n<. identified it, \1 orgenthau asserts th at it is possible to dev . . 1 rnast1Jiin. . . (and unemotional) tJ1eorv of mtem ahon a I po t1ti e cs bw;ecl on ohJect op a rarl()nal , . Jve 1 tl aw have th ei r roots i_n ~1u m~n nature. ~mce ~org enlhau '~rote th e first edi~~ 1at Politics Among .\ atums m 1948, thts search for an obJective science of ~n ()f national politics based on th e model of the natu ral sciences has been an im Inter. part of the realist and neorealist agenda. 1n he r feminist critique of th P~rtant sciences, Evelyn Fox Ke lle r points ou t th at most scienti fic c~mmu nitiec ;a~ral the "assumption that the universe they study is di re ctl y access ibl e rep s are resented bv concepts and sh ap ed not by Ianguage but only bv th e demands of' lo . e~pe rime nt .'.g The laws of nature, accord g1 ing to this view of science, are "becand th e re lativitv of language. " Like most femi nists " , Ke lle r re ie cts this view or (' _.YOnd J ,cJencc whic h, she asserts, imposes a coercive, hier arch ic al and conformist patte rn on _. entinc inqu iry. Fem ini sts in ge neral are sceptic al about th e possi bility of findi:cJ. unjversaJ and obje ctive foundation for know le dg e, which Morgenthau claimsgi: possible. Most share th e belie f that knowledg e is socially constructed: since it is language that transmits knowledge, the use of language an d its claim s to objectivity must continually be qu estioned. Keller argues that objectivity, as it is usually de fined in ou r culture, is associated with masculinity. S_he identifies it as "a ne twork of interactions between gende r development, a behe f system that eq ua te s ob je ctivity wi th masculinjty, and a set of cultural values th at simultaneously (and cojointly) elevates what is defined as scientific and what is defined as mascu lin e." 1 Ke ller links the separation of self from other, an important stage of masculine ge nder deve lopm en t, with th is notion of objectivity. Tran slate d into scie ntifi c in quiry this be co mes th e striving for the se~ ion of subject and objec-t, an impo cd: rtant goal of mode m science and one whic~. Ke lle: asserts, is based on th e ne ed for co nt rol; he nc e objectivity becomes associated With power and domin atio n. . The ne :d f~r control has been an importa nt motivating forc e for modem realIS~. To begm_has search for an objective, ra tional th eory of in ternational politics, which c::ould 1mpo~e ord~r on a chaotic an d conflictual wor ld , Morgenthau con structs an ~straction wluch he calls political man , a beast com plet ely lacking in moral restramts. Morge nthau is de ep ly aware th at re al me n, like real states, are :: m o ra l and ~al bu t, because states do no t live up to the univer sal moral laws doo govern ~e umverse, those who behave moral ly in international poHtics are med to failure because of the immoral aL'tions of ot he rs . To solve th is te nsion 0~~ ~ postulates a realm of in te rnational politics in w hich the amoral beha\' JOUr political man is not only permissible but pr ud en t. It is a H ob besian world. sepats te and disti ra ~ beas . net fJrom th e world of domestic order. In it, states k may act Ue ~ U :a l depends on a maximization of po we r an d a S willingness to fight . the validr argued that ~ personal is political, m ost fe minist theory wo~d reject ty of construc:ting an autonomous political sp he re around whtch


~ CR

,. UL 0F ~ORGE~THAU'S PRJ':'-'C!PU:.S Of



\~ Kc \1r" .' . ic., uf ,c~r li; ,..,m),. rniJ<k-~ of.<;olldlld have 1 en 1 . )0111 1clrl J( < rawn. . . 1 . I cl fi .5 . n' I (lP f'(11( l<; tl It '"t Iw < c11 , r{ 1 ' 0 l . ! "H n punlt<. a11d .m ... at e no t m ~ < fl ( a I . cl 1~ 'I \ lor('< 11 thau .s . n'< c; r,fth< l'oll1<.al bu t,o hclp.., nn It\ co fo n< 1J()U Ia n aJ nt en t au ,~~ e. fhumafl natUn' L' .ticallllar i'> a ~.rJc:ial <:omlnld based on a pa , rt1 rcpr(~en ta ton rJ p(' 1 tat !!1t ' " ' 11 a.c;k w} ,e;n the woroen were . JJoI>I . < e of natu re p rcSll rrt cl lJ I' m >e t , 0 ne Jtll . liJSL ha.,c }wen im c e<..l in rep rodu ction au . 1 ss]- ath' r than v.:ar Jiv u cn1 urcann g, r e !1<.)' ' 1 1 . f~ , ifl ifc was lo ~0 on fo r 1110rc th an on e gene 12 v1 ration. orge ntl lau's cmplaSts on <.1 tr~~on Aictu a.l a-,pectc; of the int<mational sy!:.te m contributes to a ten<..lcn ~~ - "h~eh b~eothc r rc ali ~ls , tn <..le-c i(Jpl.:lsi7 elt:m cn tc; of cooperati on an J regene rat1on w IC are abo a\ ptc.:ls of intc: rn ation a.l rdations.13 . Morge nthau s constntction of an ar~oral realm of in~ernatio~al power pol.ltiCS"~ an attempt to resolv t what he sees ac; a fundam en tal te nston betwe~n. th e m~rall~ . tJ at gove rn th e universe and the reguirement'i of successf~l pohb~ action .'n _a 1 wo rld wh ere states use norality as a cloak to justi fy the pursu1t of thetr o"vn national inte rests. Morgenthau's universaHstic mo rality po~ ;tulates the highest form of morality as an abstract ideal, similar to the Golden Rule, to w~ch s~tes seldom _adhere: th e morality of states , hy contrast. is an instrumental morality gw~ed by se lf-mte r~st. Morgenthau's hierarchic-al ordering of morali ty contams parallels w1th th e work oF psychologist Lawrence Kohlbe rg. Base d on a study of th e moral developme nt of 84 America n boys , Kohlberg conclude s that th e highest stage of human moral developm ent (which he calls ~t.age 6) is th e ab ili ty to re cognize abstract ~i ve rsal principles of justice; lower on the scale (st age 2) is an instrum en tal moraltty concerned with serving one's own interests wh ile recognizing that ot hers ha ,e interests too. Be twee n these two is an interper sonal morality which is contextual and characterized by sensitivity to th e needs of others (stage 3).14 In he r critique of Ko hlb erg's stages of moral development, Carol Gilligan argues that th ey are based on a masculine conc eption of morality. On Kohlberg's scale wom en rare ly rise above the third or co ntextual stage. Gilligan claims th at thi s is not a sign of inferiority bu t of diffe rence. Since women are socialized into a mode of thinking which is contex tual and narrative, rather than formal and abstract, they tend to see issues in contextual rather than in ab st ra d te rms. l5 In international relations the tendency to thjnk about morality eithe r in te rm s of abstract, universal and unattai nable standards or as pu re ly instrumen tal. as Morge nthau does, J et racts from our ability to tole ra te cultu ral diffe re nces an d to se ek pote ntial fo r building community in spite of th ese diffe re nces. Using examples from fem injst literature I have suggested th at ~l orge nthau's attempt to construct an obje ctive, universal th eo ry of inte rnational politics is rooted in assumptions about human natu re an d morality that , in m odem \\'este m c~lture, are associated 'v\'ith mascu Hn ity. Fu rth er evidence th at Morgenthau's prin<:lples are not th e basis fo r a universalistic an d objec-tive theory is containe d in his frequ ent re fe rences to th e failure of wh at he ca lls the "le galistic-moralistic" or idealist approach to world politics whic h he claims was largely responsible fo r bo th the wo rld .wars: Having lai d th e blame for the Se cond World War on the misgui de d mo~~lt)' of appeasement, Morge nt ha u's real7>olitik prescriptio ns for successful pohtical ac tion appear as prescriptions fo r av oiding the mistakes of the 1930s rath er than as prescriptions with timeless appl icability.

P:\RT 1 AN,\ROf\' AND 11:, LUN::>t.t.Jt. t.l\ Ll.:!>

m ea contn ution to mtemationa re alion theof\ bv stntcting an altemati,e. feminist perspecthe on intemational politics tl{at help us to earch for more appropriate olution ?

,u1J 0 tJ1er realists. i_ no longer ~bl~ to deal \\1th an mcre~_mgl~ pluralistic arra~au : of Problems ranging from e<'onomc mterdependence to. em1ronmental de(J dation , .b . . I I e-ra Could femini t theory ak

If \ltlt--:tntluus world d~" is_t>nbe-ddetl in tla t raumas of the eeo d . a~ \ \ .. r. , f\' hl. nn.'Sl'I1Illi<.>n" :-till ',ilidh " e mo,t:' fu ti her awav fr011 tlll n \\ or}d 1 . ev r . I . .... ,,ith other c..nti<.-s of reali m t e ,,ew t ut. m t rapidh- chanu.; ent? 1 :-1 t .. 1. . .J, .:- na WorJ~ l1iltllt l>e.~u h.> St'<l.n:h Ior mode. of behanour w~~rent Irom tho e pre en ""\\e ~lOn!l'llth.m. Gh en that an~ war between the maJOr powers i likeh- to I bed by .. . . . -'d . cu. ' . n "n-;.Nno ~e<.uri h l,, uK'reasmu [>O"er c:otu l)(':" utCKL '" \l oreQ\""r le nuct,..,~ th '"Clf, l ~ " e nati0 st'.tte. tht> prim~ con. tituti,e element of the _intem~tional _ y. tern for \loreen . n



I1.1 "LI )ne; )J Ill (.'01 -.idt. r.lblt' \1ICC.'t''' Ill lnll l H I!! liP I l! , t j" t "I }! ;, r \f' ' }, l \\ . '' '"' . I" I . r . ' t.' \\"ll ......, , the wC'I rld "I<.N lw.. to0 \\OIIItl t prf<.Jd ..,,!~ 1 .t cou :- ' ' tn d\!..,t ri\wd in tlu ..,<.

I i.

1 cooJwr.ttloJI

i n .t n ,U!'a



e tint c:c>OI>t.t tttn Ollkt>nll'' m JntL"nt.ttlCin,ll re\alJOI~ . . . r PIuraj t1<' Et\m h. tOllllll\1111..l H'. Il'''''ltl lw . ... . tlthonrh 1-...trl D clll'ch s notum o ,c; " ot tetm . . ' I . I I tt o commulllt\- <lllt n e such l'\,tntple wlwn pO\\ t.:t i~ .'lS~<x:ialec \\lt 1 1UI c IlL,. . . I I t. \S on . . . .. . . uh in mtcm.tlJOna re cl 101 . . ~n,incT that [)0\\r r " dommatlon IS a pen ,\SI\ e re. 1 . . I I1 t 1' U de mm ' . ' r tl-ere m: tlso inst,mcc<: ol cooperatiOn ter tate rdat10ns " HC c . H o" t>' c ' . . Tl ki u . I out 1 )owtr 1n b ob cur<.>d when powtr 1. een solek as dommat10n. 1m nr a 1 . I to e . I . h. k t f,eh- about tlu"' I10tE' ntn\ 1 this multidimensional sen::-t' ma; he p u. tot m cons rue _ all .


If the way in which we describe reali~ has an effect on the wa~ we perceive and
act upon our environment. new perspecti,es might lead us to consider alternative courses of action. \\'ith this in mind I shall first examine h,o important concepts in international relations, power and securi~- from a fe minist perspecthe and then discuss some feminist approaches to conflict resolution. \torgenthau's definition of power, the <..'Ontrol of man o,er man. is ~pical of the way power is usually defined in intemational relation . :\ ancy Hart ock arrues that this type of power-as-domination has alwa~-s been associated ,,; th masculinitY. since the exercise of power has generally been a mascu line a<:thit-y: rclrel~- ha,:e women exercised legitimized power in the public domain. \\" hen women write about power they stress energy, capaci~ and potential. ays Hart ock. She notes that women theorists, e,en when they ha,e little else in common, offer similar clef initions of ~wer which differ substantially from the understanding of power as domination. 11 Hannah Arendt, frequently cited by feminists writing about power, defines power as the human ability to act in conc-ert, or to take action in connection ,,~th others wh~ share_ similar concems. 16 This definition of power is similar to that of psychologist DaVJd McClelland's portrayal of female powe r, which he describes as shared rather than assertive. 19 Jane Jaquette argues that. since women ha,e had less access to the instruments of coercion, they have been more apt to rely on power as . persuaston; she compares women's domestic activities to eoalition building. 20 All"! these writers are portr-a}ing power as a relationship of mutual enable::~z~g~definition of femal~ power to international relations, j aquette sees . een female strateg~es of persuasion and strateoies of small states operating from "ti of eakn o al of a ~ 0~ w ess m the mternational system. There are ~ ~ beha\lOur that contain elements of the female strategy of coali tioo Conference ~eh example_ is ~e Southern African Development Coordina( ADCC), which IS designed to build reoional infrastructu.rf on mutual cooperation and llectiv o ase ~ 011 the South African eo e self-reliance in order to decre. , economy. Another is the European Commurut).


for cooperation a<; well as conAict. an aspect of international relat1uns gencr ; plcned do,,,1 b, reali m. . . tl 11 Rerlefl.ning. national sec11 rilY is anothe r wa~ in which femu11 t W00 _cou ' 11 11 1 contribute to 'new thinking about international reIabons. 23 T racl"I tiona ) _ 1 tlc:\\'est. the concept of national sccuti ty has been tied to military strength antl1ts ro1 ~ in the physic< protection of the nation state from external ~h reats._ ~,l orge~1th~u_ U 1 notion of defendina the national interest in terms of power 1s <.:ons1stent v \1th th_ dentlition. But thi~ traditioMI definitio n of national security is partial at best tn toda,'s world.2-1 The technologicall~- advanced states are highly interdependent and reh- on weapons whose effects would be equally devastatina to winners and losers ~ke. For them to defend national secmity b~ relying on ,.,ar as the last resort no loncrer appears ,ery usefw. Moreover, if one thinks of sec~rity in :\orth-Soutl1 rather than East-\\"est terms, for a large portion of the worlds population securin has as much to do with tl1e satisfaction of basic material needs as ";t}1 miJitruv threats. According to Johan Galtungs notion of structural ,;olence. to suffer a low'er life ex'Pectancy by ,;rtue of one's place of birth is a form of ,;olence whose effects can be as devastating a<; war. 25 Basic needs satisfaction has a great deal to do with women. but only recently haYe womens roles as providers of basic needs, and in development more generally, become visible as important components in development strategies.~ 6 Traditionally the de,elopment literature has focused on aspects of the development process that are in the public sphere, are technologically complex and are usually undertaken by men. Thinking about the role of women in development and the way in which we can define development and basic needs satisfaction to be inclusive of women's roles and needs are topics that deserve higher priority on the international agenda. Typically. however, this is an area about which traditional international relations theory, with the priority it gives to order oYer justice, has had very little to say. A further threat to national security, more broadly defined, which has also been missing from the agenda of traditional international relations. concems the environment. Carolyn Merchant argues that a mechanistic '\iew of nature. contained in modern science, has helped to guide an industrial and technological development which has resulted in environmental damage that has now be<.'ome a matter of global concem. In the introduction to her book The Death of Xature, Me~ch_ant suggests that, "'Women and nature have an age-old association-an affihati~n t~at has persisted throughout culture. language. and history. .Zi Hence she mamtams that the ecology movement, which is up in response to



. d tl1e women's move ment are d~, pt, mtc n:onn ected l'tl\ironmen tal threats, an . 'tl nature ratl1e r than d< m.:1 t'ng it both s 1 uilibnum w1 1 ' Both stres t\1ng m eq_ caJ nt1 in which each part i'> nH tllv det)end"(:et n., . l' . -hterarc1 u e ; ~ r '"n nature <lS c1 t\l.t~g.non. as well as many feminists, are ne" -st~.:.,~. st_mg that on\ , on the whole. Ecologtsts. f Id 11evv will allow the huma n sp<'CJC'S to surv ) 1ve aJ such a (unc1 ament <.:I1 anae o wor . ' o t . . 111A' ti 1 on the en\lronmen . . 10' . the damage 1t lS tc ic and environmen tal se<.:unty in intertle. 10111 Thinking about ~- ~:f)~e=~o~or new methotls of conflict resolution that seck pendent terms suggests aJ ther than zero sum. outcomes. One such method 6 to achieve mutually b~~c~s ~:;.k on maternal thinking.''1'> Ruddick describes comes from Sara R~ d the preservation of life and the growth of children 1 'nki r as 1ocuse on matem al t lJ ng . . t conducive to th ese goals, tranquility must be c . d estic en\llronmen To tostel a om . . fl ' t vhere possible, engaging in it non-violently and con tc prese rve cl by avotding b .t .5 'over In such an environment t h e enc1 for which . s . 00 mmuntt)' w en t t restonng l ..J= t , e foug 1t are su b01w nate to the means by whic:h they are resolved. This dispu es < n11 u 1 tion involves making contex-tual judgements rat11er than metbo? of cob Jclt resot ud ds and thus has much in common with GilLgan's clenappealmg to a so ute ~ an ar rution of female morahty. . l . . 1 t resolution of confli ct in the domestic sp 1ere 1s a w1delv Whi Ie non-vJO en . . ' es1 stance in the public realm IS regarded as deVlant. But, accepted norm , passtve r . .. . . k argue . tl e peacefuJ resolution of con AlCt by mothe1 s does not usually s, 1 as Ru dd1c. 1 1 ne's enem ies an important reason w 1y women 1ave extend to tbe cl1ildre n Of O ' b been ready to suppo rt me n's wars 29 The question. for Rudruck then ecomes how . . . . king, a mode of thinking which she believes can be c d m roun to get ma temaI t1110 . . . . nto t11e public realm. Ruddick beheves that finclmg a men as we l] as women, Out l .. . . common humanity among one's opponents has become a condition of s~rvtvalt~ the nuclear age when tl1e notion of winners and losers has beco me questio~able. Portraying the adversary as less than human has all ~oo of~~n bee~ a techmque .of the nation state to command loyalty and to increase tts legJtimacy 1n t11e eyes of Jts citizens. Such behaviour in an age of weapons of mass destm ction rnay be selfdefeating. . . We might also look to Gilligan's work for a feminist persp ective on confli ct r~solution . Reporting on a study of playground behaviour of American boys and girls, Gilligan argues that girls are less able to tolerate high levels of conAict, and mor~ likely than boys to play games that involve taking turns and in which t11e suctes~ one does not depend on the fail ure of another. 31 \hile Gilligan's study does _not~ ~ into account attitudes toward other groups (racial, ethnic, economic or nationa ), Jt does suggest the validity of investigating whet her girl'i are socialized to use diffe~e~~ modes of problem solving when dealing with conflict, and whet her such behavJO might be useful in thinking about international conflict resolution .



gency. KeiiPr argues for a form of knowledge, whkh she calls 'dynamic objectivity." "that gran~s to tlte .world around us its independent intc.:grity, hut does so in a way that remams cogmzant of, indeed relies on, our connectivitv with that work!. ~32 K~ll(;r illustrates this mode of thinking in her study of Barbar~ McCiintock, whose work on gen<:tic: transposition won her a , 'obel prize after many years of marginalization by the scientific communlty.3.1 YlcClintock. Keller argues, was a scientist with a respec t fo r complexity, diversity and individual di!Terence whose methodoloe,ry allowed her data to speak rather than imposing explanations on it. Kelle r's portrayal of McCuntock's science contains parallels with what Sandra Harchng calls an African world view.34 Harding tells us tl1at the \\'estem liberal notion of rational economic: man, an individualist and a welfare max:imizer, similar to the image of rational political man on which realism has based its theoretical investigations, does not make any sense in the African world view where the individual is seen as part of the social order acting \vi thin that order ratl1er than upon it. Hard ing believes that this view of human behaviour has much in common with a fernirtist persp ective . If we combine tllis view of human behaviour \vith Merchant's holistic perspective which stresses the interconnectedness of alJ things, including nature, it may help us to begin to think from a more global perspecthe. Such a perspective appreciates cultural diversity but at the same time recognizes a growing interdependence, which makes anachronistic the exclusionary thinking fosterec..l by the nation state system. Keller's dynamic objectivity. Harding's African world view and .\1erchanfs ecological thinking all point us in the direction of an appreciatio n of the "othe r" as a subject whose views are as legitimate as our 0 \\ll. a way of thinking that has been sadly lacking in the history of international relation . Just as Keller cautions us agai nst the construction of a feminist science which coulc..l perpe tuate similar exclusionary attitudes. Harding warns us against schema that contrast people by race, gende r or c:lass and tl1at originate within projects of social domination. Feminist thinkers generally disuke dichotomization and tl1e distancing of subject from object that goes with abstract thinking, both of which. they believe. encou rage a we/th ey attih1de characteristic of international relations. Instead. feminist literature urges us to construct epistemologies that ,ralue ambiguity and Jifference. These qualities could stand us in good stead as we begin to buiJd a human or ungendered theory of iJ1ternational relations which contai ns elements of both masculine and feminine modes of thought.

~lifT(n11t f(rnini<;t_li~r.; ratu~cs that f have r<:vie\\'cd whicl. could help us to bt'gin to lornnrlat< a fern nmt <:ptstemfJiogy of international relations. \1orgcnthau enw11 rage~ us lo lry to stantl back from the world and to think about theory building in terms of c:onstrllcting a rcttional outline or map that has unjver al applicalions. ln contrast, the fe minist_litcr~ture reviewed here ernph asizcs connection and contin -




The first part of this paper used feminist theo1y to den:lop a critique of .\forgenthaus principles of political realism in onler to demonstrate how the theol) and pr.Ktice of

J am deeply aware that there is no one feminist approach but many, wruch come out es in ....__ _ _ __ _..__v_._._.. discinHnP..s ::m~ inh:ll Prtn<> l .-o,l;. ,;,.




. t inten tahomt1re 1 tons rna" exhjbit a masculine bias. Tl1e .second. p<ut sug_ d sornc a ; . , . _1 ~este . t'Ontn.bu t.tons tlllt fPrninist theory m1ght make to 1econc:eptuw iZl11g. some nnportant < ... k. b r . . . t , fonal relations and to tlun mg a out a remmtst epistemo) t>Iements 111 111 ema J : rill . c . . ogy. . on tl tJservations this conclusron "' prese nt c1 remuust refonnuJati D n,,..,ng 1ese 0 L . . alis f , . on 1 c~rl rei~ m this paper, of Morgenth<tU s six ptinciples of po~tical. re m, out rnec. . h . 1t help liS to bemn to tlunk differently about ml<: matio nal relatio wl1rc rmg 1 o . . . h l < ns. shall not use the term realism since fenumsts b~1 ~eve t at t 1cre ~re multiple rea) 1 ities: a truly realistic picture of int~mational pohtr cs r~1~1:t re<:ogmze. ~lcments of cooperation as well as conflict. morality as weU as realpolttLk,. an~ the sbwmgs for justice as well as order.35 This reformulation may help us to tlunk m tl1ese multidirnen. sional tenllS .
1. A femin ist persp ective believes that objectivity, as it is c:ulturaUy defined, is associated with masculinity. Therefore, supposedly "objective" laws of human nature are based on a partial, mascu line view of human natu re. Human nature is both masculine and femin ine; it contain elements of social reproduction and development as well as political domi nation . Dynam ic objectivity offers us a more connected view of objectivity with less potential for domination. 2. A feminist persp ective believes that the national inte rest is multidimensional and contextuall y contingent. Therefore, it cannot be defin ed solely i.n terms of power. In the conte mporary world the national interest demands cooperative rather than zero sum solutions to a set of inte rdepe ndent global problems which include nuclear war, economic well- being and environmentaJ degradation . 3. Power cannot be infused with meaning that is unive rsally valid. Power as domination and control privileges masculinity an cl ignores the possibility of collective empowerment, anoth er aspect of powe r often associated with femininity. 4. A feminist perspective rejects the possibility of separ ating moral command from political action. All political action has moral sign ificance. The realist agenda for maximizing order through powe r and control gives priority to the. moral command of order over those of justice and the satisfaction of bas1c needs necessary to ensure social reproduction. 5. While recognizing that the moral aspirations of partic ular nations cannot be ~quated with unive rsal moral principles, a fe minist persp ective seeks to find co~mon moral elements in human aspirations whic h could become the basiS for de-escalating international confli ct and building international community. 6 A fe~nist ~rs~e denies the autonomy of the political. Since a~ton~ omy 15 associated With masculinity in West ern cultu re, disciplinary efforts ~0 construct a world vi~w which does not rest on a pluralistic conception of :~nature are partial and masculine. Building boun daries around a nar ~ Y efined political realm defin es political in a way that excludes tbe concerns and contributions of women.

goes hcyont1, uotlt masculine and fctninine perspectives. Such indusionary thinking, as Simonc de Beauvoir telb us, values the bringing forth ofLife aS much as the tisking or ]jfe; il is becoming imperative in a world in whjch the technology of war ancl a fragile natmal environment threaten human existence. An ungenderecl, or hu man, cliscourse becomes possible only when women are aclequately represented in the discipline and when there is eyual respect for the contributions of women and men alike.

To <onslru<:t tl11s feminist altemativP i~ not to dcuy the validity of Yforgenlhau\ work. But adding a feminist p<..rspe<:liw: to the epistemology of international relatiom i~ i.t stage through wb ieh we must pass ih"-e are to think abou t u mstrucling an nngendcrcd or ltuman s<:icnet. of inte:mational politics which is sensitive to. hut

An earlie r version of this paper was presented at a symposium on Gender and Intemational Relations at the London School of Economics in June 1988. I "voulcl Uke to thank the editors of Millen nium, who organized this symposium , fo r encouraging me to undertake this rewriting. I am also grateful to Hayward Alker Jr. and Susan Okin for their careful readin g of the manu script and helpful suggestions.
Unive rsity Press, 1986), p. 148. 2. In 1987 only 4.8 per cent of the top career Forei<rn Service employees were wome n. Statement of Patrida Schroeder before the Committee on foreign Affairs, U.S. House of Representatives, p. 4; Women~~ Perspectives on U.S. Foreign Policy: A Compilation of Views (vVashington. D .C.: U.S. Go,em ment Printing Office, 1988). For an analysis of women's roles i11 the American military. see Cynthia Enloe, Does }\)wki Become )ou? The Militarisation of\Vomen:s Lives (London: Pluto Press, 1983). 3. Euward P. Crapol (ed. ), Women and American F(,.reign Pol cy (Westport. Corm.: i Greenwood Press, 1987), p. 167. 4. For an analysis of the role of masculine language in shaping strategic thinking see Carol Colm, ''Sex and Death in the Rational World of Defense Intellectuals,' Signs: j ounw l of Women in Culturf and Society (Vol. 12. ~o. 4. Summer 1987). 5. The claim for the dominance of the realist paradigm is suppo rted by John A. Yasquez. .. Colouring It Morgenthau: New Evide nce for an Old Thesis on Quantitative Lnternational Studies," Britislt jounw l of Intentational Studies (Vol. 3, No. 5. October 1979), pp. 210-28. For a critique of ~l o rgent11aus arnbiguous use of language see Inis L. CJaude Jr., Power and Jntenw tional Relations (i\ew Y ork: Random House. 1962). especially pp. 25-37 . 6. These are drawn from Ha11s Morge nthau, Politics Among Nations: The Stm gglc fm Power rmd Petu:e, 5th revised edition (l\ew Y Alfred Knopf. 1973), pp. 4-15. I am ork: aware that these principles embody only a partial statement of Morgenthaus ,ery rich study of intemational politics. a shtdy which desetves a much more detailed analysis than I can give here. 7. This list is a composite of the male/female dichotomies which appear in Evelyn Fox Keller's Reflections on Gender and Science (~ew Il aven. Conn.: Y University Press, ale 1985) and Harcling, op. cif.
l. Quote d in Sandr a Hardin g, The Science Question in Femi11sm (lthaca. N.Y.: Cornell i





[$Site (\ ol. 2S. No. . . n hold Anm that these perceptions stl . 9. K eller. ap. cif .. P 130. 10. Ibid.. P 9. 11 . Ibid.. p. 9. ~ male Culture. Woman Culture and Concvptual Change: l2. Sat"'a Ann Ketchum. Fe, . Studies.- Social TheonJ and Pm et ic1 (\"ol. 6. r0 _ 2 To"~ard a Philosophy of \\ omens . ,

V gel Donald M. Bro,ermau, Fnmt :larkson <tnu 0 , K Bnwcrman. Susan H. s A Current Apprni~al. , rwl of Social , . Ing~ . ..5 x-role Stereotype . _ . h". . ~ . l1ost:'nkmn7.. e. _ -g Repltcaoon of t I!' 1es( aH 1 1 t 1 t 980s con. l' P,m 1 . . 2 l9i2), PP ::>9_,


Summer 1980). h H0 bbes's state of nature provides an ate-urate desc1 ip13. OtJ1ers have questi~ned whet er S for example Charles B~:>it~.:, Polit icol Tlteory and tion of the intemaoonal sy~tem. eeN . Princeton Unjversity Pres.. 1979). pp. 1 R 1 f'0 1s (Pnnceton. 1 -1 0 1 Intemationo e ' n. . Be ond Borders (Syracuse. ;-.l.L Syracusc University and Stanlev HofTmann, vtttws y . Press, 198i). chap. fl. raJ d . 1 pment are described and discussed in Hobert Kegan 1 KoIll>e rgs staaes o mo . eve oPmcess in Human Development (Cant I)ricl ge, Mass.:. 14. c bl . The Evolving Self Pro em ant1 . H d University Press, J982), chap. 2 . _ _ atvar .. an In 0 Different Voice: P sychological Theory and Women_ s. De~elupme 11 t 1"- Carol ~~dllig ~~ _H rvard Universitv Press, 1982). See chap. 1 ror Gtlhgan s critique (Cambn ge. ,v,ass.. a

of Koh~ber~d th t toward the end of his life, Morgenthau himselr was aware that 16. There IS e\1 ence a , _. . . . _ his own prescnp t ''ere becoming anachrorustic.. In a semmar presentation m .19t8 1ons . d f. he sugges ted that power politics as the guiding pnnctple for the con uct o - m temational . . f 1 relations had be<,-ome fatally defective. For a descnption o t liS semmar presentation see Francis Anthony Boyle, World Politics and International Law (Durham , ~.C.: Duke Universitr Press. 1985}, pp. 70-4. 17. Nancv c. M. Hartsock, Money. Sex and Power: Toward a Feminist Historical Matetialism (Boston: ~ortl1eastem Unhersity Press, 1983). p. 210. 18. Hannah Arendt, On Violence (~ew York: Harcourt, Brace and World 1969), p. 4-!. Arendt's definition of power, as it relates to international relations, is discussed more extensively in Jean Bethke Elshtaln's "Reflections on War and Political Discourse: Realism, Just War. and Feminism in a Nuclear Age," Political Theory (Vol. 13, No. 1, February 198.5), pp. 39-57. 19. Da\id McCieUand, "Power and the Feminine Role," in David McClelland, Power: The lntzer Experience (New York: Wiley, 1975). 20. Jane S. Jaquette, "Power as Ideology: A Feminist Analysis," in Judith H. Stiehrn (ed.}, Women$ Views oftlze Political World of Men (Dobbs Ferry, N .Y.: Transnational Publishers, 1984). 21. ~ examples are cited by Christine Sylvester, 'The Emperor's Theories and Transformations: Looking at the Field through Feminist lnses," in Dennis Pi rages and Christine Sylvester (eds.}, Tran.ifonnations in the Global Political Econom.y (Basingstoke: Macmillan, 1989).

T_hi:-; IS th( flr~tll ~l-Ilt maur lly Edwarcl Azar and Clmng-in ~loon, ..Third World . atJOnul ~~ runty I 0\\ arc\ a \ cw Conce-ptual F rauwwork.'' I r~tcrnnlioual I 11temctiom 1 \ ol 1 ' 0. 2 19~-1 pp. I 03-.'35. 25. Johan C' tl" 111t! \:iobKl' Pcac.e and Pcace Research.'' in Caltung. Essays in Peace Rc.~carc/1 \ ol I Copcnh.,~cn: Clni-;tian Ejlcrs 1975). 26. See. for C'x.unplc Cita Scu all(! Caren Crown. Development, Cri.'ies and Altcmwtit;e \"i.sion<; Tl11 rd \\'orld \\ (JIIll'lr ~s- Pe~peetiJ;es l :"ew Y ork: \llonthlv Review Pre s. J9S7). This i~ an example of u growing literah1re on v.:omen and dc,el~pment ,, hic:h deserves mor<> attention frou1 tire intcn mlional relations cornmunitv. 2i. Carol)11 ~l<:n:ha.11t. The Death r~{.\1ature: Women, Ecology and the Scientific Rewlution (l\ew Y ork: I larper and How, 19&2). p. xv. 28. Sarn Jh u.klic:k, " ~ atcmal Thiuking and ''Preservati' e Love and Militan Destruction: Some R(A ections on Y1othcring and Peace.'' in Jo~ce Treblicot, .\lotl,ering: F:.ssays in Feminist 111eory (Totowa, K J .: Howrnan and Alienhead, 1984). 29. For a 1norc extensive analysis of this issue see Jean Bethke Elshtain, W rmum and War ( i ew York: Basic Book5, 1 987 ). 30. This type of conHitt resolution contains similarities witl1 the problem sohing approach of Eclward Azar, John Burton and Herbert Kelman. See, for example. Edward E. Azar and John \V. Btuton, h1ternational Conflict Re.~olution: Theory and Practice (Brighton: Wheatsheaf, 1986) and Herbert C. Kelman, "interactive Problem Sohing: A SociaiPsychologital Approach to ConAict Resolution,'' in \V. Klassen (ed. ). Dialogue Toward Inter-Faith Understanding (Tantur!Jerusalem: Ecumenical Institute for Theoretical Research. 1986), pp. 293-314. 31. Gilligan, op. cil., pp. 9-10. 32. Keller, op. cit. , p. 117. 33. E,elyn Fox Keller, A Feelingfor the Organism: The Life nnd Work ofBarbora .\ lcClintock (New York: Freeman, 1983). 34. Harding, op. cit., chap. 7. 35. ''Utopia and reality are ... the two facets of political science. Sound political thought and sound political life will be found only where both ha,e their place": E. H . Carr. The Twenty Years Crisis: 1919-1939 (New York: Harper and Row. 1964). p. 10.

22. Karl w..Deutsch et al., Political Community and the North Atlantic Area (Princeton, N.J.: Pnnoeton University Press, 1957}.
23 New thin.lcin 15

onnulations under Gorbachev. There are indications that the Soviets are to conceptualize security in the multidimensional tenns described here. see p~~~ ~ Theory of lnte1'7UJtWnal Relations {New York: St. Martin's



hat 1s also being used in the Soviet Union . refi g a_ term t to describe foreign

Th e Anarchic Structure oJWorld Politics

Only throu gh some sort of system s theor y can international polities be unde rstood. To be a success, such a theory has to show how international politics can be conceive d of as a doma in distinct from the econo mic, social, and othe r international domains that one may conceive of. To mark international-political systems off from other inte rnational systems, and to distinguish systems-level from unit-level forces, requi res showing how political struc tures are gener ated and how they affect, and are affec ted by, the unHs of the syste m. How can we conceive of international politics as a distin ct syste m? What is it that interv enes betw een intera cting units and the results that their acts and inte ractions produ ce? To answer these questions, this chapt er first examines the conce pt of social struc ture and then defines struct u re as a conce pt appro priate for nation al and for intern ation al politics. A system is comp osed of a struc ture and of intera cting units. The struc ture is the syste m-wide comp onen t that makes it possible to think of the system as a whole. The probl em is ... to contrive a definition of struc ture free of the attrib utes and the intera ction s of units. Defin itions of struc ture must leave aside, or abstra ct from , the characteristics of units, their bebavior, and their intera ctions. Why must those obviously impo rtant matte rs be omitted? They must be omitt ed so that we can distinguish be tween variables at the level of the units and variables at the le\'el of the system. The probl em is to devel op theor etically useful conce pts to replac e the vague and varying system ic notions that are customarily empl oyed -noti ons such as envir onment, situat ion, conte~1:. and milieu. Struc ture is a useful conce pt if it gives clear and fixed mean ing to such vague and varying te rms.
From Kenneth N. Waltz. Tlteoi1J of Intenwtional Politics. pp. 79-106. Copyright 19i9 by \I<.Crdw-llill. Inc. Reprinted with permission of The McCraw-Jlill Companies.

. any .. Il,\\.e to omit from . definition d .' t n. if the deRl)JtJon \ r \ know\\ hat W<" I I . l . . ll . Abstmcting Irum t 1e attn H ... v i .ml~ means 1 . 1 ' 1nl J t o I>t> thcOJ<.'tiC ki Is of political le<K1 ~~ l :. ers. soc:w 1' 'I cconmnic institeaVJng t' 1 ' f n thout t 1e ne u IOns ~t$10l' ques 10 !i . . ts stttes may have. Abstra cti ng from relations 1 l octl com nutme n ' . means antl lt eo 0~1 ' . b t the cultural econom1c. po1 ..1 .tnd mllitary int .ilJCd.l, . . . a ou . . . era<:. lt'a\1ng ,\SIde questwnsI t. to be left out 'does not tm1. 1cate what ts to be put in Th . f t tes To sa, w la IS . . . non. o s a t ertheless becaus e tl1e mstruction to omit attrb e 1 utes . t 1s importan nev . < neg<ttl' e pmn d th . t . 1 . ft . 1 ted ul . e illS n ction to omit interactiOns almost always goes uno). 1 10 IS o en " ~ '.b d interactions are omitte d. what i left? The question . ened. But tf att~dut~s antl double meanincr of the term 'relati on.' AsS. F "adlSJ 0 db 1s 1 enng 1e ' < mswere ~ eo~ . cre obscUIes a distinction that is impor tant in theo e . ts out ordinal) 1 anguao . f. . pom ' d to n lJoth the interaction o 11 m ts arKl tI11':: positions tJ1ry. "R 1 " is use mea ey e abOJ~ , . th t To define a structu re requ ires ignoring how units relat cupv \'ls-a-viS eac1 o .er. 1 . 1 I e CX: ' th (l10 , the)' interact) and concen trating on 10w t 1cy stand in rela. w1th one ano er v. . they are arrangecl or pos1t10ne d). I n t e ractions, as 1 have tion to one anotller (1 lOW . . d . .I . . . k 1 t the level of the umts. How u111ts stan m 1e ation to one IJ1SJSted, ta e p ace a .. . . e arran<Yed or pos1bone d , 1s. not a P'.opert) o r tI1e un1ts. The anothe r the way they ar o . ' ent of units is a property of the syste m. arrange m tl B leaving aside the personali~ o f actors, tl.1e1r be 1 <omc1. ~err t~teractions, 1~\'lOr,_ y. ely positional tJicture of society. Th1 ee propos 1t10ns follow from one arnves at a pur . . . . . . l .. . . . F' tructures ma) endure while personality. behaVIOJ , anc mteractions ,ary th1S. trst, s d f . S d widely. Structure is sharply distinguishe rom ac~ons an d . rac t mte 10ns. econ , a structural definition applies to realms of widely differe nt sub~tance so the arrange ment of parts is similar. z_ Third, be~ use thls is so, :~~~ones d_eveloped for one realm mav with some modification be applica ble to oth e1 1ealms as well .... The ~oncept of structure is based on tlle fact that units diff:re nt1y ju.xtaposed and c:ombined behave differe ntly and in interac ting produ ce diffe rent outcomes. 1 first want to show how internal political structu re can be define d. In a ~ook on international-political theory, domes tic political structu re l~as to be examm~d in order to draw a distinction between expect ations about behav1or and outcomes'" the internal and external realms. Moreover, consid ering domestic political structure now will make the elusive international-politic al structu re easier to catch later 011 Structure defines the arrangement , or the orderi ng, of the parts of a system[ Structure is not a collection of political institu tions but rather the arrangement 0 tl1em. How is the arrangement define d? The consti tution of a state describ es 50~1~ parts of the arrangement, but political structu res as they develo p are not id~nb~ . with formal constitutions. In defining structu res, the first questi on to answer 15 this. \\'hat is the principle by which the parts are arrang ed? . . . nd Domestic politics is ruerarchically ordere d . The units- msti~b ~ns agenc ies-sta nd vis-a-vis each other in relatio ns of super- and subordm~oon. bout ordering principle of a system gives the first, and basic, bit of inform~tlo;;h , of how the parts of a realm are related to each other. In a polity the bier )ela offices is by no means completely articul ated, nor are all ambig uities abou~ rfor tions of super- and subordination removed. Never theless, political acto~s ~e tind mally differentiated according to the degree s of their author ity, and theJrth 5 J and fun<:tions are specif ied By "speci fied" I do not mean that the law of e




fully dr .,( ) t ' l~le dulif-'s that <.liffc.:nnt agenci<:!. perform, but on I) thal broad agrt'C'lll C'HI pre,mb o~ the: ta-;ks tlml va1 ious parts or a government are to unclcrtakc and 01.1. tit<. e~tcnt o! the power they legitimately v..ield. Thus Congress :.11pplies th<" 1111~1tary forces; the President co10mands them. Congre.o,s makes the law:.; the e,xcc:ulJ\'C ~J~an~h cn~orccs them: ~genc:ies admini.o,ter laws; judges interpret them. Such spcc.Aic:atJon of r<1les and differentiation of functions is found in any state, the Jllorc full y so_ a~ the ~late is more highly develope<.l. The specifkation of func..:tions of fonnall y clrfferentli:lted parts gives the second IJit of struc:tu ral information. This second p<llt of the definition adds some content to the structure but onlv enoucrh ' ; 0 to say 111ore fully how the units stand in relation to one another. The roles and the functiOJ:s of the British P1ime Ministe r and Parliam ent, for example. differ from those of th e American President and Congress. When offices are juxtaposed and functions are combined in different ways, differe nt behaviors and outcomes result, as I shall shortly show. The placement of units in relation to one another is not fully definec.l by a system s ordering principle and by the fo rmal differentiation of its parts. The stancling of the units also changes with changes in their relative capabilities. In the performance of their functions, agenci es may gain capabilities or lose them. The relation of Prime Minister to Parliam ent and of President to Congress depends on. and varies with, their relative capabilities. The third part of the definiti on of structure acknowledges that even while specified functions remain unchanged, units come to stand in different relation to each other through changes in relative capability. A domestic political structure is thus defined: Brst, according to the principle by which it is ordered; second , by specification of the functions of formall y differentiated units; and third, by the distribution of capabilities across those units. Structure is a highly abstract notion, but the definition of structure does not abstract from evel)rthing. To do so would be to leave everything aside and to include nothing at all. The three-prut definition of structure includes only what is require d to show how the units of the system are positioned or arranged. Everything else is omitted. Concern for tradition and culture, analysis of the character and personality of political actors, consideration of the confli<.:tive and accommodative processes of politics, description of tl1e making and execution of policy -all such matters are left aside. Their omission does not imply their urtimp01tance. They are omitted because we want to figme out the ex-pecte d effects of structure on process and of process on structure. That can be done only if structtue and process are distinctly defined. I defined domes tic political structures first by the ptinciple according to which they are organized or ordered, second by the differentiation of units and the specification of their fu nctions, and d1ird by the distribution of capabilities across units. Let us see how the three terms of the definition apply to international politics.

I . Ordering Principles
Structural questio ns are questions about the arrange ment of the parts of a system. The parts of don1es tic political systems stand in relati.ons of super- and suhordinaion. Some are entitle d to commru1d; otl1ers are require d to obey. Domestic


PART 1 ANARCHY ANU 11;, '-'vn..,...'< - - .


. hi The parts of intem ati' a li tical ~Yst S\ tt>ms art' rentralized and hJe:arc Fc.nnally each is the equ. ' all the otherns . t' ' stand in relattons o coordi ndation. . o required to obev. Inter n lonal svstemsers . 1d mman . none ts , are ~one is entit e to CO . 'h dering principles of the h'-'O ': 11ctures are dis. decentralized and anard uc. T e otr eh other. Domestic politK.u structures ha .J:rr . deed cont:raf\1 o ea ve tinctk WJ,erent, m d rn ' as tlteir concrete coun terparts. . ternation I In . al . tit tions an o ces . go\'ernment tns u aUed "politics in the a bsence u f' govemment a "1 . m con'"-43 has been c . <+ . politics. u ',' d . t. and in e,er-growtn g nu m I)ers. Supranational . al rgamzattons o e:os . . lntemation o . ;tivel 1 however, either t hemse1 ves acym re some of agents ~ble to act eff~ties ~f states, as did the ~e~eval papacy in the era of the attnbutes and capab . th oon reveal thelf ;,..abilitv to act 10 1mportan t wa\s except With u~ / Innocent Ill, or ey s h . cence of the princ ipal states conc erned with the the support. or at least t e acql wes ts of ~uthoritv emer ge intern ation allv are bare))r t hand Wbate\'er e emen , matters a bility that provides the foun dation ~or tl1e appearance once remo,ed from th~ ca~a . klv reduces to a partic ular expression of capabilof those elements. on~ q~~ system-wide authority, form al re lations of super ity. In the absence o agen WJ devel d su bordi nation fail to tructuop. an f ral definition states the principle by which the S\'SThe first term o a s ' . ed S tu an organizational concept. T 1 prom me nt c haracteris1e tem JS order true re 15 . . a1 l'ti'cs however seems to be th e Iack o f order and of tic of mtemation po 1 ' ' . . H . one think of international politi. cs as b emg any ~.., d of an tun orgamzabon . 0\\ can l' d f rder at all?The anarc-hy ofpolitics internationally is often re rerre .. to. I structure o . ti a1 concept the terms "structure , an d "anarchy seem to be . m IS an orgaruza on , .. . b f contradiction. If international politics is ..politics ~~ the a ~ence o government:,. what are we in the presen<.-e of? In looldng for mtemati~n_al struct~re, one 15 brought face to face with the invisible, an uncomfortable ~ostti on to be m. 1De roblem is this: how to concei\e of an order w1thout an orderer and of ~ effects where formal organization is lacking..Beca~e these ~e difficult questions, I shall answer them through analogy w1th. mJCroec:o~ omtc theory. Reasoning by analogy is helpful wher e one can move from a domam f~r which theol)' is well developed to one where it is not. Reasoning by analogy 15 permissible where different domains are structurally similar. . Classical economic theory, developed by Adam Smith and his followers. 15 microtheory. Political scientists tend to think that micro theor y is theory about small-scale matters, a usage that ill accords with its estab lished mean ing. The tenn ..micro.. in economic theory indicates the way in which the theor y is constru~ted rather than the scope of the matters it pertains to. Micr oecon omic theor y descnbes bow an order is spontaneously fonne d from the self-i ntere sted acts and interac tions of iodi\idual units -in this case, persons and finns. The theor y then ~s upon the two central concepts of the economic units and of the market. Economtc units and erooomic markets are concepts, not descr iptive realities or <.:Oncrete entities. This must be emphasized since from the early eight eenth centu ry to the present. from the sociologist Auguste Com te to the psychologist Geor ge Kato~~ ecooomic theory has been faulted because its assumptions fail to correspond \' 1 . . ~g~ lilies ~. economi~ theorists ~-nceive of an ~'Onomy opera t the ldaboo &oua its SOCiety and polity. Unrealistically, economiSts assume tha


econorn ,. worlJ i. tlw w0rld of the world . UnreaJi<;ti<:allv, ec:onomL'>ts think of thfading udl. the f:uno~:s "<:c:onomic man.- as a sin~le -;nindcd profit maxirnizC'r. They '>i ~JP 0 1 t 'In(' aspect c1f rnan and l<:a... e aside the ,.,.ondrous variet y of human life. A'> aJ) mr>denltely sensibl<: econ01ni~t knows, "economi<.: man" doe~ not exist . Anycm< whl) ask!) businessm(;n how they rnake thei r dedsions \\ill find that the assumption l.hat rncn are c<:onomic maximizerl. grossly di~torts their characters. The assurnptwn tlwt men be:ha'-eas ~onomic men , which is knO\vn to be false as a descripthe statement, turns out to be useful in the const n;c:tion of theory. yfarkets are the second major c<mcept invented by microeconomic U1eorists. T\.\'0 general questions muc;t be aske.d about marke t!>: How are thev formed? How do they work? Tir e answe r to the first question is this: The market of a decentralized economy is individualist in origin , spontaneously generated. and unintended. The market arises out of the acti\ities of separate units -persons and firm~ whose aims and efforts are dire<.:ted not toward creati ng an order but rather toward fulfilling their own intem aJiy defined interests bv whatever means they can muste r. The indi vidual unit acts for itself. From the coa~tion of like units em~rges a structu re that affects and constrains all of them. Once formed , a market be<.:omes a force in itself, and a force that the constituthe units actinO' singlv or in small numo ' bers cannot c:ontrol. Instead, in lesser or greater degree as marke t conditions v;u:. the creato rs become the creatu res of the marke t that their ac:ti\ity ga,e rise to. Adam Smith 's great ac:hiC\ement was to show how self-in terested, greed-drive n action s may produce good social outco mes if only political and social conditions permit free competition. If a laissez-faire economy is harmonious, it is so bec-ause the intentions of actors do not correspond 'dth the outcomes their ac :tions produce. What intervenes behveen the ac:tors and the objects of their action in order to thwar t their purposes? To account for the unexpectedly fa,orabJe outc."O mes of selfish acts, the concept of a market is brought into play. Each unit seeks its O\\ll good; the resuJt of a number of units simultaneously doing so transc:ends the motives and the aims of the separate units. Each would like to work less hard and price his produ ct higher. Taken together, all ha,e to work harde r and price their produ cts lower. Each 6nn seeks to increase its profit; the result of many firms doing so drives the profit rate dow-nward. Each man seeks his own end, and, in doing so, produces a result that was no part ofrus intention. Out of the mean ambition of its members, the greate r good of society is produced. The marke t is a cause interposed between the ec."Onomic ac.tors and the results tl1ey produce. It conditions their calculations, their behaviors, and their interactions. It L not an agent in the sense of A being the agent that produces out("Ome X. e; Rather it is a structural cause . A marke t constrains the units that comprise it from taking certain actions and disposes them toward taking others. The market. created by self-djrected interacting economic; units. selects beha"iors aC(.."Ording to their consequences. The marke t rewards some with high profits and assigns others to bankruptcy. Since a market is not an institution or an agent in any concrete or palpable sense, such statements become impressive only if they can be reliably inferred from a theol) as part of a set of more elaborate expectations. They can be. Microeconomic theory e>.-plains how an ec."Onomy operates and why certain effe<:t:s are to be expected ....

'--. '

--....,. . ..... .. _...


- ~

luternationitl-political s~'Stems, like ~conomi<' tllar~ds , arc. formed b , . . . t-o-achon o1 se'If- t w units 1nte stru ctu res arL d<'fined in termstht ) . . . tlw ptin ttm politicalunjts of an era, be they Cl~' tatcs: etnpt re~. or nations. St . of . j' . o tun? em erg e 10m tl te c exis tence. of states. l'\0 state' ,mtc.~nds. to r)atticiptte . rue. . . . < m . . . t tur> b, which 1 and others \>\1ll 1 C con~tra.uwd. Inte rn l' tit. t ) e Ionn aho n o1 a s llJC ... _, . cl. 'd . . . . . . a lOnaJ . .. _ like economic markets. are 111 tvl tklJ St 111 ong1n spont-tn 1 1 poIttic~u S\'Stems. . . d . t ded In both systems , struc.1ures are lormE>d 'bv thE' <c eousl) . generate d . an urun en . '. , . . .~ " oachon .... f th . . t \\fhethe r those umts U prospe t, OJ c.he dep end s on thei ve. 0 en Unl S. . , . d . . . r ow11 rt: rts Both svstems are formed and mall1tame on a pl mclple> of sE:" If-h elp tl. euo . , lat applies to the units. . . . . . .. . In a mic rotheo.ry, whether of mternational p_ol~tiCS 01 of e~o nom 1.c . the ll"lo~s "ah.on o f the ac tors 1 assun1ed rather tl1an reahsttcall y clcscnbed. I assun1e tl1at . . . . states seek to ensure their survhal. The assumption IS a radical stmplification mad e for the sake of constructing theo ry The que stw n to as~ of the assumption. as ever . is not whether it is true but whe the r it is the most sens1bl c and use ful one that ea~ be made. Whether it is a useful assumption dep end s on whe the r a them y based on the assumption can be contrived, a theory from _w~ich imp ~rtant consequences not otheiWise obvious can be inferred . Wh eth er 1t ts a sens1ble ass um ption <.;an be directly discussed. Beyond the survival motive, the aims of stat es may be endlessly vari ed; they may r~ge from the ambition to conque r the world to the desi re me rely to be left alone. Survival is a pre requisite to achieving any goals tl1at states may have, othe r than the goal of promoting thei r own disappe aran ce as political entities. The survival motive is taken as the ground of action in a worJd whe re the security of state s is not assured, rather than as a real istic description of the imp ulse that lies behind every act of state. The assumption allows for the fact tha t no stat e always acts exclusively to ensure its survival . It allows for the fact tha t som e stat es may persistently seek goals that they value more highly than survival; the y may, for example, pref er amalgamation with other states to the ir own survival in form . It allows for the fact that in pursuit of its security no state will act with per fect kno wle dge and wisdomif indeed we could know what those term s might mean ... . Actors may perceive the structu re tha t constrains the m and understand how it serves to reward some kinds of behavior and to penalize oth ers. But then agai n they either ma)' not see it or, seeing it, may for any of many rea sons fail to conform their actions to the pattern s that are most often reward ed and least ofte n punishe d. To say that "the structure selects" means simply tha t those who conform to accepted and successful practices more often rise to the top and are like lier to stay ~ere. The game one has to win is defined by the structu re that determ ines the kind of player who is likely to prosper. . ..



2. The Character of the Units


~ sec.:ond ten n in the definition of dom estic political stru ctu re spe cifies the func ~ rrfo~ed by differentiated unite;. llierarchy entails rela tion s of super- and ~ m:;:ron ~ong a system's parts, and tha t implies the ir differe ntiation. r~

rung ornestic political structure the sec ond term like the first and third, 15

slat es tl nt are the umts of intemalional-political systems arc not formaUy differentint(d h~ th<.. !unctions they perform. Anarchy entails relations of coordination UJ non g .1 sy.;;ttm \ units. anu that implies their samene ss. The second term is not needed in defining internat ional-political structure, because , so long as anarchy end ures. states remain like units. L natiomJ stru ctures \'ary only through a nter change of organizing principle or, failing that, through variations in the capabilities or lltlit<;. :\(-'VeJtheless I shall discuss thcst:: like un its here, because it is by their in teractions that international-politics structures are generated. Two questions a rise : Why should stat es be taken as the units of the syst ern? Given a wide vadety of states, how <.;an one call them "like units"? Questioning the choice of' states as the primary units of i.ntemational-poH tical systems became popular in the 1960s and 1970s as it was at the turn of the century. Once one understands what is logically involved , the issue is easily resolved. Those who que stion the state-ce ntric view do so for two main reasons . First. states are not the only actors of importance on the international scene. Second , states are declining in importance, and other actors are gaining, or so it is said. Neither reason is cogent , as the following discussion shows. States are not and never have been the only international actors. But then structures are defined not by all of the actors that Aourish within them but by the major ones. In defi ning a system's stru cture one chooses one or some of the infinitely many objects comprising the system and defines its structure in terms of them. For inte rnational-political systems, as for any system, one must first decide which units to take as being the parts of the system. Here the economic analogy will help again. The structu re of a market is defined by the number of firms zco mpe ting. If many roughly equal finm con tend, a condition of perfect competition is app roximated. If a few finns dominate the market. competition is said to be oligopoHstic eve n though many smaller firms may also be in the field. But we are told that definitions of this sort cannot be applied to inte rnational politics because of the interpe netratio n of states, because of thei r inability to con trol the emironmen t of their action. and because rising multinational corporations and other nonstate actors are difficult to regulate and may rival some stat es in influence. The importance of nonstate acto rs and the extent of transnational activities are ob, ious. The conclusion that the state-centric conception of inte rnational politics is made obsolete by them does not follow. That economists and economically minded poli tics scie ntists have though t that it does is ironic. The irony lies in the fact that all of the reasons give n for scrapping the stat e-centric concept can be related more strongly and applied to firms. Firm s <.;om peting with numerous others ha,e no hope of <.;Ontrolling their market , and oligopolistic firms constantly struggle with imperfect success to do so. Finns iute11>enetrate. merge, and buy each up at a merry pace. Moreover, firm s are constantly threatened and regulated by. shall we say, "non firm actors. Som e governm ents encourage concentration : others work to prevent it. The market stllJCture of parts of an economy may move from a wider to a nar rower competition or may move in the opposite direction, but whate,er the exte nt and the frequency of change. market stntctures, generat ed by the interaction of firms, are defined in terms of them.

nC'tcled hc<nmt each tem1 points to a po:,sible :,ource of structural vadation. The

p,i\RT l

ANARCHY ANU 11 ::>'-Vi' '""'"''< - - .

~\,;ng that the nation-state IS_J~ Id have' to be rcddinpd Th t 1cn the , . . 'f . ti0 11'1) l)OhhCS WOU . a Would b strudure o nttema ' . , . bilities cannot be separated fro m the otl . e . economiC capa 1e1 cap ne-<.--es ~u:' because . . ..; 1 frequently drawnbetwPcnmiltte::rso f higha _, a.l. . 1 b11ties of st;a tes The distmcuOl e economic means c m1 1tary and politicalnu 1O\v ror . . . placed States us f . po l1hcs IS nus . for the achievement o economic interests encJs. t . . 1d polibca1nJean.5 I ld . .. <Hld mili ru: ,u . f Kindleberger's statement may 10 : Some state A 1ded verswn o h s rnav .n amel d u as economic entities. and otl~ers not.~ _at p_ ses no prohle~ o be nearly ' "'d_5be P.. al theorv since internatiOnal pohtlcs 1s mostlv .11 c t mat10nal-po1tic 1 ., . " )Out 1 rnaJor actors, the struc:tu ror 111 e S 1 gas the major states are t 1e . alities amway. o on . mequ . a1' .. . 15 defined in tenns of them. Tl1at t I1eoretical statement is ref . 1 of intemanon po !tiCS actice States set tJ1e scene w Juc1 t1 . m 1 1ey, along '"ith rse home out m pr . . l cl cou tl . dramas or carry on the1r 1um rum a rans. Though the nonstate actors, state letr . . . , . {' . 1 . ) . t e little in the affrurs of non state act01 s 01 ong penods of may choose to m er er . I I -l . ss set the terms of mtercourse, w 1et 1e1 )y passively permittime, states nevertl1ele . . . 1 . ting m f.orma1 ruIes to develop or by activelv mtervemng to c 1ange rules that no " . . th When the crunch comes. states remake the rules b, wh1ch other longer sUit tem deed one mav be struck by t11e abilitv of \\'eak states to impede l. actors opera e. n ' " . I . tile operation of strong international corporations an ))' t 1e attention the latter pay to the wishes of the fonner. . . . States are the units whose interactions form the stmctu re of mternatlonalpolitical systems. They will long remain so. The death rate .~~11ong ~t~:~s is remarkablv low. Few states die; many finns do .. . . To call states hke umts 1s to say that ea~h state is like all other states in being an autonomous political unit. It is another way of saying that states are sovereign. But sovere ignty is also a bothersome con cept. Many believe, as the anthropologis t~. G. Smith has said, ~at ''in a syst~m of sovereign states no state is sovereign.'>t> The error lies in iden tify1ng the soveretgn~ of states with their abiUty to do as tl1ey wish. To say that states are sovereign is not to say that they can do as they please, that they are free of others' influence, that they are able to get what they want. Sovereign states may be hardp ressed all around, constrained to act in ways they would Hke to avoid, and able to tlo ha~dly anything just as they would Uke to. The sovereignty of states has never enttuled their insulation from the effects of other states' actions. To be sovereign and to be dependent are not contradic.1:ory conditions. Sovereign states have seldom led fr~ and easy lives. What then is sovereignty? To say that a state is sovereign mea.ns th~t it deddes for itself how it will cope with its internal and exte rnal problems, 1nclu whe ther or not to seek assistance t mg from others and in doing so to ,, 1 s free utmt . dom by making COmffiltments to them. States develop their own strategJes ' chart ds their own courses, make their own dedsions about how to meet whatever nee 1 they experience and whatever desires they develop. It is no more contradictory ~ that . . 15 to Sil\ . say soveretgn states are always <."Onstrained and often tightly so than Jt that free individuals often make decisions under the heavy pressure of events. J e Each state, like every other state, is a sovereign political entity. And_)~~~ ~:e differences CK.Toss states, from Costa Rica to the Soviet Union, from Gambia

. l fi ukets in terms finll~ ,,) J d fine int'-'J . . m ts c e 111C nt. '- l'l<ltio 1 just as t'eonor ~ . f states. If Charles P. "-111 1 b<'rger were . na1 . . ' t cturcc: m tellns . o . t about through a..'i .c nght 1 .11 pt) l1hc:u :; n l an P<.ouom1 unit "5 tl




re .


un il(u St,lt.~. m imtnense. ~ tatcs an: alike. and ar<. aJso diiTen.:nt. ~o are <:orpor.tlit 11s, rpks, un iw r'> itics, ancl pt:ople::. \\1c~ever wc put two or more objc>Ctl> Ill " 1 .;,urw c.:ategof) . we arc sa)ing that they ar<:; alike not in all respects but in som<'. '\o two obJeC.: ts in this world are i<lentical yet thcv <:an often be usefullv <:Ontp<H'<'d and co n: hined. "YOlt c;m't < apples and oranges" is an old sa}i ng that dd s<;ems lo hv c<;pecmlly popular among salesmen who do not want you to comp< tre their wares with others. But W(; all know that ti le trick of addjng dissimilar objects is to ~xprcss the result in terms of a category that comprises them. Three apples plus four oranges equals seven pieces of fruit. The only interesting question is whe.thcr the category that classiAes objects according to their common qualities is useful. One can add up a large number of widely varied objects and say that one has eight mil lion things, but seldom need one do that. States va1 widely in size, wealth, power, and form. And yet variations in these y and in other respects are variations among like un its. In what way are they like units? How can they be placed in a single categ01y? States are alike in the tasks that they face, though not in their abil itjes to perform them. The differences are of capability, not of function. States perform or try to perform tasks. most of which are common to all of them; the ends they aspire to are similar. Each state duplicates the activities of other states at least to a considerable extent. Each state has its agencies for making, executing, and interpreting laws and regulations, for raising revenues, and for defending itself. Each state supp}jes out of its own resources and by its own means most of the food, clothi ng, housing, transportation , and amenities consumed and used by its citizens. All states, except the smallest ones. do much more of their business at home than abroad. One has to be impressed with the functional similarity of states and, nm.v more than ever before, vvith the similar }jnes their development follows. From the rich to the poor states. from the old to the new ones, nearly all of them take a larger hand in matters of economic regulation, of education, health, and housing, of culture and the arts, and so on almost endlessly. The increase of the activities of states is a strong and strikingly unifonn international trend. The functions of states are similar, and distinctions among them arise principally from their varied capabilities. International politics consists of like units duplicating one another's activities.
' / .I

3. The Distribution of Capabilities

The parts of a hierarchic system are related to one another in ways that are determined both by their functional differentiation and by the extent of their capabilities. The units of an anarchic system are fu nctionally undifferentiated. The units of such an order are then disti nguished primruily by tl1eir greater or lesser capabilities for performing similar tasks. This states fonnally what students of international politics have long noticed. The great powers of an era have always been marked off from others by practitioners and theorists .ilike. Students of national govern ment make such distinctions as that between parliamentary and presidential systems; govern mental systems differ in form. Students of intemational poutks make distinctions between international-political systems only according to the number of their great powers. The structure of a system changes with changes in





. the S)stem's units. And chaHgcc.. iu struct . I" . bilities <t<:Joss . ]] the cUsttibutJOn o c~pa tJ nHs of th e system w1 I>e I1mc aud ahout tlle th . . . bout ho\>V le u . . I I d' rr . ehnnge ex"}>ect<\tJOOS a . ce Domes tlcal y. t w 1. etcnLJuted parts ofe .. I . . t ctions Wl 11 proc1 . u . . outcomes t 1e1r m era . k We know from observmg tl1e .\me ncan go,e a r sinular tas s. ' . . . . rn. ,stem may penonn . leuislate and legts 1 tu 1e :)Qmt>ti mcs cxecul 5 a . h t' ves sometimes e r ... e. mclnt t at execu 1 . imes perform dif,ere nt tas k ... lmt two prohlems s JntemationaUy, like umts somet should be considered. . th" . Ca ability tells us something aboul units. Definin The first pro?lem 15 ISf. tl P distribution of capabilitie seem!-. to violat~ . artlv m tenns o 1e stnctw e P , . tt butes out of structura1 de c s. t \ s. 1 remarked nmtion t ction to keeo umt a n Tl my ~ns JlJ . . a rhi hl , but not entirely abstract concept. 1e maxJJnum ol' 5 earher, structure 1 . ? ) f content and that minimu m is what is needed to b tr tion allows a mmunum o , .1 . a s ac tJ .t stand in relation to one ano tJ1er. States are uifferenth enable one to sav how 1e wu s z 1s . ' . ' And r one mav ...vonder wI1y on Iy capab1 zty . . n1cIuded in et ) ' . cl c l>lae--ed. by thetr power.d fi 111t011 and not such c1 f tJ 1aracte nst1cs as 1 eo1 ogy, rorrn of the tlmd part o 1e e 1 ' . . . . . . c. 1 ss bellicosity or whatever. The ansv.re1 ts th1s.. Power IS estigovemment, peaceru ne ' ' . I the capabilities of a numbe r of umts. At hougI1 capa biJ' . are 1ties matecl by companng 'butes of um'ts, the distri"bution of <:apabilities across umts IS not. T 1 d" ion 1e 1stnbut attn . of capabilities i~ not a unit attribute, but rather~ system.-WJd~ concept. , ._. The second problem is tJUs: Though relations defined m t~nns of mteractions must be excluded from structural definitions, relations defined m tenns of grouping of states do seem to tell us something about how states are placed in the system. Why not speci~' how states stand in relation to one anothe r by consideling the alliances they form? Would doing so not be comparable to defining national political structures partly in terms of how presidents and prime ministers are related to other political agents? It would not be. ~ationally as internationally, structural definiti ons deal with the relation of agents and agencies in terms of tJ1e organization of realms and not in terms of the accommo<i'ltions and conflicts that may occur within them or ilie groupings that may now and then form. Parts of a govern ment may draw together or pull apart, may oppose each other or cooperate in greate r or lesser degree. These are the relations that form and dissolve within a system ratJ1er tJ1an structural alterations tJ1at mark a change from one system to another. This is made clear by the example that runs nicely parallel to the case of alliances. Distinguishing systems of political parties according to their number is common. A multiparty ~tem changes if, say, eight parties become two, but not if hvo groupings of tJ1e ~Ight fo~ merely for the occasion of fighting an election. By the same logic, an ~ntematio~polltical system in which three or more great powers have split mto two alhances remains ul ~ a m tipolar system -struc turally distinct from a bipo)ar ~-ystelm,da sys m which no third power is able to challenge the top two .... fin~em. n e mg mternational l'fca) ditions hab'ts b' . -po 1 1 structures we take states with whatever tra ~ectJves, not ask'\\helth, ostates are desires, and forms of government they may have. We do er 1 ti . ideo)noical . . revo u onary or legitimate, autJ1oritarian or democ ratic, -l:j" or pragmatic We abst t fr capabilities N h' kin rac. om every attribute of states except t betr or m t m g ab t f states their feelings of . dsh~u structur~. do we ask about the relation s 0 alliances they form and ;Jen 1 and hostility, their diplomatic exchanges, cl1 P e ' e extent of the contacts and exchanges among tJ,em- We


of CXf>N:latiollS ari'iCS JriCr<:iy from looking at the type of order thal prevail s < 111011~ tlwtn ami at the dio;tribution of capabiJitie~ \vithin that order. We ab~tnwt f: IJOJ an} particular qualities of states and from all of their concrd e connec:ti<Hl' \\ emerges i~ a positional pic:tur<:' , a general descliption of the ordered O\ erall arrange utc.:nt of a so<.:iety written in tc.:rms of the placement of 1mits rather than in terms of their qualities .... what
l' 111


[We must now] examin e the characteristics of anarchy and the expectations about outcomes associated with anarchic realms .... [This] is best accomplished

by drawin g sonte comparisons between behavior and outcomes in anarchic and hierarchic realms.
4. Violen ce at Home and Abroa d

The state among states, it is often said, conducts its affairs in the brooding shadov,, of violence. Because some states may at any time use force, all states must be prepared to do so-or live at the mercy of their militarily more \igorous neighbors. Among states, the state of nature is a state of war. This is meant not in the sense that war constantly occurs but in the sense that. with each state deciding for itself whether or not to use force, war may at any time break out. Wheth er in the family, the community, or the world at large, contact without at least occasional conflict is inconceivable; anc.l the hope that in the absence of an agent to manage or to manipulate conflicting parties the use of force \vill always be avoided cannot be realisti cally entertained. Among men as among states, anarchy, or the absence of government, is associated with the occurrence of violence. The threat of violence and the recurrent use of force are said to distinguish international from national affairs. But in the history of the world surelv most rulers . . have had to bear in minc.l that tJleir subjects might use force to resist or overthrow them. If the absence of government is associated with the threat of violence, so also is its presence. A haphazard list of national tragedies illustrates the point all too well. The most destructive wars of the hundred years following the defeat of :"Japoleon took place not among states but withil'l tJlem. Estimates of deaths in China's Taiping Hebellion, which began in 1851 and lasted 13 years, range as high as 20 million. In the Amelican Civil War some 600 thousand people lost their lives. ln more recent history, forced collect:ivation and Stalin's purges eliminated 5 million Russians, and Hitler exterminated 6 million Jews. In some Latin Amelican counhies, coups d'etats and rebellions have been normal features of national life. Behvee n 1948 and 1957, for example, 200 thousand Colombians were killed in civil sttife. In the middle 1970s most inhabitants of Idi Amin's Uganda must have felt their lives becoming nasty. brutish, and short, guite as in Thomas Hobbes's state of nature. If such cases constitute abenat ions, they are w1comfortably common ones. \ Ve easily lose sight of the f~lCt that Struggles tO achieve and maintai n power, to establish Order. and to COntrive a kind of justice within states may be bloodier than wars among them.



d . ction. and death. t 11<' u d tl clisline:tion . ........ 11.m ... " Jw is identified \\ith ch<lO . s esttrutell us much. \ \'hi c.l is I ()re pree:an. . t doe no . ben,<:>en anarchy and go,emmen or of a govern men t 111 re 1 r le lts :subjects? .tlJO . the life of a state among states. A nu ou. . . cl place. 1110 some . tate s at ~ J'lll ' t1 mes. the . Tl w !\.115\\er ,.anes ,;vith bme an Jence 1S )O'"' Within some ~tates .1t some times f ex'TlP.C'ted occurrence o V . JOf ce is hjgh. The use of f11ru. or the c:on- actual Or .., - # x'l"\f\Cted occurren ce 0 ,,o1en . . . . ds 1101. distingu1sJ1111g mUrna tional from c: t l1e actual Or e "1.'- ffi t groun . t r . r of its use. are not su eten . tual use of force ma rk hot h national and st<lll ,ea bJe lestic affairs. If the poSSJ and the ac 'nction benvee n tJl e n\.0 I' s can he bl distJ l d01 c cum tJ 0 dura e .1 . . t mational orders, 1en n f force. :\o human oruer IS pro o ragainst me . f the use or the non use o drawn m terms o . ole nce '-1 . ali . differences be...,veen internal .and cxtern<li affairs one l . To disc.'Over qu tative h 1 occurrence of vJOlence. T } 1e cl'1stmction . tle c l mus t look fo r a critenon other talan alm s of politics is not roun< . t l1e use or the . m between international and .nati.C: ret tructures. But if the dan ge rs of being 'io nonuse of force but 10 clle1r diJterent s . king an eve ninu stro 11 t Iuou g1 down town 1 0 , ter say Ill a .I JentJy attacked are gr~a .' . alonu tJ1e fre nch and Ge m1an boruer. what Det roit cllan they are 111 plcn~~~cc g f . 0 sb'llcture make? Nat ion all y as. interna. cl wHerence . . ractical difference oes . t d at times issues m ,,ol enc e. The cliiTerenc:e P t erates con6JC an tionally, contac gen . al l'ti'cs lies not jn the use o f r e but m the rorc al d 1 mation po l nte between nation an c: . od 0 f garuzation ror domg something abo ut it. A govern ment. ruldiffe rent m es or f .ti. arrogates to itself the righ t to use force-t hat . b 1 ~~o~m~ . mg )' some 5 . f cti' to control dle use of forc e bv .subjects. If some / 1ts . t0 Jy a vanety o san ons / lS, app th . al to the government. A govern me nt has no se private force, o ers may appe . u tb is all too evident. An effective gove rnm ent , h0\\ ,monopoly on e use o rorce, as '- n1nate use of forc e and te l1ere means t1 1at ever has a monopo lYon tbe wgr ' . f [i ubllc a ents are organi7..ed to prevent and to cou nte r the pnvate use o orce.. cttfzens n! d not prepare to defe nd themselves. Public a~encies do that. A nati onal system is not one of self-help. The inte rnational system 1 s.




w e

re .

5. Interdependence and Integration

The political significance of inte rde penden ce varies depe ndi ng o n wh eth er ~ realm is organized, with relations of authority spe cifie d and establis hed, or rem ams formally unorganized. Insofar as a realm is formally org anized , its uni ts are free to specialize, to pursue their own inte rests without con cer n fo r develo pin g the mea n} of maintaining their identity and preserving the ir sec u rity in the presencetlo others. They are free to speciaJi ze because they have no rea son to fea~ 1 .e increased interdependence that goes with specialization . If tho se who speCJ~ ~~ most benefit most, then competition in specialization ensues . Go ods are man ad tured, grain is produced, law and order are mai nta ine d com me rce is conduct.e ' and financial services are provided by people who ever 'mo re nar rowly special ~~ In simple e<.-onomic ten ns, the cobble r dep end s on the tailor for his pan ts and 1 . tailor on the cobble r for his shoes and each would be ill-clad witho ut the sen'lces of the other. In simple political ;en ns, Kansas dep end s on Washingto n for P7~ tect ion and regulation and \Vashington dep end s on Kansas for beef and whe at.

saying tha i ILL ~illlations inte nle pe11dc#nce is close. one nc.,e d not main tain lhat the one par t <:0 1ld not learn to live: without lhe othe r. One need onl~ sa~ that the cost of'1~ the interdependent relation would he hjgh . Persons and insti t\Jtions dep e11d laavily on one anoth er because of the differe nt tas h they perform and the different good they produce and exc hange. The parts of a polity bind themselves togctl.e r by the ir differences .; Differen ces betw<::en national and inte rnational tru<;tures are reA d in the ecte ways the units of each system define their ends and de,clop the means for reaching them . Jn anarchic realms, like units <;oact. In hierarchic realms. unlike unjts inte rad. In an ana n.:hic n:alm , the units are funetionaJly similar and tend to rem ain so. Like units work to main tain a measure of inde pendence and may even trive fo r autarchy. In a hierarchic realm , the units are differentiated, and they tend to increase the exte nt of their specialization. Differentiated units become closely interdependent, the more closely so as their specialization proceeds. Because of the differe nce of stm <;tu re, interdependen<;e within and interdependence among nations are hvo distinct concepts. So as to follow the logicians' admonition to kee p a single meaning for a give n term throughout one's discourse, I shall use "integra tion" to describe the condition within nations and "interdependence to describ e the condition among them. Although states are like units functionally, they differ vastly in their capabilities . Out of such differe nces something of a division of labor develops. The division of labor across nations, howeve r, is slight in comparison with the highly articulated divi sion of labor witllin the m. Integration draws the parts of a nation closely toge the r. Inte rdepen dence among nations leaves them loosely connected. Although the inte gration of nations is often talked about, it seldom takes plac e. Nations could mutuall y enrich themselves by further dividing not just t11e labo r that goes into the production of goods but also some of the oth er tasks they per form , such as political manage men t and military defmse. why does the ir integration not take place? The structure of inte rnational politics limits the cooperation of states in hvo ways. In a seLf-help system each of the unHs spends a portion of its effort. not in forwarding its own good, but in providing the means of protecting itse Lf against others. Specialization in a system of divided labor works to e,eryones advantage, though not equally so. Ine quality in the ex-pecte d distribution of the increas ed product wor ks strongly against extension of the division of labor inte rnational!~. When faced with the possibility of cooperating for mutual gain, states that feel insecure must ac;k how the gain will be divided. They are compelled to ask not vvill both of us gain ?" but "'Who "'rill gain more?" If an expect ed gain is to be divided , say, in the rati o of two to one, one state may use its disp roportionate gain to implem ent a pol icy inte nde d to damage or destroy the oth er. Even the prospe ct of large absolu te gains for both parties does not elicit the ir cooperation so long as each fears how the oth er wiJl use its increased capabilities. :\otice t11at the imp ediments to collabo ration may not lie in the character and the imm edi ate inte ntion of either party. Instead , the condition of insecurity- at the least, the uncertainty of each abo ut the otJ1er's futu re inte ntions and actions -wo rks against their cooperatio n ... .





I . .... - -

. ,. oss ibk crain<; that Ill,\~ o1\0 J" Others I ) divistOll () p b r Jlll l J"Jl:ltll"taI pol 1 Ore t\tL' \\"l)nies aIlOU t cl . h" \ s , itics li . ' " . JC Il tl1e . tru ctu rc 0 l11tts ' . I "t ,.]f. Th ,,ti . the firstwayu L 'O s lest it hcc.:on w t J-rie t n.n _, eh P' lcltnt on otl1 . . A tlte <\ o '' . ' ers ~"' s' . t1H' ~""" tx>ration of states. . vors an d excbanges of goods cUt< 'it'n ces. That is th . d . ' ""\l '7 h C:00l)eratlV<' t'J1 ea f . tem ationaJ pO tItiC: t tS. ltW <.:Ooperation e tl11 "' b lllll f . he st1uctu re o Ill . " , . ,..,... d ww in wlu c1 t 1 y . 0""'1 ,' aJizes t1 tnor e it reli es on ouw 1' to ~upp1 the rnate. 'le . . 1. . . t . The more a stat e spec st,'l es. . . t roduc111g. Tl1e larger a states nnp orls anc exports tl1e l . . \ .an d goods that tt ts no P Id's well-beiuo wouk1 ) t' m<.-reased if an eve na s tl Th e wor o r re it depends on o leJS. developed but states wo ulcl there bv plac:e mo .. ' " e elaborate divtston of labo r were. terdependence . Som e states may not resist mor . . . 0 fever c1 r 111 ose . . ,. . . . the mselves u1 sttuabons d t the costs of dom g :so . . cxc,css lvelv lllgh aH' ~at For small and iJl-endowe . sta ese more enmesh ed "vilh oth e rs ordin~rily cl~ , . . "t ev 1 Bu t states that can resls becommcr t s tb<1t are heavilv dep cnclent , or closely ; inter. 5 so in either or boili of two ways: at]~ t eh they depend on. Th e high iltterde. 1a . dependent, worry abou t secunng t tes in question exp enc nce. or are sub jec:t to 1 an sth att 1e sa endence of states me h" 1 . terderJende nce e nta.1 L1t(e ot1 organiza-' s. 1er P biltv that tcr 1 111 1 the common vu 1 a ' ner b tl e , depend 011 or to lessen the extent of their 1at 1 > . tions, states seek to control w1 explains quite a bit o ftJ1e be I Vlor o rs_tat . . thou ht _1a es: their dependency. This Simple th g of their control and the1r autarchtc sttiv ings . perial thrusts to widen e scope
0 0 0


l 3





toward greater self-su clency.rt . behaviors and penaJize those wh o do not 11 Structures encourage ce ~ t. nallv man)' lam ent the ext rem e developm ent tl1 encouragement. l' a to ', respon d to e . d nt that results in the allocation of ever narf th di . . 1 of Jabor a eve1 e o e VJSIOJ. . d ~ And opm ecialization pro cee ds, aml its extent et s i a mearower tasks to mdivJ u f ~ ti. ; In a fOiomally organi:;,ed realm a pre mium is f the development o . socie e . sure o . , . a able to s ecialize in order to increase its va lue to ot hers . m put on each umt s beme, P . . . . em of divided labor. The dom estic Imper_ . e JS "speciahze" l I nternationallv ativ ~. . ,, ~1: lament tlle resources states spend unproductiv_ely ~or the ir own dei~:~~ a:~ the oppo rtunities the)' miss to enhanc e the welfar e of the tr peo ple th roug ~ J an ation with other states. And yet the ways o f states<.:l1an ge l"ttle . In . unorgamzeu t - f realm each unit's incentive is to put itself in a position to be able to_ tak e ea~~~ itself since no one else can be counted on to d o so. Tlle 111 t e m ationt ltmpcrative 1s G l _ 11 "take care of yoursel f'! Some leaders of nations may un de rsta nd tha 1 t t ~ ~v~ being of all of them would increase through the ir participation in a fuller ~1 151011 " oflabor. But to act on the idea would be to act on a dom estic im per ativ e, an unper. . ative that does not run internationally. What one mtg l1t wa nt to d 0 10 the absence '. their of structural constraints is different from what one is e nco ura ged to do tn d . . presence. States do not willingly place tbe ms e1 ves m sttu ah.ons of increase dependence. In a self-help system , considerations of secun.ty su1 cli nlte ec:o)Or ' nomic gain to political interest. .

ffi .

6. Structures and Strategies

That motives and outcomes may well be disjoined sho uld now be easl1Yseen-. have
Structures cause nations to have consequences the y we re no t inten~ed to ble to Surely most of the actors will notice that, and at lea st som e of them will be a

thmk (){ othc rs . H ~h~rtage of a commodity is expected, aB are collecti,e ly bet ter off if the y buy less of 1l111 ord er to n:och:~rate ptiCE! increases and to distribute shortages equ itably. But bt'ca_usc so~ne w111 be better off if they lay in extra supplies quickl y. all have a strong m cen t1 to do so. lf one expects oth ers to make vc a run on a bank, one's prudent course _is to ru~1 fast er then they do even while knowing that if few others run, th e bank \VllJ remam solvent, and if many run. it will fail. [n such cases. pursui t of individual in~ercs t produces coll ective results that nobody wants, yel ind ividuals by behaving differently will hurt the mselves without altering outcomes. The se two much used examp les establish the main point. Some courses of action I cannot sensibly follow unless we are pre tty sur e that many others will as well ... . we may well notice tJ1at our behavior produces unwanted outcomes, but we are also Hkely to see that such instances as these are examples of what Alfred E. Kah n descdbes as "large" changes that are brought about by the accumulation of 'small"' decisions. In such situations people are victims of the "tyranny of small decisio ns.' a phrase suggesting tl1at "if one hundred consumers choose option x, and this causes the marke t to make decision X (where X equals lOOx). it is not necessarily b-ue that those same consumers would have voted for that outcome if that large decisio n had ever been presen ted for their explicit consideration."!! If the market does not present the large questi.on for decision. then individuals are doomed to making decisio ns that are sensib le within their nanow con texts even though they know all the while that in making such decisions they are bringing abo ut a result that most of them do not want. Eitl1er that or they organize to ove rcome some of tlle effects of the ma rket by changing its structure for example, by brin ging consum er uni ts roughl y up to the size of the units that are making producers' decisions. This nicely makes the poi nt: So long a~ one leaves the structure unaffected it is not possible for changes in the intentions and the actions of particular actors to produce desirable outcomes or to a\"oid undesirabl e ones .. .. The only remedies for strong structural eflects are stm ctural changes. Structural constraints cannot be vvished away, although many fail to unders tand this. In every age and place, the units of self-help systems-nations, corpor ations, or whatever -are told that the greate r goo <.l , aJong \vith their own , requires the m to act for the sake of the system and not for their own naJTowly defined advantage . In the 1950s, as fear of the world's destruction iu nuc:lear war grew. some c-onclud ed that the alte rnative to wo rld des truction was world disarmament. In the 1970s, with the rapid growth of population, poverty, and pollution, some concluded, as one political scientist put it, that "states must me et the needs of the political ecosystem in its global dimensions or court annihilation.'.g The inte rnational interest must be served ; and if that means anything at all, it means that national interests are subordinate to it. The proble ms are found at tl1e global level. Solutions to the proble ms continue to depend on national policies. \Vhat are the conditions that wo uld make nations more or less willing to obey tl1e injunctions that are so ofte n laid on them?

figu re c~ 1 t v,hv. -~i 11<~ may devdop a pre lty good sense of jmt how struclures work their dk r L~o \\ 11l tl tcy not Lltc.n be ahle to ad licve:: their original ends by app ropriately <tdJust ll~~ tlwJr strategies? U nfort uualcly. they often cannot. To show why Lhi~ is ~o I ., J_t<tll ~tve only a few exa mple!'>: once the poi nt i made. the reader will eac;il)




. b ... , e ll pursuing their 0\\11 iJ.t< ' ~o">h anclt<:t e llow l'an the, reso1 t1 e tensJOn el.''sJ,own how that can be< l01 e. tit hough <ll' ng ,e l . ? 1 one 1 No 1<lS tny for tlte s;1ke of the system. r tional behador. The ,e1y prohl<:'tn, lowever . . . ,,1 ;ng tiiCtr hands atl<d I)lead lOr ra al constraints, does not Ieh< I to t I1e want. ts d ' b 1 gven stntctu tc: . f that rational e tavwr, 1 .. ed to take care of itsel . no OI1l' <:an take caree of 1 results. \\'ith each country constrau the system. 10 . nd doom may lead to a clear defit.tition or en us that rnust A strong sense of p~nJ a . ot there b, made poss1hle. ThC' possi bilitv of . ed Th . acluevernent lS n , , be aclue, : elf the ability to provide necess~uy means. It d epends even effective action d~pends 0 ~conditions that pennit nations and othe r organi7.ations to more so on the eXJsten~e_o d trateoies. World-shaking prohlems <:1y for global r U an s o '- .. l ro ow app ropriate pol.icJeSJobal agency to provi detI1em. ,, eces 1ties to not create . solutions, but the~e 15tln 0 al ses were efficient ones does not make them so . .b.uti Wishing lat nn eau poss1 L es. lished onh- b, agents of great <:apab.Li ~- That is wlw 1 t tasks can be accomp , ' . ' Grea ; a]J tl . ones are called on to do what lS neccs ary for the . . d ecr y 1e maJor , . states , an esp h e to clo whatever thev thmk nec:essarv ror their own Id" ival But states av ' wor s s~rv . can be relied on to do it for them .\ \'h~ the' achice to place . Preservation, smce no one >bove national interests IS meanmg1 can he ex1Jlamecl . . a1 t est ess the mtemation tn er "' . . 1 . . 1 ftbe distin.ction between mtcro- and mac rot 1eones .. .. prec1se y tn terms o changes in the awareness and purp o e. th e orgam. m zaSome have hoped tllat . . . . tion and ideology of states would change the quality of 1~tem~bonal l~e. O~er the centuries states have changed in many ways , but the quality of mtem ation al hfe has remained much the same. States may seek reasonable and wort hy ends, but they cannot figure out how to reach them. The problem is ~?t in their .stupidity or ill will. altho ugh one does not want to claim that those qualities are lad.-,ng. The depth of the difficulty is not understood until one realizes that intelligen ce and goochviJJ cannot discover and act on adequate programs. Early in this cent my Winston Churchill observed that the British-German naval race promised disa ster and that Britain had no realistic choice other than to run it. States facing global problems are like individual consumers trapped by the "tyranny of small decisions." States, like consumers, can get out of the trap only by changing the struc ture of their field of activity. The message bears repeating: The only reme dy for a stron g sbl.JCtural effect is a struc:tural change.



7. The VIrtues of Anarchy

To achieve their objectives and maintain their security, units in a condition of anar chy- be they people, corporations, states, or wha teve r-mu st rely on the means they can_generat~ and the arrangements they can make for them selves. Self-help is n~~ariJy the pnnciple of action in an anarchic order. A self-help situation is one of ~ nsk-ofbankruptcy in the economic realm and of war in a world of free states. It In . IS also. one in which organizationa1 costs are low. Within an economy or w1th. an. mtemation coordin al order' risks may be av01'ded or lessened by moving from a situation of . crate. ac:tion to _one of super- and subordination, that is by erec ting agencieS with euective authonty and ext din ' the fun . en g a system of rules. Governm ent emerges w]1ere ctions of regulation and management themselves beco me distinct and

~vecializerlt;t" h . Th<: c.:<~<;ts of naintainin!; a hierarchic: order are frequent!~ ignored b: thoc;_e " ~ 1 loplore Its_ah~encc. Organizatio11s ha,e at lea.-;t two aims: to get somC"Lhmg do11t and to namtcun themselves as organizations. ~l am of their acli\ilie are dirc'ctt:cl to_warcl the secon d purpo!)e. The leaders of organi7..ations, ancl political leader!) pn:<tulne ntly, are not masters of the matters their organizations deal with. They .ltav~ lwcotoc l~ad~ rs not by being experts on one th ing or another hu t by excelhng m tl~( orgam7ational e:t rts-i n maintaining <:ontrol of a group's memhcrs. in .. elicitin~ predt~t-ahle _satisfacto?' efforts from them . in holding group together. In maki ng pohttcal clet:JSIOns, the first and most important coneem is not to achieve the aims the ncmbers of an organization ma, have but to secure the continuitv and health of the organization itself. 1I ' Along with the adva ntages of hierarchic orders go the costs. In hierarchic orde rs. moreover, the means of control become an object of struggle. Substanti,e issues becon1e enhvined with efforts to influence or control the controllers. The hierarchi c orcle ling of polmcs adds one to the already num erous objects of struggle, and the object added is at a new order of magnitude. If the risks of war are unbearably high, can they be reduced by organizing to manage the affai rs of nations? At a minimum, management requires controlling tl1e miJHary forces that are at the disposal of states. Within nations. organizations have to work to maintain themselves. As organizations, nations, in working to maintain themselves, sometimes have to use force against dissident elements and areas. As hierarchical systems, governments nationally or globally are disrupted b~ the defection of major parts. In a society of states with little cohe rence, attempts at world government would founder on the inabiJity of an emerging central authority to mobilize the resources needed to create and maintain the unih of the svstem bv ' . regulating and managing its parts . The prospect of world gove'mment would be an invitation to prepare for world civil war.... States cannot entrust managerial powers to a central agency u11less that agenc~ is able to protect its client states. The more powerful the <.:lients and the more the power of each of them appears as a threat to the others, the greater the power lodged in tl1e center must be. The greater the power of the center, the stronger the incen tive for states to engage in a struggle to control it. States, like people, are insecure in proportion to the extent of their freedom. If freedom is wanted. insecurity must be accepted. Organizations that establish relations of authority and control may increase insecurity as they decrease freedom. If might does not make right, whether among people or states, then some institution or agency has intervene d to lift tl1em out of nature's realm. The more influential the agency, the stronger the desire to control it becomes. In contrast, unlts in an anarchic order act for their own sakes and not for the sake of prese an organization and fmthering their fortunes within it. Force is used for one's own interest. In tl1e absence of organization, people or !>tates are free to lea,e one another alone. Even when they do not do so, they are better able, in the absence of the politics of tl1e organization, to concentrate on the politics of the problem and to aim for a mininmm agre ement that will permit their separate existence rather than a ma.ximum agreement for the sake of maintaining unity. If might decides. then bloody struggles over right can more easily be avoided.

a u?




f-1 !tO\ em111l _J. [ C.:<;:. I. :\:1timdlv. the orce o < ~ lo,ed for the S<LKe o 1t!'. own protection .1 d r f. state 1S em P n ...mllionuk the ton.:e o a ents claim to au tl10nty: t1 ( JUestion th W\ l nl" ' . t -'1 r a govern 111 I e . d ''nt<wr. H cbds CL1a.L enhe t tes cannot settle guest1ons o allthorit)' "nd 1 ' ~ ~ \~ " 'rthtfulncss of its rule. l;u-s among 11s a. tion of gains and losses among contendc _ nEd . . .aht tbev C<U1 onlv etermine the <wocaho is the stTonger. N attorl.t!I\: relations rs. f nr . . . .. tl est]on o w I . o1 and settle for a ttme le qu . allv. onlv relations of strengl1 rc.snll. :\ ationallv authoritv are established. lntemation t. clu~tens the political ~:'Stem. Force used b: ed a 1St a aovemmen . . Y rivate force us aoau t> tl ntemational perspective. t 1 pnvate use of 1e P t I' body-is from le 1 - ' a state a puo JC t overthrow and no govemm enla~ apparatus to r force: but "' u1ere . no.govemment 0Id heaemony. the plivate use o f" orce does not rs o . .. . Sl101t of a drive tow-arc1wor litics onlv some of 1t s me m I 1s. \\'ar prts some capture. . 1al 1e threaten the system ofintematio; l:ong ~imtlarly constituted enlilie . The power states againSt others in alstruggak~ rrom assertina their claims. not because the weak t le 1 o . I f the strong may detertful we f 1 le on the pa1t of the strong, 1 o 1ut unp v because . a kind of righ ness o m ak . . .d ~gnv ..e . tan le ,.~,~tJ 1 tJ1em. Conversely, the we may Cll.JO}' cons1 erable tt L~ not sensJb~e t~f th g. f:tr removed in their capabilities from lhe strong that freedom of actiOn I eyboarethsored bY tlleir actions or much concemed by marginal h the latter are not muc e . ' increases in tl1eir capabilities. . . . . strat10n, and of law. lnter:\ational politics IS the realm of authority, of admm1 f .. cl . . national politres the realm of power, of struggle, and o . accommo .ahon. The rs , I a] . intemational realm is preeminently a political one. f he nationa re ~~ IS vanously described as being hierarchic, vertical, centralize d, l~etero~e neous, dJrected,. and contrived tile inten1ational realm, as being anarcluc. honzontaJ. decentrahzed, homogen~us, undirected , and mutually adaptive. The more ~en tralized ~1-e order, tile nearer to tile top tl1e locus of decisions ascends. I nternatJonally, dects1ons are made at tile bottom level. there being scarcely any otJ1er. In the vertical-horizontal dichotomy, international structures assume the prone position. Adjustm ents are made internationally, but they are made without a formaJ or authoritati e adjuster. Adjustment and accommodation proceed by mutual adaptation. 12 Action and reaction, and reaction to the reaction, proceed by a piecemeal process. The parties feel each oilier out, so to speak, and define a situation simuJtaneously v.rith its development. Among coordinate units, adjustment is achieved and accommodations arrived at by the exchange of "<.:onsiderations," in a condition, as Cheste r Barnard put it, which the duty of command and the desire to obey are essentially 1 3 :;ent. \~ere ~e contest is over ~nsidera tions, the parties seek to maintain or anp~e tll.e1r J>OSltions by m.~eu~enng, by bargainjng, or by fighti ng. The manner d .mtenSJty of the competition ts determined by the desires and the abilibes of parties that are at once separate and interacting. ,__. Whethlfe! or ~t by force, each state plots the course it thinks will best serve its uuerests. 10rce IS used by t states . to r one s ate or its use is expected the recourse of other JS use 10rce or be prepared t0 . . I ' can be made hi . use Jt smg y or in <.:o mbinatio n. No appea1 the ability t0 to a .gher entity clothed with the autl1ority and equipped wjth a<.:t foree will be usedon its own initiative. under such conditions the possibility t11at b 0 background. In po~~ or ~notl_ler of the parties looms always as a threat in the cs orce IS said to be the ultima ratio. In inte rnational politiCS

.11t 15. exercised in tlte name of n~ltt < justi . . md



forc.:e s tJVt ' t ()u)~ as the ultiuw ratio, hut indeed as tht first and c.:omtanl cmc. To li ntit .~<'~l' ,t~.~~t'lll~ ~he u~ti1:w of politic:s implies, it1 the words of Ortega y Cassct, lit< l H ."ous submiSSion of force to metl10ds of rea~on." 1 1 The constant possibility tlwt lor~C' wi ll he used limits manipulations, moderates demands. and serves as an mc:enttvc for the settlement of disputes. One who knows that pressing too hard m~) ~(:'ad to. war has strong reason to consider whether possible gains are worth .the ns~s e.ll<uled. The threat of force internationally is comparable to the rule of the stnke ll1 l~~hor a~1d management bargaining. "The few strikes that take place are n a sense, .as L1vernash has s~id, "the c.:ost of the strike option whi<;h produc.:es settlcme11Ls m the large mass of negotiations.'' 15 Even if workers seldom strik~. their doing so is alwa~s a possibili ty. The possibility of industrial disputes to long and costly stnkes encourages labor and management to face difficult tssues. to try to understand each other's problems, and to work hard to find a<;commodation~. ~he possibility that conflicts among nations may lead to long and costly wars has s1 mrl arly sobering effects.

8. Anarch y and Hierarchy

I ha,e described anarchies and hierarchies as tlloucth everv political order were of 0 ' one type or the other. Many, and I suppose most, political scientists who write of struehtres allow for a greater, and sometimes for a bewildering, variety of types. Anarchy is seen as one end of a continuum whose other end is marked bv the , presence of a legitimate and competen t government. International politics is then descJibed as being flecked with particles of government and alloyed ,,~th elements of community-supranational organizat ions whether universal or regional. alliances. multinational corporati ons, networks of trade, and whatnot. International-political systems are thought of as being more or less anarchic. Those who view the world as a modified anarcbv do so it seems for two ' ' reasons. First, anarchy is taken to mean not just the absence 'of government but also the presence of disorder and chaos. Since world politics. although not reliably peaceful, falls short of unrelieved chaos, students are inclined to see a lessening of anarchy in each outbreak of peace. Since world politics, although not formally organized, is not entirely without institutions and orderly procedures, students are inclined to see a lessening of anarchy when alliances form, when transactions across national borders increase, and when internatio nal agencies multiply. Such views confuse structure with process, and I have drawn attention to tJ1at error often enough. Second, the two simple categories of anarchy and hierarchy do not seem to accommodate tJ1e infinite social variety our senses record. Why insist on reducing the types of structure to nvo instead of allowing for a greater variety? Anarchies are ordered by the juxtaposition of similar units, but those similar units are not identical. Some specialization by function develops among them. Hierarchies are ordered by the social division of labor among units specializing in di.fTerent tasks, but the resemblance of units does not vanish. Much duplication of effort continues. All sociE-ties are organized segmentally or hierarchically in greater or lesser degree. Why not then, define additional social types according to the m~ture of organizing principles





4 . ;!I! the Ptrely ive o f son1e societies apprna' . ,.erarchi<..:. and . of st:t. qtllt:>rs1.re0c:cti g conce l_"" .J,.'l One might . the }Jure1, 11 l ' 11 the, t'lll XJU. 1 . . IHC. 0 f others 30J)I"Oi.1C ung 1oonai 'hn,es Jn anarchtes t 1e1 t 1 tc:l Jkeness 0 f 1 : r t tUU\J'C' , ,(!.ed nuxes of the two org<uliZ<bv captbi]jtv ctlone won <. < <>-.en be a re' l t., .. ,.. . . ' ' f I tions . spt:'Cul , ftl nteradion of LlllJts guh.lPd hy admini. 1 the deten11inat10n o re a . r, I , . , ciJ . . . s . 'tll none o Je J umts .mo 1 c1 entiatJon or :politi'cs and powet Wl h 'hi In 1...... ,..c!Jies the comp et<: lJet<u w h0 ll\'0 l . and conditioned by aut on.r. 1uncti'ons would prod uce. a. .realm wholly of . .. . e1r f t11 ~.: . , . tnttion . .~ I . nteraction of patts ..tOected by pohti<.:s . . p<u '~ an<J the full specification1tl1 none oftl 1e 1 ~ h ure , autl10 n'tv and administration "' orders do 110t exist' to distingLUs 1 realms b" their rand important. and power. Although sue P 0 .,,...;.,..;nu principles iS ne\ertheJess pr. pe'"'OuJd bJinu the clas ification of SOcieties . . o b 0 f categones org<U...... o Increasing the num er ud be to mo'e away from a theory claunmg explanatorv . . . . tlJat wot One who .51ng greater descnpti,e accuracy. Closer to reality. But . . h d . . aJ system proJm. tl1eOJetic Power to a less . ti1er tl1an to descn'be should resist. movmg m t at u cction if . . . ? Wl t does one gain by ms1stmg on two types when wishes to expam ra resistance is reasonable. Is Idt. 'lllba t 0 smpuh boldly? One gain~c; clarity and e<.:on1 'J [!0ur woul sb e 1 or h uld be introduced only to <.:over matters tl1at admitting t 1ree f concepts A new concept s o societies are neither,anarcIu.c or 1. 11erarchic:' omy .o . 1. d0 not reac} If some . fined b some third ordering pti nc:Iple. then we would ha,e ~nsti~g concepts. Y. if thetr structures are de16 All societi'es are mixed. Elements in them represent both third system d6 . . 1 5 That does not mean that some societies are ordered to e ne a . of the .ordenn! ?dnnc~p ~ Us ally one can easily identify the principle by which 1. l . J. I . according to a unr pnnctp e. u es . 1 . . . . sectors w1t 1m uerarc 11 rc oes not a soc1ety IS 01dered. Th e appearance of anarchic . 101 mger . d shouId not 0 bscu re the ordering l)nnc1ple of the 1 b h system, th t 1ose . r e ltllJts r alter an . sectors are anarchic onIy ' VJ' tl1111 limits The attributes and e av10r o . . . . . u populallng th ose sect ors \!VJtlu'n ..t.e larger system differ moreove r, from .what ther should be and how they would behave outside of 1t. F1rms m o!Jgopolrstic markets again are perfect examples of tJ1is. They struggle against one anothe r, but be~m~ e they need not prepare to defend tllemselves physical ly. they c.:an afford to spectal~ze and to participate more fully in ilie division of economic labor than states c~n. :\or do the states that populate an anarchic world find it impossible to work w1t~1 ~ne another, to make agreements limiting their arms, and to coope rate in estabhsh1~g organizations. Hierarchic elements within intemational structures limit and restram the exercise of sovereignty but only in ways strongly <.'Onditioned by the anarch>' of the larger system. The anarchy of that order strongly affects the like)jh~d of cooperation, the extent of arms agreements, and the jurisdiction of intematiOnal organizations....


4 . .\ fca rnt \ hrti'' ' <tU Tlw Pn~itiu Pl~tlo.~"l'h!J of lw~u\te C11111U: Freely Translated am/ Cmtdf'll'lrl. :3rd l'd. I London Kcgan Paul, Trench., 1 !1~3 1 \'ol. l pp. 3 1- '5:J (;rorg1 Katnua... Rational BPlaa"iur and Ec:CJnomic Heha\iC>r." Psyclwlo!!.ical Ret:.iC'u 60 (S('pttm kr HJ.'53,.

Prcss. HX>~ ), p. 207 . 6. S111itl 1slrould know better. Tnmslatcc.l in to ~erms that he hru. IJimsclf so t:fTecti\ely used. tcJ ~ay that l-tates are so,weign is to say that they are sf'gmt:;nb of a plural sO<:iety. S(:>e Ilis "A Strm:tural ;\ pproad, to Cornparathe Politics'' in David Ea<;ton , E'd.. Vaneties of Politic:, Theori~~ ( uglewood Cliffs, 1\.j .: Prcntice HaiL 1966), p. 122: cf. his "On Segmental) I ,ineage S~ stems." jounwl of the Royal Antlmrpolo{!.ical Society of Creat Britain a11d Ireland% (July- Dece mber J956). ork: 7. Emile D urkhcim. T he Division of Labor in Society, trans. George Simpson (. ew Y Free Press. 1964), p. 212. 8. Alfred E. Kahn , ''The Tyranny of Small Decision: Market Failure, Imperfections and Lit nits of Econometrics," in Brucc M. Hussett, ed., Economic Theories of International R.elatiuns (Chicago, Ill.: Markham , 1966), p. 23. 9. Richard W Ste rling, Macropolitics: International Re/.(Jtions in a Global Societ y ork: Knopf. 1974), p. 336. (;\e\\' Y 10. Put differenlly, states face a "prisone rs dilemma." If each of two partjes foll ows his 0 \\11 interest, holh end up worse on than if each acted to achieve joint interests. For thorough examination of the logic of such situations, see Glenn H. Snyder and Paul Diesing, Conflict among Nations (P1inceton , '.J.: Princeton Unive rsity Press. 1977): for brie f and suggestive iJ1ternational applications, see Robert Jervis, "Cooperation under the SecUJity Dile mma,'' World Politics 30 (January 1978). 11. Cf. Paul Diesing. Reason in Society (Urbana, Ill.: University of Lllinois Press, 1962). pp. 198-204: Anthony Downs, Inside Bureaucracy (Boston: Little, BrO\vll. 1967). pp. 262-70. 12. Cf. Chester l. Barnard, "On Planning for World Government.'' in Chester I. Barnard, ed., Organi::.ation and Management (Cambridge, Mass.: Hruvard University Press. 1948), pp. 148-.52: Michael Polanyi. ''The Growth of Thought in Society." Economica 8 (Nove mber 1941), pp. 428-56. 13. Bamard, "On Planning,'' pp. 150-51. 14. Quoted in ChaJmers A. Joh nson, ]~evolutionary Change (Boston: Little, BrO\'vll_ 1966). p. 13. 15. E. R. Livemash, "The Relation of Power to the Structure and Process of Collective Bargaining,.. in Bruce M. Russett, ed., Economic Theories of International Politics (Chicago, Ill.: Markham, 1963), p. 430. 16. Emile D urkheim's depiction of solidary and mechanical societies stiJI pro,ides the best explication of the two ordering principles, and his logic in limiting the types of society to two continues to be compelling despite the e fforts of his many critics to o,erthrow it (see esp. Tlte Division ofLabor in Society ).

ale I 5. Charlcl\ 1~ 1\:indlcbc-:: rgcr. American 13uc;iness Abroad ( ~ cw . Iaven, Ct.: Y}


NadeJ. i ~~'pp. ln~TheoryofSocial Structure (Glencoe, Ill.: Free Press, 1957), PP 8-ll.
3 William ed.

Fo~f -~OX. 'ihe Uses of International Relations Theory" in William T. R.'t\' 0 , A...,.,-., of
-~ p.International Relations (t\otre Dame, Incl.: UniverSJ ' 1959), 35.


otre Dame Press,

Anarchy and the Strugglefor Po wer

"'te Great powerS, I ar5' are alwa)IS . searching for oppo1tunities to gain power over . , . their rivals, with hegemony as theu final goal. Thts pers pec~v<.> does not allow for statu s quo powers, except for th~ unusual state ~c~u e_ve~ ?r~po~deran <:e. Instead, the system is populated wJth great powers that h~ve Je~s JonJ st Intentions at their core. This chapter presents a theory th at e.Kplams th1s competition for power. Specifically, 1 attempt to sh?w. that ~ere is a compelling logic behi ntl my claim that grea t powers seek to max1m1ze thetr share of world pm.ver.. . .




My explanation for why great powers vie with each othe r for powe r and strive for hegemony is derived from five assumptions about the inte rnational system. ~on e of these assumptions alone mandates that state s beh ave competitively. Taken together, however, they depict a world in which states have conside rable reason to think and sometimes behave aggressively. ln particular, the system encourages states to look for opportunities to maximize their power vis-a-vis othe r states .... The first assumption is that the international system is ana rchic, which does not mean that it is chaotic or riven by disorder. It is easy to draw that conclusion, since realism depicts a world characterized by security competition and war. B y itself, however, the realist notion of anarchy has nothing to do with conflict; it is an ordering principle, which says that the system comprises inde pen den t states that have no central authority above them. Sovereignty, in othe r words, inhe res in states ~ecause there is no higher ruling body in the international system. The re is no government over governments." . . The secon.~ assu~ptio~ is that great powers inherently possess some offe nsive military capability, which giVes them the wherewithal to hur t and possibly destroy each other. States are potentially dangerou s to each other, although some states have more mJnt~ 1 ht th an .. mg others and are therefore more dangerous. A state's military power_ usually identified with the particular weaponry at its disposal. IS althoug~ even If there were no weapons, the individuals in those stat es could still use theJr feet and hands to attack the population of another state. After all, for every neck, there are two hands to choke it. tions~e ~?assu mption is that states can never be certain abo ut othe r states' int~n . pec.:ifically, no state can be sure that another state will not use its offensive

their offensive capabilities. The fom th assumpt ion is that sUJv ival is the p1i mary goal of great powe rs. Speci Ficallv, states !>eek to main tain their territorial integritv and the autonomv of , , their dom estic political orde r. Survival dominates othe r moti ves beca nsE:, once: a state is conquered, it is unlikely to be in a position to pursue othe r ai ms... . States can and do pursue othe r goals, of course, but security is their most important objective . The fi fth assu mption is that great powers are rational actors. They are aware of their exte rna] environm ent and they think strategically about how to survive in it. In parti cular, they consider the preferences of othe r states and how thei r own behavior is likely to affect the behavior of those othe r states, and bow the behavior of those othe r states is likely to affect their own strategy for survival. Moreove r, states pay attention to the long term as well as the imm ediate consequences of their actions. As emphasized, none of thes e assumptions alone dictates that great powers as a general rule should behave aggressively toward each othe r. There is sure ly the possibility that som e state might have hostile inte ntion s, but the only assumption dealing with a specific motive that is common to all states says that their prin cipal objective is to survive, which by itsel f is a rather harmless goal. l\' evertheless, when the five ass11mptions are married together, they create powerful ince ntives for great powers to think and act offe nsively with regard to each othe r. In partic:lLiar, three general patterns of behavior result: fear, self- help , and power maximization.

divine with 100 p<. rc:cnt <:C"'rtaiuty. Tlwr(; arc nrany po!., iblc causcs of aggr<-~!.ion, and no ~ late cr.u1 be sure t l1 another slate ic; not motivated by one of the m. F urthe:rmo al rc' , intentions can change ynic:kly, so a state'!> intentions can be benign one day and hostile the next. Unccrtainty about inte ntions is unavoidahl<:: , which means that slalcc; can never be sure that othe r stat<.:s do not ltave offe nsive inte ntions to go alon~ \\itlt

rn ilitar~ cap;tbilil\ to attack tit<.. fir<it stale.:. This i'i not to say that slates rrcctssar il} have lto~till intentions. IHdcetl, all of the slates in tlw syste;nl rnay be re:li:~hl: henij4n but it i~ impc,.,!.ibk to be 5u rc of U1at judgmC:'n t bc,cause inle ntiow. arc impossiiJie to

Great powers fear each othe r. They regard each othe r with suspicion , and they wony that war might be in the offing. They anticipate danger. There is little room for trust amo ng states. For sure, the leve l. of fear varies across time and space, but it cannot be reduced to a trivial level. From the perspective of any one great power. all other great powers are pote ntial enemies. This point is illustrated by the reaction of the United Kingdom and France to German reunilication at the end of the Cold War. Despite the fact that these three states had been close allies for almost fortyfive years, both the United Kingdom and France immediately began worrying about the potential dan gers of a united Gennany. The basis of this fear is that in a world where great powers haYe the capabWty to attack each other and might have the motive to do so. any tate bent on survival must be at least suspicious of other states and relu ctan t to trust them. Add to this the "911 -

~1 by John J. Mearsheimer~se:~r PolJ~U:s_ by John Mearsheimer, pp. 29-40, 46--53. Copyright 0

Y permiSSion ofW W Norton & Company, Ine.

EXL-erpts from Tragedy ofG





the Con~~ently, states pay close attention to how power is distr ibuted among m, an t ey make a special effort to maximize their share of world power.

war aggressor. Bec-ause 1t IS sometHnt tes and to be prepared for . mtl1 ~ .\.m. th ie reason not to trust o er sa f f: 11' g victim to aggressiOn furLlH amplify the amP The possible consequences 0 . a 10 .' force in world po1. . . c ~ powe rs do Jt1Cs. '' edt . portance of fear as a motv~ftJ~g ly un h as 1 111 temahonal politics were mere. an cconomi<: 1 not compete witl~ ~~c, ot er tition among states is a much more dange rous busimarketplace. Poht1 C~ 111 ?e cal . ,. the form er can lead to .. ,. an d war often war. JC mtercourse ness than mere econom fi ld ell as ma<>s murder o r CJVJ an . I n extreme .. w means mass kiJlmg on the battle e ast'on of states. The horn'b) e <:ons. ences of d E'qu I d to the estruc I , . cases war can even ea . each other not just as <:ompet1tors, but as , . to VIew . d b . war sometimes cause states p )jti al anta<Tonism, m . . short, ten to e mtense, potentially deadly enemies. o c o because the stakes are gr~t. . t 1 also aim to guarantee their own survival th 10 temationa1 svs en . States m e at' tlueats and because there is no h1gh er authorihr tl states are pote nt:J ' 'J Because o .Jer cue wl th ''al 911 states cannot depe nd on othe rs for h 1en ey w , to ~ome to t e~r r~ach state tends to see itself as vulnerable and alon e, and therether own secunty. d r 'ts own survival. In intemational politics, Cod helps fore it aims to proVJ e 10r 1 . This emphasrs on self-he1 does not preclu de states p those who heIp t lJemseIveS. . r . all' . s But alliances are onl)' temporal) ' marn.ages o f converuence: from Iormmg tance. . d , . todays aluance partner might be tomorrow's en~m y, and to ay s en~my m~g ht be tomorrow's alliance partner. For example, the Umted States fought With Chma and tile Soviet Union against Germany and Japan in \1\Torld V II, but soon therea~ter Var flip-Hopped enem ies and pa1tners and allied ""ith West Germ any and Japan agrunst China and the Soviet Union during the Cold War. States operating in a self-help worl d almost always act according to their own self-interest and do not subo rdinate their interests to the interests of othe r states, or to the interests of the so-called international comm unity. The reason is simple: it pays to be selfish in a self-help world. This is true in the shor t term as well as in the long term , because if a state loses in the short run, it migh t not be arou nd for the long haul. Apprehensive about the ultimate inten tions of othe r states, and aware tl1at they operate in a self-help system, states quickly Wlde rstan d that the best way to ensure their survi val is to be the most powerful state in the syste m. The stronger a state ~ relative to its pote ntial rivals, the less likely it is that any of thos e rivals will attack 1 and threaten its survival. Weaker states will be reluc tant to pick fights "'~th t more powerf~J states because the weaker states are likely to suffer military defeat. Indeed, the btgger t11e gap in power between any two states the less likely it is t11at the weaker will attack the stronger. Neither Canada nor' Mexico, for example, ~oul~ coun tenance attacking the United States, which is far more pow erful than 1ts ne1ghbors The ideal s'tuati 1 I on IS to be the 1 mon in the syste m .... Survwa 1ege would then be almost guarantee d.

an turn fo. 1 . traJ autI1011-1 tl , .,m- the absence o( a ccn . tJ\te to fear eac1 o ler. v .. 'V(r, there is 1 ll'" pro > help -and states have even greater 'ncen nterest ol tl1Lfd partic," . . " I Il ls 1 an bJe self-i ling no mechanism, oth~r _tl1an th~ p':s~fficu lt to deter pote ntial ag,_grc~!i.r , <;tates have

n. to which a threa tened c;t ~



varie ty of liiC."Hll\ <:<:onomie, diplomatic, and milita1 v-to shift th<.; halanc:c of power in thr-ir f:t\(>J f'>~l'~l if doing so makes othe r stale~ !>uspicious or even hostile. Because one stat '.' s ~c.u n Ill powe r is anot her states loss, gre::at powers tcnd to have a zero-sum meutaht} whcn ucaling witII eac:h other. The tric:k, or course, is to he the winner in this <:onpPti tion anc.l lo dominate the othe r states in the S}Ste m. Thus , the claim tbat slal~s ma~indze relative power is tantamount tc, arguing that states are disposed to thmk offensively toward othe r states, even though their ultimate motive is simply Lo survi vr. In short, great powers have aggressive intentions. Even when <~ gre~t powe r ach ieves a distinct military advantage over its rivals, it continues lookmg for chanc:es to gain more powe r. The pursu it of powe r stops only when hege mony is achieved. The idea that a great power might feel secu re withou t dominating th e system, provided it has an "app ropriate amount" of power, is not persuasive, for two reasons. First, it is difficult to assess how muc h relative powe r one state must have over its rivals before it is secure. Is twice as much power an appropr~at e threshold? Or is three times as much powe r the magic number? The root of the problem is that power calculations alone do not dete rmin e which side wins a war. Clever strategies, for example, some times allov., less powerful states to defe at more powerful foes. Second, dete rmining how much powe r is enough becomes even more complicated when great powe rs contemplate how powe r wiJJ be distribute<.l among them ten or twenty years down the road. The capabilities of individual states vary over time, some times markedly, and it is often difficult to predict the direction and scope of change in the balance of povver. Remember, few in the \Vest anticipate d the collapse of the Soviet Union before it happened. In fact, during the first half of the Cold \ Var, many in the West feared that the Soviet economy would eventually generate greater wealth than the Ame rican economy, which would cause a marked power sh ift against the United States and its allies. \Vhat the futur e holds for China and Russia and what the balance of power ,;o,lil] look like in 2020 is difficult to foresee. Give n the difficulty of <.lete nnining how muc h power is enough for today and tomo rrow, grea t powers recognize that the best way to ensu re their secu rity is to achieve hegemony now, thus eliminating any possibility of a challenge by another great power. Only a misguided state would pass up an opp01tunity to be the hegemon in the system because it thou ght it already had sufficien t powe r to survive. But even if a great power does not have the wherewithal to achieve hegemony (and that is usually the case), it will still act offensively to amass as much power as it can, because states are almost ahvays bette r off with more rathe r than less power. In short, states do not become status quo powe rs until they complete ly dom inate the system. All states are influen<:ed by this logic, which means that not only do they look for opportunities to take advantage of one another, they also ,.,ork to ensu re that othe r states do not take advantage of them. After all, li ,al states are driven by the same logic, and mos t states < e likely to recognize their own motives at play in the W a~tions of othe r states. In short, states ultimately pay atten tion to defense as well as offense. T hey tltin k abou t conquest themselves, and they work to check aggressor

Spec i ficall~ tlw: lflok fi >r opportu11 ities to alt(r the balan ce of power h~ acqu iring additional lll('rt 11 tmts ()I power at ll1c expe11sc of p<>l(ntial rivals . States employ a

llll' ur t .tds tu a world of (: n t,\1\t ~ '1.'Uti~ competition. wher<' sl~ttes art. WJ~Im g ,-. at, and use br t\.m'(' tf 11 heiJ l~ thl'lll ,,ain ctdvantagc over tleu mals 1\. Cf' i1 011e de fine !Jle t ~ s lat l't-1lh.'\'pt as,\ st:ttl' of tmnq uili ty or uwtu aI coneorcl . IS not IJ k' 1: to break out in tlhis


~,unill!! po,w r

a t their expense.



" wld .... 1t should be apparent from this discussion that sayiu~ tlmt slates are puw nHL mizer s i . t<ltamount to aying th_at _th~y care ahuut rc.>lative pow<' r, n:: \i U absolute powe r. There i an important distinction he1 bccanse states concerned :e. about rclati,e powe r behave differently than do states _nl (..r~~stec~ 111 absolute power State that maxjmize relati ve powe r are concerned pnma nl)' w1th the distrib uuo~ of material capabiBties. In partic ular, they by to_ gam a large a power advantage as po ible 0 ,er potential rivals, be_cause po,~er ~~ the best_ means to survival in a dc mgerous world . Thus, states mot1vat~d by 1ela~ve 1~owc ~ conce rns are likely to for<JO large gains in their own powe r, 1f such gams glVc nval states even greate po,~e r. for small er national gains that n~ve_rtheless provide them with a powe~ ac h-antaae 0 ,-er tJ1eir rivals. States that ma.xtmJ absolute powe r, on the other hand, ze care onlv about the size of their own gains , not those of other stales . They are not motivat~d by balan ce-of-power logic but instead are conc erned 'vid1 amas ing powe r without regard to how much po~e r otl_1er sta~es control. ~hey \\'ould jump at the oppo1 tunity for large gains, even 1f a n val game d more m the deal. Power, according to this logic, is not a means to an end (survival). but an end in itself.

There is obviously little room for status quo powers in a worl d where states are inclined to look for opportunities to gain more powe r. Neve rth eless. great powers cannot always act on their offensive intentions, becau se behavior is in fl uenced not only by what states want, but also by tl1eir capacity to reali ze these desires. ,ery state might want to be king of the hill, but not eve1 state hac; the whe rev.~th al y to compete for d1at lofty position, much less achieve it. Much depe nd on how military might is distributed among tl1e great powers. A great powe r that has a marked power advantage over its rivals is likely to behave more aggrE-ssively, because it has the capability as well as the incentive to do so. By contrast, great powe rs facing powe rful opponents '"'ill be less inclined to consider offensive action and more conce rned with defen ding the existing balance of power from threats by their more powerful opponents . Let the re be an opportunity for those weaker states to revise the balance in their own favo r, however, and they will take advantage of it. In short, great powers are not mindless aggressors so bent on gainin g power that they charge headlong into losing wars or pursu e Pyrrhic victories. On the c..'Ontrary. before great powers take offensive actions, they think carefully about ~e balanc:e of power and about how other states will react to their move s. They weJgh the costs and risks of offense against the likely benefits. If the bene fits do not outweigh the risks, they sit tight and wait for a more propi tious moment. Nor do states start arms races that are unlikely to improve their overall position. States

sometiJO<' lm.:t tit fc l''i< ' 5pcnd111~ (;itlwr lwcau sc.: '-P<nclinl!; 11H1n \\OnlciiJritag tac strategic- acl < nt;1 or h<~C<ll i V.: spencH11g rnon" wq11ld weak<'n tht:- C'<:CJ110III~ aud g< undem lini llw ~tat t ' powe r in the long n111. To paraphrase: Clint Ea<stwood. a '>t<ttl has to know 1ls lirllll tlions to survive in the i11 ternal ional svst<:tn. 1 evcJihr~less, gn-at power~ misc<tkt tlat<:; frotn tim~~ to lin1c lwca:-.t the\ invarjahly 1nake intpnrtant uel:isions on tlte: basis of' intperf(ct inforr nation . Stat~' hardly enr l1avc w111plete information about any situation the;y confront. Then are two dimensions to this probl em. Potentia.! adversaries have incentives to misrepresent thei r own stre ngth or weakness, and to conceal their true:: aims. For example. a "veaker state t1 ing to deter a stronger state is likely to exaggerate its y own pow<'r to discourage the potential aggressor from attacking. On the other hand, a slate bent on aggression is likely to emph asize its peaceful goals while exaggerating its militaty weakness, so that the poten ti al victim does not build up its own arms and thus leaves itself vulnerable to attack. Probably no national leader was better at practicing th is kind of deception than Adolf .Hitler. But even if disinformation was not a problem, great powe rs arc often unsure about how their own military fo rces, as welJ as tbe adversary's, \.rill perfonn on the battlefield. For example. it is sometimes difficu lt to determine in advance how new weapons an d untested comb at units will perfo rm in the face of enemy fire. Peacetime maneuvers and war games are helpful but imperfect indica tors of what is likely to happen in actual combat. Fighting wars is a complicated business in which it is often difficult to predict outcomes .... Great powers are also some times unsure about the resolve of opposing states as well as allies . For example, Germany believed that if it went to war aaainst France and Russia in the su mme r of 1914, the United Kingdom would probably stay out of the fight. Saddam Hussein expected the United States to stand aside when he invad ed Kuwait in August 1990. Both aggressors guessed wrong, but each had good reaso n to think that its initial judgment was con-ect. In the 1930s , Adolf Hitler believed that his great-powe r rivals would be easy to e>..'ploit and isolate because each had little interest in fightin g Germ any and instead was determined to get someone else to assume that burde n. He guessed light. In short great powe rs constantly find themselves confronting situations in which they have to make important decisions with incomplete informaUon. Not su1 prisingly, they some times make fauJ ty judgments and end up doing themselves serious harm . Some defensive realists go so far as to suggest that the constraints of the internatio nal system are so powe rful that offense rarely succeeds, and d1at aggressive great powe rs invariably end up being punished . .. . They emphasize that 1) threatened states balan ce against aggressors and ultimately crush them. and 2) there is an offense-de fense balance that is usuall y heavil y tilted toward the defense. thus making conquest especially difficult. Great powe rs, there fore, should be conte nt with tl1e existing balance of powe r and not try to change it by force . ... There is no question d1at systemic factors constrain a&:, uression, especially balancing by threatened states. But de fensive realists exaggerate tho e restraining forces. Indee d, the historical record provides little support for d1eir claim that offense rarely succeeds. One study estim ates that there were 63 wars betwe en 1815 and 1980, and the initiator won 39 times. which translates into about a 60 pen.e nt


4RT t Pn



how~. t k!t oil use someti . 1powt>r rnax.irniz rnes 1 L . _ , 11 t'\.'t't w am1snmct11nes cloes not The trick for a sop mtlcal<'< -er L~ to fi,nrrt' out whl'n to rai e and when to fold. ::-

I rt the historical record r.tll'. . . . l n s ro .

-,., I

C. at powers as 1 have emphasized, strive to g~n power over their rivals and 1 e -t lOpe fuIJ L ' v >ecome 1egenlons Once a state achreves that exaJtecl position , r 1 . I ' ecornes a sta tus qu0 power More needs to be sard, however, ahout th e meaning b of hegemony . . . . . . a state tllat 15 so powerful that 1t dommates . 11 t 1 othe1 states in a 1e A I1egemon rs .. . . . . . the S\stem . .\o other state has the mrhtary wherewrthal_ to put up .t senous fight .. / t agamst 1 . In essence, a hegemon is the only great power m the system. A state that . . . . . is substantially more powerful than the other great powers tn t1 s~stem _ not a 1e 's hegemon, because it faces, by defin ition, oth~r great ~owers . The Umted Kmgdom in the mid-nine teenth century. fo r example, IS somettmes calJed a hegemon. But it was not a hegemon, because there were four other great p~wers i~ Europe at the time Austria, France, Prussia, and Russia- and the Umted Krngdom did not uominate them in any meaningful way. In fact, eluting that pe riou , the United Kingdom considered France to be a serious threat to the balan<.:e of power. Europe in the nineteenth century was multipolar, not unipolar. Hegemony means domination of the system, which is usually interpreted to mean the entire world. It is possible, however, to apply the concept of a system more narrowly and use it to describe particuLa r regions, such as Eu rope, N01theast Asia, and the Western Hemisphere. Thus, one can distingui sh between global hegemons, which dominate the world, and regional hegemons, which dominate distinct geographical areas. The United States has been a regional hegemon in the Western Hemisphere for at least the past one hundred years. No othe r state in the Americas has sufficient military might to challenge it, which is why the United States is widely recognized as the only great power in its region ... .

fear <UTt'lll2: gn<ll pov.ers derive~ fro~rt the fact that thev invariablv ha\l' \ome offensive lltihlaJ') (apahility that ll1 (.;<tO U againSt ea<.:h ~ther, and the r~Ld that ey Se one can never be C'('liHirl that other ~tates do not intend to use tbat power against oneself'. Moreover, because states op<~ratc in an anar<.:hie system, there i!\ no night watc:hmau to \\horn they can turn for help if another great power attacks them. Although anarchy and uncertainty about other states' intentions create an irreducible level of fear among states that kads to power-tnaximizing behavior, they cannot ac<.:ounl for why son tetimcs that level of fear is greate r than at other times. The reason is that anarchy and the difficulty of disc:trnin g state intentions are eonstant facts of life, <.tnd constants cannot explain variation. The capability that states have to threaten <::ac:h other, however, varies from c:ase to c:ase, and it is the key faetor that driv<::s fear levels up and down. Specifi<.:ally, the more power a state possesses , the more fear it generates among its rivals. Germany, for example, was much more powerful at the end of the 1930s than it was at the decades beginning, which is why th e Soviets became inc:reasingly fearful of Germany over the course of that decade . ...


Su1 vival is the number one goal of great powers, according to my theory. In practice, however, states pursue non-sec:urity goals as well. For example, great powers invariably seek greater economic prosperity to enhance the welfare of their c:itizenry. They sometimes seek to promote a particular ideology abroad, as happened during the Cold war when the United States tried to spread democracy around the world anu the Soviet Union tried to sell communism. National unificatio n is another goal that sometimes motivates states, as it did with Prussia and Italy in tJ1e nineteen th century and Germany after the Cold War. Great powers also occasionally try to foster human rights around the globe. States might pursue any of these, as well as a number of other non-secmity goals. Offensive realism ce1 tainly recognizes that great powers might pursue these non-secu rity goals, but it has little to say about th.e m, save for one impmtant point: states can pursue them as long as the requisite behavior does not conBkt with balauce-of-power logic, which is often the case. Indeed, the pursui t of these nonsecurity goals sometimes complements the hunt for relative power. For example, Nazi Germany expanded into eastern Europe for both ideological and realist reasons, and the superpowers competed with each other dming tl1e Cold \Na.r for similar reasons. Fu rtl1ermore, greater economi c prosperity invariably means greater wealth, which has significan t implications for security, because wealth is t_he foundation of military power. Wealthy states can afford powerful military forces, which enJ1emce a states prospects for survival. ... Sometim es the pursuit of non-security goals has hardly any effect on the balance of power, one way or the other. Human rights interventions usually fit this description, hecause they tend to be small-scale operations that cost Little and do not detract fiom a great power's prospects fo r survival. For better or for worse, states are rarely willing to e>q>end blood rutd treasure to protect foreign


That great powers fear each other is a central aspect of life in the international ~-ystem. But as noted, the level of fear varies from <.:ac;e to case. F or example, the Soviet Union worried much less about Germany in 1930 than it did in 1939. How much states fear each other matters greatly, because the amount of fear between them ~~ely determines the severity of their security competi tion, as well as tl1e ~robab~Jty that th~y will fight a war. The more profound the fear is, tl1e ,~ore mtense 15 the secunty competition, and the more likely is war. The logic is strru~,t ~orw~d:. a scar~d state will look especially hard for ways to enhance its secu~~ :md at will be disposed to pursue risky policies to achieve that end. Therefore, tt 15 ~mportant to understand what causes states to fear each other more or less


. . t'o n:.- from t- "~ "' b. C" ' including 'gcnoeidt> F! r 'lotll<e , despite c:la trro us ~ . fuse d with I' tx)p\11 ,\' . . 11 JliO f<ll St.J ~. J .. ulin (1 992 ) 11'rls m f >cr 11!1 01IC\' lS I . 1 93 l l\,lt :\ mt' nc: Oll t-. dre d year in whi <1t 1 ") soklie . - . ishe t . l . trthep tsto ne 1 lUll . <nth m t.mre ounne . . A'ld in tltl t ea. < the loss of:~. 1ts'' ere klled '. n . . 1 . 1 'tarilll 11'HSSJOI1. I , c et in action on a tumam ~ ~ . Octobe r 1993 so traum,ttizcd Amen e elghteen . . . 1t 111 t'amous fu-eng1 m c:an pol soldiers m ~n . , ulled all U.S. troo ps out o r So malia ll)' m;.tkers that they immediated ~ the spri ng of 199-!. wlw n eth nic Hutu l) and then . d t ene in Rwan a m . re Iuse to tn en . t thetr Tutsj neighbors. Stoppm g tha t genocidwent on e .la} pacre atYru.ns .... ' a genoc1u ram o 0 d 1 ,ould have had drt ualh- no e ffe<:t on the Would 't .. an " . b een re lathelr ec\5)' tl balance of pow . POSttio ha,e . er. Y not hmg was don e. In h n et d 1e f tlle um'te. States m t prescribe human s o d rights inte ne ntio ns. it does not n ort' although realism oes no eces. Sarih -proscribe them. aJ .a 'tl . rsuit of non-securin go s cOILDJC ts \\1 1 balance-of-po But sometimes t1 pu 1e - . th cl' . . h' 1 . tates usuallv act according to e 1ctates of realism. wer F loC11C m w JC 1 cc\Se s .din d . o d . tb U s commitment to sprea g em ocr acy across the globe or .1 e.xample esp1te e 1 I d ~,erthrow democratically elected gO\em me nts a~l d em b :ac ed a number,of he pe . . . during the Cold \Var, "'hen Am e ncan poltcymakers felt that authontanan regtmes . . . Jd help contain the Sov1et Uruon. In \\ rorId \\ ,ar 11, the liberal d1ese actions wou . , . ies ut aside tlteir alltipathy for com murusm and formed an alliance with democrac P . the Soviet Union against Nazi Gertnall)' "1 can 't take comm~rusm , ,. Franklin Roose,elt emphasized, but to defeat Hitler "I would hold ~1a_nds Wl~l the De,il." In the same way, Stalin repeatedly demons trated that wh en hrs rde ologtcal preferences clashed with power c:onsiderations, the latter won out. To tak~ the most_blat~t example of his realism , tl1e Soviet Union form ed a ?on -ag gressiO n p~c t w1th ~azi Gennanv in August 193 9-- the infamous ~~lo lotov-Rrbbentrop Pac t-m hopes that tlte agr~ment would at least temporarily satis~' Hit ler's tenitori~ ambi_tions in eastern Europe and tu m the \Vehrmacht toward Fra nce and the Urute d Kingdom . When great powers confront a serious thre at, in sho rt, they pay littl e attention to ideology as they search for alliance partners. Security also trumps wealth whe n those two goals conflict, because 'defence,~ as Adam Smith wrote in The \Vealth of Nations, 'is of mu ch mo re importance than opulence. Smith provides a good illustration of how states beh ave whe n forced to choose between wealth and relative power. In 1651, Engla nd put into eff~ct the famous Navigation Act, protectionist legislation des ign ed to dam age Hollands ~~ merce and ultimately cripple the Du tch ec'Onomy. Th e legislation ma ndated that goods imported into England be carried either in Ena lish ships or ships owned b~ 0 . mally . the country that ong produc ed the goods. Sin ce the Du tch produced few ~ goods themselves, this measure would badly dam age the ir ship ping, the c:ent n . gredient m therr eco . . m nomic success. Of cou rse the Navigat ion Act wo uld hurtf England's economy as well, mainly because it wo cld rob Eng lan d of the benefit~ : free trade. "The act of navigation ," Smith wro te, "is not favorable to foreJ c~ co~merce, _ to the growth of that opulence tha t can aris e from it." Neve~le]es~f or Smith COilSldered the legislation "th e wisest of all the com mercial regulaoonsli\ h England.. ~ it did more damage to the Du tch eco nom y tha n to the Eng ~r ~ and m the mid-seventeenth cen tury Ho llan d was "tb e only nanll po' which could endanger the security of England." ...
c c




One mh! hl emu:lucJ,. from lhe prc~ceding dist'u sion that my theol) do<>s not allow for any cooperation mno ng the gre at povvers . But this conclusion would be wrong. StHtes can cooperall-. although cooperation is sometim es difficult to achieve and always difficult to S in. Two fa<:tors inhibit cooperati011: con side rations about l!Sta relative gaim. and con<:em about cheatin g. Ultimately, gre at powers live in a fundamentally com peti tive world whe re they view each oth er as real , or at leas t potential, ene mi es. and they therefore look to gain power at each oth er's expen e . Any two states contemplating cooperation must consider how profits or gains will be (lisbibuted between them. They can think about the division in terms of either absolute or relative gains (recall the distinction made earlier between pursuin g either absolute power or relative power; the concept here is the same). \IVith absolute gains. each side is concern ed vvith maximizing its own profits and <:ares Uttle abo ut how much the oth er side gains or loses in the deal. Each side cares about the other only to the extent that the othe r side's behavior affects its own prospects for achieving maximum profits. \Vit:h relative gains, on the other hand, each side considers not only its own individual gain , but also how well it fares compared to the other side. Because great powers care deeply about the balance of power, their thinking focuses on relative gains when they consider cooperating with other states. For sure, each state tries to maximize its absolute gain s; still, it is more important for a state to make sure that it does no worse, and perhaps better, tha n the other state in any agreement. Cooperation is more difficult to achieve, however, when states are attuned to relative gai ns rath er than absolute gains. This is because states con cerned about absolute gains have to make sure that if the pie is expanding, tl1ey are getting at least som e portion of the increase, whereas states that worry about rela tive gains mus t pay careful attention to how the pie is divided, which complicates cooperative efforts. Concerns about cheating also hinder cooperation. Great po,.,ers are often reluctant to ent er into cooperative agreem ents for fear that the otl1er side will che at on the agreement and gain a signjficant ad,antage. This concern is esp ecia lly acute in the military realm, causing a 'special peril of defection, " because the natu re of military weaponry allows for rapid shifts in the balan<:e of power. Such a development couJd cre ate a window of opportunity for the state tl1at cheats to infl ict a decisive defeat on its victim. These ban iers to cooperation notwithstanding, gre at powers do cooperate i n a realist world. Bruanc e-of-po,ver logic often causes great powers to fon n alliance s and cooperate against common ene mies. The United Kingdom, Fran<:e, and Rus sia, for example, were allies against Germany before alld during \\'orld \Va r I. Sta tes sometimes cooperate to gang up on a third stat e, as Ge rmany and the Sov iet Union did against Poland in 1939. More recentl y. Serbia and Croatia agreed to conque r and divide Bosnia between them , although the Uni ted States and its Eur ope an allies prevented the m from executing tl1eir agreement . Rhals as well as allies cooperate. Afte r all, deals can be stmck that rough!~' refl ect the distribution of power and satisfy concerns about cheating. The \'arious arms con trol agr eements signed br the superpowe rs during the Cold \\'ar illustrate this point.


CHY AND ll~ ~v~ ....... "' ~~ - -ANAR

. s t , howe,.et, 1 that cOOJ')eration takts 1, " m a world that is he estates have poweriul inc '1 11\'C.:S to take 'ld l'hl' l'IOttOIII llll. c van. . . ''--one \V l . .. . t . araphically htghhghted h. tl1 state of Euro '-'nnpl'ltll\'l:' at tt. ('OlC . 15 o pean 'S . t~\ t c Thts pom rore \ Vorld \ Var I. T1te grull powers cooper t . 1_ 1 lCI tl\~1.' ~, f ot 1 a ed f . 1 [!0 .+.. ,ears ue' l)(.'litk~ 111 l te . ".1 . od but that did not stop the m rom going to war frequent!~ dunng thts ~end St, tes and the So-viet Union also coope rated consid on er. 1 . 19 14. The UJUteb t a coor)eration did not pre,ent lte outbreak ofth that IJ :\.U2:\lSt l. e ar , u . du~ng \\orId \ V Germanv and Japan were defeatecl . Per1 1aps most amazing! , .. abl~ . ' and milit~rv cooperation between Nazi Germany Cold \\ ar shortJ~ after Y - 1 . ificant economiC there \\ClS stgn . d . tl1 two years before the \Vehrmac:ht attacked the Red and the Soviet Umonf unng . t~on can eliminate tbe dominating logic of sccurih l) . ~ . ount 0 coopera 1 . or a world in which states do not compete for power is . Am1~ .. o am ' l. tition Genwne peace, ~~~~~ely as iong as the state system remains anarc 1JC.

Anarchy Is What States Make oflt

Classical realists sud1 as Thomas Hobbes, Rein hold ;\iebuhr. and Hans J. Morgenthan attributed egoism and power politics primarily to human nature. whereas stmctural realists or neorealists emphasize anarchy. The difference stems in part fro m different interpretations of anarchy's causal powers. Kenneth \"faltz's work is important for both. In Man, the State, and War; he defines anarchy as a condition of possibility for or "permissive" cause of war, arguing that "wars occur because there is nothing to preve nt them." 1 It is tl1e human nature or domestic politics of predator states, however, that provide the initial impetus or "efficient" cause of conAi<:t which forces other states to respond in kind .. .. But ... In \~altz's Theory of lnternat:ional Politics ... the logic of anarchy seems by itself to constitute self-help and power politics as necessary features of worl.d politics.2 ... Waltz defines political structure in three dimensions: ordering principles (in this case, anarchy), principles of differentiation (which here drop out), and the distribution of capabilities.3 By itself, this definition predicts little about state behavior. It does not predict whether two states will be friends or foes, will recognize each oilier's sovereignty, will have dynastic ties, wilJ be revisionist or status quo powers. and so on. These factors, which are fundamentally intersubjective. affect states' security interests and thus the character of their interaction under anarchy.... Put more generally, without assumptions about the stru<:ture of identities and interests in the system, Waltz's definition of structure cannot predict the content or dynamics of anarchy. Self-help is one such intersubjective structure and, as such, does the decisive explanatory work in the theory. The question is whether self-help is a .logical or contingent feature of anarchy. In this section , I develop the concept of a 'structure of identity and interest" and show that no particular one follows logically from anarchy. A fundamental principle of constructivist social tl1eory is that people act toward objects, induding otl1er actors, on the basis of tl1e meanings tl1at the objects have for them. States act differently toward enemies tl1an they do toward friends because enemies are tl1reatening and friends are not. Anarchy and the disttibution of power are insufficient to tell us which is which. U.S. military power has a different significance for Canada than for Cuba, despite d1eir similar "structural" positions, ju~i as British missiles have a different significance for d1e United States than do Soviet missiles. The distribution of power may always affect states' calculations, but how it does so depends on the intersubjective understandings and expectations, on the "disoibution of knowledge," that constitute their conceptions of self and other.'1 If society "forgets'' what a university is, the powers and practices of professor and student c:ease
From Alex<mdcr Wcndt ...Anarchy Is \ Vhat States ~lake of Jt: ihe Social Construc:tion of Power Politics." lntemational OrganiUitiun, Vol.. -1.6. No. 2 (Sp1;ng 1992). pp. 391-42.5. @ 1992 by the World Peace Foundation and the Massachusetts institute oi'Technologv. Reprintt>tl by permission ofTht MJT Pres:-. Portiom of the text and some footnotes have been omitted.


1 . . 1 St t and Sodet Union decide t. \('\ are no I tt1 ~i t if thl' l mtel a es . onoe l ~ .J , ... It is <..'Ollective meam n):;s t ?.t \lllStitute th <:> r e;', Ht:' mh.s, "tlw Colu \\ ar s 0 ' e1. e struc. t ". or<l'mize our actions. .f U n.: ~ \\ l:'l" .d ti'ti' -relative)v stable. role-speCilC 1J nuerstanding -\ctors ncqlllre 1 en es . lf' b . rti.cipating io sue1 eo11ecti\ .c me(ltungs. ldennn s and 1 e n'lP('rntions about se - )' pa 1 r es are . Xr- . . . . al "Identity, with its attac m'lCn ts o psychologkal rea] m.herenth rel.ltiOll ,c ocally const mded world ,. p t B it\' is ah~,ws identity within a specmc, ~ 1 . . . a1 I , e er erger . _ l ,. identities linked to mstitu tion ro cs. such as brat} amues , Each person Jas man. . . ~ .. d .. sJmilarlv a state may I1ave mul bp1 as "soverei 1er.: f' son. teacher. an C itizen. J c. ld ,. ... nperial powe r,, and so on. ~11 1 com mitment to andgn, e th < Jeader of the tree wor . u . . .nh . ul .d tities vl:lrv but each identity IS an 1 erent1 social definiti e v salience ofpartic ar 1 en ' _ ,, ll . l h Id . l . d d . . the theories wh.ich actors eo ectwe y o a )OUt themselvon. of the actor groun e m f tl a1 Id es and one another and which c:onstitute the structu;e o ~e SOCl' " wor[l Identities are the basis of interests. Actors o no~ aveda thportdo 10 of interests d inder:>endent of social context; mstea , ey efine their interthat they carry aroun . . . f defining s1tuations. . . . 5 ometimes SLtuations. are unpreceests on tl1e process O . . ence and in these cases we have to construct tl1etr meaning. and dented m our expen ' ft tl .t ts by analogy or inven t them de no\o. More o en 1ey have routine th us our m eres , b . f u d fi qualities in which we assign meanings on the _a~ts o m~titution a y e ned roles. \1\'hen we say that professors have an "interest'. m t~achin.?, resea rc~; or going on leave, we are saying that to function in the .role 1?entity ~f professor, they ha\e to define certain situations as calling for certrun actions. Th1s does not mean that they will necessarily do so (expectations and competence do not equal performance). but if they do not, they will not get tenure. The absence or fail~ re of roles makes ~eEining situations and interests more difficult, and identity confus1on may result. Tl11S seems to be happening today in the United States and the fann er Sovie t Union : \Vithout the cold war's mutual attributions of threat and hostility to define their identities, these states seem unsure of what their ..interests" should be. An institution is a relatively stable set or "struc ture" of identities and interests. Such structures are often codified in fonnal rules and nonn s, but these hare motivational force only in virtue of actors' socialization to and participation in collective knowledge. Institutions are fundamentally cogn itive entities that do not exist apart from actors' ideas about how the world works. This does not mean that institutions are not real or objective, that they are "nothing but" belie fs. As t'Olle~~-e knowledge, they are experienced as having an existence "over and above the in@,d uals who happen to embody them at the moment.'>6 In tlUs way, institutions com~ to confront individuals as more or less coercive social facts, but they are still a func:tion of what actors collectively "know." Identities and such collective cognitions do ~01 exist apart from each other; they are "mutually constitutive." On this view, inso~ ~on is ~ proress of internalizing new identities and interests, ~ot sorne~! occumng ~ts1de them and affecting only behav:ior; socialization 15 a cogruti . ~ not J~ a behavioral one. Conceived in this way, institutions may be c<>?r:. ative or conflictual, a point sometimes lost in scholarship on international regun which tends to.equat e in....:tutions W1'th coope . u..,u ration. There are importan t derences . JJI I ble !:hahween.. confli~ and cooperative institutions to be sure, but all relatively sta ~relations even those of"enemies"-are defined intersubjectively.



i ,

Self-lwlp J~ a11 institution, one of variow. sln1ctnrc~ of idt.:ntitv and int<-r('\l thnt , may exbt umlt-r tnard t). Proccss~s of identity formation under an.m: ll~ art" concerned first and f(mmost v.:ith preservatio n or "security'' of the sdL Concept'> of secu rity there{ow differ in th<: c~xtent lo which and the manner in "vhich th<' s<lf i'> identified cognitivelywith the other, and , l want t<J suggest, it is upon thb wgnitivc variation that the mc:aning of anarchy and the distribution of power dPpc:n<ls. LC'l rne illustralC' with a standard continuum of security systems. At one end is the "cOJn j)Ctitive" security svstem in which states identirv ' , ' negati vely with ea<.:h other's secutity so that ego's gain is seen as alter's loss. ~egative identification under anarchy constitutes systems of "rcaJist" powe r politics: risk-averse actors that infer intentions from capabilities and worry about relative gains and losses. At the limit- in the Hobbesian war of all against all-c ollective action is nearly impossible in such a system because each actor must constantly fear being stabb ed in th e back. In the middl e is the "individualistic" security system , in which states are indifferent to the relati onship between their own and others' security. This constitutes "neoliberal " systems: States are still self-regarding about their security but are concemeu primarily with absolute gains rathe r than relative gains. One's position in the distrib ution of power is less important, and collective action is more possible (though stiJI subject to free riding because states contin ue to be "egoists"). Competitive and individualistic systems are both "self-help" forms of anarchy in the sense that states do not positively identify the security of self with that of others but instead treat security as the individual responsibility of each. Given the lack of a positi ve cognitive identification on the basis of whkh to build security regimes, power politics within such systems wiJJ necessarily consist of efforts to manipulate others to satisfy self-regarding irJterests. This contrasts with the "cooperative" security system, in which states identify positively with one another so that the security of each is perceived as the responsihility of all. This is not self-h elp in any interesting sense, since the ,.selr' in tem1s of which interests are defined is the community; national interests are international interests. In practice, of course, the extent to which states identify v.itb the community varies from tl1e umited fom1 found in "concerts'' to the fuiJ-blown form seen in "collective security" arrangements. Depending on bow well developed the collective self is, it will produce security practices that are in nrrying degrees altruistic or prosocial. This makes collective action less depe ndent on the presence of active threats and less prone to free riding. \'loreover, it restructtues efforts to advance one's objectives, or upower politics," in terms of shared nom1s rather than relative powe r. On this view, the tende ncy in internation al relations scholarship to view power and institutions a'i two opposing explanations of foreign policy is there fore misle ading, since anarchy and the distribution of powe r only have meaning for state action in vi1tue of the unde rstandings and expectation s that constitute institutional identities and interests. Self-help is one such institution, constituting one kind of anarchy but not the only kind. Waltz's three-part defirution of structure there fore seems underspeci fied. In order to go from structure to action , we need to add a fourth: the intersubjectively constituted structure of identities and interests in the syste1n.

\ \m h,,, .111 . liiiJlt lll .tll t 1111p 1 '.1t'11 >11 f'111 till' \\':1\ 111 \\ Itu l1 ' ' ' <011 ''1\c 0 1\ l : ct t ~ l 1< 11 , "( t h , (,\ 11' t ) I. 11:1l til l' 1 1 ' tl 11'11 !it<!( t'tl('cll tltl tt' wtllt t: wl t otl u I Hcc11 11_., 1 '' ol t 1' \ ti() ll' Ill( ' . 11' I 0 tl 11 ' I 'lttl l ll ttl ' .\1'('11 111 \ IIJIC II'S (\ lll'll( ,. ' "' 1 h,,,, . t-\ HH'' ' PIIIl ll' nl w a ll < 11 ttlldt dH lll l till' \ l' of tttc t1 : \V, l'tlln 1 (0 0 1 pnot hi ttt h'tnd lllll \\t ' 11"''' 1111' ' . ' < o,,,.,,,. 11 ' '"' li kv 1 \\l th \\'alt t thnt , lll \'i l111t' nf' :tlltll't'lt.\', "illll 't'ltlllill ttll l , oltt w:al '\.' p '<'1111tJtr1it \\I .d~q "" ""H II )() 1 I ' t ' ;t l( ' I0 t I 11 I < Il l\ tltc wwt iOIIIII 'lclf-l't'l!,llldtt tg "'"" \\,ll M ' I . I ,. . 1,. j . :111 1 ~l ntt '.' 111 I 11 ' ' l:tl t <d tta ltarc 11 1vv,s "' lt\\ll' I' I 1 \\I ' :11'~1 11 ' I 1!1 1' Ill 11 1 1 0 .111 \ ~I{.. . . . 1 1 . .. .. '\(!I~ 1 t" . ..... 11'1 dlh' tll lttn. II H :-c <l. tllrt:oi' P1<' 11 J>POS' < :1 ltt. sto1ol l y 1v f' I,\l't1 11 11111 1)1 .\( ' I , , ., , " , , , . ' \' () . 1 ' tltlc'l ,ldHlll Ill \\'1 l ' 1 !I(' 1Ill'\ 1'1, .,. 'tt'cJtll,ll'cl stlft.., h t< h ttl tl ll ' .\l tcltttll t1,1v' t,11 Il I , llt'j . , ( 1 t'll 1 .1J,..,t 1 ndio ll lro 111 h 1 ~l :11tcl .\1'( '1111< 1 llll:tg c fal'llll's) tit lllh'l':l('tll)ll :\Ill :) I 11 ' ' . .. ' ~'Y . .l ' Ill )I ll l whiC'It In lutS(' .\11('11 dlrttttl ll l1 c)l' '('" :l lld otiH t .,. 1\ \\'t)ll Ill 1,1\'(' 1\l) C\)Jt' l1t'll( l , . 11 . . 11,1111.11111 11 lo sttl t:-. in ti H .\l!tl c n l ll:tltt n cp a itH '\ tl 11 tl :tl t 1 .l,:o.III IH' ot Iwtwr:.c I' c . , . '}' .. . .,,,,1,.1 ,~ . Stll' hl'll) j.., an ttlslltttll tJ II , 1101 :t t'C III.\It(ttl tvt lv:1t11 n.r ell .,tv t:all 011 Iy posstss 111 ' anan ll\ . ''1 1. tlll,11 . is a co~ts ti lu tiv<" lt:ttl tr'< of. tht ~l :tl t oltt: tt tlr<' 1H ' f'on ittl crat'l iott? ' \ ., m . I I. I 'l\vo tlt in!];S an ll'l'l if w strip awnv tllm t' [H'OJWI'lits oft H S<' ' . "' 1idt pns 11 ppo!lc inltrad i<;ll wi tlt otltcr, . Till' fi rst i~ ti lt' li iHlt' rial ~ ~~ b..,tr: tl <: ol :.'~< t~t v. i 11dud i 11 ~ it:-. t.inl ri tt-"ic t:tpnhilit it's. Fctr httl llll tt bl'i ngs. t.lti:-. i~ tl w l>odv: f~ >r ~t.:t l t~ . it 1; au . Ol'~ltltiltlt lott:tl appanttiiS or ~()V('I'tlall ('(> Ill tflttl. I :11 11 Stl~{ ' \ 111 1~ fm ri H'tOtkal ptupo:-.ts that the mw u~:tl <rial nul of wltic h I ~ H'tt tl >< r!l o l ti H' ~l~tlt ' )-"1<111 :tH tonstit11 ttd i-; tnalld by clonw:-.tic sol'itl l)('lo n ~l n l ts cnlt r tir e t'OII S til ~tliv< pnxt ~~ ofittl l'l'ttalionu l socit'ly. altho ugh th b p ro(T SS illl pl its mitlt<'l" ~tah k l<'rt ilo rialit \ ttor .\IJ\ tni)111 )'. wltidt an intcna:ttiollally ll c'gotiat< d tt 'I'll I ~ of illdi vidt~ ality (as disuu11scd liutlt t r lwlow). Tlw :-;t<'Olld is a dL to pnst tv< thi s II IH sin I('rial suhstmlt', to 'ilm~w. This dot-; not <' lllail "~w l f'-r<'garcl i ii~ H'i~ ... ltm cvc r.< :ut()J'~ du IIUI lt:t\'(' u wlrplior I0 j n t (' ract ion wi tit :t not her: ho\ tl wy viI '\V tltc Jll( a11 in~ mad l'tquhcuwuts of' this stuv ivalt lttn l(m dl'IH'uds 011 I he prm:l'ssc s I 1 \ovl1 id1 )' ( '(11\('( ptions ~d r( volvt>. This may all SC('ll l V('l) ' IU<'l UIC' , but there is a11 illtpo rl:lnl isstw at stalu : Ar<" ti ll' forti~n poli<.y idtntiti('Sand int Pnsts of' stal<s (">gtnotll> or cndo~<uo us to tl w .\lall' syslt m? Tht l(mtwr i'i tl w auswer of' an individ11al isti<. or tlliO<.' rSn('ialii'.t 'd ~ r:-.ll'lll ic tltPory l(,r whkh rationalism is npproprial t; tlw latte r is t ltc a11swtr of' a r1111 ' sorlnl iZt'<l sys~t:m~<- thlOJ y. Waltz Sl'{'ll IS to on~r tl w Iall< r ami p rop oM s IWO ll Hd l(lllb lil~. <'0111 JM'hhl~H and sodalizat iou. hy wlaith str11t:lnre <:ond itions stal e actio n." Tl w t'(l~lkllt of his argt~mc'n.t uhout tlds r ondilioning, ltowcvn , pn.s uppos< 'S a s<:ll'-lwlp systtm thl~t Is not Jts<>ll a (.t)nstitutiv< fcutu n of' anard ry. As Jant <..s Mo rrow poi nts nut , Waltz~ two lltt<:hanisms <.'ondition hPimvior. not idt.~tt ity and iut corcsl. .. . w If sl'llhtltl is not ''\<.'<)I IS I U t'I V( ('eatttJ 'l' 0 f' t'l .. anun.:II V. it II1 11Sl l ' llH'I'g t' l ';I II S<I llv fnnn nn)('tssts in wlli<:l1 .,,, I 1 ~<.: a P uys 011 1 a JWnuissivcrole. This r<"flc: l'l s a sccwH I y y p~nr~plc orfc:onstmctivism: tlaut tlac lllt!iillill~S in IW'Ili S of whiC'h act ion is orgu niY.I'd llnSC out o mt<rattion . ... tonsid<r two actOI'S- (. , ) . I I . ti 11 F 1 W llllu a t(!f~ n<:ouutcring c 11 s ac:h other for t Iac r. I me. .ne l wants to su1 viw 1mtl l . ~ 11tl' ocior l 11 1 a1 has <.:t.rtam material ('ll]Xlhiliti es, hut 11('1 1 IlL, l 0 t)gic. or domc.sti tht.'n" is uo lai"l f . l IIIIJ>crutJws for powcr, glorv, or COIH (I u si . all< I " orv o 5t><unty or in .1111't 1 1 Realists would 1, II S(.'( Y >etwttn the two . What shou ld t1cy l I0 ? 1 ro Ja) y argue tl.ut tach should ac t on tl< hasis of wors t-case

Wl Nl:rt I ANA RC IIY 1:.-. WIIAI 'J IAr I C., MAKI ()I 11



:' "'""" ' '"''' I 1 1 tl c otlt~l' lall rll rnn ' J""t tl yuw '> He 1. "' .dltt 11d ,,., P"" '' 111 l11 vicW ul tl 11 j 1 1 tLIIth of dc,\11, l111111 tr1:1k 1tt~ ,, tnt,t. ,kc ~11t lt .t ('11\'> t!Hiit~ ,d \.';t', 1xi..; ts. I '\ I 11 '" ( tVII \111' 11 I ) III ,W t'\l' t, " '' "''Y ld lw itt apt 1,,il ,lt t1 jll'tlplr 11 1 o~dt woat dt thlo" ' (>'"' I t/11 llac la.,.j.., 11l Wl> l''tl l'1\t' pn..,,i!Jil ilit..,, lll, l<ad tii0\1 dc< i;,JrJII.., at :1ncl ' ''""I d h lll.dt '"' ll w IJ.t\1\ of p1111 J:tbilitrt \ :tlld tlw.,,. '" '' pc,d1Hr d h) inlt t: H'IIIIII h\ 'lt:al .tl ' (l)l t, r/u . In tlw IJc ~tr11ci 11g i" q~o's ).',1'\111 1<' wlaith 111ay t'lllt\ i\1 lor c.'<:1111pl< qf :lrt a d v:u l<'l ' a 11 111'.11 .11JI:tlld i. . l,iiiJ41lf :u!IJ\ , : t l ayit~g dow ll of :u111\ , n1 '" :1lt:JC'k . F c11' cgo. tl1 i' gt\ ltttl ' n 'jlt'l '\1'111\ t ill' h:l\i.\ otr wltil:l a it j., prvp:t rc d to n... pc111 lo :tltr. d ' J'It i~ l1 > i\ j, tlla krHIW II lo all tr laowcvt r, u11d M> il 11 111\l lll:tk< an inl'c < t o r .11lri :t!' r ut l11tl inn" :.l 1o1tl q~o' iflll'tllio!l s :t lld , in p:ttl k 1tl:u, giw1 Ll tal tl1i., ;, anarchy abo11 t 1 wlrd l cc r q~o ;.., :t t l t~ (: ll . 'I'll(' < l1Jtl < lt i-. ill f'cn:trtl willl a rgc:ly dq><fl d 011 twn '<111 lt <otrs id('l': il ic n t~. Tl w fir-;1 i.\ tltl' gt'iltrlv:, a11d cgo\ phys i< cpml il il'' wl. id t an :al ill I""'' ('Oil l riVI'd hy 1 ' )41) :urd wll idt int'lu d< LIJC dirl' d ic)ll or III()V tr lt:ll l, nobc. t tllttlt l>l'r~ . :111d irt lll r< alt l'IJIIS I'qrtt ~llt'< '.\ ol' ti H' g< slii i'C'. TIIC' " to nd ton<; id<ra l io11 di < 'tlllt'( 'rtl.\ wlt:cl alt er would i11L(1ad l.y o; ud t ll'l:clitil'S W itlo 111akt 'llld t a gt\l ur< <n ils<ll'. Alt11 lll:ty t11ake an all rii HIIiotral "trro r" i11 it.\ in l'tnllc( aho11 t <go'c; i11l< ' 11l, b11ttlw rt' is al so lt< l r<!aSIJII l'o r ill o n.s.S111111' a priori IH.f'orc ti iC g(.'S(II n> - L q~o is I1al tlllcal < llillg , .sii!L it is only tltrougl a prot v~s of' siguali 11g and ill lcrpr<li ng that 't' ti ll' ro~t -" :uul prohabilitic -" o l' lwin g wrottg can I><' d<"ltnnii iL:d. Soc:i al th nat" arc C'O II Slrt I(' (( d, Ill) I ll:t[ll r: d . C:oll ~id<"r all (X:llllpl< vVould \V(' a. 11111<' , a pricJr i, that \V(' W r( abou t to I>< . % L a ll a~; h d if wt an <vc r t ottl a<:t<d l1 ttH'II d l('r~ of' att ali ' 11 c:ivili zal iott ~ I tlaink n ()l. y W<; wo11 ld lw ltigltly :.dvrt , ol' tom sv, b11 t wl1vtl1cr we pla< (J ilt' militat) ' for<.:(:!) nn :cd al<' rl or la1111dlvd att :tlta('k would dqw nd on how w< int ('rpr<'t<d the intpo rt of' thl'ir rirlll ~tslu n: I(H' 0 11 1' st<.:urity- if only to avoid lllak illg an itt lll l('dia l< l: fl my 0111 of' wl1 at lliH IH' a d:tt lgt rotts adve rsary . Ti lt' poss ibility of' erro r. itt olht r word\ . Y do<s not f(m:L tts to act on tl1 assn111pt iu11 tl1al tlw alitns an llt real<'ll ing: Actio n e dqw nds on tiH: pm ltahililic!s w< assign, and tiles(' art i11 key part a ftii H ;t'iott of' what tlw ali< ll~ do; prior to tlwir gcsl11 n , w <. haV<' 110 sy~tc t nic l n.L\i~ 1 assi~ ning proba()1' hiliti('S. If' tl1dr first gcsl11n is to appPar with a thousand spa<.t~l tips and dtst ruy < York, we will d< fitt < tiH si t11ali 011 as tlt r<.at< 11ing and r<.;spo 11J acco rdi11g l '. But 'W ' if' thl'y app< 'ar with Oil( spal't'Slt ip. sayi11g w lt:1t sec.111s lo h<.: "wC> com e in pLaet." Wl' wi ll fctl "nassund'' and will p roh~1 bly n spoud wi th n gcst mc inletu.k d to nass un tlll'lll, cv< n if litis gcs tllrc is not llL't < ssarily intcr prdltl by them as ..;udt. Tllb pr()(.'t'SS or si~llal iug, intcr pr<'ling. aud nspo ndi n~ <:otnpiC'IC'S a "social act" and h< ns tlw proc(ss of' creating iu ters11hjct: tivl' ll ll'at tin~s. It adva nce~ the same gi way. Tht" first social act en at es l 'XJH' <.:l ations on hntl1 sides ahuut each oth(r's fut urc b,.J,avior: pol<nti:tll) nli.sta k<.n and cNtainly tenta tive. but C\pc<.:tation!<l non thckss. Basl'd on this tt ntntiv< knowkdg<, t"go nwkc:s a JWW gestu r<', agajn signil~ving th<! basis 0 11 \vltidl it will respo nd to alt er. and agai n alt <.r r<:spond~. adding to tht pool of kuowlt:dg< cad 1 has ahn11t th(' olhC' r, a11d so un owr lime . The nwc hau is11l he re is niulc>rcL ll; iut cmc tion rewards a<tors lill hold in~ <Ltiaiu ideas about each othe r 'IIIt'l "' and d istou ragt' s tl1c' n1 frotll hold ing othe rs. If' rcpt~tt ed long enough. thes t "rcci protaltypificalions" will (.'r<'at< r('lativd y st:~hl < <:Onl'cpls of S<:lf and otlu.- r regardin~ tlw issttc at stake i11 thl' in1N aclio ll . 1 ~

tnwhil . . . . , thei r secnntY tllll .tlct< ) l 1' ' ' .tten~ these. . C:t ldliCC r !Ilt l rlo1ts of tC.'tOI"$ to en 1 < d a lienation. 1~ he 101 Ill' ol denti tY a '<:urt"' of . . '' , . t tatin{l' chst rust an . nu tnte t1 t)t Ilt rs. pe qJe l ll) e> I , are th emsel\'e ~ o wuinu r.Jo [r rest 1 . ucb dilemmas, lOWe' er. :'"< b "' ec:ts of con ttute . . t. . identities are prouuc:ed m and through .. t . not , to the tntetac 1on. st ttat d 1 1 l \O~t'l c~.t ~ d our relationship with the aliens ilt a security dil e 1 <Wthitv. '"' \Ve o not uegm . ., emtna : . . t oiven bv anarchy or ll < c... . . . ltuJ ecuntv dilemmas ate no 1::r d f . t1 of' identity formation ts a cru c a<.:count o how the pt The mmor 1eorv . 1 ll ocess . . .d .0. a'n(l interests might v"ork, but 1 c ocs not tc us why a syste t of creating 1 enti es . m . abl)' our own-wouJd have en d ccl up wtt1 self-regarding. of 1 states-sueI1 as, rugu . . ffi lnd . . .d bties In this section, I exammc an c c1 cause, prcdan ent not co11ecb'e 1 en . . on .} . . b'on with anarchY as a pernuss1ve cause, may aenerate a self-h 1' w htc 1, m conJune , . .. . . . ep 1 t 1 structUJ e of 1dentittes d . however I show the key role t 1at 1e S\stem. In so d omg, , ' , . an ~terests plays in mediating ana_rchy s ~xplanat~?' ':le. , . dator arQl.lment 1s strrughtfmwcud ,md compellmg. For v.rhatev, o . . . . The pre . er . ogy, dotrtestic politics or svstem1c victmllzat!O n-some states 1,.. reasons-b101 ' ' . . . ..ay become predisposed toward aggression. The ~ggressivc.l~e haviOr of th~s: predators or "bad apples" fo rces other states to engage m competitive power pohhcs, to meet fire with fire, since failure to do so may degrade or destroy them. One predator will best a hundred pacifists because anarchy provides no gu~rantees. This argument is powerful in part because it is so weak: ~ther th~n making ~he_ strong assumption that all states are inherently power-seekin g (a pU! ely reducborust themy of power politics), it assu mes that just one is power-see~ing and that the others have to follow suit because anarchy permits the one to explmt th em. In making this argument, it is important to reiterate that the possibili ty of predation does not in itself fo rce states to anticipate it a priori vvith competitive power politics of their own. The possibility of predation does not mean that "war may at any moment occur"; it may in fact be extremely unlikely. Once a predator emerges. however, it may condition identity and interest formation in the following manner. In an anarchy of two, if ego is predatory, alter must e ithe r define its security in self-help terms or pay the price .. . . The timing of the eme rgence of predation relative to the history of identity formation in th e co mmun ity is therefore crucial to anarchy's explanatory role as a permissive cause. Predation will always lead victims to defend themselves, but whether defense will be collective or not depends on the history of interaction within the potential collective as much as on the ambitions of the p redator. Will the disappearan ce of the So,~et threat renew old insecurities among the mem bers of the Nort h Atlantic Trea~' Organization ? Pe rhaps, but not if they have reasons independen t of that threat for identifying their security with one anothe r. Ide ntities and interests are relationship-specific, not intrinsic attributes of a "portfolio"; states may be com petitive in some relationships and solida.IJ' in othe rs .... The source of predation also matters. If it stems from unit-level causes that_ ~e immu~e- to s~emic impacts (causes such as hu man nature or don:esti~ ~~: taken m lSOiation), then it func:t:ions in a manner analogous to a "genetic tnut 11 oonstructed world of the state system. Even if successful , this trait does not se e<:t r other pre dators an evolutionary tl1er states 0 or m sense so much as it teaches o

r. ter'1ction are.prone to sec 111 (\Hnlwtilh s, , tcntso 111 1 JI

lt'l~'trt mas ,..



respond in lnn :1, bHt <;iitee traitc; cannot he unlearnetl. the other state-; \\ill c:onli11ue compelitt 't.; hch t\~Or until lhl: predator is either tlestroycJ or transfonned from withi n. Ho\\ I cr. in the rnore likely event tl1at predation stems at least in part from pdor syo;tc rni' iult'raetion- perhaps as a result of being victi mizcd in the pac;t (one thinks hcH' of ?\azi Germany or the Soviet Union)- then it is more a responsC' to a learn('tl identity and, as ~uc h , might be transformed by future social interaction in the form of appeasement, n:.:assurances that secUJity needs will be met, syste mic effe<.;ts on do1ncslic politics, anti so on. In this case, in other words, there is rnon hope that process can transform a bac.l apple into a good one .. . . This raises anew the guestion of exactly how much and what kind of role hu man nature and domestic politics play in world politics. The greater and more destructive this role, the more significant predation will be, and the less amenable anarchy will b:: to formation of collective identities. Classical realists, of course. assume<l that human nature was possessed by an inherent lust for power or glory. My argument suggests that assumptions such as this were made for a rea<>on: An unchanging Hobbesian man provides tl1e powerful efficient cause necessary for a relentless pessimism about world politics that anarchic structure alone, or e,en structure plus intem1ittent predation, cannot supply. ... Assuming for now that systemic theories of identity formation in world politics are worth pursuing, let me conclude by suggesting that tl1e realist-rationalist alliance "reifies" self-help in the sense of treating it as something separate from the practices by which it is produced and sustained. Peter Berger and Thomas Luckmann define reification as follows: "[It] is the apprehension of the products of human activity as if they were something else than human products-such as facts of nature, results of cosmic laws, or manifestations of divine will. Reification implies tl1at man is capable of forgetting his own authorship of the human world, and further, that the dialectic between man, the producer, and his products is lost to consciousness. The reified world is ... experienced by man as a strange facticity, an opu.s alienu m over which he has no control rather than as the opu.s proprium of his own productive activity. "14 By denying or bracketing states' collective authorship of their identities and interests, in other words, the realist-rationalist alliance denies or brackets the fact that competitive power politics help create a ve1 "problem of order" they are supposed to solve-that realism is a y selffulfilling prophecy. Far from being exogenously given, the intersubjecthe knowledge that constitutes competitive identities and interests is constructed every day by processes of"social will formation."15 It is what states have made of themseh-es.

l. Kenneth waltz, Man, the State, and War (New Y ork: Columbia Uni\'ersity Press. 19.59). p. 232. 2. Kenncth Waltz, Theoi1J of International Politics (Boston: Adclison-\\'eslev. 1979). 3. Waltz, Themyoflnternational Politics, pp. 79-101. . 4. The phrase "distribution of knowledge" is Barry Bames's, as discussed in his work The Nature of Power (Cambridge: Polity Press, 1988); see also Peter Berger and Thomas

Luckmann, The Social Constntction of Reality (~ew Y Anchor Books. 1966' . ork:


a~J)!t.'r. "ldt"ntity as a Problem in UH. Sociolog) or K11()\'.. ltd:.!

I u ropea 11 j


c if

. 'l<'i(l/o~y. i . 1 t}966). lll. 6. Be1v~r :lnd Ludm1:mn. P " '

ritv Dilemma." \\ (Jiid Politics 30 (Janua0 l; t _:_16 ' - . 9 w:utz. TheonJ aflnternatimwl Politics. PP 14- t t . . , . ee Jawes >.1ono\\, sOCic Choice aud Svstem Structure 111 \\oriel Pohti<:s," \Vo1lt flo/. :~ . u 10. c ifics 41 (October 1988). 89. . . . . . . ~ . . . 1 mctapboncal m wo1 le1 po1 1ttcs. s1nce throughout 11 .. 1 Storv 11 This Sttuaoon 1s not en0 re Y . h . states ba,e "discovered" eaeh other generatinu an mstant anarc ,. as tl were. A 5 , . . o . . )stem. . . ca1 d . f first contacts would be interesting. aoc empm stu ) o k . - ., .. . ___ ,.I ifications ., see Beruer and Luc rnann. pp. :>-r-;)8 . 12. On rectproca.t t)p , I w J ''$'tu t I .t" 13. See C. ~orman Alexander and Mary G enn 1 ey, J a e~ f \ cr1\1 . an d I clentity For. . maoon... . . oms Rosenberg and Ralph Turner. ed .. Socwl Psychology: SociolcJa 1 mM . btca Perspectives (New York: Basic Books, 1981), PP 269-89. 14. See Berger and Luckmann, p. 89. . ... 15. See Richard Ashley, ''Social Will ru1d lntemational Anarchy, 111 Hayward Alker and Richard Ashley, eds., After Realism, work in progress, Massachusetts Institute of Tech. nology, Cambridge, and Arizona State University. Tempe, 1992.

7. Wal~., Tht'0'1J of Jntcnrativunl Politics. P 91. . . .. . .. d \lnr and Robe1t j erv1s. Coopl r,thon Under th S hc State. nn . :t"t.' Waltz. . \lnu. t . -a) - _91-l e E'<:u.


The Conditions for Cooperation in World Politics


Nations dwell in perpetual anarchy, for no central authority imposes limits on the pursuit of sovereign interests. This common condition gives rise to diverse outcomes. Relations among states are marked by war and concert, arms races and anns control, trade wars and tariff truces, financial panics and rescues, competiti\e devaluation and monetary stabilization. At times, the absence of centralized international authmity precludes attainment of common goals. Because, as states, they cannot cede ultimate control over their conduct to a supranational sovereign, they cannot guarantee that they adhere to their promises. The possibility of a breach of promise can impede cooperation even when cooperation would leave all better off. Yet, at other times, states do realize com mon goals through cooperation under anarchy. Despite the absence of any ultimate international authority, govemments often bind themselves to mutually advantageous courses of action. And, though no international sovereign stands ready to enforce the tem1s of agreement, states can realize common interests through tacit cooperation, fonnal bilateral and multi-lateral negotiation, and the creation of intemational regimes. The question is: if international relations can approximate both a Hobbesian state of nature and a Lockean evil society, why does cooperation emerge in some cases and not in others?

From ''El(p1aining Cooperation under Anarchy: Hypothesis and Strategies" b)' Kenneth A. Oyt> from World Politics. pp. 1- 22. Reprinted by pemtission ofJohns Hopkins University Press. Portions of tlw text and some footnotes have been omitted.








ar. question. to guarantee adherence> to agrc~c ment . 1 1011 clllj? Gi\'l'll the lack of a centra au )ermit states lo bind lh<?lii S<"'Iv~s to rnut:us. 1 . f . . t' s encourage or .r what features o srtua ron . ? \' 'I t fe'ttures of situations preclud" coope ration? .11 benenc1aJ c .r ;u > cow-ses of action 'v ta t < ~'oster the emergence c~ cooperation b d to o Y Sccond, w I10 1 st1oteuie.(j ca11 states .r. op !> Governments need nol ne<:essarily accept e ..,1 .1' . Ley altenng t7 c1rcu.mstances t1 C01':Jtro, are situational impedi men ts to coope ration te nt . . as <riven To w1 ex e 1at . cm:umstrul<:~s e dul . . ? Through what highe r order stratcgtcs can states eresubject to wt.llful mo cation. . ? .h nditions for cooperatron . . .. ate t e preco . tantial dimensions se1ve both a~ proximalc expla1 submit t11at t1 c1rcums uee . . d t <~ets oflonger-term strategtes to promote coope ranations of cooperation an as aro . . -0 . sections of this piece cle fi nes a cl'1mens10n. explains tion Each of tIle tJ1ree maJ I . . . . . ts for the incidence of cooperahon ancl conf!rct m the how that dtmensJOn accoun . . _ . r . . .. d. thority and exam tnes assoctated str ategt es 101 enhan c:mg absence of centralJZe au the prospects for cooperation. A . In the section entitJed "Payoff Structure: ~tutu al and Co~ 1 ctmg Preferences," I discuss how payoff$ affect tl1e prosp~ts for coo_pe rat1on ami present strategies to improve the prospects for cooperaho~ by alter~ng payoffs. Orthodox game theorists identify optimal ~trategies giv~n or~nally defined. ~lass~s of games. and tJ1eir familiar insigh ts provtde the stmting pomt for the dt scuss1on. Hecent works in security studies, institutional microeconomics, and international political e<:onomy suggest strategies to alter payoff structures and th ereby improve the prospects for cooperation. 1 In the next section , entitled "Shadow of the Future: Single-play and Iterated Games," I discuss how the prospect of continuing interaction affects the likelih ood of cooperation; examine how strategies of reciprocity can provide direct paths to cooperative outcomes under iterated conditions; and suggest strategies to length en the shadow of the future.2 In addition, this section shows that recognition and control capabi~ties-tbe abiJity to distinguish between cooperation and defection b~ othei'S and to respond in kind- can affect the powe r of reciprocity, and suggests strategies to improve recognition capabilities. In the third section, '!\umber of Players: Two-Person and N-Pe rson Games.' I explain why cooperation becomes more difficult as tl1e numb er of actors in~re.ases; p~esent strategies for promoting cooperation in N-acto r situations; and oJfer strategJes for promoting cooperation by reducing the number of actors necessary to tbe realization of common mterests. G ame theorists and oligopoly t1 ts teons have long noted that coope ti b . . h . . . ra on ecornes more difficult as numbers increase, and t elr~nsJghts provide a starting point for discussion. Recent work in political economy IOCuses on two strateoi {! . t' . r. . . o-es or promotmg cooperation in thorn y N -perso n SJtuarons. mnctionalist analysts of . l'k lib od d b regimes suggest strategies for increasmg t1te 1 e o an ro ustness of coo " ti ad hoc ba . . . . pera on gwen 1 arge numbers of actors, 3 analysts 0 f rgammg m mtemational politi aJ and relrional d .. c economy suggest strategies of b11 al ateJ . o ecompos1tion to reduce th l1.ation of some m t a1 . e nu mber of actors necessary tothe realr cooperation .... 4. u u mterests' at the expense 01 the magnitude of ga111s frOJ11

. 1 "] . d l. both ext> Iana t ory .111 d [)rescrir)tivc aspe<.:l~ of tlris IWrcnn 1 ra lSchom:; ,\t ress fi. <theemergenceofco<> iwn . JJCrtJ
F" 1 / 1circumstances nvo 1 IfS . tc w tJ ty




The stmcture of payoffs in u given round of play- the benefits of mutnal cooperation (CC) relali\(: to mutual defection (DD) and the benefits of unilateral defection (DC) relative ro unrequ ited c:ooperation (CO)- is fundamental to the anal}sis of coope ration . The argum ent proc:ectls in three staaes . First. how does payofT structure affect th<:: significance of c:oopcrahon ? .\~t ore narrowly, when is cooperation, denned in terms of conscious policy coordinatio n. necessary to the realization of muht al interesls? Secon d, how clot!s payoff structure affect the likelihood and robustness of coope ration? Third , tl1rough what strategies can states increase the long-term prospects fo r coope ration by alterin g payoff structures? Before turnin g to these questions, consider briefly some tangible and intangible determinants of payoff structures. The security and political economy literatures examine th e effects of military force structure and doctrine, economic ideology, the size of currency reserves, macroeconomic circum stance, and a host of ot11er factors on national assessments of national interests. In "Cooperation under tJ1e Security Dilemma," Robe rt Jervis has e~:plain ed how the difFusion of offens ive military technology and strategies can increase rewards from defection and thereby reduce the prospects for coope ration. In "International Regimes, Transactions, and Chance: Embedded Liberalism in the Postwar Economic Order," John Ruggie has demonstrated how the diffus ion of Liberal economic ideas increased the perceived benefits of mutual economic openness over mutual closure (CC-DD), and diminished t11e perceived rewards from asymmetric defection relative to asymmetric cooperation (DC-CD). In "Firms and Tariff Regime Change, Timothy YlcKe0\\11 has shown how down turns in the busin ess cycle alter national tastes for protection and thereby decrease tl1e perceived benefits of mutual openn ess relative to mutual closure and increase the perceived rewards of asymmetric defection... . 5

A. Payoff Structure and Cooperation

How does payoff structure determine tJ1e significance of cooperation'? Ylore narrowly, when is cooperation, defined in terms of conscious polic} coordination, necessary to the realization of mutual benefits? For a mutual benefit to exist, actors must prefer mutual cooperation (CC) to mutual defection (DD). For coordinatjon to be necessary to the realization of the mutual benefit. actors must prefer unilateral defec tjon (DC) to unrequited cooperation (CD). These preference orderings are consistent with the familiar games of Prisoners' Dilemma. Stag Hunt, and Chicken. Indeed, tl1ese garnes have attracted a disproportionate share of scholarly attention precisely because cooperation is desirable but not automatic. In the e cases, the capacity of states to coope rate under anarchy. to bind tbemsekes to ll1utually beneficiaJ courses of action '<vithout resort to any ultimate central authority, is vital to the realization of a common good . ... In the class of aame s-inc ludin g Prisoners' Dilemma. Stag Hunt. and Chick en-where coop~ration is necessary to the realization of mutual benefits. how does payoff structure affect the likelihood and robustness of c:ooperation in

I Ill.

\..Vl'' <IUIIIU NS



. .11 b less likelv in Prisoners D ilt ll!rn t.m in St 1 tln. , ~ sit mllion ? Cooper.ttion "' le ' de.r each of thes<' g.u tr s in lOTIJ\I11<;.-: ag , . k -r d tand w w cons IJOn I hmt or Clue en. 0 un ers .d t1 . derhe tJ1eir names. . . f om -.vht 1 1e) \\i th the tllustran,e tones r . are susoected of a 11ajcJr <-rime. l'h . . 0 .1 0 . Two J Jnsoners r t Pn onet 1 em m cnon on onh' a minor 1. IJ:tr:.?:c. 1f ne1 tt .. d to secure com 1 )(:r authonttes possess evt ence li ht entence on the minor chargr CC). I f <l s ne al s. botl1 " 'ill draw a g ails the rat will uo free 1 DC and the suckc pru oner sque . a1 cl the other stonew o r pnsoner sque an . (CD). If both squeaL both wilJ dnm a mouerate ,,;n draw a very hea') s~nten~ 'erence ordeJing is: DC > CC > DD > CD I f pre . ( DD). Each Imsoners onlv~ one time, each pnsoner \\1tt I)C better . senten~e . . .. lav" off the pnsoners expect tal P _, matter what his partner chooses to do (DC > CC oJ 1' th tonew mg, no squeaung an s . f the rat payoff and fear of the su(ker payoff will d DD CD) The temptatiOn o . an > . . Dil n'lrnas toward mutual defection. Unfortu nateh if . 1 _ lay Pnsoners e /' d n\e smg e P h' oning they will draw a mode rate sentence on the both prisoners act on t IS reas . ' . . h'l ation could have led to a light se ntence on the minor maJOT charge w I e cooper , iJ . . 1 'd JI . ' DD) 1 sngle-play Prisoners D e mmas. JllUJ V ua ~, rational I charge (CC > n 1 actions roduce a colJectively suboptLmaJ outcome. . St~ Hunt: A group of hunters surround a stag. If all co~perate t? tlap the stag. all will eat well (CC). If one person defects to chase a passmg rahbtt, ~he stag will escape. The defector will eat ~ghtly (DC) and none of th~ ot~ ers ,~JJ eat ~t. all (CD ). If all chase rabbits, all will have some c~anc~ of catchmg a rahb1 t and eatmg li htl (DD). Each hunter's preference ordenng 1 CC > DC > DD > C D. The s: ;ut~ interest in plentiful venison (CC) relative to all other outcomes militates strongly against defec:tion. However, because ~ rabbit in the hand .(D C ) is better than a stag in the bush (CD), c..'Ooperation will be assured only 1f each hunter believes that all hunters will cooperate. In single-play Stag Hun t, the temptation to defec-t to protec-t against tl1e defection of others is balanced by the strong universal preference for stag over rabbit. Chicken: Two drivers race down the center of a road from opposite directions. If one swerves and the other does not, then the first wiJl suffer the stigma of being known as a chicken (CD) while the sec.'Ond will enjoy being known as a hero (DC). If neither swerves, both will suffer grievously in the ensuing collision (DD). H both swerve, damage to the reputation of each wiJJ be Hmited (CC). Each driver's preference ordering is: DC > CC > CD > DD. If each believes that the other will swerve, then each will be tempted to defect by continuing down the cent er of the road. Better to be a live hero than a live chicken. If both succumb to this temptation, however, defection wiU result in C'Ollision. The fear that the othe r driver may n_ot swerve de~Teases the appeal of continuing down the cente r of the road. In smgle-playChicken, the temptations of unilateral defection are balanced by fear of mutual defection.
In ~es that are not repeated, only ordinally defined preferences matter. Under smgle-play. eonclitions, of mteNal-1 evel payoffs in ordinally de fined categones ~:=~ot (m theol)_') affect the ~elihood of cooperation. In the illustra~ions ons of dominant strategies do not hinge on the magnitude of iliffer= o n g~ paJoffs. Yet the magnitude of differences between CC and DD een an CD c be large or small, if not precisely mea~urable, and can :an

increase or c. :t. Changes in the magnitude r clifr . I o 1 <::rences m t 1e value J '>latcd H on outconH ., l Ul n ''en<:c the prospects for coope t' I h . J I I ra ton t 1 rou~ two path F1r"t. c '<ngts 111 t H: \'a ue attached to out<:o1n .. tr r . . . 11 1 1 .1 1 es cc1n an 10rm sttuatton from one ordma ' l'.lll<:u <:ass of vame into anot11e f 1 . "C . under the ~ turit~ Dilemma,'? Hohert Jet'\is ders ?br dexahm~ edi. ffil~ 1 oo~eratto~ .. , . en e ow cu t Pn oncrs Dilemmas ma\ C\ .oh e m to less. chaHenginu Stag H un ts 1 .L uruns from mutu a1 f UJe . . , cooperation t CC mcrea.<>t relat1ve to t11e 11ains from expl0 ta t:i on I DC J. H e re1 cl ate f pm offs to trac.Utional concepts of rre 1 . ilie structure o : (' . 0 11 ns1-.e an d derensl\e domJ nance, and of fensi\'C:: and defensive dominance to techno1 a1 ancl doctnna1 ogtc shifts. Emst 11aas. -"larv Pat \\ 'illiams ' and Don Babai ha'e emphas1zecl t he . ... ' unportance o f cogmtlve congruence as a determinant of technolog a1 C:OOpera0 On. The f . .IC di ffus~on o c~mmon concepti~ns of the nature and effects of technology enhanced percetved gams from cooperatiOn and diminished perceiveu gains from defection. and may have transformed some P1isoners' DiJemmas into Harmonv.6 s.:c~nd, u.n.der .i ~erated conditions, t11e.magnitude of differen;es among payoffs w1 thz.n a g1 ~lass of ga~nes can be an Important determinant of coope ration . ven The mo~e s u bs ta~tt a~ the gams from mutual cooperation (CC-DD) and the less substantial the gams from unilateral defection (DC-CD), the ureater the likelihood of cooperation. In iterated situations, the magnitude of the dJference between CC ~nd anu. between .and CD in present and future rounds of play affects t11e likelihood of cooperation m the present. This point is developed at length in the section on llie shadow of the fu ture.



B. Strategies to Alter Payoff Structure

If payoff structure affects tl1e )jkelihood of cooperation, to what extent can states alter situations by mowfying payoff structures, and thereb\ increase the Ion a-term likehl1ood of cooperation? .\If any of the tangible and int~gible determin~ts of payoff structure, discussed at the outset of this section. are subject to willful modification through unilateral, bilateral, and multilateral strategies. In "Cooperation under the Security Dilemma,'' Robert Jet'\'i.s has offered specific suggestions for altering payoff structures through unilateral strategies. Procurement policy can affect the prospects for cooperation. If one superpower favors procure ment of defensive over offensive weapons, it can reduce its own gains from exploitation through swprise attack (DC) and reduce its adversary's fear of e:\p loitation (CD). Members of alliances have often resorted to the device of deplo)ing troops on troubled frontiers to increase the likelil10od of cooperation. A states use of troops as hostages is designed to diminish the payoff from its own defection- to reduce its gains from exploitation (DC)-and thereby render defensive defection by its partner less likely. Publicizing an agreement diminishes payoffs associated 'vith defection from the agreement, and thereby lessens gains from exploitation. These observations in international relations are paralleled by recent de,elopments in microeconomics. Oliver \ ViJliamson has identified unilateral and bilateral techniques used by firms to facilitate interfum cooperation by diminishing gains from e~loitation. He distinguishes between specific and nonspecific costs associated With adherence to agreements. Specific <.'Osts, such as specialized training, machine



1\Nf\ K\...1 11 "- -

b ecovetcd tn the e,e nt ol tlw I f ~do wn of an . c011Stntction, c.:annot e ' nt t'ncur hiuh speci fie.: co;.;t . , pu cliation r to<>Is. Ill cl . . aureeme .::1 nt \\'hen partteS to an c;all Finns can thus reu u n the n gains froo ;tl,!;f('en l . . . . .. ' l'lt.tll'lt'l)tS \\'lU ental 1Stl bSt <1ntr osses. uiJing deLiiC.:ale cl <lSS( t S. tha t Setv m {' acq comt e 1 1iqu e o as ~'f)loitation through the tee 11 . Nonsoecific assets, such as gctte ral- pu '11osc r I hostages to contmumg coOJJeratJOn. fa greements bre ak cown: f ins <:an bl lt rcduc: .. . . salvauea e t f .. e: . . .. , I e ~ . . . nu the use o non spe ctn c.: d<;S t Is, )LJt such t tucl--s and au1)Jan es, are . . . . their fear of bemg expIOtted b)' maxJll11Zl ~ . )alation b) servmg as hostages. - U ntlaterll . . . l . < assets cannot dunuusI1 gains from exp f coor)eration by re ducmg IJot l, t I1e t costs of . . strate<ries can tmp tove the prospec so exploitation (D C). Tl1e new ]tteratu . f OJn re Oll 0' . d (CD ) d tJ,e r . being ex.-plmte an . an old questiOn on the c.:os ts of unt.late ra) . 10 direct1 rruses . y interfinn cooperation : . nterna . . 0 Jeration 1 1 tional relations. . 0 . strate~es top~ omote eo lnilateral actions tJ1at limit one ,s gam s from exp loit ation In man y mstances, u . e's vulnerabtlity to expl.oitation bv oth ers . For ffect of increasmg on . al . / . mav l1ave t I1 e e d . . . s f'rom defection from libe r mtem at1onal eco. J t t coul 1 t <Tall'l JJm examp e, as a e . . th expansion of sectors of comparati ve advantage and ~ .. ai nomLc no rms by penmtnngf. e c.ent sectors. Because a spe cr tze d econom . .d . effi 1 y is a b tting liqw ation o m Ypernu . ti. ~'economic cooperation, tllis strategy would uny u esti onablv hostage to mtema onat . al . th dibJty f the nation. commitment to tbe1 tsm. I t . 1 1as the s 1 ,1 so 1 mcrease e ere I I o . . .. . . . easmg the nations vulner abdity to p1 ote ct10n l)y otl1e1s. In effect l1owever, of mer . the tr~ps-as-hostage example, the governme nt that stat10n s ti:o ops m ~y p_ro m_ote . ng cooperatIon b)' dimm LSh1 an ally's fear of abandonm ent , bu t m so domg 1t rruses its own fears of exploitation by the ally. . . . . Unilateral strategies do not exhaust the ran ge of option s tha t sta tes may use to alter payoff structures. Bilateral stra teg ies -m ost signifi~n~ly sti~at~ gi_es of issue Linkage-can be used to alter payoff structu res by com bm mg diss1mtlar games. Because resort to issue linkage generally assumes iteration, ana lysis of bow issue linkage can be used to alter payoffs is presented in the section on th e shadow of the future. Furthermore, bilateral "instructional " strategies can aj m at alte rin g another cou ntry's understanding of cause-and-effect relationships, and res ult in alte red perceptions of interest. For example, American negotiators in SALT I sou ght to instruct their Soviet < nterparts on the logic of mutual assured des tru ctio 8 :ou n. Multilateral strategies, centering on the formation of intemational regimes, can be used_ to alte~ payoff structures in two ways. First, norms gen era ted by reg imes ~y be mtemalized by states, and thereby alte r payoff structure. Sec ond, info rmation generated by regi.mes may alter states' understanding of the ir inte res ts. As Emst Haas argues, new regunes may gather and distribute information tha t can highlight cause-and-effect relationships not previously understood. Changing per cep tions of means-ends hierarchies can tu . . , m m, res uJt m changmg perceptions of interest.1:1





influenc.:e de H '.' ' n the P~(sent is fu nda me ntal to the emergence of coo pN alion among cgotP.;b. !\" ll~~ prevtots sc~ tion s~tgg<:.sts. stat es confronting strateg ic \itn ations ~hat re sul lhl l' sm.gl_-play Pnsoners Dil emma and , to a lesser extent e , sin t;kpl~y Stag ''.'~nt _and _Cinc~en , ar~ _consta~t ly tempte d by imll'Wdiatc gains fro m unilate ral de lec_lwr:- ,u~cl fea rf~J of ~ ~m ediate losses from unrequ ited coo peratio n. How doe ~ cc~ntmu~ ng m~~ rachon affe~t prospec.:ts for cooperation? Th e ar ~m cnt proceeds_ l_ou: s_t,lges . ~ tr~t , wh y do 1terated conditions imp rove the prospe m cts [()r cooper ation ~n l .nson ~rs D~l emma and Stag Hu nt while diminishing the prospects 111 Ch, ~ke for cooper~lton n Second , how do strategies of reciproci ty imp rove the prospects for ~oop~r at1~ n under iterated conditions? Third, why does the effec.:tive ness of rectprocrty hmgc on conditions of pla y-the abilitv of acto rs to distinguish reli ably between cooperation and defection by other; and to res pond in kind? FoUJt h. through what stra tegies can states improve conditions of play and length en the shadow of the future? Be fo re turru ng to these questions, c.:onsider the attributes of iterated situatio ns. First, states mu st expect to continue dealing with each otl1 er. This con di tion is, in practic.:e, not pmticularly restrictive . With the possible exception of glo bal the rmonuclear war, inte rnational politics is charac terized by the exp ectation of fu ture interaction. Second, payoff structures must not change sub stantially ove r tim e. In other words , eac h round of play should not alter the structure of the gam e in the future. This condition is, in practice, quite res trictive. For example, states consid ering sUipris e attack when offe nse is dominant are in a situ ation that has ma ny of the characteristics of a single-play game: Attack alte rs options and payoffs in future rounds of interaction . Conversely, nations consid erin g increases or dec reases in their military budgets are in a situation that has many of the charac teri stics of an iterated gam e: Spending options and associated marginal increases or dec reases in military stre ngth are likely to remain fairly stable ove r future rounds of inte raction. In inte rnational monetary affairs, governments considering or fearin g dev alu ation under a gold-exchange standard are in a situation that has many of the cha rac teristics of a single-play game: Devaluation may diminish the value of anothe r state's foreign currency reserves on a one-time basis, while reductions in holdin gs or reserves would diminis h possible losses on a one-tim e basis. Conve rsely, govern me nts considering inte rvention und er a floating system with minimal reserves are in a situation that has many of the characteristics of an iterated game: De pre ciation or appreciation of a cur rency would not produce substantial one-time loss es or gains. Third , the size of the discount rate applied to the future affects the iterati\ en ess of games. If a government places little value on future payoffs , its situation has many of the characteristics of a single-play game. If it places a high \alue on future payoffs, its situ atio n may have many of the characteristics of an iterated game. Fo r example. political lea der s in the ir fina l term are likelv to discount the future mo re substan" tiaJiy than political lea ders running for, or ce1tain of, reelection.

A. The Shadow of the Future and Cooperation
How does the shadow of the fut ure affect the likeliliood of cooperatio n? Under single-play c.:onditions ,~~,~thout a sovereign , adl1erence to agreement s is often

The distincti be on tween cases in h. h 'mil . unlikely to be repeated cl . w 1.c SI ar transactions am ong part1es aJe
an cases m which the expectation of fut ure inte ractjon can


t'tu\ J &

IU ~ t u~ -

. s Dilemma. Eacb p11 t:r t11r is better r . . t! . . . . ~t.,;otnl Con~ader the s111g e -IJlcW Pnsoner. to squeal. In t I1c ,1I>'-lll( .c. u {' con tm,,o I . d ,.5 1n. ~ . . tncr c ect e 1 . .. . . .OYlinl'1 whether or not 1us P<0 . tl domimll1t strategy. Becc\u-;t tl ,e pn:-;onlllg . qu"'" c- .t l e ers . 't"ntclion defection wotuC eme rge <L~ 1 nrorcement of an agrc'enH nt to coopcrat . r lh " ( ' . Jl . . "et'ther rum to a central au thonty..ore 'eter prese nt de reel ton. cooperation,, .e . t can '' . . . f etahation o u l .," 11 nor relv on the anttc tpatiOn o r di . lf the prisoners expect to w plnc:ecl in sitni. un d . 1 plav con tions. . . . . be w 1tikel) er smg e- ' t c0 r coop eration 1mpt.ove E'1)c,ll mental e,,; " Il . . \ar situa6ons in the fu ture, .the prospec sners' Dile mma the JnClcIencc' o r cooperatiodp . dence suggests that under.rterat~b ~~~of centralized authority, tacit agreeme nts . J .,.. rises substantiaJl) to Even 111 tlle a se a are frequent)' reachec1 . I mar.ntamed alli. .m< . . . I cooperate throug1 mu tual stone'vc nootential defector compares t1 nnm ediat 1e . ' Dilemma, a P . e Under iterated Pnsoners 'bl acrifice of future gatns that mny result frorn . th the poss1 e s gain from squeal rng Wl H t each hunter is tempted to defec:t . order to m squealing. In singl~-play Stag .b~~or'defection by othe rs. A reputatio n rcliabildefend himself agamst tl.1e pos~ ~stile likelihood of defection. If the hunters are a ity, for resisting tempdtation. ret ul nt together again , the imm edia te gains from unit 0 1I}> an expect o 1u permanen gr ' . t equited cooperation must be halanc:ecl against the t al defection re1 . , 1 en . . . ed ative o unrn the a ti' future. In both Pnsoners Dil em ma and Stag cost of cLmtmsh c:oopera on l f . . c . . tJ esent decreases the likeliJ100d 0 coop erat1011 . tJle futu re Hunt derection m le pr . m . th r t rati'on improves the prospects for coope rabon. In Chrcken In bot1 ere1ore, 1 e 1, . . .. . may d ase the prospects for cooperation. Under smg le-pl ay conditions' iteration ecre .. dle temptation of unilateral defection is ~alan~ed by the of the colh sron that fo~lows from mutual defection. How does 1teration affec t this balance? If the game 1 s repeated indefinitely, then each driver may refrain from in the preser~t to coerce the other drive r into swerving in the future. Each drive r may seek to acqUire a reputation for not swerving to cause the other driver to swerve. In itera ted Chicken, one drive r:~ defection in the present may decrease the likeilllood of th e othe r driver's defection in the future.




B. Strategies of Reciprocity and Conditions of Play

?f conti ~uing interaction has varying effects on the likelihood of coope ration in the 1 Uust_rations abov~ , an iterated environme nt perm its resort to shat egies of recipr~lty t~at may rmprove the prospects of cooperation in Chic ke n as well as in Pns~ners Dilemma and Stag Hunt. Robert Axelrod argues that strategies of red~rocrkhave the effect of promoting cooperation by estabLishing a direct connecotnr Til -IOr-ttwteen andi ac.tor's present behavior and anticipate d futu re bene fits. a , or con 'tional coo ti . tion by h th fu pera on, can mcrease the likeliliood ofJ oint cooperas apmg e ture conseq f In iterated p . , . uences o present coop eration or de fection. future consequen~:ne;s Dilemma and Stag Hunt, reciprocity unde rscores tile sented abov... ..L . 0 p~esent cooperation and defection . The argu men t pre.--u w Iteration enhan th game s-rests on the . ces e prospects for coop erati on in t l1ese I: L_ .I assumption that det ti th wwoo of COOperation in the fu 1 ec. on m e present will decr ease the l'kematching stonPU~<~Il:~.... .th ture. Adoption of an implicit or explicit strategy of --~ oouu"b Wt stonewalling alin . . g ' sque g With squealing, rabb1t chasll1
It is at this juncture that strategy enters the explanatio n. Alth ough the expectation

effects of n.-> l.tlJO'M; c.onsJuerations on the prospects fo r cooperation. Recall that in iterated Cl ne~'- n, <- atlt ~riv~ r may ~e:f'rain front swerving in the present to coerce t}1 e other dn\'('r !~le~ swc1v1ng m the futm e. Adoption of an implicit or e>qJiicit strategy of tit- for~tat 1n tterated.gaJ~es of Chi<.:ken alters the failure stream of benefits associated wJth pn:\t 'lll defection . If a strate6'Y of reciproci ty is credible, then the mutual losses <L<isot:iate~ with fuhtre collisions can encourage present swer\ing. In all thrc>e games. a prom1se to respond to present cooperation \vith futurc coop eration and a th reat to respo nd to present defection with future defection can impro,e the prospects for cooperation. The effeetiveness of strategies of reeiprocity hinges on conditions of play- the abili ty of actors to distinguish reliably between cooperation and defection by othe rs and to respond in kind. In th e illustrations provided above, the meaning of "defect" and "cooperate" is unambiguous. Dichotomous choi ces- betw een squeal and stonewall. chase the rabbit or captu re the stag, continue down the road or swe rve- limi t the likelihood of misp erception. Further. the actions of all are transpare nt. Given the definitions of the situations , prisoners, hunters, and drive rs can reliably dete ct defection and cooperation by other acto rs. F inally, the definition of the actors eliminates the poss ibility of control problems. Unitary prisoners , hunters, and dJive rs do not suffer from facti onal, organizational. or bureaucratic dysfunctions that migh t hinder implementation of strategies of reciprocity. In international relations , conditions of play c:an limit the effectiveness of reciprocity. The definition of cooperation and defection may be ambiguous. For example, the Soviet Union and the United States hold to markedly diffe rent definitions of "defection" from the term s of detente as prese nted in the Basic Principles Agre eme nt;J 1 the Eu ropean Community and th e United States differ over whether dom estic sectoral policies complise indirect e>;port subsidies. Furt her, actions may not be transparent. For example, gove rnme nts may not be able to detect one another's violations of arms control agre eme nts or indirect expmt subsidies. If defection cannot be reliably detected , the effect of present cooperation on poss ible future reprisals will erode. Together, ambiguo us definitions and a lack of transparency can limit the ability of states to reco gnize cooperation and defection b, othe rs. Because reciprocity requires flexibility, control is as important as recognition. Internal factional , organizational , and bureaucratic dysfunctions may limit the ability of nations to implement tit-for-tat strategies. It may be easier to sell one unvarying line of policy than to sell a strategy of shifting between Lines of policy in response to the actions of othe rs. For example, arn1s suppHers and defense planners tend to resist the cancellation of weapons systems even if the cancellation is a response to the actions of a rival. Imp industries tend to resist the removal ofba.niers to imports, even if trade liberalization is in response to libemlization by another state . At times, national decision makers may be unable to impleme nt strategies of reciprocity. On othe r occasions, they must inves t heavily in selling reciprocity. For these reasons . national decis ion makers may display a bias against conditional strategies: The domestic costs of pursuing such strategies may partially offset the value of the discounted sheam of future benefits that conditional policies are e>:pected to )ield . ...

with rab~Jit c 1 ~n " 11 d cc:o.lwrative hJnling "'ith <.:oopcrativc !tuntin~ \aJicl,ttcs the assumpt1011. fn 1t. r tl r,d ClJ~c:ken , a slrat eey of reciprocity can off<.et the pen er~<:






c. Strategies to Improve Recog

the Shadow of the Fu ture

nition an d Len gthen

. peration hv creating l.t,nrable c:ondi ts oromote coo r To \,l''\t extent can goven1men . the sIladow of the futu., ? 1~I le I' atn re on inte re. tln < .. rtion , of play and by lengtherung . es for creating favornhle <:onditton s of play. rr national regimes ouers 5everal techmqudefinitional am b'tgUJ'tY. Th e ''Cl~' ac.:t f' _ . limit o c:larope rative and unc:oop<ralJ\'e behavior E :\'l)licit codification of not rns can 6 . ""' d f de mng eo . .. ifvin<r standards of con uct, 0 , . t es of reciproctty. l'ur ther. pro,tsJons l'or . . rr . . . . can pemut mo re e11ec ti've resmt to stra egifor verincauon 111 a1.m;;. c..:on tt.o1 . agreele mecharusms .1 surveillan ce- for _exa ~p on on the nature and effects o f uo~11 Csllc: sectoral ments or for shanng tnformati I . . parency. n P1actice ' the goal of enhan c:mg recognition policies--can mc rease trans tiations under anarch y. capabilities is often central to ~eg? . aJ microeconomic lite ratures offer several f . a1 d msrrtu 0 on< 1 The game- th eore 1C I >rac ter of sihtat'ions. Th on1as Sl:helling and . . u the tteraove c 1.... approaches to mcreasmo . f d nposition over tim e to lengt1 t1 sha 1en 1e dow t tactics o ecot Robert AxeIro d sugges taoon to defect in a deal promising thirh bill~ F0 mple t1 temp of the future. - ~ ~x.a . ' I 1e foil may be reduced if the deal is sliced up ' into a lion dollars for a bJIJtodn bdarli:e s Coo peration in arm s reduction or in territorial f . ents an e ven es. s~nes o pa) m . be difficult if the reduction or dis engage me nt must be achieved disengagement mayd . .l: . If a re uction or wsengagement can be sliced up into increments, the 111 one Jum p. . b endered more tractable. Finally, strategies of issue may e r . Problem of cooperation It . .. .1 voff structures and to . t]ect e Iem ents o f Iterativemte linkage can be useu to a er pa; . . 1 1 t ti'ons Relations among states are rar e Iy limite d to one ness mto smg e-p ay SI ua . e of .nw g impmtance \i\T natwn s con front a smgle-p Jay .l=n hen single-p]ay 1ssu ovet _ . . e, esent defiection mav be deterre d by th 1.eats of 1etaliat. on game on one Issu pr JOn other iterated issues. In international monetary affrurs , for mstanc:e, a go~e1 nment fearing one-time reserve losses if another s~at~ deval~ es its cur ren :y may link de,~l uation to an iterated trade game. By establislung a dire ct connection between p_resent behavior in a single-play game and future benefits in an iterate d game, tactt or explicit cross-issue linkage can lengthen the shadow of the future .. ..

Prisone rs Di ' 111 11 1a giws way to the pessilnism of analyses of' cooperation in thC' provision o_f Jl 11 !,lic- ~oods. A~J~lic:ati~>ns of Olsonian th~ory to prohlem<; ranging from cartdv.a Jl1 to the.; proVlSJOn ol pub lic: goods in alli anc:es undcrsc:nr<" the si<rnifi cance of "free-riding" as an impedime nt to cooperation. 13 ln inte rnation al rel; tions, the nu1nbc rs prohl~~ has bee n c:entral to hvo debates. The longst anding controversy on r the stab tltty of bipolar versus multipolar svstcms reduce s to a debate ove r the imp act of the number of significant act~rs on interna tional conflict. l l A more recent <:ontroversy. between proponents of the theory of hegemonic stabiJity and advocates of inte rnational regimes, reduces to a deb ate over the efl'ects of large numbers on the rob ustn ess of cooperation. l5

A. Number of Players and Cooperation

How do numbers affect the likelihood of cooperation? Th ere are at least thre e impor16 tant channels of influence. First, cooperation requires recognition of opportu nities for the advancement of mutual inte rests, as well as policy coordination onc e these opp01tunities have been ide ntined. As the number of players increases, tran sactions and inform ation costs rise . In simple terms. the complexity of ~-person sitl tations militates against ide ntification and realization of common interests. Avoidin g nuclear war dming the Cuban missile crisis called for cooperation by the Soviet Un ion and the United States. The transaction and information c.-osts in this particularly harrowing crisis, though substan tial , did not preclude cooperation. By contrast, the problem of identifying significant actors, defining inte rests. and negotiating agreem ents that em bodied mutual inte rests in theN-actor case of 1914 was far more difficul t. These secondary costs associated vvith attaining cooperative outcomes in :\-acto r cases erode the difference between CC and DD . More significantly. the inbinsic difficulty of anticipating tl1e behavior of other players and of weighing the value of the futme goes up with the number of players. The complexity of solving N-person gam es, eve n in the purely ded uctive sense, has stunted the development of fonnal wo rk on the problem. This complexity is even greate r in real situations, and operate s against multilateral cooperation. Second, as the number of players increases , the likelihood of autono mous defection and of recognition and control proble ms increases . CooperatiYe behavior rests on calculations of expected uti lity -m erging discount rates, pay ofT structures, and anticipated behavior of other players. Discou nt rates and approa ches to calculation are likely to vary across actors , and the prospects fo r mutual cooperation may decline as the number of players and probable heterogeneity of actors inc reases . Th e chances of including a state that discoun ts the future hea ,ily , that is too weak (domestically) to detect, react, or implement a strategy of recipro city, that cannot dis tinguish reliably between cooperation and defection by other states, or tha t depatts fro m even minimal stan dards of rationality increase with the numbe r of states in a game. For example, many pessimistic analyses of the conseq uence of nuclear pro life ration foc us on how breakdowns of deterrence may becom e more ukely as the number of countries with nuclear weapons increases. Third, as the number of players increases, the feasibility of sanctioni ng def ectors diminishes. Strategies of reciprocity become more difficult to implement withou t


Up to now, I have discussed the effects of payoff stllJc:tu re and the shadow of Ule future on the prospec.-ts of cooperation in terms of two-person situations. Wh at hap~ns to ~he prospects for cooperation ac; the number of signifi can t actors rise s? In thts s~on, I explain why the prospects for cooperation dim inish as the number of players mcreases; examine the function of inte rnational regimes as a respon se to the toblems cr~ated by Iru:ge n.umbers; and offe r strategies to imp rove the pro spects r cooperntion by altenng Situations to diminish the number of signiflcan t play ers. OlsoT?ethnumbe~ problem is central to many areas of the social sciences. Mancu~ n s eory OJ collective acti c Dile Th . . on IO es on N-person ver sions of- PnsOJ1ers cus mma. e optimiSm of ou . 1 -1=. wscuss1 .ons r ear 1er of cooperation under 1terated




., . . . n tmposc. costs on aJI Parties m an. ~"' -per r hJty of sm<:tiOJLIIlg de le< lor'i <:rc:atcs the, ' '' . . d .-d The mreas I ' . po,. re<:rprocrty 1 un ermtnt fwe increase the llltn rhtt nl adors in t1 1e it . 'I' fr w ng. What 1 t ()? Confession by any one oltiH'm c:ould lcatlt 1 lappens I . cr. stbt rty o ,rce-n f o 1 mma rom 2 o 2 atecl Pnsoners J e . h ge therefore the thr<at to rdaliatc a1,.,; o c ar ,.,""ll!\l he c:onVJctron Of a11 on t!Je maJor oon in ,the future 'will impm< tWits on all prist t .th d r~ l f, . the present W1 erec.: me e cct10n 111 J aJ 'ef'ecoon in subsec1 uent round\. f ur exam[Jie llntll: . . d Id lead to w 1 es r u 10 . . ' r ers, an cou f. lu taJiation against one mcrnber ol tlw al hant:e wa') llr tl 1g14 system o a ances. re " le. r a1 ti. g"i ust aJI. In :\-person games, a stratcgy of conditional e(IUJVaJent 0 ret Ja on a "' 1 . h tJ ffi t of spreading ratl1er tJ1an conlam111g, de f'cction.
defection can ave 1e e cc '

111 two-person ga11H " I t1t !t ; -tal work\ w<.:ll 11 e of c:ooperauon. tri~enng a <..'0 ap . r used on onl) or1 ot "' p..t1.\ Jf defet c f defectiOn are ,oc . . . ron bcc:au'i<.' tl1C' costs o . ,1 son game. howcvcJ, tlu po\ 1 r of st rater!i( f . .,.


J ,., ,., (

fs l

B. Strategies of Institutionalization and Decomposition

. be. ol["layers, what stratemes can states use to increase Lhe likcJvcn <l 1 arge num 1 ' ~::r . . C lihood of cooperation? Hcgimc creation can increa~e the lkel1hood of ~oo.p~ration in ,. arnes first conventions provide ruJes of thumb that can dmm1 transsh 1... -rwrson g . , . . ac.:tion and information c-osts. Sewnd, collective enforcement mechamsms both decrease the likelihood of autonomous defection and permit selective punishmenl of violators of norms. T11ese two func.'tions of intemationaJ regimes directly address prohlems created by large numbers of players. For example.' !ap~n and the members of NATO profess a mutual interest in limiting Hows of md. tanly usefu l goods and 1 technology to the Soviet Union. Obviously, all suppliers of militarily useful goods and technology must cooperate to deny the Soviet Union access to such items. Although governments differ in their assessment of the military val ue of some goods and technologies, there is consenstt~ on a rather lengthy D of prohibited items. By faciJitalst ing agreement of the prohibited list, the Coordinating Committee on the 0Jn!mltative Croup of NATO (CoCom ) provides a relatively d ear definition of what exports would cxmstihJte defection. By defining the scope of de fection, the CoCorn list forestalls the necessity of retaliation against nations that sh ip technoJogy or goods that do not fall within tJ1e~ consensual definition of defection. CeneraiJy, cooperation is a prerequisite of regime c.:reation. The creation of mles of thumb and mechanisms of collective enforcement and tJ1e maintenance and adminL<;tration of regimes can ~em~d an extraordinary degree of cooperation. This problem may limit the range of Situations susooptihle to modification through regimist strategies. What strategies can reduce the number of significant players in a game and thereby render cooperation more likely? When governments are unable to cooperate on a gl~bal scale. they often turn to diseriminatory strategies to encourage bilateral or r~nnaJ cooperatiorl ""ac..,; . . of. decomposttion across actors .. . --oJ i ucs can, at h. mcs 1mprove the prospe<.:ts ~or cooperation. Both the possibilities and the limits o1strat, e. gaes to .reduce the number ofplayers are eVJ'dent the discussions that f.o11ow. F'rst ~ 1 . m ~educt:ions m the numbe f ' I 'tude f . r o actors can usuaiJy be purchased at the expense of llC magm t:lwl th ,o gams from coopcration. The benefits of regional openness are srnaller . & nomicale gamffis om global openness. A bilateral dearing arrangement is less ece>l ye CJent than a m11ltilatc a) 1 r c eanng arrangement. Strategies to reduce the

.t 1111111 llC'r of pi t\'C'r\ in a "arn~ v<.-n<.:rally ] flllll!>rs t h(;; gai r 11 . tlwy' lllf;ro-t~t tlw lik< liltf,Od and robwt . f m rom <;oopcratron while ' nC\\ 0 COOp<.:rati S J red11<"C' t!J, 111Jrrd,cr of players gcn<:rtll)' . b on. C <:On< . .,trategies to .. ' unpos<: su stanti l . t h..1 ThC'SC'C>-tllnailtusm~y motivatetl ' cl . . a CC>ss on t nu parties. ur parttcs to undermine th , t .t d f c:ntt ioll or 111ay '>trvc as art impct 11s for , tl .. 1 e ltnr C:: area o coop1 nru party to cnla . ' th tion. Jn tiiC' IH30s, for example, whoksalc res0 rt t , t . . r~e e zone of <.;oopera , l c:nalion of cx<.:lnsivc :r.onc~ of cotn n . o uL'><.:n mtnatory t ra0 [X>li c1 Jacililatct mg es 1open ne~!! Whf'n eo f t 1b . ' erc.1a a shrinkrng 111arket <;),art, Cno:at Britain .1dopt cl , I l'b 1 er and more n ron e< Y . aJ 1 . ' c a ess discriminatory colltltH'rcr po r<.:y tn order to sc<.:ur(: prer.. tial . '~ren access to tls 1 und('n llim: prc.:fercntial agrcerncnt') between othe . tri th empl~e anc to 1 . . r coun I1an. cIJJrltntS1 ' t1e Lin1tcd States adopted a m 1 ts. As cl 1e Amenc:an markets . . . 1 lCu, 1 . ore J1 . alan more dlscnmma:x:r tury <:01111 n<rc1 po 1cy to mcrease.: it\ access to e rt . k a . . d th _ . xpo mar ets. It 1s not possible howev<.:r, to.rt. uce e ''umber of players in all .ituatons. ["or exarnp1 cor npare the ' . . 1 [;' , e, example o r lnn1ted commerc.:ial openness with the examp1 of a 1 ' strate!llc: , e 1m1teu . emhargo. fn rc<.lllcethcnu mber ofactorsJnatrade"ar, mar ket a<.;cess can s1mp1 ~ y ut: v offt:red to.only one .<.;Ountry and withheld from others By eo t t d r ti b 1 ~ . . .. . . . n ras , e1 on y on y ec one supplu.:r <.:an pcrmrt the target of a strategic embargo to obtain a c.:riticaJ technology. These pro?lems may li mit the range of situations susceptible to moruncation througll strategtcs that reduce the numbE:r or players in games.
al .

l. F'or examples, see Robert Jcrvis, "Cooperation under tbe Security Dilemma.," World Politics 30 (January I978), pp. 167- 214; Olivcr E. \Villiamson ...Credible Commitments: Usiug J lostages to Support Exchange," American Economic Rf!1.iieu: (September 19'3.'3},

pp. 5J 9-40; John Cerard Ruggie, "International Regime , Transactions. and Change: Embedded Liberalism in tJte Po twar Economic Order." in tephcn D. Krasner. ed .. ltltenwtion.(l[ Regimes (Ithaca, N.Y.: Comell University Press. 1 983). 2. For orthodox game-theoretic analysrs of the importance of itemtion, see R. Duncan Luce and ll oward Raiffa, Games and Decisions (New Y Wiley. 1957), Appendix 8, ork: and Oavid M. Kreps, PauJ Milgram, Jo!tn noberts, and nobert Wilson, "Rational Cooperation iu Finitely-Repeated PrisoutJr!> Dilemma.' jollnlal of Ecor wmic Thecry 27 (August 1982), pp. 245-52. For the rC'sult!l of laboratory e>.~rimen~, ce Robert RaJlow, "An Experimental Study of Cooperation in the Prisoners' Dilemma Game,~ journal of Conflict Resolution 9 (June J96S). pp. 221-27. On the importance of indefinite iteraliou to I he emergence of c:oopemtion in business t r.tnsactions. see Robert Telsor, ''A Tht'O'Y of Self-Enforcing Agrcc1 ncnts.~ j ournal of Husin e.>s 53 (January 1980). pp. 27-44. 3. Sec HoiJI'rt 0. Keohanc, After Jlegc:mony: Cooperation and Discord in tlte WorUf Politic:all:cmwmy (Princeton. N.J.: Princeton University Press. l984 ), and Krnsuer(ln. 1). 4. S(:( Juhn A. C. Convbeare, "Jnt<rnational Organization and the Theory of Property Bights," J11 ter1wticmdl Orga 11i:::afion 34 (Summer 1980). pp. 307-34, and Kenneth A. Oyc, .. Brlicf Systems. Bargaining, and Breakdow11: lntemational Political Economy 1929- 1936,'' Ph. D. diss. (llarvard Uuiver!lity, L983). clulp. 3. 5. Sec Jcrvio:; (fn. J ); Huggie (fn . l}; 'Jin 1othy J. McK eown, "Firms and Tariff Regime Ch:mgc: Explaining the Demand f(>r Protection ... World Politia 36 (January 1984),



I l ."' . f 1) n, s(>t.lhl' cnnduding ections of Jcn1S( n. . I I 11 I if. , 1 'r 0 .. . . ' d B 1. Scientists and \Vorlc 0 w W\ l ecltnica[J<J 6. ll aas. \\ !Ihams. Io a t. l.S (Berkelc\': l' niH'I sil) of ( 'ali l'ornia Press. Hn~l)IV/. ) 101

. r fprefC.nnc<.;' ')) >tS-3~ On the C'lfects o f am lJtgut Y 0

tl e PHl\()C't:l !) of t:ooperatio .

:lll T{!/1111~:(/ !;ll, ed<~e in lntcrllatwruJ


, .

7. \\'illiamson (fn. 1). Tl1 St0 ry of SAL'/ I ~I'\' York: Holt H' . ee John ~ewhouse, Cold Dawn: e 'ne1 ~ l<trt

Kant. Liberal Legacies, and Foreign Affairs


\\lnston, J973). C 9. See Haas, "\.Vords an

'-l t
r 01

Y Or Who Said \ \'hat to\\ hom About ou


Hcgime " .



Krasner (fn. l ). d Alb rt Chammah Prisoners' Dilcm11w (Ann Arbor: Univ . 10 s A atol Rapoport an e ' l .re: .IJ ers1~ ee . n . , ) and subsequent essays in Jmmw o; .cntp tct llesolut;011 of~ IC hlgan I ress. J965 . M o i1JU u.s.-socuJt R. l ry.. p n >J l '1)1.~. '~ .n sts p rcuen(' . . 1W If . 11. See Alexander L. George, ' ana>:> o 1 on . 1983). . (Bou IdeJ. Colo... Westview, rfi' t (Cambridge. Mass.: Hanard Un . if C lv<"rstty Press, 19601 12. Schelling, Strt1tegy o on .tc '' PP 43-46 I J .,.1 LocJiC of Collective Action: Pu./;lic Goods and tlte Theorry 0r 13 See Mancur 0 son, r.. 1 ze b 1965) M ~ b .d M . Harvard University Press, , an d . ancm Olson and Groups (Cam n ge. ass .. 11. . 11 , . , .r . . "A11 Economic Theorv or A . lt\JlCCS. n e l.iW W 0; con o ml c.~ 0!1([ Richard Zeckh auser, ' 1 d . . ( 6) PP :266-79. for a recent E' cgant u r111na1y an extension of Stat1sttcs 48 August 196 ' . " . dilemmas or collective actJOI1. see J\U SSe11 ll arc] Collective Ill, the large Itterature on . . 989 ) . (Bal.,; e Jol1ns Hopkins University Press. 1 - ActiOn umor Id .. D l f 93 (S 14 See Kenneth N. Waltz, ''The Stability of a Blpolar \>\:or . d haecF liS '' } u mme 964), a lr~ and Richard N. Rosecrance, "BipoJaritv uJtipolantv. an t e utu re, OII17W o1 Con.,, \11 _ , ict Resolutio11 (September 1966), PP 314-2 ' . . fl 15 On hegemon)'. see Robert Gilpin U S Pou;er and the M11ltwatwnal Corporation . tl l . ( v k B Books 1975) pp. 2.58-59. On duopo1 see Ttmo 1)' 1 y. E vlcKeown, ;'liew , or . as re er .. 1 1 111 urope, ntematrona1 ''Hegemomc Stab'''ty Tlteory and 19th-Centurv Tarin Leve s .m _ Organization 37 (Winter 1983), pp. /3- 91. . 16. See Keohane (fn. 3), chap. 6, for extensions of these pomts.

re . .

What difference do liberal principles and institutions make to the conduct of the forei~n aff:~rs of ~iberal states? A thicket of conflicting judgments suggests that the legacres of hberalrsm have:: not been clearly appreciated. For many citizens of liberal states, l.iheral principles and institutions have so fuiJy absorbed domestic politics that tl1eir inAuence on foreign af-Tairs tends to be either overlooked altogether or, when perceived, exaggerated. Liberalism becomes either unselfconsciouslv patriotic or inherently "peace-loving: For many scholars and diplomats, tl1e r~lations among independent states appear to differ so significantly from domestic politics that influences of liberal principles and domestic liberal institutions are denied or denigrated. They judge that international relations are governed by perceptions of national security and the balance of power; liberal principles and institutions. when they do intrude, confuse and disrupt the pursuit of balance-of-power politics. Although liberalism is misinterpreted from both these points of view, a crucial aspect of the liberal legacy is captured by each. Liberalism is a distinct ideology and set of institutions that has shaped the perceptions of and capacities for foreign relations of political societies that range from social welfare or social democratic to la.issez faire. It defines much of the content of the liberal patriot's nationalism. Liberalism does appear to disrupt the pursuit of balance-of-power politics. Thus its foreign relations cannot be adequately explained (or prescribed) by a sole reliance on the balance of power. But liberalism is not inherently "peace-loving"; nor is it consistently restrained or peaceful in intent. FUJthermore. liberal practice may reduce the probability that states will successfully exercise the consistent restraint and peaceful intentions that a world peace may well require in the nuclear age. Yet the peaceful intent and restraint that liberalism does manifest in limjted aspects of its foreign affairs announces the possibility of a world peace this side of the gr~ve or of world conquest. It has strengthened the prospects for a world peace estabhshed by the steady expansion of a separate peace among liberal societies.

From Michael \V. DO)'Ie, "Kant, Liberal Legacies, and Foreign Affairs.. P:m L" Philosophy & .Pu~lic Affairs, Vol. J 2, No. 3 (Summer 1983), PP 205-232. Reprinted br penn~"SJOn of Blackwell Publishmg.


ent 1e LihlraJisrn ha - h<(~n re1 ifi d WJ'tl1 essential j)rin<:il,l , - !. im[)Ortance or the an . . . . 1 'd I.1-et.!t1 o f t1 . d'''1 u al AlJove all thLS ts a L>elr el 111 c 11 lporta ne;e of 1110ra) om 1e 111 I. .ree d f t I1c ngI t to be treated and a dutv to treat oL 1 u~ t t lJ<.:al su biec.:ts . l H I om o ' . 1 pnnclpJe has generated 1 1 t s and 111 nJ . and not a s obiec:ts or means onlv. T 1JS ., . s tutions . J tl eefold set of rights forms the foundanon of liberal .b A_~ommaUttms ~o rt tro n re"e01do1;,, from -arbitra ry autholi t)', ol'lell called "negative r'rsern. L1 e nu1Sll1 C< 11 [ 1f <::.. w I1rch meI udes freedom of conscience, a ree pre s am ree speech, equalty . . d om. h gilt to hold and therefo re to exc ange, property with 1 und er t he Iaw, and th e n . out r f arb'JtraJY serzu re. LJ'beralism also calls for those ng1 necessary to prot . , ear o ' 1ls " .. . ect 1 pos1trve freedorns ., d . t the capacity and opport unity for freedom , t H.: an hpro~alo e d Sue socr an economc nghts as equalitv of opportunitv . in educa tion anti rig!1 1 ' J. . . ' . to health care and employment, necessa?' for_ ef fecti~'e self -expresst.on an~ particiation, are thus among liberal rights. A tlnrd hberal nght, cle mocrati<; participation P repres en t auon, Is necessarv to guaran tee the other two. To ensure that moral! ..: or .1 . f . l . Y autono mous individ uals remain free m those areas o so~1a achon_ ':"here public: authoritv is needed , public legislation has to express the w1ll of the Citizens making ' laws for tl1eir own community. These three sets of rights, taken togeth er, seem to meet th e challenge that Kant identified:





bv indi''" l d .tr 'Jllisition (fc>r txa1nplc h 1 1 ) h . l rtilit\. "J ill t M:h rdcs state socialism l>',. s~tatae)Or ?tr l y socbJal agreement or SlJ<:ial , c:ap1 a rsm ut 1 't .1rk(-t c..oc ali .. Jn or variou ~ forms of' th" m d ' nee d not exc1 cle u Jn ... rxe econom F rth ~ions ar< l'f'l domina ntly ~haped b)' the force f I Y dodu , eoonomtc <.!eel0 supp Yan s emand d. and intt!flhtiJOtaiJy. <tiiU an: free from strict cant I b b . , omest'rea11y ro Y ureaucracJ<::s....

In foreign afl:tirc; )jberalism has show11 as it has in th . d l . .k l'b . , ' t: omestc ream,senousweakncsses. But 1111 Jr c.; 1 erahsm s domes ti<.: realm its rore f h ' . . . ,. . , '' rgn a a~rs ave expenenced startlmg bu t less than fully appreciated successes 'Togeth th h . . . . . 1 er ey s ape an unrecogmzed dilem ma, for both these successes and weaknesses 10 1 rt f .. ~ - tl . " . . . . arge pa spnng rom the same cause. 1e mtematronalunpli<:ations of liberal principles a .t tituti' . tu] f'l b Tl1c .l)ac;J .pos . ate o i eral international theory holcl- that t t nu msth ons. c . .u; s a es have e ng1 to 1t ~e free from. f~re1~1 mterventi_on. Sinc..:e morally autonomous citizens hold rights to lrbert)', the st,ttes that democratically represent them have the right to exercise political independence. Mutual respect for these rights then becomes the touchstone of international Uberal theory. When states respect each other's rights, intlivid uals are free to establish private international ~es withou t state interference. Profitable exchange between merchants and educational exchanges among scholars then create a web of mutual advantages ~nd commitments that bolsters sentiments of public respect. These conventions of mutual respect have formed a cooperative foundation for relations among liberal democracies of a remarkably effective kind. Even though lib1

To organize a group of rational beings who demand generai Ja,~s for ~ei r survival , but of whom each inclines toward exempting himself, and to establis h thetr constitu tion in such a way that, in spite of the fact their private attitudes are opposed, these private attitudes mutually impede each other in such a manne r that [their J public behavior is the same as if thev did not have such eviJ attitudes.'

But the dile mma within liberalism is how to reconcile the three sets of liberal rights. The right to private property, for examp le, can conflic t with equality of opportunity and both rights can be violate d by democ ratic legislation. During the 180 years since Kant wrote, the liberal traditio n has evolve d two high roads to individual freedom and social order; one is laissez-faire, or "conservative," liberalism and the otl1er is social welfare, or social democratic, or "liberal," libe raHsm. Both rec'Oncile these confucting rights (though in differing ways ) by succes sfully organizing free individuals into a political order. The political orde r of laissez-faire and social we lfare liberal s is marked by a shared commitment to four essential institutions. First, dtizens possess juridical equality and other fundam ental civil rights such as freedo m of religion and the pre~s: Second, the effective sovere igns of the state are repres entative Jegislatur~s denvmg their authority from the consen t of the electo rate and exe rcising their authority free from all restraint apart from the require me nt that basic civic rights be pr~serv~d. Most pertine ntly for the impac t of liberalism on foreig n affairs, the state 15. subje<.1' t~ neither the externa] authority of other states nor to the intemal authonty.of s~1a) prerogatives held, for example, by monar chs or military cast~s over foreign ~hey. ?flird, the economy rests on a recognition of the rights o~pn vate property mcludmg the ownership of means of produ ction. Prope rty is justified

eral states ltave become involved in numerous wars with nonlibeml states. constitutionally secure liberal states have yet to engage in war u.,ith one 01 10ther. 2 No one should argue that such wars are impossible; but preli minary evidence does appear to indicate that there exists a significant predisp osition against warfare between hberal states. Indeed, tl1 reats of war also have been regarded as illegitimate. A liberal zone of peace, a pacific union, has been maintained and has ex"Panded despite numerous particular conA.icts of economic and strategic: interest. ... Statistically, war betwee n any two states (i.n any single year or other short period of time) is a low probab Uity event. War between any n.vo adjacent states, conside red over a long pe riod of time, may be somewhat more probab le. The appare nt absence of war among the more dearly liberal states, whethe r adjacen t or not, for almost two hundre d years thus has some significance. Politically more significant, perhaps, is that, when states are forced to decide. by tl1e pressure of an Lmpinging world war. on which side of a world contest they will fight. liberal states wind up all on the same side, despit e the real co mplexity of the historical, econom ic. and political factors that affect their foreign policies. And historically, we should rece:ill that medieval and early modern Europ e we re the warring cockpits of states, wherein France and England and the Low Countries engaged in near constant strife. Then ~ tl1e late eightee nth century there began to emerge liberal regimes. At first hes1ta.nt an.d confused, and later clear and confide nt as liberal regi mes gained deeper domestic founda tions and longer international ex-perie nce, a pacific union of these liberd.l states becam e established.






f . . tional relation wh ic.. ~;' I J, ., a plau'i'IJ The realist model o . mtem~ty of states offer~ lit tlt ,, J l rl' in cxplai~' <: exvlanation of the genentl 1 ~secu~ Realism .in its dasr,ic<tl I r .&t on. hold~ t:Og 1 the pacification of the Hbe;a "a1 r c ereJgn ' effectivch un b0 dr tl bv indiv.1 I(Jtl 'o .1 1 SO" d h Id b rorm v uua the state lS an S OU e Jl ' f determining itS 0\\1 1 ' op 0 <LU tiJority {'fl . rights nationally and thus c:apda J e ~ ticaiJv oligarcllicall~ . or a llttJ{ ratically.).lnt ~ 1 ' . , . . . < determrnation c:an be. made emocra an ,anarchicall!ocwl~ 111 , . I 11c:I1 1t h radicallr. . . ts in . all the sovereign state ex:Js . na t 1 on y. d d protected by intematlOnaI . Ia\\ .. or tn:ntie~ } independent. neither boun e Jblo er b e of the se,entecntl-crnlll rv foun<ue orf u . d h ce insecure Ho s, on f . r~ o uties, ~ en ' dr~~ the international implicalion'i o rcaJism when he the realist appro~, c.:e of international anarchy. the very independence of statt:s, argued that the eXJsten ti.h. n the fear and the temptation toward preventive . ' . . be5 t acco unts for the .compe 0 relations. Politics among nation~ 1s not a contin. . ahonal var that chara.c..tenze JJltern ' b b . . . thjs view a "state of war ... a tract o f time, w herem the will uous eom . at, ut It IS JO 3 nd b battle is sufficiently knovm . . . . . to con:edin yh t all states including liberal states. do engage m war, the realist Fm g t a ' ( r th Jb I concludes that the effects of differing domestic reg es Jlw e e ' 4era or not} h,~h t~. dele b th nternational anarch, under w JC a states ''e. . .. But the are ovem n v e 1 d c h . ends that shape 'the international state of war are decree ror t e real 1 by the st anarchy of the international order ~d the _fu ndamen_tal q~est fo r ~ower_ that directs the poky of all states, irrespec--tiVe of differences m the1r domestic reg1me:s. As Rousseau argued, internationaJ peace therefore depe nds on the abolition of international relations either by the achievement of a world state or by a rac.lic.:al isolationism (Corsica). Realists judge neither to be possible. Recent additions to game theory spec:ify some of the circumstances under which prudence could lead to peace. Experience; geography;_expectations of c~peratio~ and belief patterns: and the differing payoffs to cooperation (peace) or conflic:t as_soc:lated with wrious types of military technology all appe-ar to inRuence the calculus." But when it oomes to acquiring the techniques of peaceable interaction, nations appear to be slow, or at least erratic, )earners. The b-alance of power (more below) is regarded as a primary lesson in the realist primer, but centuries of experience did not prevent either France (Louis XN, Napoleon I) or Germany (WiJhelm TJ, Hitler) from attempting to conquer Europe, twice each. Y some, very new, black African states et appear to have achieved a twenty-year--old system of impressively effec.:tive standards of mutual toleration. These standards are not completely effective (as in Tanzania's invasion of Uganda); but they have confOunded expectations of a scramble to redivide Africa6 Geography-..insular sec..'llrity" and "continental inse<.'Urity"-m ay affed for eign policy attitudes; but it does not appear to detennine behavior, a~ the beiHoose reoords of England and Japan suggest Beliefs, expectations, and attitudes of leaders and masses should influence strategic behavior.... Nevertheless, it would be diffkult to detern.ine if llbernl leaders have had more peaceable attitudes than leaders who lead nonlibeml states. But even if one did make that discovery, he aJso would have to =~';:~these peareable ~des onJy appear to be effective in relations with (~ wars with nonliberais have not been uniformly defensive).. Seoond, at the level of~ ~etermioants , some might argue that relations among any group of states With s1m1lar social struc.tures or with compatible values



would l/ " ful. Bu t again , the c:vide . r nee or odetic\ ' ' t <.cx:1etJec;, or s<x:iali!.t <rv.1 ti , d feudal soc1e hes, <:ommunist S " a t .L Feudal " a r'~ .:a'> rrtqu<::nt and verv ''- 1:! es oes ne> support ull"> conclusion. rnuch Ther< lta.:c ,' c. h('en enou~h truly totalitarian, ~~ of tht, monarchs and nobility. long e1101J~ to tesl fai rly their pacific.; ,.,. bJ~ powers (nor havt they lasted vvrnpati 1 ,..., but ft . wider !>CHV ' of nationaiL<>l, capitalist rnilita _,. h; ~ a'i<.:tst powt-rs in tbt . powers ry "" L )930s. Col nlfdHllSl havt engagtd n u.Jt1:ators 1ps fr>u& 1t eac11 Clther . the m 1 wars mC>re rt<:entl E . .l we have uc)l bad enough SO<.ialist societies t 'd th Y10 astAsia. Anu . ''-t ac.:ification. The more aI)stract category<f oJoonsl er- e rtlf:vance of S0<:1am. a1 W}(:S n t ffi P ) p ur ISm Ce . I Germany was pluralist when it engaged in war with liht a) 0 ~u ce. . rtam Y well in J94J. But they were not liberal. r states m 1914; Japan as And third, at the level of interstate relatione; 'th nc reg.onal attrib. . a11 . neJ er spec1 utes nor h1 stonc - 1ances or fnendships can -:.''"'>u t c th .d -rh e peace ~ n ror e WJ e reach of the liberal pea.c.:e. extends as far as and no furth th h 1 . . , . . 1 d' . ' re liberal states, not me u mg non Liberal states in an oth er an, t e 1 ations among l'b .. . . . h ,. . eiWise 1 era reg.on (sucb a<; the nortI1 AtJantic m t . e l 930s) nor exc:ludmg liberal states a nontL- a1 . m mer reCilon (such as CentraJ Amenca or Africa). o At this level, Rayrnond. Aron has identified three ...11'\P f t st . . . _ l) r -s o m er ate peac:e: 1 ~mp1re, hegemony, an.u ~uihbnum. An empire generalJy Sllcceeds in creating an mternal peac.:e, but th1s IS not an explanation of peac.-e among independent liberal states. Hegemony can ~reate pea~e by over-awing potential rivals. AJthough far fr~m perfect a~d ce~amly precanous, United States hegemony, as Aron notes, ~mght ac~unt f~r the mterstate pea:e in South America in the postwar period durmg the he1 of the Cold War confliCt. However, the liberal peace c.~.ronot be attribght uted merely to effective international policing by a predominant hegemon-Bri taio in the ninet~enth century, the United States in the postwar period. Even though a hegemon m1ght well have an interest in enforcing a peace for the sake of commerce or investments or as a means of enhancing its prestige or security, hegemons such as seventeenth-century France were not peace-enforcing polic..:e, and the liberal peace persisted in the interwar period when international society lacked a predominant hegemonic power. Moreover, this explanation overestimates hegemonic control in both periodc;. Neither England nor the United States was able to prevent ilired: challenges to its interests (colonial competition in the nineteenth century, ~iddle East diplomacy and conflicts over trading with the enemy in the postwar period). Where then was the capacity to prevent all armed conflicts between liberal regimes, many of which were remote and others strategically or e<.:onomically i:nsignifi<.'ant? Liberal hegemony and leadershjp are important, but they are not suffident to explain a liberal peac:e.... Finally, some realists might suggest that the liberal peace simply reflects the absence of deep conflicts of interest among liberal states. Wars occur outside the liberal wne because conflicts of interest are deeper there. But this argument does nothing more than raise the question of why liberal states have fewer or less fundamental < :onflicts of interest with other liberal states than liberal states have with non~beraJ, or nonliberal states have with other nonliberals. We must therefore examine the workings of liberalism among its own kind-a special pacilication of the ..state of war" resting on liber-alism and nothing either more specific or more general

\lost lilwral theorists haw offered inadequat<' .~ 11 d:tu t it t ~tnderstnndin (''\ct:>ptionalnnturc ofliberal pacification. omc h:t\C' .u,.!;utd that d<mocratic 1-;.tht would be inht'rently peaceful si mply and solely lw cau:w m these stales citi~en:t,ttcs tlw polit' and bea; the costs of wars. Unlike monarch:-. tttizens ar<.' not . rttl~ indulgt> their atTarcs ive passions and have the c:onseqwnev-; suff'crcd h)' so~ > e to ~ . . . . .. 'tl. . . . "lse ... Other Jibenls hwe .umted that I<Ussez- I~,urc, cc1p1't> tsm um tatns an in)neone . . ' ' . o . . ,. . . . . . .. . . . lerent 1 11 tendel1 C\' toward rationahsm. and that, smcc ' ' ell ts 111 .lttont~ , 1>era) capitalis >will be I;acillstic. Others tilL such as Montes(jtticu. claim that eotnmerce is thn~ .J e cure for the most destmctive prejudi ces. ano "Pcac<.' ts Llle nalmal ~:>ffcctt of . trade.''!! While these developments can help account f01: the hb: ral peace. they do not explain the f~K't that libE~raJ states are peac~fu~ onlr m ':c'lat1om with other lib. eral states. France and England fought e)\11anst?mst, col~n 1 aJ W<~s. throughout the nineteenth centllly (in the 1830s and 1840s agamst Algc' na and C hma); the United States fought a similar war with Mexico in 1848 <~nJ intervened again in .19 14 under president \ViJson. Liberal states are~ aggr~ss1 ve < war prone as any other ll:d fonn of government or society in their relations w1th nonhbernl states. Immanuel Kant offe rs the best guidance. "Perpetual Pe<\Ce,'' written in 1795, p redicts the ever-widening pacification of the liberal pacific uni on, explains that paciBcation, and at the same time suggests \\'h)' liberal states are not pacific in their relations with nonliberal states . ... Kant sho,vs how repubHcs, once established, lead to peace ful relations. He argues that once the aggressive interests of absolutist monarchies are tamed and once the habit of respect for indhi dual righ ts is engrained by republican government, wars would appear as the disaster to the people's \ve lfare that he and the other liberals thought them to be. The fundamental reason is this:




Th ' l do JHslit' l'f'[Hihlitan rcstr... 1 I 111 " 1, '1 wo11ld no ht warltkc'. whkh i-; Ftr 1.<> ~tlto not end wa r. 1f L) l<.,Y<utl, 1 wra.l slates 1 ~ t ' m c <:asc 1h <.1 --1 tion' inl'l tt 't' nlmollan:lti<.:al<.:al>ri<:. 1 .1 I t) o tntr~>ull<.:<' Kants .. canw. .I )C Tc\ wan. arc only fm (11 t r I 1 '1 c p11111'N'' fo M't' hl)w this n mo\'c!> tl ral -. . , 1,., 1 orpopu ar. 1 > . lC cx:caston ol wars . 1 n I :llld Jlll, \ar' 1wtwccll libcral <~n<.l noul'l ...1 . at lOng '>rra slat<:'s I )t:r,u slates \V(.> need to sh' fl ll t' from tott.lltllltutml law to int<.matin,nlla K t' . l . I our a en ton 1 , . ' " an S secont source. Colltp <m<. nttng tlw < :OII!>ltlutioual uu t f . . ,., a ran ec o cant ton, mtenwt imwl lmc I ad<s a M'<.'< Ht< I 'i011Tl'C- a guarant<c of resl)e -t r1 . 1 1 1 c 1e (.;paration of naliom that ;\S(K'ta ~()('la )I lt: \.'ll('OIIragc>:-. i!> ninforno h ' th . I> I f . . 1 I . . ~ <: t ~vc opment o scparatt lan~r11age~ ,UH Tr 1g 10ilS. 1 1 ll'l>f' further 0'\larantroe . . jl f o . . . ,., " a wor u o s(~paratc slat ''i-m esscnltaJ eon dttton needed to avoid a "vlohal S Oltl-lc"s ,espot'tsm ." y et, 1 l . . !'> ' " at t le stunc tin 1<.>, tile ab(J lllO rallv tnlcgrah~ lilxral statb "as cult '\' . d d 1 I ure P' ogresse!l an men ~ra . uall)' cOIIl t' t. oser toget t<:'r toward a {"1 crr<"ater agree 111 ~ 11 t 01 1 I' 1 ., 10 . ' 1 p n net p es or p<.:act> ant t111tlersta1H I11111 _ As rc c:,.:.t sutti(;") cu ' n. . p11hhes cmc rae (tle n .., . . .. ., 1cl a cult u rC' prognsses, . . r; an llnclcrstandl n_g of the lcg1htnatc rights of all citizens and of all reptthlics comes into pl ~t~; a11d tl~ 1s, now that tau~ion characterizes policy, sc:ts up the m<>ral founJa-

If the consent of the citizens is required in order to decide that war should be declared (and in this constitution it cannot but be the case), nothing is more natural than that they would he very cautious in commencing such a poor game, decreeing for themselves all the calamities of war. Among the latter woultl be: having to fight, having to pay the costs ofwar from their own resources, having painfully to repair the devastation war leaves behind. and. to fill up the measure of, load themselves with a heavy national debt that would embitter peace itself and that can never be liquidated _on account of constant wars in the future. But, on the other hand, ill a constitution winch is not republican, and under which the subjects are not citizens, a declaration of war is the easiest thing in the world to decide upon, because war does not rcqui re of the ruler, who is the proptietor and not a member of the state, the least sacrifice of the pleasure of his table, the chase, his country houses, his court functions, and the like. He may. therefore, resolve on war as on a pleasure party for the most trivial reasons, and wil_ h perfect indifference lea,e the justification which decency requires to the diplotnatJc corps who are ever ready to provide it.9

~ne_ couJd add to Kant's .list another source of pacincation ~~eci.fic to li~~~
constitutions. The regular rotation of office in liberal democratic politJes IS a nontn de\ice that helps ensure that personal animosities among head'i of government provide no lamng, escalating source of tension.

tions for ll~<' . hher~ll pcmx:-. Correspondingly, illtt rnational lnw higl ,1ights the imporl < cc ol ~anltan publl(.:1ty. Domestic:ally, publicity helps <'IISure that the offi_ m cials o~ r('puhl1 cs_ att accor~ing to th e principles the~ profess to hold just and accordtng to tlte mtcr?sts <JI the electors they claim to rf:'prc.: enl. lnt<:rnationally. free spee~l~ and the cl fective communication of a<..curate conceptiou of the politicctllife of loreign peoples is essential to establish and prest'rvc the ttnderstanJing on which the guarantee of respect depends. ln short, domestically just wpublic . which rest on <.:onsent, presume foreign republics to be also consen ual. just. and therefore deserving of accommodation. The experienc-e of coopt:.ration helps engender fu rther cooperathe behavior "'hen the consequences of state polic)' are unclear but (potentiHily) mutually beneficial. 11 Lastly. cosmopolitan law adds material incenthes to moral <...'Ommitments. The cosmopolitan right to hospitality permits the "spirit of commeree" ~oone r or later to take hold of c,ery nation, thus impC'lling tates to pro111ote peace and to try to avert war. Liberal eC'onomic theorv holds that these eo mopotitan ti<.s tletive from a ' cooperative i11tern:..1tional division of labor and free trade according to <...'OmparativC' advantage. Each economy is to be better off than it wonld have been undt' r autarky; each thus acquires an iucentive to avoid policie that would lead the other to bre;k these ewno 111 ic ties. Since kE-eping open markets rests npon the assumption that the next srt of transactions will aLo be determined br prices rather than 1 coercton, a scnse o f mutu;t1secunty ts "'t.lI to' vo' tl securitv-moti"ated,searches for . . economic autarkv. Thus. avoiding a challenge to another liberal states sectt_nty or "a enm en hanci ng each ot 1 rs secunt) h~ m" 11s of allia.nte nahtrallv lollows 1r ecot tonti<: interdependence. h . ti J ket 11 A fmt her cosntO )){)lit< source of lihent.l peace is t1 t e th m tat ltem.~ om lmar f . . 'tribution from e wrect ~1> 1ere o rLmmes dillkult dec:b:ions ol produ<1mn an d Jis .,. r h .1 t . nP}lr directh re. ponsiU1 tOr t ese e state pohcy. A fore1 state thus uoes 110 apr~ gn _ t ti . 1 con en ous ..d [ . . d t some demee <l00ve, t lCS<:' outcomes stalt.s can stand <1St e rom. ,m 0 e" . F I . th . to resoh-e cnses. urt 1 ennore, e tnarket rivalries and be ready to step m




d tb connections of !>l.ttl o l1t d\ he! mtl"rdep\'lldt>nce of c.-onuucrce an c . . . P <:rtatc . l .., al ..,es that serve as Jobb1es f01 ntutl . tl .trtomrnotlat <.'rosscutt mg ransnauon . u . . . IOn 1 11 1 1ransnationa]. Accordine1 modem liberal scholars, intema~onal fin~u ~ ' ' ' , " oti. d domesoc otganizations create Inte rests m fa, OJ o, tt<ornmouat ' ureaucra c. an . 'On b _ , b th . ~an'ety that no single conflict sours an t Ill r\' r<'lationsl11 12 __ . ~m d I1ave e nsureu y e1r v . . p. ,. f b ti'tutonal international o r cosmopollt.m sour<.;cs is al 1 J' O one o t ese cons . one:: . b th ( nd onkwhere together) t11ev plau tbh con1l<'cl the ell u ffi ctent, ut toge er a , . . '. . cl 11, . . ar. 1 acteristics of libe ral politics and economtes w1~1 s~stame ., H r~l peae:e. LibtraJ r. om the realists' "secunty dilemma, till' m~ecuritv e:a1 d 1 d Lr states have not escape l>e by anarchy in the world political system. considere~ as ~ whole. ~3ul th~ e~fects of international anarchy have been tamed m the relatJO.ns.among slat~s of~ stmilarly liberal character. Alliances of purely mutual strategtc mtere~t among hberal and non liberal states have heen broken, economic ties b~tween .ltbcral an.d non liberal c.raoile states bave prove n ri 0 - , but the political bond of liberal nghts and mte rests J,.,~ . , ....., proven a remarkably firm foundation for mutual nonaggresswn. A separate peace






exists among liberal states.

1. Immanuel Kant, "Perpetual Peace~ {1795), in The Philosophy of Kant. ed. Carl J. Friedrich

(i\ew York: Modem Library. 1949), p. 453. 2. There appear to be some exceptions to the tendenc:y for liberal states not to engage in a war with each other. Peru and Ecuador, for example, entered into conflict. But for each, the war <.-ame within one to three years after the establishment of a liberal regime, that is. before the pacifying effects of liberalism could become deeply ingrained. The Pales tinians and the Israelis clashed frequently along the Lebanese border, which Lebanon c:ouJd not hold secure from either belligerent. But at the beginning of' the 1967 War, Lebanon seems to bave sent a Bight of its own jets into Israel. The jets were repuJsed. Alone among Israel's Arab neighbors, Lebanon engaged in no further hosti lities \\~th Israel. Israel's recent attack on the territory of Lebanon was an attack on a country that had already been O<.'<.:upied by Syria (and the P.L.O. ). Whether lsraeJ actually will with draw (if Syria withdraws) ami restore an independent Lebanon is yet to be determined. 3. Thomas Hobbes, Leviathan (New Y Penguin, l980), T <:hap. 13, 62, p. 186. ork: , 4. Kenneth N. Waltz, Man, the State, and War (l'ew York: Columbia University Press, 1954, 1959), pp. 120-2'3; and see his Theory of Ir~ternational Politics (Reading. Mass.: Addison-Wesley, 1979). The classic sources of tbi'i form of Realism are Hobbes and, ~ore particularly, Rousseau's "Essay on St. Pierre's Peace Project" and his "State of\<\'af m A Lasting Peace (London: Constable, 1917), E. H. Carr's The Twerdy )ear s Crisis: 191~1~ (Lond?n: Macmillan & Co., 1951), and the works ofHans Morgenthau . .5. JeMS, Cooperation under the Security Dilemma," World Politics 30, no. 1 (January 1978), pp. 172-86. 6 Robert H. Jackson and Car! G. Rosberg, "Why West Africa's Weak States Persist," _ World Politics 35, No.l (October 1962). I. ~ Aron, Peace and War (New York: Praeger 1968) pp 1.51-54 8 of JncomnM-ibiJity ~f dernocrdC)' and war is forcefully asserted by Paine ' ' in The Ri.gI'1 . The rf Mon. The connection between Jjber a1 capttal ISm, democracy, and peace is argue d bl' r

as m1t1gattng e secunty 1 emma caused by anarcbr. Also, expectations (including theory and historv) ~ inAuence beh av10r. maki ng u ra1 . ''be . , states expect. (and fulfill ) pacific policies toward each other. These e11ects are exp1 d rr . " . .. . ore at .a the?retical level 10 R. Dacey,. Some Implications of Theory Absorption for Econom1c Theory and the Economtcs Infonnation." in Philosophical Ditn.erl.~iom of Economics, ed. J Pitt {Dordrecht, Holland: D. Reidel, 1980). 12. Karl Polanyi, The Great Transjonnatio11 (Boston: Beacon Press. 1944), chaps. 1-2 and Samuel Huntington and Z. Bnezinski, Political Pou;er. USAJT.jSSR (:\ew Y Vikina ork: Press, 1963, 1964), chap. 9. And see Richard Neustadt, ALliance Politics (:\ew Y ork: Columbia University Press, 1970) for a detailed case study of interliberal poutics.

ai!JO"" I) Jowpt. Sc:humpctcr . f . .I -d In n,perw/i~ t . Mem. " JJ J' ~~~~ :vl outescuicu s,,. 't he z . . anGbkSocial Cla.s~e" (''' ew york m ,m oft ., is sur ';' ;u1r I analvtkd hy Alhe: t 11 . ~ws 1 . 20, chap. 1 Th'IS l't , . , 1 . ' r lrS<:hman "R I 1 erature: Socwty ',j \ I mng. De~tructivc, or Feeblc":>n . lva lntt;rpretations of \tarkct (D<:c('rn . t t ICJf:>2) . journal of Econorruc Literature 20 1 9. J uliJW. 1J f' Kaul, "Pcrpc:tual Pcact: 'rl . n 1 1 Enlt~ht Silnon & Se Illl~tcr, J!.J74), pp. 790-92. ? enment, ed. Peter Ca} ( ~~;;w York: 10. Kant, Tlw Pldlo.~oplty of Krmt, p. 454 Th ' est fac-tors also L I I lty D eu~c: I'~ "<.CJrnpa t.l) t CJ f values" and u d'b . nave a 1. . u~aring on Karl .g' 1 slytIZ<.'cl vt:rl>tOn of this e:ffc:c.:t c.:anpre rICta dllJ"' ofbeh a"10r." Iy . . I l. A ~n b 1 ., r e . 1 ga1 There:, a 1a11 of rnutual tnJst and th oun m the real'st's "Pmoners Dilcm 1 na~ n<.>. urc e mc:entives t 0 h . produce a non<.:oopcrative solution that m k b ~, en ance one~ own position . . . erat1on , a con llntlm(:;nt to avoid txploiting a es otn parties worse oIT. Contranly, <.'OOp tl 1e ot1 partv p d er . . . nifkance o r tlw garne in this c:ontext is the h . f.'' ro U(;(:s JOtnt gams. The sig ' I r I c o are prcsurncu to Je C on ious unrt:lated apartaracterh . tts participanus. Tl 1 u pnsoners" f c . . , rom t elr partne h'1 . . . II 111g 111 mutual trust-<:ompetitive nation-stat . . r~ P m c:nrne, ant ack. es tn an anarchtc world A 1 benveen 1ratcrnal or sororalnvins-K t'5 bl' 51111 ar game an repu JCs-would be l'k 1 1 ad "f-r 1 ent results. See Hobert Jervis "Hypotheses on~~ . ~ e Yto e, tow crj r . . 1sperception, World Politic.~ 20 ~ 0 3 (Apn l 968). ror an expos1t1on of the role of presum r1005 . d "C . .. o1 .. L P an ooperatwn under the Se<:unty 1emma, Wond Politics 30, ~o. 2 (Janu-ary 1978) 10 1. t1 f r . th d'l 1e ac-tors reahsl\ see







Liberal regimes and the pacific union (by date "liberal")a

- ---~








Liberal regi mes and the pacific union (by date "liberal")a Switzerland, the United S Canada, Australia, Nev~a~:, Great_ Britain, Sweden. Mexico and, Finland, Ireland Uruguay 1973 Chile 1973 Lebanon 1975 Costa Rica 1948, 1953_ Iceland 1944France 1945Denmark 1945Norway 1945Austria 1945Brazil 1945- 1954, 1955_1964 Belgium 1946Luxemburg 1946Netherlands 1946ltaly 1946Philippines 1946-1972 lnd1a 1947-1975, 1977Sri Lanka 1948-1961, 1963_1977, 1978Ecuador1948-1963, 1979lsrael 1949West Germany 1949Peru 1950-1962, 1963-1968, 1980EI Salvador 1950-1961 Turkey 1950-1960, 1966-1971 Japan 1951Bolivia 1956-1969 Colombia 1958Venezuela 1959Nigeria 196 1-1964, 1979Jamaica 1962Trinidad 1962Senegal 1963Malaysia 1963South Korea 1963-1972 Botswana 1966Singapore 1965Greece 1975Portugal 1976Spain 1978Dominican Republic 1978-

18th century



Swiss Cantonsb French Republic 1790-1795 the Un1ted Statesb 1776Swiss Confederation, the United States France 1830-1849 Belgium 1830Great Britain 1832Netherlands 1848Piedmont 1848Denmark 1849Switzerland, the United States, Belgium. Great Bntam, Netherlands Piedmont 1861, Italy 1861Denmark 1866 Sweden 1864Greece 1864Canada 1867France 1871Argentina 1880Chile 1891Switzerland, the United States, Great Britain, Sweden, Canada Greece 1911,1928-1936 Italy 1922 Belgium 1940 Netherlands 1940 Argentina 1943 France 1940 Chile 1924, 1932 Australia 1901Norway 1905-1940 New Zealand 1907Colombia 1910-1949 Denmark 1914-1940 Poland 1917-1935 Latvia 1922-1934 Germany 1918-1932 Austria 1918-1934 Estonia 1919-1934 Finland 1919Uruguay 1919Costa Rica 1919Czechoslovakia 1920-1939 Ireland 1920Mexico 1928Lebanon 1944-


Total number





~ 1 ha~e drawn up ~his approximate list of "UberaJ Regim~ acc~r~ing to the four institutions ~scn~~d as essentral: market and private property econom1es: poht1cs that are extremely soverergn; CJttzens who possess juridical rights; and "republican" (whether republican or monarchical).



ARCHYANUII;)\...VI '-1'-'<- - - ~

l'\:t't\'St'lll31l\'t> g~we1 nment This latter includes the requireme_ ' tl .. : I t le ~!ali ve branch have n ' potentially or ,\0 ellt"\:'ll\ e tult' m public policy and be formally and compelltl '--

~ t 1~tcJ Ft1rthermore 1have taken into account whether male ~ullru <' vr ... pen to JChH!\ cment by mhabitants (for cxampl~. to poll W' I "


TABLE 2 a cr.ntlnued) FAJRS Palestine (194&-1949) Hyderabad (194S) ~~adagascan (1947-1948) irst F Kashmir (1 947-1949) Korean (1 950- 1953) Algerian (i 954- 1962) R~sso-Hungarian ( 1956) 1nai S (1956) T~betan (1956- 1959) Smo-lndian (1962) Vietnamese (1965-1975) Second Kashmir (1965) ix S Day ( 1967) Israeli-Egyptian (1969-1970) Football (1 969) Bangladesh (197 1) Philippine-MNLF (1972_) Yom Kippur (1 973) Turco-Cypriot (1 974) Ethiopian-Eritrean ( 197 4-) Vietnamese-Cambodian (1975-) Timor (1975-) Saharan (1975-) Ogaden (1976-) Ugandan-Tanzanian (197&-1979) ino-Vietnamese (1979) S Russo-Afghan (1979-1989) lrani-lraqi (1 980-1988)


:1: that is, 30 actually, s r national 0 1 metropolitan territory. Female suffrage IS granted V!lh n " S- c ration of it or 1 , demJnded and representative government is internally so_ erelgn f ex .mple. inctu/ be ng lng and .-.mtXiallv over military and foreign affairs) as well as stable (111 exste"-e f r ::n least thre e years) . . . ~ - -"" t-There are domesuc \'ariations within these liberal regmes. For exam;;le S\ tzerland wa r liberal only north ?r the 1\tasor. Oixon line only m certain cantons. the Umted States 86s, "hen it became liberal throughout. These hsts also exclude anCient reoubllc<- smce non lil Kant's criteria see stephen Holmes hAristippus in and out of :\thcns AmencaJ'e ~hOc!

'.uuseholder~)ercent) ~he ~



un~il~~eral ~PPear

a eview 73. No. I (March 1979). SCience R ' selected list. excludes liberal regimes with populations less than one millton. SOurces: Arthur Banks and w. overstreet, eds.. The Political Han~book ~flhe \Vorld. 1980 (New Yor . o; Foreign and commonwealth Office. A Yea1 Book oj the. Commonwealth 19} McGraw-Hill. 198 1_(LOndon: Europe. 1981 l: \V L. Langer, An Encyclo (London: HMSO. 1980); Europo Yearbook of State, Coumry dia of world History (Boston: Houghton-Miffim. 1968): Human Righrs Practices (\.Vashington. D.C.: u.s. Government Pnntmg Office. 198 1); and

1 9~



~ Reports~

Issue, No. 54 uanuary-February 1980).


British-Maharattan (1817-1 818) Greek {1821- 1828) Franco-Spanish (1823) Ftrst Anglo-Burmese (1823-1 826) Franco-Mexican ( 1862-1 867) Ecuadonan-Colomban ( 1863) Second Poltsh (1863-1864) Spanish-Santo Dom111ican ( 1863-1 865) g-Holste111 (1864) Second Schlesw 1 Lopez ( 1864- 1870) Spanish-Ch1lean ( 1865-1866) Seven Weeks ( 1866) Ten Years (1 868-1878) Franco-Prussan ( 1870-1871) Dutch-Achinese (1873-1 878) Balkan (1875-1 877) Russo-Turkish (1877-1878) Bosnian ( 1878) Second British-Afghan ( 1878-1880) Pacif ic (1879- 1880) British-Zulu {1 879) Franco-lndochinese ( 1882-1884) Mahdist (1882- 1885) Sino-French ( 1884- 1885) Central American ( 1885) Serbo-Bulgarian (1885) Sino-Japanese ( 1894-1895) Franco-Madagascan (1894-1895} Cuban(1895-1896) ltalo-Ethiopian ( 1895- 1896) First Philippine (1896-1898) Greco-Turkish (1897) Spanish-American ( 1898) Second Philippine (1899-1902)

Japanese (1 825-1830) Russo-Persian (1826-1828) urkish (1828-1829) Russo-T First Polish (1831) First Syrian (1831-1832) Texan (1835-1836) First British-Afghan (1838-1842) Second Syrian (1839-1840) Franco-Algerian ( 1839-184 7) Peruvian-Bolivian ( 184 1) First British-Sikh (1845-1846) Mexican-American ( 1846-1848) Austro-Sardinian (1848-1849) First Schleswig-Holstein (1848-1849) Hungarian (1848-1849) Second British-Sikh {1848-1849) Roman Republic (1849) La Plata (1851-1852) First Turco-Montenegran (1852-1853) Crimean (1853-1856) Sepoy (1857-1859) Second Turco-Montenegran (1858-1859) Italian Unification (1859) Spanish-Moroccan (1859-1860)

soer< 1899- 90LJ Boxer Rebe on - 900) Hinden (1903J Russo-Japan~:se ( 1904-1905) Central Amencar (1906) Central Amencan (1907) panish-Morcccan (1909-191 O) S ualo-Turkish ( 1 911-1912) First Balkan ( 1912-191 3) Second Balkan {1 913) world War I (191 4-1918) Russian Nationalities (1917-1921) Russo-Polish (1919-1920) Hungarian-AIIies ( 1919) urkish (191 9-1 922) Greco-T Riffian ( 1921- 1926) Druze (1 925- 1927) Sino-Soviet (1929) Manchurian (1931-1933) Chaco (1932-1935) ltalo-Ethiopian (1935- 1936) Sino-Japanese (1 937-194 1) Changkufeng (1938) Nomohan (1939) World War 11 (1939- 1945) Russo-Finnish ( 1939-1940) Franco-Thai ( 1940-1 941 ) Indonesian (1945-1946) lndochinese (1 945-1954)

. Table 2 is from Melvin Small and J. David Sin er Reso ~s~ PP 79-80. Copynght ~ 1982 by re to . .g Sage Publications, Inc. Re rinted b ~~~~~s~~~~fl Sage Pub~cations. Inc. This is a partial list of international wars fought between Y o .- n Appendices A and B of Reson e Anns. Small . . I and Singer identif largely domestic ~i~~lta aof 575 wars m this penod. but approximately 159 of them appear to be w rs. re defi 'f era!Thisim m IO~ of war ex~ludes co~ert interventions, some of which have been directed b libthe ChiT es laga.mst other liberal regmes. One example is the United States' effort to desta6'mze tions a:an e ectlon and Alle~de's government. Nonetheless, it is significant ... that such intervenpursued pub~lcly as acknowledged policy. The covert destabilization campaign against ~h~lot _ e . 1e .1s recounted m U.S. Congress, Senate, Select Committee to Srudy Governmental .espect to Intelligence Activities, Covert Action in Chile. 1963-73, 94th congress. 0%t ~atl~ns With R 1 ess1on (Washmgton. D.C.: U.S. Government Printing Office, 1975).


ltalo-Roman (1860)
ltalo-SICilian (1860-1861 )

--- ... ~..c.::,

Alliances: Balancing and Ba11dwagoning




. I ' the < 1Jclfl ,. l 'C<liiSe the weaker sid reases the new member' 1 l 1 . the stro11 ,id". bv contnst g e has .. s nAuence , ' . tves t1e n , greater need ~ r a-;s1stanc All . within 1 relativeh 1t ' to the coalition ) an cl eaves member little inA e\\ e. .t I uence (bee ymg v.ith . 1 1 vu nerable to the wh ~~~se 1t adds Joini ng l1 \Waker side should b " h w e t e preferred choice. lms of Jts partners.

Se<c 11

je>i" i"g the w elk"

.' . '" r Slut Jnc

. f t 1 1 . a srgnificant extemal threat. ~tates lllH) Pitlwr balance < )r 1en . . w \\ 1 con r011 re .,, s de~ned as alhin<T w1th others ag<.un:st t 1 pn.~ varlmgr threat 1 . n . bandwa<Ton.. BaIonclll{!.to aliQ1lment "vith bt11e source of clang 'r. Thus two distir~ct, tr . u:ogow11g rerer!i . . . . .' 0 baudtheses about how states wiJI select the1r ~Jj;;u1ce P<~rtr~c,'1.. .. b<, .rc.lcntinc<l on h' scan .:po . f 1 tl . tll"' --tltes aU against or w1tb the pnncJr)al extcmal threat 1 l th e ><l"ls o '"' 1e 1er , " ' . . . . These nvo hypotheses depict ,ery different worlds. 11 halanc.: mg 1 more coms mon than bandwagon ing, then states are m01:e secure. b_ecause aggressors will face combined opposition. But if l~andwagoning ~~ tl1e dommar~t- tende, then securitv is scmce, because successful aggressors wJ.JI attract adtlit10nal all res, enhancing their power while reducing that of their opponents . ...

The belief that states fonn alliances in order to prevent stronger powers from dominating them lies at the heart of haditional balance-of-power the01y According to this dew. states join aJliances to protect d1emselves from states or coalitions whose supe rior resources could pose a threat. States choose to balance for two main reasons. First~ they place their survival at risk if they fail to curb a potentiaJ hegemon before it becomes too strong. To ally with the dominant power means placing one's trust in its c :onti.nued benevolence. The safer strategy is to join with those who cannot readily dominate t11eir allies, in order to avoid being dominated by those who can. A5 \Vinston ChurchilJ explained Britain's traditional alliance policy: "For four hundred years the foreign policy of England has been to oppose the strongest, most aggres sive, most dominating power on the Continent. . .. [I]t would have been easy .. and tempting to join wid1 the stronger and share the fruits of his conquest. However, we always took th~ harder course, joined with the less strong powers, ... and thus ~~f~ted the Contmental military tyrant whoever he was."2 More recently, Henry K.issmger advocated a rapprochement witl1 China, because he believed that in a tri angular relationship, it was better to align with the weaker side.

ReprintedU f~'_Stephen ~1. Wait. The Origins of Alliances, pp. 17-21 27-32. Copyright t98i by Comell oD'IiHed. l..... . have be-esannemty Press. Used UJ pennJSSJon of t1le publisher. Portions 'of the text and some footJtOtes

The beli<'f .that states will bllance 1s uns . c .. examp Ies1o f slates joining togeth er to resiSt urpnsmg, .given the many 1 . a th r aJn1 1 desp1te t .1e powerful evidence tl1 t .h1 reatenmg state or coal't yet a storv p d . I 10n. hypothcs1 th e belief that the 0 pposJte respon es m support of the balancmg . ' rovJ . . ' . s. Accordmg to one scholar "In 1 matiana pol'ti se IS more likelv is ,Vldes)read. nte . . . . I 1 1 Morn enlum accrues to the gainer and . . cs, ~othmg succeeds like success. 1 . .. --1 . of m eve1s1)J'I'rty . I11 g<tins enfeebl acce erates his movement. The appearance rn .s . ' more. 'fl 1e l)ancIwagon collects one stde. and sf rmu Iates lhe other all the th .d .1 1.. .1 . on e sr elines '3 1 1e udnuwagonrng hypothesis is es eciall , . . justifY overseas involvements or incr pd _ popular With statesmen seeking to liY . . ease rnL tarv bud t F admmtl AJ fred van Tirpitz's f.unous n k tl < s 1eol)rrested ge s. or example, German th 15 ing a great battle fleet, Tirpitz argued German coul on type oflogic. By buildor alliance 'vith her by posing a threa~ to En l~d's vi~~orce .~ngland into neutrality . Bandwagoning beliefs have also been recurrin mantime supremacy. War. Soviet e fforts to intimidate both l\ . . d g them.e throughout the Cold reveal the So,iet conviction that state~rw~ an Turke)da not ~oinjng ~ATO . mto alth l tl ' accommo te readih- to threats . oug 1 1ese moves merely encouraged ~orwav and Turke t ali . I I, 'tl tl v1 t o.1 s ffi . . " o gn more c ose ,. Wl 1 le ves . OVJet o CJals made a similar error in b li . th tb . Soviet mili . _ e evmg at e growth of . ta.ry power m the 1960s and 19 tOs would lead to a permanent shift in the correlatiOn of fo rces against the West. Instead. it contributed to a Sino-A 11 in the 1970s and the largest peacetime increase in u.s. power m the 1980s. I) . American officiaJs have been equally fond of bandwagoning notions. Accordm~ to ~SC-68, the ~lassified study that helped justi~r a major U.S. military b~~dup m the 1950s: In the absence of an affim1ative decision [to increase U.S. m~tary capabilities] ... our friends will become more than a liability to us. they w1U become a positive increment to Soviet power.'>5 President John F. Kennedv once claimed that "ir the United States were to falter. the whole world ... would inevitably begin to move toward the Communist bloc. -6 And though Henry Kissinger ofte n argued that dte United States should form balancing alliances t~ contain the Soviet Union, he apparently believed that U.S. allies were likely to bandwagon. As he put it, "If leaders around the world ... assume that the U.S. lacked either the rorces or the will ... thev will accommodate themselves to what they will regard as tl1e dominant trend.''; Ronald Reagans claim ...If we cannot defend ourselves [in Central America] ... then we cannot expect to pre,ail elsewhere. . . . [0 ]ur credibility will collapse and our alliances will crumble.'' reveals the same logic in a familiar role tJ1at of justi~ing overseas interYeotion.~







_ 1

d . g are usuaJJy framed solei~ i11 tcrn" <J c.:apahili t', Bthncin,r :md u:ln wagollln aker side. ban d wagonmt:; wt " ' t 1w res. , . ~ .I h o;lrongcr Tl . Ba1 1. lf.,rnmber1t w1t 1d e we a}' " .. . :mcmg . t I ,e,er to account for the otltr l~t tor\ that stat 11~ . esronc..-ept ton s1 uld e rev1se . 10'' 1 t aUy Althouvh l'ow,.r i<; a11 im w 'ti 'd I l deciding wt 1 w lom o . ..., portant men cons1 er " 'ler It s more accurate to that stat . art of the equation it is not tile on1 one. ' y . ~ cs tend P . . h ~ . . po"'er that poses the grcatec;t titre at. I or cxampl , to ally \\lth or agrunst ~f OJet~ otiler strona states if a weaker pr)W<'r is morf dare, 1 state may balance by )l n; '~ ~ coalitions iliat defeated German~ in \ Vorl<.l War ; T aerous fo r other reasons. tl . sur)erior in total resource~. 1 l 1 came togethe us 1 o 1ut W\ and \\'orld \\"ar JI were ,as }aggressJ\:e auns of . . ' dJ1c . . Lutd ;\Jaz1s posE:d .. . r . th ti1 the \\ . 1mm~!-. : when rt became c1 ear at e ancing and ban dwagonmg a1.c, mor.c accurately . bal ti1e greater danger. Because ts it is important to cons1der other . factors that will tllrea ' viewed as a response t 0 . h affeet the level of threat that states ma}' P?se: aggregate power, geogJap re; proxiJn-


i ' offensive power, and aggressive int~ntJons. . . .

.. , . . ty B defining ti'le basic hypotheses U1 terms of threats rathe r t~an po~ver alone, ~ lete picture of the factors that statesmen w11l consH1er when we gam a more comp . . . . chot'ces One cannot determine a pnon, however, wh1 ch sources of making all 1ance . . . . . threat "'~JJ be most important in any gtven case; one can say on ly all .o.f them are likely to play a role. And the greater the threat, the greater the probabd1 ty that the "ulnerable state v.ill seek an alliance.


The two general hypotheses of balanci ng and bandwagoning paint starkly contrasting pictures of international politics. Resolving the q uestion of which hypothesis is more accurate is especially important, because each implies very di ffere nt poli<.:y prescriptions. What sort of world does each depict, and what p olicies are implied? If balancing is the dominant tendency, then threatening states wil.l provoke others to align agrunst them . Because those who seek to dominate others will attract widespread opposition, status quo states can take a relatively sanguine view of threats. Credibility is less important in a balancing world , because one's allies will resist threatening states out of their own self-interest, not because they expect others to do it for them. Thus the fear of allies defecting will de::cline. Moreover, if balancing is the nonn and if statesmen understand this tendency, aggression will be discouraged because those who c."Ontemplate it will anticipate resistance. In a balancing world, policies that convey restraint and be nevoJence are best. Strong states may be valued as allies because they have much to offe r their partners. but th~ must take particular care to avoid appearing aggressive. Foreign and ~ense polic.1es that minimize the threat one poses to others make the most sense m such a world.

world. b(c_ ' .t. < tdd,tJonal defec:tions and f exneded ~.I .rrr>v<"r, i/' ~tah::Smtn b 1 ah urther dtcline in p<><;iti r c ltve t at b 1 on are to h< 1 will be mon 11c 111cd to use fc>rc:t ThIS tcnden an< wagoning is wj-L. .1 , ucspr<:ctU, thcv status quo DO'.\ crs. The former will use r ' ~k wrce b C) IS true for both aau . th ~re~~ors and will be 11111! .l h to balanc:c against them an d be . hey will ~sum th t h ecause ' -:: a <>t ers through bC'Ih~f'fC:nc:e or brinkrn<.~nship Th I . ca~e t ey can attract mor<: all' 1 t atter \VJIJ ~011 11::s fear t I ga111~ t WJ r oppone:nts will mak b 1e . ow suit be<:ause the 1 11 11 e Y appeann r yw F111 a \ , JTII\pcrc:eJvmg the relative . g powenul and resolutE: ,, propens1ty to b 1 gerous, becau\C the policies that are ap . . a ance or bandwagcm is clan. propnate for 0 . )1er. f f ~lat<::S inen follow the baJan . the ot . nE: ~Jtuatton wiJI hackfire in (;]ng prescnpf1 b o their moderate re~ponse:s and relaxel VJew 0 f l h reatc; nn a an<.lwae~cming world In . u ., . defect, leav1ng thetn isolated against an h . \\'I encuurag<; th~:ir alii<::~ l.o overw e1 mmg . 1 . . )owing the banclwaguning prescription in a world of coa Jtlon. Conver e::ly, foland th reats fn::quently) will lead others to 0 balanc.;er (employing power These c:once::rns are not merely theoreJp~s; more and more vigorouslyY' ogn.ize that he r all ies in the Little Entente ea n e 1930s, France failc:d to recwere prone to b d that French mi litary and diplomatic poJ,c::Jes . r d an wagon, a tenoent:}' rem1orce As noted 1 S . attempts to intimidate Turkey and Norway aflte " ' ld W. . ear Jer, OVJet r vvor ar f I reveal the error; they mertly provoke<.! a greater US eomffi.J ment to t11 .t oppostte esc rt:cic n d cemented thei r entry into :\ATO. Likewise' the seIf-enc:1rc1 1 lit;) . an . mg ) t 1 ty of . . . cos1 Wilhelmme Germany and Imperial Japan reflected the ac;sumptJon, prevalent in ' . both states, t hat bandwagonmg was the dominant tendenC\' ternationa1 aff:aJrs. . mm




T.hese examples highlight the importance of identifying whether states are more ~ely to balance or bandwagon and which sources of th reat have the gre-atest 1mpact on the decision . . .. In general, we should expect balancing behavior to be much more common than bandwagoning, and we shou ld expect bandwagoning to occur only under certain identifiable conditions. Although many statesmen fear that potential allies will align with the strongest side, thjs fear receives little support from most of international history. f or example, every attempt to achieve hegemony in Europe since the Thirty Y ears War ha.c; been thwarted by a <.lefEmsive coalition formed precisely for ilie purpose of defeating the potential hegemon. Other examples are equally telling. Although isolated cases of bandwagoning do o<:c.:ur, the great powers have shown a remarkable teudency to ignore other temptations and follow the balancing prescription when ~e<.>eSSaJ). This tendency should not surprise us. Balancing should be preferred f~r the simple reason that no statesman can be completely sure of what another will do. Bandwagoning is <.langerous because it increases the resourc.-es a\'a.ilable to a threatening power anti requires placing trust in its continued forbearance. ~use pe~ ..: cl __ c. to balance agamst potential uons are unreliable and intentions can tange, 1t lS sater . threats than to rely on the hope that a state will remain bene' ~ . But if balancing is to be expected, bandwagooing remams_a poss,brlit}. S~eral factors may affect the relative propensity for states to select this c.:ourse.


a:. ~oning they, those

ascendaoc.y of



world, by contrast, is much more competitive. If states tend who seem most ~gerous, then great powers will be rewar~ed if both strong ~d potentially aggressive. International rivalries WJ.ll be because a sm~e ~feat may signal the decline of one side and ~e the other. This situation is especially alannjng in a bandwagonL!Ig





Strong versus Weak States

than tl le 111 ore likch r, . 1e ln 1!\'nt.'t~\L tlw weake r t 1 hatutc.'\.'. Thi ituation occut because weak t ates add httl~ lt th(;; stre ngth of a -''t' . c"'w 1011 lmt.mCUJ' tl~e wr:tth of the n1ore th reat' t un~ states nonethcless 1 < ' ~n ot' 1 1\'l' .ttle to affect tl1e outcome (<mU n.av -;ul fer gtievousIv . . 10 Bl'<'<tu e we~\k states can do ll n t choose tJ1e winning side. On l whe. tltu r decision c:an "' f~ ., (ec:t tl te proc~ss ), t11ev mus c . 1 raaonal Jl01. tJ1em to J'oin tbe weaker alliance. B, contras t.' Sttong . } outcome IS :t t 1e on . r 'ti'on into a wmrunco one . And bee .ms e. lhe 1r dt:c1s1 ,.... ay . g coalJ osm tates can tum a 1 arn p1 y 1,erence beh ''een \'ictorv and defeat the v ar(" l1 kely to be e mean tl1 cun . .. . rewarded for tlleir conhibution. \\'eak states are also likely to be especJally se nsJtJvt, lo ~J ~~Xl mate power. .. . . . ' ere ,,'I1 great powers have both global interesttJ and .global 1,c:apabtlii1es, \\'eak tates . s . \ v , ,1 be concemed pn 111 ariJ. with events m 1eu JI11Jlle<.uate V lC:m tty. Moreove r, \.vu k states can be expected to balance whe n threatenecl hy states with rouah]y wea al capabilities but tlley will be tempted to b~ndwagon \\'~le n ~hre aten ~d by a equ great power. Obviously~ wh~n the gr.e~~ pm~er ts c~~~ ab_J e ~~ rap td ~nd effcc~ve action (i.e. , when its offens1ve capabilttles ar e esp ec.:1all) st10 ng), thts temptation '-'~11 be even greater.
sta te.
l' , " .

it i~ tt h.~ .. ,h, .t<ron r1tJ1er

Tah ,!,, tiHr, these factors hel cxp1 . lte fo t' f 1 grea t powers Altl' p I amt L rma ron o sphe res of B ce s urro 1 mrun ~' ll' m uen oug 1 strong nei II to bala ncf'. ' tall and wea k neighhors of the P g I lors of strong tates are \jkch .g ~at d ers . " bandwagon.. Because they wi ll he th e F'. . vt<:hmspo~v, may be more in<.:lined t'o 1rst . aI)I 1 the cap ttH's to stan d alone, and because a defeJn cxpansron, hec.:ause thtv lack . . all' rate too . o em slowly. to <.1 t I1 much good accom modatrng a th srve . ranee ma)' ope .t rea enrng great power may he temptmg.
Peace and \Va r

The Availability of Allies

es will also be tempted to bandwagon whe n allies are simply una,aiJable. This Stat ement is not simply tautological , because states ma~' balance by mobilizing their stat resources instead of relying on allied support. They are more like ly to do so, own however when thev are confident tllat allied assistance ,..,rill be avai lable. Thus a further prerequisite for balancing behavio r is an effecti, e syste m of diplomatic commwtication. Tbe ability to communicate enables potential allies to recognize their shared interests and coordinate tllei r responses. If weak states sec no possibility of outside assistance, however, they may be forced to accommodate the mosl imminent threat. Thus the first Shah of Iran saw the British withdrawal from Kandahar in 1881 as a signal to bandwagon with Russia. As he told tl1e BJitish rep rese ntati e, all be had 11 -no thi ng else... received from Great Britain was "good advice and honeyed wo rds Finland's policy of partial alignment with the Soviet Union suggests the same lesson. When Finland joined forces with Nazi Gennany dw ing World \Var 11, it alienated the potential allies (the United States and Great Britain ) that mig ht otherwise have helped protect it from Soviet pressure after the war. ~f course: excessive confidence in allied support will encour age weak states to -nde, relymg on the efforts of others to provide secmity. Fre e-riding is the free mal polic.y for a weak state, because its efforts will con trib ute little in any case. opti Among ~e great powers, the belief that allies are readily available encourages k-_ assmg; states that are threaten ed strive to pas s to oth ers the bur dens of buc p :nd ing up to the aggressor. Neither response is a form of ban dwagoning, but th suggest. that balancing behavior is more likely to occw whe n mero al bers of an alliance are not eo ced that the1r par tners are unconditionallYI0Y nvm
' J

Fin al!)', the context in which alliance chot<:es are mad '1 1 f~ c . ance or bantlwagon. States are more like) t bal . . ~ WJ a e~t decic;ions to halstages of a war, as they seek to deter 0 ; d~feat~~= tn pea cetim~ or in the earl y D~wers posm g the greatest tltreat. But once the outcom e appears certain the losing sidE> at an opportune mom ent Th, so: e~; e te.mpted to defect from allie with Nazi German y initially and then aband~s dote umam~ and Bulgaria as thd ne ermany ror the '1\llies bb d f 'd e t1 es o war e e_ and. flowed across Europe in World War Il. . . . The restoration of peace, however' restores th e mcentiv.e to baJance. As man v . . . . . observersf Ila\.e. noted, vtctonous coalitions are like! Yto d'tsmtegrate wtth the con' . ). . 1 . ce. e us1on o peak . I romment exam ples mclude Austria an cl p russra aft er tl1e1r war -..1.. 1 th s u w1u 1 Den ma r m 186 4, Britain and France after World War , e ovtet nton f' ds tL~tes a ter World War II, and China and Vietnam after the u.S. 1e ar~ d t1 Untte \~thdr~'~al from Vte tna_m. ~his recu~ng pattern provides further support for the p10pos tt10n that_ bal.ancmg 1s the dommant tendency in international politics and tllat bandwagomng IS tlle opportunistic exception.


Hypotheses on Balancing
I. General form: States facing an external threat wilJ align with others to

2. 3.


5. 6.

oppose the states posing the threat. The greate r the threatening state's agg regate power, the greater the tenden cy for others to align against it. The nearer a powerful state, the greater tJ1e tendency for those nearby to align against it. Therefore, neigh boring states are less likely to be allies than are states separated by at leas t one other power. The greate r a state's offensive capabilities, tlle gre~ ter tl1~ tenden~~~ for others to align against it. Therefore, states \\~t11 offensl\:ely o~ented -~rhtary capabilities are likely to pro"oke other states to form defensi,e ~oahtions. The more aggressive a state's perceived intentions, the more hkely others are to align against that state. Alliances formed during wartime \\~11 disintegrate when the enem~ is defeate d.

l 0:!



Hypotheses on Bandwagoning
Tlw hypotheses on hanth'<\goning are the oppo!'itt of tl u on h\lanci 11 g.


- -



1. Geueralform: States facing an will:.'!) \vith the most tl'teat ening power. .. . 2. The greater a state aggregate capabd1ttes, lhe grt.\ter the tendcnc: . 'tl 't y for otJ:1ers to tgn W J 1 1 . al 3 . The nearer a powerful state, the greater th(' tendency for those nearb - 1= 't Yto augn WJ'tl1 J . . .. 4 ' The o <Yreater a state's offensive capabilities. the greater the tendency r lOT





others to align '"~th it. 5 The more aggressive. a state's percei,ed intention . the less like!) tl . o 1er states are to align agamst 1t. 6. Alliances formed to oppose a threat will disintegrate \ovhen the threat becomes serious.


Hypotheses on the conditions Favoring Balancing or Bandwagoning

1. Balancing is more common than bandwagoning. 2. The stronger the state, the greater its tendency to balance. vVeak states will balance against other weak states but may bandwagon when threatened by


great powers. 3. The greater the probability of allied support, the greater the tendency to balance. w hen adequate allied support is certain, however, the tendency for free-riding or buck-passing increases. 4. The more unalterably aggressive a state is perceived to be, the greater the tendency for others to balance against it. 5. In wartime, the closer one side is to victory, the greater the tendency for others to bandwagon with it.

'ih~')(} ( X ew York, 1970)' . . . . aml t ,f.d \\ ar: l94.'5-1949 ( X p 3BI , andCelrLundestad An . :-._ <;C ( ... L' nitr~ <.l States ObJ fcv. York, 1980), pp. 308-9. ~enca, Scondmat;ilJ, C.a< 11 ar cJ Et;..--nId. ContainmentJve;s 404 Prog ec p and s .rams r,or :\ational l. Se::curitv") r . d . all(l-1 3 ;. ' . . lmilar passages can be founu' . epnnte m Qu(J!('(] in Sevom Browtl "'l F on PP 389.414, . . .' ' te aces of p . Ioretf!.ll :ohr:y from Tnwum to Johnson (~wer:) and Clumge in United Stat Quoted 1 U.S. House Com ln'tt n ew or . 1968). p. 917 e.s I ee on Fo ff: 1 11 World Watershed ill Great Power p . :~ ~ A airs, The Societ (.;11 ion and th Th. d 1 Neu: )'o rk Times , April28 198 Ao rcy . ttb Cong.. 1st scss.. 1971 PP 1- - : :~ rr . . . ' p. 12. ln the sam 3 , . .., h.IO. tra I Amenca r were to fall ' wlaat wou Id t I1e , conseq e speech. Rcagan also said. "lfC enbe -> for E urope an d JOr alliances such as :\ATO'. . . . \\ .u~nces wh' our position in A.sia d ' ?" . . hJch allY 1ch fnend wouJd . an 1cn. t tl y 1 . m~t us t 1s wort 1 h h that Napoleon an d fJlt1 underesti l noting . . .I I assurnmg t at t eir potential ene . -J . er ma eu t 1e costs of aggression b . 1 . 1 . mles wou1 bandw d ,r Y 1t er ulSI111Ssed the possibility of .. agon. J 'lter Munich for example 1r ., . opposition by claim tl B . . ' men were !Jttle worms.'' 1\apoleon a arentl. . ~~g lat ntJsh and French statessonablv make war on us unaided'' dpp ) beueved that England could not "rea' an assumed that th p f 1at Eng1 and had abandoned its 0 . . e eace o Amiens guaranteed cl . d. ppos1tlon to France B H' 1 . e1 !eve m a bandwaO'oning world th . . . ecause lt er and :'\apoleon b 1 This situation is anal;gous to Rob~rt ~!~~~e~x~ess ~el~ eager to go to war. and the spiral model. The fonner calls foro s~ction between the deterrence model 51 for appeasement. Balancing and band P.po onthto a suspected aggressor. the latter wagonmg are e allianc a1 fd an d appeasing. See Robert Jervis p . . ~ eqUJ\' ents o eterring lies (Princeton, N.J. , 1976), chap. '3. erceptron and ~lrsperception ill lntemational Poli-


11. Quoted in C. J. Lowe, The ReluctaTlt lmperialist.s (!\ew York. 196- )' p. ;). I

credits it to Stephen Van Evera) in his Theory of I~ternational Politics {Reading, M ass. 1979). Amold Wolfers uses a similar terminology in his essay "The Balance of Power in Theory and Practice," in Discord and Collaboration: Essays on Intematimwl P olitics {Baltimore, Md., 1962), pp. 12~24. 2. Winston S. Churchill, The Second World War, vo1. 1: The Gathering Stonn (Boston, 1948}, pp. 207-8. 7 3. W. ScottThompson, "The Communist International System," Orbis 20, no. 4 (197 ). 4. ~or the effects of the Soviet pressure on Turkey, see George Leuczowski, The Midclle ~ rn World Aff irs, 4th ed. {Ithaca, 1980}, pp. 134-38; and Bruce R. Kuniholtn, The Orig~'~ a (If the Cold \ the Near East (Princeton, N.J., 1980), pp. 355-78. For the Nof\\le@; response to SoVIet pressure, see Herbert Feis, From Trust to Terror: The Onset of
1. My use of the terms balancing and bandwagonin o follows that of Kenneth Waltz (who


The Future ofDiplomacy



... Diplomacy [is] an element of nation~ powe~. T.he .i :nl~o.r~a.nce of diplomac:: for the preserYation of intemational peace lS but a l:artl~ul~n as~cct of that general For a diplomacv that ends in war has failed tn tts p nm ar:- objecthe: th . f unet10n. e I I. I , promotion of the national interest by p~aceful m~aJ:s: T 11 las i'l \vays been so and is particularly so in view of the dest:ructi:~ pote ntialities of total war.. Taken in its widest meaning, compnsmg th e whole range of foretgn policy, the task of diplomacy is fourfold: (1) Diplomacy must de termine its objectives in the light of the power actually and potentiaiJ)~ av~able for the pu~suit of these objectives. (2) Diplomacy must assess the obJectives of other nations and t11e power actually and potentially available for the pursuit of these objecti\es. (3) Diplomacy must determine to what extent these different objectives are com patible \\~th each other. (4) Diplomacy must employ the means suited to the pursuit of its objecthes. Failure in any one of these tasks may jeopardize the success of foreign policy and with it the peace of the world. A nation that sets itself goals which it has not the power to attain may haYe to fat-e the risk of war on two counts. Such a nation is likely to dissipate its strength and not to be strong enough at all points of friction to deter a hostile nation from challenging it beyond endurance. The failure of its foreign policy may force the nation to retrace its steps and to redefine its objectives in view of its ac-tual strength. Yet it is more likely that, under the pressure of an inflamed public opinion, such a nation will go fmward on the roa~ toward an unattainable goal, strain all its resou rces to achieve it, and finally, confounding the national interest with that goaJ, seek in war tJ1e solution to a problem that cannot be solved by peaceful means. A na?on will also invite war if its diplomacy wrongly assesses th e objectives of ?ther ~~ons and the power at their disposal .... A nation that mistakes a polic~ of ~mpenalis~ for a policy of the status quo ~ be unprepared to meet the threat to tts own eXtStence which the other nation's policy entails. Its weakness will invite . . attack and may make war mevttable. A nation that mistakes a policy of tJ1e statuS r the ..: 11 quo ror a policy of unpenalis m ww evoke through its disproportionate reactiOn r: t of . very clangerB war which 1 15 trying to avoid. For as A mistakes B's policy ror Jtn pe alis h both r mtght mistak A' d 1 . n m, so e s e ens1ve reaction for imperialism. T us d nations each . t m ent upon forestalling imaginary aggression from the other SI e, '
o f. From Politic8 Amo11g Naaom 5th . . . u:tibyon by H.~ J. Morgenthau, oopyright 1972 by Alfred A J{n ~. adi"ision of Random HOUse permJSSton of Alfred A. Knopf, a division of Random, '


to~ a will n l'>:l l 1 .rm!-.. Similarly, the eo f . . . f. n USton of one tv . I . 1 1 ' . I spl oportionate reaction and th , 'Pe o Impenahsm with another may ea As t r 'Is< assessment of the po . f us e\oke the risk of war. .h v. er o other naf I)e equally fatal to the cause of IOns, eit er to o,errate or to underratC' lt may B, A m a~ prefe r to ~'ield to B's demands until fi pe~ce ..By overrating the power of di~. nall), A 15 forced to fight for its ven. existenc< undN the most unfavorable . h con tions Bv und erratmg t e power of B d A ma, Ilec:ouw overconfident in its as superi tv sume ' . on ' A may advance demands and irnpo e conditions upon B which th I Unsu pectin~ B's actual power of resistan~e a~%~~. ~uppo edl~ too weak to resist. either retreating and conceding defeat 0 0 r' cl . . e faced ''-1th the alternative of r a vancma and riski a , t1 n " ar. d 0 A nation 1at seeks to pursue an intelli li gent an peaceful ~ :y oreign po c cannot f cease compaling its own obiectives and th 1. . e o )jectives 0 othe ti . th li J r na ons m e ght of . bl their cotnpatibiJity. If they are compatible n0 . t pro em anses If thev a , , re no compati. al ble. nation A must determine whether its obiecti to itself tl t th ves are so '-'1t J ey must )a 1 . be pursuec desptte that incompatibility\ 'th th 0 b . e Jectives of B. If it is found that Ns ;\1} ,1 . . . . b . c VJtct mte1ests CUl e safeguarded without the attamment of these obiectives thev ' b' . J 1t )e oug1 to 1 aban doned. On the other hand if A finds th t th a ese o Jectives are essen ' aJ r: 1 tic. tOr tts VJta mterests, A must then ask itself whether B's ob. . mc.:ompat:ible ~ect:ives . . w1th tts own, are essential for B's vital interests If the anS\,.er seem to be the negam . . . a1 ~~ tive, A must try to mduce B to abandon its obiectives, 0 rrenn g B eqw,a~ents not '1.t m J _ to A. In otl1e1 words, through diplomatic baraainina the o;,e and take of compromiSe. oo o . . a way must be sought by wh1ch the mterests of A and B can be rec-onciled. . Fi~ally, if the i~compatible objectives of A and B should pro,e to be ,;taJ to etther st~e. a way m1g~1t still be sought in which the ,;tal interests of A and B might be redefined, reconctled, and their objectiYes thus made compatible \\ith each other. Here, however even pro,ided that both sides pursue inte!Jjaent and peaceful policies-A and Bare moving dangerously close to the brink of war. It is the final task of an intelligent diplomacy, intent upon preserving peace, to choose the appropriate means for pursuing its objectives. The means at the disposal of diplomacy i'Ue three: persuasion, compromise, and threat of force. ~o diplomacy relying only upon the threat of force can claim to be both intenigent and m peacefuL No diplomacy that would stake everything on persuasion c d compromise desetves to be called intelligent. Rarely, if ever, in the conduct of the foreign policy of a great power is there justification for using onJy one method to the exclusion of the others. Generally, the diplomatic representati,e of a great power, in order to be able to serve both the interests of his country and the interests of peace, must at the same time use persuasion, hold out the advantages of a c-ompromise, and impress the other side wit.l1 the military strength of hi~ count:J:. . The art of diplomacy consists in putting the right em~>hasts at any particular moment on each of these three means at its disposal. A diplomacy tltat has been successfully discharO'ed in its ot.ller functions may well fail in advanc~g tbe national 0 f t interest and preservmg peace 1 1 stresses .persuasion when d1e gt' eAand 1 . of di take . 1 P ?ma~ compromise is plimalilr required b\ t.l1e cJJ'cumstances oft1e ~. n'se when the milit~n: m1ght of f , k ' th at r)uts most of its eggs in t11e has et o comprot 1 . -1 ln: 'lit> nught when the 1 1 the nation should be predominantly displayed. or stresse.s ~ -,, . promise will bke\\lse frul. .. d . . .. a] . poIJtlc sttuatton calls for persuasJOn an eon1



Pr\RT 1


S'l 1

The Promise of Diplomacy: Its Nine Rules

DiplnnHtl'\' could rt.>dHifit\\ouldpart with [the.~ \ i e. t;l'lt in rc<:ent ye . . . . . cl .f . . I . . l ars ha wll1-m~h destnwcd tts usefulness, an 1 tt ''on <.t r~ vut<..: 1 H. technique .ve . . I' . . . . '. . ; . s whtch haH~ controlled tlw mutual relations o nation c; ~lllet tnh mmernonal. By d . so. howc' er. diplomacy' ould realize onl~ one of th<.' P1 ,..' '' 011 ditions fo r the reOtng vntion of {1C:lct'. The <:ontribution of a re\'jved chplot l! K\ to the cause of ser. t peat-e f 1 - use. . . would depend upon the methods and purpose~ o \\"e have aJ read" formulated the four mam ta k' w1Lh wht ch a foreign oJ must cope successfullv in order to be able to promott the national interest an~ lt)' setYe peace. It remai;s for us now to reformulate those taskc; in the light of the ~re. "t" Id po)1 t c~. con front d"tp lornacy.... spe. cial problems with ,.vhich contemporary ~vor The main reason for [the ] threaterun~ aspect of contemporaty world politics [lies] in the character of modem war, wluch has changed profoun~ly under the impact of nationalistic universalismo and mode~n technology. ~he effe<:ts of mod. ern technology cannot be undone. The on I, L~1at r~n~ams ~ubject to delib. erate manipulation is the new moral force of nat10nal~stJ<: umversalism. The attempt to reverse the trend toward war through the tecl~ntques ~fa revived diplo. macv must strut with this phenomenon. That mean , m negative terms, that a re"i~ed diplomacy wi.U have a chance to preserve peace on I> when it is not used as the instrument of a political religion ai_ ming at universal dominion.



n I tiJstrusc in its ten n~ 1 .. .t . IS metaph)-sical . a1 Cllll .ancI tttalfairsofme n . all condt" I I assertJon. Itisnevertn 1c. bccaus\ . i~ are d 1t

poht ., Jll'>l think what an abommat1on in 1t . lOne and relative x"'" toturn 1 k lo ~ "" >ac any L. I>OhlJ< I u1or editor can ' at moment p ts atecraLt an abstract doc....:n' IIIUS t 11e 1\ n}' " .: in tht doctrine and applaud it hccause t'hev hear extension on it The """pi" u a new . .. . . the 1 . r- '" acqu1esce tht pohh<.1ans and editors repe at '" because the po lti<:Jans and editor.. rcntat t t I . th k . . tI ' anc ma) llll'all anythu1g or nothimr t Y 10 Jt s popular. So it grows . o a anv moment cl Jt actC'd< to 1t now. v..~thin the vague 1 ~ of w iat ' an no one .knows how it will he. you . . 1m1ts 1 have t(J accede to 1! tomorrow wh tl you suppose t to he; thcrcfore )'Oil ,.~11 en 1e same na d ~ ~ou ll<:"cr have heard or thought of. lf ou allow me~. rna e to cover somctJJing whidl you Wl ll awaken some day to find .t st Y_J l a pohtH.:al e:at<:hword to go on aucl grow . I anumg over you th . I . . wh1c 1 V0\1 are powerless a~ men. ear Jlter of your <.lestiny against ' , are power1 a<Y . t cl 1 . ess contrarY to sou nd statesmanship and oru.ns e USions.. .. vVhat can be more . I~ I tJon w 11<: 1has no definite relation to. common sense thaJ1 t o put r0 rt I1 an abstract a~scr. 1 any mterest of ou , k any number of possibilities of prol ,. . . rs now <lt sta e, but which hao; in it uucmg comp1 teation h" I which are sure to be embarrassing \vh th . ~ s w .JC 1 we <:an not foresee, but ' en ey anse!"

Four Fundamental Rules

Diplomacy Must Be Divested of the Crusading Spirit This is the first of the rules that diplomacy can neglect only at the risk of war. l n the words of William
Graham Sumner:
[f you want war, nourish a doctrine. Doctrines are the most frightfu l tyrants to which men ever are subject, because doctrines get inside of a man's own reason and betrayhim against hiJnself. Civilised men have do11e their fiercest fighting for docttines. The recon quest of the Holy Sepulcher, "the balance of power,'" "no universal dominion.'' "'trade follows tl1e,'' "he who holds the land will hold the sea," "the throne and the altar,''.clle revolution, the faith-these are the things for which Jllen have given their lives ... Now when any doctrine arrives at that degree of authority. the name of it is a club whicl1any demagogue may S\ving over you at any time and apropos of anything. In ?rd~r to describe a doctrine, we must have recowse to theologicaJ language. A doctnne tS an article of faith. ft is something which vou are bound to believe, not because )rou ban~ some rational grow1ds for believing it ~ true. but because you belong to such and s_uch: 11 church or denomination .... A policv in a state we can understand; for instance, ' ".1) the polk:y of the United States at the end of the eighteenth century to get ~e: navigation of the Mississippi to its mouth , even at the expense of war with S~~~l. fortll policy had reason and justice in it; it was founded in our interests; it had positive. its and definite sc'Ope. A doctrine is an abstract principle; it is necessarily absolute tn

The Wars of Religion have shown that the attem t . , . gion as the only true one upon the rest of the world is~ ~t~~po~e .ones own rehtury of aJmost unprecedented bl dsh cl, cl . e as tt ts costly. A c.-en. . oo e evastahon, and barbarization was neede d to1convmce the contestants that the tw0 retlgJOns couJd live together in .1 .. mutua to eration.. The two political religions of our nme have taken the place of . t he two great Chnshan denominations of the sixtee th d . ill h 1 1 . . n an seventeenth centunes W t e po ttlca rehg10ns of our time need the lesson of the Tb"m; v \e d tJ 1 . . 1. ., ,ears var, or 'vill t]1ey n 1emse ves m time of the uruversalistic asp1ranons that tablv tssue ' 1 meVI m mconc us1ve war? ' Upon_ the an~ver t~ that question depends the cause of peace. For onl)' if it is answered m the affirmative can a moral c :onsensus, emerging from shared convictions a~d com mo~ values, develop-a moral consensus within which a peace-preserving diplomacy will have a chance to grow. Only then will diplomacy ha,e a chanc-e to fac..-e the ~oncrete political problems that require peaceful solution. If the objectives of fore1gn policy are not to be defined in terms of a world-embracing political religion, how are they to be defined? This is a fundamental problem to be solved once the crusading aspirations of nationalistic universalism have been discarded.

pc-~ and to eac:h nation's claim that its own ethical code would serve as the basiS of
ocmduct for ..JI nations.]

Edito~ ~otc: By this tenn Professor Morgenthau refers to the injection of ideoloro: into ~n:::~

The Objectives of Foreign Pol cy Must Be Defined in Terms of the National i Interest and Must Be Supported with Ad.equate Power This is the second rule of a peace-preserving diplomacy. The national interest of a peace-lo,ing nation can only be defined in tenns of national security, and national security must be defined as integrity of the nationaJ territory and of its institutions. ~ational security, then, is the irreducible minimum that diplomacy must defend with adequate power 'vithout compromise. But diplomacy must ever be alive to the radical transfonnation tl1at national security has undergone tmder the impact of the nuclear age. Unti.J the advent of that age, a nation could use its diplomacy to purchase its security at the expense of another nation. Today, short of a radical change in the atomic balanc:e of power in favor of a particular nation, diplomacy, in order to ~-ake one ~tion secure from nuclear destruction, must make them all secure. \\ 1tl1 the national

.....~ . I trictive and transcend.'JJl 't r 11 cliplomacy 1 inll"'n.''-t defincu lJl 'IUt 1 re , nu,t oh el"\e the third of its rule T _ ,_ r 1 the Political Scene from the fPoint of Vieu; 0'J Oth er Diplomacy .\lust VJOI\ a ,. . .t.. : . f: .._. to a nation as an extr<:Oit: <1 <if:'1 -partJalitv. and th \ ations - \ ou una IS so aUSJ all I ,. J , e ' . 'd . of what otJlers wiJJ natur ~ sopP Or tear. \\ nat are th o total ""ant of <.'On ' eration [ . I . cl th . f otJ nations in tenns o nationa S(Um~ an are e\ comp tc a 11 ationalwterests o 1er . , ible\\1thonesown.? Tl definition of the national.inte rest 1n t<'m1~ of national ~ec:u1e . . . . d tJ terests of ilie two opposmg nallous arc more likely to h nh IS easJer. an le 1.0 f lh al e ' tibl : . b' 1ar svstern than in anv oilier sy te m o e n ance of powE: compa e m a po ' . r f th . r. . l ..,.,. m as we have seen, is more unsa1 rom E: pomt of view of e The b1 ar S\ ;>Le , po .. th ace than an'. other, when boili blocs are in competitive c:ontac:t . roughout th : . ] f pe ' b'ti'on of botJl is fired b\ ilie CrtU arun~ zea 0 a UntversaJ mission worId an d tbe am I ' . . a1 . .. .. v 'tv or nearness of situation, c'Onsti tutes nations natur enemJes. <~ vJClnl ' ' t11 . et once ey have defined their national in.terests in terms of national seeu. . b ntv. t. ev can dra"' back from their outlvmg positions. located close to, or within . "' th~ sph~re of national security_ o~ th.e oth~r s1de, and ret:reat m~o. the1r respe<.:~ve spheres, each self-contained w1thm 1 ~rb1t. T~1~se outlymg positions ad~ nothmg ts. to national seeurity; iliey are but liabilities, po~1tio~s that cannot held m case of war. Each bloc will be the more secure the w1der 1t makes the distance iliat separates boili spheres of national security. Each side can draw a line far. rustant from each other. making it understood that to touch or even to approach 1t means war. \Vhat tllen about the interjacent spaces, stretching behveen the hvo lines of demarcation? Here the fourth rule of diplomacy appli es.




automatic ppl't:dtion. lt i~ onk th gh . t. h , TOUCT a COntin supportc<- .;otts -'' Rrmness and self tr . uous prex:ess of ~\..~tation ' -res runt, that . . '-"'t' issues Ut'l be made to work. It is h01 . . . compromiSe on seconda.t-v . f. j . , Jroa<. he~ \\' . . act 1tate or hamrv-r. th"' \:\er. possible to 10dicate a priori what f arl , . r ..._ success o t : f Fir t (Jf aJ], tt lS worth notjng to what t thpo ctes 1J compromise. is. compliam:e ,,;th the fourth mle-de ex ~t e success ~f compromise-that three rules. which in turn are similarlv intpend ufn comphanc:e with tl1e other . .L " er . . second ru Ie cIepenu.s upon the realization ofepenuent.As the ~mpliance \\lth the th fi its realization from compliance with the sec .de rst. ~ the thJrd rule must await . :on . \iew of . nation al . Its mterests after it has parted . A nation. can on!~. take a ration al a poutical creed. A nation is able to consider thcompan\' ''lth the crusadinu spmt of ti 'a] . ,,~tJl objecti\ity only after it has become secure~ na haton. mt~~ests of the other side . C . mw mterests. omprom1se on am ISsue. howe\'er mino Jt constders its own national 'bl . .L , r. lS 1mposst e so lon(l' as both sides are not secure m UJeir national interests Th ti . r h rul . us na ons cannot hope to G'Om 1, w1tl1 the rourt e 1f they are not '"illina to corn 1. th h h p~ . di o P' ot morahty and expe ency require compliance with the,se"-'1 tfi e da er three, Both r tal 10 ur un ' men ru es CompJj ance makes compromise possible but it does t . no assure tts success T0 give compromise, ~ade possible through compliance \\ith the first three rul~. a chance to succeed, five other rules must be obsened

Five Prerequisites of Compromise

. Give up the Sh~doU: of W~rt~less Rights for the Subst.ance of Real Advantage A ~~p~omacy that thinks m legalistic and propagandist:ic terms is particular~ tempted to ms1st upon the letter of the law, as it interprets the law. and to lose sight ofthe consequences such insistence may have for its own nation and for humanit\. Since there are rights to be defended, tllis kind of diplomacy thinks that the issue ca'nnot be compromised. Y the choic-e that confronts the diplomat is not bet\,een legali~ and illeet gality, but between political wisdom and political folly. !he question ''ith me.~ said Edmund Burke, "is not wheilier you have a right to render your people miserable. but wheilier it is not your interest to make them happy. It is not what a lawyer teUs me I may do, but what hu manity, reason and justice tell me 1 ought to do.-6 Never Put Yourself in a Position from Which lor1 Camwt Retreat Without Losing Face and from 'Which You Cannot Aduance Without Grot:e llisks The violation of this rule often results from disregard for ilie preceding one. A diplomacy that con founds ilie shadow of legal right with the actuality of political advantage is ukely to find itself in a position where it may have a legal right. but no political business, to be. In other words, a nation may identi~ itself with a position, which it may or may not have a right to hold, ~ardless of the poli~cal consequences. And again rompromise becomes a difficult matter. A ~tion cannot retreat from that position \vithout incurring a se?ous ~~~ of ~restige. It can~ot advance from that position wiiliout ex'}X>Sing itseli to_~litical mks. perha~ :\e~ the risk of war. That heedless rush into untenable pos1tions and, more pamcularl}. the stubborn refusal to extricate oneself from them in time. is ilie earmark of . ~apoleon Ill on the mcompetent diplomacy. Its classic examp1 are Le policv of es m ,

Nations Must Be willing to Compromise on All Issues that Are Not Vital to

All government, indeed every human benefit and enjoyment, every virtue and e\e~ prudent a<..t, is founded on compromise and barter. We balance inconveniences; we give and take; we remit some rights. that we may enjoy others; and we choose rather to be happy citizens than subtle disputants. As we must give away some naturallibe~es. for the advantages to be derived from the communion and fellowship of a great emprre. But, in all fair dealings, the thing bought must hear some proportion to the purchase paid. None will barter away the immediate jewel of his soul.5

Here diplomac:y meets its most difficult task. For minds not beclouded by the c:rusading zeal of a politicaJ religion and capable of vie,Ying the national interests of both sides withobjec:tivity, the delimitation of these vital interests should not pro,~e too difficult. Compromise on secondary issues is a different matter. He re the tas~ not to separate and define interests that by their very n ature already tend to~ separation and definition, but to keep in balance inte rests that touch each oth~r at many points and may be intertwined beyond the possibility of separation. It lS an immense task to allow the other side a certain influence in those inte rjacent spaces wi~t allowing them to be absorbed into the orbit of the other side. It. ha:~: less unmense task to keep the other side's influence as small as poss1ble . r~ .. --- to . . ""'lfr' .., ~ ones own se<..'llnty zone without absorbing those reg10ns 10 to ones [l own orbit For the perfonnance of these tasks, no formula stands readY or





. \\' f L870 .111 d the poli<:ic:> ol A u:.t: i:. <UH.i Germ a t ' H ' ol the l nuK -Prnsswn o ar o ' . . . .. .. 1 . . 11)' . . . I I ,, .. . Tl1cse ex<1111l)Jcs also~ 1 n_. .. r- O'l 1 t h~ risk f Hh\ : nn tht' l' ' 'l' of the Fu ~ t \\ ot ( a t. o ".tr is ~tllil'd " i th the violation of this rule.


lJ l


hJStOty, and beyond th jncak ulahlc r)u,Sl.Jihties of the future In th e V ds of tomorrcJw it anticipat tl tctory . .. e wor of 13ol b k es w mg ro e:
< 'i L
r) ('

a fl


k All to Make Decisions f or You S IC .n::!_ 1utions that are \ eve, Allow a ' ea_ _ . ., particularly ~usceptible to vi )l<ttiug th is 0 11 Y es e e. b l' . . to tl1c l)rececltngf I u1. ellb identifying tlt<'i 1 own nation.lI in lcrcsts (;On) 0 l\10US Y "' 1'] . lose the.i r freedom o adJOn Secure in the support of Its powcl{ ll llrienu tl1e .le: I 11 1 1. m tl ose of the wea , a v. . . p ete : W l 1 . the ob'ecti~es and methods of ilS (orel~1 poltC:)' to Stlil itself'. weak ally can cho.ose (i J I that it must support interest!. not it U\-vn and that it 1 The powerful natiOn t_len t~c s s that are , ita I not to itself. hut on I~ to its alJy. . able to comprorruse on rssue < t fi l. ts un . exam )le of the violation of t11is ru le rs to )C oun<. m the way in The classiC di h 1. d of Great Britain and France on tlte eve of the which Turkey force t e Jan d 11 - 1 . . 8 - 3 Tl Conce1 of Europe ha "' rtrt< )' ag1eec upon a comurot l Cnmean\Var m l b l C k 1 T k L . r . tJ fl ' t between Russia and Tur -e~' w 1en ut ey, K nowmg that mise settling l e con l C . I :> d'd . . 11 supr)ort it in a war w1t 1 h uss1a, 1 Its 1 )est to provoke the \Vestern powe1s wou t . . . . . . . . h ._ . . ve that war an d t lms mvo 1 d Great Bntam and France 111 rt a.gamst. t en wtlJ. Th us . . . cJjna 1 Turkev went f:ar Ln dec1 o the issue of war and peace for Gdeat Bntarn and . . ' cJj to ts ,.vn national interests. Great BntaHl an F rance had to France accor ng 1 0 c] an accept tha t decJSI even tJ1ough their. national mterests di not requrre war with . . . . . Russia and d1ey had almost succeeded m preven:mg 1ts O~ttbreak. The) had s_urrendered tJ1eir freedom of action to a weak aLly. wb1ch used Its control over tl1err policies for Hs own purposes.
. \li

11ere ltt u on!~ "ay. that the glory of taki 11recl b\ tl11 Jtility that rcs11l ts frorn tl' o n~ tov,~s. and winning battles is to b . . ' . e arms. u1a\ lmng si mrne to the coun<:ils se vtc:tones v tones that brino, honour tme~. . t<: tL . o1a natron . To . 1, o o ne a W ln glory or a _gc>JJC'ral, anti of an army. . . . ut the gIow of a n c\ ),lttle, to take a town 's the . . . " u she proposc ~s. to Iter mte rcst anu her stre th th ' atlon IS lo proportion the ends , and t 1 VJgour she exerts toboth . ; e means e ng propost"s, 1c si crnp1 to the ends heoys 7

The Armed Forces Are the Instru:num.t of Foreign, Not Its Master No successful and no peaceful foreign policy is possible 'vithout observance of this rule. No nation can pursue a policy of compromise with the military determining the ends and means of foreign policy. The armed forces are instruments of war; foreign policy is an instrument of peace. It is true that the ultimate objectives of the conduct of war and of the conduct of foreign policy are identical: Both serve the national interest. Both, however, differ fundamentally in their im mediate objective, in the means they employ, and in the modes of thought they bri ng to bear upon their respective tasks. The objective of war is simple and unconditionaL to break the -..vill of the enemy. Its methods are equalJy simple and unconditional: to bling the greatest amount of violence to bear upon the most vulnerable spot in the enemy's armor. Consequently, the military leader must think in absolute terms. H e li,es in t11e present and in the immediate future. The sole question before hi m is how to ''~n victories as cheaply and quickly as possible and how to avoid defeat. The objective of foreign policy is relative and conditional: to bend, not to break, the will of the other side as far as necessary in order to safeguard one's own vital interests without hurting those of the other side. The methods of foreign policy are relative and conditional: not to advance by destroying the obstacles in one's way, but to retreat before them, to circumvent them, to maneuver around them, to soften and dissolve them slowly by means of persuasion, negotiation, and pressure. In eo~,. sequence, the mind of the diplomat is complicated and subtle. It sees the issue Ill

To surrender tl1e conduct of foreign affai t .. 'l) .,. rcompronllSe and thus surre rs 0..Lt1 mtlttarv then, .ts to destrov . 1e the poss1 I tty o cl ,, n er ure cause 0 f h , mind knows how to operate bet\veen the ab 1 t f . peace. T e militarv . so u es o v1 ctory a d d ~ , noth ing oJ that patient intricate and su btle m . f n e eat. It knows aneuvenng 0 di 1 , h puq)ose is to avoid th e absolutes of victory and cl f cl P om,tcy, w ose main the middle ground of negotiated comi)romise A f~ e~t an 111 eet the other side on . . JCY tarv men accordina to the ru les of the miJjt . orergn I10 1 c:ond uct ecl 'bYmu J0 ; aryart can on lyendin w f .. h prepare for is what we shall get."S ar. or w at we For nations conscious of the potentialities of m 0 d c . em must be the I . 10re1gn po1. . Foreign policv must be cond war.dpeace h f t 1elr goal o ICies. t . ; uc e m sue a \V as to av make the preservation of peace possible and not make th b ak 'r bl e. 1n a society mev1ta of sovereign nations military r0 . e out re o war . . ' I' rce IS a necessary mstrument of foretgn policy. Y the instrument of forei!m policy should t b ' et c . . . o no ecome th e master of 1 gn pohcy. As war is fought in order to m"ke peace poss1'bl r 0re1 . . " e, rore1 gn p~hcy should be .conducted m _o rd~r to make peace permanent. For the performance of both tasks, the subordination of the military under the civilian authorities which are constitutionally responsible for the conduct of foreign affairs is an indispensable prerequisite.

The Government Is the Leader of Public Opinion, Not Us Slave Those responsible for the conduct of foreign policy will not be able to comply v.ritl1 the foregoing principles of diplomacy if they do not keep this principle constantly in mind. As has been pointed out above in greater detail, the rational requirements of good foreign policy cannot from the outset count upon the support of a public opinion whose preferences are emotional rather than rational. This is bound to be particularly true of a foreign policy whose goaJ is compromise, and which. therefore, must concede some of the objectives of d1e other side and ghe up some of its own. Especially when foreign policy is conducted under conditions of democratic control and is inspired by the crusading zeaJ of a political religion, statesmen are al~ays tempted to sacrifice the requirements of good foreign policy to the applause of the masses. On the other hand, the statesmen who would defend the integrity of tllese requirements against even the slightest contamination 'vith popula: pass~on wo~Jd seal his own doom as a political leader and, witlt it, the door~ o~ Ius foretgn policy, for he would lose the popular support which put and keeps lum 10 ~wer. The statesman then is allowed neitl1er to surrender to popular passwns nor disregard them. H~ m us~ strike a prudent balance bet\.,.een adapting himself to

11 2



ea<.:<: JTI II"l


I . 1port of his polic:its In "' ''<" d. lH lllu~t 1 them and tll:trshaling tll<' 111 10 1C sl~'l t 1 ttlo!-.llla ll 'iltil> trirnnr11 rg lw. '>ails to tl . . . I . I . rh est ea 0 s . ~ . I( l<.'ad. 1le musl peri O l l htl ug . I t- J ctr rv tlw sl11p lo tl p t o( ~<,od 1( , tll . I . .. whil<' usmg t tCtrl < ' 1 r\\ind!> of poptl ar p.L 'iston I I t. 11,- ' zig'l-<J~' a ccJII nc. " l'ign poIit) , on I10W(~ v._ r rount a >Oil u ,..,


. irt. . tl peace wIHCI1 w,. havt' outl i11r d <:aJII Int C'll111p!'lc in inc;J> ' The: road to tntem.ttiOilc . I I 1 ci ll'llinr1 fcmn11 lac that lo1 a <c ntnry and a : anc n . I. . . Wl'tl1 the. sunJ'> e f .as veay world. Tll<:'n' J'> so111d I11.11g SJ1t'cta<:utional qua 1t1es . _ f'! 1 1 amna11on o a war ' half11aV<' nrcc l 1c un t; f r . I tllat with one swec M'<ms to dispose' of tk p . I . 11ic.: lty o a tor n" a . I I I1C ra d13 sunJ Jar in t . . . II Tl . 1as bee 11 the- promise of Sll<' 1. . o ut io11s as fre1 . r onC'Cand lm .l . liS 1 I I pro ble n1 of wa t . llm..:ve securi ty, uni v< rsa soeia iSIII, i 11 ter ' to ......~, . . l> 1011 tlisanllamcn I' state The re is nothmg spcdat11lar, !ascinat trade, ar , . ent a c t 1 wor u . 1 . . . 11 1 11 11 nat10na govcllll ' r . t hrge in the busiii< o! <1tplotlla<.:y. 'SS . . , 1t least 1or t 1C peop 1 a < l e ' 111 ng, ' fn g, or lliS P . ,ver th at tl1 solutions, inso far us they deal with esc We have rnade t1 point, 1 te towc , I .."' oms, t I not mere y with some of .its SYl npt11 I presu pposc .,..H'Iexisthe real problclrl ant '. . .. . .. t'10nl soci e~v wh1ch act ua y < ocs not cx1 1 o >ruw st. . . of an integratccI mtPrna ... ; o lcrH C ng, acc:ommodating 1 an .11 tern atoml society andketT>itinbeicl lllc . . . . ' . . into exjstenc.:e sue 1 I I r om<st1c soc1ety .tnd 1ts '. .,. . te<:hnic] IIC of UIP1 S omac.y are re quired. As the Jlltegration. o I . . r tl e spcctacular and al most unnotiecd clay-by-t ay operations peace deve Iop I rom 1 1111. I. . I I r. c of t IIC' teeI1mqut:s o f ac:c"'mt-.-.(>c.l:~tion and chanue. so. the u tun ale 1 ea o lnter'' " < n . . . a11r , . ts nat1on uc--tl 1at rs, I<> trms'end itself in a supranatJon al soecty- must awa1 t 1 ' '" rcali;.-.ation fro 111 the techniques of persuasion. negotiation, and pressure, which are the traditional instruments of diplomacy. . The reac.ler who has followed us to this point may well ask: But has not diplomacy failec.l in preventing war in the pao;;t? To that legitimate question two answers can he given. . . . DipJoma<.:y has failed many times, and it has succecclecl many tun~s, 111 1ts peace-preserving task. It has failed sometimes because nobody wan t~d ~t to sut:ceed. We have seen how different in their objectives an cl me thods the l1 mLteu wars of the past have been from the total war of our time. Whe n war was the normal activity of kings, the task of diplomacy was not to prevent it, but to bring it ahout at the most propitious moment. . On the other hand, when nations have used diplomaey f()r the purpose~ ol preventing war, they have often succeeded. The outstanding example of a successful war-preventing diplomacy in modem times is the Congress of Be rlin of 1878. By the peaceful means of an ac<.:ommodatiug diplomacy, that Congress settled , or at least made susceptible of settle ment, the issues that had separated Great Britain and Russia since the end of the Napoleonic Wars. During the bette r part of the nineteenth century, the conffic:t between Great Britain and Russia over the BaJkans, cl1c Dardanelles, and the Eastern Mediterranean bung li ke a suspended ~word over the ~ ~f the world. Yet, during the fifty years fo11owing the Cri mean War, though hostiUties between Great Britain and Russia threatened to break out time and again, they never attually did break out. The main credit f()r the preservation of

gt~l collar l<dn rirjl iC ol' a 11 at:<:Oifllnod-t' d' I :S p tile C lllt)4n '""II lkrli n. Whctr Brili-;1, Pri mc. ,vaJ nt~ler 'P.rOIIIa<:ywltidt c.:uhninat<:d \A~ '.n~ D ~r . . 111 I 1 cnc~s lo I .o11c 011 < clcd~1 rcd with pride: tl . t h - a<.: 1 returned fro1 that n 1 ,., C() I f' I I . la C: W<t\ hrinv t I .. th ho1 a or. In .11'1 w 1ad l>r<)IJght I)cac:e: fc)r lat . l')tng 1omc peace ... 1 1 "" " J a..., !we 1 1111 var 1 JCtw<c Cr<:al Is n r1 . d er generaliCm<;. l O<J; r10 r a centurv " 't , R . 1 t Il er" <I d ll IISSta " \ Vc have J.m, r, cr ntognizcd the precarious , . .f . . ign natio11s .I I 11' co11l . 11 i11 g Sll <:(e~s of dip le ,.. ncss o peace.: tn . s<X;H;ty Cl f sov111 . a c re lt..a<.:y 10 preserving ... , d d . WC' !la"< \rcll "JlOII extrar>rdinarv Jlloral att'J . t 11 ' al pe:acc < n ~. :pc a.'> ' " m t e<.:tu rtualir , t1 11 1 leading part ictpattl!. 1 11 11~l pos~<;sc;. A llli-;lakc in tl e . al . 1 tes tal a l 1 e . . . I cv tnculs 0 f. naliCJII.t I powtr, 111atIt I> one: or tl 1 other of tl uation. of one of the elt-y e I .,. 1 l C: C:a( mg Iif ('I'<'JI(;( >t'I W('(' II pta<.: e: anu war. So ma)' an .... d. 1 . '\ta t C:S in(ll . may SJclllllc ( ) 'I' powe ac.:<.:t tn l sp01 tng a plan or a r ca1 1 o.n. cu at1 . Dij11 omac:y ~~ tiK hcslneatl!'. of p re:scrvincr pea<.:e \''1 I . . . .. . ,.., ' 1lt: l a sot:tety o1sovcretgn nations ha~ to offe r, b11L, espcctally under the <:onditions .,f,.0 t Id _ . . . . ' ... 11 emporary "vor po 1 itics and of conle ~p<>r ary war, .Ills not good enough. It is only when nations lmv<' surrendered L lughe r <.1 11lhonty the means of <.lestru<.:tion whi<:h rMdc techuoloa ;rn ogy h<<; put i11 th eir hands- when they have given up their soverei~nty-lhat inte:rL national p<-> ilC <:an be ~~ t<~Ue <'i sct:ure as domestic peace. Diplomacy <.:an make (; L peace more: secu r: t!.tLn 1t rs today. and the world state can make; peaec: more set:ure than it would be if nutions were to abide by the rules of diploma<.:\'. Yet, as the:re can be no permanent peace without a world state, there:: can he no ,~orld stale without the peace-prcse:rving and community-building prcx:esscs of diplomacy. For tltc world state to be ll1 than a dim vision. the accommodating proce ses of diploorc macy, mitigating anc.l minimizing conflicts, must be revived. \\11atcvcr one's conception of the ultimate state or in ternational affairs may he, in the rcx:ognition of that need a11d in the de mand that it be met all men of good \\.ill can join.

11 3

l. We by no mcans intend to give here an exhaustive account of rules of ruplomacy. We propos1 to dis<:uss on ly those which seem to have a special bearing upon thr contempo rary situation. 2. "War." Essays of \Villiam Cralwm Sumner ( 1ew Haven, Conn.: Yale University Press.
1934), Voi.I.pp. IC:>9fl'. . . . ~ . .. 3. F:dmund 13urk<.!, "H<111arks on the Polic:v of the AI lie!~ w1th Respect to Franee ( l t93 1 Works, Vol. IV (Boston: Little. Brown at~d Compan~' 1889), P 447.

1'he Fedemlist. No. 6. . . .. . \ l 169 0 1 11 ~Edmuud 13urke "Spee;ch on the Conciliation wtth Amenca, loc. cri.. , 1 11 .. , ' ' . ~ Speech 0 11 Conc.:iliation w1th the Co1 . .. ( 1- 7,...) The Wm*~ urclmwul Bu,-,.;e. \ o . omes 1 <> 'J (Boston: Little, Brown and Company. 1865), P 140 , b . C b d Univcrsitv 1 7. Bolit1gbmke:fi De.fc11se of the 'freaty of Utrecht (Cam ne gc: am n ,ge ' Prtss. 1932). p. 95. 8. William Crahan1 Stumwr, op. cit., p. li3.
4. .S. 6.



The uses and Limits of

International Law

111 ust
mi sp



. a1 I 10 f 111t mation aw " ,1 extmines its . functions in the present dent o e d . . 1 r ern IJolicy' of The stu states \\1ll, unl ess he lakes refu ae in . t ti al system an m tne ,oreib. m em a on . r. a1n. that the pure theorv o{' Iaw once pro,ided be th fortinO' seclusion ,rom re 1' ; e corn b f ...,tu~ s He will becom e a cvnic, il- 1 c1 scs to stress, 'like 1e wo reduced to one o tl ree alll ce 1 '.I , . ~ 1 Gat the waY in leg<u cl,ums. ell.e s1. ed to support 1.1p Cir.mdotLx m1 wer at t le es. 0 . I al . . r . d s seful or necessarv on non eg g10 uncIs. 01 1 I1e gets fas. an)' position a state eem u . . b' ti of cacophony and silence th at<:1 acten. ern ahonal 1ar zes mt 1 comf cinated by t le maldon blic order. He will become a h)pOciite. if he chooses to law as a S)'Stem o wor pu

e IlOW convergmg e ffiort destined to lead to some such system end.owed \nt h suffictent . . . stability and solidity, or else if he endorses one pruti~uJ~ constmctto 1 n. (t 1at of his own s~tesmen) as a privileged and enlightened contnbut10n to the ach ievement of such a system . He ,\il l be overcome by cons~e ma~on. if~~ ~ reflects upon ~1e gap between, on the one hand, the ideal of a world 111 wh1ch traditional self-h elp w1ll be at least moderated by procedures and rules made even more indispensab le by ~~ proliferation both of states and oflethaJ weapons, and, on the other hand, the realities of inexpiable conflicts, sacred egoisms, and mutual recriminations... . 1. Some of the functions of international law constitute assets both f or the policy maker and from the viewpoint of world order, i.e ., of providing the international milieu with a fram ework of pre dictability and with pro ced ures for the transaction of interstate business. (a) International law is an instrument of communication. To pre sent one's claims in legal tenns means, 1, to signal to on e's partne r or opp onent which "basic conduct nonns" (to use Professor Scheinman's exp ression) one considers relevant or essential, and 2, to indicate which procedure s one intends to follow and would like the other side to follow. At a tim e when both the size of a highJy heterogeneous international mHieu and the imperatives of prudence in the resort to force make communication essent ial and often turn international relations into a psychological contest, interna tional law provides a kind of common language that does not amom1t to a common c:ode of legitimacy yet can serve as a joint frame of reference . (One
. .

. . . .ona1 e1'tl1er tJ e conflicting inte'lJretations and uses o f 1 bv states as a aw rati 1ze som 1

Reprinted b d . Y~nnJsSJon fJrom lntenwtional Law and Political Crisis edited by Law ~thoa..,,dandWilkinson. eels.. pp. x:i-xix. Copyright 1968 by Uttle. Brown and Companrence Scheinm an y (Inc.). Portions 0 e text all footnotes have been omitted. 114

no guaran tee agaitl\t ndm tl 1 r,tJJCr side's res rJonse to tl cmg comn umc.:ated ntay we I I dcter,e message rr.. .. "th en ,ul un<.l(.m.tandin g of the situat th . we com IOn at t1 atens t~ . mun icate to 1 or goa b- h kc 0 11 r in ternretation of th . S1re l' c war tn outh v t tell )a'>H.: values aggrt'SSiOn-there wi JI be no J t frat n f r .Oin le nam a,<; cas of e o re.ercnce at all d a r e h corrqwlitioi t may be<:om e fiercer. ) an m act l t: (h' Inhmational law affords mems of cl . < . umneL contiu:t r cJ . mg inevitable tensions and clashes from the e rt t r 'J' _--o 1'er tm r so o rorce \\ h 1 g hav e bee n str ong ind epende nt reasons for a .cli . d en~ ver l lere . . 1 . mte rnatJOna svs tem m which the superpowevot ng arme conflict-in an , . r ." ~ . , . rs m parttcu1 1 e excelle nt reasons Ot manag mg them confrontation s eithe b k ar .lav . . . r v eepmg them non\10by usmg proXIes-inte rnational law has pr'o d len t, or cl t tJ1 . 1b r . t . r . e w1 )Oth ,11 IS 01 snunrung LOrce and wtth altematves t vl 1 s atesmen 1 . 1 both the So'1ets and the West shar)ed their moves no \10 ence... . In1Berlin . . ' 1 suc.:h a way as to eave to the other s1de_ f~ll res_pon sibiUty for a first use of force, and to ~void the kind of frontal colhs10n \:vtth the oth er side's legal claim that could have obl iged the opp~n ent t_o reso~ to force in order not to lose power or face . Tl1Us, today as m earlier penods , law can indeed... se1ve as an alternative to confrontation whenever states are eager or forced to look for an altemathe . 2. I nternationallaw also plays various useful roles in the policy proces s, which ho.,.vever do not ipso facto contribute to wo rld order. He re, we are concemed with law as a tool ofpolicy in the competition of state,isions, objectives, and tactics . (a) The establishment of a network of rights and obligations, or the res ort to legal argum ents can be use ful for the protection or enhancement of a position: if one wants to give oneself a full ran ge of means with which to buttres sa threatened status quo (cf. the present position of the \Vest in Berlin; this is also what tre aties of alliance frequently are for ); if one wants to enhanc e ones power in a way that is demonstrably authorized by principles in interna tional law (cf. Nasser's claim vvhen he nationalized tl1e Suez Canal, and Suk arno's invocation of the principle of self-detennination against ~lala~-sia) ; if one wants to restore a political position badly battered by an adversary's mo ,e, so that the resort to legal arguments becomes part of a strategy of restori ng the status quo ante (\iVestem position during the Berlin blockade: Ken nedy's strategy during the Cuban missile crisis; Westem powers attempts during the first phase of the Suez crisis; So. .iet tactics in the US . General Ass embly debates on the flnru1cing of peace-keeping operations). (b) In all those instru1ces, policy makers use law as a wa~r of putting pressu re on an opponent by mobili:ing international support be~lind_ ~e .~egal ~es invoked: law serves as a focal point, as the tool for ..intemationalizing a na~ ~nal interest and as the cement of a political coalition. States that ma~ ha,e_p olitical misgivings about pledging direct support to a certain power 'vhose mt e~ only pattJy coincide wit h theirs, or because they do not want to ant agoruze er tl1ereby, mav find both . 1d l"""ful to rallv to the defense ailO ther pow tt eas1er ai ...... . nciple in ' mamtenan v v of a legal pn whose ce 0 1 promotion the. ma, ha,e a stake.

l'H : r rc:me tnh ~r 1 that cor n . . ' ' muntcation 1 Cjtt lOn . and 2. that what is h . ~






b0 1 ores international la'' i<~ .-' tl1 e field < f . (C) A polk~ maker w gn gai-rnanipuJation optn t h (lPl)Onents , l t11roug1 e . 1or political-competattOn.d s one of tJ1e nu mer(lls 1 ~shoard.y 0n ri\ctlS. lntemational Jaw pr0\1 e which state contests occur. t 1. tJ1at to the statesmen iut rn aticmal la w IY t1115 10 dicates no on ' a guide for action. l11tl a).,o that tl1is tnol 3. Ob' ious h tb;n 'd nstrument rat er I . prova es an ' J rt to it wou.ld hamper t 1<: slat< s interest as is often not used. w Jen reso de fined b~. the policv maker. . ntemationallaw often serve-~ a' <1 techni<jue of .1 1 . . (. ) One of the reasons w 1~ a .. . .s th appeal of reciprocan.: ) ou mmt suppoti mv invoJj ca1 01obilizahOn a e ; . po ti . vou the be 1 agamst 1,-m. because if , let cl . mle I I .\10lated at mv cation of the ru e a b . . . . . . , be breached at yours; an we ol .l la\ c an mterest e>.'TlPilSC som eda} tt ma} . . ':'- ' . ., B t rermrocity cuts both ways: .\1)' u mg a certam legal m tts presen-atJon. u ~r 1 ,. case auainst him ma, encourage 11m, now or later an!Ument to buttress m, o l r b . ' tooresort to tJ1e same argument against me' I mav t 1ere1ore e umVJse to play f 1u criven on a c hessboard u1 ' vt.'ch , 0- the solemn and.abstract nature o legal rights . . . and obugations. I mar not be able to make the kind of distinction between my { good} case and your (bad) one that can b_es~ be made by resort to ~cl hO<:, litical and circumstantial e,idence that a melenu1t or ruled out m legal s ~mentation. Thus ... during the Cuban crisis, when the united States tried to distinruish between So"iet missiles in Cuba and American ones in Turke, in order to build its and get support, America's use of the OAS {Org~ization of American States] Charter as the legal basis for its "quarantine~ established a dangerous pre<:edent which the So,iets could use some dav. against the U.S. or its allies. on behalf of tl1e \\ arsaw Pact. And in the ~gioomedy of the battle over Article 19 of the {..;. ;\. Charter. one reason why the li.S. 6na1Jv climbed down from its high legal horse and ga,e up the attempt to depri,'e the SO\iets of their right to vote. unJess they paid their share, was the growing awareness of the peril which the principle of the exercise of the U.;\. ta\ing power by the General Assembly could constitute some day for the United States if it lost control of the Assembly. (b) One of the things that international law "communicates" is the solemnity of a commitment: a treaty, or a provision of the Charter, serves as a kind of tripwire or burglar alarm. \\7hen it fails to deter, the victim and third parties have a fateful choice between upholding the legal principle by aU means, at the oost of a possible escalation in violence, and choosing to settle the ~~te more peacefully, at the cost the Legal issue. For excellent poli~cal reasons, the latter course is frequently adopted ... in the form of droppmg any reference to the legal principle at stake .... {c) The \"ery ambiguity of international law which in manv essential areas ~lays either gaping holes or conflietin~ principles, all~ws poli'?' ~m _ emergency to act as if international law were irrelevant-as tf an it were neather a restraint nor a guide....



4. Tlw 1 "' L lo le:gal anrurncnts IJy 1 . p<J ICV mak 10 1 b worltI o~n I('r 1/ti(L f hPrehy counterp d ' . t;r<; ~ Y <:; detrirrumttll i f) ro .ur..tH.;e f or th arg111nc Ill\. J' e 'Ylate that ur,(;cl such (a1 j 11 the k:gaJ V(l(;lJ 1J m C>r COn f 1 . I IJ.\ton Wrtt<:h pr<: ,~ .1 . . tates a<; 111trma war or th<: 11 c;e o f c . L ' al s m ~rea.s a.s vital to . J OrC<:. eacn st t . t . . . duct \.tt I1 le~al rationalization., -rh" . . a<: ne~ tr1 JU t tf- it!. con<, . . -- resut iS a)(j cl r . . ') 1 and cmmterrlrmn<>, who!>e: C<JOS"ot1e . n ~c11latwn of clailn.t " 1 nc:e, m turn i botl 1 . 1 r tion of antemattonal law and a "cr\;;Ua 1 atv ga .. ' th a 11rtl,er devalua, ,JbJ 1 statt-!\ \o\ ho ltave de:bar,ed tl ,. P at e expenS<: c,f th(,\f= 1.., <:urren<. . America' y ath . . . resort to higldy de:batahle legal argurle t t !> r er tndascnrninatc: 1 n s supjXJrt 1~ \ .. ~ . . case an pomt. fh e unsuhtle redu<.t iun f . t . s tetnam pok-y ts a o m emationaJ law to house of convenient ex post justifkations (as in the cas . .a. m~rt: storetion at Su<:z, or A merican interventions . S e ?f Bnh<;h mtervt:nan , anto 0 ommgo a11 d v t undermines the w pretensc of contnLl ti to ::.ry Id a namJ e I. u ng wor order w'th h. h 1 these statt:s ltavt tritd to justify thtir unilateral a<.-ts. w ae: (b) \lluch of <:Ontemporarv international law a th . ~ . . .~ .. ~ . u onze~ states to zncrease I 1 t1e1r r>ower. Jn t11s connection. ~asser's naticral fth S ' 1 12a0 on o e . ut:Z Canal Company was probably quatc legal, and those who accc:pt tht rather tortured argument put forth by the State Department legal advisers to ti'fv th . . JUS e Cu ban "quarantmc.. have wncIuded that thio; partial bLv.Jcad " .1 . . u.~ e \\."35 au by the OAS Charter and not in contradiction with the U..:..; . Charter. y,e t t t.S a ob\~Ous that a full exploitation by all states of all permissions granted b> international law would be a perfect recipe for chaos. (c) Attempts to enforce or to strengthen intemational/au;. far from c.:on.sol.idating a ~~tern of desirable restraints on state (mis 1behmior, m~ ac-tua.Uv backfire if the politi(."a] conditions are not ripe. This is tlle central lesson of th~ long story of the financing of u S . peace-keeping operations. American selfintoxication \o\~th the importanc.-e of the rule of law. fed by misleading analogjes between the U.~. Charttr and the u.S. Constitution. resulted u.ltimately in a weakening of the influence of the \Vorld Court 'which largely foUow d .-\.mere ic:a's Hne of reasoning), and in an overplaying of America's hand during the "non-session" of the General Assemblv in the faH of 1964 and \'rinter of 196.?.



selected H not mean that one will forgo legal rationalizations of the JnO\ es

m does action. iDtemationallaw does~ in a crisis, really restric-t ones freed~m

~~~ ~there is a legal chessboard for state eompetition.

ere~ come to the last set of considerations about the role of l,a\v:

These are sobering considerations. But what they tell us is not. as so many political scientists seem to believe, that international law is, at best, a farce, and. at worst, even a potential danger; what they tell us is that the nature of the il1ter 1Ultional sy.stem condemns intemationa/lau; to all the u;eoknes.<;es and pen;enions that it is so eas y to deride. International law is merely a magnll)ing mirror that reflec-ts faithfully and crutlly the essenc-e and the logic of international politics. In a ~agmented world, there is no "global perspective- from v.hich anyone c~ a~tbor atatively assess, endorse, or reject the separate national efforts ~ making mternationallaw serve national interests above all. Uke tlle somber umverse of A .lbert Camus' Caligula. this is a judgele s world where no one is innocent.: . The permanent plight of international law is that nO\'' as before, lt shows on.tts body of rules all the ~cars inflicted bv the international state of The tragt-d~ of contemporarv international Jaw is that of a doubJe di\"'r<.-e: first. between the old . of libera] dreaJ~ of a world rule of law, and the realities of an international system


tl ~, ten to become Jlla,iur <;: mphes; seco d multiple minidramas that always H~eqa uirernents of mode rat 11111 which in the ~ ' <:tr. . .. l cIteam . 1d the new J ggest a down -pla!JIII{!. oJ 1 ormal law in tl aJ bctwe(;'n the ok 1e . rtl1e present system1 su upuraclin{!. ol, more fl e:-.1.11 L' l(e Illll(lues, unrl 1 J . . cumst<ulces o I I! r I . b al d wJr ISsues, anu Tl e interest of internatiOJHt t\\ or t 1e l)O)iti caJ realm o peace-an - ' . . ne . b 1 less 6erce. 1 I1as eco . no better way of grasping the conlllllJJng clifferenc:~;:s tl1e syste1~1 scientist 1s that t~le~e ~ national soci~t')' and the fragile ordC'r of international d h n states use legal language symbols, and tlot:u. between order WJU11ll afTairs than to study how an w e rnents. and with what results.

Jntern(Jtjonal Institutions: can Interdependence Work?


ti. J. . . nt To analyze world pol ities in the [current era] is to discuss 1 erna ona 1nstttutJons: . 18 rules that ~~vcrn elements of world poli~~s and the organizations that help ~ 1mplernent tl1ose ndes .... : Under what conditions should China be admitted to the World Trade Orgamzat1on (vvro)? How many billions of dollars does the Jnternational lv!.o~etary Fu.ncl (IMF) need at its disposal to remain an effective "lender of last resort for countnes such a'5 Indonesia, Korea, and Thailand that were threatened in 1997 with financial collapse? Will the tentative Kyoto Protocol on Climate Change be renegotiated, ratified, and implemented effectively? Can future United Nations peacekeeping practices-in contrast to the U~ fiascoes in Bosnia and Somalia-be made more effective? These questions help illustrate the gro\ving importance of international institutions for maintaining world order.... Superpowers need general rules because they seek to influence events around the world. Even an unchallenged superpower such as the United States wou ld be unable to achieve its goals through the bilateral exercise of influence: the costs of such massive "ann-t\,~sting" would be too great. International institutions are increasingly important, but they are not always successful. Ineffective institutions such as the United ~ations Industrial Development Organization or the Organization of African Unity exist alongside effectual ones such as the Montreal Protocol on Substances that Deplete the Ozone Layer and the European Union. In recent years. we have gained insight into what makes some institutions more capable than others-how such institutions best promote cooperation among states and what mechanics of bargaining they use. But our knowledge is incomplete, and as the world moves toward ne\: fom~s o~ gl~bal regulation and governance, the increasing impact of intemat10nal mstltutions has raised new questions about how these institutions themselves are govemed.


Academic "scribblers" did not always haYe to pa~' much attention to intema~on~ institutions. The 1919 Versailles Treat')' constituted an attempt to cons!ru~t an mstitution for multilateral diplomacy-the League of Nations. But the reJection of the
ork?From Hol>crt 0 . Kcohane 1ntemational Instihltions: Can 111t C-dependenre r:W r Foreign PolictJ. ational Pcac:t> l ' pp. 82-94. 1998 by tJ1e Camegte End0\~1mut 10r n1em . ISSue # 110 (Spring 1998). Reprinteu with ptrmission.



US S, . t , ensured that until \\'or!< ' \\ ' fl thl' lllOst L(';\guc Cowu ;lllt by tl~e ,, c:.;Jil~ '(: - f'roln the secret C crriiH \IIS~ia n deals or t . ti l , Ill \\'01' C po l lCS important nego tut Ol . r . I .., ' ~lllluc h cont erenc e-took place on :m ., ! . . basis. On)Y the 1920s tot1 19-> 1e r cJ d. , 45 (', I U . I Ntcn" was tOUJl e Ill 19 ' with strong support lrom the United . , ... ;uter t le 11llt'C t a 1 > f . 1 l 'ge 11 ,. 1 perfonni ng dJII<tcnt tasks 1 ., es d , 1 1 t\' > specw tZec a. '" , States and a mu tJp tCI l . rt'l sttl)stantial intcrnat1ou,1l tllention. u . . a1 n t' s begu1 to <:omm.. " mtemat1on mstl 1 lOll. no t [JO'''e rful state. !came toj 1d) increasin(11 y c. 1945] howe,er even t 1 1 H~ [Atter r-, : . _ ' . F.01 n the lute J960s onward, thC' Treat) on thr Non1 on internationalmst ltutiO\I~S. . U'e chief' vehicle (()r el'fotts to pr<'ve ntthe dan.,. . p ro L eratio f N clear\ eapons was 1 n u e d of nuc1 weapons. ,, \TO was not onlv the most succt>ssflll mu ltilate ral car ,, ge~'OUS spr ~ b .l h eneral, a allrluce m I11Story ul a so t e most hicrhl)' iJ1Stit11 tionalized. with a sec:re tmy-gr. n . l.fT d I b te rules goveming relatiO nS among nw tn1 .s. t' rom its )Ct perm~el~t st<4:..rutl1 e al otlra Urucnla\' Round that concluded in 1993, the General 9 founding ll1 1 I Jroug 1 1e o . r cl I rr -er, . cl Trade (c \Tf) presided over a senes o lra e roun<.s that Agreement on tatlns an ,, . . . ". b ". . JoJt tarifTs amon<r indusb1abzed countnes )' "P to 90 pe1cent, I 1ave reduced 1n 1 t:t . tl . . al t ade Af'ter a shakv start m ilie 1940s. 1e 1M I. 1 .I 1 t11e 1au- Jy boosting mterna 0 on r ust 1960s- become tJ 1 centorp1 ofeffi>1ts bv the maJor cap1ta. clemocrac1 to rege " ece . . . .es . .,rr.,;rs ex1ble 11 I ate tIle JI monetaJ). nllcu . When tl1al function anoph1ed wrtll the .onset of H 't L excIH 0 rates the 1970s. 1 1Jecam e tl1eir leadjng agent for finan<:mg and prom otuwe m . . ,.nt ;,... of mg econonuc de' e1op111"' ... A1iica' Asia and Latin America. The sheer number _ . inter-governmental organizations also rose dramatically -from about 30 m 1910 to rO in 1940 to more tban 1.000 by 1981. The exchange rate and oil crises of the early 1970s helpe d b~ing perce pti?ns in line with reality. Suddenly, both top policymakers and academ1 observers m the c United Stat~s ;ealize d tl1at global issues required systematic poJicy coordination and tl1at such coordination required institutions. In 1974, then se<..: retary of state Henrv Kissinger, who had paid little attention to international institutions, helped estabUsh the International Energy Agency to enable west ern countries to deal cooperative ly with tl1e threat of f11ture oil embargoes like the 1973 OPEC embargo of the Netherlands and United States. And the Ford administration sought to construct a new international mone tary regime based on H exible rathe r than pegged exchange rates. Confronted with complex interdepend ence and the efforts of states to manage it, political scientists began to redefine the study of international institutions, broadening it to encompass what they called ''intem ational regimes"-stru ctures of rules and norms tJ1at oould be more or less informal. The international trade regime. for example, diu not have strong fo rmal rules or integrated, eentralized management; ratl1er, it provided a set of interlockin g institutions, including regular meetings of the GAIT contr acting parties, formal dispute settlement arrangements, and delegation of technical tasks to a secre tariat , which gradually developed a body of case law and practice... . I~ the 1980s, research on intemational regimes moved from attempts to descn?e the phenomena of interdependence and intemational regimes to closer analysis of the conditions under whkh countries cooperate. How does cooperation occur among sovereign states and how do international institutions affect it? From ~e stan.dpoin.t o~ po.litical realism , botJ1 the reliance placed by states on certai n. mtemahonalmsntutions and the explosion in their numbers we re puzzling. Wh)


internatin!l.t! in~t itut inns <::xist tt . 11 ' a a t"S? Tllis qtwo.;ttot l S<.'<'med unanswenblc~ ifIll t'tworld dommated bv sovert:ign . . s t<~ '" .. < 1 . bove the state h11t not tl they were \i~;:wcms d uhons were seen as )osed to as . op1 or a ' . u ev1 to h 1 ces ' their objcc:ll,es. . e p states accomplish The new l"f'"''arc:l t 0 11 mtern aliuna l insft f l ali m- li te vi<'W that law <.:an be effe<.:tiv<:: reg:r~J ~ons f bro~~ decishely v..ith e~ll as with ti1C' id<'alism associ ated with the fi ~~~s 0 .P~Ilhcal c.:onditions- as we: . ted the asslllnpltons of' real ism aceepting th e us ongms 1nst ead, se1 t .1 . 1olars at[op ' a IC ative stat d tin<T interests wen key factors in world politics b t h e power an compe o I . f . ' u at l e same ti d . me rawmg new conclu sions about t 1e m luen<:e of institutions 011 th e proces .. s. Instituti . ,1te tJ1e capabd1ty fo r states to <:ooperate in muhl tlly b fi Cla1 ons <:re< < . . the costs of making mtd enfor<.:ingag reeme nts-w hat ecene .' ways b), lec]u<.:mg I' .. 1 ... 1 . onom1 'lC costs. T 1ey I ell e y engage m centralized enfor<:eme tstsfre er to as "tran s_ tion t f . . . the)' do renHorce pr..a<:t'tces. of re<:1pro<.:1.ty wh1ch provid .n .0 agreements. but .. r . . ' e 1 govern0r ments to keep the1r ovvn com rmtme nts to ensure that othe . 1 . . rs uo so as we U Even . Jowedul states have an mterest, most of the time, in following th 1 f . r 1 . . e ru es o we11 _ establi sI1ed mtern atlon a mstJtutlons, sin<:e general confonn1 to 1 tv ru es makes the behavior of other slates more predictable. This scholarship Jrew heavily on the twin concepts of uncertaintv and credibility. Theo rists incre asi_ngly recognized that the preferences of state~ amount to "private inform ation ''-tha t absent full transparency, states are uncertain about what their parb1ers and rivals value at any given time. They naturally respond to uncertainty by being less willin g to enter into agreem ents. since the, are unsure " how their pa1tners ,viJl later interpret the terms of such agreements. International institutions can reduce this un<:e1tainty by promoting negotjations in which transparency is encouraged ; by dealing "vith a series of issues o,er many years and under similar rules, thus encouraging honesty in order to preserve future reputation; and by systematically monitoring the compliance of governments with their commitments. Even if a government genuinely desires an intemational agreement. it may be unable to persuade its partn ers tl1at it will, in the fuhtre , be "illing and able to implement it. Successful international negoti ations may tJ1erefore require changes in domestic institutions. For instance. without "fast-track" authority on trade. the United States' nego tiating prutn ers have no assurance that Con~ess "~1 re.frain from adding new provisions to trade agree ments as a comlition for .the1r r~tifica tion. Hence, other states are reluctant to enter into trade negotiations ":tl~ the United States since they mav be confronted. at the end of tortuous negotiations. 1 _~__ WJtI a redesigned agree ment less favorable to t1em than tJ1e wa ft tbev ini tialed 1 By the same token vvithout fast-hack authority, no promise by the U.S. g~vemk n1en t 111 to ab'1de bv negotiated tenns has muc1 eredib'l'".''' due to the presidents lac 1 of COntrol over Cong ress. . ili In short this nev., school of thought argued that, rather than unposmb emg 1 ' nd se ves on states, international institutions shoul d responcl t~ tJl e dema..... ;\' states . d the fo uncertrurh r COOperative \Vays to fulnll their own purposes. B~. ~educmgh't t. ns J et anstates al . ' J c .s 0 f' making :ost t ernattt)11' ms u 1o 1 1 . and enforcing agreements, m achteve co11ective gruns. . .
shoU Id



I 22




RS[f.S 1989- 95



thout jts critics. who f(x: :ceJ ei r atta<.:I<Js on . vas not \ \1 . 1 . Th is nev: institunona ISm ' gs F 1 st, the) claimed that mtct. n tJOllal institut1cm, r n three perc:eived shortcomm . . states wield the on.I\' rea 1>0\' t r in world 1-oI. . . . ficant smce feffo rts bv the Li:'\ or Ur..t't, w of .\:ation to k gm art' fu ndamentally msJ .. j wea ness o . I1asr~d thea ainst aggression by . gre~t powc::rs. am. they pointt'd bJ . itk~. They em~ 'b tors in in ternation al econontK organi:t..atic n adueve collective secun~ g J ~. cl ., J of maJor contn u . institutions were attnnute 1nore to tt honal . the dommant roe nt a t5 of these mtema. ., b k than to the institutions them elves. Hen<.:e, any euec at power ac ers . ffiorts ofthe1r gre ated. Of c'Ourse. great powers such as the L; nited e . . al This argument was overst ce ,\ithin intemation . 'tu tions. But t I1e nnlkit msti . fi uen .. 1 ~ . se Statcs exercl enonnous. m ~tutions are different from those that the Cnitecl State~ th b that emerge from ese msu agreement }' man} states IS necessary ed 1, t rall)' . \.Vhere f 1 fi . 1. would have adopt unua e f'~' ....: e even the United States nu.s 1t use u to compromise on . 1 d .. r al I Je for pohc:y to 1 e Je<:'uV . ecJsJcm-making } , tJ1e 'nstitutional seal of approv .. T .1ere10re, t 1e 1 . substance too Jtain - - ..1 . ' affect both pruu;uures and rules of international rnstitutions matte r. T 1ev fpoliG-v and the degree to which other states accept 1t. f h ,. th b the substan ce o .. ., The second countcrargument focused on anarc ~ : ~ ~ se nc~ ? a world government or effective international legal system to wh1ch VJC~1ms o.f tnj~Jstice c:an appeal. As a result of anarchy, critics ar~ed, states pre fer relatJVe gan~s (Le., doing better tJ 1an other states) to absolute gams. They seek to protect the1r power and status and will resist even mutually beneficial c:ooperation if their partners are likely to benefit more than they are. for instance, throughout the American -So,~et arms race, both sides focused on their relative positions-who was ahead or threatening to gain a decisive advantage-rather than on their own levels of armaments. Similar dynamics appear on c:ertain economic issues, such as the fierce EuroAmeri<:an oompetition {i.e., Airbus Industrie versus Boeing) in the production of large passenger jel~. Sc:holarly disputes about the "relative gains question" were intense but shortlived. ft turned out that the question needed to be reframed: not, "do states seek relative or absolute gains?" but "under what conditions do they forego even mutually b<! cooperation to preserve their relative powe r and status?" When ther~ are ~nly two major players, and one side's gai ns may decisively change power relation~~lps, relative gains loom large: in arms races, fo r example, or monopolistic competition (as between Airbus and Boeing). Most issues of potential cooperation. ~owever, from trade l~beralization to climate change, involve multilateral negotia tions that make relatwe gains hard to calculate and e ntail little risk of decisive r)W~r shifts for one side over another. Therefore, states can be expected most of h.e time to seek to enhance their own welfare without being worried that others ~ .also make adv~nc:es. So the relative gains argument merely high1igJ1tS the diffic:u ~es 0 ndecoor<:ration where there is tough bilateral competition it does not by any . means u rmme prospects c ror cooperation in general. . . Th hi . e t rd obection to theones. of cooperation was less radical but more endur J Th . 't . mg. eortsts of C()( ar10 h cl 1 emerges out of W:~ ~ a recognized that cooperation is not harmonious: 'r an takes place through tough bargaining. Nevertheless,

the, daJ d ctt thl p<Jte:ntial Joint u . f -:>runs rom such . ~h . . .<-'OO{>(::ration explained the dramattc 11 r es I!'J e number and S(;()pe of <:oo perah,e multilateral instituti Critics pom tPU rnt, however, that barg . . ons Id a.Jllln u problem 5 . achie,;ng JC . 11~ '!Gt.Jm. f or in tance wh~....L otb e K\'Otc pcou produce obstacles to . ewer . agreemeut '" (~uc:c;tHmable in part he<.~use de\.E:: 1 . J <:rotocol wtll lead to a global . opmg . . h . . .. . :nuntnes refused to accept bincling )umts c,n t c1 r emrss1ons and the {.j S S . . ts . . . not <.:ont<unmg such enate declared 1 un.,,illmgness to . an', . agreement ratlfv . (;{JffimJtments b . cl . I . k d .. . . h' d . ) E:\ e opmg oountries. des Both s1 sld. e out tough bargaining positi ons. m en ne: ef[l0 rts at credible c Ito f these bargaining probl :omth promise. As a resu bt erns, 1' fatt that pos51 deals . cl . could e . produc:e J<>ln~ ~(1Jns oes not assure that cooperative sol . tactic of political actors and the info rmation th . h uti~ns "ill be reachecl The e' ave.available abo ut one another are both ke, aspec.-ts of a process that does not nec:essan ,. ead t . 11 0 tutions rnav help pro,ide "foc.:al points~ on \11h' h <.:ornpetinu cooperation. l nstiIC ' o ne,... issues often lack such institutions. In this <, both the ~<:tors may rurree, but pace and the extent of cooperatjon become more problematic.


The general problem of bargaining raises specific issues abo u t ho" . . mstitutlons . . a1 . . c afrect ~nternation . nego~atJ.o ns, which always involve a mixture of discord and ~o~ent1~ coopera~~n. Thmking about bargaining leads to concerns about subjectiVIty, smce bargrumng depends so heavily on the beliefs of the parties involved. And th e most Fundamental question scholars wish to answer concerns effectiveness: \Vhat structures, processes, and practices make international institutions more or less capable of affecting policies~and outcomes-in desired \\<l\ 'S? The impact of institutional arrangements on bargaining remains pu;linu. \\'e plorations of bargainunderstand from observation, from game theor: . and from ex ing in a variety of contexts that outcomes depend on more than the resources aqtiJable to the actors or the pay-offs they receive. Institutions affect bargaining patterns in complex and nuanced ways. Who, for example. has authority over the agenda? In the 1980s, Jacques Delors used his authority as head of the European Commission to struc:ture the agenda of the European Community, thus leading to tJ1e Single European Act and the Maastricht Treaty. What ,oting or consensus yoto Conference. arrangements are used and who interprets ambiguities? At the K agreement on a rule of "consensus did not prevent the conference chair from ignoring objections as he gaveled through provision after pro\ision in the final session. Can disgruntled participants block implementation of fonnally ratified agreements? In the GATT, unti l 1993, losers could prevent the findings of dispute r~solution panels from being implemented; but in the \\TO, panel reco~menda llons take effect unless there is a consensus not to implement them. Asking such questions svstematical ly about international institutions may well yield significant . ' new insights in fu ture years. Institutional maneuvers take place ''~thin a larger ideological context that helps c., denne which purposes such institutions pursue and w h' h pr.actices the, find accept. . ' JC able. The Mandates System of the League of Nations depended m part on specific

I ~.a



NIIV " "" - - .


, fitnchunenlal wa.-; till' 1 mderstan'' ,11 t hut tne>t"t. 1 institutional :1rmn~ctll< s. 11011 _p: 11 ropcan pcnp t lcceptable.UJng N l t: ut ru 1 uvct c 11 thtt l.ontit lltt'C I ~tlrOJ>L Eu r<>J 1eans c:on L I'C:'I11'\ ' ' ,l ,.nnate after ~I o . l E ,. 1 1c ..,, .tc.m of rult '' ' u ropt. at1~ over Inon- ,cu... following \\ ,orJ< ' ,, r , ... . ' . . , I ; n , t le 15 , , , <:olltjJSt' of that t'Onscnsus < uJ ~ h.011 .11instilulions <:r >nl m 1 'nnationul st ' . .. 1 1 s ol intema ' . . rue. The.' p1'0C'Cdut es .tm ru e . . . are acce Jtabk a~ t11{' b,lsts (or redu,,. 1 .. . . , 1hat [JnllCl p 1 es .. . . '-ltlg IctcJJlllll<.l " h1 rt'S. Thev < .. . . I I . ' cnunc.rtt,u, ~'ctions are legllliH<lt<.' <' r 11 1 ~.tu nate. Conse. , ctmflil'ts ;uld w 1 tet gu . T'S. ex xct-ations. for instan ce. I ra< I< c(n.lltcts arc increas. et tjuent k they help sh<~pt> <tc:to { l. teson (T in the \\1'<>--pronu:-:ing tough action o ' . . itttrly 1itmlhzecl Ill .'t P'oce~;s. o pw. ,,; 1tt1 oin c uasi-Judicial cltsputc resolution pro""n . r- 1 dustnes eug1 c 5' 111 . --c behalf of ones own .bl ."-"tnl)lainint1 about defeat wlwn necess;uy. Ther . . o e Imnung neto" 1fpossl r. 0 1 "v rnstitutionaUzcd prCX:t'!>SC. usually releg tte dure . c ' b. .,.,,, 1 1 d ~1 y ut re!!\1 1 " is much sou ne an ." . ' . 1:' es'51 lustitutions there b' ('rent e di Fk'rentia.t'e,l -011 . 1 1t ramat1c e~v' ' conID et to 1:1 rea 111 0 .. 1< t et t]l"' lanmtat1e directed tow;U'C] ..outsiders.. a d tC . ..I ders cm m erpt 1 1 '" . t- o infom1 ation. nst ~ to. t" T lJt"<Ot or manipulate. others belie(-;. 1derstcu1 1 lll!!S 111 " c use tllelr own uJ I .' ,.. ~ lttional institu tions continue to try to unde rstand Fint.ll v stuc ents o I ll elt < . . . ": . . . much more e ffecbn ' tI1<-tfl otIw rs. \'anatt.on 111 the wlw some mstttuttOns ,lre s0 . r . . . . . . . . 0. ' JOlic, or me 1nbet-s conrorm1t> Wlt I1 mstltutiona1mles is coherenre o1utstitu Oll;u I . I I li t . . d r b . the degree of c'Ommon . mteres t anc t 1e c stn )Uhon of )artiaU accou nte 10r ~ " l .1 l be rs. Institu tions "'hose membe rs s 1are socta 1\"luues and have . powe r mnong mem . . . L n S111ll.klf po1 Ca1 S)'St""tlls- such 'as KATO or the Europe<m Umo n-;u e bkeh-. to be I " . . . . . . stronger thau those such as the Orgamzation . fo: Secu nt: < Co~p e ration m md Europe or the Association of South East Asiat.l Nat~ons, whos e mor~ dtverse met~ bershi p does not necessariJ~ have the sam~ .kind ol deep comm on t~ter~sts . Addi. tional.h-. tht' character of domestic pohtics. . . . has a substantial tmpact on intem ~tional institu tions. The distri bution of powe r is also impo rtant. Institutions dominated b" a small number of members-f or example. the I~I F, with its weighted \' system -can typicaJl~,. take more,e action than t110se where influe nce is more '"idely difl'used, such as the UN Gene ral Asse mbly.


E''en as scholars pursue t11ese areas of inquiry. they are in d<mger of overlooking~
major nomtati"e issue: the democratic defici t" that exists in many of the world s most important international institu tions. As illustrated most recen tlv bv t11 far.. . e reaching interYentions of the LMF in East Asia, the globalization of the "orld econ omy and the expanding role of international institutions are creat ing a powerful foml of globali"E'gUlation. Major international institutions are ill creas inoJy laying do"11 rules and guidelines that goYemments, if they ''ish to attrac t foreigni~,estment and generate growth, must follow. But these international institutions are managed by t~ocrats .and super.ised by high governmental officials. That is, they are run .by ~lites. ~y m the most attenuated sense is democratic control exercised o,er major mlemational ora:mi,.~nons K negoti . . tJ e, atiOilS m 1e \\TO are made in c I osed sessions. The IMF nPCMtl~es in sec t 'th last Cl ~""-' re Wl potential borrowers and it has onlv begu n 111 the ew months to provide the conditi . . ' ons 1t 1IDposes on recipients ....J

:o--.-.u .

central banks , ('aJ I only be effective if it is ig ,entdcafrn he made that the IMF like 11um . l""o. su ate , trol. Ever sJnc<'. _to t , however, practitioners and th om. di. clemocraticconJect authoritati,e d<<.:tSJOn makin g can be combin d . h eonsts have explored how e Wtt 1 indire ct demo crattc.: con trol. The U.S. Constitution isaccountabiJ ty to publics and b ti l . ased onsuchathe orv-t he ldea th at poptd ar sovereignty though es sen through rathe r elaborate institu'tions. An issue tha ' 1s best exerctsed ' :tly, b I mdirec how to de\is c international institutions that are na~ts~~ a~s should now explore is but also accountable, at least ultimately to den . . Y bo l~petent and effective 1 ocratic pu 1 c:s One poss1blc respo nse IS to say that all 15 "ell . ationa are responsible to go,emments- which in turn a smce .mternbl . l titutions e ms r ae:cou to their own peop Ie. International regula simply dd . nta elimkdemocrac:les tion h . l3 I .. . . of delegatiOn. ut ong ch ams. of delegation m whicbatJ s anot erer n to the chain bli . ' le pu c auects action onlv at se,eral removes, re d uce actual public authoritv If tl t f A . . le erms o multilateral cOOJ)erahon are to re ect the mterests of broader democrati . .. . .. c publ'tcs rather than JUSt those Of llal ~ OWelites, l:raditl?naJ patterns of delegation wiJl have to be supplemented by oth~J . means of ensun ng greater accountability to public opinion. . One prom1S111g ap.~ro ach wo~d .b~ to seek to imigo rate transnational society m the form of ne tworks among mdl\'l duals and nongovernmental organizations. The gro\\t h or such netwo rks- of scientists, professionals in \-ariOUS fields. and human rights and emir onme ntal activists-h as been aided greatly by the fax machine and tbe Inter net and by institutional arrangements that incorporate these netwo rks into decision makin g. For example, natural and social scientists developed the scientific conse nsus underlying the Kyoto Protocol through the Intergovernmental Panel on Climate Change (JPCC) whose scientific work was organized by scientists ''vho did not ha,e to ans\\'er to anygovemments. The K yoto Protocol was negotiated , but gmernments opposed to effective action on climate change could not hope to re negotiate the scientific guidelines set by the IPCC. . . . Therefore > the future accou ntabilitY of international institutions to their I publics may rest only partly on delegation through forn1al democratic institutions. Its other pillar may be ,-oluntary pluralism under conditions of maximum t~sparencv. Inter national policies maY increasinul)' be monitored b, loose groupmgs b . of scientists or ot11er professionals. or by issue ad,ocacy networks such as Amne~' Intern ational and Greenpeace, whose members, scattered around the wo~l_d be linked even more close ly by mode m inforn1ation technology. :\cc.:ount_abili~ " 111 be enha nced not only by chains of official responsibili~, but b~ the r~c~Uir~m~nt of transparenc,. Official action s, neuotiated among state represent:atnes m mternational org:mizations will be subj~cted to scrutiny by transnational n~tw~rks. ' Such transp arency, howe,er. represents. nongo'emmental orn:amzations and 0 . rks fromelite oet\vo networks more t11an ordinarY people, who ma~. be as excluded . . . be ther - circle h as are from 0 t10vemme nt . T at IS. transnationaJ chilili o' ma' . a 'SOCJet-. . . . necessary but insufficien t condition for democratic accou.ntab t-.. . .emocrnctes . '. __ _: m . a1 0 rQaiuzations mamtam sw1J.Uent sIlould insist that, where,er feasible. intemation. .o
J '

Ad mi ttccll~ dcmoc.:racy does not alw . t ., ays work reguhu:ly eugagt Ill c lntn ~e:>s again st international instiwell.. American politicians ma)' of a ~ocal ~c t;l nen t o I thti r electorates at the ex . tu.hons, playing on the <lisin the U tllted \at tons. More seriously an ar c:esstve number of foreigners


l 2.5






" . ..I . ., '. cs for transnatlon(U networ k. or advocacy gro11ps , \l "Sli(: lcgisJ at ors, pnt JH l , of demoe:ratir. btcs to e,aJuate their actions. But 1 . ti " <Uld democra . c pu 1 .I so le<.'Ounta shouId a1 seek c:ou nterpa1t s to the mec l <m.s 'IS \ tontrol embed . . . ' 'ded in national democratic mstitutions. Governors of the Feu( .J a 1 Reserve Boarcl . 1)\ the Senate, even iftJ 1 . . ' cy . are tfter all nom mated toy the pres1'dent and confirmed cli tl 'ty l ng their tetl'ms of office. lf Ma son. llan11lton, und }ly '' . '. 1 , J<, 1, . z 1 great au 1on c un exerctse n . 1 1arusrns mvent md'1rect mec1, . of popular con tro 1 t H.: ex. c1 a tst Pa ,.-, e~8 could . ev1se comparable . centunes ago, . s] Jd no t be beyond our competence to c1 tt 1ou two mechanisms at the global level in the twenty-first centwy.

rhe UrJI[Cd Nations and International Security


of troops deploy~d 111 UN pe.a~e-keepm~ acti~iti~s and by an increase in the types of roles they p~~f01 m. At long la'it, the Urut~d 1 ations seemed to offer the prospect of moving dec~J~ely away fr?m the anarcluc reliance on force, largely on a unilateral gn basis, by indiVldual sovere1 states. _The United l\ations has, and will probably continue to have, a far mor~ centTal _role.m sec~ty issues than it did during the Cold War. However, the Untted NattOns muJtifaceted role in the security field faces a huge array of problems., Almost every difficulty connected with the preparation. deployment, and use of force has re-emerged in a UN context and does not appear to be any easier to address. Excessive demands have been placed on the United Nations, which has been asked to pour the oil of peace-keeping on the troubled waters of a huge number of conflicts, to develop its role in preventing breaches of the peace, and to play a central part in defeating aggression and tackling the aftereffects of war. Arms control, too, is embroiled in controversy, with \-arious statesIraq and North Korea bei ng the clearest examples-challenging what ther see as a discriminatory non-proliferation regime. Above all, the increasing role of the United Nations in international security raises two cenbal questions: First, is there a real coherence in the vast array of security activities undertaken by the United Nations? Second, is the re a danger that the elemental force of ethnic conflict could tlefeat the United Nations' efforts? ... This article advances the following propositions about the United ~ations' post-Cold vVar role in the field of intemational security:
l. The United Nations has become seriously overloaded with security is ue . for good and enduring reasons. The extent to which it can transfer these . . .. responsibilities to regional organizations is debatable. 2. Most conflicts in the contemporary world involve an el~ment of C!Vll w~t or inter-ethnic struggle. They are different in character fr~m those conAtct essentially interstate, that tbe United l\ations ,, established to tackle.
ll ~ E .xt'Crpt~ from Adam Hoberts "The United :\ations and lntemation aJ Secunty.- Sun:ivll dTlu . __ ().. . r .., >n 5 " "t:uarterly Vol. , , "'o 9 ($ UllllllCJ. 1993) PP. ..r-JV. Reprinlt:t.~J b'' pc:-nn~:.lon ofthe author an Ta\1 V. F
' raucis Ltd. http://www.tandf.<

. I n recent years. there l1as been a remarkable growth in dem an cls r tb e serv1ces ol. ror . . ( r) h . UN 111 t e field of mtemational security Th 1g91 h . d the Uruted Nanons aut on:~..e e . kl f' action in Iraq was ~jUJC , y ollowed ~~ 1992 by a fivefold increase in the numbers




ANARCHY AN U 11:, l..Vt~o:~L'-l"''" -




1 ng the major po ' ll ll the bas 3. There i. only limite? agreemen~ an~:ited shared iul . ' Prtsu rin ~~~f international ecunty and onl) a . t d g t at . . ' effectively ttnp 1 emen e . Ultemaooncu nonns are . C il inducuna tlw s t( n of Rve r th e Secuntv ouncin. danaer hof Jo; Veto. 4. The structure o b. . its legiti m 0 Jd' nent mem ers, lS ac.v. \\1e mg perma f mbership or powers '' 1 ' he ve ry hard t 0 Although a form~ chhang~ ~i~s procedures and prttrticc>s 1nay be bot~ achie,e, changes m t e oun 1 desirable and possible. . the practice whereb\ enforl't'ment ha., take 5 Tl ome ad\antaaes m r' n . 1ere are s . d 0 i]"tarY action by groups o statcc;, rather than corn. the form of.authoU~~e :n~and as a ]iter~ reading of tlte lJ .\" Charter would ing under direct eo


suggest. . d 1 nons role is increasing, basic questions about <.:olN 6 Although the Umte a . . There is no prospect o f a genera1system of <.:olle<:lecbve secunty remam. . . . c 1 nng existing strateg1 aJTangements. twe secunty supp an .. are in no way intended as cliticism of the increased These propositions . . . . h r . ] . f . . tl United Nations and its role m t e 10retgn po tc1 o many es emphasis gJVen to le b fl I I states. Rather, they constitute a plea for tl1e so er asse~smtle1~t k~ Jot 1 t 1e merits b and defects of an increased role, as well as for construcbve 1111 1ng a out some of the difficult issues it poses, and a caution against. the has~ abando~ment of some still-valuable aspects of traditional approaches to mtematwnal relattOns.

European ' ' ., With responsibilities in th . . . . Conferen<.:C'P.I vcurityandCoop .eratio . E.e secunty field is notorious Th . r. h n m urope (CSCE) 7\ T .. e Comm um~- L t e Western European U . . ' !\ 0 , the European Cooperatior, ( o tncil ( NACC) all play roles f~on .(\V~U ), and the :'\orth AtJantic development" nlargi ng the international o vru!'mg Importance. ... Despite such easier said than done. These organization5secuntyb of regional organi7.ations is h role 1 cl and me mbPrs 11ps, an they often have great ave a. ewildenng.vanety of purposes diffi . 1 taking action . .\I any :c:g~onal bodies are seen as t~u ~n rea<:hing.decisions and in is often far fr01n c;clf-eVJdent whi<;h regional b cl a] to one s1de. Moreover, it addressing a ~iv<:n problem. The United ~a~ Y s houJd ; ave the prin<.:ipal role in bodies to hand le crises only to Rnd that importa o;s as 0 ten encouraged regional withjn its own domain. n aspec.1s of the problems remai ned



Many of the conflicts in the contemporary world ha dio h 1 1 . ve a very nerent c: aracter from those t 1at t 1e Umted Nat1ons was designed to address. Above ll h . a t ose who . . framed th~ UN Charter ~ad m mmd the problem of international war, wa ed b well-organ1zed states. Th1s reflected the view, still common today, that aggress1on g y . . . . and mtemahonal war constitute the supreme problem of international relations. Although the pr?blem o~ inte.rstate war has by no means disappeared, for many, civil war-whether mtemabonal1zed or not-has always represented the deadlier threat. Some of the twentieth century's principal political philosophies have underestimated the significance of et.hrticity. however defined, as a powerful political force and source of conflict; this is now changing through the pressure of e\ents.. .. In the overwhelming majority ofU~ Security Council operations today, there is a strong element of civiJ war and communal con.Bict. For the United :l\ations, involvement in such a conflict is hardly new, as the long-standing and continuing problems of Palestine/Israel and Cyprus bear witness. The collapse of large multinational states and empires almost always causes severe dislocations, including the emergence or re-emergence of ethnic, religious, regionaL and other animosities. The absenc-e of fully legitimate poHtical systems, traditions, regimes, and state frontiers all increase the likelihood that a narrowly ethnic definition of"nations" These difficulties are compounded by the fact that, for the most part. the geographical distribution of popuJations is so messy that the harmonious realization of national self-determination is impossible. Confli<.:t-ridden parts of the fonner Yugoslavia and the fonn~r S<M:et Union are merely the two most conspicuous c'Ontemporruy e~ples of u~perial collapse leading to inter-ethnic war. In both cases, the taboo against <.:hangmg old "c'Olonial" frontiers has been undennined much more quickly and seriously than oc-curred in post-c.'Olonial states in Africa and elsewhere in ~le de<.-ades _following European dec.'Oionization .... It is by no means impossible that mtemal conlhcts could drag the Unjted 1\: ations down; its inability to prevent a resumption ~f war in Angola following the September 1992 elections is an ominous indicator 0~ ~type ~fhazard Internal <.:onflicts, especially those with a communal or ethmc dimension: p~es ent special risks for international engagement. whether in the fonn of mediation,

... Reasons for such a heavy demand to deal with wars, civil strife, and oilier crises are numerous and persuasive. vVhatever difficulties the United ~ations may face in the coming years. these reasons wiUnot suddenly disappear. Three stand out. First, the impressive record of the United Nations in the years 1987-92 has raised ex'Pectations. The United ~ations has contributed to the settlement of nu merous regional conflicts, including the Iran-Iraq War, the South African presence in Namibia, the Soviet presence in Afghanistan, and the Vietnamese presence in Cambodia. It provided a framework for the expul~ion of Iraq from Kuwait. Second, given a choice. states contemplating the use of force beyond their borders often prefer to do it in a multilateral, especially UN, context. A multilateral approach helps neutralize ~omestic opposition, increases the opportunity that operations have limIted and. legJ~ate goals, and reduces the risk of large-scale force being used by adversanes or nval powers. Third, the United i\ations has some notable advantages ove~ regional organizations in tackling security problems: It is uni versal; it has a reputation, even if it is now under threat, for impartiality; and it has a more clear set of arr:mge~ents ~or making decisions on security issues than do most regional organizatio;, mcl~~g even the ~orth Atlantic Treaty Organization (NATO). ... ~.ogruZing that the United Nations is seriously overloaded much iliought has been gwen to the question of ' 'T'l.. . c1ea th th . cooperation WJth recrional security organizatiOns- 1 'ne at e Umted N ti d o 'bilitv for security he a ons an reg10nal institutions could share respons 'r seems to emergmg aibe hesttantly, in E urope. The proluerati'0 n o . .c It

""""' ru~u INrt:nu

First internal cc dh t '1. nd lo b . l''-"\ct-'kt'Cping. or lorccful mtlttmy '. , c.: 1' t 1 ~., , 1 <'tl<lllrintr .. t\cl.!.... l>nt l .!>...1. antI lo '\11 .. md thev leave contnHHlltiC~ Wll . l,. , 1 1 t'' 1 . < I 11 . . h 1 . . . . .. rienccs an< con ..:. proxmit mutual susptcions basecl on traurncltiC expe I . . , Y. . . .11' to tay ,. .hat may be a ver; Oil.., ' 11 !>< Second lntl:'rYt'ntlon reqmre.s a"'' llngnesd t I nder the leadership ul nc~n gm<'n~tncntl' nnJica r . ... m tt., conR ts a re ; 1 . ~ con uc ec u see reat advantagts m tit< degree of tc . or se rni-govemme ntal enotws, whJch may g . s a nu .J .. 1)e ttnwJ!J mg ( )r ~ . . 'th u N' 1epreselltalJv(' ) cl . t mvo l' 'ed u1 negohatJnu wt . tecogm1on ;.., . Tl rd interntl conf l~<::ls h n) .. 11 . unable to carrv out thc tems. of aureements. . . .,, b . 1attons, tJ1t1s. 1 11 1<.;.1 u -' . y .. 1 . l' "' l . ., nst the cvthan pop )Ccorn 111 involve the use of force <.. Jrectec agru I M . . . cliffi lt roblems relatetl to t w protecllon o r d'sperscd espectally bttter ~~-po~tnF g lcu.1 mal confucts are often conducted with small and vulnerable cJVihru1s. ou t 1 n e . d 1 st's match. It i verY dir-e: 1 to cnntro1t 1 use ncu t 'lE: rif1 s kn 1ves an t 1e arsol1l weapons: e ' b b' , 5 embargoes or formal methods of an ns control I1 ,eaponn.r by om m g. arn ' . . 0f 1 sue ' ' t1 '. there is frequentlv no tern tonal status quo ante Finally in cases sue1 as lese, l 11 ' C fi. . nd other agreements are ...,. 1 nera) e to tJ1e charge to which to return. ease- n es a . '11 . .. . . 1 c0 .1 and that tlwy create 11nposs1) y comp 1 -cc 1catecl that t11cy lemomv_..e t1 use o 11 le . b ,. _ .. aJ anangementc; based on ethmc tern tonal un1ts that are . "Ieopard-spot tet nton . . . small and separated and, thus, difficuJt ~o defend .. . .' . .". . ssues ctbout the cntena used m Commun a! and eth nJc c:onBicts nnse awkward 1 . . . . . l'tical no'ties as states and in favonng their admiSSIO to the Umted recogmzmg po 1 e . . . cn c . Nations. When the United \ ations admits member-stat~s, tt_J~ m 1 con ~e.rnng a act pruticuJarly important form of rewgnitio~. and it ~ also nnpbc1tly und~nvnting ~1e inviolabj]jty of their frontiers. Y tl1e Umted Na_ti?ns do~s n?t apperu to _be taki_ng et, sufficient account of trailitional c1iteria for recogmtion, v.rhJch melude careful consideration ofwhetl1er a state really exjsts and coheres a~ a political and social entity. Many European states also forgot these traditional criteria in so~ne of ~eir ~e~ent acts of recognition, many of which did not involve setting up diplomatic miSSions. If tl1e resuJts of recogrution are risky security commitments to purported states that never really attained internal cohesion, public support for UI\ action may be weakened. Such confucts also raise issues about the appropriateness of cettain principles derived from interstate relations, including the principle that changing fronti ers by force can never be accepted. This principle, which is very important in contemporary international relations, has been frequently reiterated by th e international community in c.'Onnection with the Yugoslav crisis. A successful armed grab for territOl)' on largely ethnic grounds would indeed set a deeply worrying precedent. Yet, it must be asked whether it is wise to express this legal principle so forcefully in circumstances in which existing "frontiers" have no physical existence, in which they lack both logic and legitimacy, in which there are such deep-seated ethnic problems, and in which almost any imaginable outcome will involve recognition of the consequences of frontier violations.
. .. ti 11 1nten en . .

I '

remain f ' r. fntm wderenccs of both t there . m erest and be enougl.1 to p .<. the Security CounciJ from reach in r der_<:~ption. These may not they can frustre:t P Jlts to turn decisions into actions ing CCl~ons ~n key issue , but Difference"- of 'nterest amongst states a . fast-changmg situations .... rundamental natu re of re complem E;nted by differences in rceptions aho 1t t Id .. pe t . . . al . wor ir differen t :usto1 1c expenences some st t pohtJ<:s D ependi ng lar{fely on . I t1 e ] . ' a es VJew col . 1 d . b . ""'perialistn as lle ;nost senous problems in int . ona ommation and .... ] as th e most ( angerous threat to intematio emational relat'ons; others see civil ] . ar w i. . na sec:unty )'et 0 th . ve ers v1 aggresew st conquest. anc mternahonal war as the centra1probl ems. . . . . . . Such se nons diffe1 ences of perception and t .I! r th . m erest are of eo B the proceewngs or e U)1 Security Council 0 } Id ' urse, re ected 11 1 . ne s 10u not ne . ril relations among .maJOr po~vers to he good, and there mav be ces~ _) ~>;pect sons why counb"les perce1ve major sec:uritv probl diN-} vahd rea. lth , ems , China's world-vew, a ough undergoing important h ~t erentlv [For exampIe,] , . . l .1 c ' . . c anges retams disti elements- 111C Uumg a T 0 fore1 subversion, a Stron b )' f'. , ear gn ~Ctive 1 c t' . 1 .1 g e te n state soveretgntv 1 and some Juentlllca 1011 WJt 1 ueve aping states-which could set . . ,' Security Council members. lt agamst other






... If the United Nations is indeed to have an enlarged role in securitv affai t .. _, rs, 1 s f deciSlon-maki ng must be seen to be legitimate. system o The powers of the Security Council are, in theory, ve1y extensive: "The Members of the United Nations agree to accept and carry out the decisions of the Security Council in accordance with the present Charter." In practice. the Security CouncU cannot impose its will on the membership in the way this statement implies and. despite the absence of any system of fonnal constitutional challenge. there is no si211 of the emergence of a doctrine even hinting at the infallibility of U\ Security Council pronouncements. However, these U mitations on the power of tl1e Security Council do not mean that states, having successfully retained considerable sovereign powers in security matters, see the existing aJTangements as satisfactory. The criticisms of the composition of the Secmity Council involve several el~ments: doubt about preserving unaltered, half a century later, the special position_ of those countries that were allies in the Second World War; concern that three of those powers-France, Britain, and the United States-make most of the agendasetting decisions in running the Security Council; irritation, especially on the part of Germany and Japan, about "taxation without representation ," and frustration that the views of the non-perman ent members of the Security Council, and indee? _o~ the great majority of the 181-strong General Assembly, count for uttle. Th~se ~ntic~sms could become much more serious if events take such a tu m that the~ comclde wtth a ~)erception that the SecUJity Council has made serious misjudgments on central


: It~ und~Ie, and_ve~ welcome, that there is more agreement among states about mtemational secunty Issues now than there was during the Cold War. However,

In the history of the United Nations much more has been achieved by changes in practice, rather than Charter revision. More thought '"ill have to be ghen to how the Security Council might develop its procedures and practices: for example. by



. .. . "VI ~;:,


-permanent 1nembers to rdltl't t 11.! ' con LJibtttiOtl\ . . trl'ngthenitw the selection of non ' I ing more regular Scc:u rt) f ) mc:tl cons~tlta. t " U 'tj' Nations' work and de,ede op ....hes Such change<; lit~J difficult t to trw m eu . st pcu u o tion ,,;th major state. and. tnter:e way to\\'ards meeting th~ ..t~<n~ conc:em!l of implement. might gn at least so f decisions that affect them \lt, 11) certain states abottt being left out o




tions is central to almost <.'\'Cf}' discussion of The issue of orgafilzmg enforcement . gs out the con H I>etw<::en "Cl1arter lunda. b ac t<:t the United ~ations' future role.,It -~ s to be organized precisely in accord "~tit . .. . ld like sue l acuon .. . . mentalists, \\ho wou . .. mmon Jaw approach. '' ho believe the tnost the UN Charter, and those '~'lt 1 a eo , . important gwde 1S u1\ prachce.. . . miJitarv action authonzecl by the United in the UN era. maJOI . .~ . . . . 95 ; . . I " . Tlree tunes s t u~ command: Ill KOJ ea Ill 1 . 0-53. h aq Ill Nations has been ~ ~~- The~e episodes suggest th e emergence of a sys1 1990-91, and Som~li_a Jn Nl' . authorizes military actions, which are then placed . } c] the umte aeons' . d tem tn w 11 1 f tates There are important a vantages to cl th control of a state or group o s . un er c F' t it reflects the reality tl1e1t not all states feel equally such an arrangement. IT'S ' tary actiOns rcqutre extremeIy . . c nt action Moreover mw invohed in e''ery en1orceme ' .' . . .1 1uy func:.J , _ _ ..; be...veen intelligence-gathenng and operations, a smoot cIose coorwmsuOll '' f ki . decJSJon-maki11g 1nacl1 and forces with some expenence o vvor ng ine nonmg . . ..~: dangerous and c:.-omplex tasks. These tlungs are more likely to togetI1 to penonn er . d ili' 1 be achieved through existing national armed fo~ces, alhances, an _m taty re ah~nships, than they are ~thin the structure of a UN command: As habits of cooperation benveen armed forc.-es develop, and as the United Nations Itself grows, the scope for action under direct Ut\ c'Ommand may increase, but this will inevitably be a slow process.... Experience seems to show that mobilizing for collective security onJy works when one power takes the lead. However, as a result of the e ffort, that same power may be reluctant to continue assuming the entire burden of collective security. After the Korean War, the United States tried to set up regional alliances to reduce its direct military obligation. After tl1e 1991 Gulf War, the United States was mani festly reluctant to get entangled in Iraq and to underwrite all security arrangements in the area. ... The issue of UN versus authorized national command arises in non-enforcement connections as well. As UN-controlled peace-keepi ng forces become involved in more complex missions, in which neat distinctions betweet~ ~~-keeping_ and enforcement are eroded, the adequacy of the United Nation~ ~Xlstin~ machmery for controlling complex operations in distant countries 15 mcreasmgly called into question ....


Is it ~~ble to ~y that out of the rubble of the Cold \Var a system of collectl\'(' security JS emergmg? The t .. 11 . enn eo eetive security" nonnally refers to a systetll

in lhe system ac<:epts th t tJ . . a 1e secuntv of 0 . h 1 .,JI anc..l ai!;fl't 1 1n m a coll ective respon t .. ne 1s t e C."Onc:em of " cl 1 . se o aggress1on 1 th nct fro m <:' d 1\ e e ense or alliane;e svstem . b n ts sense, it is dist1 11 . , s. m w t<:1 arou 1 f each other. p_n ."a ~- a?.amst possihle external threats. .., ps o states ally ,,;th "Coli< <"ll' ~ ec:un ty proposals have been : . . m CJrcu 1 ation smce th be f the modern ,.l_ ,, ._ ~- ltc-1~ an d were indetd aired at the ne . . e gmnmo- o ~otlahons that led 1648 Pea<.:(' of \\ e!)tphaba. The attractive theo ry 0 f eoll ectJve securitv h to the . against Olll<' Ims1c c1ueshuns, ofte n reveals som f d . w en tested . . , 11flc'twe secunty~ There is alwa,s e unk amental flaws.' Whose eo a tl. . 11 , ns tem will I)C see11 as protectmg onlv e;ertain counhie lat. a eo ectwe securitv S\"S . . . 1 h .. ' s or mtercsts or as privil cert.Un pnnc:tp e<; at t c ex-pense of others. Some eo t . , r egmg . b . un nes mav. 1 wbate,er rea0r son feel exc Iuc:l e cl from tts enefits or threatened br 1 Th . '. . t ' l cl . 1 e ai1Xleties expressed bv some ~?u'll~riJesl tn tile e' e opmg world reaarding the C"Oncept of the ~:\ ew World Order. w 11 e t 1ev 1a,e not yet crvstallized into definite 0 t ,c:. .d f ppos1 1011 to any specmc Ui\ actton, are en ence o conc:em on this point. Cm~ there .be consistent re~punses to security probleo~~: Although the u~ system_ 1 the fir~t tmly glohal tn~ernational sy tem and although it involves the s. subscnpttot~ of \IJrtually all count~es _in the world to a common set of principles. it is not yet eVIdent _that the same pnnctples and practice could or should be applied consistently to different problems. countries, and regions. Difficultie can arise both from the consistent application of principles to situation that are fundamentally different and from tl1e inconsistent appucation of principles. It is al o not vet apparent that collective securit)1 can operate as effecthely for East Timor or T1bet as for Kuwait. The widespread perception that Israel has successfullv defied U!\ Secmity Council resolutions while other states have not, although arguabl~ faciJe in certain respects, illustrates the explosiYeness of emerging accusations of'double standards" at the United 1\ations. The political price of apparent inconsistency could be high. Against which fljpes of threat is a system of collectiue ecurity intended to operate? There is no agreement that collective security hould appl~ equall~ to the following: massive aggression and annexation; cro -border incursions: emironmental despoliation; acts of terrorism; human rights ,;olations within a tate; communal and ethnic conflict; and the collapse of state stmctures under assault from internal opposition. In 1990-9 L many people anrued that i_t was t~1e p~cul~l~ flagrant nature of the Iraqi invasion. occupation. and annexatto~ _ofKuwcUt that JUStified the coalitions response; even then, the international mihtary respo~se was far from unanimous. The fact that this argument was so \\idely u ed underlines the point that in cases in which aggression i not o blatant. it m~ 11ht be much ~ard~r to secure an intemational military response; a state caught up m such a confuct nu_ght . I1ave to look after its own interests. Smce l 991 ed partlv bv tJ1e establishmsptr ' . edJ ment of "safe havens" in northern Iraq and pattly by a trend of ?pinion, adm~tt ~ . - -~en f:< f'TOm universal in favor of democracv, the1e Jlas 1IJt' some mcreased ad\(>cacy. ll . . . not least in Franc~ and the Unitcd Stat;'i, of a right of intervenhon ~~ stat~s even~~ the absence of a formal invitation. This remains a deep!~ con~entious ~s~uecr:anrts ..ard which collective secunty euo serves as a useful reminder that the ends to\\ 5 might be directed are not fixed.
l '<'

.,1 which

.t; tt


Hou; cvllcctive does rmjorceme11t ho1Je to he? Is col1lpJetc ;u;n: I\ i 111pos .1 to ;\ttain' esnt'!ci'\lly n the cu;e of miUtarv action? Is the re still 'ome ss' >le:. tate ,. - s ( ' to he neutral? In practice, there has never been. on the gl:ld J,v. '. a truly "colle .' tive" caS<' (id alone sy,;tcn>) of c'OIIcctive secunty. In the (, ull en'" <>I 1\llJ0- 91 , 1 key UN Security Council resolution .-oided the calllur al l<s tu takt action. Instead, it merely authorir.ed "member~~lat~ eo-operatlll~ '' 1th the Govern'. mcnt of Kuwait'' to use "all necessary means to !lllplemcnt n:lc' .lll t UN r solutions. This implitd that it was stilllegitil1late for~ state t~ haven ~tat u~ of neutrali r or non-belligercncv iu this<<mflict. lt marked an mteresti ng "','d '' ali,tie i tion of some optiu1istic provisions in Chapter VII of the UN Charter. How ca a system of co/Je(,ti~ sccutihJ act~vely deter a particular tlneat to a 11 pmticular coulllt~J? In the wJke of .the 1~] G~.U1 Viar. there was n111ch discussion as to p<>!;sible menns by wlaich, in the fullJre. mva.stons c lcl be deterred he fore disast ;ou struck. ... Following ,,unanimous Council decision of ll December the idea was implt>mcnted by the United Nations for the first time in Macedoni ' lronic.illy, a slate that until April 1993 remained a non-member was cllUs receivi: prolc<:tion from a state. Yugoslavia. tbat was still. for most practical purposes, a mtmbcr. De pite remarkable progress, the idea of "preventive deployment" i fmught ,v;tl, < Ufficui(V. There is tl1e risk that large number.; of states would request its that it would be insutllc:ient to discourage aggression , and that it tniO'ht be used b ' go\'emment <IS an ultemathe to pro'iding for its own defense. It should not, however ya Jx. taken for granted that militan' det)lovrnents are absolute!)' essenti'al Tl1e re mav ' also be some residual deterrent "alue in the lessons of Korea (1950-53) and K : (1990-91); under UN auspices. the United States has led coalitions gone to t~e defestse ~f JO\'aded tates to which the United States was not bound b . fonnill alliance <.'Omnutments and in which it had no troops deplored at t11e ti' Tlus c: r I, l , me. 11nous act ma) not L entire y lost on would-be aggressors. Yet there ar b d le be ~~m w1 1 ~me k' 1 f preventiYe UN deployments, of which Macedonia isto .. uc: 1 ' e oun a me so barbmger, are t'OnStdcred necessary. . t' " rlha pays. for col/('C a tive se~rity ? The question of burden-sharing in iJlterna on rUN NATO secunty matters is notoriousIY comp1 as sbown by the experience of k . ex, crisis ir~ 1992 pctlat-e- eepalmg. and of the US-led operations in the 1990-91 Gulf te annu c.'Ost of UN ... t>\'er-about S2.8 billion. Un . 'd ron . pe.ace-k!eepmg activthes was the highest tions in September 199 sto!fat trib~b?ns towards UN peace-keeping operafigure \\'liS reduced to 2 about $6~ _ million. but bv the beginning of 1993. this increasE'<~ costs of k . 0 mtlliou. States have responded well to the peaceoperations on badJv. th eepmg Howeve r if more Ut\, peace-keeping (or other) Id ""- do not, there are probl o, , ere cou added be d diflicu1 m securing pa)11lent. E\en If ty . . ems . . -' paagn. Bill Clinton while indi to a dressed D unng t l1e US presidenti<u caJll to the United 6ns edlcating that he would ad on pavment of the US debt . . repeat called new agreements for sharing the costs v maotaining peace and or of kEeping costs be reduced :;:,~ed that the US apportionment of UN peacecountry most~... in- L- .l ~ :~% to 25%. The extraordin~rv IJaradox of tht ~J its majo ( milita.v support IOr an mtemational orgamzatt0 ~~m - ' . . ~ iDISbation of the gulf betweer thoughth steadily repa)ing) defaulte r is vet one more . n e theon. 0 f eollectave security and its pr.1ctiCf . . . -J

~pac(' ~.,,


h .




35 f{owever, fu1 1 ~ 1 uvtn <.: nt difr., 1 SECURITY . L ' ncu ties m ical of, Scc:un ( <niiH:il decision ay come from stat . \Vhat ;., tlu. plru ,, of disa~ment es not mvolved in, or critsecurity? ~Jo.t proposab for c;o IJ ective and arnlS cont roz tn a y . . . . . . consistent vnth the needs of mtemal sccunty C'all for 1 . y.stem . ower lev J50f of collectit;e However, th<' uutPcl :\ations h secunty and int . e annaments its efforts in the fi<.'ld of u sar mament to d -, as yet work out a , .....heematsonal obligations ' 1 ""'"' rent h'J ''ArmS eontrol'' is ~till s<.:en by m an anns control in th P osophy lo guide develop guidelines fo r c;onvent. a as a suspect, melionst e post-Cold War era. a]ny . n arm t r concept A 10 ing Chsna. The rationale for ar ms reuucf ransrers have man Ysharp <: ' f ttempts to ...1 s r nuclear non-p rolife ration efforts all '?ns, or control of arms trann se:s, mc:ludrefined. This is especially important . vi ' m stiJI need to be ...... 'uIJv exa and for '"'rer f h d control arrangements are discnrJllnatory o t e common . . ew ~ , and s tb at ex1sf Secu ri ty Counci l assumes a mo re central -earsnon couJd be exac:erbated arms ". mgif the ro.le in that J r -pro ~re ration matters.




~rers, ~me



cl 1at~:;

be .



W ith the end of both the Cold War and the Soviet uruon , tl1e n1ghtm are of an . all-out nuclear war b~t:vveen. the superpowers that so dominated world politics since 1945 enu?d. It IS not lik~ly that a new danger of the same magnitude will arise, at least for the econom1caUy developed democracies of North America Japan, and Western Europe. Indeed, for the first time since the formation of thes~ nation-states, the <.:itizens of these countries may live out their lives without worrying that they or their children will have to die or kill in a major war. This fact, however, does not mean that we should no longer be concerned ~'ith how states use force. Even if the optimistic prediction is correct, we still need to understand previous eras in which warfare played such a large role. We cannot understand the course of the Cold \Var without studying the role nuclear weapons played in it. Moreover, an understanding of the role that nuclear weapons played in that era is central for determining the role they will play in this era. This is true for no other reason than that national leaders' views of the present are hea,ily influenced by their reading of the past. Furthermore, even within the developed rich world, where a great-power war is unlikely, military power still remains useFul to the conduct of statecraft. If it were not, these states would have already disarmed. They have not because the use of force must always be available, even if it is not always necessary. For much of the rest of the world. unfortunately, circumstances are different. Threats to the security of states remain real, and war among them bas not been aboJisheu. For all states, then-those likely to enjoy peace and those that ,..,iJJ have to endure war-what has changed is not so much the utility of military power as how it can be usefully employed.


The use of force almost always represents the partiaJ fail ure of a policv The excep' h . . of course is the case in which figh ting va1ued for its own sake ul en tt 15 15 tion, a d w ' believed that war brings out heroic values and punses . dhiduals anthc . tu res, or m. d .--- l . e mcreasc.:u hen fighting is seen as entertainment. ehanges 1Il 5tates. ,-alues an C""' lo'cts as the w d t0 rs to v ew arme ,vi ru 1 destructiveness of war however have led state ac tual f 1 ' ' di 1 f maneu,ers ac use o ast resort. Threats are a second choice to P onla JC for<.:e follows only if t11e threats fail.




l1iE us~"' 0~FFORCE L.>


. f 101 nee its use is temper .I 1,: ...~traints and Because of the high cost ~ ' e th~' could alway be bloox' t r. Brutalitie bar~ainin g. As bloody as 1110 t "~rs ar~- sba~ed intere t _ if not h\ their scruples t are limited in part b~- the combatantsto ,.., r 1 doe not nece arih- follow that t) ,~ .::\ . . ah to go '' .. dirr . --:h e,en tJlincr that 1S good for one side is hadJ uer enou q~n if two tate . e. ,-" en sts On1 . t ha,-e nu comm~_n m ere ..' . ~tion , do the opponents cram nothmg IJ~ bargaining. 51 for the other .a zero-sum ~ outcomes are clearh bad for both 'ide.: therefore. . . . certaJO h de hares an interest in a,oid.111g them. In most case . ho'e'er. . war eac ' thougl1 th e) are at ftl . . st as Thomas Schelling pomt out. stem from th" e,en " 1e mtere . 1 The_s1 ed_lature otrov ilian to create. Force can be u ed to take or to bargain. d ~ar t need ,-our ach-ersan cooperation and do d fact that 1t 1 eas~er to es . d . -di . If mu can take what ,ou want, ~-ou ono . _th hi -\ COUJltn ma, use force to e1ze pute temton Just t not hrobavebeo ~~ ~n to;; ~-oun,~all~t. \ lo tofthe things people and nations ~"ant. ak 1 r ma~ NU \OU o , as a be -' lO thiswa,..-\ nationnot on)'''(l11tsto t e temtory, Itwants . . . [ID\.en cannot hO\,e-er menacmg 1t; 1t ma,-even ' d 1 t t . nation mar want others to tophi . th ' . to crovem an exp Ol I "' 0 e ts d t 1 -'ues Brute force alone cannot ac e'e h e tl goals. A nation oth fi want ers to a op nu menacincr it ma, not want to g t 1em order to m c... the rs u ,0 m ' o that wants to stop o ~ri- A ~n..,,. 0 that wants others to adopt its ,-aJue cannot impose them cl reffiO\'e the th IC;.cu . II<UIV solelY through conquest \\'here the cooperation o~ an ad,~rs~: lS ~ee e~. b~gaining "ill~nsue. The robber does not need the cooperation of his \1Ctim tfhe kills him to get his wallet. Howe,er_ the thief who must obtain tile combination of a safe from the hostage who carries it only in his bead does need such cooperation. The thief may use force to demonstrate that the hostage can lose his life if he does not surrender the combination. But the thief no more \\ishes to kill the hostage and lose the c-ombination than the hostage ''ishes to die. The hostage may trade the combination for his life. The bargain may be unequal or unfair. but it is still a bargain. The mutual a,oidance of certain outcomes explains why past wars have not been as bloody as they could have been: but an analysis of why wars were not more destructive should not blind us to the factors that made them as destructi ve as ther were. B~- 1914. for example, all the statesmen of Europe belie,ed a war inevitable. and all were ready to eAploit it. ~one, however. imagined the staggering losses that their respective nations would infUct and bear in the field, or the extent to whic:h nonet . combatants would be attacked Y bv the second ,vear of the war the same men were accepting the deaths of hundreds or thousands for a few yards' gain in the front lines; and by the end of the '""-ar, they were planning large-scale aerial gas attacks on each other's major cities. The German bombing of Guernica in 1937 and Rotterdam in 1940 shoc-ked statesmen and dtiz.ens alike, but by the mjddJe of the war both v.'ere accepting as routine tbe total destruction of Gennan and Japanese cities. large? account for the increasing destructiveness of the \\'afS of Three factors _ ~last two centuries. ~trst was the steady technological improvem ent in weapo~ry. \\~~as madline guns, submarines, poison gas. and aircraft made it feasJbl_e to m:am' ~ kiD large numbers of people quickly. The rapidity of destruction that 15 ~:! 0 ~ weapons is only the most recent, albeit biggest, advance. Sec ~ eA~ m the capac..ity, and thus the need, of states to field ever large~ _ the JUR:e5. As states became more industrialized and centralized. tb~ de\.-eloped the administrative apparatus to move men 0 3 acquired wealth and

d scale. c( -col' .itant with the increase t'n mili' tarv pote tiaJ ~ \s soon as one state ex'nanded th. r n ~-as the necessitv to - ~ ..1;'7P this por , . Th us w hen Prussiarinstitutede rorces at tts ,1;....,~-- ' all other r l' ~ """", .. states had to tCJ ,o\ smt. tl . adunr~ersal COns<:ription and the neral-staff sysJ t:rn and then demonstrated 1eJI vantages 1'th .:r... Cle 0 1 fr.m<.:e, the rest of the contine t . kl S.\' u{ \ic:tories " er Austria ant qut<: v ado t d n . O' '. th ~ e tts methods. An ocrease in the potential power of states led to an in 1 Third was thC: ~adual"' democratization of w~~~ m ~r standing power. e d hence the indi~cri mi nate mass killincr of non eo bat expanslOn of the battlefield an ti m ants. E,eno . fj ghting and dying.oWorld War Il 'th . ' n~, <:1 zens and soldiers alike, began power. marked_not the debut but the zenith of this ~:s ~~>itensJv: us~ of airg. As ''ar changed from the pr0\1nce of the princes to the burden of th e masses_ the distincti on f d \l between combatants an cl noncombatants increasinoh- bl urre . . ost o the v.-ars of o, th . tu . di'd . . . h _,_, tmpmge upon the citizenrv. but ainl . fi the e 1g teen cen ry ,.: nan<:. . ' di' e d m th em. \\1l' th th e '"1clespread use of consc . m .~ th .1a11 few ti chilians np on m e runeteenth h . LL and t\ventiem centunes , . ov.ever, more citizens became soldiers. \\''th th e advent I . . . . . .I'd i ofindustn alJZation and w1th the mcreasmg dh-sion oflabor thectizens who w not 1 . . fight remamed behmd to produce weapons. :\ow a nation not onh- had to conquer . b - '- h cl cl ' its enemys anmes . ut cmo a to estroy the industrial plant that supp~ed their weapons. Graduall). the total ener~ of a country was diverted into waging wars. and as the costs of wars mcreased , so did the justifications gjven for them and the benefits clai~ed to derive from them. The greater tile sacrifices asked. the larger the victory spoils demande d. Because wars became literally wars of. by. and for the people. governme nts depende d increasingly upon the support of their citizens. As '"'<IfS became democratized. so too did they become popularized and propagandized. The readings in the fi rst section explore how forc-e has been and can be used in a changing world. Robert J. Art notes that the threat and use of force has fou r distinct functions and shows how their relati\'e importance varies &om one situation to another. Thomas Sc:helling examines the differences bet\veeo the uses of conventional and nuclear weapons and the links between force and foreign policy goals. Robert Art analyzes the concept of roerche diplomac y-the resort to force short of all-out war-an d demonstrates why it is difficult to execute. Robert Jenis argues that the extent to whic:h states can make themselves mote secure ,..,;thout menacing others depends in large part on whether offensive postures can be distinguished from defensive ones and whether the ofTense is believed to be more efficacious than the defense. Terrorism has never been absent from world politics, and Bruce Hoffman discusses its changing forms and purposes. and distinguishes terrorism from guerrilla warfare and criminal actiit).


It is a mistake to examine the possible use of force in a vacuum. As Clal&Se\\thitz li ca1 oa1s Its utilitv as well as e cl, st u{e~se force is an instrument for reaching po ti g . ed be ' ' 6 0 f fi hting ~~~: ho~ the elihood of its use, depends not only on the costs and percel' hut on the general political conte~t. the values statesmen and Cl alternative policy instruments available, and the ob,iPrti\'es sought. r--

l .Jtl



. f'tltl' readi 1ws in P;lrt Oth ' t., . :> 'tht' iJn))or c ~<tnc ~ ) n ' ent('Cl u1 mam o . . this ~sect io n d Llll llns tr ! ~: t h<." conti . > l" di 1 l. m . rq ~ ,c;t . 1 . pO\Vl'l. The two rea ngs .ut Robctt f. Att a rgth '' milihu) p nu1nu ~:> I nu tt'cU) u is ower . t0 l>OHtical outcomes. r: a . late J'k I . 1111nber of <~o:tk t'' ,r 'nr re lC\"<\m-e o f 101~ (: the be used tO l'C<lC l cl I . Robctt <:>. Pitp, ' mws tit ~ I li . ) . . .t 'bl lun~ e t 1at 1s. ' can c: mJes. 1 t 1 ks strona "tatle cne.. ....e of case of lhic; uni\Htunate [)hpo lll'al . ..J S , t> U uit~ , tatcs tla ac cnorne-. . 11 stnvevs tle un1' e , . f . d . tl . t .. it oa\ 'S because it h<ts libe ral dem ocI'<\ logtc o s utCt e terronsl I . non from 1980-2001. anc1argues l cl des to compromise. .

- -- -



e . l C ld \ V nuclear weapons. it w;'lS argued , hel.ped makI eompetition Icl o tl1erw1 e l1av e )een. That i ar. Dunna t le o ers safe r than it wou t> . . .. . betw een the two superpow two sup erno,.vers sca re d . not . cuce, ~m d t 111~. restramed tl e J . r e ' nuclear weapons madrr v that if it pushed tJ1e othel. tOO c. ., l11c t.SCU ll l0 get out of ltte r<ll . . them. EacI1 I1a d to wo.. .r ar war Each leam ed, espect,l11 ' . ellctc r tl1e Cu ban M. tssile . c nucle . d hand an f. escaIat e toto push tJ1e othe r to the pomt whe re 1t race d the choice of t .. d 1962 no Cnsts o upping ilie ante and riski ng loss of control , or backing down an n sking humiliation. Rul es of the road between tl1e two superp owe rs graduaU~ develop ed, and tlleir subsequ ent competition prO\ed safe r in the last twe nty -eta ht yea rs o[ the Cold War than it had bee n in the first fifteen. How relevan t for today is the superpower ex-perience witl1 nuc lear weapons? \\ftll states that experience inte nse poutical conflic ts witll one anotlte r be deterred from pushing one another too far? Or \\ill they be less ~estrained than were the supeqx>wers and find themselves in the horror of escalating to the use of nuclear -per pons? How valid a model is the US-Soviet ex ien ce for dyadic conflicts today? wea Scott Sagan and Kennetll Waltz anal.yze what is today the most dangerous politica1 conflict between two nudea r am1ed states- the Pakistani-I ndian conflict over the state of Kas hmir. They look at the 1999 shooting conflict ove r Kargil and draw oppo site con dusions from it. Sagan argues that we should take no comfort from the [act that a large war did not ensue because the re were too many nea r misses and bec-ause time the hvo states might not be so lucky. \\la]tz ar!m es tha t the limited use of next force by both sides in 1999 shows clearly how the mutual possession of nuclear weapons causes states to restrain their ambitions and reign in thei r mili tary. The Kargil case serves as a good exemplar by which to extrapolate the oth er possible con flicts hetween nuclear ann ed adversaries that the world may experience in the future.

The Four Functions ofForce


In view of wha t is like ly to be before us ' it is Vl'tal t o thmk carefuIlv and l . prec Y about the uses and limits of mi)jtary power. That is th e purpose essav tse is It IS . c 1 kd dd . .' . mten . e as a )ac ' rop ror pohcy debates ' not a presc npt'ton oI' specfic pohc1es. It t . I . .or consciOusly esc 1ews . elaborate det ail on .the requisite mI'litary trorces 1 scenanos . r a . . . n and rocuses mst ead on what militan pmver has and h not done can and . as , cannot do. Every model of how the world works has polic,, 1mpl.ieat'tons.' But not , . . clre 11 eve i} polic) ts bas ed on a clear \iew of how the world works \\< tl1E'n. . t he uses to w~tch m 1 htar~ power can be put ~ How have nuclear weapons affected these uses. And what IS the futu re of force m a world of nuclear parih and increas ing economic inte rdependence?


The goals tl1at states pursue range widely and ,,.uy considerably from case to C'ase. Military power is mo re useful for realizing some uoals than others. though it is generally considered of some use by most states for all of the goals that they hold. If we attemp t, howeve r, to be descliptively accu rate, to enumerate all of the purposes for which states use fo rce, we shall simply end up with a bewildering list. Descripthe accu racy is not a virtue per se for analysis. [n fact, descriptive accuracy is genemll~ bought at the cost of analytical utility. (A concept that is descripthely accumte is usually analytically use less .) The refore, rather than compile an exhaustive list of such purposes, 1 have sele cted four categories that themseh-es aual~tically exhaus~ fu nctions that forc e can serve: defense, deterrence, compellence, and "swaggenng.


ng 1980). From ''To What Ends Mili tary Power" by Robert J. Art. in lnt~matianol Security. \'oL4 (Spri pp. 4-35. Portions of the text and the fooh1otes bave be<-n omtted 141




ART I THE FOUR FU~ . "'-ON $ OF fOo..-r

at ~ posture. In fac.1:.. usu ~- ~-e more than two functions h on ' nn titen ' tL fll.!. . t r- c.~tJ.u ~r~at ones '"M._ militarY forces that can. se aller po'' .ers, not ,;s-a-\"15 . ~ ne L . , Jital' forces must m: 1 .ade relativ"" to . m1 i.s achie\ed onlv ns~li-' 15 sm "' ' bt of a states 1 measure of the capa Jhes th [I nee to some absolute c:calt. \ state that Cln those of another state. not WJ de~=n~:gai nst it and u uall~ d,.~~r .1 . \ state that tcu, t thereb, autom atieaJl~ dc~cr Qr cornpel it A c:ompel another state <.an also defend against another state ~thnnot ha"ing' the abilitv to either ddend again st or ' th state w1 ou er ,is-a~\is another ma~ or ma~ not be abk to per. state can deter ano er _ relative to it. \\her e feasihlc. defense is ttH<: ..h rompel il A state that can 5'\\fun-agg f th 0 ther three \A,on . ~ fi t If defense is not possib le. deter rence is generaJh fonn any o e 0 c.unc;tion mo t difficult to pin d0\\11 analvtic.:all~all states rum ~ ~ the 11 goaJ that , t\ Swaggenng IS the next pnon whose achie,emen t is the most difficult to dewonstrate; corn. nstrate but among the harde st to achie,e. The follow. dete rrenre, the ~ne d c..ill . peUence, the easJest to emo . . develops these pomts more n ~. d.i . mg ~_1n su;e use oft".or-ce iS the depJo\-ment of militarY power SO as tObe abJ(; . . . , da " ,. . The ur-;en. mage to oneself if t vard off an attack and to m1mmtze 1 th r . to do two unngs o ' 0rces ~gam st . ose of a attacked For defenshe purposes. a state will dired 1ts 1 ential or aL-tual attacker, but not against his u~armed population. For defensiq: pot . a state can depJo, its for(.-es in pla<:e pnor to an attack. use them after an ~ o<:eurred to re~J it or strike first if it believes th~t an attack upon it is immjnent or inB'itabJe. The defensive use of force can thus mvolve both peaceful and ph~caJ employment and both _repe~ent 'soc-ond, strikes ~n? ~ffen~he <6_~ strikes. 1f a state ~trikes first when 1t belie,es an attack upon 1t JS 1mmment, 1t JS Jaundring a preemptive b]ow. If it strikes first when_ it believes an a~ack is\.itable but not momentary, it is launching a preventive blow. Preemptive and pr~ent:i\e blows are undertaken when a state calculates, first, that others plan to attack it and. second, that to delay in striking offensively is against its interests. A. state preempts in order to '-\Test the advantage of the first strike from an opponent. A state launches a preventive attack because it believes that others will attatk it when the balance of forces turns in their favor and there fore attacks while the balance of forces is in its favor. In both cases it is bette r to strike first than to be struck fim:. The major distinc.tion between preemption and prevention is thecalt11ation about when an opponent's attack will o<:cur. For preem ption , it is a matter of hours, days, or even a few weeks at the most; for prevention , months or even li few years. In the rnse of preemption , the state bas almost no control over the timing of its attack; in the case of prevention , the state can in a more leisu rely way con template the timin g of its attack. For both cases it is the belie f in the certainty of war that governs the offensive, defensive attack~ For both cases , the maxim , "dte best defense is a good offense," makes good sense. The deterrent use of forre is the deployment of military powe r so as to be able to pu~ntao adversaty from doing something that one does not want him to do and~ he might Olberwise be tempted to do by threatening him with unacreptable puniSh ifbe does it Deterrenoo is thus the threat of retaliation . Its purpose is tD prevent IOIIII!tlliag undesirable from happening. The threat of punis hmen t is direc-ted at the

ntlwe ~ot all four functions are nec.-essa__ / r- rs ha\'e the ,,.J.en.""'' 1al to dev"l c "' 'P 1 all 00 1 the gre<u

ri.], well or equaJI.' ~er t'' :

l:!ive-n rnilita n

e cl"rectiveness d~nds upon <- tat s ahliJ ty to oom1 nee a nr.t......~, . adver~that '--- of the thTeat t~'"""ll4:11 I . . . h"'~" it taS both the\.,iJI and powe r top ; no severe y 1fhe undertakes the un ~! terrenCC thert ,(;rt employs for(;{; nr_.,,.,.l:._uV. Jt . thedesirable action in Cmo:-<tir~ J)e ,-....uu threat to IS l~lU . th . L resort to force in rder to pum~u u-... t 1 ~ e essenc.:e of deterren' If th ~eterrence by drfmil.ion has failed A deter r: thr e ~reat has to bt (;affjed out, intent that it wiJ) not have to be carried out. Threats ar~n: made precise~ .vith the Je to pre<>ent ac:tions f rQm L.mJ<T undertaken. If the threat has to be ,....ple 1 has mented, th "' (JI;"' o ~ready been undertaken. Hence deterrenc-e can be judgoo SU<:ctSsful e .~on or~ If the retahatory threats have not been implemented. Deterrence and defense are alike in that both are mtended to th . pr~t estate or its closest allies from pbysic:aJ attacks. The r~ 0 fboth.ts di.ssuasJOJ\-nP~ ..- .l:.... ~ r~-.1 ~- 'others not to unue rutKe ac:tions harmful to oneself. The deii . .- r-3UCSU1Jlg ens:J\.e use """'not es ----,1 b, (<1 an adversarY that he '......., conquer one's mili of~ force . -' " OJS5w:au , that . lar) or~. i 1be deterrent use of force dhsuades by c:om., ncing the adve population and territory will sufT~r terrible damage if he initiates the und~le di~~ense dissuades by presenting an umanquishable rnilitarv forre. Det bv errenre ting the c-ertaint,v of retaliatory devastation. ' presen bl Defense is possible without. deterrence , and deterrenr... 1S poss1 e \\1thout ...., . defense. A state c::an have the militarv where \\ithal to repel an m'-aswn ''1thout also ~ being able to threaten devastation to the invader's population or teniton. Similark a state can have the wherewithal credibly to threaten an adYersary \\-ith ~c:b ~-~~ tation and yet be unable to repel his invading for<..-e. Defense. therefore, does not necessarily buy deter rence, nor deterrence defense. A state that can defend itself from attack, moreover, will have little need to develop the wherewithal to deter. If physical attacks can be repelled or if the damage from them drasticallv minimized the incentive to develop a retaliatory capability is low. A state that ~not defend itself, however, will try to develop an effe<-1ive deterrent if that be possible. ~o state wiJJ leave its population and territory open to attack if it ~ the means to redress the situation. \\'hether a given state can defend or deter or do both "is-a-\is another depends upon two fac-tors: {1) the quantitative balance of forces between it and its adversary~ and {2) the qualitative balance of for('eS, that is, whether the extant military techn ology favors the offense or the defense. These two factors are situation-specific and there fore require careful analysis of the case at hand. The compellent use of force is the deployment of military power so as to be able either to stop an adversary from doing something that he has already undertaken ?r to get him to do something that he has not yet undertaken . Compelleoce. m Schellings words, "involves initiating an action . .. that can cease. or be<:-ome ~ less, only if the opponent responds.'' Compellence can employ force ~ther ~~ <:ally or peacefully. A state can start actually harming another wtth ph}~ destruc.tion until the latter abides by the fonner's wishes. Or, a ~e can take actiO~ 0 reqwre -~ latter ~ ~ against another that do not cause phy-sica1 hann but ~ bombin~ 0 some type of significant price until it changes its beha-..10r. Amencas ~orth Vietnam in early 1965 was an example of physical oompellenre~ T=tt 's building of a Gem1an Heet aimed against England's in the two ~ uU:~ World \\far I, an example of peaceful compelleoce. In the first case,

ad...-ersarys p01 -

an~or industrial infra;trucwrE:







11iE USES OF i'UI<l-t.



\ i<.' tcong force iJ1 outb Yieh1am. ln t11e latter case .cc "'ll:\:t; h ult a battleAc . le Enctland m :JI\.!t~r to compe11 tt ed ~ l<:r to . that in an engagement threaten to cnpp . ~-~- a aene political settlement advantacteous to C.crm<m". !n both case . one m""'e 0 stnte initiated some type of action against another preclsel~: " a.." to ~e able to ~to p it. to bargain it awav for the appropriate response from the P 11 ~ npon state. The distincti~1 between compelleoce and deterrence ~~ ?ne between tht )) r . and passtve use of Il0 rce . The success of a deterrent .threat IS measured I , 1.~ active be usecL The suc:cess of a compellent action IS. measured b" h0\v . . . . not havmg to doselv and quicklv the adversal)' conforms to ones stipul at~d ,,,shes. ln the case of a negative. to show why e'nce / . som . . " succ:essfu1 det erre , one is trving to demonstrate thing did not happen. It can ne,er be clear whether on~ s actions were crucial to. or irrelevant to. why another state chose not to do somet~ung. ln the case of successful compeUence, the clear seque~ce o~ actio~s and r~act10ns lends a _c~mp?Uing plausibility to the centrality of ones ac-tions. F1~re 1 1llustrates the clistincnon. In successful compellence, state B can claim that 1ts pressure deflecte~ state A from its course of action. In suc-cessful deterrence, state B has no change m state A's behavior to point to, but instead must resort to claiming that its thre~ts were ~esponsihle for the continuitv in Ns behavior. State A may ha,e changed 1ts behav10r for reasons other than ~ate B's compeUent action. State A may ha,e continued with its same behavior for reasons other than state B's deterrent threat. "Proving'' the importance of B's influence on A for either case is not easy, but it is more plausible to claim that B influenced A when there is a change in Its behavior than when there is not. Explaining why something did not happen is more difficult than explaining why something did. Compellence may be easier to demonstrate than deterrence, but it is harder ent to achieve. Schelling argues that oompeU ac-tions tend to be vaguer in their
"' ra1

~t.\h..~ . htrted bombing ~orth Vietnam in order to com.t)(l


"tr~p a-;sisting the


(I) A is doing something that Bcannot


(2) 8 in~ action against A in order to get him to stop his intolerable actions (3) Astops his intolerable actions and Bstops his (Or both cease


(I) A is presently not dojng anything that .r . B finds intolerabJe (2) B tells A that if A changes h1s behaV1~ 1 and does something intolerable, BWl d5 . punish him (3) A continues not to do anythmg B fin intolerable

objecti\'C::S t dc:lt:rrent threats and for tl attain If ad,ersan h.t' 'ltard .time un<.lerstan-tln g wlhat reason more <.liffit-ult to an . at it is th u . ''-'1shed him to do h' compHan<.:e wJtlt ones \\1Shes is made more diffi cult. at one . . IS Th I . ere IS, howe\er. no inherent reason " ~~ t compeiJent ac-tion must be v regard to ho"' dear!~ the ad,ersan under ta dsag~ber ~an a deterrent threat \\ith L, s n '' at.IS want d ~ . I . attack me .. 111 11ot any c earer in its ulrlmate meanmg th e "trom njm. "Do not stop attacking mv friend." A state can be as confused or as clear about what tan } " . . 1 wts les to pre,ent as it can be about w I1at tt w1shes to stop The I . . . ~ c anty. or lack of 1 0 f th . , cornpellent act1ons and deterrent threats does not van ac " di e obJeCtives of b oor ng to whether the ~,en action is c:ompeUent or deterrent in nat of particularities associated \\ith the given , ~ti:e, ust rather acc.:ording to a welter ac on. orne ob f . r ~ec . er are inherenth' clearer and hence eas1 to perc:e1.-e t han 0 th 1\es. ror examole. .. ' ers. communicate more clearlv than others Some states have mor Some statesmen b. . . e tu ti o:. C bear for a gi\en objecti,e than others. It is the speCLil CS ora !11ven~\\er to nnCJ to a o intrinsic difference between t:ompeiJence and deterrenc:e thatSldete on. not any th . . . . . rmmes e ' clanty mth whtch an objective is perceived. . \\'e must, therefore, look elsewhere for the reason as to w h\. compellence 1 s . 1 h d : comparative y ar er to ach1eve than deterrence It 1 not m ,,.hat one asks tes, . ethin another to do, b ut tn hou; one asks. With deterrence. state B ks as som ak 0 of . . A tn tl11s f:ash'ton: " D o not t e action X; for if ,ou do 1 "ill b- h . . th state as \ OUO\er e , "' 1 th' 1 b , \\" b head w1t 1 ts c u . " 1t compellence. state B asks something or siate . m this b . . "I f:as h10n: a:n now gomg to ash you 0\er the head with this club and will continue to do so untd you do what I want." In the former case, state A <..'ail easih- denY Rith great plausibility any intention of ha...i ng planned to take action X. 1~ the' latter case, state A cannot deny either that it is engaged in a given course of action or that it i~ being su~jected to pressure by state B. If the~.. are to be successful, compellent actions reqwre a state to alter its beha...ior in a manner quite ,isible to all in response to an equally visible forceful initiathe taken b, another state. In contrast to compeUent actions, deterrent threats are both easie; to appear to ha,e ignored or easier to acquiesce to 'vithout great loss of face. In contrast to deterrent threats, ent compeU actions more directly engage the prestige and the passions of the putupon state. Less prestige is lost in not doing something than in clearly altering behavior clue to pressure from another. In the case of compellence. a state has publicly committed its prestige and resources to a given line of conduct that it is now asked to give up. This is not so for deterrence. Thus. compellence is intrinsicallr harder to attain than deterrence, not because its objectives are '-aguer. but because it demands mere humiliation from the compelled state. The fourth purpose to which military power can be put is the most difficult_~ be precise about. Swaggering is in part a residual category. tl1e deplo~rrnent of m~ tary power for purposes other than defense, deterrence. cornpellen~. Forre 15 not aimed directly at dissuading another state from a~tcki_ng. at repelliJlg ~tta~ks, nor at compelling it to do something specific. The obJectwes ~or swaggermg more cliffuse, ill-defined and problematic than that. Swaggenng almost ah\ays involves only the peacefui use of force and is expressed usually in one of t\~'0 ways: displaying o~es military might at military exercises and national demon~ons an~ t buying or bui1ding the era's most prestigious weapons. The swagg,er use ofti~rrehlS OPOP . al d 0( r- - e or to sa / e th e most egoistic: It aims to enhance the nation pn e a





l -Hl




. I n 1('1 onw am ll o1l S of lts ru ler A state or states man swagg.;er.., ' ord~r to look anc) 'S I . rtnt to be taken senous1 1 ot 1 1s 111 I 1 eottnc ils . v >' 1l w fe(l more powe rfuI an d unpo a ' 1 : . . . Id . . ._-=no- to enhance the nation s unagc Ill t1I<',' cs of.others ol' .1 L mtemationa eciSIOll l11cuu o ' . 1 11 . ..;on's ticfense deterrent an<. c:o111pt ' nt <:aj)ahiliti . it. imarre IS en1 l:II1Cecl ' tlle nau l . ' . . . E'~ l" 1. . . d l>tlt 5..,vao-celi nu 1 not uncle1taken soleI) r H. < ' t.n p1. s 11nan J 01. vr mav also oe en1 1ance , oo o ' specific Swaggering is pursued because 1l o r1 .. to 1 ng ))restig et-; m ' these puq:>oses. cl . t he cI .. swagge 'ng J s r>ursued becau se of tl iC fll n HIJI('IIlal vean1iJI{T C' on 1eap. n . . /:') 0 f' states and sta tesme n f.OJ' l'"' SI)ect and r)restigc. Swag_g;enn g IS more '>0111Clhing to b~> "" .. -1 .. r ]f.! to 1 e etOI)]O)'CcJ for a S11eclflc. COI1SC !OUS \ l 1 enjoyed 1or 1 tse u 1a.J1 1 . . HJ111!)11 - 0.lll cnd. And vet, the instrumental role of swaggenng CatlnO~ be lc~lally dJs<.:ounted becau se of the fundamental relation between ~orce anti fo:cJgn poll<:~ that it oht ain~ . 1 on m an anarc1 . envtrm111ent Be'"'LISe there JS a connect1 . btlwe en tlte mil ittn 11c ...... . < J . . cht tmght tb at a na no11 15 tllot10 to possess and the success tl1 at Ll achteves. 111 attnini11,g . . ob.1 ecn ves, the enJ1ancement of a states statUI e 111 the e)es ol. othe1 can alwa,s 1ts . . . . . . _ , . ,. be 'ustified on realpolitik lines. If S\.vaggenng eauses other tales to lake ones 1 nterest~ more seriously into account, then the i~terests of the state \- i ll benefi t. v Even in its instrumental role, however, swaggenng 1S unde ttaken less for any rri,cn end than for all ends. The swaggering [uncti on of powe r i thus at one and the same time the most comprehensive and the most diffus e, the most versatile in its effects and t11e least focused in its imme diate aims, the most instrum ental in the long run and the least instrumental in the short run, easy to j~sti fY .on hardheaded groun ds and often undeJtaken on emotional gro~nds. Sv.r~~enng m1 the rational .xes and irrational more than the other three functi ons of m1htruy powe r and, for that reason , remains both pervasive in intern ational relations and elusive to describe. Oefense, deterrence, compellence, and swaggering-these are the four general purposes for which force can be employed. Discrimina ting among them analytically, howe ver, is easier than applyi ng them in practice. This is due to t\.vo factors. First, we need to know the motives behind an act in order to judge its purpose: but the probl em Le; tJ1at motives can not be reac:UJy inferred from actions because several motives can be served by the same action . But neither can one readi ly in1er the motives of a state from what it publicly or officially proclaims them to be. Such statements should not necessarily be taken at face value because of the role that bluff and dissimulation play in statecraft. Such statem ents are also often concocted with domestic political, not foreign audie nces in mind , or e lse are deliberate exercises in studied ambiguity. Motives are important in order to inte rpret actions, but neither actions nor words always clearly deline ate motives. It is, moreover, especially difficuJt to distinguish defensive from compeU ent actions and deterrent from swaggering ones unless we know the reasons for which they were undertaken. Peaceful defensive prepa rations often look largely the s~me as peaceful <:.-ompellent ones. Defensive attacks are nearl y indistinguishabl e fro m oompellent ones. Is he who attackli first the defender or the compelle r? Deterrence and swaggering both involve the acquisition and display of an era's prestigious weapons. Are such weapons acquired to enhru1c:e prestige or to dissuade all attack ~ Second, to make matters worse, consider the following exam ple. Germ an~, launched an attack upon France and Russia at the end of July 1914 and there~) began World War I. There are two schools of thought as to why Germany did t]us.

holds that ; . II!Otl\ <:~we re acrrress . 147 . l:'lb vc-ternto . I n ifl and elc,atJ ll .,. the slat us of' a world . . na aggrandizement g... ' . enp1 A th re . economic; e prevc n lJ\'e "11' 1 1 nee c.lcl(;nsive Sh w no er holds that h, . wer , _ .. ._ . e struck first b . <:r motl\es lernent, slo'' slt ,u ):. d.ttJon , and then inevit 11 ecausc he feared ~ . c I 1 I' I a) e attack b . h s foes w wm 'l u e t were daiJy incre - h . ) er two powerf11 t enclr. 1OJ > -<ts1 t e1r l t ng . ne1 1 g 1was. She struck whde !>he had th e chance to 'Win 1111 I ary ll11ght faster than she It is not simpl e to decide which school th 1 e more ne 1 5 1 . can marshalI eVJc en<:(' to budti a powerful cas . ar ycorrect because hoth I 1 .1t the secon (l 1s <.: oser to t 1e truth . There e. th . Assume for tl1C 1nomcnt, tJ 1ough tl1' I en two bJ (1 ) Germany launc 1eti an attack becau se aleWCJ.<; the .poss1 11 to consider:, it ties lanning to attack her ultimately anti Germa h d ~e that htr foes were P ,. ny a t1 \'.d e (2) Germany felt she ),ad reasonable evitie nce of her ~ , .e 1 enc:e to prove it: or tual ly but in fact her evide nce was wrong bee ohes 1 ~tellt to attack her even. aust: s e mtsperc:e , d h . . 1 l e1r mtent from 'their acbons. I f tI1e fi rst was the case the "e . , , . n ask 15 responstble was Germanys cliplomac:y in the fifte we must b r th question.. How en vears . and blundering as it was, i11 breeding hostility in her ~ei b eore 1914 aggress1 ? ve in the knowledge that they would eventually ha,e struc~ orb ~ferm~y attacked s. r di pIomat1 c record was a SlgJ1tfi cant 1actor in c:ausing th er, Iut 1hher fifteen-\ear t ' . em o ay t ese plans must we coneIude t1 Germany m 1914 was merely acting d r . ? \ 1at . d b . e,enstve1 .. lust we confine our JU gmen t a out th e defenstve or aggressive nature ofth y . h 1. . to or even the year m w 11 1t occurred? If not how man)' )ear b ekact hi t e month . .ch s ac storv do we m go m ?rder to make a judgment~ If the second was the case. then we must ask this queshon: If Germany attacked m the belief, mistakenly as it turns out, that she would be attacked, must we conclude that Germany was actino- ~1 ust we confine ~u r ju.d& ment about the defensi\'e or aggressive nature of the a~t simply to Germ anys beliefs abou t others' intent, without reference to their actual intent? not easy to answ er tJ1ese questions. Fmt unately. we do not have to. Asking them IS enough because it illustrates that an asse sment of the legitimacy of a state's motives in using force is integral to the task of determiJting what its motives are. One cannot, that is, specify motives without at the same time making judgments about thei r legitimacy. The root cause of this need lie in the nature of state action . In anarc hy eve1 state is a valid judge of the legitim acy of its goals because y there is no supra national authority to enforce agreed upon ruJes. Because of the lack of unive rsal stand ards, we are forced to examine each case ,,; thin its given context and to make inc:Uvidual judcments about the meaning of the particulars. When individual judgm ent is exer~ised, individuals may well diFfer. Definitive answe rs are more li kely to be the exception rather than the rule. Wher e does all of this leave us? Our four categories tell us what are the four. possible purposes for which states can employ milit;uy power. The attributes of each alert us to the types of evide nce for which to search. But because the c:ont~xt ~.f an action is crucial in order to judge its ultimate purpo~e. these f~ur c~tegones ~anna t be applied mindlessly and ahistorically. Each states purpo ~e ul usmg fo.rc~ 10 a given instan ce must faJJ into one of these fou r categories. We know a p~n what the possibilities are. vVruch ooe it is, is an exercise iH judgn.lent, an exercthlSe that depends as much upon ase the particulars of the gJVen ~- as tt does upon e general featu res of the given category .. . (See Table l).
0 ne


l -t '




ve Defens1




Chat act< ristics



The Diplomacy ofViolence


Fend off attacks Peaceful and physical and/or reduce damage of an attack

Primarily military Secondarily industrial

Defens1ve preoarations ea h n ave . . OISSUa!.IOn value Defens1ve preparations can look aggress1ve; F1rst stnkes can be taken for defense.


Prevent adversary from initiating an action Peaceful and Compellent Get adversary physical to stop doing something or start doing something Peaceful Swaggering Enhance prestige


Primarily civilian Tertiarily military

Threats of retaliation made as not to have to be carried out Second strike preparations can be viewed as first strike preparations. All three with Easy to recognize but hard to achieve; no clear Competent actions can be ranking justified on defensive grounds. None Difficult to describe because of Instrumental and irrational nature, Swaggering can be threatening.

The usual distinction between diplomacv and fo rce . not merelv n th . ts , 1 in the relation between adve , . . h ' 1 e mstntments, words or bullets, )Ut rsanes-m t e i 11t I f . c erp ayo motives . and the role of communic.:ation, understanclings <: and restra 0 t o1 1 . . . . " macy JS barg<unmg; 1t seek<; outcomes that though not I'deal r ett herparty . P ob ror . arc et1an ter for both t1 some of the alternatives In diP1 1 omacy eac1 party somewh t a . . 1at controls w1 tI1e otl1cr wants, and can get more by comprom1se. exc1 1ange, or col. . . , . , . gn . laborahon than by tctkmg thmgs m h1s own hands and 1 onng the ot1 ,s w1shes 1er . The bargammg can. he pohte or. rude' entail threats as well , oners, assume a status as er . quo or 1gno~e all ~1ghts. and pnvileges, and assume mjstrust rather than trust. But whether pohte or mtpohte, constructive or aggressive, respectful or v1CIOU.S, w het1 1er it occurs among friends or antagonists and whether or not there is a basis for trust and good,vill, there must be some .common interest, if only in the avoidanc-e of mutual damage, and an awareness of the need to make the other partv prefer an out' come acceptable to oneself. With enough military force a country may not need to bargain. Some things a counhy wants it can take, and some things it has it can keep, by sheer strengtJ1, skill, and ingenuity. It can do this forcibly, accommodating only to opposing strength. skill, and ingenuity and without trying to appeal to an enemy's wishes. Forcibly a country can repel and expel, penetrate and occupy, seize. ex1:errninate. disarm and disable, confine, deny access, and directly frustrate intrusion or attack. It can, that is, if it ha<; enough strength. "Enough'' depends on how much an opponent has. There is something else, though, that force can do. It is less military, less heroic, less impersonal, and less unilateral; it is uglier, and has received less attention in ' Westem military strate!:,Y. In addition to seizing and holding, clisarmil1g and confining, penetrating and obstructing, and all that, military force can be used to ~t~rt. In addition to taking and protecting things of value it can d~stroy ~lue. In additiOn to weakening an enemy milit;uily it can an enemy plam suffenng.


eone o;,e it to ~ulThere is a difference between taking what you want ano llla.J~.ulg som t you, - ~-=~ , }ou, between fending off assault and ma.J( so111 eont> afraid to assa take it c. d t . h d makinu t em amu o .1. . l b tw 0 e een 1olding what people are tl)'lng to ta.J~.e ai1
.I - 1..: .

,.. Go vriuht 1006 b) Yal~ Uni\'ersi~ . F rom Thomas C. Schelling, Arms and Influence. PP l-34. P. e Reprinted by pem1ission of Yale University Press.


' "'"

....,. ..... - -(l

- .. O..VI11\\,. y

OF VlQl .,...,

. IJtt smmoue can rorciblv take and gid11g it up t .t\oid risk Ol ~ l'~'[\\ t't.'l"l l0 :->111f; \\ ' ' l (' _utd dCl<.' ITC'llt'P het ' l ' I brute (' . , . I di 11' ,1 , 1we between t e enst: ' . 0 1<:c;.\~t'. It t, t le c t t . nd blac:kmai I. between ac im 1 tlld L rcats h ] . - . . l t' l ctween congues cl .. 1 .mu tnhtnlt a ron. > . aJ .. 111tdipl omatic" rec:onr'>t h1 c;tnrwtl 1 . t is the diflt'rence between the 11111 1ater: "' ctnd , . ] I . . b s don the power to hurt. t"'t'r{'t\'e c tp onMC) ,t e _ Tl ely "militm) ;' or "uudiplomatic:" recou rse t e,cru1 1e pUI T I1e contn\S ts are s ' ' . t1 encrth no l c n cn l\ inlcnsls; tht . , o . '11 . s e> <.:ot r{orct) e action IS concerned w1th enemythe vel) ' CXJ tJ u]l 15 . 1Ioitation of e nl'l11\' wa nts . , f -} er tO 1 wrt, lOLio ' . . . ' 111 1 C (;' use o t le pow 1\' th . aJiy measured relative to enem.' st rength , tl1r on , r Al cl b te stTeng IS usu e tears. 1 ru ile tl. te power to hurt is t\pi callv not ncl11<:cd bv th" 1 .J: . tl ' O posinutheot1 er,w1 1 - 1 ~ twec ~ P o . turn Opposinu strengths ma) c<mce c~ach olhcr. pain enenws power to 1 m re twt o , l: 'I] ss. to bwt tlte ere dbJ o f a tl1rcat. antl t Iw ahilih, 1 1 rtv . d ef do not T 1e w1 mgne ' " '' an gn . h t . 1 indeed depend on how much lhP achc rsarv <:an u to ex.T'1]01t tl1e power to u1 "' ' . . . -~b l tttJe or nothing about an ac1 ary pam or gncf lhat ,er. hurt m return ut t 1ere rs 1 . d . ,..,n T..vo sides cannot both overcome eac:h otlwr "~th dLrect1 re uces ones 0 . )' . strengtl1; tl1 111 ay both be able to hurt each ot I1er. \\,.1tl1 strc>ngth they c:an supenor ey tl . .n spute ob~ect s of va1 e ; ,t]1sheer violence thev can des trO\' lt>m. u ' ' . di And brute force succeeds when it is used, wh.ereas the pow: r to hmt IS most successful when held in reserve. It is tl1e threat of. dan"lage, ?r o! more dam ..~gc to come, that can make someone yield or comply It 1~ latent ,,~)e~cc that can mAuence someone's choice-viole nce that can still be \\1thheld or ~~Hc~ed or that a ,ictim believes can be .,vithheld or inflicted. The threat of paw tnes to structme someone's motives, while brute force tries to o,ercOJ~e his strength. Unhappily, the power to hurt is often communicated by some perforrn ance of it. \\ heth er it is sheer tenoristic violence to induce an irrational response, or cool premeditated violence to persuade somebody that you mean it and ma~ do it again, it is not the pain and damage itself but its influence on somebody's behavior that matters. It is the expectation of rrwre violence that gets the wanted behavior. if the power to hurt can get it at all. To exploit a capacity for hurting and i.nfucting damage one needs to know what an adversary treasures and what scares him and one needs the adve rsary to understand what behavior of his wiiJ cause the violence to be inflicted and what will cause it to be withheld. The victim has to know what is wanted, and he lll<l)' have to be assured of what is not wanted. The pain and suffering have to appear co11t in gent on his behavior; it is not the threat alone that is effective-th e threat of pain or loss if he fails to comp1y-but the corresponding assu rance, possibly an implicit one. that he can avoid the pain or loss if he does comply. The p rospect of certain death may stun him, but it gives him no choice. Coercion by threat of damage also requires that our interests ancl our oppo ne~t's n?t be. absolutely opposed. If his pain were our greatest delight and our satisfac:ti~n Ius great woe, we would just proceed to hurt and to frustrate each other. It 15 when his pain gives us little or no satisfaction compared 'vith what l~e can do for us, and the action or inaction that satisfies us costs him less than the p<Wl we can. cause, that there is room for coercion. Coercion requires finding a bargain. for him to be better off doing what we want-wors e off not doing what we want-when he takes t he th reatened penalty into account. ..

This cliffcrc.:lt c b( tw<t n e:o(rc:ion and h t . r . 'l ,..strumcnl. .i P In rat cI0\\11 Comancl a and tc- orcc ~ soft<:n In the intcJat . e nl . tl1e 1" their . o cxt<:rmi . t . 1 a\ "' .,;cl V11l<i.!<-" to 111akc th<'m behave na <: ltern was hn1t,- r, to r.... w~ e:oerc:i\. cl' I , ,, >r<<: wer to hutt. The pam and !()ss to the lndians . - I t! 1p c>tnat). based on tlw , po mlg l t la\'( look cl way as tJ ollwt; t1 t ,.rr 1e 11: 1 Nen<.:e was one of e . nuc:h the sanw one cl because l IH?\ \\'C'rt' . I PHlp<.lse and lt r 111 t w way or somebocl <: ec:t. r [nJ jan_o; \\E!rc: kilje . ' . ' Y . clesprured of makmg lhe m behave and <.:o I' wanted th C Ianc1 or the authori. tr , oes u ~ 1 was pure unilateral force u not <.:on fine tlem anc1decided to terrninate them, t 1at lf . . . ex 1 l ther Indians behave. that W<LS coercive violence ~0111e ncl1 11s were killeel to make . cl < o . . Th G , . . . --or 't was effecb\ e. e et mans at Verclun I)erc:e1 . cl lnten ccl to be, wheth er. or not h 1 ' e t em selves t0 b h ndreds of thousands of French soldiers in a gru . .. e c ewmg up l)l1 esome ose was to eHmi.nate a military obstacle-the Fren h. r meatgn ncler.,. lf t he purP C: IOJantnman r tary "asset" rather tl1;m as a warm human being-the ffi " . vie~ecl . a rrn'j1as 1 r ve uniJateral exercise o f 111 1 1tary 1 0rce. If instead the ob O ens1 at \ erdun was a t . " JeC was to make the Jo 0 f oung men-not o f 1mpersonal effectives," but of sons h b' ds r 1 s ) . . , us an , 1athers and the ride ofFrencI1 man I l-so angUtshmg as to be unend 11 1oot k P . f' l .I l . ura) e, to ma e surrendE:r a welcome relre anc to spor t 1e foretaste of an AJlied victory th 1 t . . 1 d 1 en was an exerc1se m coerciOn, 111 app 1e v1o. ence, mtended to offer relief Llpon ac, commadation An cl of . course, since any use of force tends to be brutal' thoughtless vengeC..1 l: b , 1 w, or p run o stinate, the motlVes themselves can be mixed and confused. The fact that heroism and brutal!ty can be ~oerci:e diplomacy or a contest in pure strength does not promise that the distinction WJJI be made, and the strategies enHghtened bv the distinction, every time some vicious enterprise gets launched. . . . War appears to be, or threatens to be, not so much a contest of strength as one of endurance, nerve, obstinacy, and pain. It appears to be, and threatens to be, not so much a contest of milltary strength as a bargaining process-dirty, ext ortionate, and often quite reluctant bargaining on one side or both-nevertheless a baruain mg process. The difference cannot quite be ex'Pressed as one between the use of force and the threat of force. The actions involved in forcible accomplishment, on tl1e one hand, and in fulfilling a threat, on the other, can be quite different. Sometimes the most effective direct action inflicts enough cost or pain on the enemy to serve as a threat, sometimes not. The United States threatens the Soviet Union witb virtual destruction of its society in the event of a surprise attack on the United States; a hundred million deaths are awesome as pure damage, but they are useless in stop~ing the Soviet attack especially if the threat is to do it all afterward ~nyway. So it rs worthwhile to keep the concepts distinct-to distinguish forcibJe action from.the threat of pain-recog nizing that some actions serve as both a means of forcible accomplishment and a means of inRictinu pure damage: some do not. Hostages tend to entail almost pure pain and dam;ge, as do all forms of reprisal after th~ fact. Some modes of self-defense may exact so little in blood or treasure as to ent~J negligible violence; and some forcible actions entail so much dolence that thelr threat can be effective by itself. . . 1 lS The power to hurt though it can usuallv accomp1 1 no tl ung directh' . potenIS " rIaJ1Y more versatile than a straightforward' capacity 1 'bl cromplishment ' for orcl e a . h . By al . orce one we cannot even lead a horse to water-"e ha\e to drag hun-nlUC



. . . tion any collaboration d n ~t anything, Out I . ' .J . . k A . am nnatJve ac . requtres t 1at an opp()nent or a \IJ<;. . less make him wu1 . ll) . or e>.tennu1 a~on ' . u I P h,-sic~u exclusion, expu swn. I)' to stop or get out. The. threat of pru.11 anu dal"' ., age .1 . 1f on d ytlliJ10' he can do IS polenha lv sus<;eptiblc tim do sometlung. even . l) . o . Iurn wruI t to do It ' an . an omplish what requtres nt) eo. aI)Oration 'fhlO !Till)' make e y (I e can on 1 ace inducement. Brute ore ' l . e of unarmed eo m bat: One f'an disable a m . 1c l1l.m lo Jai l one has an . . Ie ts ill us t ra ted b)' a tee uuqu l killing blows, hut to ta . i p1 nc1p or fractunng.'C e along" holds are those t l1at threaten pai o r bv various stunnmg, 1 1 or efwrts om / exploit the ma~ ~own . f j0 n as tile victim complies. giving hi111 the option of' disablement, gMng rehe as g 1 own legs to getto jail. e pain and damage can be use d or threatened t . usmg us o d . al. . ence-pur Th fact that VlO1 to paralyze . t. date and to blackmruJ, to emorc IZe and e ' . l . b coerce and to deter, to m mu f dirty bargaining, does not y any means 1mp y that vio. . in a consctous process o and meaningless or, even w1 purpos1ve, m danger of 1en t 1 ft . "r lence LS not o en wan 1 wars were often qUJ.te "tota1 101. .t he 1 er, the men os f J d A1.1 ent . 01 d tl an getting out o 1 t t d tll the women sold as slaves, the boys castrate , 1e cattle slaugh. bemg pu o ea ' 1 . f tered and the buildings Jeveled, for the sake o ~evenge, JUStice, persona gam, or ,y custorn. If an enemv bombs a city, by .des1gn or by carelessness., we usuallv . , mere1 bomb his if we can. In the excitement and fatigue of warfare, revenge IS one of the few satisfactions that can be savored. . .. Pure violence, like fire, can be harnessed to a purpose; that does not mean that behind every holocaust is a shrewd intention successfully fulfilled But if the occurrence of violence does not always bespeak a shrewd purpose, the absence of pain and destruction is no sign that violence was idle. Violence is most purposive and most successful when it is threatened and not used. Successful threats are those that do not have to be carried out. ...






Pure violence, non military violence, appears most conspicuously in relations between une~~al c:ountries, where there is no substantial military challenge and the outcome of military engagement is not ill question: Hitler could make his threats contemptuously and brutally against Austria; he could make tl1em, if he wished, in a more refined w~y ~ainst Denmark It is noteworthy that it was Hitler, not his generals, who used this kind of language; proud military establishments do not like to think of themse~~es as extortionists. Their favorite job is to deliver victory, to dispose of opposing ~ilitary force and to leave most of the civilian violence to politics and diplomacy. But ~there is no room for doubt how a contest in strength will come out, it may be pos ible . S to. bypass the mili'tary stage altogether and to proceed at once to the coercive bargammg. A t}pical confrontation of unequal forces occurs at ilie end of a war between '~c ce ' . tor and vanquished Wh was vuln rable aft . ere Austna was vulnerable before a shot was fired, Fran are th ~ wh er Jts military shield had collapsed in 1940. Surrender negotiations fteere the threat of civil violence can come to the fore. Surrender oego~p are o n so one-sided, or the potential violence so unmist'akable, cllat

. ' ctininc stK't'('l'ds ~11 1d tlt(' vi olenc:~: re 111 ains tn resene B l th f: . be:l l o" n r th. cuware ac.:t thatJOost of the . -tual dam<tg(' was dl)lj (' dming the military st . . age o . l . ,tc pnor to \lctorv and r. at re>t does not .rue,m I 1 vrolc uc.c wa'i idJ( m ti IC a, term tl ' I . de r~ " ' 1 a on Ythat it was latent and the th real ~>f it stt<;<;t.><>sflll. . . . l The Hussm11s <;rmhed Budapest in 1956 . d c.:owec Pola d an . n There l'>Odng coun tn es. 1 I was a lag of ten yetrs betwceu m.. ta and other neigh. lll ' , . ry vtc.:tory and this ..)1ow of violcn<;f', Hl t l w prin<:iplc was the one [ t] I . . M'l' llned JUS exp c . 1 cl :> .t . d h r 1 tary v1c.:tory . 5 often th e prC' u e to vJolcn<;c, not the encl of 1 ruJ vto an t c rac:t th t . 1. h a ~uce:ess vc lence is usually held m rese1 should not deceive . b us a out t t role 1t I . . 1 . What about pure v1o en<;e during war itself the . Ht<.tion of pa Pays. f~ . d 111 ' . as a military techn1que? Is the threat of pain involved 1 ..111 h m a~. su enng on) t e poltttc.:al use of victory, or is it a decisive te<;hniq uc of war itself? . Evidently hetween tmequal powers it hac; been part 0 f . r wanare. Colomal con... .. . quest has orten been a matter o1 punitive expeditions" rath er tl 1 uenume mJJ 1tary r tJ tn.b esmen escape into the brush you c,ln anurn thetr villages . . .. b 0 engagements. I .1e . . . ' w1thout them unt1l they assent to re<;eJve what' in strikingly modem language used , .. ' to be knov.rn as tI1e Queens prote<;tion.". . . Pure, as ~ miJitary tactic, appeared in some of the militarv actions agai~st the pl~tns I nc.lran~. In 1868, during the war \vith the Cheyennes.' General Shendan dectded that hrs best hope was to attack the Indians in their winter camps. His w~s that the Indians couJd maraud as they pleased during the seasons when ilieu pon1es could subsist on grass, and in the \vinter hide awav in remote places. "To disabuse their minds from the idea that they were secu re f;om punishment, and to strike at a period when they were helpless to mo\e their stock and viJiages, a winter campaign was projected against the large bands hidin(J' awav 0 1 in the Indian tenitory." . These were not military engagements; tlley '"'ere punitive attacks on people. They were an effort to subdue by the use of violence. \vithout a futile attempt to draw the enemy's military forces into decisive battle. They were "massive retaliation" on a diminutive scale, witl1 local effect<; not unlike those of Hiroshima. The Indians themselves totally lacked organization and discipline, and typical ly could not afford enough tary am munitions for target practice and were 110 mW match for the calvary: their own rudimentary strategy was at best one of harassment and reprisal. Half a century of Indian fighting in the West left us a legacy of cavalry tactics; but it is hard to find a serious treatise on American strategy against the Indians or Indian strategy against the whites. The twentieth is not the first century in which retaliation has been part of our strategy, but it is the fixst in which we have systematically recognized it. Making it "terrible beyond endurance" is \Nhat we associate with ~geria ~d Palestine, the crusrung of Budapest, and the tribal warf~r~ in c : ntral Afnca. But 111 ~he great wars of the last hundred years it was usually rmhta:ry victory. not the hurt1~1g of the people, that was decisive; General Sherman's attempt to make war hell for the Southern people did not come to epitomize military strategy .for the ce~t~ry to follow. To seek out and destroy the enemy's military force. to adue\e a cru~hmg Victory over enemy armies, was still the avowed purpose and the central ~m of ras seen as an altemattt:e to t. .1 Ame tary n<;an strategy in both world wars. M11 ac ton " barg ammg, not a process of bargaining.






The reason is not that chilized ~untries aJe 0 a,er t ;o_lt~ trti n? people Lh r .. I 'lit . wars (:\or were all of the p.u llUIMilt<; m th at tl1ey pre1~r pure y m1 a~ . ese 1 tc<.IlJH> 1 . . geogr wars . u Og\ c1nd enti re I,. Cl\'l zed .) The reason is apparentlv that t 1e " tk ap y 1po\\'ers < uring the<.: 1 of 1 _ ,..r. 1 tc ar between anvthing 1 ~e equa wa1 1 re, at eas ror a w. .a . . . f. b . 1... . . enturv endin(J in \\'orld War Il, kept coerCI\e vtolen~e rom emg < CC: ISI\ c before milita~. 0 . . \1Cton was ac1 ueved . Blockade indeed was the ''hole cne mv natic not .li . m. . 1 1 .ts .:l=t'" n l . . concentrated. on 1 nuu cu:' forces' ilie Gennan c1 ans " 10 < ecl of inJluen7.a 10 1 e . tl1e F trst \\'orld ,, .ar"ere ,;ctims directed at thew 10j ('Ountn.. .lt has never l.. . "' . . . x:cn . c1 w he th blockade of the South m the C.v1 \\ ar or ol the Central p qwte ear er .c . B . . ow. ers in both world wars, or submarine wanare agrunst ntmn- was expected to make war unendurable for t]1e people or just to weaken the enem~ forces by uen\. . econonuc su pport- BotJ1 arguments were made. but the re was no need to L.. . lfl<J Je cl~ about the purpose as long as eiilier_p~:rose was re<Jarded as legitimate and either might be sened. "Strategic bo~bmg ~f e~em_y bome~an~ was also <X:<:asionallv rationalized in tenns of ilie pam and pnvation 1t could 1nf- et on people and h the chil damage it could do to the nation, as an effort to displa~ either to tlle population or to the enem, leadership that surrende r was better than persistenc:e in \-iew of the damage th~t could be done. It was also rationalized in more 'military" tenns, as a way of selecti,ely den)ing war material to the troops or as a way of generally weakening the economy on which the military effort rested. But terrorism -as ,;olence intended to coerce the enem~( rather than to weaken him m~itari1y-blockade and strategic bombing b~- themseh-es ,,- not ere quite up to the job in eiilier world war in Europe. (They might ha,e been sufficient in the war with Japan after straightforward military action had brought American aircraft into range. ) Airplanes could not quite make punitive, coercive 'iolen(;e decisive in Europe, at least on a tolerable time schedule , and preclude the need to defeat or to destroy enemy forces as long as they had nothing but comentional explosives and incendiaries to carry. Hitler's \'-1 buzz bomb and his Y-2 rocket are fairly pure cases of weapons whose purpose was to intimidate, to hurt Britain itself rather than Allied military forces. What the V-2 needed was a puniti,e pa~load worth CaTI')ing, and the Germans did not have it. Some of th e expectations in the 1920s and the 1930s that anotller major war would be one of pure civilian ,;olenc:e, of shock and terror from the skies, were not borne out by the a"ailable technol~: The threat_of puniti\e violence kept occupied c.--ountries quiescen t; but the "-ars ~ won m Europe on the basis of brute strength and skill and not by intimida ~ not ~- the threat of ci\ilian "iolenc:e but by the application of military forc.-e. Military "1ctory was still the price of admission. Latent violence against people was reserved for the politics of surrender and occupation. The great exception was the two atomic bombs on Japanese cities. These were weaponstbei of tenor and shock. They hurt, and promise d more hurt, and iliat was u._,;;ull' purpose The few .. mall" ,]: ect 5 -- bu weapons we had were undoubtedly of some wr vcuue t their e ad = :tan sense.a.L~ U .ted nonnous vantage was in pure violence. In a mw . rities u~e a m States could gam a little b}' destruction of two Japanese dustnal in cMiian m . Himduma was tfu; se~, the Japanese could lose much. The bomb that lut DOt the dead ; at all of Ja~an. The political target of the bo~b "':~ or the factones they worked in, but the sUIVJ'ors

TokyO- The tw< hr> s '>~ wtre in the tradition of Sheridan , .. , 1 Sherman iu ( ,('OrAia. Whether in the en-1 th . agamst the Comanches <J ilC sa -ted them. japaHesc t1\'(:;S or American lives u h ose two .bombs ved I'1ves or th . was I r ,w punttive {;()(: . I . uglier th an strah; 1t orward military force or ,.,e er .. . rcvt V!o enc:e L S h .. ore CIV!1 t7.ed ,vhethe t . . ore or less human(" t an military destruction w ' r error IS Ol . I d . . ' e c:an at 1 perceive th t th east ombs on Htros urna an :\agasaki represented viol . _ a e b 1 k nd not mam y an atlac on Japan . material strengthence agrunst the <.'()untrv itself s Th re a . 1 e e 'ect of the bombs and their purpose, Wa!> not rnam y the military destruction they accom lish cl b ' pain and the shock and the promise of more. P e ut the


Man has, it is sajd, for the first time in hLstory enough military power to eHminate his species from t~e. ea~h, weapons against wbic:h there L'i no conceivable defense. War has become, 1t IS srud, so destructi ve and terrible that it ceases to be an instrument of national power. "For the first time in human history," savs ~ax Lemer in a book whose ~tle, The Age of Overkill. conveys the point, "men' ba,e bottled up a power .. . whtch they have ilius far not dared to use." And Soviet military authorities, whose party dislikes having to accommodate an entire tbeorv of historv to a single technological event, have had to re-examine a set of principles that had been given the embarrassing name of "permanently operating factors" in warfare. Indeed, our era is epitomized by ,.,ords like "the first time in human historv," and by the abdication of what was "permanent. For dramatic impact these statements are splendid. Some of them display a tendency, not at all necessary, to belittle the catastrophe of earner wars. They may exaggerate the historical novelty of deterrence and the balanc-e of terror.2 \fore important, they do not help to identify just what is new about war when so much destructive energy can be packed in warheads at a price that permits advanc:ed countries to have them in large numbers. Nuclear warheads are incomparably more devastating than anything packaged before. What does that imply about war? It is not true that for the first time in history man has the capability to destroy a large fraction, even the major part, of the human race. Japan was defenseless by August 1945. With a combination of bombing and blockade. eventuaU invasion. y and if necessary the deliberate spread of ilisease. the l,;nited States ~uld probab~ have exterminated the population of ilie Japanese islands \\1thout nuclear weapons. . .. It is a grisly thing to talk about. We did not do it and it is not imagin~ble that we would have done it. \Ye had no reason; if we had had a reason. we \\ouJd not have the persisten ce of purpose once the fury of war had been dissipat~ in _victory and we had taken on the task of the executioner. If we and our enemies m1ght do h uung to each other now, an cl to others t SUc a a.1.., as we1J 1 is not be<.-ause nuclear weapons have for the first time made it feasible. . .thin "' ear .-.uc1 weapons can do it qwckJ)'. 1io com press a catastrophic war w1 the Ji . f the span of time that a man can stav awake drastically changes the po tics 0 ~'. " Process of decision , the possibility of c.-entraJ contro1and restraint the moti"'3tiODS





. tl ' k ,.,.,cl reA " h,J r ,<; il1 I) ect t.ll l>l ll)lt' in charge. and the capac1tYto Wl " rogres . ' 2oo 000 000 Huss1ans 11 f tl ' s. lt 0 is imaginablt> that we might destroy I . ' lC prcse: ... . .. . of the past. t 1 no n 1 magi nab! . nt, s thoutrh not 80 000 000 Japanese sn a was ... . c e, 1t is t' . imagined. It is imaginable because Jt couId be done m H mor n 111 the l\;VJn kLmg of m e\'e at the last trumpet... fl -1 ' . ' h 1 . . little discussion o JOW an ;u -out war might b Tlus may be w y t lere IS so "b I . I e . . le do not e.>:pect it to be roug 1t to < c osc, but just t brought to a close. Peop ryth' has been spent. It is also \\'h~ tlw idea of ''lirnt ~ come to an end when eve mg li er wars. 1 k 1 1 c: l 1 \\ oriel \ Va1 eu ecent vears. Ear 1 e 1, 't . war has become so expuCJ m r b . 1 rs 1 .. War were umited y termmatwn , )y an ending th and I I or the Franco- Pruss1 an . . 1 . at 1 negotiation th . d 0 f greatest potential vro en<:c, )}' occurred before the peno c 1 1 cl at . and privation to bear but orten prec U( c the snassh h reat of pa:m brought the t . 'e . f . .. . l e \Vith nuclear weapons avar1 ablc. t1 restramt of 0 1e ,~ o CIVI1 v1o enc ran .. . 1. . . . . ~coJne of a contest of miJ1tary str engt 1. 1 estramt, to occu r lence cannot await the OUl


I 57

at all must occur during war itself. . This is a difference between nude~ weap.ons and bayon~ls. Jt .Is n~t in the b f 1 tl: ev can eventually kill but 1n the speed WJlh wh1 rt can be eh ndum ~r ohpeop eral~.,:tion of decision in the divorce of the war from political t one, m t e cen 1 process, and in computerized programs that threaten to take t 1e war out of human hands once it begins. . That nuclear weapons make it possible to compress the fu I)' of global war into a few hours does not mean that they make it ineaitable. \ Ve ha,e stiU to ask whether that is the way a major nuclear war would be fought, or ought to be fought. Nevertheless, that the whole war might go off ~e one big stri ng of firecrackers makes a critical difference between our conception of nuclear war and the world wars we have experienced .... There is another difference. In the past it has usually been the victors who could do what they pleased to the enemy. \\far has often been "total war" for the loser. With deadly monotony the Persians, Greeks and Romans "put to death all men of military age, and sold the women and children into slavery," leaving the defeated territory nothing but its name untiJ new settlers arrived sometime later. But the defeated could not do the same to their victors. The boys could be castrated and sold only after the war had been won, and only on the side that lost it. The power to hurt could be brought to bear only after military strength had achieved victory. The same sequence characterized the great wars of this century; for reasons of technology and geography, military force has usually had to pene ~te, to exhaust, or to collapse opposing military force to achieve military Vl~ory:-hefore it could be brought to bear on the enemy nation itself. The Allies m World War I could not inflict coercive pain and suffering directly 00 the Germans in a decisive way untiJ they muld defeat the German army; and the Germ~ could not coerce the French people with bayonets unless they first ~at the Allied troops that stood in their way. With two-dimensional warfare, there 15 a tenden? for troops to confront each other, shielding their own lands while attempting to press into each other's. Small penetrations could not do major cia_rn : totheythe people; large penetrati.ons were so destructive of military organizllt!On usually ended the military phase of the war.
u..<& '

l'Juclear W<'P' t~ tnake it possible to cl0 thout first <K 1 vtc.:tory. With nuclcar . monstrous VJo1 11<:\ m~ cnce tu th<: cneny \VI '" eapons and t0 cla , ery, one expects to p< net rate an enemy homeland v.rith . ys me;m~ of dtliv., ..v force. \\'hat nuciC'ar wvapons have don out first <:ollapsing his militw 1 . fi I e, or appear to d0 1 .5 '..:nd of warfar< to ro,t p ace. :\1 lH.:Iear weapons tl to promote this 1 . nd are respoustJ)Ie r t IlC lowered status of" .1.reaten to make war Icss militarv or . m1 1tarv v1 <:torv" 1 '' ~,,.ctory is no l ml!!,c:r n prerequ~sitefior hurting tl " ' at t c prcsE:nt tirnc. ' te enemy An 11 . 't against being terribly hurt. One need not wait til h h < IS no assuran<:e rrfl_icting "unendu rablc damagc::s on his enem uOn e as won the war before t ost the war. There was a time when the assu Y . ne fneed not was't unti1 hc has . l . ran<.:e o I rna k nationa.lleaclers not just willin \lctorv-false or genume assurance coulc c but . . . tic about ,:var. Not now. g somet1mes enthusiasNot only can nuclear weapons hurt the enemy b r1 th . e ore e war has been lec:1 . and perhaps hurt ( stvely enough to make the mlt won, . . 'cl I . . l h . . r 1 ary engagement a<:ademic but 1t rs 'vVI e y assu mec t at m a maJor war that is all tl: d . .' . cl I I . ley can o. \llaJOr war rs often discusse as t 10ug 1 tt would .be only a contest in nati 1 l . ., ona cestruction. If this is indeed the case-d the destruction or cities and their popu1 ti'ons has become . a with nuclear weapons. the pnmary obj'ect in an all-out war th . . . . - e sequence of war' has ,been re\ersed. Instead. of destroyrng enemy forces as a prel d t . . u e o 1mposmg ones "'rill on the ene_my nat1on, one would have to destroy the nation as a means or a prel.ude to destroyr~g the enemy forces. If one cannot disable enemy forces without ~11rtually destroyrng th~ country, the victor does not even ha,e the option of spanng the conquered nation. He has already destroyed it. ,en with blockade and strategic bombing it could be supposed that a country would be defeated before it was destroyed, or would elect surrender before annihilation bad uone far. In the Civil War it could be hoped that the South would become too weak to fight before it became too weak to survive. For "all-out" \var, nuclear weapons threaten to reverse this sequence. So nuclear weapons do make a difference, marking an epoch in warfare. The difference is not just in the amount of destruction that ean be accomp)jshed but in the role of destruction and in the decision process. Nuclear weapons can change the speed of events, the control of events, the sequence of events, the relation of victor to vanquished, and the relation of homeland to fighting front. Deterrence rests today on the threat of pain and extinction, not just on tJ1e threat of military defeat. \1\le may argue about the wisdom of announcing ..unconditional surrender" as an ai m in the last major war, but seem to expect unconditional destruction" as a matter of course in another one. Something like the same destruction always could be done. With nuclear ~veapons there is an expectation that it would be done.. . new is the rdea that major war might be just a contest in the killing of countnes, or not e,en a t'Ontest but just two parallel exercises in devastation. . That is the difference nuclear weapons make. At least they nwy make the difference. They also may not. If the weapons themselves are ~~erable to attack. 0,r the machines that carrv them, a successful surprise might ehmmate tJle ?Ppo~ents means of retJibution. ' That an enormous exp1 can be packaged m. a smgle os1on bomb does not hy itself guarantee that the victor will recehe deadly purushment.




f: .,.,, ,.h < in a \Vestem town had ,!Jl nc1Uf;Sticm,d c.~ ~t en acmg ...-; JL "' 11 ( 1.r 1 :!IJnfiul,t>tfY<l.tih -'1 __ _..t.~. .. ._ .J:d uuarantee that both .\ \ CJlJ ( m " lo 'JU .1 one anuu~r. tl"' w not o ~"~ - (Jith L ~ ' - , ... eaw, nerrrnttmg ;t~, l!IJHr,d '"'<: 1,1 \ 1 , tnc sJl1Wer o f the two. u:;,:)s d-- ..lty weannn<;' r r. . ru"Jt back before be dled. .njght have been more concJU(1\ C to a rr rttiiiJJ~ t ~~~ . Th PrV efficiertc.:\' of nudear wcaprm~ u, d rnala tJu. 1111.1J ~ tenor: or ofcaution. e "'-; ,. '-u 11 t: . 'f h ddcnJ1.eliminatethc encrmc, <:<~pa~Jllv lfJ ,.Iux .. t 1 OT starting war 1 t e)'' can su , ' " " ':t( }' .~.. ~ rl(o.L'"Jt..:Jitv AJtd tm:: re '. a oon ttal)' r """ m , that nudcar '''<:apom arr w Jl. vulrlfraiJJ < tt, IS . ' ~ attac' and prcNe ntA to be tum'bl' effaiive agamst C=ach othtr. J)lJ'>IIlO fi(J nr<cl,,,, .. , ..., . Id r r.....,r the:.v will be ~trO\'Cd before tlH:' ::tre I:JJJndud . .J ~.~ . . at~ 5boot the m qwc y mr '. ..L lab'- vu ""'Stt:matic destnJchon of the (rtc-rm C(Jilnt" < ru1 nfl WJus DO tas,_ avaJ .e L- t the '"J .. ' necessarv r~ to do it fast rather than slowly. Jrnaginc that nudca; dc"\lnll um bad to ;, !:lowly--t}'tat the bornbs oould be drc>p~..d C>nly (~nr~ p<~r d~y. Ilw pn1'>~< 1 would~ Yi different, 50mething like the most terron'>llc ~utnlla warftm: tm it massive stale~t happcms that nuclear W'df doe not have to 7/' slov.rly; bt;t it nt:t} also ntJt have to go ~peedjly. 11ae mere ~Men~ o~ nudear ~:ap<>n'> d<J<;o, not it'o(lf determine that every1hjng mu,&t go ofT m a hlmdmg flas],, an} mcm. tha~, tl.w ,1 must gPslowfy. Su,:le'dTWeaJ)C)OS do not simpHfy thin~ cjuite that rnuc.:h.. .. In \\Orld Wan 1 and 11 one went to worlr on en<-rny military fc>rcc'>. nr,t hil people. untiJ the enemy~ mi)jtary foroos had been U.ken care of Ul(:f<: wa; typieally not anything dec.ifii\'e that one oould do to the enemy nation il\(-;lf. 'n tr Germans did not, in World War J, refrain frorn bayoneti ng Frendt citizens hy tht milliorLt in the~ that the Allies would abstain from ~hooting up tht Ccrrna11 population. 'fhey could not get at the French c.:iti7.ens untiJ they had brtached tht Allied lines. Hitler tried to terrorize Londcm and did not mak~ if. Thl.: Alli('cl air foroos t.ook the Wdf Hraigbt to Hitler's tt:rritory, with at le-a.o,t some thouglJt of doir.~ in Germany what Sherman r~i7..ed he war~ doing in Georgia; but with th( bombing tedmok~ of World War JJ one could not afford to bypass tl 1C trcKJf>S and tp eu:Jusivcly for enemy populations not, anyway. in Germany. Witl nudc:ar ~one Jw that alternative. To ~mcentrate or the ~my's military ir~tallations while deJlberc:ttdy ltolding M reterve a JruiH.ivecap'd(it')' for d~tmyiug his dti<;S, for <:Xtcrmjnatirtg hi~ fX."~Jpit and elirmoating his society. on oondition that the enemy observE: similar restraint wifh resped to orM:i own society is not the "conventional approach." fn World Wan I aod lJ ttJe fint order of bu.~iness was to dc~troy ene my armed forc.t'S betcMte that was the on~ promising way to rnal: laim surrender. To fight a pure:~ military~ ..aJI-<Jtrt" wbHe h(Jiding in reserve a decisive eapacity for violence, 00 ooodition the enemy doli~. is not the way military (}pc~rations ha\C:

~ .wo gun... t



been Jlf'Pfoacbed. ' In the f'R'Ient era oonwmhatants appear to oo not only delihcratc targtLS ~~targets, In fact, noncombatants appeamd to j,e primary targets al die ends f1f the k'ale ofw-.ufare; thennormclear war threatened 'to be~ a contt)t in Jl o( cities aod ~~; and, at tl1e other end of the scale. insur 4L, terroriltic. We live in au era of dirty war. ...,._, -10P b war properly , military affair among comhatant5, 1 a and 1s r ~ to the twentieth century that we cannot keep it wifhjn de('ent --~dirty?

national Comrr1tttee had m rmnd. ~ raat tue ~~from abtJu l 164Js {{) th(.: ~ ~1\(Jl(;.(mi<.: tra w <>r m rrHJ(..L c.L 'n 11 WE:St E 4 -' . '- .."JTJ urope W' '> SOmt:thing superimJX_,.y;.cJ on SCK...ietv. It was 4 umt.-:-t"' "nza~~ m 11)' "' bi c _._ t,b<' tlaat were:: rnea">urt:cl in territf>rit:.i> anl '""~". ~ Tfi(Jnan.. es J if 1 ~ ' u, ""'-4Slon:ul} rncmfo'\. d:v . cJaims. l11e tnY'P~ were mo\tly rru.. -r<;.E;r.aries and the. ri ..a. ; ... ~ ; - ; or . 'J'laSti<: . '-' I JC, u\ au(Jfl lOT 'A-aT 'l.d,S fined to the anstocratic elite. ~fonarchs foucrht r,,, bi f te . l oon' .. r . !Y' dents of d15puttx.~ terram were more um<:erntd wiLt ~ ''>~~mtvrv. out tht resipr . . . " th, ut c.~~Jnr E:lr <.-r:ops and their daughters from marauding troop\ than with ~tlorn lh . ~ -''- _. . I C) CJ\.. C:C1tU!eUd-Wl{:;E: to. They were, as QUJnc.:y V\ nght rema rked m hlS dasstc StutLJ of War litt1t iJK:eJTJ.ed 00 that the territory in which they lived had a n~ SOVf:Tei'.rn ;; FurtJ >of.. itrroore, as r as rar the Ki~g of Pru~sta and ~e Emperor of Austria Nere c:rm<.-emed. the lo-.~ and enthus1asm of the Bohemtan farmer WC=re not dec:.1sivt Cf.m5!..a . rations. Jt exa_o;r. ~Ut: iS an geration to ref~r to E uropean -~ar duri~g.lhic; period as a sport of kin~. but oot a gross exaggeration. And the rmHtary logJ.stlc.:s of those days <."'nfinf:d mihtasv operatjons to a scale that did not require the enthusiasm of cs rnultitudt. ' Hurting p~le was not a dE:<.isive instnJment in warfare. Hurt:in" people or destroying pmperty only reduced the value of things that were being fou~ 0\er. to the disadvantage of both sides. F urtherrnore. the munarchs who c.:.ond.uctrd wars often rud not want to discredit the sodaJ institutions th~ shared with their ene " mies. Byp~sing an enemy monarch and taking the war straight to his people would have had revolutionary impHcations. D~trO} the opposing monarchy was often not in the interest of either ~ide; opposing SOYereigns had much more in c.'Omroon with e<U;h other than with thei r own subje<-1:5, and to discrerut the claims of a moruu-chy might have produced a disastrous backlash. It is nut surprising-()T, if it i~ surprising, not altogether astonishing- that on the European continent in that partic.:ular era war was fairly well confined to military ad:ivity. One oould still, in those days and in that part of the world, be con,:erned for the rights of nonc:ombatants and hope to devise rules that both sides in the war rrright ~rve. The rules might weJJ he obSE!rved because both sides had som~ to gam from preserving social order and not destroying the enemy. Rules mJght be a nuisance, but if thcv restrk-ted both sides the disadvantages might cancel oul This was changed during the Napoleonic wars. fn ~apoleon's F~ce. people l"artd about the outcome. The nation was mobilized. The war was a national effort, not Ju"t au ac.tivity of the elite. ft was botra poHtical and military geni_~ 011 the part f'f Napol(;on and his ministers that an entire nation could be mobilized for war. Propaganda became a tool of warfare, and war became "uJgarized. Many writers deplored this popularization ofwar, this involvement of < :rdtic ITtas5es. Jn fad the horrors we attribute to thermonuclear war were after .~ for;eseen oy many commentator L ' it. but s, some before the F:-+ \\'0 rid \Var and more u;x . ~ millions of the new "weapon" to which these terrors were ascribed was r"t'-
-,. <U ' ...J

1io Wl.S\-wr !.11 ::-<!tm it 1~ uv:ful u1 dhtirl?IJlS h t}rf*- ~ I .J. ~-,.nmb-ct~tr:t - ' J' p.<tJn r" t-,nlr and Hwtr nr.., - 1 : . ~ts m mt of ' "' ' """' r.JS~~l<Jrt~ 10 th: r ,..,af:S are wort h d.s m;;m ftifl~ Lut dH-'i r "<;.Csw _. ''Jl) rJf'N"<:tf ~ ~ } t ., ... -net: lS rOt$"')v dt--.c . . .r r: rt!TV! during t W p<t-<; htt<-k huocJrE;t)vea.r!> r c L ~' :nptJW <>1 \\'~tern J:.<Uv r , 1 :~ a fl.i5tf1 nt:<J a)' . rl'>:~tfe is d1 in whir h the fX:tJpll may '' <:t IUJrt ~ . a.t : .1. ..,--'1 ~r' t7~11'JJ1 The fir>t ;>""b"" L, 1., - . 1 t 'Y ~m~JUeT'<tie oomlrr" ......- Th . status tnat pt(JP1 rJau un nn~ tnf!F>rvJirxl "'f .... . .~..l r ao.a.~Ju, ~ tS the . . . ,., r-., t'l\<11 r~A;u 'N'<ttfar tt , L







jt n a C tht 1 " test for I . na0-ona1 w:1rs ' spend ing . 1 . ' pt'Opl ~. passionatelY eugagec m 1 ;'~tha t a small .' total ' ictorv and de~perate to avoid total delee~t. ~oday '~ea;< 111 0 1 )U rn tens of rt l can cany enoug enet gy number ol: highlv trai ned pilots . . I the buildi n s t11ey Hve in; two or t 1ref' ~ t- . lons ago ther . . gfpeople using bayon ets and b,u d wire' man] . e nullions of people an<. f .... lJne ' ce111 tl1at tens o nlL ons 0 ~ I r;.,ord er ki d of destru ction anc " as eon . ns and shrapnel could create the same n . e d t ge in the relation of peopl1 to " ,r. the second tn h' gu rI That was t e secon s a Etlrope since tlle middle of the seven teenth cent~(Yd ~, t ~e lrst star peopl e had been neutral but tlleir welfare might be disre?a.r le ; tn le ~e<.loncdstage people 't . their war. Some Joug lt. ome pt o< uce mate,ials of b . were mvoh'edd ecause Id was some took care of cl''d ren: but tIwy were all part of w d d [4 n tJer attacked Poland in 1939. the Poles had reason t" war, som~ pro ~ce u~ho , aH1 var making nation v \ en When Churchill said the BJitis1 would fight on the 1 ~1.. " care about we outco me. 1e k [I the Btitish and not for a merce nary aml). T 1 war was about l ;~ lehat e er beac~uung t spomattor ed If people would rather fight a dirty war Lhan lose.a dean . b some n gove mments. 1f people between nations and not. JUStd etwee J . one, the .war ' vill be vhether the war IS continue or on t 1e te1 m o[ a h11cc makRuence on ' .' . . . . have an m s h rt eople serves a purpose. It IS a dirty purpose, bul war ttself 1 often . h bg m t e war ~1...~ gpdirtv The Poles and the Norwegians, the Hussians and the British, a out somewm r had reason to believe tllat if they lost tl1e war ilie consequences would be dirty. This is so evident in mode rn civil wars- civil wars that involve popular feelings-that we expect tJ1em to be bloody and violent..To h.ope tb~t tJ1ey wou Id be fo~gh t cleanly with no violence to people would be a little hke hopmg for a clean race not. There is anotller way to put it that helps to bring out ilie seque nce of events. If a modem war were a d ean one, the violence would not be r uled out but merely saved for the postwar period . Once tl1e am1y has been defea ted in the clean war, the victorious enemy can be as brutally coercive as he wishe s. A clean war would determine which side gets to use its powe r to hurt coercively after vict01 y, and it is likely to be worth some violence to avoid being the loser. "Surre nder" is the process follov.ling military hostiuties in which the power to hurt is brought to bear. If surre nder negotiation s are successful and not followed by overt violence, it is because the capacity to inflict pai n and dama ge was successfully used in the bargaining process. On the losing side, prosp ective pai n and dam age were averted by conce ssions; on the winning side, the capacity for inllicting further harm was traded for coneessions. The same is true in a successful kidnap ping. It only reminds us that the purpose of pure pain and da mage is extortion; it is latent violence that can be used to advan tage. A well- beha ved occupied coun~ 5 latent violence 1 is not one in which violence plays no part; it may be one in which used so skillfully that it need not be spent in punis hment. This brings us to the third stage in the relation of civilian viole nce to warfare. If the pain and damage can be inflicted during war itself, they need not wait for the surrender negotiation that succe eds a milita ry decision. If one can coerce people JJ d tbetr governments while war is going on, one does not need to wait untt 1e has an . h' ed VIctory or risk losing that coercive powe r by spend ing it all in a Ioswg"-ar . ac Iev General Shermans march th rough Georgia migh t have made as much sense' possJ d bly more, had the North been losing the war, just as the Ge rman buzz bombs an

I 6L rocke ts can ilf \f-tst umcnts to Kct the war . opped before suffering r:tlitnrv O< feat. . , f a, sm<;c: al least lh<; . Ea!>t-W t In t I1e p resc maJor .1 . . , . .1. :ap es ,...assive cv mn ' ... ' nee u n ng war itself' beyond an rth powers are c .llJ leof . '" cond Wol]' \\'.u I 1 oc:c.:~ . fo r restr . t d available dll rmg tlle ) mg lon u . . le . . ~ Se . an oe~nota . . . . .. JC . . . ilit<~rv V tOJy 01 t1uc:e . T he pnncipal restramt dunng th Wa1t the ac:hitvemen l o1 S cl - 1 "' orld War was a temporal boun a_ry. the d~tc of surrender. In the re e CC:Qnd_ W era dramatically restralllccl cl11nn g war itself' Th K P \sentw we.find the violence .. I' ar e orean V t1 1. . . ltOusly all-out" in the fighttn g. not on ) on le peninsular battle R Id b . as U ut m the e 'd ' 1ough on!} \VJ'th some dr resourccs usec.l b}' s1 es. It WclS ..1II-ou t,., l1 bot11 . . 1n ' . ~matic restrrunts; no nuclear weapons, no Russmns, no Chinese territo ry, . bombing of ships at !)ea or even airfields on the U . ryd n_o Japanese tenito1' no nJte ~ation s sil of th e ne. [t .,. . ue m ta1 was a contest m 1 1 y shcng tb circum scribed b tl1 tl civilian violence. Korea may or may not be a good y d ~ ~ 1reat of u~prec:eden ted war in th e age of nnclc ar violence but it was clramoti e ~rdspeculahon on limited ma c ev1 enc:e that tl ' ty . . le c:apac1 for violence can be consciously restrained even under th ~~~,o~ti~~ 0~ war tl1at measu res its miHta ry dead in tens of thousands and thatel ) P eoccu ptes two of the largest countri es in the world. A conse quence of this third stage is that "victory' inadequate1 expresses what v t f a nation wants rom Its 1111 1tary forces. Mostl'v it wants m th ese 'times, the . mBu' c d 1 Orce. 1at ence t.1 res1 es 111 atent L It wants tlle bargaining po' ve. that comes from . 1ts ' . 1 ct _ capac1ty to nurt, not JU St the Ulre_ consequence of successful military action . ,en :e tot~ Vlctory over an enem~ proVldes at best an opportunity for unopposed ,;olenc ag~ns t the en_emyp op ulat1on. How to use tllat opportunity in the national interest. or m some vVlder mtere st, can be just as important as the achievement of victorv y milita 1 science does not tell us how to use that capacity fdr but ~tse.lf; _ tr~clitiona~ _ mflicting p~m. And 1f ~ nation , victor or potential loser. is goinu to use its capacity for pure VlOlence to Influe nce the enemy, there may be no need to await the achievement of total victory. Actually, this third stage can be ana1yzed into two quite different ,ariants. ln ooe. sheer pain and damage are plimary instru ments of coercive warfare and may actually be applied, to inti or to deter. In the oilier, pain and destruction in war are expected to serve little or no pu rpose but prior threats of sheer violence, en'n of automatic and uncontrolle d violence, are coupled to military force. The difference is in the all-or- none character of deterrence and intimidation. Two acute dilemmas arise. One is the choice of making prosp ective violence as frighte ning as possible or hed!'ing with some capacity for reciprocated restraint. The otl1er is the choice of making retaliation as a~tom atic a'i possible or keeping delibe rate control over tlle fateful decisi~ns. The ults oth choJces are determine d partly by govemments, partly by technology. B \Glflc are ~haracterized by the coercive role of pain and destruction- of threaten~ (not tnflicted) pain and destruction . But in one the threat either succeeds or fails alto 1e 1er. . . gether, ail< 1any Vlolence ts gratm.tous; m t1 otJ progressive pain and dam. 0 age may actually be used to threaten more. The present era. r COlUltries possessmu ror od nuclear weapons, is a complex and uncertain blend of the two. em .t:. _ _ . W<mare, but rror the United States. m cti\l' , rhe powe r to hurt is nothi ng new m . . . technology has drastically enhanced the strategic imp01t an<.:-e of pUt e. unconstru

SCHELUNG I THE DIPLOMA CV OF VIOLENCE . tllo1 1ghl of ', LS t;O' e; V<;: mstr -.1 'I . .




0111 o.,, 11 1 l t fc 11 ~,. ~n . I1 tJ1er used against tt.s or in I . 11~ . . mMl'tJUi itive pain ami damage, w e Jf . ~of war and t11reats of wa1 as t<< lllHJ IL' < tn fluen"'~ . J . """' 1 1f f lt> m htm Cll laJl<.:C.'> t 1 1mportanc:e not o conqut<,l a11 < l ' t lll.ot-'; uf' bar. . cl -'eterreuce. f . . not of dtStlllction; o c:ocrcJon .Ul u ~ning :md intimidation. .k . t a contest of strengt1 . \ ' 'ar aru 1 t1 1 of w '' mnk 1 ar ks J1 e JUS . \\'ar no Iongcr Ioo _, risk-taking of pa1n and (ndurwc . '>mall w . ar, . .1 . ' are more a contest of nerve anu ..t h I t f . larger war thev are not JUSt rm ll..u; cug.tgcJ11cnt~ hut 1 J ' em houv l e t trea o a ,. Th th t of war has always been c;onu-"" wn underneath . . . ' . I 1 1 e rea s cnsL dip omac:y. b t (!0 r Americans it is now rnuc 1 ncanr t ll' S11rfac:c. l.ik<:: . aJ d' I . r . fdi mlemanon 1p omac:y. u t11e threat o vorce 111 a tamdv dic;p,Jtt 'k . d stnaJ relations ". . ' 1e t h reat of a stn e m m u tl 1 t 1mat of violcnc.:c . th party at a po~ticaJ convent1on, l 1e f I')O ltmg e 1 . . 1 .. or t I1e th reat o er poutics. i\e1t 1 strcngt 1 nor goodWill b s international . . oontJOUOUS 1y CJf<.'Ur11SCn e. Id r . procures immunity. Military strateg)' can no longer be tl1ough~ of, as 1t cou ~c!r some co11ntries in some eras, ao; the science of mjJitary victory. It 1s. now equally, d not mon.:, the art of t 1 u . . . coercJon, of 1n t'm'da:on and deterrence. The mstruments of war are more punitive than acquisitive. ~iJi tary strategy, whether we hke 1t or not, has hccnme the diplomacy of violtmce.

coercive tJiplomacy

ork: ~ acr nillan, J934), p. R2. 1. Paul 1. Well man, Death 011 the Prairie (l\ew Y 2. Wjnston Churchill is often credited with tlle term, ''balance of terror," and tlle following quotation succinctly expresses the familiar notion of nuclear tnutual deterrence. This. though, is from a speech in Commons in ~ovember 1934. "The fac:t remains tllat when all is said and donE: as regards defensive methods, pending some new discovery the only dire<.1 measure of defense upon a great scale is the certainty of' being able to inAict simultaneously upon tlle enemy as great damage as be can inRict upon ourselves. Do not let us undervalue the efficiency of this procedure. It may well prove in practiceJ admit I cannot prove it in theory-capable of giving complete immuni ty. If 1:\vo Powers show tltemselves equally capable of inflicting damage upon each other by some partiC.1Jiar process of war, so that neither gains an advantage from its adoption and both suffer the most hideous reciprocal injuries, it is not only possible but it seems probable that neither wiJI employ that means ... " 3. (Chi<:ago: University of Chicago Press), 1942, p. 296.

coercive diplomat} is, in Alexander Ceorge's words "ft . attempt to get a target-a state, a group (or gro ) ' . L~rceful persuasion'": the ups \\'ltum a state hange 1ts o1 or a nonstate 'th h )J<'CtiOnable hehavior throu h 1 actor-to c actual use of li mitcd force. It is a strategy that ~ eks tr t e threat to use force or see to per~uade a ' th cease hjs aggress1on ra er than bludgeon him into sto in : , . ~ op~ment to can include, but need not include, positive induceme~~ ; d~~erc.,~e cJ,pJomacy , can invoJve either a transfer of resou rces to the target th n rr ese 1 ~ducements or e o 'er of thmg~ that d0 . r b . I er not mvo ve resource trans1 ut that are nonetheless of ta 'bl b . ngJ e enefit tohthe tar Coerc1ve cl'1p1 . . mten ded to be an altemati" t omacy IS get. ve o war, even t ough it .. Ioyment of miJ1tary power to adueve a stat d . cl b' . involves some emp eSlre 0 . . r ath.evmg ob. . on the cheap" and e s all be Je<...tive. h . . . h ~ectives 1 It IS a tee n1que or 1t as . ts 'th b'1g resu1 . w1 small costs (to the coercer) J\ext to tri hture 1c-ause 0 u g war, 1owever. Promises . . coerciVe diplom_acy re~rese~ts the most dangerous way to use a state's milltarv power because, 1f coerc1ve cl1plomac.y fails, the state that tries it then faces two stark choices: back down or wag~ war. Th_e_ first risks loss of face and future bargaining ~wer; the second, loss of life and miJ1tary defeat. Because botll outcomes are posSible, a state s~oul~ never undertake coercive diplomacy ~ahtl~ .. we distinguish betwee~ coerc1ve d1plomacy and coercive attempts. The feature t11at dL~tinguishes tlle two IS th~ presence or absence of the employment of force. Coercive diploma~ has as one of its essential features, and often its onJy feature, the threat or tlle limited use of force. Coertive attempts utilize levers over a target. but tllese le\'ers do not involve the threat or use of force. Therefore, we have excluded from our cases of coercive diplomacy those coercive attempts that involve only the use of economic sanctions, only the withholding of benefits to a target, only the cessation of benefits that a target currently enjoys, or more generally any coercive attempt that does not entail some employment of military power. Clearly, all these actions are coercive in ~ature, but they do not constitute coercive diplomacy as we have defined it. In distinguishing between coercive attempts and c:oercive ruplomacy, we follow the convention set by George: coercive diplomacy must involve the threat or li~ited ~of force, even though it can also include some of these other types of coercJve actions. Because it entails coercion, coercive diplomacy is a fonn of compeUenct" a tenn fir~t coined by Thomas Schelling in order to distinguish it fro~ deterr~nce. f or &:heDing, the distinction between oompellenc;e and deterrence IS the difference between an action "intended to make an adversary do something"~mpellenc:e


obert J Art and Patrick ~1. Cronin. di d b . R From TJ U . . ~ le rilled States and Coercive DitJlomacr; e te r Peare Used \ \.1'th pcnmSSJOJI b' . the . ' &>-. b . ght '1:1 Ythe Endowment of the United States Insutute 0 UOJted St8 t 1 cs nstrtute of Peace Press, Washingum. D.C


I .. .11 starting somel li ng :1 terrcnt:c. Tl k<'el) IlllO ! 0111 ' . . l. . IC 1 cl to ttltl ul nelion "intenc e ll lee can be 111<.1111 Iest<'t 11 ' e o f two way . , I t b ' compe 81 . I ' lum''l' in behnvior soug I .> ~t..:.," it is not now dOJ ng. ot I I,, tH t, t rsa.rv stor) ( ~"" d tna someuw o I , s either the nd\'ers<try sttut s o -:' E'tller wav, the atlver ar~ <: l,ll l'~t., t s hchavior . ' l I doiua ometIltng t s now domg. d . 1ed, to pre,ent ctn acl'< .1.s,t ..., ' t ltn <:I1a ngln . ~"" . t is a stnttePV estgt a g Deterrence. 111 contnLS o.. .ti hr\(1 an action. Oetern:> m (' st <"s to CT(!t th . . . . di it from IOl aw o . .. I .n e ..... behanor bv dissua ng . tl at IS. to continue nol <1 > '' hat tt is n<t mg 1 '" t behaviOr- l ) ad,ers<U)' not to dumge t s_ ms to alter an adversary's behavi~r: dct<~rnn<:e. to keep doina."' Thus, compellenc:e ai all . . . 1es only tlueats to use force '' hen'<lS <:otnpel ~ aener \ mvo \ it tJ1e same. Deterrence o " force and the actual use o f ron.:c:. 1n a deter_ . Ive botl1 tlle threatb use to f' . . lence can JO\'O ed out t]1en bv d P 1111tJ0n, I1 ad"crsarv h"s H' . if 1 tl1 . t has to e cam ' ' ; <" 1 rent situa~on. t ~ ~a d terrence has failed. In contrast, because compellenc:e changed tts behaVJor andd ~ al of force compellence has not nc:<:essarily fallcd can entail botll threats an actu use ' if the threats are carried out. d ellence are analyrt:kally distinct slrategies they ' 1 1 l1oug1 deterrence an comp At .a d d'sputina parties contest the leF, iti ma<.:y of the star b . e connate w1 1en J o usu allY ecom all d The deterrer defends the status quo because of tus quo, which tlle{ genler- t Ygeot. tries to overthrow the status quo be<:ause of the .~... b efits it conrers; t Je ar 1 we en . . } t ews dle deterrer's attempt to mamtam t 1e status quo , 1 mfi1cts. T 1e targe VI 't . . !OJUI) ll "V coercing me (the taro-et) to accept a SituatJOn that beneas compe ence: 10U are o " If the target attempts to alter th e status quo, 1 10wcver, then fits you but not me. . . 11 VIew. tllat attempt as compelJence: 'You are atte mptmg. to. coerce the deterrer \Vl . .. . me {the deterrer) to stop defending the status quo and accept a J evJSIOn 1111t that 1 s less beneficial to me." In such a situation, deterrence and compell~nce become t 1 d sm Jarly deterrent threats can become transformed mto compelm errrung e . 1 1 . , . . . .. . lent actions in situations where deterrence has failed, for m that case the would-be deterrer must decide whether to carry out its threat. If it does so, not for purposes of revenge but to get tl1e adversary to stop its objectionable behavi~r, the~1 , by definition, execution of the deterrent threat becomes a compellent action. Fmally. the deterrer may calculate that deterrence is weakerung, even though it has not totally failed, and may decide to bolster deterrence by engaging in actions that are compellent ill nature. In that case, compellence is exercised to deter. Compellence can come in three doses or forms: (1) diplomatic use-the issuance of tl1reats to use force against an adversary if it does not change its behavior, (2) demonstrative use-the exemplary and limited uses of force, and (3) full-scale use, or war-the use of whatever amount of force it takes to get the adverswy to change its behavior. The first fonn of compellence does not use force physically against the target state but only threatens use. The second form uses "just enough force of an appropriate kind to demonstrate resolution to protect one's interests and to establish the credibility of one's detennination to use more force if necessary. The third fonn is to be understood as war-the large-scale use of military power to make the adversary change its behavior. In this volume we follow Alexander George and defin~ coer<.~e diplomacy to encompass only the first two forms of compelleo~ the diplomatic and demonstrative uses of force. The third form-war-is coerCion but not coercive diplomacy, even though diplomacy is never totally absent from war. The meanings of th reat and war are clear. Threat can involve mob'l:....; g and . JJ,U.J.n movmg large amounts of military force to make the coercer's seriousness of purpose






l 65


shte lr 't )edible. as po~ 11 lt, .t) tlar target < 1 can simpl , as c . wamings. r he ot . hill g tlmat does not mean is the)~~'' the ~~suance of \'Crbal nif)St the targc.. l \\ ,tr Ill\ O]\'CS SU Stained larg ) a<:tuaJ physl<.:aJ I ISC of r0 ag..... . , < e-sca e eo b '' rc:c taraet, with tlw W': I nl ( 1 ther militarily defeating it b~ a~ operations against tht 0 rt of achic' ing Jtllpielc victory over it E'th or nngmg about its surrenclt;r sJlo I er wav w 1 that is masshe, at le<\~t to the: target. " ar mvo "es the use of force The meaning of demonstrative use is m dif'C! . nc:ult to cle1 nonstrahve should meanore the.. . ~m. down. AJthougl1 George argues tl1at onlv we have used a somc\\'hat broader meaning of dem' ~~lte hmtted~ use of force, . , l onstrath e use H h enouo-h forcc to c.e monstrate resolution and e.stabJ h credibltvow muc is "j'ust o . JS .. tuat1 to another and depe d5 on tl u , can vary enor010 usly from one 1 n on 1e nature 0 f t} oals on the one 1 1 and on the militarv capablti 1anc.. d. 1 coercer's e g , I ' t J es an mtensitv 0 f . tJ1 target, on the ot 1er. We have therefore defin d d e . . mtercst-; of cl e emonstrative use t 1 d Iary ~n cl I. . .. use. Exernpla1 use serves as both a both exemp IIOJle y ~ me ~ e of what can or v.IJII come: 'You did not believe myth t b .model and a wammg rea ; ere LS an exam 1 f, to chew on of what 1 can do to you if you do not cha Pe . nge your ways." Exemnla can encompass a one-t1rne employment of force or a r . . f L' f) use . . I . . ' 'e"' mstances o use but tl major constramt 1 t l<lt 1t ts at tl1e low end of force emplo s t I th ' le ymen , c ose to e boundary between threat ;,m cl use. Exemplcuy use meam movin st b . gju eyon d th e border of threat . mak cIear b)' the actions taken that the coercer IS dead1y senous about to -e . . escalating tl1e use of force 1f tl1e target does not complv. In this vo1 te . , ume li m1 d use can mean anytillng from one to several steps beyond exemplal) use. The f . - '1 meanmg o lim1ted ~e 1 t111s: "V f:'aJJe cl tot~k both my threat and my exemplary tt.Se serioush-; s 10u e you obv10usly need more persuading; let me now give you a better idea of tlle COriS~ quences tl1at your continued noncompliance will bring." More force i.s used but not so much such that the boundcuy to war has been crossed. A central point follows when coercive diplomacy is conceived to encompass o~y threat and demonstrative use, but not full-scale use: coerche diplomac.y has failed when full-scale use occurs. Wherever one draws the line between limited and full-scale use, if the coercer has to cross that line to achieve its objectives. then, by definition, coercive wplom acy has failed. In this case, war. not coercive diplomacy, produced the change. Any employment of force beyond tllreat. exemplary use, or limited use signals the failure of coercive diplomacy, even though the subsequent fuJl-scale use of force may succeed in accomplishing the original objectives. ~sa consequence, exactly where the boundary between limited and full-scale use Is dra\vn becomes crucial for cocling cases in which limited use in,olves escalatory steps that skirt the boundary. Such cases can be coded as either successes or failures of coercive diplomacy, depending on which side of tl1e boundal) it is placed. Categorizing such cases becomes an exercise in qualitati,e judgment.


There are good theoretical reasons why coercive diplomacy is ~cu~t. In pa~cu lar, four factors which stem from the inherent nature of coerctve diplomac> and which therefor~ operate in every such attempt. explain wh~ this technique is hard to pull 0 ff 1n acldthon , depending on the spec1 c s1ru ti'o11 two oilier Factors can .. . fi a

t 66




" hen t he)..< e present' they r11ak< !hv ~ut ~"<'!isful exercise U m;mi l'c>st thcmsdn:-s. and w nf t'Oerc:h-- diplomacy e,en more difficult.

compellence Is Difficult

Corm of comrJellenc:e an d. as I lton las Sclte!J . . l rng S e1c1 e ditJ omacy 1 a 1 F' t dete rrence I t i\ intri nsi<:aiJy,..., d topulloffthan . trs co " ... ore . 1ar ed compellence 1s 1 er b ts h ge 1 behav:ior than to keep 1ts heha"ior as is C o serv ' orn. . . diffi lt t0 get a target to c an 11 . cu tions reqUJre that tl target alter its behavtor 1 a m:uuwr quite visilJIc to 1e O I 15 'bl 101"tiat:ive taken by th e C <' r<:c r. ln . pe11ent ac contrast, . n all 1 res onse to an equal)' " I e c er 110 r the target to appear to have Ignored or to acquics . . P ce . . . , . t weats ai e eas1 deterrent t1 I . fr. e In deterrent sttuatJOns tne ta1 get c,m c:l;:um IJiausib! . . e f 1 . 1 1 . . . 111 wtthout great oss o .ac . bJ c . t. ng tltat it had 110 intention o c 1cmgmg 1ts11C' 1av1or 111 the first 1 1 l dema _1ty mam am1 . . 1 . pear to ignore the deterrent t 1 reats w 11 e not changing its .' . il' . place or Jt can s1mp Yap ty . 1e . '. Tl1e ta1get llas no such plausible deruab 1 ll1 t 1 case or con1pellence laVJor. e1 b rt brn ss n is required. Greater. face IS thus lost when a target, . . . . . because 1ts ove su 1 10 to wh1 . Jt has co rnrmtted .1ts IJrestige _. " . , . . ch . . un der pressUJe, reverses a course of action , and devoted resources than when It Simply peiststs m the same beh<lVJOr. Fmally compellence more directly engages the ~~s~ons. of. the target _state tl1an does deterrence because of tl1e pain and humiliation ml1ct: d upon It, but passions, once engaged, are dangerous and produce b~omerang e ffects: they cause the government to mobilize domestic opinion agrunst the coercer, and they increase domestic support for the target government. Both effects perversely make the government more popular after it becomes subject to coercive action than it was before, with the ironic result that the target becomes less susceptible to coercive diplomacy. For iliese reasons, compellence is harder than deterrence.

Denial, Punishment, and Risk Strategies Are Hard with Diplomatic and Demonstrative Uses of Force
Coereive diplomacy is a form of coercion, and coercion, as Robert Pape has argued, can be applied in a denial, punishment, or risk fas hion.' Denial strategies seek to change an adversary's behavior by thwarting its military strategy. Denjal takes aim at tl1e target's military forces in order to undercut their e ffectiveness, seeking to stalemate tl1ese forces rather than bring outright military victOiy ov~r the~. A su~s~ful denial strategy is one that prevents the target from achie,~ng ItS poutical obJectives with its military strategy. Punishme nt strategies seek to change adversary's behavior by raising the costs of its continued resistance. Punishment ~poses pain, either directly to the target's population or to those assets tllat are ~rnportant for the population's or the leadership's quality of life. A successful pun IShment strategy 1S one that causes the target to give way not because its militar)' strategy has been thwarted, but because the costs to its ~opulation have betome too~~ Risk strategies seek to change an adversary's behavior by raising the pro I Ih~ that it will suffer ever-greater punishment in the future if it fails to <:omp y. RISk means escalation, and risk threatens more pain to the population or to


. ts valuable as~(' \ 'uc:cessful risk str t et egy 1s one tl1. 1 at causes the t\rg""t t . . . mvinc:ed th -'t th Vay becali'>C st ), cOHl('!-. cc 0 1'11V(;' ' < tt \ 'JI fr e pam '"' V VJ su er from loomintr p '? ment is not wort ' tl < nhjcc:tives it seeks I~ 110 IS1 1 TO the CAL<'llllrtat it is applied to prod . . k 11 o fr l>C<:alt!-.e risk strategieuce ns ' coertivc di PIotnacy is inherent! . p difficult to pu } s, ape tells us . . h 1 ,, ._ f' , are sn crentl) diffitult . The}' fau or scvera rca\ons. For starters r k t . es rateg1 a ' IS s 1 .re successful to the extent tJ1at they create m t te target's mind fear of future punishment sufficicnth costlv . that the target <:hanges its behavior. As Pap ' e pornts out howe . 1 1e from damage clone m t1 present is greate th h ' . . ver, t Jt pain suffered . r antepam magmed from damage . b. 11 done in the fiut11re. T I s happens because h h r uman emgs dis . c~unt t e uture, which means they value the present more. Hisk sh Id b e conceived - ou . .1 r . d un.agmc~ 1 utUJ_'e pain hurts less ilian resent . , ac; f.uh 1rc pun1sh, ment, an poHtical consJcleratwns, mk strateo-ies are gene PU . ~~am: Moreo, er. be<.:ause of ') rayapp ledmcrementll V.'Jt1 the o ay. t d Th coercer gradually ratcheting up tl'le pain inR.i hprodduccs tnore perverse c effects: the target has time to adapt its tactics t e d o re uce . . . h e pam bemg mflic:ted, and time to mobili t de amage d.o~e, f me to . get used to t l the foreign intrude r- all of which make the target b ttze obm estic olpmlon against e er a e to to erate the pain b t1 . d I cl the J tl bemg o e out y 1e coercer. Finall)' wben pam ' . . 1 ]J.m1tec1 w hen mAlcted, as is the case by defi ti IS on y lreatened or 15 . 'tl . severe y nJ on, WJ 1 coercive diplo' . . macy, the~ a_coerc1ve nsk strategy becomes all the more difficult. -~'ffi For stmilar reasons, punishment and denial are w cu1t to execute . . . . 0 w1tl1 coercive diplomacy. After all, it is hard to infuct much pu 1 hment \\1'tl1 coer. 115 . . . . CJ~e dipl~macy: t~e limited use of force produces only limited punishment. Deliv~nng ~ m1.ted punishment is not likely to cause a target iliat cares a great deal ahout Its obJectives to change course. Similarly, the threat to denv is not denial and the li~nited use of force can produce only limited denial. Stri~tly speaking. 'coercive diplomacy cannot employ denial in the sense that it cannot use enough rorce to ~talemate a Instead, to the extent that coercive c.Uplomacy aim at denial. It employs demonstrative denial." Through limited military action tl1c coercer demonstrates to the target that the coercer can, if it so chooses, undercut the cffec:tiveness of the target's military strategy but without actually undercutting it. . . Whether the coercer intends to employ its mititaty power to manipulate risk, mA1ct punishment, or execute denial, all tl1ree are hard to bring oA when ilie employmen t of military power is severely constricted, as it is with coerche diplomacy. To the coercer, its threats and limited use are intended to signal its firm resolve to es<:alate the use of force for risk, punishment, or denial purpose - unless the target knuckles under, but the target, especially a highly motivated one, can just as easily see threats and limited use as signaling weak resohe. After all. if the c:ce~-er cares that much about its objective, why pull its punche in tl'le first place?." . hat looks to the coercer as steely detennination can appear to ilie target ~ an unwilli~tg ~ess or inability to employ large-scale use of force to attain its goal. Threa~s ~d lunJted use are not unequivocal in tl1eir meaning; they can be int~rpreted to. 51gn1 ~ botl'l firmness and weakness in resolve, depending on the perspective. of the '~ewer. j . Some of these dynamics appear to have been at "'ork m t!le Kosovo \\ ~ Before the war, the NATO allies thought iliat Slobodan .\Jilosevtc would cave m after a few days of bombing, because iliey concluded. incon ectly. tltat he had done






ber 1995 afte r a period of short ii ttJs\ive hornh 0 _ e ~ orest~ er e~ . 'n Bosni.a In 199 9, howeve r. ~1 JIO'C\ onc kp rovtd tl ln~ stn kes aga.mst elr than bo e~eld that he. could ri de out a fe,, d I\ of hombi ng~e~nl erli~e \\TOng. Appare nt y, e ' . an< al I t d tl at the alliance could not hold togethe r Ir ll engage d lll a sustaineu c cu aboe b~ . . 1 ld heavy m mg camprugn ag.,;1st him. Beli eving he c.:oucl <m tl.t'-t~ th e allianc:e 1 c.u . . 'd NATO to . rt to an extensive a1r camprugn an uJtJn1 .llc lv to t1 rc:'lte 1e 1 forc e reso < " a around campaign in order to win the war. . .. . . o .h . i6'cant punishment nor stgmficant dc111al 1!-. possible witl In sum, nett er stgn . . . di l Tl efore what coercive dip lomacv can mo sl e~L'i dv c..:o1nmun 1 coercive p omacy. 1 er ' ~ ' . th 10 creasing probabiJitv of mo re pums h men t to c:om(! if it faill cate to the target lS e . /. ~ .S nsk, and also som e mdi.cat10n o f the denta] pov.,crs o f' the coerc:cr . to comply, whicl1 1 .

b [!

S t


Jikel)' to har dc n , nCJt weaken , und tr the im , t f. are l' k . r0 rce are 1 e I\ t o Ic.acl to exe 1 nplary . .paccl u a coerClvc: attempt: threa~~ t use . exer use r' . ~ an . full-scale use of f(r<;e. flacrefo rE:. a'\ Glen S _ nplary use: . likelv to ...., o IS lead 1 to cl . I I k . n , nyuer and P I . ~ non deJ stra te , c.:nscs t tatl oo hke: garnes of c:h1<.:ken haw: a .auL Otesmtt have I r , ~"> ~trd escalatiou >~rorc t 1e:y are resolved htcaus .I n lruaerent dynamic: to-. f I. . t 2 None o L11s . lo arg ue that coe rcive . 1 'pl e ne1t 11::r. party WJ.11 gvc wav at the ~ tse . ou s t .1 only that the odd!> arc not in its favo r when w omac:v 1 ,m.e~. ~bl_Y uoorned to fail. the resultin resolves of the parties . g c:nsJs hardens the initial



credibility and Power Are at Stake

Fourth, coe rcive diplomacy is diffi cult bt:cause the ta ~t h . rge wor effects of a con f ron tat1 not on Iy on its credibilitv stak as toals n about the on b ' . bl k . " es nt o on 1 power stakes. Cre d1 1 1 sta es concern reputation: power stak .. b'l' . ts . h ty , I I . . involved m t e targ ets ea c:u atJons about whe ther to stanes,ficapa I ltles . Both are d . j d rrn or gwe '-'CiV to t.l-u:: coercer. Credib1 1_ty c:onst eration s _ e compromise difficult enough for 'the: tarmak get because they mvo lve the foJl,o,.vmg sorts of issues if the target . . gJves wav on t b'ts matter, Will th1s ?~ the coe rcer s. last demand, or is it onJy the first in a ~eries of dem ~nds? Eve n If the coerce r w1ll not dem and more, what effects will gl'ing way to thts : oer?er have on. ~th er would-be <.:oe rcers? In this regard a target is in the same s1tu at10 n that Bn tJsh leaders were in when dealinu v.ith Hitler in the late 1930s: will app easeme nt satiate Hitl er, or will it only whe t hi appetite and that of other potential coe rce rs as well? Action s in the present always set precedents for the fu ture, and the targ et can never igno re how its reactions to pressures from others will affe ct its repu tation. Power stakes are equally, if not more, imp01t ant. Giving way to the coercer is usually not cost free for the target's power. For example, when the {;nited \ations began to push for rep rese ntative councils in Somalia in \farch 1993, \1ohammed Farah Aideed , the most powerful of the Somali warlords, understood that he would lose a lot of territory and hence power if rep rese ntati,e council were to emerge in a reconstructed Som alia. He the refo re resisted the establis hment of these councils. Similarly, both Iraq and North Korea would have faced a significant weakening in their military power iJ they had acceded to U.S. demands to give up thei r progra~s to acq uire wea pons of mass destruction . ~orth Korea demanded a deal_m ~eturn, and Iraq tried to do everytl1ing to thwart U inspectors. Hen ce,_w~e_n gt'mg way mea ns that the target's fu ture capacity to resist is significantly ~ mlS~ed. its ince ntives to stand firm go up dra mat ically. In these si tuatio~ a coercwe di~l o matie demand looks to the target like unil ateral disarmament: actions that are bemg demande d of the target wea ts ken its future power. r~ nus, gi'ing wav represen. . a . , . double wha m my for the target because both its reputation for resoln. <U1d 1 ability t to stand firm ar e un clerc ut . .k . le The inh ere nt rufficulty of both compeU and ns, strategJes: the formidab.ts ent [i taskof estimating resolves. before and during crises, and the target concem hr lcl l o power and cre ar dibility stakes -all make everv attempt . t coe rcive dipt:omacr when a k rth r not e bring off. Two ' 1 asy, to other factors can camp l'cate the tas C'\en u e they are present.

Estimating Resolve Is Difficult

The third reason that coe rcive diplomacy is har d _o ex~c..:u te li~s in the fac~ that estit . 1 both berore and during a coe rc1 dip lomatic atte mp t lS a tricky ve ~n g~ow '' . affair and therefore easy to get wrong. Resolve refe rs to the stre ngth of a partys will to revail, and the balance of resolve refe rs to whose wi ll- the target's or the:; co !cer's-is rhe stronge r. Before the fact, the coe:cer can nev~r ~now for <.:~rtain whose resolve is the stronge r- its own or the target s. Indeed , th1s ts the fu nction of the crisis produced by the resort to coe rcive dip lomacy: to test the reJa tiv~ strength of the two parties' resolves. Coe rcive diplomatic atte mp ts are games of ch1cken that reveal to the target and the coe rcer which one cares more about som cthj ng and just how much more. Afte r all, if the relative stre ngth of the parties' resolves were known before coercive diplomacy began, the n the re would be no reason to begin it. If the target knew, for example, that it cared much less tha n the coe rcer and knew that the coercer was inte nt on getting its way no mat ter wha t, and if the coe rcer also knew that the target cared less than it did , the n the target wo uld mos t like ly relent at the first signs of serious inte nt by the coercer. In seri ous disp ute s, howeve r, this does not happen because each party cares inte nsely about its resp ecti ve goals. Hence, the resuJting crisis serves the function of demonstrating who car es more. Even if the c cer accurately estimates the relative stre ngt h of the two parties' :oer resolves before the crisis gets under way, this is no gua ran tee tha t the ir resolves "~U remain the same once the cris is begins. Indeed , once it begins, resolves can change and usually do~ but generally in the direction of gre ate r firm ness by both parties. Each party digs in, in order to see how strongly the oth er care s . .\ttoreover, when threats are made, and especially whe n some force is used, both sides are likely to harden their initial positions even more, because the use of force engages passions and almost always causes both the target and the coercer to stiffe n their ,...,iJls. As a consequence, both will bear more sacrific e in ord er to justify the pain alre ady suJfered.

~eady unpel them to pour in eve n more resources. The ir motto is, sacnfices illl' borne j"ncr+-if., those ..._.., (,'UJTen tly be' made." As a conseq uence, .. aJ resolves mg LnJh

Economists argue that sunk costs should be ignored whe n making cur rent deci SIOils. ~r motto is, "Never throw good money afte r bad ." Sta tesmen, howev~:: 1 ~ f?lore sunk costs because of political conside rations: the costs ~e.a '

M ul tip le Coercers and M ul tip le Targets C om pl ic at e Coercive Diplomacy . r.v C cw n.: l\'(' c1 1 11ac:y lJeco nC" '"' cn rnore dwna rH ju g in srtuuti(lfl \ i11 whi t}, .J r o1 ' . ~ th an a singl< C.<)Cr<:e r an d a stn gJe target arc prcMJtt. Jf t1H r< ,., ' l <.: . 1 . rru>rt ' < oa ltu ,, of . mav 1 't 1 1 C'\><'r<.'(rs, 1t ll >~ un1 ec 111 1ts c vera goal but more cl lc. n th fl rt II <J t the <.:<> ! . ' . ' a tl11>r1 will be divided over th e means to achi eve .th e goal. Scm Hlnllc tit< ' < li tion will _'" :ou . t<lc: d tl e g< itself If eith er 1 th e c;a.<,e, . lal evcn I> d IV S c on t .. . tlwu are r<.:(jllin cJ t<1 1 J 1c kee p t 1 nwmb ers. um'te< 111 th eir c:ffort . Tlte rub 1te~ w rc: :J.<.:tJ< ms tak(.;n le) 1 I' .. .. . t n t0 geth er catl degrad e th e rmbtary and dt. 1 0 t Ile (.'()'al I IO pl on1at1e c:fk<.:tlv<.: n(;:ss ,11tt u . . .. ' .ve atte t rf s v"'ral tarcre ts are present, 1t be r1 ; coerc1 mp . e "' <:o1n es rn on diffi cult tc cltc.iu < ~ 1 .ons t hat . n aJ ac;tl coerce tLL. 1 'I So metim es steps th at coerce on e of tl1 parU <:.., ,..1")11 ~~~ e; . . . - 'ly .... n e th.., ott1nrs to resist. Otlw r l11 n ~s 11 may he " ' actua~ encourag ; necessary l<J lav,> v r one of th e targets in order to induce th e othe rs to. coop~rat e. Th e p: es< <.:<: of tw1, :n . regu1 s that th e c:oerce r de v1 act1 1S tit at re or more targets th us .sc . 01 ult1 mately ind 11, . .. all the targets to change th ej r behavior. N e1th ...... ~~ ho ld1~ g toge th er a coal1 t10 n while maintaining its military effecti ve ness nor dev1 smg ac tJO~ s to al t~r th ~ hchavior of all the targets is an easy tas k to acco mp~j sh , an d ~~c~e~s lll .such s.~ ~~Jati~On s rcquirc:s . en fi nesse. d1pJoma"':y pat1 ce , comprom iSe' and, oftcn ti1nes, d uphc1ly. rlw !l, two or ... . more parties at either th e target or th e coerce r end, or bo th , compb <:ate what i5 already an inherently dif6cult task. Five of our cac;e s- Bosnia , Kosovo, Iraq , So mali a, an d even North Kor(;:ainvolved more than two parties, and compl ications ensued as a conscq ucii(:C. Steven Burg and Paul Shoup report that NATO 's bom bing in late Au gust and early September J995, which helped end the Bosn ian War, was a two-edged sword: ll helped bring tile Serbs to the negotiating tabl e, but it al so e~ couraged ~h e Mu sh ~!~ and Croats , who were a1lied at th at tim e, to co ntinue to resiS t and achwve all theu goals through continued battle, using NATO ~L~ th eir air force. As a conscquene;e NATO had to walk a fine lin e hetwcen hombjng th e Se rbs enou gh to bring th em to the negotiating table and bombing them not so much that it pushed th e Croa tia~ Musum alliance away from the negotiating tabl 3 e. T he Kosovo War witn essed senous conflic.1:s among the !\ATO rnembers ov er th e seJection of targets for tl,e bombing campaign. The United States wante d to escalate both th e scope and the intensity of the bombing more quickly than did many of its E uropean allies, an~ these confli<.1:s threatened tl1e cohesiveness of th e coalition and probably th e efficienc,y of the bombing campaign, even if th ey did not ul tim atel y degr ade th e campaign's military effectiveness. The coalition against Jraq . unite d during the ~ar, began to fall apart during the 1990s as members be came more aud more disaffected with the continuing costs of the san<.tions. Th e in tervention in Somalia ha~ to deal with dozens of factions vying for po wer, with two of them , one led by Nl Mahdi Mohammed and the other by Mohammed Farah Aidee d, being the most im~t. Ali M~ cooperated more with the United N ations th an did Aideedd who ~ed the Uru~ Nations and the a<:tions it took as hurting his inte rests an benefiting Ali Mahdi s. This set the stage for th e armed <.:onfrontation that led to tbe ooJlapse of the United Nations' mission in Som alia. Finally even though the United States was the only coercer of North Korea , it could no~ ignore the vieWS

ART I COEROVE DfDI ,.. ..- L.V.MACY 17 J I i nt er c~ t~ of .otl. C., oq t}, Kcm;<s. and Japan ' t I an<. ' ' 1 \ <.: ()~(; aJh.e:s in th . "S vw \-.. rnadf' IJ ~ . :m ~tnt< P\ on ~orth Ko r<; a\ llll I" r .,. . c n:gum. rh<:ir : : the fear th ey ,.,~. id If lc:a<se radi ()a<.1:iv<; mate c car ac1 1t1 s rial that CCJIJt ~ t<:u 1 hoc~.w~ of l n and So u llt Ko1 r: t a.'> wc:ll a~ rt- iunit<; the Ko 0 c p<rc ap \V hs .c fJY'(.:r both ~ J a Tf'an ar.

o n

Belief in "C ount e r- Coe rcion~~ Techniques Can Foil Coercive


cinally if the target belic.;vc.-. that it has th e ahili ty to <.:o un te tL r .'. . . , h . . ~ . . an d rmhtary pr<:'>'>Urc'> , l en coerc1 dipl oma<." r n (: cr.>Cr(;{:rs diplornatie; ve ~ k r h . .,' O("..<.:Orn<;s s) d'lm t1J 1t tl led . ,...,.J1 gen. ( , l! ally fail. fh c ta.o, or t e <:oerccr IS to convince er tht: taro,.. th t ffi . tl . c {' I SU ffering will emue 1 t 1etarget dc>es not cease itc; at.tiono, ~lth ataStl,. <.:Hntpamand If f.: L r.,.,.., foil or s1 1 cantly mttl.gate the <.:oerce rg= , 1J<:ve~ lua t it rX: gn H r's mea.~ res or in turr1 JX . .._ 1 J~ l1Sl\!> on t he 10 coercer, thro ugh wh at are call<;d c..:o untcr-coerc. :ion" te<:hn icues, which c:an L . . _ . 'I' . ical, ew.nom lc, or, 1n11 ry m. n~tu re, t hc~ the 1 ~~ lt ta . target i~ muc.:h less likely to give:po1). wa ~ilosevtc must have ~ ad e s11 ch caJcul ~ti on s when contemplating ;\: \T O's air ,,ctr again st him ,. because m th e months before th e war, Serbian military fiqures ,isited lraq to see if th ey c:o uld lea ~ how to thwart U. S. ai rpower. One the 5erbs hit upon was not to fire most of th eir surface-to-ai r mLc;stlt:s (SAM e; ) but instead to hold them back. Trus fo rced ~ ATO pilots to fly at hig h altitudes over K ovo and impede:d os th eir abili ty to kn ock out Serbian arrn or placed there. Th us. believi ng that he had the means to ride out an air war, Milosevic did no t hack down under \ATO's threats. Matters becom e especially vexing in those si tuation~ in wh ich th e targe t \,;u not reveal its co unte r-coe rcio n techniques fo r fear that doina so will negate their effective ness . Not aiJ eounter-coer<:ion tech niques are undermin ed if th e target makes th em known be fo rehand to th e coer cer. In deed. th ere are often strong incentives to make such measures kn own ah ead of tim e if doing so '"ill deter the coercer fro m unde rtaking his actions. For exam pl e, before the on set of th e Persian Gulf War, Sad dam H ussein argued th at Am er icans could not suffer hca,y casualties the way that Iraqis co uld, th ereby attempt ing to make Jraqi "illi ngne!IS ~o uffer mor e th an the Am erican s a w unte r-coerc ion tool for deterrence. ln the ~osovo conflict there is credible evidence that \1 tlose vic belie\'ed his threat to ex-pel large numbers of Albanians from Kosovo to surro d. t t der to destabiHze un mg sa es tn or them wouJd de ter the NATO alliance from lau nching the air war a~ai n ;t him . The counte r-coe rcio n measures that the target m .. . ..a1 are those that the cocrcer usl conce . . can quickly design around once th ey hec:ome kn th ereby enabling 1t to make ovm. . the threat of significant punishment once agrun d'bl In the e C'clSt'S the target ere 1 e. . has no in centive to forew f 1 counter-c:ocrcon techruque t:s arn th e coerce r 0 , . ,.;n bC:cause such foreknowledge wo uld d . cl th target's de1 ses ag... st the en egra e e coercer's at tack no t enhance deterren ce of it. . ' """ teehruques produ....,.,' a pef\'e rse The target's failure to reveal its counter-coerc.1on . di Lo It: not onJy does it make th e coer rt to coerctve p t"'IU'V more likelv~ it resu cers reso .. 1.. ~~ ru riJJ faiJ Ignorant that ~~ 1 ' -1 measures ()a l cm:o ra1ses the probability that the attem~t ~\ ~ive di lomatic 111aneuver. Con6rmght be thwarted, the coer<.-er ,vilJ begin Jts ~r 1 !unter-coercion methods to .ts dent that it can de fe nd itself the target then ap p tes c ' . clermine the <.'Oercer's gambit once 1t 1-wnin The c:oercer then works furiously to un IJ'"''s.
I,.(U . .. '- .




to the tugets counter!>; the t an~et 1t''f'On<l., in kind ; h<Jtl, i\( more into tlt<ir n "JX'c.L efT' . < dtvi$l' it$ own coun e_rs arts. In llhort . r IX' t-sist in their respectl\'e achons. pmmng \er. . . b ] . . take O and the COil I ront.t~rnn nscualates frorn tl . . VJ~~ t age of ~ercion-\ntr. TIH I~ '' lwn a target I .;:<'at, the d~'t1anucs of cnSIS tJe l,~ :.tnd exemplary use to. le lrr t s coercion techlliqtres that cmrnot be revcue<)lCI<:v~.:s 1 1 \,,;11 < 1 1 1 ll t 1 J >ssesses effective c:oun er 't ( Ja use o . forC'e>. crec 1 ><'. antl full -se: I use., . . I< l<l of dc . not find t11e coercer's threats o.r exemp 1Y . . . . ed if tire coercer wants to get ItS~v<ty. . .. . r -xpl_,;,1 why coercive d1plomac~ IS Jd'flcul t. TIH f st ou f01 ce will be teguu . r 11 these s1..x ractors e 111 .r whv, . f: ctors'' and explain c C\ Cli. if executed well coertlvc sum, . _1 are "pem1anentJy operating a Jsent esenl _,,_ c TlIe I t two are occasione:u ractors. p1~rl fi or ac1 r dcl)<'ndi ng . . as diplomacy often r~. IC' rst me ractors rn k .t c . J . of the coercive diplomatic attempt.~ ~ I r on the exact nature . diplomacv inherent1 \, hard but not 1mposs1 1 e. I 1 1ek ast. tactor-thc targct's h. I J h . J coerCive . fr . . tt belief that . has eneenve co;nter-coercion tee mgues- ma es Jt lg 1 y likcl)' if 11ot guaranteed, t11at coercive diplomacy will fail.



Based on t be Iogtc and evidence IJresented, what guideli nes')can we give to poli<:v, . . . . makers who are contemplating resort to coe rcive diplomacy. SIX seem m order.

First, coercive diplom.acy is difficult and has a relatively low success rate. Second, it is difficult to esti1rwte the likely outcome of any given coercive diplomabic ganwit. Third, possession ofmilitary superiority ovm the target d.oes not guamntee success at coercive diplomacy. Fourth, positive inducements seem to enhance the likelihood of success, but only if they are offered after the threatened or actual use of exemplai1J or limited force. Fifth, derrwnstrative denial worh better than limited punishment f or coercive diplomacy., never resort to coercive diplomacy unless you, are pTepared to go to war should it fail, or unless you have devised a suitable political escape hatch if war is not acceptable.
Because the fourth and fifth guidelines have already been cliscussed at len!rth in the preceding section, I deal here with only the other four.

andu ' 1 ( this 32 pc~n:cnt fi 11urc: I k SL ac s <:ontext b t f,ortunateh- we have ,., I 1 . u Ornc statl'>ll<. on . 1\\ ot 1er instnnnent!. and tr ' s ategtes <:om . . 1 s pare-not as much as to provide some <:ont t .tl . ve would Ilk( mt nou~h ' f coercive cIrp J<>mc.lcy. T IH.:: eviu(~nce conce ex WJ 1111 which l < pIact the record J . . rr11ng economtc o .:!cause li te i111 position of ~- . . s.tncllons provides a useful sl;,ui i n~ point. bc ) an is done to tplomacy: somethi 11 san<:tions IS rough1. . a1 ogous to what ha ppcm 111 cocrcJvc: J. try to for<:~ a change in a t<Hget's bdta,ior without going to war again~t it \I 1 . ms . lllenl o f.ea<:Il ( Jrr (1111ltarv versus eeonom,c) th . ;bt wugl, the pnman . 't ruI JCrs .,. e o ~ ectJves and , Or Procedure of botJ , are roughI) the sarnc Tlle e,,.dence on san genera1method k f C fflons rs remar diplomacv data Garv H n. abl)' simi lar to our coercive d re' < ' ~ tud . uced the most comprehen umauer, Jc da SchotL an_ u.we prot1 Ki mberly E 1 1JOt } ' h Slve . . sanctions. I1avmg revte\\'ed 120 cases in the twentieth cent111 ~ to le of economtc sanctions were w d u imposed. They cone1 dec] that sanc:tions worked 33 percent ''oft! ent 1e 1111e to pro uce . . , . r I d cs .1 modest c 1ange m tne target s polrc1 and 25 percent )f. th t e me to pro uce a < . , b .1to tJ1 e uppe d 1 maJor IJoiicy change. These figures are not identical r an ower oun~ . of the success rate _f~r our coerciv~ diplomac:y cases. but they are <:lose enough to . on support the proposition that coerc1 short of war through either economc 1 0 r eco 1 ilital) ' means Is dicfi cu It and fails more than it succeeds. nomic-m Still, the data on sanctions do not prO\i de sufficient context because it could be tlle case that most other types of diplomatic gambits also fail more than the,succ.-eed. identiary benc:hmark that shows this do~s not have Fortunately, we have one other ev to be the case and t11at helps put into comparative perspective the suttess rate of coercion sho1t of war. It concerns extended deterrence-a diplomatic gambit wherein one state (the defender) tries to protect another (tl1e protege) from attack by a third party (the attacker). Paul Hutl1 and Bruce Russett studied the unherse of cases of e:\tended deterrence-fifty-four such instances-fTOm 1900 to 19 0. They found tJ1at extended deterrence was successful in thirtY , -one of these cases. for a suecess rate of 57 percent. The Hut11-Russett study not on1~ shows that some politicalmilitary uses of force can succeed more than tltey fail but also pro,ides empirical support fo r the proposition that deterrence is eac;ier than compeUence. Table 5 displays the success rates lor these three tlata sets. Taken together, the evidence about economk sanctions, coercive diplomacy, and extended deterrence gives us greater confidence that coercive diplomacy's success rate is not due i mpl~ to flawed execution and, therefore, that coercion short of war is indeed an inherently difficult enterp1ise . . ..


I. Coercive Diplomacy Is Difficult

As I argued earlier, coercive diplomacy is difficult because coerdon in general is difficult and because coercive diplomacy is the most diHicult form of coercion. Th~ case studies [examined] ... show that coercive diplomacy has a success rate~ 32 percent. Even tl1at figure may be too high because two of our cases (Bosnra 1995 and Haiti 1994) were borderline successes and because one (Iraq 1993) ma)' 0 ~ .be a success at all. A tougher coding could easily rank al] three as faiJures, driving the success rate down significantly.

'TYPe of Strategy or Instrument

Coercive diplomacy Economic sanctions Extended deterrence

success Rate
32% 25-33% 57%

I i~



2. Estimating Outcomes is Difficult

\\'e han! seen that positive inducements and denial ~l ratl'~it" tnhan<:<;: the . 1 lihood of' success at coercive diplomacy, but they alonl. <an not g11arantec: su 'ke. because the size of the disparity in resolve between lh~ C< H n< ' I .md th<: targ~t ~~c~~ a ignificant role in detenni~ing t~e outc?~~e. Then.:~~m dt lt. l:lll. i~1in~ bcfor~ ~~~ ssf11l1s also 1 fact whether an)' given coerc1ve dJplomatiC attempt" 111 be !>lit< c 1'ff' lard (" I I and again the problem derives from the 111 leJcnt tu lCU l) 0 ('Slunating relativ~ resolves. First. before the coercive diplomatic cri~is ~e~ins, tllC' c~><.:~tc~ c~uJnot know the strength of the target~-; resolve compared With lls own. nol cttn Jt lully know I 1ow 1 r .I . . . 111 g to t t'J I'nu . Oltentim . . strongly the target is attached to t l1e interest lt JS tl) such situations, the coercer may not be full~ ce1tain :1l~oul thP ~trength of' its~\\:~ resolve and how firmly it is committed to the mtcres ts 1t 1s clefeuchng. The same ma be true for the target. To complicate matters, the cocr<.:er will finu it dif'R<.:ult / credibly communicate to the target just how strong its resolve is, because only reso~ to war will reveal the full lengths to which it is prepareulo go to prevail. War, how. ever, is generally not desired by. the coercer.since it has chos~n <.;oerc:ive diplomacy in the hope of avoiding war. It 1s therefore mh~rently !1ard.lor th e coerc:er to persuade t11c target that it is prepared to wage war 1f coerc1ve d1plomacy fails when the coercer has chosen a step sh01t of war to sign ifY the seliousness of its intent. Fmthermore, for its part, the target will. more often than not, believe that it has effecvant to hide tive counter-coercion instruments available to it, some of which it will v and some or all of wruch will cause it to believe that it can persevere and ultimately prevail in the test of wills. If the coercer cannot fully reveal its resolve anu if the target believes it can counter the coercer, then accurately predicting the outcome of the encounter is difficult. Finally, under the impact of events, wi lls can change. standing interests can take on different values, and new interests can be formed. All tl1ese factors make ex ante estimation of outcomes extraonlinruily difficult. As a consequence, tl1e United States should be wary of putting high confidence in estimates abo~t the likely ot~tcome any given coercive diplomatic atternpt. This also means tha! Jt, shou~d put llttle ~nfide~ce in the notion that the less clemanding changes in a targets behaVIor are eas1er to bnng about than the more demanding ones.



3. Military Superiority Is No Guarantee of Success

In ev~ry one of our cases, the United States possessed mil itary capabi lities far sup_tmor to the target's. If military superimity alone guaranteed success, then the I' 1 d 100 percent success rate. The fact that it had on~ Umted a 32 States should have la a , ~rcent ~uccess rate shows that the shonger adversary does not nee essan y prevrul at coercive diplomacy. 'f . . , The reason why is clear C 1 u the \\1 s (reso1 and kill ompare coerc1ve diplomatic gambits to war. In war, ves) 5 0 f th e opponents are equally matched, then the _out~ .. C.'O me is decided 1 I ~ strengtJl prev 'Is >~;et]ative mthtary capabilities, and the party with superior milit~ le weaker party hac; the stronger will, it still cannot pre\'ail ru

1 . -' ... ag.,;nst a strOill~tr at ,crsarv unless that adversary ea . res ,ke St" that it ,,,p Iqtlrt tbc Fight ' whic:h is us ua11 v not the .so httl<; about wl'at IS at ' . c:ase After all, stal<,., < 110t. wage war over things .~. ' one 0 b h onc:e the war hc:gms. u1at I1mgs that thcv both <: "'ge b are a great d ra] ot <:arc little aho\tt th e~ , ~ ,..... war onh onr t e a out. In coen..:J'-'"<. tll plomacv, in contrast .11 counts mo h . \VJ ' . Clearly, capal)lht~ counts to a degree. Both artie . re C;:aVJiy than <:apahilitv. hu~ th 5 do make estimates of ea~h other's power, cal<.:ulati~1g how much it can e other, how wtll it can def<:nd itself from th e blows of the other -1nd to wh t 1 a extent it c [! 1 h c. t c othtr's militarv an strategy. The11e power estimates do aftect I ve reso1 esp : 11 believes it possesses effective counter-coer<:ion t h '. ecla )' w 1en the target thwart l e mques that gate the <:oercer's miJitary power. However, mJ_ictary <:ap-btncan'11 k or miti'.! tJ . a 11 S ta es second place to will in coercive diplomacv situations b . military power are dif"fc rent fro~ the actual ecausef. es~l~ates about the efficacv of . '. tary use o mt 1 p . . 1 a coerc1ve diplomatiC gamb1t the coercer employs onlv th . t d . . ower. n IJ / rea s an 1 .,. .. f 1ts m11tmy capabilities. Its purpose is to si al Jmlted force not tl1' r 'u. . panoply o 1 1 is "~lling to go in its use of force v.ithout having to gn tot e ~rget JU~t how far 1t 1 use muc1 Iurce. \ \:ith miJit . 1 I ,.. I d . . aJ) h b . power pa1 t y s 1eat 1e , 1t IS the value tl1at each partv t ' pu s on t e o Jectlves at stak 1 e tbat large Iy determmes 10w many risks each \\'1.11 take and 1 manv costs ea h u 1ow . . c ''1 bear. True, confidence about its military capabilitv can st. gth 1en en a target s resoh-e ' t . but the target must have a strong resolve to begm ""th because 1 can nt\er b. tere . . . . r ta.m that 1ts counter-coerc1on techruques will work as enVIso ned It mu t thererore 1 . . . on ts value the objective h1ghly enough to take the risk that 1 co un ter-coert1 tee h. ises mques may ~veil n?t '"'o:k. A~ their core, then, coercive diplomatic c1 are akin to ch gan_1es of chicken m wh1 wills more than capabilities are being tested. In such situati_ons the more the target values ~ts objectives, the more pain it is ,,;Jiing to bear to acrueve them , and hence the less likely the coercer is to succeed. . l1~ all t!1e cases ':e_s_tudied, the United States faced targets that, although militanly mfenor, were ll1Itially more highly motivated than the United State . After all, in the bulk of ou r cases, the issues at stake between the United States and the target were vital for the target, but not for the United States, with the 1994 ~orth Korean, 1990-91 Iraqi, and 2001 Afghanistan rases being the clear exceptions. How much military power the United States is prepared to commit, and therefore how far it is willing to go in signaling its intent to commit its \ast re ources. depends on how much the United States values the interests at stake. The United States can always militaiily overwhelm such a target, but to win at coerche diplomacy, it must convince the target that the United States cares more about ''inning than does the target, and that tJ1e United States will use a sufficient p01tion of its supelior strength to prevail. This is not easy to accomplish in situations in which the target views the issues at stake as vital. As a consequence. target with strong wil~s but infe1ior military capabilities may well believe that their superior determination will offset their capabilities' deficit. For all these reasons, the United States should never bank on the fact that being militarily stronger automatically brings v1ctory i~ _coerd,e diploma~c encounters the way it can in wars. were that the case. the Cruted States could dispense with war and do only coercive diplomacy.





4. Do Not Res ort to Coe rcive Dipl oma cy Unk <:~, Sho uld It Fail You Are Prepared to Go Dow n the Path of Vv'ar c)r You Have ' Prep ared a Suit able PoJitical Escape Hatc h This advice does not imply that policymakers should e.;l'hcw coercive d.hlo 1ndel:'d. it is reasonabl e for U.S. national security <ledson make rs lous e it[ it1 mac:y. orde s1 0tt 0 f war. {' t works, is chea r 1 to achieve their objectives beca use coerc.on Per than wa2ing war. Because the odds are ag<unst ~ucces!'. however, the United States o start down the road of' coerC. wp1 ess . ts 1 l\'e .J' omacy un 1 It . wi 1 ing tor , should not esortt 11 t . I strategy tIlat WJn t:na) e . to hack d (} war, or unless it has devised a poUb ca 0 ~vtl "itJ1out too much loss of face, should coercive diplo macy fail. To resort to eoCrCI\'e l L... f. tJ.1e OIJJe cbvc w len coercive diplomac: , diplomacy and then to abandon pursuit o fails, if done too much , weakens tbe techrnqt1 e for future use ancl may well cli > credit it. AJtJ 10ugh it is true tJ1at the obj~ctives at hand d.(te nnine. to a gr~at degr:~ how both the targe t and the coercer v1ew the coercers dete rmmation , repeated use of coercive diplomacy, followed by hasty retreat~ whe n strong resistance is encountered, cannot but have a negative effect on the coer_cer's repu tation and, b extension, on its use of this technique. For tJ1is reason, if for no other, resort t~ coercive diplomacy should be unde rtaken only when the objectives sought are worth going to war for, or can some how be easily discounted politically to the U.S. public and its external audiences, shou ld coercive diplomacy fail. There is ob,~ ously a tension here, however, because if the objective is wortJ1 going to war for, it is more difficult to disco unt politically. Thus, although the temptation to try coercion on the cheap is great, tJ1e United States should not try it unless it is prepared to go the expensive route or can find a suitable escape hatch.

offense, Defense, and the security Dilemma


. Another approach sttllts with the central point of th e secun.ty dilemma th t . , . --:- a an increase m one states secunty de<.:reases the security f th the cl .o ~ conditions unde r whi<.:h this proposition holds. Two cruCJaI ers~banles exa~mesved are mvol vana d .. . c . whether clerens1ve we.tpons an pobctes can be distin gwshecl from offensJVe ones h f'I' c .. I1etl1er tI1e derense or t e o en se has the advantage. Th e defu11tions are n0 t' and w . always clear, an cl many cases are difficult to J. udge, but th ese .... . vanables shed a lWO . . .~11 great dealof ltght.on the question of whether status-quo powers \-vu adopt compat. 11 h ] . Ible secunty po teres. A t e vanables discussed so far leave the heart of the rob. lem .untouched. But when .defe nsive weapons differ from 0 cr ve ones p.1t IS nenst . ess ' ossible for a state to make 1tself more secure without making ot11ers 1 secure. < P when the defense has the advaotage over the offiense, a large And , tn mcrease one . . state s secunty only shghtly decreases.tJ1e security of the others, and status-quo pow . . ers can all enJOY a lugh level of secu nty and largely escape from the state of nature.


1. Robert P. Pape, Bombing to Win: Airpower and Coercion (Ithaca, .Y.: Com ell University Press, 1996), pp. 18-19. 2. Clenn Snyder and Pau] Diesing, Conflict among Nations: Bargainin g, Decision Alakin~, and System Stmcture in International Crises (Princeton , N.J.: Princeton Univer sity Press, 1977), pp. 118- 122. 3. Steven L. Burg and Paul S. Sharp, The Warin Bosnia-Herzegoai:no: Ethnic Ccmflid ami lntemational lnteroention (Armonk, N.Y.: M. E. Sharpe, 1991), pp. 354-355.

When we say that the offense has the advantage. we simply mean tl1at it is easier to destroy the other's army and take its territory than it is to defend one's own. When the defense has the advantage, it is easier to protect and to hold tl1an it is to mo,e forward, destroy, and take. If effective defenses can be erected quickly. an attacker may be able to keep territory he has taken in an initial victory. Thus, tJ1e dominance of the defense made it very hard for Britain and France to push Gennany out of ~ranee in \Vor ld War I. But when supe rior defenses are difficult for an aggressor to improvise on the battlefield and must be constructed durin g peacetime. they provide no direct assistance to him. The secUJity eWe mma is at its most vicious when commitments. strategy, or technology dictate that the only route to secwity lies tluough ex-pansion. Status-quo powers must then act like aggressors: the fact that they would gladly agree to forego the ~pportunity for expansion in return for guarantees for their security ha'i no implicatio~s for their behavior. Even if expcmsion is not sought as a goal in itself, there will be qUtck and drastic changes in the distribution of tenitory and influence. Cooversel)~
7 F'rom ~Cooperation Under the Securitv Dilemma" from Wvdd Pvlitic.s. \'ol. 30. ~o. 2 (January 19 8). 86-214 by Rober t Jen'is. Reprint~d with permission of Johns Hopkins L'niversity Press. Portions P~ 1 of the text and some footnotes huve been omitted.






hrus tile advmtaue status-quo stat< ". t '.lil 111: kc th,. 111 1 " 'I n H~ ' o . . . '"' ~e Vt~ 1 1 St'cure \\'1'tl lOU t gravel. endangennu others. fndc~cl. If thC' ' . le 11\V h 'L'> ,.1ougl 1 niJt~< b . .... f . l .. t: < lltt > ac ' un age an d 1 the states are of roughly equal c;m .' or dv will th C!. S<.:c:u . 1 r .J=} ""ase to t11bit status-quo states rrorn coopfratlll f~. 1. wmrc'isr< m Will 1 111( 111 ru cmma C.:.c '"'~") 1 rr lttt\(1 uni 111 . t. next to rmpossll) )e. tllus rendering international anarc '' . . \' Porta11t )(: states cannot conquer each other, t11en the ~ack of so' cnr!!;nh_ altlrcmgh it prc:,e, If e Its pro blems. of collect'I' goods in a number of areas. no .loncrcr forcc11 states to d<.:vot . . . . c: tJ. .r pnmary attention to self-preservation. Although. 1f lorcc "Ne not U\ahlt:, th<:rt 1e1 f Jj . wo uJd he r rewer restr,;nts on the use o nonm1 tan .m ~ln ~t n vntc; ' tltesc arc.: rarl"l . "" ' .-r. } powenu1enough to threaten tbe vital interests of a maJOr :-,tatt. 'r. .awo quesnons of tJ1e offense-defense balance can be ~eparated. F'irstdocs the Jl 1 .CJJ '>J.' '" r.orces to offset Cae:h state have to spend more or less than one do ar on c cf doUar spent by the 0 t]1er side on .for~es tha't ~?u.ld be ltSe<_l to att~c~? ff the state h<i!. one dollar tO spend on increasmg ItS ~eCllllty, shouJd ll p~lt. lt llllO offe~tsive Or defensive forces? Second, with a given mventory of forc: E-:s, IS rt better to attack or to defend? Is there an incentive to strike 61st or to absorb the other's blow? Th c~ . two aspects are often linked: If each dllar spent on o ff ensc ~an o vercome eae:h _o dollar spent on defense, and if both s1des ?av~ the sa~1 e de fense budget~. then both are likely to build offensive forces and find 1t attractive to attack rather than to wait for the adversary to strike. These aspects affect the security diLemma in different ways. The first has its greatest impact on arms races. If the defense has t11e ad\'antage, and if the status-<juo powers have reasonable subjective security requirements, they can probably avoid an anns race. Although an increase in one side's arms and securi ty wi ll stiiJ decrease the other's security, the former's increase will be larger t11an the latter's decrease. So if one side increases its arms, the other can bring its security back up to its previous Je,el by adding a smaller amount to its forces. And if t11e first side reacts to this change, its increase wiJI also be smaller than the stimulus tllat produced it. Thus a stable equiUb riurn will be reached. Shifbng from dynamics to statics, each side can he quite se<..11re with fon::es roughly equal to those of the otller. Indeed, if the def~mse is much more potent than the offense, each side can be willing to have forces m t tch smaller ~1an rle other's, and can be indifferent to a wide range of the other's defense policies. The se<..'Ond aspect-whether it is better to attack or to defend-inliuences short-run stabWty. When the offense has the advantage, a state's reaction to international tension wiiJ increase the chances of war. The incentives for preemption and the "reciprocal fear of surprise attack" in this situation have been made clear b~ analyses of the dangers that exist when two countries have first-strike capabilities: There is no way for the state to increase its security without menacing. or .e\:en ~king, the other. Even Bismarck, who once caUed preventive war "comm.tttrng 1 SUJ. de from fear of death," said that "no government, if it regards war as inevitable 1 It 101 even r does not want it, would be so foolish as to leave to the enemy the c1 ceof time~ occasion and to wait for the moment which is most convenie~t for~ enemy. In another arena, the same di1emma applies to the policeman m a d ~confronting a suspected criminal who appears to be holding a weapon. Thou~' ~ may_in~ be present, the security dilemma can account for many of the agtc.shootings of mnocent people in the ghettos.
r t I1<.' cIercnse




difficult. Second, because wars are expected to be~ ~'If ma e this particularly "'~11 be incentives for high levels of arms and qu ~ rdequcnt and s~ort, there , . 'I ' tc an strong reaction to th<: others mcreases m arms. 1 1e state eannot afford to wait til 1 . . 'd 1 L h . ous ev1 ence t 1at tne o t er IS bUJlding new weapons E .un 1 t 1ere ts unamh1gu. . . . .. . ven arge states t1 have 1at fruth rn thetr econom1c strength cannot wart' because the war Wl11 be over be1ore . r their products can reach the army. Third, when wars are quick t t tl h all , d . 4 h . . s a es Wl ave to recruit 1es rn a van<.:e. W1t out tht opporturuty for barg";nrng an cl reaJgnments cu 1 during the openmg stages of hostilities, peacetime diplomacy loses a degr f the fluiru ty that faci litates balance-of-power policies. Because alliances mu:~ secured during peacetime, the international system is more likely to become b~polar: It ~s hard to say ~hether war therefore becomes more or less likely. but this b1polarity mcreases tens1on hetween the two camps and makes it harder for statusquo state~ to gain th e benefil5 of cooperation. Fourth. if wars are frequent, statesmen s perceptual thresholds will be adjusted accordingly and they o,a,i lJ be quick to perceive ambiguous evidence as indicating that others are aggressive. Thus, tllere wiJJ be more cases of status-quo powers arming against each other in the incorrect belief tllat the other is hostile. When the defense has the advantage. a11 the foregoing is reversed. The state that fears attack does not preempt-since that would be a wasteful use of its military resources-bu t rather prepares to receive an attack. Doing so does not decrease t11e secu rity of others, and several states can do it simultaneously: the situation will therefore be stable, and status-quo powers will be able to cooperate. When Herman Kahn argues that ultimatums "'are va5tly too dangerous to give because ... they are quite Hkely to touch off a pre-emptive strike.''5 he incorrectly assumes that it is always advantageous to strike first. More is involved than short-run dynamics. When the defense is dominant, wars are Hkely to become stalemates and can be won only at enormous cost. Relatively small and weak states can hold off larger and stronger ones, or can deter attack by raising the costs of conquest to an unacceptable t:vel.. State~ th~n approach equality in what they can do to each other. Like ~he .~-caliber ptSt~l m the American West, fortifications were the "great equalizer m some penods. Changes in the status quo are less frequent and cooperation is more <.-ommon wherever the security dilemma is thereby reduced. . . .. . Many of these arguments can be illustrated by ~e maJOr J>O'~ers polictes m the periods preceding the two world wars. Bismarck s w~ su~nsed st~tes~en by showing that the ofTense had the advantage, and by bemg qUJck, relaovel) cheap.

Beliefs abr11rt tire cou rv of a war 1n wh' h t h<:: off, uc: h c. rther deepen tt c 'tcurity dilem ma. Wh"n tl . ense as the advantage ru k 1 . . 1ere are rncent' . suCcessful attar '-\11 uc;ually !><> wcakt:n the cJth er SI'de that :trves to11 stnke first' a . ck blooJJc<,s ,lfld tlcc.:isive. 1t is in thel>e pe . ds h vrc ory \VJ be relatively qUl ' no w en con ttractive that statt 'I consolitlatc power intern 11 f, . c1uest IS possible and a .1 a y- or mstance h d . reudal baron.s-anu cxpantJ cxternaJJy TI e ' Y e~trOYJng thc '' . ' re are S(::Ver I decrease the chanc.:t of cooperation among status- uo staa co~.sequenccs that profitable for the winner. Th<.; <.:osts will be low and~he be~:sfi FJr~t, war ,..,;JI be losers vvill suffer; the fear of losing cou ld intluce st t t f,ts hrglt. Of course, . "' t I h. . a atiVe arrangemen s, )Ut t c temptat1on of es o.try to konn ~table cooper11





. .. . into a common error, observer ,}.('.: ('ted thL., a and qmte dectstve. Fallmg ti s had several e{f( '<: I trst stat p ttcrn into the future.6 The resulting explecta o~f the Fra nco-Pnt" . n '\ ctr. Napesls ought . t all'e s In the ear y stages . o senu-permanen 1 , of time to recrui t A11 : 1 to ,bis sid eon IJJ .. 1 had thought cllat th.ere wor~d~; sli:nJ~ista ke. Second. def~ns. budgets We~ ~~~w, others we re not gomg to P. ed h I) to mcreases on the ocller sic.le. lt h not surpnsmg thgb and rea ct qmte s arp f fits this period we U Third . mo!!t c.lecision m k at . Ri hards 's theory o anm races c on ar would not cost mucl1 blood and treasure; a ers thought that the next European ~v seen as inevitable and wit; mass opin ion Th .was at reason why war was gener J IS one med likelv there we re strong pressu res to preern so bellicose. Fourth, once war see ~ 'd th .d b lieved that whoeve r move firs t could pen etr ate the oth cl pt. . Bo SI es . e hili' ti' and cllus gain an insurmountable .eep (Th enough to disrupt mo za on e use of naval forces. AltJ1oug1 Ch urc I',J made an ere 1 w illwas no sueh be )jef about th . "d vised h sa . that if German slups o not come ou t , d figh t m time of an they~be d: ;u t like rats in a hole,~s everyo ne kne''~ that subm~rines, tnines and coastal forti6cations made this impossible. So at the start of tl1e w~ ~ac h na')' prepared to defend itself rather than attack, and the sho rt- ru ~ des tab Jhzm g forces did that Iaunched the annjes toward each otherh not operate.) Fu rthenn ore, each . . 'd knew that the other saw the situation t e same way, th us mc reasmg th e per~~ed danger that the other would attack, and giving each added reasons to precip itate a war if conditions seemed favorable. In the long and ~he .sho r: run , the re were thus both offensive and defens ive incentives to strike. ThlS Situation casts ~ght on the common question about German motives in 191 4: "D id Ge rm any unl eash the war deliberately to become a world power or did she sup por t Austria merely to defend a weakening ally," the reby protecting her own pos ition ?10 To som e exten~ this question is misleading. Because of the perceived advantage of the offense , war was seen as the best route both to gaining expansion and to avoiding drastic loss of influence. There seemed to be no way for Gennany me rely to ret ain and safe guard her existing position. Of course the war showed these beliefs to have been wrong on all poi nts. Trenches and machine guns gave the defense an overwh elming advant age. The fighting became deadlocked and produc ed horrendous cas ualties. It mad e no sense for the combatants to bleed themselves to death. If the y had kno wn the power of the defense beforehand, they would have rushed for the ir own trenches rather than for the enemy's territory. Each side could have done this ,,;thout increasing the other's incentives to strike. War might have bro ke n out any way; but at least the pressures of time and the fear of allowing the other to get the first blow would not have contributed to this en d And, had both sides known the cos ts of the ~ they~uld have negotiated much more seriously. Th e obviou s question is why the states did not seek a negotiated settlement as soon as the sha pe of the ,var became clear. Schlieffen had said that if his plan failed peace should be sought. 11 Tbe . 1~. ' answer IS compiCA, uncertam, and largely outside of the scope of our concern s But part of the reason was the hope and sometimes the expectation that break ~~-be made and the dominance of the offensive restored. Without ~at L-- - 'be political and psychological pressures to fight to a decisive victor)' 011ght lliiY e ~overcome.



The poli tiC'-. ';{ t h( in t<rwar period were sha ed b ous wn lli<.:t. nd till' belief that any Futtrrc.. P mem preVl . . "'war wou Id rY.the bl . ories. of the . ,,:tar'V lessons n:mforced each oth(;r in . ese m e 1t. Pohti<:al and rnw 1 . rating . . . cause it was b( hcwcl that the First Worldarne 1tOh c.l b the ~ecunty d1lemrn \ \T Be. a. 1I . ar a een a . t k h mts a e t at could I1ave bee n avoidc:c J~ skillf~tl conciliation both B . . ntam and t0 1 ..,.,..,ce were highly sensitive to the possihilin, th t . t a esser extent f ~~ 'J a m crwar Ger threat to peace. and ak: rt to the danger that reactin uickJ man y was not a real arm s could create unn ecessary conAic:t. And becauseg q. . Yand, strongly B to her . ti clom the de fense to con n11c to inate, they conc:ludecl nta1 . and f ranee exnected th tn . a 1t was safe to -r t ado more relaxe d. an cl nontI1reatenmg military posture.12 Bri tain als ~ I . P a ntam t'gl1t all't ance hot11. Th ' 11, , )j 1 0 mru e t .less need to ~. e hwes mi 'tarv post th ' sliuht danger to Gennany.; 1 the latter been content ure then constituted onh: a 1ad 1 'tl . o r b t 'cl Wl quo t have been easv ror otn s1 es to have felt secure behr'nd th1 e. status r :fi1 would err 1 of 10rt i cations mes Of cou rse the Ge rmans. were ~ot ~on ten t, so it is not surprising that thev dev oted their n~oney a~1d attention to fio~mg ways out of a defense-dominated ~tal emate. Blitzkn eg tactics were necessary rf they were to use fo rce to cL,. th t .. uan .e s quo The im tial stag~s of the war ~n the Western Front also contrasge withatus Firs. ted the t World War. Only w1th the new atr ann were the re any incentives to strike first , and these forces were too weak to carry out the grandiose plans that had bee n both dream ed and feared. Th e arm ies, stiJl the mai n instrument, rushed to def ensive positio ns. Perhaps the allies could have successfully attacked whjJe the Ger mans 13 were occupi ed in Poland. But belief in the defense was so great that this \vas never seri ously contemplated. Th ree months after the start of the war, the French Prime .\1iniste r summe d up the view held by almost everyone but Hitler: on the We stern Front the re is "de adJ ock. Two Forces of equal strength and the one that attacks seeing such enormous cas ualties that it cannot move with out endangering the con tinu14 ation of the wa r or of the afte nnath." The Allies were caught in a dilemm a they never full y recogn ized , let alone solved. On the one hand, they had very high war aim s; although unconditional surrender had not yet been ado pted, the British had decided from the start that the removal of Hitler was a necessa n conditi on for ' peace. IS On the other hand. the re were no realistic plans or instruments for allo wing the Allies to impose the ir will on the other side. The British Ch ief of the Imp erial General Sta ff noted, "The Fre nch have no inte ntion of carrying out an offe nsiv e for years, if at all"; the British we re only slightly bolder. 16 So the Allies looked to a long war that would we ar the Ge rmans down, cause civilian sufferin g thro ugh sho rtages. and eventually underm ine HitJer. The re was ~ttl e analysis to support this 'iew - and indeed it probably was not supportable but as long as the defense ~~ d~m inant and the number s on each side relatively equal, what else could the Allies do. To summarize, the security dile m;na was much less powerfuJ afte r World Wa r I than it had bee n before . In the late r period, the expected power of the ~efense allowed status-quo states to pursue compatible security policies and a,o ld races. Fur the rm ore, high tension and fear of war did not set off short-run d)n anucs by which each state, hyi ng to increase its security, inadve rtently acted to "_lake war more likely. The expect ed high cos ts of war, howe,er. led the Allies to be~ eve that no sane Ge rman leader would run the risks entailed in an attempt to donuna te the Continent, and discouraged the m from risking war themsel"es.

lR I

m _ns





Technology and Geography

vo main f~t<:IOr' Lltal <h ..,.. 1 1 an d geog'' 11')h)' are tl1e ,t' As Brodie not<-.. <)t 11Il ,.t'l lw wlrC'! Irt, tl I< ~ aec 1110 ogy . ,,., ica llf'vcl . .I . . . l\ o1icnse or the uetrensc Il as the advantage. kcr hut ntan" r cJ l.I1 cIl'fcHdt... 'f'l a tt a\ 1 r >rs . . ru Ic, rew p1 . al r t< fa vor I tC a a<.: -fk c:1 1amctcn...,tl( 1ll Im-. lro 1J ..J. I(l 1ystc< tac er 111 1 llll( r 1 11 I tl , advantage o1c.:ov I.,. detcncer usua y 1as IC , 1 crosses op(ll grotrt H. Anytl1ir r tl . 11 . f' I it . hilc Ius opponcn . I . ,., r.u some form o s 1e et ~ 1 t 1 attackN has to cros..,, m llliJ><< ts IJis progr s~ increases tl1e amoun t of arounc 1c .111 while crossing. illtn .t.'>('' l I1c adv .< ~"> e 1110re Vlt 1 nera 1111 .tgt Ium across it, or makes . r" s"par tcd IJ\ l>arricr<o tlal prod11<:c tl t . . d r \\11Cn states a " " ' " 11~~ accrurng to the ecnse. . . d . . dilemma 1 ease , sinee both can . Itaw IOJC'c ~ advc Ill <it. flJr S he secunty cffeets, t . k Jmpenetrahlt lnt rrl<'r~ "<>1<1 a<.:l11ally pnvc 1 dcfense without being abk to alttac 1 settle for a lfoocl U<'al 1<~~. Hufl(r zon<<; sic,'1 ' . 1. cl : makers 1ave o ~ . . 'v war; m reaJty. ecJSJOn 1 1y g,ve the defcntln tlltH' to prepan, i~t~;n"s. they t 1t>1<' ) "" < the attacker's progress; d tile ntlmher of soldiu:-. available ((Jr tiH fi . uce 11 '1 . . f I 'ti . . . 11 rcc1 problems o ogs cs, tll . t tll cent11 rv A1tlwr Balfo11r not ed Afghauishn'1 Jf ; ,, It. At tJ,e encl < t w nmc een assau . "$ 1 gas it possesses few road!,, and no raiIroads it will .. . d cting" < Jua!JtJes. o 011 1 ' non-con u . k effecti ve use of her great llll lllenca Slpt:riorit} tt 1 l . . ss'ble for HussLa to llll:l e I 11 f'f' r ' Je nnpo . , 1t0 tile Empire.'' The Hussians va uc< > crs or the sw1<' 1 yVJta point immed1ate1 . cJ 1 . I . any .. eac;ons it is not surpnsmg ll 1at when Persia was bcmg IVlt1cu tnto H11ss1an and r ., h {J British m eres orlllllllencc S()Jnt')'earslater' the Hussianssoug1Lassnrances thattllc 1 l 1 ilding potentia.lJy mcnac111g nu ., ' . s 1nr t 1e1r sphcre. British ..- uld ref'ralll frem1 bt 1 WO I roac . _ . ctkJn radica11 altered the a > rbes o conntncs to )' . Jncleeu, smce raJ1 cl c.:o11stn 1 roa . 1 . . .1 r oerencl 'ttl emse1 ancl to att-:tck otJ1ers' many dplomat1c notes an<Ill 11111<.: h 111[(11 ves c ' I. ence activity in tJ1c late nineteenth century centerc< on tlt tS su ">JCd. . 1 . g Oceans, large rivers, and mountain ranges serve .the same .f11nctJon <L'i bnffcr w nes. Being hard to cross, they allow defense aga1nsl supcr~~r numberl!. Tlte defender has merely to stay on his side of tile barrier a nu so cart utd!:;r.e aiJ th~ 111cn he C'dll bring up to it. Th<' attackers men, however, can c:ross o.n ly ~ l~w a.t a ti me, and they are very vulnerable when doing so. If all states were scll.-sufBcient 1 sland , anar chy would be much Jess of a problem. A small investment 1n shore defenses anJ a small army would be sufficient to repel invasion. Only vey wc<tk states would be ~ul nerahle, and only very large ones cou ld mena<.:e others. As notcu above, the Un1ted States, and to a lesser extent Great Britain, have partly been able to escape from the state of nature bet-ausc their geographical positions approxi1natcd this ideaL Althoug~ geography cannot be changed to conform to borders, borders can aud do change to conform to geography. Borders across wh ich an attlli:k is <.'<L'i)' tPnd to he unstable. States living within them are likely to expanu or be absorbed. Frw piC'nt war~ are almost inevitable since attacking will often seem the best way to protect what ouc has. This process wjJJ ~op, or at lea'it slow down, whe11 the state's borders rcac:h- hy expansion or contraction- a line of natural obstades. Security without attack will then be possible. Furthem10re, these lines oonstitute salient solutions to bargai ui ng pro~ lerns and, to the extent tJmt they are barriers to migration, are likely to <~vide ethuK groups, thereby rdising the cost~ and lowering the incentives for conquest. . 1 . Attachment to one's state and its land rei nforce one q uac;i-geograpluc<t aJd to the defense. Conquest usually becomes more difficult the deeper the

kcr pu!>IH's JlJ L CJII< turritory. aliow . ht r\ 1 attac . advatH 'lng ,o Iy 1 a 1~rn '>pure, th<; dcf ' on cngthtJI'I the attack<; . I . <:uucrs tCJ fi gl.t Jer har 11 nf:11 nili rs Ill<:\ : ' 1 CJftc" d<:vastatcu land th ~upp .y 1 1 takes him gh 1ut th rou . s at req 11 m.: trc>< ~ . . )ps or garnson T ht'SC stabr lr/ ng dyrr<U ni ~;s will not O])(:r te I duty a '' , t,.;ricl is sitnat ccl 111 ar its honkrs, or if the pconl I(}Wf.!V(: r' I r tJl C cJ t: ~en U I er's war rna c:: r. c c only about 1l l'lll!.; on I 11<.: win11it rg c;id<.:. In Sllc.: h ~ o not <:~trc aho11 t thClr state, .. t I )IJ r 'IJL 'Ill at work and illl I cI ell'als wr ne i11S11rrnountabh 111 ses, pnsJtiVC.: rccdback Wl. )e tra lrnitating gr<1graphy, trcn hav<.: triccl to create ba . r . . I . r r demilitam:N zonE-s on IJOllt s1ues of. th<: bordt:r altlrnE;! rs. r<::atl<:s rnay.proVI'dc h . 1 SIJ<: zones .....,,IJ rarelv JO roug I deep cnougl1 to proVl'd c rnore than warning. ' Even tl . . ." b e . 1 wa\ r.urope but the Hus~wns auoptcd a gauge: for thei r railro .11stl. t not poss1h!e 10 l d ~:. . . aus 1a was Jroa er than tl at of the nc1ghbonng 'itates, thereby complicating th' l< g . . . ll f I . . 1 , 1ICS pro ) C::ms O an)' "' > IS attacker- mcludmg H.uss1 a.

l &'l

Perhaps th e most ambitious. and at lt:asl temporarily suceessful attempts to construct a syste m that..wou ld a1d the defenses or both sides were the intcr.var naval treaties, as they affected JapancseArneric:an relations. As mentioned earlier the problem was that ~e United Stal(;S c.:ould not J efcnd tile Philippines ,vi thou~ denyi ng Japan the ab1hty to prote<.:t her home islands?>On 1941 this <.IHemma bec:ame insoluble when Japan sought to extend her contr()l to ~1 alava and the Dutch Ea'it Indies. the Phi~ippines ha~ bee~ invulnerable, they could have provided a secure hase from wh1ch the Un1ted States could interdict Japanese ship ping hetween the homelanu and the areac; she wao; trying to eonqucr.}1n the 1920s and early 1930s each side wou ld have beeH'vi lli ng to grant the other sec.:urity for its possessions in return for a reciprocal gran t, and tile Washington l':aval Conference agreements were designeu to approach this goal. As a Japanese diplomat later put it, their country's "fundamental principle" was to have "a strength insufficient for attack and adequate for defense."21 Thus Japan agreed in 1922 to accept a navy only three-fifths as large as that of the United States, and the United States agreed not to fortify its Pacific islands. 22 (Japan had earlier been for<:ed to agree not to fortify the islands she had taken from Germany in World War 1.) Japan's navy would not be large enough to defeat America's anywhere otJler than close to the home islands. Although the Japanese could stiU take the PhWppines, not only would they be unable to move farther, but they might be weakened enough hy their efforts to be vulnerable to counterattack. Japan , however, gained se~;urity. An AmeriC'an attack was rendered more difficult because the American bases were unprotected and because, until 1930, Japan was allowed unlimited numbers ~f cruise~, destroyers, and submarines that could weaken the Americ.:an A as t made ts eet way across the ocean.2.1 The other major determinant of the ofTensedefense balance ~ technology. When weapons are highly vulnerable, they must be employed. before they are ~t~acked. Others can remain quite iuvulnerable in their b~es. The fonner ehamcterJStics arc embodied in unprotected mis iles and many kinds of bombers. (It should be notcu that it is not vulnerability ]Jer e tllat is crucial but the location the vul11 ner-rtbtty. Bombers and missiles tl1at are easy to cles t r0 , onl, after ha...wg been . . launched toward their targets do not create destabWzing dynamics.) JncealntiVes tko stnk< f' are usually absent for naval forces tl1at are tl1reatened bv a nav attac lrst

H :





hen in the b pmt ,, ed _ the, are usually '''eU r . 1r ;l~h d d iles m harden wos, ~ Uke mtss lvl::!s suc:ce . -...;;s. be prepare to erenc I . 1 Both sides can then sJmultaneous) conditions, forts, trc - ' .J.nd small ssfuJiy. I n ground ' varfare under some off large numbers o r., r r~. Less frgroups of bold .. men in prepared.postt:JOns:endefenses. By and large. it 15 1 C test hetwe;;;~en~~ on the one hand, ;d ' 10hility a d hOrtifi. a few attad.-ers can stonn n ea,. rtinO' liO'ht weapons d . ler cations an suppo o o Ji th ttack on the other. As tJ,,. Erroneous . ea I the way or th s no simple waY to dPtcrmi ne whVIews heId l . weapons that c ear JC 1 IS d " 1 be two world wars show' ereooth and predictable like' those of . orn. before t a sw10gi iUations are not sm "[T]h . ;,... both extent and time. Some occur in t}te c:our nfg ese osc mant. u duJ urn. The}' are uneventhers in the course of a war. still others during ase 0. a . pen senes also be detected: single battle or camprugn .o . a of w rs.n Longer-tenn osctllat:ions can tl:le tv:el.fth to the late tllirteenth cen tu rv, 'vith its wond ..r f . e.,uj ". . I. h h . . . The early Goduc age, rom dunng w 1JC t e attackers m Europe gen. c . d 1 ces was a penod . , . . cathedrals an d 10rti 6e Pa 1e because t 1 rmpro\iement 10 the strength d. creasing difficulties, . . . fd erally met senous an m o estruction. Later. With the spread of e advance in the power c of fortresses outran tb fifteentl:t centurv. old tortresses 1 t1 power to resist An ost 1crr the end oftbe f ' firearms at . a bich the offense possessed, apart rom short-term setbacks 11..... ed d unnc v. ' "" all r. age ensu . g the seventeenth centmY. especr y atter about 1660 and . . _ : advantages. Then. dunn til least the outbreak of the War of the Austrian Success1on m l t40, the defense ~much of the ground it bad lost sinc-e the great medie,al fort~esses~had pro"ed tmable to meet the bombardment of the new and more numerous art:illery.2A . Another scholar has continued the argument: "The offensi,e gained an advan. tage with new forms of heavy mobile artillery in the nineteenth century, but the stalemate of World War I created the impression that the defense again had an advantage; the Gennan invasion in \Vorld \Var IL howe,er. indicated the offensive the field."'25 superiority of highly mechanized armies in The situation today with respect to con,entional weapons is unclear. until recently it was believed that tanks and tactical air power gave the attacker an advantage. The initial analyses of the 1973 Arab-Israeli war inclkated that new anti-tank and anti-aircraft weapons have restored the primacy of the defenst'. These weapons are cheap, easy to use, and can destroy a high proportion of the attacking vehicles and planes that are sighted. It the n would make sense for a status-quo power to buy lots of 820,000 missiles rather than buy a few half-million eUdollar fighter-bombers. Defense would be possible even against a large and w equipped force; states that care primarily about self-protection would not need to engage in anns races. But further examinations of the new technologies and the history of the October War cast doubt on these optimistic conclusions and leave us unable to render any firm judgment26 .Cooceming nuclear weapons, it is generally agreed that defense is impossible~ btntpb DOt ci tbe offense, but of deterrenre. Attack makes no sense, not beca~ it can be beaten off, but because the attacker will be destroyed in rum. ln terms under oonsideration here, the result is the equivalent of the ucteme. F"IJ'St. SJoity is relafu.ely ,.J.~ ...... Less than one percent of the G.~

l SS5

. ooceivahlc cm't .,. rr<:H'>. Second ' hc>th sd(;S <:an SliT JJt I C " ancously gain sccuritv in ;apability Tl tird d e form of sdor1d \tnkc < t11 , atcd t th c an re1 ~> e roregoing, ~cond.d in the r. . f Strike capahrlrh cm he .rrwintai ncd . h . ace o WJ e vanati .., ons rn t e othe:r sides militarY postun'. I w rc rs no purely milita r v b each s.I cle has l() reaet quickly ' reason ' I l ( 'f'> increases in anns. An w~ y(jj L ' and strongly tOtrr Ot t-rving to achieve fir'>l-strike <:apability can 1 y spen l' ng that the other devotes to . , )e neutraJ7..ed b th . ~;b Y e states spenclmg much smaller suml. on protecting its second ,_ .k capa ilitv F ourth, t here arc no -Sl n e . m .,. incentives to stri'kc first . a crisis. Important problems remain, of cou rse. Both sides h . f h ave mterests that go well beyond defense of the homeland The prate t' c 1on o t ese inte ts . . . . . Ricts even 1f ne1ther s1dc destres expansion. Furthermore the .res ereates cond ' .shift from defense to deterrence has greatly incrca'>ed the importanc e rity now rests on each side's belief that th th e an Pld rceptrons of resolve. Secuh'gl . ks f e o er wou prefer to run 1 1 ns o t A total destruction rather than sacrifice ils vital . t m eres s spects of th e secunty dilemma thus appear in a new form. Are weapons p . index rocurements ') . Ive.? Must t hey be so used. side fails to respond t th used as an 'Id th , b of reso o eo t r s Ill up ' k d h b - . 11 1 1t appear '"ea an t ere y mVJte predation? Can b th d . 1 w s1 o f 1 h 1 or is there a zero-sum elem tSI es 1 mu> taneously have 1rnages o 11g reso ve en mvo ved':. Although these problems are real, they are not as severe as those m th e prenuc1 era: ear . . There are m anv md1ces of resolve and states do not so much JUdge 1mages of " . r~solve L~ the ab~tr~c.:t as ask how like~y it is that the other will stand finn in a partic:ular dispute. Smce states are most likelv to stand finn on matte rs wb.1<:h concern , them mos~, it i~ quite possible for both to demonstrate their resolve to rote<:t their p own secunty Simultaneously.

The other major variable that affects how strongly the security dilemma operates is whether weapons and policies that protect the state also provide the capability for attack. If they do not, the basic postulate of the security dilemma no longer applies. A state can increase its own security without decreasing that of others. The advantage of the deffmse can only ameliorate the security dilemma. A differentiation between offensive and defensive stances comes close to abolishing it. Such d.ifferentiation does not mean, however, that all security problems will be aboushed. If the offense has the advantage. conquest and aggression will still be possible. And if ~e offense's advantage is great enough, status-quo powers may find it too expenstve to protect themselves by defensive forces and decide to procure offensive weapons eYen though this will menac-e others. Furthennore. states "ill stiiJ ha,e to wony that even if the other's military posture shO\vs that it is pea<.-eful now, it may develop aggressive intentions in the future. Assuming that the defense is at least as potent as the offense, the differentiation betv.'een them allows status-quo states to behave in ways that are dearly different from those of aggressors. Three beneficial consequen<.-es follow. Fi~ status~uo pow~rs can identify each other, thus laying the foundatior~ for ~r~t:ion.. Confh<:ts growmg out of the mistaken belief that the other side lS expansiOnist will be less



cleQeJ to


>~ attack on the United States most o rtdtmdaot systems to provide a lot of insuran~ against the \\'0

.J:_~ a uuea:


t f 1 is .-at







_, n obtain advance \\'Jrtl ' lt '~ 11 other 1 t. second status-quo states ww 1 rrequen k . ._ . to dc,elop and deplm nl n~ ivc w ,. ' p an . B c . tate can attac , 1t oas <: ap<Jn\ aR.gresston. eL oreas t be disguised and le:.l~ <>. . . a1110 tl1ese weapons C'.lnno lllll of 1f procure men t o f tate vvi 11 have the tJtllr to take . . . aJ 1 does a status-quo s t:o11 nter. time, as 1t most a ways : . ._1 level of defensive arm~ <l!\ Ln!!; a! its l)Olc . 11 . I cl t 1nai.ntam a 1 11gL . 1 1a1 mea'iures. t nee no ful posture (Although be~ng so anncd shmdd aclversanes are . dopting a peace t d below aJarm other -.tat ns-r[tiO po 11o1 a . .J h rt t ex.ceptton no e wcrs ) w1t 1 t e one tmpo an . t'on "o tctions thal thcv l>cliC'\'(' would not he lak d . t y srJecta1atten 1 l ' < ' < ('ll States o, m ac pa b t:he, 14 that state e~hibitiug S\Jd l beha,ior .,. reel ' u by a status-quo 5tate .ecause de~elopmeot of transportation facilities will \Jar e sive Thus tl1e seJZure or . .r 1 . 1 n aggres if tl r . . . ha~e no commerc1al ,aluc, anu t 1e1 efore c;an onlv be 1ese .acJ 10 es J d . 0 th rs more e il' . In I906 tJ1e British rejectc a RussLan protest about wanted for m ltai)' reasons. ' . . ~ . . . . . . cl' t t of Persia b) clannmg that t1us are a "as on 1 , of lstrategic] ~ the! r actiVities m a IS ne . . , . [ h R . ansl if the)' WJsJ1ed to attack t I1e I nd'. r . . or to put u.ssJ d 1mnortance to t e k 1an..,-rontJer, .. c-I aking us think tl1at they in ten to attac 1t. - ressure upon us JY m< P The same inferences are drawn when a state acqw.res more weapons than c eded 140 r defense Thus. the Japanese spokesman at the 1930 c observers ree1are ne London naval conference said tl1at his country was alarmed. by _the Amencc:n refusal to give Japan a 70 percent ratio (in place of a 60 ~ercent rat:i~) m heavy <:rutSers: "A s long as America held tl1at te~ percent ad~IQ11tage, 1t was poss1 for her t~ attack. So bl~ when America insisted on stXt)' percent mstead of seventy pe1cent, the 1dea would exist tltat iliey were trying to keep that possibility, and the J ~pa~ese people could not a<.:cept that."'2.1! Similarly, when Mussolini tol~ Cham_berlru~ m Janua~y 193~ _that Hitler's arms program was motivated by defens1 consJderahons, the Prune M lllJSter ve repJjed that ..German military forces were now so strong as to make it impossible for any Power or combination of Powers to attack her successfully. She could not want any furilier armaments for defensive purposes; what then did she want tbem for?''29 Of course these inferences can be wrong-as they are especially likely to be because states underestimate the degree to which they menace otbers. 30 And when they are wrong, the security dilemma is deepened. Because the state thinks it has received notice that the other is aggressive, its ov.rn arms building vvill be less restrained and the chanc.:es of cooperation wiLl be decreased. But the dangers of inc:orrect inferences should not obscure the main point: \Vhen offensive and defensive postures are different, much of the uncertainty about the other's intentions that contributes to the security dilemma is removed. The third beneficial consequence of a difference bel:\veen offensive and defensive weapons is that if aiJ states support tl1e status quo, an obvious arms con~~ agreemen~ ~ a on weapons that are useful for attacking. As P:esid~n~ oosevelt put lt m h1s message to the Geneva Disarmamen t Conferenee 111 1933 "If_all nations will agree wholly to eliminate from possession and use the weapons which make posSJ'bJe a succ:essm1 attack, defenses automatically will becorne . .r . . Impregnable, and the frontiers and independence of every nation will become secure "31 The f: th h ~ act at sue treaties have been rare the \Vashington navaJ agree mentsthat sed above and the anti-AB M treaty can be cited as examples-sh o": either states are not ai willin t1 at 11 is hard to distin . h ofli ":ays g to guarantee the security of others, or 1 gws enslVe from defensive weapons.



Is such a cLshd lon possible? Salvador de \1 a . . ve in the disan: .1:nent negotiations of the inte dariaga, the SpaniSh statesman l:l ac . lwar years, . 1 c . according to which d f. thought not: "A weapon . tJter ofTenSJ\ ~ nr c erensJVe 1s e1 . . en o 1 you a 1 ki t french Foreign \l m1ster agreed (although French olic di re oo ng at." The . w) "Every an n can be employed offensive!)' dp . Y_ d not always follow this 0 r e1 v1e I ensJ . y to discover whet 1er arms are intended for pu 1 d r vel)' in t urn. The only . ,va . . . re Y eensJVe pu h . a spirit of aggre~swn 1s m all cases to enq,;re . t th . rposes or are eld tn ...... m o e mtentio f th ncerned.'' Some e"idence for tl1e validity of this ar . n~ o e country CO . . th gument 1 provtded bv th f: s that much time ~~ ese unsuccessful negotiations was de,oted t . :' e act 0 ve from defens1ve weapons. Indeed no simple and b' separating offenSl . . ' unam 1guous d fi ti . r e_ ru on IS possible and m many cases no judgment can be reached. Be1ore tl1e Amencan entry ~ d . intO World \Var I, \\ oo row \.VJison wanted to arm merch t . . th 1v w1 guns in an men on th the back of the sh1p so ey could not initiate a fight but thi d: c ' s expe 1ent cannot be applied to more corn mon rom1s of armaments.32 There are. several problems. Even when a differentiat0 n 'bl 1 IS poss1 e, a status. quo power will want offens1ve arms under any of th ree conditions: (l) If the offense a great adv~ntage over the defense, protection through defensive forc~s WJll_be too e~ens1ve. (2? Status-quo states may need offensive weapons to regam temt~ry lost m the open mg stages of war. It might be possible, hO\vever, for a state to wru.t to procure these weapons until war seems likelv, and thev might be needed only in relatively small numbers, unless the aggressor ~vas able t~ construct strong defenses quickly in the occupied areas. (3) The state may feel that it must be prepared to take the offensive either because the other side wilJ make peace onlv if it loses territory or because the state has commitments to attack if the other makes war on a third party. As noted above, status-quo states with extensive commitments are often forced to behave like aggressors. Even when they lack such commitments, status-quo states must worry about the possibility that if the~' are able to hold off an attack, they will still not be able to end the war unless they move into the other's territory to damage its military forces and inflict pain. }.lany American naval officers after the Civil War, for example. believed that "only by destroying the commerce of the opponent could the United States bring him to tenns."33 A further complication is introduced by the fact that aggressors as well as status-quo powers require defensive forces as a prelude to acquiring offensive ones, to protect one frontier while attacking another, or for insurance in case the war goes badly. Criminals as well as policemen can use buUetproof vests. Hitler as well as Maginot built a line of forts. Indeed, Churchill reports that in 1936 the German Foreign Minister said: "As soon as our fortifications are constructed [on our western borders] and the countries in Central Europe realize that France cannot enter German territory, all these countries wi ll begin to feel very differently about their foreign policies, and a new constellation will deveJop.":34 So a state may not necessarily be reassured if its neighbor constructs strong defenses. . More central difficulties are created by the fact that whether _a weapon lS offensive or defensive often depends on the particular s~tuation-~or ~nstance, th~ geogr-aphical setting and the way in which the weapon lS used. !an~. ~ headed the fateful Gem1an thrust through the Ardennes in l940. but tfthe Fn:dned had cltsposed of a properly concentrated annored reserve ' it would have prOVl








. . n and into .1 the best means for thetr cutting 0 ff the penetratiolmill'' ,t . . -~~ \-; <. di vhe sastcr became instead an oven M : An r for the Ge rmans w Ita t . b sed they nm sl \, ..u. {iJ r lhe tt-ai rcrat tl eu 0 1er sid bvtou s1 defens1ve.-to ttack on ' Israel in weapons seem o Y 1973 '<m id hav t: to come to the m But the Egyptian a e bee . . . . ff tive air defenses that covere d t Ile I,,;tt Ie f' Id. Neverth 'n w unpossJble Wlthout e ec . . I s J h Simon, then the British Fop ' . C..J : 1f n less , som e ws ti neti 0 ns are posstb e.. d ,0 I.ier e1 Se(' . . stated tha t jus t because a f'gn . . to the vtews C eal , ite < retary, m response "th . reason for saying that the re '' Pr c not strelneeltne tch f could not be drawn , at was no . . d , . ry . er cl h. I allr)racttcal me n an v. omen kll< .v. , to 1 we ll on s 0 . )e ter nto on eith SI e w tc 1 thts .cl f h . we or th at s1 e o t e 1me.,. Altllough tJ1ere are alm ost no 1 apon~ and stra tegies that c kin tJ1ere are som e that are a 111ost cxc 1ustve 1 defensi y are useful only ror attac g. . . . ve. for protection but a state tI1at re 1ecI mostly on the 1 Aggressors couId want t11e111 ' . cou ld not.menace otll ers. More frequently. we c~nnot "cl[etedn]nd t1 abs m u_le le olute sc:o characte r of a weapon , but [we can ] make a compan son .. . an. 1 . . ver wh ethe r . t1 s or not the ofc 1ve potentialities predommate, whe 1e1 a weapon 1 mo re useful rens . in attack or in defens e. "36 . . The esse nce of defense is keeping dte oth er s1de out of you r ter ntory. A pur ely defensive weapon is one that can do this '<vithout ~eing ~bl e to p~n etr ate the enemy's land. Thus a committee of miUtary experts m an_,_nterwar d1sarma ment conference declared tJ1at annaments "incapable of mob~,1ty by me ans of selfcontained power," or movable only afte r l?ng delay, we re only capable of being 3 used for the delimse of a State's territmy." ' The most obvious examples are foJtifications. TI1ey can shelter attacking forces, esp ecially wh en they are bui lt righ t along the frontier,38 but t11ey cannot occupy enemy territ01y. A state with only a strong line of forts, fixed guns, and a smaJI arm y to man the m wo uld not be mu ch of a me nace. Anything else that can serve only as a barrier against attacking troo ps is similarly defensive. In this category are systems that provide warni ng of an attack, the Russian's adoption of a different railroad gauge, and nuclea r land min es that can seal off invasion routes. If total immobility clearly defines a system that is defensive only, Umited mobility is unfortunately ambiguous. As noted above, short-range fighter aircraft and anti-aircraft missiles can be used to cover an attack. And, unlike for ts, the y can advance with the troops. Still, their inability to reach dee p into enemy ten itory does make the m more useful for the defense than for the offens e. Thus, the United State.s and Is~ael would have been more alarmed in the early 1970s had dle Russians prOVIded the Egyptians with long-range instead of sho rt-range aircraft. Naval forces are particularly difficult to classifY in these terms , but those that are very short-legged can be used only for coastal defense. Any forces that for various reasons fight well only wh en on the ir own soil in eff~t ~ck mo~ility and therefore are defensive. The most extrem e example would ~ passtve resiStance. Noncooperation can thwart an aggressor bu t it is veJy hard or large num~rs of people to cross the border and stage a sit- i~ on anothe r' s territory. Morocc:os recent march th S . h success de ded . ~n e parus Sahara approached this tac tic, butitS . pen on spe stan to the extent to which itcial crrcum.vili ces. Similarly guerrilla warfare is defenstve .an ' og on} in o . re~wre.s Cl Support that is like ly to be forthcoJlll. , y pposttion to a foretgn mvasion. Indeed , if gue rrilla warfare were east!)


able aud if il touk t cn defender!, to destro . l . export' I . J eac 1 gtternlla th l Id not onh lw one w ll< h could be used a<> eas1 t k en th' weapon wou Y o attac the ot1 JS . 1 ren de'' d one's 0\\1 1 .n t one Ln which the offense had the advan ler s terntory as to tag t1 dilemm ct would op<.-> rat e l'Specially strongly. ' e: so 1e secu rity lf gue rrillas aw unable to fight on foreign soil, other kinds . illing to do so. army imbued with lhe idea h t I d of a_rmles may be \n t W un r a on ust wou ld figiJ t }('ss <'1 tctively, if at aH, if the goal' were cony uefens 1v e . ~wars ~~~re J . , Jack both the ability and the will for aggress ion Th I est. Citizen m1l1 tias met) . . L pon ort term of sem ce, tue tim e requi red for mobilization e weah s .employed th cl . , e sI1 I :.I d . Ill d I ,t t .ttacks on t11 e 1om e an . a en t 1emselves mu ch mo re an "'r e spmt of repellin(l o a 't '39 o <.1 se th an to attacks e'en 0 0 foreign ter n Ol)'." Les s ide alistic motives can produce the sam e result A l di edieval warfare J1as d cscn'bed the armies of that period as r ea n(J.. student of m U b 1 d' 1 0 ows: Assem led with difficul ty, t.nsu )Or mate, un able to maneuver ready to l""elt . fr . h . ' " awa) om 1ts standard t}1e mo ment tl1at tts s .ort pen.od of service was over a re dal r . . , 1 u 1 e presented ' 0rc an asse mb_lage of unsolclie rltk e_q ualittes su<: h as have seldom been kno wn to coexist. Primarily mten ded to defend Its 0\V ll borders from t11e Maovar the : orth . . . bJ , man the Saracen .. . , tI1e mstitut10n was utte rly unadap ted to take the offensiv , or e."40 Some political gro upings can l?e sim ilarly described. International coalitio ns are more rea~iJy held toge~ er by fea~ than by hope of gain. Thus Castlereagh was not being entirely sel f-se f\.~n g '~h ~n m 1816 he argued that the Quadrupl e Alli ance "could only have owed 1ts ongm to a sense of common danger: in its \'erv natu re it must be conservative; it cannot threaten eith er the sec:UJity or the liberti~s of other 41 States." It is no accident that most of the major campaigns of expansion hav e been waged by one domina nt nation (for example. Napoleon's France and Hitler's Germany), and that coalitions among relati,e equals are usually found def ending the status quo. Most gains from conquest are too unc:ertain and raise too many questions of future squabbles among the victors to hold an alliance togethe r for long. Although defensive coalitions are by no means easy to maintain-confli ctinO' national obj ecti ves and the fre e-rider proble m partly explain why three of them dissolved before Napoleon was defeated- the common inte rest of seeing that no state dominates provid es a strong incentive for solidarity. Weapons that are particularly effe ctive in reducing fortifications and barrier s are of great val ue to the offe nse . This is not to deny that a defensive pow er will wan t some of those weapons if the other side has the m: Brodie is certainly correct to argu e that while the ir tanks allowed the Ge rmans to conque r France, pro perly used Fre nch tanks could have halted the attack. But France would oot have needed these weapo ns if Germany had not acquired the m, wh ereas e~'en if France had no tanks, Ge rm any couJd not have foregone the m since they ~ron~ ed ~~ onl~ chance of breaking through the French lines. Mobile heavy ~llery 1_, sun s ~arl)~ especially use ful in des troying fortifications. The defender. whilt> needing ~le~ to fight off attacking troops or to counte rattack, can usually use lighter gw lS smce they do not nee d to penetrate such massive obstacles. So it is not surpris ing that one of tl1e rew th.mgs c that most nati. s at the terwar dis.......mament conferences On m . were au e to 1.1 ~ agree on was tha t heavy tanks an d mo bile 1ea' )o s were particularlv 1 . valuable to a sta te planning an attack. 42







. I . de end for their cffpcti\('nc>~" on surpr \\'<.'a pons and stTatewes t ,,at p . . ed bv ~omt ol t, ' d I . . . se are . rr . . . . <.gates ] I ays oenstve. That fact was. recoglllz e principle lwlqr u! l th . to tl1 ~ amostaw d 15 th 1 t conferences ctn C' <.:ornn mte rwar cJsannamen A . lier rC'fJresc 'ntuli\C o! tlab 1 1 11 . Il I I weapons. n eat . wcespre nattona >an on concea et , PI 1il delJ>hia newspapt t ltat anr11 d .. <ltI 1 1 teenth-centur; a \le'-'\' was t 1e mre -nme . d d canes arc entire h ''"clc~s n e . As .1 Th ' f d [i k es d1rks an swoJ ey are fi measure o e cnsed, alnll' I . tta~ks are of murderous chan.wll 1. \\'hoe\cr <:a ... t 1 only for attack. an sue 1 a . . . 1 1:J ltes , . I . ared ltimself for bolmctc e. . SU h a weolp011 laS p!ep C . '} ] t0 ffistirwuis h h etWCCll forces that . It is of course not always P0551 ) e t::o 1 for! itare most ' ' 't nd forces optimally des1 gne<. Sll<:h c ct've f'or holding tern ory . am{ weapons" a e f e 1 , d fo in E distinction could uot have been ma e 1 the strateuw urope . d b tween d1e Franco-P russ1an \\ ar and World Wa 1 he peno e durino- most oft d'l 1 f 1 r . o r . t ~;cal air forces ca n be rea 1y <: ass ll'( rn these tenn 5 Ne1ther nava1 torces 1101 acu . . t 1 1 . such a distinction 1 poss1) e. t re ccn t ral characte 1. S But the pomt 1ere IS tl1 t ",lten 1 . a _ . . f th 'lemma no longer holds, and one of the mos t troublesome tstic o e secun'ty w consequences of anarchy is removed.


Offense-Defense Differentiation and Strategic Nuclear Weapons

In the interwar period, most statesmen held the reasonable position that '-'Veapons that threatened ch~lians were offensive.4-l But when neither side can protect its civilians, a counter-city posture is defensive becaus~ the state ~ credibly threaten to retaliate only in response to an attack on itself or 1 closest ~es . The costs of this ts strike are so high that the state could not threaten to use it fo r the less-than -vital interest of compelling the other to abandon an establish ed position. In the context of deterrence, offensive weapons are those that provide defense. In the now familiar reversal of common sense, the state that could take its population out of hostage, either by active or passive defense or by destroyi ng tJ1e other's strategic weapons on the ground. would be able to alter the status quo. The desire to prevent such a situation was one of the rationales for the anti-A BM agreeme nts; it explains why some arms controllers opposed buikung ABMs to protect cities, but favored sites that covered ICBM fields. SimiJarly, many analysts wanted to li mit warhead acc:uracy and favored multiple re-entl) vehicles (MRVs), but opposed multiple independently targetable re-entry vehicles (MIRVs). The former are more useful than single warheads for penetrating city defenses, and e nsure that tl1e state has a seco~d-strike_ capability. MIRVs enhance countcrforc:e capabiiHies. . . . . What IS most Important for the argumen t here is that land-based ICBMs are both offensive and defensive, but when both sides rely on PoJmis-type systems (SLBMs), offense and defense use different weapons. ICBM s can be used either to de~troy the other's cities in retaliation or to initiate hostilities by attacking the others . oc and stratecri missiles. some measure s-for instance, hardenin g of mjssile sites . wamOthmg systems-are purely defensive, since they do not rnake a first strike easier. ers are predo . I 0 f"~' . . . defenses and hi mmant Y rei1Slve-for instance, passive or act1ve cit\' , both ' ghly a<.'CUrate warheads. But ICBMs themselves are useftU ~or purposes. And because states seek h' a 1gb level of insuranc e, the desJre lor

tection tL'i \H: " t H ' <:on tcmplati<m of a cou t c pr jsition of cxt :- '' ,(l>I 'k arge numbers of miss, le nSenorce stJ' e can the . . acqU o very 1 er's intention s f10111 . mil ita1 postu re. Eachs. .d 1t IS .. cHfi cult to infer the 1Ls y , otl1 . rin.' by proc.:nn n).!; ll lOH' missiles decreases t SI es efforts to . mcrease its own seCu ; extent ve efficacy of lh<: off(use.;: and the dcfense .tho an h . detcrmmed by the reiab . e s1de's . case when both o;tdes use SLBMs. The point .ot er th secunty. T hat is not the l I IS at less vulnerabl e t tan and~base d ones (this bears on not of~ sea~based syste~s are the tl at SLBMs are dcfc-ns1ve, retaliatory weapons SL ense defense ratto) but . \ trument of attack aguinst other SLBMs Th h. d BMs are not the main JO k . .. . . e ar est problem c f . tate that wants to ta e tls <:1tles out of hostage is to 1 .. t h on ranting a s oca that requires not S LBMs but anti-submarine weapon e t e other's SLBM s, a JO b A . to attack the other's submarin es (although othtr weapos. . statl dmtghbt use S LB~s e 1 . pro efficient), but w1t 10ut antJ-submanne warfare (ASW) .ns .wou th ably be more .1. . . capa1 1ty e task cannot be )I Jerformed. A status-quo state that wanted to forecro ofc I b rens1 capabJ1 could ve 1 ty ply forego ASW r~search and procurement. . . . stmWhen both stdes rely.on ICBMs' one side's missiles ea tt k th oth . n a ac e er's and so the state cannot be mdifferent to the other's building prog am B t b r u ecause one ' side's SLBMs d o not menace the other's, each side can build as manv as it wants and the other nee_d not respond. Each side's decision on the size' of its fo rce depends on techmcal questions, its judgment about how much destruction is eno~gh to. deter, a~d the amount of ins_urance it is willing to pay for- and these cons1deratwns are mdependent of the s1 of the other's strategic force. Thus the ze crucial nexus in the arms race is severed....



The two variables we have been discussing-wheth er the offense or the defense has the advantage, and whether offensive postures can be distinguished from defensive ones-ca n be combine d to yield four possible worlds. The first world is the worst for status-quo states. These is no way to get secuJity without menacin g others, and security through defense is terribly difficult to obtain. Because offensive and defensive postures are tJ1e same, status-quo states acquire the same kind of arms that are sought by aggressors. And because the ?ffense has the advantage over the defense, attacking is the best route to protectmg what you have; status-quo states will therefore behave like aggressors. The situation will be unstable . Arms races are likely. Incentives to strike first will tu m crises into wars. Decisive victories and c'O nquests will be common. States will grow and shrink rapidly, and it will be hard for any state to maintain its size and influence Without trying to increase them. Cooperation among status-quo powers "ill be extremely hard to achieve. There are no cases that totally fit this picture, but it bears more than a passi~g resemblance to Europe before World \~Tar I. Blitain and Gennany. although m t~any respects natural allies, ended up as enemies. Of course much of the explanation lies in Germany's ill-chosen policy. And from the perspective of o~r tlleo~' the powers' ability to avoid war in a series of earlier crises cannot be easily ex"})lained.




1 ute u


'" J T


acv .. : .:sge ---------~m~e:a~d~v~a~nt~a~ge:_______________ ---- ---- -2


T ~AB~LE~t~ _-----:~~;hc;;------- D....- has the ! Offense haS

Doubly dangerous

Offensive posture not distinguishable from defensi ve one Offensive posture distinguishable from defensive one

3 No security dilemma, but aggression possible Status-quo states can follow different policy than aggressors warning given

Sec; t 1 dilemma. but ::l ' l'Y requirements May he compa tible 4 Doubly stable

much of the behavior in this period was the produ ct of technology Neve~ eles~, t ma tined the security dilemma. Decision make ~~ th?ught that the and belihfs la d tage and saw little difference between oHensJVe and defene dt b. gi offense a a tg a van . . . ili" tur The era was charactenzed by arms races. An cl once war s1 m tar)' pos es. ve bilization races created powe rfu1mcen bves to stnke fi rst. seemed likeIy, mo t 1: C!-st world wou Jd be one m wmc1 eac1 SI e reue cl on 1 1 cl In th e nucIear era, the m . vulnerable weapons that were aimed at si1~ilar forces and each s1d~ underst?od the situation. In thjs case, the incentives to strike first would be very h1gh- so htgh that status-quo powers as weUas aggressors would be sorely temp ted to preempt. A~d since the forces could be used to change the status quo as well as to preserve tt, there would be no way for both sides to increase their security simultaneously. Now the familiar logic of deterrence leads both sides to see the dange rs in this world. Indeed, the new understanding of this situation was one reaso n why vulnerable bombers and missiles were replaced. Ironically, the 1950s would have been more hazardous if the decision makers had been aware of the dangers of their pos ture and had therefore felt greater pressure to strike first. In the second world, the security dilemma operates because offensive and defensive postures cannot be distinguished; but it does not opera te as strongly as in the first world because the defense has the advantage, and so an increment in one side's strength increases its security more than it decreases the other's. So, if both sides have reasonable subjective security requirements, are of roughly equal power, and the variables discussed earlier are favorable, it is quite likely that status-quo ~tes can adopt oompatible security policies. Altho ugh a state will not be a?Ie to JUdge ~e ~er:s in~entions from the kinds of weapons it procures, the level of spending will gwe Important midence. Of course a state that seeks a high ]e\'el 0 anns ~t be not an aggressor but merely an insecure state, \vhich if conciliated''~ redua: Its anns, and if confronted will reply in kind. To assume that the apparentl} ~ level of arms indicates aggressiveness could there fore lead to a respo~ would~ the dilemma and create needless conflict. But empathy and skillf ful statesmanship can red this dan uce ger. Furthermore, the advan tageous poslrton o


.lefense mu n f t!' t a status-r1uo state ca f' . . n o ten mamt . h' ty with a }{ c: ! <t'"l n!. lower than that of ts am a 1gh degree of secUn I expe<.tec.l ad demonstrates tha t ,r k~ the ability or desire to alter the s versary. Such a state esent tirne. Tl w :nr 1 ~~~th of the defense also all tatus quo, at least at the p1 r ows states to tr"int when ther ~Par that others are tne;na,.- th react s1 1 anc.l with res .., -:mg em S0 a1 l ow y . ers wiJI to som' ext(n t he threatenjng to othe th t tough status-quo p<>W . This world is tlw one that comes closest to rs, t at.extent \VJube 1lmttecl. 1 ma C1lng most ) . ds. h Attacking is usuall y lic.m.ler than defending be<.:ause of the str I eno ~~ ist_ory. and obsta des. Bll t purdy defensive postures are rare! oss~~ gth of fortifications ons are usuall y supplemen teu by armies and mob"! YP h~ because fortificatt w ck In the nu<.:1 r era, this world would be oneI .e guns h b1ch can support an ea . . atta m w1 c ot1 d 11 1 1 1 relatively invuln erable JCB Ms and believec.l that't cJ SI es re tee on 1 . 1 war was irnpo \ AJR e sible. Assummg no . v1 Vs, . would take more than nuc ear k tt . . . s, p one attac mg mlSstle to destroy one o r th eaclversarys. reemption is therefore unattr t Ifb h "cl . . I " . have large mventones , t 1ey can tgnore all but drastic increas. ac tve. th ot . SI es th . . . . es on eo er stde. A world of e1ther ICBM s or SLBM s m wh1ch both sides adopted th . 1 ofl1 . nuclear war wou Id P'o bably fi t thi s catego rv too The me e po ICY . m1ted m / ans of preservmg the status quo would also be the means of changing it, as we discussed earlier. And the defense usuaJ iy would have the advantage, because compellence is more difficult than deterrence. Altho ugh a stat~ might succeed in changing the status quo on issues that ma~e r much more to_ It than to others, status-quo powers could deter major provocations unde r most ctrcumstances. In the third world there may be no security dilemma, but there are security problems. Because states can procure defensive systems that do not threaten others, ~e .dilem~a need not operate. But because the offense has the advantage, aggressiOn IS posstble, and perhaps easy. If the offense has less of an advantage, stabtlity and cooperation are likely because the status-quo states will procu re defensive forces. They need not react to others who are similarly armed, but can wait for the warning they would receive if others started to deploy offensive weapons. But each state will have to watch the others carefully, and there is room for false suspicions. The costuness of the defense and the aJiure of the offense can lead to unnecessary mistrust, hostility, and war, unless some of the variables discussed earlier are operating to restrain defection. A hypothetical nuclear world that would fit this description would be one in which both sides relied on SLBM s, but in which ASW techniques were very effective. Offense and defe nse would be different, but the former would have the advantage. This sit uation is not ukely to occur: but if it did. a status-quo state could !ihow its lack of desire to exploit the other by refrai ning from threatening its submarin es. The desire to have more protecting you than merely the other side's fear of retaliation is a strong one, however, and a state that knows that it would not expan d even if its cities were safe is likelv to believe that the other would not feel threatened by its ASW program. It is e~y to see how such a world could become umtable, and how spirals of tensions and conflict <.i>uld develop. . The fourth world is doubly safe. The differentiation between ofTensJve and defensive !i)'Stems permits a way out of the secu rity dilemma; the advantage 0~ the defense disposes of the probl ems discussed in the previous paragraphs. There 15 no







. to be tempte <.l to procur e <.l'rl'li-.iw fcm.:cs reasun for a status-quo po~e~ . ns b ' the posture the' .l'" pt. Jndeetl .: .and au~ssors lTive notice of the1r mtenno ) '. , 11tl1e ::-o o d f, . . t enough. there are no secunl) proLie rns. The 1 ad,antage of the e cnse IS gJea 1 l ll ' f h to alter the status quo" 011 < L ow greater 5 oss of the ultima.te form o t]j~ powe:ans and probab ly wouJclt eud to Freeze the d~otpe for the exerc1 of nonrm tary me se IS n. bution of values. . Id Id I existed in the fi rst decade of tl I< t\' t te tl1 century 1 "en r Th1s wor wou 1ave ail bl 1 1 r h .. k I. the dectston ma ers 1ad un derstood the av a e . , b Oil:). tn t at case' the . 1. r0 Jlowed differe nt po1 es ot 1 111 t1 long run .,, cl tc1 1e European powers wou Id have 11 1 f' 1 . . . tl f Even Germany, facin g power u cnctm es on both side Ill le summe r 0 1914. d r F' s, could have made herself secure by developing strong e emcs. ra~ce could also have made her frontier almost impregnable. Furthe rmore, when cnses arose, no ncentives to strike first. . There would have been no comr)eti one wou Id have J1ad 1 . . . . mo biJ' t' tive 1za ton races reductn g the time available for negot1at10ns. . . In the nuclear era, this world would be one m whiCh ~h~ supelp owers relied on SLBM s, ASW technology was not up to its task, an.d limited nuclea r options were not taken seriously. ... Because the problem of v10le~cc below the nuclear threshold would remain , on issues otJ1er than defens e of the homeland, there would stil1 be security dilem mas and security problems. But the world would nevertheless be safer than it has usuaJly been.

1. Thus, when Wolfers argues that a status-quo state that settles for rough equality of power with its adversary, rather than seeking preponderance, may be able to convince

the other to reciprocate by showing that it wants on ly to protect itself, not menace the other, he assumes tbat the defense has an advantage. See Amold Wolfers, Discord a11d Collaboration (Baltimore: Johns Hopkins Press, 1962), p. 126. 2. Thomas Schelling, The Strategy of Conflict (New Y ork: Oxford University Press, 1963), chap. 9. 3. Quoted in Fritz Fischer, War of Illusions (New Y ork: Norton , 1975), pp. 377, 461. 4. George Quester, Offense and Def nse in the Intemational System (New Y e ork: John Wi.ley, 1977), p. 105. 5. Herman Kahn, On Themumuclear War (Princeton, N .J.: Princeton University Press, 1960), p. 211 (also see p. 144). 6. For a general discussion of such mistaken learning from the past, see Jervis, Perception

and Misperception in Intenllltional Relations (Princeton, N.J.: Princeton University Press, 1976), chap. 6. The important and still not comple tely unders tood question of why ~his belief formed and was maintained throughout tl1e war is examined in Bemard ~r~c, War and Politics_ (New York: Macmillan, 1973), pp. 262-70; Brodie, "Techno<>glcal Change, Strateg~c Doctrine and Political Outcom es " in Klaus Knorr. ed.. Hiftorical Dime io of li.T z ' . ' .ty Kans n.y ns nationa Securtty Problems (Lawrence: Univers1 Press 0 f as.' 1976), PP 290-92; and Douglas Porch "The French Army and the Spirit of the Offiens1 1900-14 " B vc H 1 &M . ' ' m nan Bond and Ian 'Roy, eds. War ancl Society (New york o mes e1er, 1975), pp.ll7- 43. '

1. some were not ~ ) n~,.trllstic. Grey's remark is well-kno'Wn ''Th over Europe: wt hd.!l not sc>c them lit again in our life-tim~ .. e lamps arc going out all ter, Bethmann ll:>llwt ~. also feared the conseguen. f . hThe Gennan Prime ~linis vieW was that it 'J1'] certainl y pay for the winner. ces 0 t e war. But the controlling Quoted 8(Boston:in Marti n Gilbert, \Vinston S. Churchill Tl1 Th, , H ou~;l1ton ~1iffi in. 1971). p. 84 . ' ' e ChallengeofWar, 1914-1916 9~99. ,r ~ . 9 Quester (fn. 4 ), pp. p c: Robert Art, The Jnifluence oJ corerf!.n PoliCij. S . S (J3everly Htlb: . age rD1eSS1onaJ Papers in InternationalS d' . 0~ eapou;er, 11 26-28. tu les Senes, 19r3), pp. 14-lS, "The 10. Konrad Jarausc:h. 191 "lllusion of Limited War: Chan. ll or Bet hmann k I . Holl , Cal cuJated Ris , Ju y 4, Central European Hi.stony lJ c:e h ( \A, . . I . 1. . , .vtarc 1969): p. 50 wegs .. ~., 11. Brodie, Wm an~ Po tttcs (:''llcw York: Mae:millan , 19-t3)' p. )Q. . . 12. President Roosevelt and the American delegates to the Lea e of ~ . meot Confere nce maintained that the tank and the mobile h: , arUIIations DlSarmalished the dominance or tht: offensive th us making disarm ') ery had reestab. efi . . :. ament more Boggs, Attempts to D ne and Lz.m1t Aggressive" Arnwment in v urgent f \larion . M' OnUJCtj and [coIumb1a: u mvers1ty or . 1ssoun. Studies , XVI, ~ o. 1 1941]: lP131 c) b Strategy th ' t' cl ' pp. 10o . Ut IS was a mmonty post 1on an may not even have been believed b th . . Y e nmencans. Tbe reduced prestige and mfluence of the military and the high press . . . ures t o cut go"emment spending throughout th1s penod also contributed to the lowering of defense budgets. 13. Jon Kimche, The Unfou.ght Battle (;\ew York: Stein, 1968); ~icholas \\llliam Bethell The \Var Hitler Won: The Fall of Poland, September 19.39 (1\ew York: Holt, 1972): AI~ Alexandroff and Richard Rosecrance, "Deterrence in 19:39," World Politics, XXJX (April1 977): pp. 404-24. 14. Roderick Macleod and Denis Kelly, eds., Time Unguarded: The Jronside Diaries. 1937-1940 (New Y ork: McKay, 1962), p. 173. 15. For a short time, as France was falling, the British Cabinet did discuss reaching a negotiated peace with Hitler. The official history UO\vnplays this, but it is CO\'ered in P ~1. H . . Bell, A Certain Eventuality (Farnborough. England: Saxon House. 1974' . pp. 40-40. 16. MacLeod and Kelly (fn. 14). 174. In flat contradiction to common sense and almost f;\erything they believed about modem warfare, the Allies planned an e~pedition to\ia to cut the supply of iron ore to Gennany and to aid Finland against the Russians. But the dominant mood was the one desc1ibed above. 17. Broclie (fn. 11), p. 179. 18. Arthur Balfour, "Memorandum," Committee on Imperial Defence. April 30, 1903, PP 2-3; see the telegrams by Sir Arthur Nicolsoo, in G. P. Gooch and Harold Temperley. eds., British DocU'ments on the Origins of the War; V 4 (London: 1-I.~vl.S.O., l929). ol. PP 429, 524. These baniers do not prevent the passage oflong-range aiJcraft; but f\en in the air, distan<:e usually aids the defender. . . 19. See, for example, the discuss ion of warfare among Chinese warlords in Hst-Sheng Chi, ''The Chinese Warlord System as an International System.'' in Morton Kaplan, ed., :Vetc Approaches to Intematicnwi Relations (New Y S Martin's. 1968). PP 405-25 ork: t. 20 Some American decision makers, including military officers, thought that the best "ay out of the dilemm a was to abandon the Philippines. 21. Quoted in Elting Moniso n, Turmoil and Tradition: A St11d!j of tire Life and Times cif Hen,ry L. Stimson (Boston: Houghton Miffiin, 1960), P 326.
0 0


" .1ne
,:,,-, "'I"L -





U S - c: ed .. 00 der limitations on Hawaiian defcn<.! ., ~wc# thc..:se wor~ . , . ,o DSl . . rems ;r. The Vmted Stale1 \ 0t 'J i 11 the r"aci.J IC ~L "UJJ no mrcat t0 Japan. ,,:,n=am Braisted. _ posed . 1909-1922 (Austin: Unh-ersity ofTexas Press, l~ tl}, P 6 12. 23. That is part of the reason why the Japanese admtrals strongly obJ.N.I<d whc,, the t.i\ilian I .ded to ~t a seven-to-ten ratio in lighter craft m 1430 Stcphen Pe z _ . p . d U01 --- ..leaders deCJ ., Race to Pearl Harbor{Cambridge, Mass.: Ha:a r . "erst) re'>~ 19~4 }, p. 3. 1963 : ~ ll>;,. AI o see il>ul , 24 . John ~ef, War and Human Progress (~ew York: 1\ort~n. .: dcde ~e.~ (Ithaca, :\.Y PP 237, 242-43 , and 323: c.w. Oman. The Art of War 111 the ;\f1 ,e, C<>meU University Press, 1953). PP 7~72; John B_:eler, H.a~are m ~eudo/ Eur01 .: (Ithaca. N.Y ComeU UniversJty Pr~ss. 1? d ). pp. 21:-14; ~1 chacl lloward. 307 1200 War in European History (London; Oxford Unwe~t)' Press, ~9t6 .' pp. :33-.'37. _ Quincy Wright. A Study of War (abridged eel.: Ch~go: Untverst') ~>f Chicago 25 1964), P 142. Also see PP ~70, 74-75. T~ere are tmpo~ant exce~hons to thc:se gen. eralizations-the American Civil War. for instance, falls n the mtddle of the period Wrigbt says is dominated by the offeose. . , 26. Geoffre) Kemp. Robert pfaJtzgraff. and Un Raanan. eels., The Other An m- Race (Lex. ington. Mass.: D.C. Heath, J975); James Foster, lhe future of Conventional Arms Control" Policy Sciences. So. 8 (Spring 19'77): pp. 1-19. 27. Richard cballener. Admirals, Generals, and American Foreign Policy, 1898-1914 {Prioceton. !'.J.: Prioceton University Press, 1973); Grey to Nicolson, in Gooch and Temperley (fn. 18), p. 414. 28. QucXedinJamesCf'O\Io-iey,Japon's QuestforAutonomy (Princeton, K.J.: Princ.-eton univeasity Press, 1966), p. 49. American naval officers agreed wi th the Japanese that a tento-six ratio would endanger Japan's supremacy in her home waters. 29. E. L. Woodward and R. Butler, ed., Document~ on British Foreign Policy, 1919-1939. 3d ser. ID (London: H.M.S.O.. 1950), p. 526. 30. Jervis (fo. 6), pp. ~72, 3.52-5.5. 31. Quoted in Merze Tate, The United States and Annoments (Cambridge, Mass.: Har"ard UDMmity Press, 1948), p. 108. 32. ~ (fn. 12). pp. l5, 40. 33. lenneth Hapt. American Gunboat Diplomocy arul the Old Navy, 1877- 1899 (West port. Coon.: Greenwood Press, 1973), p. 20. 34_ Wmstoo CburcbiiJ, The Gatlrering Stonn'(Boston: Hougbton, l 948), p. 206.



4-'3. 44

197 DllfMMA .. Quoted m Cl.1arl \\ n~tc:r, 7he Foreign Poli uf . . cy Ca.rtlereagh, 11 , 1Jf1.5- 11122 (IJmdon: G. Bell and ~Oil'> I!1fl'3 ,, p. -~J 0. aoggs (fn. 12}, pp. 1 4~ 15, 4 t-48, 60. . Quoted i~ PhihE JtmJan Frontier Law and Order . . .

Press, 19 t0), p. 1, abo sec: pp. 16-17. . BoggS (fn. 12}, pp. 20, ~ -

(Linwln: Urnvtrsity of :\1': hraslca

35. Bnxlie, War and Polities (fn. 6), p. 32.5. 36. BafP (fD. 12}, pp. 42, 83. For a good argument about the possible di.fferentiation betwren &IIIM aod defensive weapons in tbe 1930s~ see Basil UddeU Hart, "Aggression and tbe PJ\lllkm of'Wearms," Enf!)bh ~55 (July 1932): pp. 71- 78. :rr. Quoted in ~ (fn_ 12), p. 39. 38. Oa tl..ose pnds, the Cenuans claimed in 193-2 that the French forts were offensi\'e (lhld.. P 49). Similarly, fonifiOO fon.wd naval bases can be necessary for launc:hjng an
.U') 22). p. ~. 38 'l1le FAI!Pdi made this argument in the interwar period; see Richard challener. ~ tl/the Nation in Ann1 {New Yorlc Columbia University Press. J ~' ' .FiA 44 4pp.IBI-82. 'l1le Ces wns dimgnm see Rnn..., (fo 12)


lee Brl?ted (fn.


40. 0..(fa.24).pp.57~



Wha t Is Terrorism?






everyday vocabnlal): ~' " , ble part of the argot of th e late twe nti eth centur ysimilarly becom e an tnwspe nsa .d or impression o f what tenonsm 1 , but 1 k a more ac: most_people have;~at~~ le ealanatory definition of the word. This im precision ha-; precLse,con~rete tl Y ~dern media. whose efforts to communicate an oft en bee n abetted part1 1 dJe :sa ge in the bri efest amou nt of airtim e or yl JY print spac:e complex an con vo ute m . bJ h d t the promiscuous lab elling o f a range o f v1 Ient acts as "te rroro e e IJOSSI. p kave 1 o Japel or tur ]evl 'tl n on the te Slon ancl- eve n wt 1111 t11e same ism .' 1c up a news! . . . . . age -one can fin d such disparate acts as the bo mbmg of tl le Same I) b roadcast or on . a building, the assassination of a head of state, the massacre o f cJVJ!ta~ s by a military unit, the poisoning of produ ce on sup ermarket she ~v~s or the de hb er~te contamination of over-tl1 e-counter mecilcation in a che mtst s sho p all descn bed as incide nts of terrorism. Indeed, virtuaLly any especi ally abhorre nt act of violence that is perceived as directed against society- wheth er it involv~s the ~ctivities of anti-governm ent dissidents or governme nts the mselves, orga~tze~ ~n me syndicates or common criminals, riotin g mobs or pe rsons engaged m mthtant protest , individual psychotics or lone extortionists -is often labell ed "te rroris m. " ... Terrorism , in the most widely accepted contemporary usage of the ter m, is fundamentally and inherently politic;al. It is also ine luctably abo ut powe r: the pu rsuit of power, the acquisition of power, and the use of powe r to achieve politi cal change. Terrorism is thu s violence- or , equally important, the threat of violen ce- used and directed in pursui t of, or in seJVice of, a political aim. With tlU s vital point clearly illu minated, one can app reciate tl1e significance of the adrutional de fin ition of "terrorist" provided by the [Oxford English Dictionan;]: "Any one wh o attempts to further his views by a system of coercive intimidation ." This de finition unde rscores clearly the other fundamental characteristic of terrorism: that it is a pla nned, calculated, and indeed systematic act. Given trus ~el atively straightfo~ard elucidation , wh y, the n, is terrorism ~o difficult to define . The most compelling reason perhaps is be cause the meamng of the term has changed so frequently over the pas t two hu nd red years.. ..

d n? Few wo1 s have so ins iruously wo rkecl thei r way into 0 1tr L1a ts C nsr . ' ikc "Jnternet"- another grossly over-used ler JTo m th at has
t t

From with Inside.Terronsm 1 Bruce Hoffman, pp. 13 44. Copyright 199 . )y 8 Bruc.:e Hoffm an. Repnnted pennJSsJOn ofColumhia Uni versity Press.


rp risingly. .a. 1 . , '''<.ming and usage of the wo d h l Not su ' I 1 r ave c: 1ang(d 0 . "'odate tbc no. i!C cl vern acu lar and discourse [ h r cco111 1' 1. 0 eac suc cessi ' C titn <' lo ~ haS proved inC ff u',JT?~ly elusive in the fac e of att em t~ to COn . \ ~era, ten~rtS l defini tion. At oH' tlm c. the terrorists themselves ~v r . . struc:t fl' one conststent . il j . cl ere raJ mo re o this endeavouJ ,;:m l lCY are to ay. The early practi tion ers didn' coope rali v . _ e I ds or hide be hmd the scrnan ti<: camoufl age of m d t mtn ce theu wor fi I " . " I ore ano "freedom g tte r m ur )an gu ern.lla. , The ninete th- . yne labels such . as exa b 1 cl! 1 en <.:enturv anarchtsts c mple, un a as tE: y proc atm ed the mselves to b t . tO r ' .. .. . b . . e erronsts an d fran klv i\ proclaime d then tacttcs to. e terron sm. The me mbers of 1 arodnay . : d _ . th 1 a Volya 1 ly displaye no qu tumshm usmg ese same words to desc b th. f . ves . deeds. However, sue rankn ess did not las t The JeWJ's] tn e emse1 and tJ1eli . . 1 erronst gro . up oftb e 1940s kno wn as LehJ (t1 He brew c onym for Lohamei Herut Yisrae 1e lcr l, the Freedom Fighte rs for I~rael , mo re popularly k:nov.rn simply as the Ste m Gang after their fo un de r and first leade r, A~ r~ am Ste rn) is thought to be one of the last terrorist gro ups act~aU y _o ~e scnbe 1tself publicly as such. It is sig t nificant, however, that eve n Le h1, wlule 1t may have been far more can cild tha n its latter-dav counterparts, chose as the na me of the organization not "Te rro rist Fiahte;s for Israel," bu t the far less pejorative "F reedom Fighters for Isr ael.'' Si: ila rk although rnore tha~ twenty years later the Brazilian revolu tion ary Car!~~ Marighela displayed few co mpunctions about openly advocating the use of"terrorist" tactics, he still insiste d on depictin g himself and his disciples as "urban gue rrillas" rather than "urban terrorists." Indeed, it is clear from Marighelas writings that he was well aware of the word's undesirable connotations , and stro ve to cilsplace them with positive resonances . "Th e words 'aggressor' and 'terrorist, ., \1arighela wrote in his fam ous Handbook of Urban Guerrilla War (also known as the " \1iniManual"), "no longe r me an what they did. Instead of arousing fear or censure, they are a call to action . To be called an aggressor or a terrorist in Bra zil is now an honour to any citizen, for it me ans that he is fighting, with a gun in his hand, against the monstrosity of the pre sent dictatorship and the suffering it causes .'' This tre nd towards ever mo re convoluted sem antic obfuscations to side-step terrorism 's pejorative overton es has , if anything, become nwre entrenched in recent decad es. Te rrorist organizations almost without exception now regularly select names for themselves that consciou sly eschew the word "terro rism" in any of its forms. Ins tead the se groups actively seek to evo ke images of: Fr eedom an d lib eration (e.g. the National Liberation Front, the Popular Front for the Lib eration of Palestine, Freedom for the Basque Ho me land. etc.); Armies or oth er military organizational stn1ctures (e.g: the ~at i~nal Military Organization, the Popular Liberation Arm y. the Fifth Battalion of the Liberation Army, etc.); Actual self-defence move ments (e.g . the Afrikaner Resistance Movement. the Shankhill De fen ce Association. the Organiution for the De fen ce of the Free Peop le, the Jevvish De fense Organization, etc.):




Righteous '-engeance (thtle O anization for the Opp ~~" l or. E a rth , the rg Justice Commando o f 1e Arm an Genocide. the Palcs iut ,m Heveno em -::.e Organization, etc.) : th ta re decidedl) neutm1 an d t . - o r else deliberate~ choose na erefo re bertft mes or Of all but tl1e most innocuous sugges on_5 association, [<--.g. tht Sh injnu Path D ' Front Line . ai-Dawa ("'The Calf'.) A]faro u ,es- amn It 1-:ach "Thuo," al -G a n 1 at . ..) tlle Lantero Y uth ~l ovc, o nent. e tc .] . al-lslamiya (l be Islamic O rgam za:~;is tllat terrorists clearly <.lo no t sec or regard \\ b a t all these examples suggall l am ., amih' man: til e archtcrr d0 " \})0\"e orist Carlos a themselves as others : If . -d bed Wmse to a French newspaper follo"in g his capture in 1 "he Jackal. escn r . tuallvon the deie ns1 ., and forced to take up arm to protect them ,.. 1994. Cast perpe . -. . . ed stituents o n k terronst p e rce1,e h selve.s and their real 0 ~ 1 ma&J~d co t e m elve as b n tion-and lacking an) \i a b le a lt e m a th e -to reluctant wam.ors. dri,en. . , espera .1 .,: ... <+arepresst, e st ate . a predaton 1;,-al etlmic o r nationalist group , or v1o en ce ag<u.Ju . . nal rd This percei\e d ch tematio o er. aracten. . o f se lf denial also stic an unresponswe m . fro othe r t\ n PS o f political ..J:..&.:-.nnc:hes th e terrorist extre m ists as well as from m ~UJas........- . r. . 1 e d . w ~- ~ ,; olent a, . ocations. A communist o r a re similarlv 10\ec>rv m egcu. ,olupersons . uld likeh readih- accept and ad th h r mtt at e IS m tact a comtio~- for exampl1 - " 0 , d _:...,~ m e_ ~ ,. would doubtless take particu munist or a l"C\U ution~ . 1 lar pride in Jl c:ou., ' . . th . il ) . f t1 . claimma eJ er o lose appellations fo r iliemseh-es. Stm a r ~ , e 'e n a person en in illegal wholl~ disreputable o )l!- h . I . .. h r e ntirel} se Le> n o e nt acti,~ti es. su~ as banks or ~ing out contract killi ngs_- would probably_admi_t to bemg_a bank ri b e r or a murderer for hire. The terrorist. by contrast, " il l ne v e r ackno'' ledge that he is a terrorist and mo reo ver will go to_ e~ t lengths to ev gr ade ~~ obs~~re_ - r: rence or conn~on amsuch me The te rronst '"ill alwavs arQUe th at tt tS socteh 0 o r' the gO\-emment or the socio-e conomic syste~ " and its _ ws la th~t are the re~ "'terrorists... and mofeO\-er that if it were not for this op p ression . he " o u ld not ha' e felt tbe need to defend either himse lf o r the population h e claims to rep resent. .. On one point, at ]e as t, e\'eryon e agrees: terrorism is a p ejora~ ve term. _ is a It word with intrinsically negative co nnotations that is g e n e rally app lied to ones e~ e mies and opponents, or to tho se with whom one disagrees and would o then\1se prefer to ignore. \\ fi a t is called terr orism," Brian Jenkins has writt e n, "thus seems to depend on one's point o f view. Use o f the te rm implies a mo ral judgemen~ ; and ifone party can successfully attach the label terrorist to its o p p on 1 e n t, th e ~ _t has indirectly persuaded others to ad opt its moral vie-.vpoint." H e n c e the d ects~on t~ caD someone or label some organiz ation "t errorist" bec.-omes alm ost un a, oJdabl) RJbjective, depending largely on whet her one sympathizes \vith o r opposes the peuoulgroup'crnseconcemed . If o ne identifies w it h th e victim o f the violence. for "2mple, then the act is terrorism . If, however. o n e identifie s w it h th e perpetrator. the violent act is regarded in a more sy mpathetic , if n o t positive (o r, a t th e worst. an -abivaleat)light; and it is n o t terrorism . ... 1t.e approach where .identiiication with . the \i c ti m determines th e classifica ofa \UaMt act as tenorism is evident in bon th e conclusions o f a parliamen tary worki ng (an orgaumtion comprised o f lon g-established, status quo \Veste rn 1 \e final report of the 1989 North A tlantic Assembly's Subcommittee 00



II;A ..U '

srn states: ~~~ . . -: kidna ppi_ng , arso n an d other feloruo '[e~nal beha 'io r, , -. .un: no . n- \\ este m natio ns have . d us acts COnstitute crifi'IJJ'l te rrorist a(; "" .l c1t th ev consider to b e struggles pfrO\e. reluctant to condefi'III as ning. tile d eL;-,, 1g h , c aracte ristic o f terrorism is 0tll nationalliberati " In '- 'c reaso f . u,..., . c on .....,0tj vatio ns o. tS tlfication 10 e o \lolence 1. 1f r or reasons behind .t t th e " ' B . . tse , h no . el satisfactory solution e it t er, since it fails to diffe rentiate cleaut th tS ts not ent:JI Yrpetrate d bv states a rlv betw ~n nd b, non-state entities such as te . ' een \l olencel ~nto the hands o f te rroris~ and their apolo~ ..~ who wou7don t p ays I sts. A<thccorthding\? l "l . h'' I o difference benve . .ocn t le O \\ -tee te rr? n st ptpeat e 0 bomb placed argue rubbiserbits in the rowded marke t that wantonly h n and mdiscriminatelv kills or m . atthi~ a radius me asured in ten a s o f feet and the 'mgh-t~cb" prec ,~, n . r isioru~ ~ded onde ery e d b y rur 1orce figl1te r-bombers nance dropp n gw or . from a height of 20 000 ceet d. . that achieves th e s~me ~,a n to n Ii or more an mdiscriminate effects on ilie , crowded marketlace far below. T his rah?~ale thu s equates the random \iolence in flicted on enemy P ulation c e n tres b y milttary fo rc e s- su c h as tlte Luftwaffe's raids on Warsaw and b!-entty, the Alli ed fi re -bombin gs o f D resden and Tokyo. and th e atomic bombs dropped by th e Unjte d States on Hiroshima and 1\agasaki during th e Serond World War, and inde ed the cou~ teiVal ue st rategy o f the post-war su perpowers' strategic nuclear policy, which d eliberat ely targe ted the enemy's ci"ilian population-with the violence com mitte d by subst ate entities labelled "terrorists," since both invoke the infliction o f d eat h a n d injury on non-combatants.. .. It is a familiar argum e n t. Te rr orists, as we have seen, deliberat ely cloak themselves in the terminology o f mil itary ja rgon. T hey consciously po rtray themselves as bona fide (f ree d o m ) fi gh te rs, if not soldiers , w h o -t h o u g h they w ear no identifying uniform o r insi g n ia -a re en title d to treatm en t as prisoners of war {P O Ws) if captured and the re fore should n o t be p rosecuted as common crim inals in ordinary courts o f law. Te rrorist s fu rt h e r argue that, because of their num erical inferiority, far more limited firepowe r a n d paucity o f resources compared " it h an established nation-state's massive d e fe n ce an d national secu rity apparatus, th ey have no choice but to operate c la n destinely, e m erging from the sh a dows to carr y out dramatic lin other words , bloody a n d d est ru ctive) acts o f hit-and-run ,;olence in order to attract attention to, a n d e n su re publici ty for, the mselves and their caus e. . But rationalizations such as th . ese ignore the faet that, even while national anned forces have been responsjble for far more death and destructio n than terror~ might ever aspire to bring about, th e re nonetheless is a ~damental _qualita~: difference b e tw e e n th e two typ es o f violence.. .. In theory, 1f no a.t. . t always ID practi ~rules o f w a r- a s o b se rv e d from the early seventeen th ce ntury~"he~ thP\ wthe -,_ ere first Proposed b y th e D u tc h jurist H ugo Grotius and subsequently co CamOUS Ceneva a n d H a g dified~ d . _..r u e Conventions on \Vanare of th 1860s 1899. 190 t, an e , l949-not only g ra n t civilian non -combatants immunity from atta ck. but also Prohibit taking civilians as h o red or surrendered Impose regulations goverrtinstages ; g th e treatment o f captu soldiers (P O W s); , Outlaw reprisals against e it h e r ci Yi.lians o r p O \\ s; ~ _ Recogniz gh f tizens of neu tral >Y~Les.. a_.J e n e u tr a l te rr it o ry a n d the n uu ts o Cl - - - -- -



"~l'se nt Uphold the in..,iolability o[d. Iomats . d other accrc<.htr .tp an <lhves. . u Even the most curson reVIew of terronst tactics and targets v\' ' r1 artert~nt\uY re,eals that terrorists have violated all these rules. T1 ll) ; :i r..-quently have le)' .. 1 . h0 m in some instances the~ ha' e then brutall Taken hostage C lvtJans, w Y . tl c Italian prime minister Ald0 \11 oro cli1 CI the Cermm executed (e.g. 1e Iormer ,0 . ,1 k, ,. . . . .. Martin Schleyer, ,. v.rere respec "e :' ta en capt1 and .,ho ve mdustrtalist Hans . d d the Red Anny Faction) \ . cl d b the Red Bnga es an ' ateJ mur ere d dered kidnapped militaJ officers-eve n when y Similarly abuseo an mur . . UN-sponsored peacekeepmg or truce supervisurv they were sen'lng on ' . . th A can Marine Lieutenant-co1oneJ \vtt1am Higgins 1 miSSiOns (e.g. e men d 1 1 ' ~ was der of a U truce monitoring etac 1ment, w 10d) abducted the comman tl h by Lebanese Sb''a terrorists in 1989 and subseguen y ange ; 1 1 ft Undertaken reprisals against wholly innoc~J~t C Ja~s, o e~ m,countries JVl far removed from the terrorists' ostensible t e~tre o ofp~~at1on , thus disdaining any concept of neutral state~ or the ng 1ts o CJtiZens of neutral countries (e.g. the brutal 1986 machine-gun and hand-gre~ ade attack on Turkish Jewish worshjppers at an ~stanbu~ s~nagogue earned ou~ b~ the Palestinian Abu Nidal Organization m retal1ation for a recent Israeli ra1d on a guerrilla base in soutllem Lebanon); and . . . . Repeatedly attacked embassies and oilier diplo~a~c ~stalJations (e.g. the bombings ofthe US embassies in Beirut and Kuwrut City m 1983 and 1984, and the mass hostage-taking at tlle Japanese ambassador's residence in Lima, Peru, in 1996-7), as well as deliberately targeting diplomats and other accredited representatives (e.g. the British ambassador to Umguay, Sir Geoffrey Jackson, who was kidnapped by leftist terrorists in iliat country in 1971, and the fifty-two American diplomats taken hostage at tile Tehran legation in 1979).
1 :-


..,..;Jla warfare is ~o > 1 place to start. Terrorism . . c cueJ 1 .h IS O rtell confus d . I r treatet as ~~"' H._. ;nous W 'guerrilla warfar Th . it e or tqllattd wJtl , o ft I th e. I .. . e uuerrillas o en . 'P O) e same tactics (assassinatiIS ISnot entire~.S\Jrpnsmg. kid sine I;) I h . . . f blic gatl1enng-p accs ostage-taking, ete:.) for th on nappmg, I:>nm I)JO~s 0 pu erce thereby afiecti ng behaviour through the .e samf Pt-pose (to intimitlate odr c1 on both teJ:roJists and guerrillas wear neither ~nr~~tsa 0 fe~r) as tetTorists. Jn l~ ac ' .I~ IIOrm nor 1dent:i:f\ 1 and thus are often inu.istinguis ~able from non-combatan ts. How . /mg 1 ~s1gnia . clination to lump botl1 tcrronsts and guerrillas into th e. er, desptte the JJ1 . catch-all <:ate . f "irregulars,, tl1e1.e are none th e less fundamental dif~ e same b gor; o . , c . 1 . . . erences etween the tw0 ''Guernlla, ror examp e, m 1ts most w1dely accepted usa . ak . I f. d. . . ge. numencally arger group o anne mdmduals, who operate as IS t .1,en to refer to a c d . dl enemy military rorces, an se1 an 1old territorv (even if on! a mtutai) umt attack ze 1. all . 1 . Y epnemer y dunng ), wh1 e a1 exerc1 . some form1 of sovereign"' daylight hours so smg l . hi . 1 , d . . ; or contra over a defined geograp ea a1 ea an 1 population. Terrorists howeve d . fu . . ts . r, o not d uruts, generally do not attempt to seize or hold territ .nction m the open as arme d l'b _ d 1 c . Of\. e 1 er ately avOJ enga~g en~ my m11tary r_orces m combat and rarely exercise ~v direct control or sovere1gnty etther over temt01y or population. It is also useful to distinguish terrorists from ordinary criminals. Like terrorists, criminals use violence as a means to attaining a specific end. However. \\'h~e the violent act itsel! m~y be simil_ar-kidnapping, shooting, arson, for example-e_ the purpose or ~ohvahon clearly 1s not. \Vhether the criminal employs violence as a means to obtam money, to acquire material goods, or to kill or injure a specific victim for pay, he is acting primarily for selfish, personal moti,ations (usually material gain). Moreover, unlike terrorism, the ordinary criminals violent act is not designed or intended to have consequences or create psychological repercus ions beyond the act itself. The criminal may of course use some short-term act of ,iolence to "terrorize" his victim, such as waving a gun in the face of a bank clerk dur~lg a robbery in order to ensure the clerk's expeditious compliance. In these mstances, however, the bank robber is conveying no "message" (political or other~vise) y lrough his act of violence beyond facilitating the rapid handing over of his ~oot. The criminal's act therefore is not meant to have any effect reaching be~ond either the incident itself or the immediate victim. Fmther, the "iolence is neither ~nceived nor intended to convey any message to anyone other than the bank clerk himself, whose rapid cooperation is the robber's only objecti,e. Perhaps most fun?amentally, the criminal is not concerned with influencing or affecting public opin~on: he simply wants to abscond with his money or accomplish his mercenary task tn the quickest and easiest way I)OSsible so that he may reap his reward and enjoy the . . . . fru 1ts of his labours. By contrast ilie fundamental aim of tI1e terronst .s vtoIence : ulti~ately to change "the system"~about which the ordinary criminal, of course, ouldn t care less. . .. an We ~ay therefore now attempt to define terrorism as_ tile de~berate crea~ion ~ ~xplOitation of fear through violence or the threat of viOlence m the pursuit of po tical change. All terrorist acts involve violence or tl1e threat of 'iolenet>.

Admittedly, the armed forces of established states have also been guilty of violating some of the same rules of war. However, when these transgressions do occurwhen civilians are deliberately and wantonly attacked in war or taken hostage and killed by miJitruy forces-tile term "war crime" is used to describe such acts and, imperfect and flawed as both international and national judicial remedies may be, steps nonetheless are often taken to hold the perpetrators accountable for these crimes. By comparison, one of t11e fundamental raisons d'etre of international terrorism is a refusal to be bound by such rules of warfare and codes of conduct. International terrorism disdains any concept of delimited areas of combat or demarcated battlefields, much l~ss respect of neutral territory. Accordingly, terrorists have repeatedly taken the1r often parochial struggles to other sometimes geographically ~t~d . , ' P~ countnes and there deliberately enmeshed persons completely unconnected With the terrorists' cause or grievances in violent incidents designed to generate attention and publicity... . s ate defi Are we to conclude t11at terronsm 1 unpervious to precise much 1 accuJ ? ess mti~nh ~ot entirely. If we cannot define terrorism then we' can at least usefully . nguts tt fmm other typ 0 f 1 ' tl at make terrorism the disr . es vio ence and identify the characteristics 1 met phenomenon of political violence that it is.





1to hwe ~~tr-rcac.:hing p:o.' I, ' 1 ~..:,.cal f'fccts b f <tl e terrorist all,tc: k. t j, . ll'i\11 l to .. . ~~YOnd ll.' rt\1tiSil1 is sped flcaJly d<'S1~11~< 1nst1 fc . t, . t' (.;) or ohrect o 1 . Ih' 111\llh'l1 l VK 1111 . 'cl . "ttrttct audi<'1H'< . at '''11?;h l induc) . . lr . t:l t e ( b I tl , . b ' jntitn1datc . a WJ etntrvanational t;o\c tlll'lllorpolitica] <~nvaJ . I. .. \\ lttc) i t' thuk or religious group. aaln c~bJ e c?tn is,desiP"ned to tr< ;,r, llowc r wl,cre ll parl), . 1erc i5 n . . renen J. en-onst . 111 11 the pnblicitvg is verY li ttle. Thnw! tc or putl 1 opuuon g 1 tJlere . en er . )ower w 1ere oliclate 1 . k to obtain the Jc,eragc influence and . none or t o cons power ' olence terronsts see 11 t1 r " 1 liticaJ change on either a local or an intcrnati fl' all~ >~ 1e1 0 ona1 thev otherwise lack to e ect P scale. .

--- -


rhe Fungibility ofForce


There are two fundamental reasons why military power remains more essential to statecraft than is commonly thought. First, in an anarchic realm {one without a central government), force is integral to political interaction. Foreign policy cannot be divorced from military power. Second, force is "fungible." It can be used for a wide variety of tasks and across different policy domains; it can be employed for both military and nonmilitary purposes ....


... I have argued that force is integral to statecraft because international politics is

anarchic. By itself, that fact makes force fungible to a degree. Exactly how fungible an instrument is mi litary power, however, and how does it compare in this regard to the other power assets a state wields? In this section, I answer these questions. First, I make a rough comparison as to the fungibility of the main instruments of statecraft. Second, I present a counterargument that force has little fungibility and then critique it.

Comparing Power Assets

Comparing the instruments of statecraft according to their fungibility is a difficult task. We do not have a large body of empirical studies that syste_mati~y a~~yze ,~h~ comparatjve fungibility of a state's power assets. The few studies we do ha,e, e e though they are carefully done, focus on onJy one or two instruments_ and are m?re concemed with looking at assets within specific issue areas than With c:ompanng
(Sro Robert J. Art, "American Foreign Poli<.)' and the Fung~b h "~;ner l996), pp. 7-42. Copyright l996. Reproduced by pE'nJUSS!On ttp

. ili

fF0 Security Studie-1. \'ol. 5. No. 4 ty~ . ~1; i r & Fr.llldS Croup. LLC.



. .

.. .




we lack a conseque nceThrouuhsurficicnt videnec to c:om pare . I ~)'IC. however ;t:' s ('h; acro 'b'lih Cl btt (. 1 '' o r , we C:at1 pmnr assets accorwng to tl t ett' fun<1l I l ' o l . cU [ me uc atwn . Pl'O\ 1C1 :<0. 111 c b 'l,>ark estimates. a state O\'1)1$ Tlle\' . 1 lt l)t)l)U 1 . - tl1e sr . ts . Constder what power asse . .. , . . u of 1ts CtUL.A- ' geogyal) ,L1)'-t1 SIZ(. 1OtcllJOll . anu natule, ,.;.,.,olllV' le a) 0 le\ eI ski s Ulcl r eclucation . e the effecthc1tcss of its potti ru resource endowment of the state; govemanc r tl .. . I c, ~)'S11ts 1 eoloO'\ and r b\ and stands ror. 1e nature o es c . \~ 'uee-tlt e norms a state li \ l OJ te111 aJ " . . altl: -the leY el. sour<:es ' anc nature of 1s pro. t tl f a a1 to fore1gners; we 1 0 l e ~:\tent tts ppel cl h' the political skill of its leaders and the number of skill ducn\e ec :onomv; ea ers tp-. cl .. . . fulleaders 1t , ' and mili'tary, po\~er-the nature, slZe, an composition of 1ts militan nas: . ..L . .r c: all rorces. Of th ese assets. ,vealt11and political skill look to be UJe most versatilegeog. rap1y a11 d go\emance tile least versatiJe. because both are more m the nature of 1 givens th~t set the physical and political context \vi~lin which tl_1e other assets operate; ' .,.e o vaJ ues and poptu...ti011 cu hicrhl)' va1iable.' depending, respectJVely, on the content .of ( . . .. tile values and on the education ru1d skill of tl1e populace, and mdJtary power bes somewhere between wealth and skill on the one hand, and geography and governance on the other hand, but closer to tlte former tl1an to the latter. In rank order, the three most fungible power assets appear to be we~~: poli~cal ski!J, ~nd military po~'er. Economic wealth has the highest fungJbility. It IS the eas1est to convert mto the most liquid asset of all, namely, money, which in turn can be used to buy many rufferent things-such as a good press, topfught international negotiators, smart ]av.yers, cutting-edge technology, bargaining power in _international organizations, and so on. Wealth is also integral to military power. A nch state can generate more military power than a poor one. A state iliat is large and rich can, if it so chooses, generate especially large amounts of military power. The old mercantilist insight that wealtl1 generates power (and vice-versa) is still valid. Political skill is a second power asset tllat is highly fungible. By definition, skilled political operators are ones who can operate weU in different policy realms because tl1ey have mastered the techniques of persuasion and jnfluence. They are equally adept at selUng free trade agreements, wars, or foreign aid to their citizens. Polltically skillful statesmen can roam with ease across different policy realms. Indeed, that is what we commonly mean by a politically skillful leader-one who can lead in many different policy arenas. Thus, wealth and skiJI are resources that are easily transferable from one policy realm to another and are probably the two most liquid power assets. Militazy power is a third fungible asset. It is not as fungible as wealth or skill, but ~hat does not make it illiquid. Military power possesses versatility because force is mt~ ~o poutics, even when states are at peace. If force is integral to international poli~cs. It_m~st b~. fungible. It cannot have petvasive effects and yet be severe!! restricted m Its utility. Its pervasive effects, however, can be uniformly strong, umformly weak, or variable in strength. Which is the c.-ase depends on how milit~' ~er affects the. ~any domains, poucy arenas, and disparate issues that come with_m Its fi~ld. At tll~ muumum, however, military power is fungible to a degree because ItS 1 phys_ use, Its thre~t~ned use, or simply its mere presence structure expect~~ons ca1 and mflueru:e the_ ~litical calculations of actors. The gravitational effects of mJlitar)' ~e~ mean that Its mfluence pervades the other policy realms, even if it is not dornmant m most of them. Pervasiveness implies fungibility.
. . ~ L'i \s ~ l~ ll e eli te . f .

e case ' l 1 n ubt .ny power, moreover great ~ In tl1 . er amounts 11 pomt, more of' it is therefore . ~f .mc:rca\e its bility Up to (I t: 01.1<1 ).C tt fun@ d uable to be nllttanly powerful than m'''tariJ k bettc:r than le\li. lt is re es m v wea ~'llt 'I motes l1ave ocrreatN eh 1t m world politic:s than militaril\J ak ' 11an ~ powerful we ones :\Ill 1 1 sta are less uhjc et to th e influence of other stat~s tl1 . . . tan Ystron~ 1 states 1 b rr an ffilttanlv weak , , 1; .1 powe1fu '-latcs can etter o 1 protection to th )' er one!\. \hvtan . . ..1 , . .A o er states. or more s I . threaten them, 111 orucr to m uence their bebavior th . . . . enous } 'li 1. . r 1 an mtlttari.h weak FinaiJy. 1111 tan~ po'"enu states are more secure than <:an ril .1 . ones. b 1 b' mt 1 \' weak ones ta ,e more clout. to c ess su Ject to the ""ill of others to b . To ha\ . . 1 ." e m a stronger po!li. to offer 1 orote<:tion 0 1 l 1l eaten 1 1arm, and to be secure 1 , worId \\'here others n a oon .. nsecure- these are po1 1ti<:al advantages tl1at can be diplom ti ll are 1 . a ea y exp1 . 01ted, d thev can also str engthen .the will, 1 esolve, and bargamjng sta11ce or th e state that an ' h 1 . 1 has them. Thus, alt 1oug m1 tt~l)' power ranks behind wealth and skill in tenns of its ,ersatility, it can be a close thnd behin~ those two, at least for those great powers that choose to generate large amounts of 1t and then to exploit it.

conflating Sufficiency and Fungibility

The view argued here-that military power possesses a relativelY hicrh decrree of . 0 0 fungibility-is not the conventional wisdom. Railier, the commonly accepted ,iew is that put forward by David Baldwin, who argues that militru;- power is of restricted utility. Baldwin asserts: T"o of the most important weaknesses in trarutional theorizing about international politics ha\e been the tendency to exaggerate the eiTec heness of military power t resources and the tendency to treat military power as the ultimate measuring rod to which other forms of power should be compared. 1 Baldwin's view of milita1y power follows from his more general argument that power assets tend to be situationally specific. By that he means: vrhat functions as a power resource in one policy-contingency framev.,ork may be inelevant in another." If assets are situationally or domain-specific, tl1en tl1ey are not easily transferable from one policy realm to another. In fact, as Balch~rin rugues: p~! power resources ... tend to be much less liquid tbru1 economic resources ; and although power resources vaty in their degree of fungibility, 'no political power resource begins to approach the degree of fungibility of money.''2 For Baldwin, two consequences flow from the domain-specific nature of power resources. First, we cannot rely on a gross assessment of as 's o,erall. power assets tate . tn order to determine how well it will do in any specific area. Instead. '' e must assess the strength of the resources that it wields in that specific domain. Seco~d. the ~en :rally low fungibility of political power resources explains what B~d~'m calll> ~: paradox of unrealized power": the fact that a strong state can pre,ru1 m one ~ area and 1 m another. The reason for this, h 11 . e te s us, 15 stmple The. state .o..U JSSue ose . d t has strong assets in the domain where it prevails and \veak ones wh~l"E' 1 ~ . ~ no t On the face of it Baldwin's argument is reasonable. It makes mtwti\e senseot~ argue ' d atir1 armies than the, are at prom . or example, that armies are better at e e g th 'ti . that the more mg stable exchange rates. It also makes good sense to take e posi on fi . e the t'arefuU b to bear on a SJ>!C1 c tssu Ywe assess what specific assets a state can nng





ART I THE FUNGJBil ..,..,, ._.. , OF FORCE

~ :_.,;_


~' our !<.>e1\\1'Li l>e o 1 w hat tlte state cm n.'ah<.:ti, tlh 1 'I 1t)lish on t hat < domau1-~r 'l)et'ih<: 1o 1 'l'('L is tllcreorc r ~-sue . To denv t11at ,JJ po\\'er assets are t J E 'i:-.t'l'tto ~uro . quaI.Jy a bsurd . howe,er are the positions that .all <L. t!l oU"C' ( omain-S[)ecific . . . . .. . . d ,. . d t]1at a moss uwentor" of a state~ O\ ( r >m\ ll .tssets 15 not . t 0 tJ ll:' ,une egree. clll ' . . 'k I I . a .t r y . ,..,,;de rel1aoIe. e,en 1 on1 a rough, e....... to how well the late 1s h l ' ! >-. o m anv c. n . ' oive 1 . onKUJl. A e t s are .not equalt'n funoibility and fiJ1'-hll1111!! c ov' not mean dramati s I:J d c-.,t.lh alteJing assessments. . .. .1. . . ~ .-:> . \\"hat does all t11is mean fo r t1te fu ng:tbiJJt;' of nH It.u; po\\ (; I. Should we . . t? accept BaId wm .s VIew about 1 . r argue that ,,e should not. To set wll\', let us look . in ~rreater detail at what else he has to say. . e Baldwin adduces four examples that purport to .demonstrate the illnited versa. . f ili't . ' ' til1tv o m ary powe J. 3 The examples are bypothebcal. but are nonetheless useful 1 to ~nalp because tJ1ey are equivalent to thought expenment . T 1csc arc the ~ examples:
mort' Rm-t\ull'd

Possession of nuclear weapons is not just irrelevant to secudng the election of a U.S. citir.en as UN secretary-general: it is a hind ranee. . . . The owner of a political power resource. such as the means to deter atomic attack, is likely to have difficulty converting this resource into an?t!1er resource that would. for instance. allow his country to become the leader of the 1 lmd World. Planes loaded with nuclear weapons may strengthen a states abili ty to deter nuclear attacks but may be irreJe,ant to rescuing the Pueblo [a U.S. destroyer seized by the North Koreans in earh- 1968] on short notice. ' . The ability to get other c-ountries to refrain from attacking on:~ homeland is not the same as the abiJity to "win ilie hearts and minds of the people m a faraway land [the reference is to the Vietnam War).4

Seemjngly persuasive at first glance, t11e examples are, in fact, highly misleading. A little reflection about each will show how Baldwin has committed the carrunal error of conflating the insufficiency of an instrument with its low fungibility, and, therefore, how he has made military power look more domain-specific in each example than it really is. Consider first the United Nations case. Throughout the Un ited Nations' histo.t)', the V nited States never sought, nor did it ever favor, the election of an American as secretary-general. If it had, money and bribes would have been of as little use as a nuclear threat. The Soviet Union would have vetoed it, just as the UnHed States would have vetoed a Soviet national as secretary-general. Neither state would have countenanced the appointment of a citizen from the other, or from one of its client states. The reason is clear: The Cold War polarized the United Nations between East and West, and neither superpower was willing to allow the other to gain undue inAuen<..e in the institution if they could prevent it. Therefore, because neither super power would have ever agreed on a national from the other camp, botl1 sought a secretary-general. from the ranks of the unaligned, neutral nations. This e>.~lains wha col~ war secretaries-general came from the unaligned Scandinavian or Third Worl ~ (Dag Hammarskjold from Sweden; U Thant from Burma, for example), particuJarly during the heyday of the Cold War. This arrangement, moreover, served both superpowers' interest. At those rare times when they both agreed that the

. C' r The Third \ Vorld CX<tmplc ts <-'q uallv misleading To see w1IV. 1 t u pcnonn a . sunpie 'thou(J'ht e v enmcnl. Although .a Third World lead.,., t,,.t1t 1 d arnH:< tis '"' . ' " la >II state with nud ea r weapons mtght not nse automatically to the top of the Third W pack, he. or she ''~ould bccon:e a m~ghty impottant actor nonetheless. Think orld 1 trhh. ld 1 1ow less wct.0 ;" C hn1a and Ind1a, whtch have nuclear wepons, wo11 appear o . . . .. . . _ . "' to other states I r tiH) dt d not posses~ them.; and tlunk of how Iraq, Iran. or Libva. "hjch do not have them. would be vtewed 1f they did. For the former set of states. nuclear weapons add to their global political standing; for the latter set, their mere attempts to acquire them have caused their prominence to rise c:onsiderabk Bv themseh-es. nuclear weapons cannot buy the top slot in the Third \\.orld or.els~ where. Neither economic wealth, nor military power, nor any other power asset alone. c< buy top dog. That slot is resetved for the state that su rpasses the others m in all the key categ01ics of power. Although they do not buy the top position. nuclear weapons nevertheless do significantly enhance t11e international inOuenc.:e of any state that possesses the m, if inB uence is measured by how seriously a state is taken by others. l n this particular case, then, Baldwin is correct to argue that nuclear weapons are not readily convertible into another instmment asset. Although true, the poi nt is irrelevant: They add to the ultj mate resou rce for which all the ot11er assets of a state are mustered- political influence. The Pueblo example is the most complex of the Clses, and the one. when reex< amined, that provides the strongest support for Baldwin s general argument.5 Even when reexam ined, this strong case falls far short of demonstrating that military power ha<; little fungibili ty. . The facts of th e Pueblo case are straightforward. On 23 January 1~, ~orth Korea seized the USS Pueblo, an intelligence ship that was fitted ,,; th so~hlsticated electronic eavesdroppi ng car)abiJities and dtat was listening in on ~orth K orea. and al nost vear er did not release the ship's crew members untiJ 22 December 1968 1 ''\ . . art. the hacl been captured. North Korea c1 d tl1e sup ''"" patrolling 111S1de tts } ~,.. Y aune .. . twelve-mile territorial waters Hmit; the United States<l tile dt~m b~mse. tts radi 0 "fix" on the Pueblv showed that it was patrolli fic.t ng 11een 'ctnd a hall nautical .nu ed .e from the nearest North Korean land point. Immediate!~' after the ~eizuredi Ul ~'ooot , ~he States b r cl . r ~. eere up 1ts conveutional and nuc1 1orces 111 East As1a. sen ng. , ll , ear . Navv d . d..J' . a1 -aft to Soutl1 Korea, as v. e :l.S ; an Air Force reservists and 350 a wtion <UfCl . . r t 5 mo, t1 d. k [i w1tJun a lt'W nnnu e . 'lng le aircraft carrier USS Enterplise an tts tas . orc-e S K '"Ul b~es A lng t' }' f' the (urcwn sent to out1 or.. , ... c.. t tme of Wonsan, North Korea. Some o

r~ions tot dd I" lw lpl'ul. 1'\ mediation wa<; 1,., I I N' . 'tet I.J11 1 j .. , t,u-v-fJ't'l lt'l I t I rat ,,ac, ll('lllral, not al igned ooclut more en' l'rw I)<'tau <'it C<: . I <l sec I t . ~ . . . .. . . . hll . 11 ,. r ,cn t1 \ nil ne .re, rndrtat'\ powtr htd not) . cl f m<l ' ltng to o \\it! 1 1 .. . le 1' J we ... 1w11ld nnl <.:oncludc that it llas r10ll11ng to 1 c .l'l'hn~ \tlr('. s-<tC'l 1 ta!1e .c- "ithin tlte in,l1l 11lion. Alllcrieas 11 rct>minen. . . to \\Jtt Arncrita' . <:e \\1t1 thr t.; 1 1111 1 11tbu'lg 5l<t ".1 clear. So. loo. b the lm:l that this stems fron A . . nrtt< :\atiow. h' s eel 1 ment"t p r 051 ron a. the ll! . trorwest nation. a position detking from both .t . . . \ 11s s 'Of c t1 <: . 1 nn<:I<'<IT weapons cannot bu, se s econom 1 and mt.1 '' .A-Jr Th us. altho11~ 1 t ta0 trenc.' . . , ere an -gene raJ P lC' f . iUtarv powN 1 mngs great 1111iucnce in an internati l . . . <: IOns. ttreat n1 . . . , . . I' , .. 11 . . . . . ona 0 1ganl7.ation. oue of o "' 111,un ptt tl)O'itS, ,1 t( t ,\ , IS to lth1eve collective secu 'ty tl 1 the thr(:at or wtOSoc 1 n . 1rou~ 1 use 0 fforce. .



, ...



,., ,



... ,... .....

"' lli .,


:! J u




r '> idt>n t J<> l

' :.uH I llc>St' on the 1 !t] w<rc " "clcar WjJahle. Act orcl! , I -.rllt

.. . '\('\< ralnulttal~ npllon<; wer e tonsrd cr ccl hut ultimat eh a rlllll lll!! \\'on~an Iwr 1 tor: 111111111.., otlrc r .



North Korean harlmr..- lilt\ Hlk ting <.:O't.\l l J I . I f ' . . . \ K . ,. ship striking -;ek :ctc t :\ort 1 ,u11.,, rarg Ns hy a s liJl .. pmg: ' <.: 11.111g a ort 1 0 1( <Ill 1 . k uaval ~nnfire;. In Padt case we deci Jed that th~ n'i wa'i t(JO ~r<a t and lh< aJr <ll td , pos~iblr .I all "J c not want to wul the arj.{IIIIH ut .md lose th< . .. lo a~co 1np Ir~ ww11t too \Ill 1. ><I< 1 ' () con sbtcntly war n etI H t)' advrscrs.

f~tct , rigltt from the outset of the crisis. tl~e North ~orcan ncg~)trators nadc clt>a r that onlv an Am erican t:onfessiou that rt had spu~d on 01 t 1t Korea and ha u iutruded into its territorial waters would secure the crew~ release . For elev<'n 111 onth s the Uni ted Statcs continued to insist th at th e:: P11eblo was n()t engag~d in illegal activity. and that it had not violated No1th Korea's ter:Jitorial waters. Only 011 22 Dece rnhe r, when General Gilbert Woodward. the U .S. rep rese ntative to tlw negotiations. signed a statement in which the U.S. govcm me nt apologized for the espiouage aud tl1c iutrusion. did North Korea release th e ~n.: w. Th ~ American admission of g 11 ilt, however, wa'> made und er protest: !111 medratcly before signing the statement, the government disavowed what it was abo ut to sign ; and imm ediately after the signing. the governmen t disavo,.ved what it had jus t admitted. Alth ougJ, the facts of the Pueblo case are straightforward, the interpretation to be put on them is not. Thi!:i much is dea r: Neither nuclear weapons, nor any of An1 clica's other military assets, appear to have se<;Ured the cre w's release . Equally dea r, however, is that none of its other assets secu red the crew's release either. Should we then conclude from this case tha t military power, diplomacy, and wha teve r othe r assets were em ployed to secure the crew's rele ase have low fung ibility? Cle arly, that wou ld be a foo lish conclusion to draw. The re was only one thing that sec ured the crew's release: the public humiliation of the Un ited States. If nothing but hu miliation worked , it is reasonable to con clude tha t humilia tion either was, or more likely. quic kly became North Korea's goal. Wh en an adversary is firmly fixed on humiliation. military postu ring, economic bribes, dip lom atic pressure, economic.: thre ats, or any oth er tool use d in moderation is not likd y to succeed. Onl y cxtrerne measures, such as waging war or economic bloc.:kade, are like ly to he suc~essful. At that point, the cos ts of such actions must be weighe d aga inst tire ben efits. One clear lesson we can draw from the Puehlo case is tha t sometimes the re are tasks for which none of the traditional tools of statecraft are sufAcient. These situations arc rare, but they do on occasion occur. The Pueblo was one of them. There is, howeve r, a second and equally important point to be drawn from this example. Although it is true that Am erica's military pow er did not sec ure the crew 's release , nevertheless, there were oth er reasons to unde1take the mi litary buiJdup the United States subsequently engaged in. Neitl1e r tl1e United States nor Sou t!l Ko~ea knew why the North had seized the Pueblo. Preside nt Johnson and 111 5 ~visors, however, speculated that the seizure was rela ted to the Tet offensive 111 VJetna.~ t~t began ei~1t days after the Puehlo s cap ture . They reasoned t.l ~at cl1e PueJJ_Lo s seizure was deliberately tim ed to distract the United States and to fnghten the South Koreans. Adding weight to this reasoning was the fact tha t the Pueblo was

Tire American govern ment's denial, its military n.l<'a'>nn-. and its suhsequ< nt diplomatic eflorts, were to no ava il. North Korea rclu scJd lo nl~'asc the t:rew. In

nng t 1e1r hvo di,iSJO The Puehlo's sci:.urc.: th11s raised three problem s forth < Unit. 't" crew and sh1p back; how to dete r the \'<>rtl f cd hO\"' 11et r ' 1 rom engtgStates: f rth to :r n , ..1tive acts ; and how to reassu re the South Kore ffi . < mg m u er rovo..., . . . r uld keep t11eir troops 111 South Y1etnam. A strong c:aseans su btlentlv so that h . ll w-o cou < e made tl 1 t th t 1C) tasks not th e first, were the primarv purpose<; [(Jr th b a e ast 1'\VO . c. ' jtarv buildup 111 Eas t Asm. Arter all the UnHed States e su sequent Americ-an tl'd not 1111 t , ' 1 nee cl additional c rorces there to press me the r orth md1taril)' to release the erew. There wen: alre d . . . . bout 100~ 000 Am encan troops m East Asia. A<itarv builclup h . ,t lda ) .. . ; , owe" er, wou be a useful s1gnal for det en,e nce ~f further provocations and rea<>surance of its alk Unti l (or if) North Korea.s arc hives are opened up, we cannot know whether dete~ rence of further provocation wor k~d, because we do not know what additional plans the North had. \1\lh at we d.? kn~~ ~s th~t the reassurance function of the buildup did work: Sou th Kor ea kept tts c.hVJSJons m South Vietnam. Thus. America's militarY buildup had thre e purposes. Of those, one was achi eved, another ,..,as not, and th~ third we cannot be certain abou t. In sum , it is 'vvrong to draw the conclusion that the Pueblo case shows that forc e has little fungibility, even though military po turing appears not to have got ten the crew released. Baldwin's final example is equally problematic if the point is to how that militarY power has little fungibility. Y it is true that preventing an attack on ones homeland is es, a different task than win ning the hearts and minds of a people in a distant land. Presumably, however, the point of the example is to argu e that the latter task is not merely dille rent fi-om the former. but al.'io more difficult. 1f tltis is the assertion. it is unexceptionable: Compelling anothe r govemment to change it beha\'ior has always been an inherently more difficult task than dete rring a given govemment from attacking one's homelan d. Not only is inte rstate compellence more di:ffi tult than interstate deterrence, but intrastate com pell ence is more difficult tl1an interstate <:OmpeUence. Fore ing the adversmies in a civil war to lay dovm their arms and negotiate an end to t~~eir dispute is a notoriously difficult task, as the Chinese ch-l w :r in the 1940s. the.\ teti a namese civil war in the 1960s, and tl1e Bosnian civil 'var in the 1990s all too tragJ~l~ show. lt is an esp ecially d.ifficlJt task in a situation like Viet11am. where the .ou~Jde power's inte rnal aJJy faces an adversary that has the force of nationalism on tts tde. (Ho Chi Minh was Vie tnam 's greatest nationalist figure of the twentieth centwy and was widely recognized as such within Vieblam. ) It is hard to prevail in achil war wl~e~ the aclversaty rnonopolizes the appeal of nationalism. Equally important, ~s hard t0 . . 'ted prevail m a civil war witl1out resmt to 1 e. Tlle um States cotJd not a\e ore. . . WOn ' y b cl m tetnam by forc e alone hut it would bave a no chance at alJ to Wtn Wlthout lt. . h th N0 h ' t oug htful anal)'St of military power wou 11 tltere fore disaQ'fee ) \\lt. . e 0 c folio Jtary Wing propositions thclt can be teased out o f tIle c th example:al(1 tnJ1 t rour P<>wc k canno 9 r wor s bet ter fo r defense than for conquest: (- ) militarY .power onea1 e is . . gtlara t 1 (3) 1 an po" er on mht n ec padfication onc e con quest has take n P ace:

1\vo day<; earl ier thirty-o . . t an r. . ' J . . ' ne SfX;< I no. t cl into St(' ' ~~~H :.;ot WJl llln one-half rnile of th ;IaJ .':\ ortt K n . orea aaents . filtra e l'all 1 1'} . . . la -o ,n re ovc rcon tc li t c. J<;Jr tYIISSron w ts t"' k'll e prc>sdentiI pa. <.:e before p . ( I ey we ared that t 11rou gI lI tese two inci t1 r dent" anl I hresi dent park. The ljnite tes re ' u per aps 0th u Sta vas tryin g to d 1vC'rt \ merican mili tary n~so urt: , f ,. er!) to com e, :\orth 11orea" frJciently ner es ~ ak tJ1e South Korc'<m:-. ~u mus that tl rom \ 1etnam to Korca and n1 e _ 1evwouldh 1 . 1

sc>lated inCJC 11t


~ ns Rghting in Yielnatu hack home.'







l'r a.rra~'S ' se ' ., ,..., nclliO . re' I><' msu fr : >Ol to win hut !'0 ton. ~.) ' ll',\rl v all th nal. . isrn, not ouly will lo roe: ' e ot Itcr . too Is Cif. slatccra ,.t- monc.:y. rc>litic:tl .s kill l)r<)j)Hgundct, :111d \(1 on . In Slldr ,... < ' -..t~es .,. fl' . (' , tll" S ''llllt' insll f hCIC:' ll<.:\' <I " tiH olllt . 111~lrt tlll('n t" r1 mt rt ;tJY power s11 e r ton . ' " rat makes .it no 111orc. hut no less. fung ible tlran tlrcy arc.. . S < 1 o11r ( \ll I. o j' l3 Cl IIWlll . ('X'I111'f) leo; de 11 wnSIT;Il<' :Ill l 111 pOli , lllt fact aho11 t lllilit lt\' 't "'llll < cll ie' c rmuw thin gs. SliJ'{I~ . lht" t'i ,m llll i)OJ >t tant l> ,,,'t <> pOW(" I': USl'( I a] OIH' , I (;.-. o , : . , ,.. ' to J'Cll1<.>lll IJC'r, 1 l . . 't 01 l e tltlt js f>CCIIU:J r to Imlr taiy pm' <I .tloll l or tlrt~tt)rovcs tll< tt )U IS I c : . l Ill e r 'b'l't . n:-.t 1t l<L'; 1 1 1111g 1 I I y'> s r"h' not Jndccd. no srng le 1 , nt~ n<ll l of state~.:ra ft is l'' < . ll " . . . 'r .CICil t t0 Htt<lln et nv simtificant f(>reicm f?OIIcy. ob1ect 1\'('- a fact I shall term' ''1-lsk . Sl.l rr1 , ~" n . , . < . f:r.: . ncv.,s ' [I , e ., 1.(, {\.VO r eL-;o 11s for thrs. l1rs t. a sl <tt <.'sman mus t untJcipat< ti e msu ncw 11 < < 1 . takc n hv the states he rs t1 mg to mflu ence Tile)' y . cou ntera<:t 10ns tit:\t ,,'lJIIJ<' unde 1 . . . will atlCillj)t tO C ter his stratagems 'Nitl~ tho~ e or therr OWI1 ;.11tC'Y W OIIJl JJills e uif'f'c rt:nt t')ves of insh11 ments to o flset the one.~ he IS ~sm ~: and tl_1c~ wt ll a.t~e mpt to corupt>nsate for their weakness in Oil<' area w1th ll wtr str engt h Ill anot hc. 1. A well -prepared influence attempt therefor(' requires a multi-in ~trt~me~tal approac h to deal with ~le likel v counters to it. Second. < irnpo tt(tnt policy Itself has mal l)' fac-ets. A multilacUlY etel policy hy necessit)' requires "'any instrume nts to impl ell)c nt it. l~<~r both reasons, aiJ truly important matters require a statesman to mus ter s?ve ral , rf not all, the instruments at his disposal. rven though he may rely more heaVIly on som e than on others. l n sum. in state<.:raft no tool can stand alone. For miljtary power, then , as for the othe r instrume nts of state craft, Fungibility should not be equated with sufncie11 cy, and insufficiency shoutd not be eyuated "'ritb low fungibility. A given instJ1Jme nt can carry a state pa1 of the way to a give n goal, t even though it cannot carry the state all the way there. At one and the sam e time, an instmment of statec.:raf t can usefuJiy contribute to atta ining many goal s and yet by itself l:>e insufficient to attain any one of them. Thus, care ful consideration of B aldwin's examples demonstrates the foiJowing : (1) mil itary pow er vvas not sufBcicnt to achieve the defined ta-;k; (2) none of the other traditional policy i nstrur nents were sufficient either; and (3) milita1y power wa-; of som e value, eith er for th e defin ed task or for another task closely connected to it. What the examples did not demonstrate is that states are unable to transfer miJjtary power from one poli cy tac;,k to another. Indeed. to the contra1 Each showed that military power can be used for a vaJiety of y: tasks, even though it may not be sufficient, by itc;elf. to achieve any of them.

j1l f ' ''0\'( ' l"llll1 , 11()[ su ffidt lll l n <:Oil tpd a pop u IUl'C' t 0 ..l l <:C tll(.' f('trit illl.' r . .., t: lll; anu rr 1 l 1. ll clcte iT<'II<:< 1 1 .. l\'' t~o1 r 1 11 \ I) < ptll ,ntl' i" m11r<' clllllC: Il l l<l .'lllll ' ' > <. state. .~ tllis eXcll11pft .. tl l<' llls. lltt'rc is. lw w<'V<' I', a 1 '' ffth thtl slrould bt dr.t. so 1 < . . t , tr 11 1 1 c.: ivtl war ou I ht \ >wr stdv ol . . ~ .- ) \\'he n < o11tside pow l.ll


n lll r . thl' wrond <:1'-<' 1 arv p<> r i'i 'lc-II t'Oil'ii'-CIIIC ncts mr tlCllltnilih rv 11ilit w< I 1 11 . '. ~ 1 l<.rat ( y link I . ' 111' rh lite P "l'['fl <;l qf ..,tIc 'llgll ><' Ill n g a state\ b'lr ,. . . I l:C to et norHnilita.ry . JL'. \VI . . I < uflln . tSS' . I C.:' L'il' for( <'' >tS<'l ,lg<U.n'il f'orcc" in tlrc: se<; g<.I I' g I'VNa [r. nn IIat 1\'> lll' ln ._,..,c Ic frrs < ' ' . . ll r11 both (.'<1 \l'!' llllhtar~ p owcr h ('{;OJl i('S fllngemu, Iort< i., link(;(I With allrlii.Hr il ll . .. . trl'. . 1 . e ) Ct:awiC' rt . ,ss l tltcs trict ly u11hta'f) c C> tli<Hn. l <'xpl ainlrowc> ;HIt . tl .: ks prm1 cflccts net:~ otrtstt c I. pa 1 W<Jr and ill~tstrat< hotl w exarnp o . ith '
rrll ' . tlCI'S

c.:otl rtuYCl lll <'l 1


c:. an Ot lll:OllJC witlr l-.illnific.. t

2 13

spm -over Effects

'litu . . Amr y en<.:ottn.tcr. w ltdl l('r peac.:<'l'ul or l'o rcef11 1 }'l'elds . '.csu 1 l 1:tl c.:an he . . . . a entia! to the Jl1te racl rons and the outc:o 111 es that tak I . .t J <:onseqll . . I I " I ' .. 'I I .. . e ace rer . rs 11 1t wine: 1 tCI II t I H spr -ovm elft: <.:t .. 1s ({)() ol'tc' p - rn Ol 1 dotn ains " TlliS I ' .. . . , . . n1 orgo tl<>n." Militarv-lo..1 .tarv encountt"ts do no t ptodu<.:c only 111rht ar)' res 11 1 :rt' I .. 1 ts-- till I , .". I . . . . . . . --... tcs aJu wac;tc. an n u:~s ferted enc nm s s11h< ne d. dtt,lc.:ks pt t:VC . allt<'<; prot < 1 . . <;te 11 I cc c ' I . f . 'Ilted c. 1C\' aI'oo I . .l bout 1 ic.:al e ffects t 1<1 1' stgr nlcantly rnfl11 e:nc:e events in utJt ... . do : , .1rmg Jolit . c , f . . . .. " 1 ma1 ns. ,\ 111 tlarv powe r ach 1 c~~s mnc h o ItS fung rbtlr ty th ro11gh this effect: The political shc~<:k waves of a mrl1tar~ c.n<.:oun~('J' rcve rhc~atc beyond the military domain and cxt~.; nd into the othe r P? lrcy dom ams as well. fh t: exercise o[ succ.:cssful dctc: rren<:c, tornpellence, or ddc> nsc affects the overall political fram ework of relations between two states. Bc<.:aUSl' '.11 poli cy don1 ~\.ins arc situated within this ovcrarc:lting frame' w what happens m the latte r all<cts what happens in these domains. Spill -ov~ r ork, effects define with morC" prec ision why forc.: e acts akin to a gravitational ficlu . Aspill-ove r el'fr d can he understood cit her as a prcr cquisitc or a by-prcxluc:t. As a prer equ isite , the result produced by the ac.:t of force checking force creates something that is delibe rate and viewed as essential in orde r to reach a given outcome in anot her do main. As a by-produ ct, the encounte r produ~.:cs sonw thing in another domain that may b< heneRcial b11t is inc.:idcntal or evcu unin tend ed. Of course, what is by-produ ct nnd what is prerequisite hangs on what outcome arc valnC'd in that othe r domain. Tvvo examples will illustrate how the spill-ove r effect works and how it manHe sts itsel f eith er as a prerequisite or a by-produ ct.


If ~iJitary power is a versatile instn.unent of statecraft, then exactly how does _H achteve its fungibility? Wlmt are the paths through which jt can influence events 111 other domains?

r ..

rv,lttics In the first ,,.,,e :l t!ln' ~ mu1 ...... 1 power encounters miiHary power, but tcrorn thrs :

~l~re are two p~th s. The first is through the spill-over effects that .~~ilitar~ ~~ ~ 011 other JX>~c..y domains; the second, through the phenomenon of hnk~

Examples: Banking and Cold War Interdependence Tht first example has to do with banks; the second with rece nt history. Th( hanking Pxample dell lons tratc s th e role forc: e plays in solvency: the historical example, the role that U.S. mili tary pow N played in creating today's economic interdcpe nd~n c.:e. First, th(' hanki ug example. Begin wit!, this question , Why do we depo~t our llloney in a hank? T he answ er is wr put our mon er in a bank because we thrnk '~<: C' take it out -..vh~never W C want. We believe the mon er is dO there when W(' want rt. In short, we believe the bank to be solvent. . . 1 n . 1\ lnnk is . okent SoIv~nc . .' y rs usu ally thought of sole ly rn ec<~no~r c . _e ~:. : : . an' called. tn hecause rt ltas enough assets to meet its financJal Jrabrllttes rf tl.le. . . afi' Solvcn cy, howe . . . . C! ver, IS a (unc tion , not sunp Iy or nn.tne'"' but of .l)h\'Sit .al .s, ct} ...... . . . Abank' I , rt~ s so Vt rrcy U<.' t)ends on the- fact bot 1 t1 . < ts rxc(ed rts lrahrhtres. (.1 . 1at rts L'iSe 1 balan., 1 1 . <:c s reet is in the black) and that its asse ts are P1~,st~allv l'<.:llre (not .easl} . . . 1 d\'tolen ) PI . . . . . ')'Srcal secu nty is thcrpfore as . 0 . l)'urk olvenc' a.s rts rqUI unpo r1.u lt 1 '1 ' ,d, . ~~ , r t 1 wlwn we resr f 111 ' ' cvt~n though Wt:' ge1 lly take the form er ror gran <( wra

::! ] -4






willJed at wil l t} state . , 1en it . 11 e t lo1nest1c oreIcr. If the banks within a A tcuuiJkl '' 1 sl.l 1 1Ysically securs .ttzcn. wou Id not pu t their mone)' in them. ft sta e rna '( ' . "~ p1 . cl d c1 . 1 1 1 robbers and t(; b' nsinl1 its militarv power to deter an? e en agamst \\nut.- )e {'tJ t ai"e back the funds if a robbery takes pl.t((' (assuming th o ey lre ,d's etnll ~ o recovered) Through its use of its legitimate monopoly on ta e COI11JI>et ld fr . caua1 an 1e tUt . 1 0 11e S<.> i%urc If th t te seeks to neutralize the tJueat o 10rc1 > fr e state f . b ks . . u. e o rorce, a s a ucceeds in estabLishing the physical secunty o Its an . lt prod uces one of th . e r ,. . . , . two prerequisites reqmred tor a ban k's sohenC)' ln sum , in a welJ-ordered state. public fo~ce supp~e.sscs pnvate force. The effect of this suppression is to create a generalize~ sta.~)lht: ~hat sets the context within which all societal interactions take place. Th1s eHect spills over into nu me~ f.~stations, one o.f which is confidence ous other domains and produces many. mani about the physical security O.f banks. This confidence c~n. be v1cwed _as a by-product of the public suppression of private force, as a prerequ1s1te to bankmg solvency, or, more sensiblv, as both. A good J 1jstorical example of the spill-over effect of military powe r is the eco. nomic interdependence produced among the free world's economies during the Cold War. In a fundamental sense, this is the banking analogy writ large. The bank is the free world economies, tJ1e potential robber is th e Soviet Union, and the provider of physical safety is the United States. During the Cold War era, the United States used its militruy power to deter a Soviet attack on its major allies, the \.Vestern Europeans and the Japanese. American military power checked Soviet military power. This military-to-military encounter yielded a high degree of military security for America's allies, but it also produced several b)'-products, one of the most important of which was the creation of an open and interdependent economic order among the United States, \.Vestem Europe, and Japan. Today's era of economic interdependence is in no small part due to the exercise of American military power during the Cold \Var. A brief discussion wiU show how American military power helped create the economk interdependence from which much of today's world benefits. America's forty-year struggle with the Soviets facilitated economic integration \vithjn Western Europe and among \Vestern Europe, North America, and Japan. ?bviously, American mj~tary power wac; not the sole factor responsible for today's mterdependence among the major industrialized nations. Also crucial were the conversion of governments to Keynesian economics; their overwhelming desire to avoid th~ ~tastrophic experience of the Great Depression and the global war it ts b~ought m 1 wake; the lesson they learned from the 1930s about how noncooper ative,. ~ggar-thy-neighbor policies ultimately redound to the disadvantage of all; the willingness of the United States to underwrite the economic costs of setting up the ~-stem an~ of sustaining it for a time; the acceptance by its aJJies of the legitiof Amem:an leadership; the hard work of the peoples involved; and so 0 '~ portantas all these factors were, however we must re me mber where economiC ' openness fj rst began and where it subsequently flourished most: among the great powers that were allied with the United States against the Soviet Union. How, then, did the Soviet threat and the measures taken to counter it belp e produce the modern muac1 of economic interdependence among Arn ericas


allies? A1 Hl l1o'' cxae;lly, did America's 'l' . lt'i . 'I l ln Jt 'J rf] 1ere were ro. ml Jtary power <tnd , ()vers(Jas odustnG . I tl ur wavs. . jitar)' I)resen<.:c colltn Ill<. 1 F' st the spc:unt"' proYKkd hy the United Stat ~ . fl1 es created a 0I' . 1 I If , P Jtl<:al stahilitv . crucial to tiH on er y development of tral . ,, ations As l _ wng re1 k . I . that was wscusscd at tset of th1s ari1<. e. 111ar ets <.lo not exist in Jo--' vacuums th lh po 1 li . l l I .I . the ou ' rk best when ern )<"< c eu m po tlcal frameworks that vield . . . , ra er, ey 0 v American militan pm.C:;r deployt<.l in the Far East ' d pltdl<.:table expecta'' an on the E ' tions. uropcan <:on..1, t brought these 5lahlc expectations first, b)' . 1 prO\lwng the psycho! < 1 . I ' ' < hnen oo ca rance that the Eu ropcans and the Japanese need d t 1 . e o re nuld the 1. . . . reass U mse 'tS by contrn umg to prov1<.le them thereafter \ 'th . . d, second, a sense of safetY th t VI . . an ' a es c bled their econom1 energ1 to work their will. Indeed 1 Id ' 1ncrnher . we s 1ou re r x ~ro ena . was 'ormed was ps)re;hological, not rn1.1 , 1 t the prime reason l" A 1tary: to mak(; . . . ,. . " . . u1a e the Europeans feel secu1h enough agamst the Soviets so that th e). wouId. 1 the lave .d political w~ll to rchud t e ~selv~s economically. The initial purpose of ~ATO is the key to 1ts .( ~nu to the U . S.-Japan ~efense treaty's) long-lasting function: the cresland am1dst a turbulent international sea. ation of a pohbcally stable 1 Second, AmeJic~'s provisi?n of security to its allies in Europe and in the Far East dampen ed the11. resp~ct1ve wncerns about German and Japanese rnilitarv rearmament. The Un1ted States presence protected its allies not onk from th~ Soviets, but also from the Cer~ans .and ~e Japanese. Because Cerma~ and Japanese military power was contamed m alliances that the United States dominated. and especially because American troops were visibly present and literally within each nation, Germany's and Japan's neighbors, while they did not forget the horrors they suffered at the hands of these two during the Second World \\'ar. ne,ertheless, were not paralyzed from cooperating "ith them. The success of tile European Common Market owes as much to the presence of American military power on the continent of Europe as it does to the vision of men like ~lonnet. The same can be said for the Far East. America's military presence has helped "oil the waters" for Japan s economic dominance there. Third, Ametica's military presence helped to dampen cone-ems about disparities in relative economic growth and about vuJnerabilities inherent in interdepend~nce. both of which rue heightened in an open economk order. Freer trade ~enefit~ all nations, but not equally. The most eFficient benefit the most; and econ01m.c ef'fi<:Jencies can be turned to military effect. Interdependence brings dependencu:-s. all the greater the more states speciaUze econmnically. Unequal gain from tr~~e and trade rulitary eiTects. dependencies all too often historically have had adverse political and 1 11 ' t . . Th rough 1ts provision of military protection to tts a111es. t1 'um ed State mitie-ated le the security extemalities of interdependence and enabled the Ce~ums ~~d _the \\lth, all' es to their econom1c orbtts 1 ca1 ap J anese to 1 )ring their neighbors (Amelicas 1 ) m . ili't~n. conquest.or po Jti . . . o t th ed m cu.' u ose netghbors feming that German or Japanese e d I 'th t}1 economiC l)r Olni l_ . uummation would follow. Wicl1 the securitv issue ea t Wl ' . c th i!Thbors to swallow. f n rt 0 f b. 1 anc-e o the Germans and Japanese wac; eas1er LOT etr ne o 1- h . . . Fmally, Ame rica's military presence fostered a solictli1 . that came ' ' h ue lped . of sohdanh', m tu m, e 1 hein g partners agai nst a common enemy. T 1<.\t sense . the ine\itable develo th d ssaJY to O\ert'Ome d vill nece ' ''J . ., eaects of rnilitarv P e etermination and the goo ' bnng. The -spu -o,er u econo d sputes that inten.lependencJes rruc 1




'n '

I 111:.



Political stability. protectiOII from polcntial c .e rt11 <ll.) and Japaii(!!.C: militay !'( !Sur. .1 . u of'cc>ne;cms aiJollt relatve gauts and depcllclt.:Jcics .111 , 1tl genc.:e, 1IH1 ui:I II IJ'e!l lllln ' . . . . < ' 1 e . . so.H1 .1y- a11 of'Lhese were < tJdeJ by the A l rl< 'IIC<IIl Jt1rl11ary r1 resew . 1 an sense oI '-'' 111 Europe and the Far East.

'hould.nol lw wttkresli nwtc..;d, tholgh Uleyl :tre t ll ~~lllltl o l'll'' lll tl .and <Jllalltif'y l l I ( [ so I'd r'"' anc :~()()( W... 11l .t l .lcl ll. '( Ill d <;orrnnon. Surdv. howcv< tire S('nsc o r, I <t ; . , I ltl 11 se spi ll-over c:.:ffccts. J 111.1 y I " Ill <'d lo [1re . ' t"tll "(' hred 11111St haw laC . . 1 I' serve: a .. . I . 111 e lle iiiY f'lll lllnlts (JII 10\\ ill l1IC alli es ., 11 I I 011 u111 teJ front' agan1sl t w eo m ' c 11 IC ' IIC' < u to ITiait < .t a . . t Ullll(d Stai<"S, W()ll 11per.Ill I't their cconorn Jc . di Sf'llles to go. ( . I I . . I ell I I a . I' . I .,. . r 11 honndcd the lllC'vt la) (' I:COIIOIIIJ(' dl~[> lllus and llllJic d po 1l1<.a - Jnt' I OI I . 1 . . Pr< . . . ., l lin 11 into a dow11warc -sp1 ra HI}.!; t:L'O IHIIIII( nation 1 Vf'lll C'<I lJCnt 1 10111 esc..;.l <~ n li~ ''J. I

t'OO[HrttiO II a~rd liSt

. . . Lle politic..;< tl will lo ~w.ta ,. rrmotrdc 01 l]IC' SovJCl!i UJJ JI I . I t:n rl(:~s


Linkage Politics
The Sl.:c.:ond way for<:C' <xerts inlluen<.:c 011 other clom ~i ns o ! ' pc~lic.;y .is lhrcmglt lh~.; pCJwcr of li nkage politics. 1n politics, whether ~Jon1 cs_t1 c.; o ~ forc1 1s~~es arc us11 . gn, , ally linked I() Oil<' another. The link can .be e1the r func::tJon~tl or arb fdal. ]f' two issues an link<:cl f'uncliomtlly, then there 1.'-i a causal con nec::t1on bctwC;:en them: A changfl in (J JW produces a change in the othe~-. Tlw pric:e of th e ? oll ar (its excl1angc rate value) and the price of oil imports, for example, are fun c:tionally linked, because the global oil n1arket is priced in dollars. (Not only L hat, oil can o11ly be bought witl dollars. ) A decline iu tl1e value of the dollar will increase the cost of a given amount of oil imported lo the United Stales. Similarly, a rise in the value the dollar wj)l decrease the cost of' a given an1ount of imported oil. As long as oil remain.<; priwd in dollars, the functional tic between exchange rates and energy cannot he ddinked. Mor~over, as the oil-do.llar example illllstrates, functional linkages generally have corresponding spill-over eff(;cts. That is, weakness on one issue (a weaker dollar) produc.:es more weakness on the othe r (more money spent on energy irnports); aH<l strength on one (a stronger d<JIIar) produces greater strength on the other (cheaper euergy imports). Thus, function al linkages produc:e causal effects that either magnify a state's weakness or adclto its strength . Wheu two issues are linked artillciaJJy, th(;re is no causal connection between them. Achange in one does not automatically produce a change in tl1 other.J nstcad. e the two issues hecomc linked because a statcsmau has made a connettion where noue hc~f(m~ existed. Usually, hut not always, this wi ll be done to gain bargaining leverage. B making a link between two heretofore uneonnec.;te<.l issues, statesmen y try to briug ahout politically what is not produced fimctionaU They make a link in y. ordl:!r to t'fJJH} Jeusatc for W(!akncss on a gjven issue. Their method is to tic an issue where th~y are weak to au is~ue where U1ey are stroug. Tl1eir goaJ is to produc;~_. a


y 1 . . . )lea< of' ..1 slaH -. rP.all, l'" '' it. is IIC>I. ar1 lc ss nal or any !<.;ss err . I 1'" 1. 'Ill c:<<llll [>k cd :t 1arg ;ul nll g lll, kagc.; below l-C'lJvp ;1.} a rco;111l rovH e ' . I .(' . . . . I p A thN ftJJ a I) cll<JII:l llr :u1 1 ll'lal , J SSIIC linka(I(S j . v ! C l . . h . l:tV<: a <:rncial . , .,11 ,l!ysis ~md I t< cxfrc.rsc of sl at1~ j)OW cr IA f.<. cottsc -qllt;nc< f'nr V\ -can l1e I)O t11 1lw ..ssncs ttr< c<JIItiC'd<d , domains caniJot' b(; wl ll p11t .1k porrlt rtl<>n.: stronaly: , f3 CtiUSC 1 ' 0 Y I 1 1 (;df101 (; JI1 h c an not lw cklillk<'cl, tl ,c11 w< sho1dd fi(Jt vi< tl . n one anuth1~r. 1cy c..; 1f t 1 1 . w Jem 111 iw,l r r .~ 1'hemfon:, wy r:xp <UJatiOI J ol an O lc.:onw ill' . . ' ll . ..uon ron one 1 < U1CJll 1e 1. 1 . 1 l . . a gtvcn urJJ mun that i~ h- .. nc.:om['llct . r cL <;UJ only . t 110es on 111 l 1.11 c Olllc\111 w 1 .dways he 1 Ilt.l n on w I. . I I . (; , I I lOt<1 ri"'lt w own J sue lin kag<~s 11 n tt I I C'f'Xpanatory pow<~rof.tdcHna . Fl roug. n . k . . . . ... I su1n, JS. . . I ' m-rcstn<:tc;d analy'>is 1 c~g< s Ill P< t l<.:ll clr IT!i-1<<.: state w J3arg::unmg Ill ll ;scts IC!Ore r L I ~ I . . . k l'l .. . . . r. f'. ungm t lJdO thc:y rrll' ' llt I . wise h~. J ,JII ag< po I I(.;S I.S cl <ld () International pol'ti . I l'r 1-)' o i1CI . . . ' ea ue. We shou ld not eel othc rwls<'. Stalcs1nen ar c 011l to rnake the hest deals lt1 . . 1 _ t~xp . . . .. , 1 -! . . . ey can J)' cotnpenc;ati 11 , r weakness 111 one> ,u ca v. 1t 1 sl rc ngt t 1n others. Powerful <t 1 .. .. h _ ~ ,or . ., a c:s can ettcr c:nuauc ill n n IJ cMrlt1CIISatOI)' 1t 11 kages than can weak (JIIes. Th(y 'Lre strc Jl1Ker .m JJ10re areas < h tltcy tl1an they arc weak; Wll!W( jiJCnlly, r . .can more easily utiljze tl,,. 1 . t (J ' nakc " P r r t I1eu (Ie l <:Jlln the: weak ones Crett.pow .everage in t e . . .:IT a1 o slrcJI11' area<; n . . . ers are , so 1 1e:tter . able 10 shift assets ;;unong 1 ss11e are<-L<i "' order to bnild positions ofbargainingstrength when nee;ess~1y. They c~m , .for exarnp.l~, more <;a~ily ge~1erate military power when d1ey need to rn order to !1n k 1t to non mdJt< tasks. fherelore:, bec.:ause powerful states uy can link issues morC;: ea.<>ly than can weaker ones, c:an cornpensate lor dcf1c.:iencies better, can generate more rcsonrces and do so more c p1i<.:kly when needed, and <.:an shift assets around with greater ea'ic, how powerful a state is overall rernains an essential determinant to how successful it is in tcrnationally, inespec:tive of !tow weak it may be at any given moment on any specific issue in any pa1tic1 domain. [n sum. linkage dar politics enhances the adva11tages of being powerful anJ boosts the fungihility of f()rce by enabling it to cross domains....

Examples: Deficits, Petrodollars, and Oil Prices

Three ... brie f examples show the range of state goals that <.:an be served b) constructing such linkages. The first involves the relatiou between A111eric:a's large and continuing balance o payments deficits and its global uli'iaJJce system. Througho1.'t most of the ~~Id r War era, the Unitecl States ran an annual large balance of payments defic1ts. l listoric.:ally, no nation has been ahle tu buy 1 nore ahroad than il sells abroad (import more than it exports) in as huge a vc;lum~ and for as long a perio.d as ltas 11e Unitcd States. Then! were rnt:lny reasons why it W<L'i able to. rangi11g from the hqllidity that defi cit dollars provided, whicl1enabled world trade to gr~w, to ger~ cral <.:on{identc in the Ameriean ewnomy, ,...,hie:h c.:aused f(Jrcigners to nw~st theH dollar holdings i 11 tl1e Uuited States. Pmt of the reason tllat foreig, continue~ .~0 take A , . ll 1 lnen cas continuing How of c:1 ars 1 o 10wever, \\,.IS. a11 1np!Jc1t' ,f not err,lklt, , "L tradcoff 1 ' f . .. l()U 's ('defic..;it dollars). t1 le : 11 return for their acceptance o AmeHcan 0 "itcd States provided the largest holders tl1 (the Ger:rmms. t!le,Jar~~~se~ en1 and th0' Ja11u1 militcuy protection ag1unst t l1 . . e1 "' . l s) . temies Amenca rm ltary .. e11 ' ' . " sbt>rwth compcrlsateu {or its lac:: k o I"C! I (.]rscp j11 l e. 11 t') nsca (', 12 Alter. tl, e oil fJrice A so . I ,. (' "trodoJl<U"s. 'k " COl i( example involvt:s the recyc mg 0 pc G lf wer<: hI e . f' l aJJ .s () t lt~ 1970s, the ()}JEC producers, especr y t IlC p.-i'S.Ian u menH.x: . v

rnorc ~csJr~Jie outtome in the weak area either by th to do sowethn 1g. undesirable m tllc strong area, or by promising to do somethi ng beneficial ther~. 11 they C'4fl make the couuection stk:k, then the result of an arti ficial Hnkage 15 11 strengthening of a states overall position. Unlike a functional lin kag~. wJ1ere weak ~~.!><-.ogcts weakness and strength hegel<> strength, in an artif1cial L inkage, streng~h u~.~tas weakruss TiuJs' an art'{' 'a) 1 L~ is a bargaining conne<.:tion t l1at IS 111 ade JO . lCI' 111Kage


. ""'-I:


invc~t !' , \Vh ., JlCUl the). cotLld prontabk . 'I . t ere to put .K'<'tllnulclting more dollars t 1 l < "> utdis 1 ' ciaJ decision especta.J. . . r c ~ .' w 10 were . thost' dollar~ \\':l'> ~m important nnan ' tal t!\1den<:e that !!'t'llC'nl.tintt th<' lamest dollar surpluses. There ts ~trong Ctl <.:tll tlle atJdi~""'agreed ~park a sizable portion of tJ~err \.. s .. Treasu ry bills (T-bills) in part because of an explicit Amen.~ proposal k nm,, de a secuntv .i ~ . \.,;) tor um b re11a r tl1e Gulf"IJ ,""'Da,id Spiro notes: .B, tJ1e fou rth quart1r of 1977 Sauui :\.rabia accounted for twen~ percent of a.ll holdings of1 reasut~ untes and bon<.Is bv foreign central banks.'1o~ The Saurus also continued to agree to pric:e oiJ in doll.; ratJler than peg it to a basket of curr~ncies .. A1thouC1h ther~ ~,Lrc dear finan <.:iaJ incenthes for both Saudi decisions, the mcenti,es are not .uffi~tent to explain Saudi actions. The Kuwajtis, for example. never put as m<~~ of thetr petrodollars in the United States. nor as many in T-bills. as did the Saudis . .\loreo,er. an internal U.S. Treasurv studv concluded that the Saudis would ha,e done better if oil had been pegO'ed to a b~ket of currencies tJ1an to dollars. Indeed, OPEC had. c..l ecided in 19-r;) . I ~tJ octo price oil in such a basket. but never follo~ved 1rou~ 1. " ~ men~a s provision of security to tJ1e Saudis was an important, e,en tf not suffic1ent, mgredte nt in persuading the'm both to price oil in doUars and then to park the dollms in the United States. BotJ1 decisions were of considerable economic benefit to the United States. Parkin(T Saudi dollars in T-bills ga,e tJ1e American government 'access to a huge pool of foreiQll capital''; pricing oils in dollars meant that the United States 'cou]d print monev 0 to buy oil. "16 \1ilitary pov.:er bought economic benefits. A third example, again im-ol'<ing the Saudis. concerns tJ1e link between American militaiy protection and the price of oil. The Saudis ha,e a long-tenn economic interest that dictates moderation in oil prices. \\'ith a relatively small population and with the worlds largest pro,en oil reserves, their st:Iateg_v lies in maximizing re,enue from oil over tJ1e long terrn. It is therefore to their advantage to keep tJ1e price of oil hiah enough to earn profits, but not so high as to encourage imestment in aJtemati,e energy sources. Periodicall}; Saudi Arabia has faced considerable pressure from the price hawks within OPEC to push prices higher than its interest dictate . American military protection has strengthened Saudi willingness to resist tJ1e hawks. A specific instance of this interaction between U.S. protection and Saudi moderation, for example. occurred in the fall of 1980, with the onset of the Iran-Iraq war. Iraq attacked Iran in September, and the two countries proceec..led to bomb one another's oil facilities. The initial stages of the war removed about four million barrels of oil per day from ~orld markets and drove the price of oil to its highest l~e) e"er (842 per barrel).1' As part of their balancing strategy in the Gulf. tl1is tim~ the S~u~ ~ad allied themselves with I raq and, fearing Iranian retaliation agamst their ~il fi~J~. asked for American military intervention to deter Iranian ~cks on. thear oil fields and facilities. The United States responded by sendll1g A\\ACS au-craft to Saudi Arabia and by setting up a joint Saudi-American nanu task fort-e to guard against Iranian attacks on oil tankers in the Gulf. 18 In return. the Sau~ increased ~eir oil production from 9. 7 million barrels per day (m bd) to 103 which w~ ~e highest level it could sustain, and kept it there for the next ten months. Saudi actions had a c:onsiderable effec1: on oil prices, as Safran argues:

hand the. 1a s1

t lte

developmcnt ' '



. . t sent O1pnces s.rv..I l t)- n I ' c va"ng afL<~r tl 1 f: 11 f I l l .n ':'} < e('ision to cutp ~.1 . a o t e: rvuuc.:tion by 1 rnbd.tu

kand <l panic that had

this instan<.:e, American mjlit . . . b . a!) po" er alon~: ,., not t:C : nt to cause Sau. lt u<.:t1on~ to lower oiJ pric 'a) es, ut Jt was e . l S [j d . . su'uCie ssenti . tl ;c turbulent pcno< . aut E'C:JSJons on how mucI1 oiJ t hev w Id be<.:au e durl Jilg ,,., , ou pump were not ned solelv 1\ e<.<nomic: factors. True the S d. ....,11 . h au lS, ag , . : dete~" . amst t e de tres of the . hawks. wh1ch md ucled tJ1c Tranians had bee n pumpma mo 1 . ' . . nce P to lower od pn<:e . The Sauws had aJso violated the~ re Ol sm<.-e 19711 in order nrl'h 1979, howe,er, when they decided to cut oil od . . long-term strategy in pr uction b, J mbd . . . h ~ ftu~ ncre . Ppease Iran, a mo' e t at triggered a rapid 1 ase m oil pri Th. pn mank to a ision foUO\\ed a political c..lecision to move wplomati all . <.-es. IS pumpin~ dec c ' awa) fro th Li . . . th s 1 m e n1ted ater. however the (;{)-a'ct 1.thi th States Only a few. mon n e Saudi 1 f: . ull! '" '. mg . between an Amencan- .versus an Arab-onented strateO'V ~as reso1 .ru a c::o amil~ ,ed m o: . _ 1 cl . mplomise that le cl to a po I1tice:u reconciliation with the United St t a es; an th,s politicaJ . h l b c: ll . decision was 10 oweo y anot er to mcrease oiJ productio b , 1 b 1 . n \ m <' startma 1Juk h c: , o l979.zo Berore t he I ran- I raq war, t en, Saudi pumping d ..- . eusJons were affected bv . . . b . . . po)jticaJ calculations a out the1r secuntv, m which the strategJ<:. <:onnection \\1th the .I d . Amencans playe a promment role. If this was true in peacetime , sure'. tt was so . 1. m 1, wartime, too. The ffillJtary protectwn announced bv the :\menr...,n.- 0 n September """.., 30 ' .. . . r 1980 was a necessary conwtion for the Saudi mcrease in oil production th t 1011 ed 0\\ a in October. Agam, mili' tary power had bought an economic benefit. I~ sum. these ... exam.ples~ ... A~erica's ability to run deficits. petrodoUar r.ecycling, an? ~oderat~ oil pn~~s-all illustrate just how penClShe hargairuna ~kages ~~ m mte mational pohtics and specifically how military power can be linked politically to produce them. In all cases, military power was not sufficient. Without it, however, the United States could not have produced the fa,orable economic outcomes it achieved.
As in the other

<..''"' s. i~t

aJdwin first ork: 1. David Baldwin. Paradoxes of Power (New Y Blackwell. 1989), 151-02. B olitics :31, 1 developed his argument in his "Power Analysis and \\'orld Politics,.. \l orld P ou. (January 1979), 161-94, which is reprinted in ParadoxesofP -er. . ; Quot~s from BaJdwin, Paradvxes ofPou;er. 134-35, 135. and 136. respero,~lr ln frumess to Baldwin, these examples were not fully de,elopecl but c-onstSt of onl) a

PhysicalJy, the Saudi increase of 0.5 mlxl was hardly enough to make up for the sbo~~ caused by the war. PsychologicaDy, however, the Sauru action was crucial in preventing

sentence or two. 1\evertheless. the, are fair aame bec-ause BaJch,in used them as illus0 trations of his more generaJ point ~ut the limit to the utility of nwitary powe~ The fact that he did not de,elop them further led him astray, in my view. He was ~~o<J to show with them that miutary power is less effective than comm?nl~: ~ou<rht. 1 re.mterpret these examples to show how ,ersatile military power in fact LS. :~\either B~d\\m nor l.howe,er. can put a number on the fungibilif\ of military power, and 1rert~~ ~~ 1 "' th him that no political power resource be~ns to approach the degree of Jungtbilit} . -) . of mone " (Baldwm, Paradoxes of Pou:er. 135 . Baldwi Y 4 hnso The . n, Paradm:es of Power, 134. 135. 133. 5 R yndon Bames Jo . n.l or the facts and interpretation of this case. I have relied on L ork: J969 ~ ~ Y Holt. Rine 1art. Vanta p, . {- C\ ge omt: Perspectives of the Presidency. 1963-




.. , " p1 l('n S. Ka I P <lrt \' mstou. , ~" . . . . . . d -c rces os a Polilica/lll \ l nJI!I , :\ tl-b in~ton D , ' C.: ) 1 1 17] b' Force \\'itlu111t Wm : U.S. Anm ro -n - 0 ) -18 'Ill(1-1- I if,. ]:>'chard P. Steb ill S anc "'. UIH II ' t I, OCl/ltUmt. ()lt ~ 'U . 1: 13 roo,_..ugs..19 1 '~ ' . I -69 (:--Jew York: Simon & ~< u 1 1972). 292-: . ' 302; t\m('rirrm hm!IJ!.Il Rr>lotrons. 1968 _ 1111es I 11",le~ J968 ' 32-36. and the .\'eu: ),ork 1' . . _9 6. Johnson. 536. K,tplan. - Jo I1nson. 535; Bleclumtn and tJ th' t .... t He has written:" k lualh. am techn 1. tquc of rreeS Wl l IS pOI11 . f 11\\lll. o course, ag \ . lhldwin Econrn11ic . r . Ba c tl 1, 15olanon 1 rom tt. e others Sec David k r 111

:::: 3- Ban) M. 131echmu .., uI ()-I ) 3 5 38 t tnc1 " 39 1

rhe St!ategic Logic ofSuicide rerror1sm


. statccrat wor s poor)Princeton Umverstl) p ress. 1985) 1-t:3 < p . . ton ( nncc . rJt Emst Haas en:n though I am U'>illg it clirf<.rentJy than h r.. Stolecm c . ' 9 1 have borrowed th1s term nom .. . t lescribe tlte effects that cooperctholl 011 t>conornic matter s . 1 llcLSe o c 1tle p1 I' . 1 I . I . does lie uscc :c:. estem Europe could ha~Jle on t.lCIr IIJO .'~t:al . allJOI11s .. I le argued amo~<r the .states of W ons, c atters would spt over mtot 1C'II po llca re at1 indttce . ro . . 1 1 tl that coopcnttion cm economi 111 to lC po Jttca mtcgral!on of Westelll . 1 . and lead ultimatelv . t . ' . ~ e, reater cooperatton t1e1 g E . H .15. Beyond the Nation State: lm1ct1011a IS Ill ane1 Internationa[ Europ~. s.eC' (S~,s~ ..~ st'nford UniversitY Press, J964), 48. For111 aas's later assessment 1 " . Qrunmz.ttl/011 tallJOJU. "' see Emst. Haas, T Je 0 :Jsolescence of,Hegiona/ . 11-o ,,.r errects were rr . e. : . . . ' w h lOW cnecnve sp1 0f1 . '1', (Berke]e,.. fnstitute of IntematJonal Stuches, Umvers1ty of CaUfornia .' [ntegratum ,aeory


~~;:~;lcy is to be clistinguished from Liquidi~' A.ban.k solvent but not liquid. Liq-

uidity refers to the ability of a bank to meet alltts liab1ht1es upo.n demand. Most .banks are not able to do so if all the demands are c-alled at the same tlme. The reason 1s that man\' assel~ of any given bank are tied up in investments. that cannot be calle~ back on short notice but take time to comert into cash. The fw1coon of a central bank 1s to solve tJ1e liquidity problem of a nation's banking system by providing the liquidity in the short term in order to preYent runs on a bank. 11. As Gilpin put it: "Partially for economic reasons, but more importantly for politi~al and strategic ones. \Vestern Europe (primarily West Germany) and Japan agreed to fmance the Ame1ic:an b<1lance of payments deficit." See Robert Gilpin, U.S. Power and the M~tltin(lfional Cor7wration: The Political Economy of Direct Investment (New York: Basic Books, 1975), 154. 12. For this example, 1have relied exclusively on David Spiro's original and thorough research. See David E. Spiro, Hegemony Uubound: Petrorlollm- Recycling and the De-Legitimation ofAmericfln Power (lthaca: Gornell University Press, forthcoming). chap. 4. 13. Tbe quote is from an interview conducted by Spiro in Boston in 1984 with a American ambassador to the Middle East. See Spiro, 271. (All page references arc lor the manuscript versiou.) 14. Spiro, 261. 15. Spiro, 263-SG, 281-83.
16. Spiro, 259, 287.

ergin, The Pri:::e: The Epic Quest. for Oil, Money and Power (New York: Simon & 17. Daniel Y Schuster, 1992), 71 1. 18. Nadar Safran, Saudi Arabia: The Ceaseless Quest for Security (Ithaca: Cornell University Press. J988), 322, 410--U. 19. Safran,41 l. 20. Safnm, 23i.

/ l )itical objectives. I<or examp e, specta<.:ular suicide terrorist attacks 1 rcc:ently 1ave . .. . , 1.. . .. . , po been emp1oyed by P,t estin1an groups. m attempts to for<:e Israel to I)all(.1 thc West a . . on . . . . r an an Bank and Gaza, by the Ltberation T1gers of Tami] Eelam to compel tl1e Sn ... k . government to a<.:<.:ept cu1 lHH.lependenttlTamil homeland, and bv AI Qaeda t0 pressnrc r . d d States to w1t 1 ra"" rom 1e Saudi Arabian Peninsula. Moreover. such ue Unite 1 attacks are increasi ng both in tempo and location. Before the earlv 19ROs suicide terrorism w;:lS rare but not unknown. However, since the attack on ~1 e U.S. ~mbas~Y in Beimt in April 1983, there have been at least 188 separate suicide terrorist attach worldwide, in Lebanon, Israel, Sri Lanka, India, Pakistan. Afghanistan, Yemen. Turkey, Russia and the United States. The rate has increased from :31 in the 1930s, to 104 in the 1990s, to 53 in 2000-2001 alone. The tise of suicide terrorism i especially remarkable, given that the total number of terrorist incidents world,,ide fell during the period, from a peak of 666 in 1987 to a low of274 in 1998, ''~th 348 in 2.001. \~~1at accounts for the rise in suicide terrorism, especially. the sharp escalation from the 1990s om.vard? Although terrorism has long been part of international politics, we do not have good explanations for the growing phenomenon of suicide terrorism. Traditional studies of terrorism tend to treat suicide attack as one of manv ' tactics that terrorists use and so do not shed much light on the recent rise of thi type of attack. The small number of studies addressed exvU<:itly to suidde teiTOrism tend to focus on the irrationality of the act of suicide from the perspective of the ~ndividual attacker. As a result, tl1ey focus on individual motives ~ither re~igiot~s 1 ~doctiination (especially Ishunic Fundamentalism) or psychologJcal predispositions that might drive individual suicide bombers. The first-wave explanations of suicide terrorism were developed during the 1980s and were consistent with the data from that period However. as suicide attacks mounted from the 1990s onward, it has be<..'Orne increasingly e,ident that these initial explanations are insufficient to account for which indhiduals bec_-omc ..,.,tions are increas111al\' . suk'de tenonsts and, more importantly, why terronst orgmu..... 0 1 relying on this form of attack. First, although religious moth-es may matter. modem
1e . ..,

~ rrodst org<mizations are ill<.:reasi ngly relving on suicide att

ks to ac1. maJor 11eve ac

, 97' No. 3 (August 2003). pp. 34~361 . Copyright 2003 h) T r Amencan . ge .., " '.' , . eprill...,d WltI1 t Iw permission ol.Cam bm1 UIll"- 1..,,..., PrE-SS ation. H w

. m ., :\nunica 11 P> . , .. . . . 1 litiad Science ftqt;ict~. F'rom Ho! . Political Scicuct> As~ >ert A. Pape, 1 he Strat<'gtc Logtc of Swca.le Ter:ons j VoJ






uu 115 re<:, rnenlalislll. ~~- 1 . 't cl to Islamic Funda . ve e1 d , 111111 e s wc~tlc terronsm 1 not .... . \\' tem media but the world s lea l'r 11 ~ , . ,, terrorism. s . 1 . E) . . 1 . w t 1 most attention m es n~cru 1 ts from ac.'tmtllv the Liberation Tigers ofTamd ~ela.m (LIT ' ~~ gJO:~ \: ,. . onunant 1y H'u1 du rr:a 1niJ [:>opulation m north ern .mdI t.:rlS < n <,n Lank<t a'u1 . ., 1 t Ilt' prec 1 UitS st!Leninist eleme nts. 1 he LIT L a l)JJ( ,I('('OI1 Ior 75 0 r 'd Jogv I1 M <H XJ E as "'hose r eo 101111 1<. a mic Stti<:i<.l e the l86 suicide ten-orist attacks from 1980 to 200 1. 'vcn an_ tIHse ' s 'tl1 secuJar orientations accou nt..for about.a ll urd ,.ol .. . ath<.:k.. . . {' . 1 . . ~ s. attack groups \VI ers . . I. stucly of the personal ch.u.acteJrstics o . Slllll( < <lltac:k . 111.ty . 1 Second, e:tltl wug someda\ help identifY individuals ten-o~t organr7.ations are lrkel) to n<.: ruit ror th b putpo s~, the vast spread of suicide terronsm over the las.t two de<.:ad.cs -;uggestc; that tJ,ere may not be a single profile. Until recently, the leadmg exper ts 111 psy<..:lwlo~k: al J1(' profiles ~f suicide terrori sts characterized tJ1eJ'l1 a5 uneOUCi.lted. U 1l lj110ye u, SOC:iaiJy isolated, single men in tJ1eir late teens and early 2.0s . :"Jow ~ve know that sui<:itle terrOJists can be college educated or uneducated, ~am ed or srngle. men or wome n, socially isolated or integrated, from age 13 to ~ge 4t. In other words , although on I)' a tiny number of people become suicide terron ~ts. they wm~ from a broad cross se<.:tion oflifestyles, and it may be impossible to ptck them out m ad,ance. ln contrast to the first-wave explan ations, this article shows that suicid e terrorism follows a strategic logic. Even if many suicide attacke rs are irrational or fanatical, the leadership groups tl1at recruit and direct them are not. Viewed from lhe perspective of the terrorist organization, suicide attacks are designed to achieve specific political purposes: to coerce a target government to change policy, to mobilize additional recnJits and financial support. or both. Crenshaw has shown that terrorism is best understood in terms of its strategic function; the same is true for suicide terrorism. In essence, suicide terrorism is an extre me form of what Thomas ScheJling calls "the rationality of irrationality," in which an act that is irrational for individual attackers is meant to demonstrate credibility to a democratic audience that still more and greater attacks are sure to come . As such, mode rn suicide terrorism is analogous to instances of international coercion. For states, air power and economic sanctions are often the prefe rred coercive tools. For terrorist groups, suicide attacks are becoming the coercive instrument of choice. To examine the strategic logic of suicide terrorism, this mticl e collects the universe suicide terrorist attacks worldwide from 1980 to 2001, explains how terrorist organizations have assessed the effectiveness of these attacks, and evalu ates the limits on their coercive utility. Five principal findings foUow. First, suicide terrorism is strategic. The vast majority of suicide terrorist attacks are not isolated or random acts by indh~ dual fanatics but, rather, occur in cluste rs as part of a large r campaign by an organized group to achieve a specific political goal. Groups using suicide terro rism consistently announce specific politic.-al goals and stop suicide attacks when those goals have been fully or partially achieved. Second, the strategic logic of suicide terrorism is specifically designed to coerce modern democmcies to make significant concessions to national self-determination. In general, ~uicide terrorist campaigns seek to achieve spedfic territorial goals, most often .the Withdrawal of the target state's military forces from what the terroristc; see as national homeland. From Lebanon to Israel to Sri Lanka to Kashmir to Chechnya,

,, . de terrorist c,,. JMigll l'rorn I!JKO to 20C)J 1 . IJcrn w I I letS I I . t N sutC lgc< . r, 1 osc nwllt guH ''''" lrc11 lo establish or rtarutarn ~df-d. . 1y tcrrori<.l ev~; . . . . S 0 P w1 C:tc nnrnah :ncrr gro 111 unil)'~ h olltt I. ud hy <:om pcllrng an < ry. to v..rtIJdraw F 1 <Jr for urt rer, <~'<'l'V . I r (.'Otrn rrorist <.;amp tu~ll \lll<:f I~)HO ha.-, hetn hr(lr-t ' per t ' r,'' (;u ag<un.!>l 1;J.lctnat had 'a de e'C ronTI of gov<'l 'l llllt:nt . a~ t:J I soJC ocra . derT1 -rhrd duri ng lllC' past 20 y<ars , sui<.:ick terrorisIll Ila.\ L ll \ I l'l . l)(;:(' . ' ' I cac . lear11cJ tl1at it pays. Suicide t crr()n.sts <,oucrht t 1Ynsing terrorist)) have .,. I1ecause and Fre n<.:h Ill I 1tary rorc:<.:s to abandon 1 cb"nc . u,., ,., C0rnpC'I . an . ~ " m rn 19().> I ~. I ~ ,. . nerrc' A1 1 ' rl(, I .I I . Lebanon in J985, f!->raclr for<:(!') to yuil thc.; Cazt 'St np anu lr<;. rWest or<:c::~ In ' B k . _, k . leave an cl 1995 the Sn Lan an governmen t to create an.111cl ep<.;n(1 Tamil ~l 1n t e:nt .1 ~ ' J99 4 an . g90 on and the rnrkis 1 government to grant autono rny t o t1 K11 in a e:. re rm th . . 1- .1 . ' from l < L . os. Terro nst group s uru not ach reve their full obj'e<:tiVies 111 , ll t11ese: ea-.,(;\ a [T k l late 199 .1 se 'ever' in all but t .1e <.:e:L o m ey, the terrorist politi<:al <.:ause rnaue rnorc garns 1 H0\Y after the resort to su rcrut opera trons. ~1an rt had before .... . J':'ourth although mode rate su1C1de terrorism led to rnc)de rltc concesstcms < . . . . . r these more ambrtious S IIICJO~ terron sl campaigns arE: not likely to achitve ~till greater gains _an~l may well frul_comple tely. lu g~neral, suicide terrorism relies on threat to JnA1ct low to medlllm levels of punrshment on civilians . In otl1 ctrer : the wmstances, thjs level of punishment has rarely caust:d modern nation slate,~ to surrender signifkant politicaJ goals, partly because modern nation states are often willing to countenanc e high wsts !'or high interests and partly because mod<"'m nation states are often able to mitigate civilian costs by makin g economic and other adjustm ents. Suicide terro rism does not change a nation's willjngncss to trade high interests for high costs, but suicide attacks can over<:ome a country's efforts to u1itigate civilian <.:osts. Accordjngly, suicide terrorism may marginally increac;e the punishment that is in1lictetl and so make target nation s somewhat more ukcly to surrender modest goals , but it is unlikely to compel states to abandon important interests related to the physical security or national wealth of the state. \ ational govern ments have in fad responded aggressively to ambitious suicide terrorist campaigns in recent years, events which confir m these expectation . Finally, the most promising way to contain suicide terrorism i~ to redut.i:' terrorists' confid ence in their abUity to carry out such attacks on the target ~i~t) . States that face persistent suicid e terrorism should recognize that neither oflenst,e military action nor concessions alone are likely to do much good and should in\'est signincant resources in borde r deff:mses anu other means of homeland security.


..uted 10 d Most SU1<:1 e terrorism is undertake n as a strategiC e (-r rt di r"'.. towrrd achie\ing ' parti<:uIar po] goals; it is not simply the prod uct ot. ational indi"iduals or an h . . rrr tt1cal . 1 . exprcsS o f f:anatical hatreds. The mrun purpose of SUIc'de terron. sm ts to use t, e atl ion ' 1 threat 0f punishment to coerc e a target aovernment to clange pohc' especr th to . . .. . 't . terromb ne'' as eJI o t'ause d xh'b't tendeodes J! emO<.:ratic states to withdraw forces from tem O homela d TJ from 19(JO to 2001 e . t ' s I a1c 0 1e record of suicide terrorism n . ~ t with tl us stratey';c 0 "" 1n th t' e mmg, goals , and targets of attack that are con 15 en

~-~~~ - =

c-nJ'I .L

flU.. VoJ..__, ._, ,

.. - - -

- v._

I t:.I'O(QDr<'

o' hut not ,,;th imtlionaI or f. t'1caI be I1avior (1 ) timin r. w ,trf -:e ' '11Cid< atta<:k nna s (1(.xur in Of%Ulized . coherent camprugns. 11 ot as i olatcd c,r r o ' ' h. t'uned inc:i. dt~nt~. :2' na/ ionali.~t ![oals- uicide terrorist campru~L a ~' ~r ded at gain in . I t tl g con t m I o f " ' ta tc terron t ee as tLeir national homeland w rl to"' . specific:tll''at < tjediug foreiPn fcJrce from that territo~: and (3 > l~lrf!et \,. ~ ',m-RIJ sui~ide

two decades ha,e been rutned ,,+ O . lOt'racics wl, h . f . , lt: ' f'rom t1 terrons ts' po111t o vtew make tnore suitable targets 1e

. . t :::. a, - tl e hst t erron <.:ampal'"'' ' n1 1 '

Defining Suicide Terrorism

Terrorism imolves the use of \iolence b~ an organization o~1er than a national ao,'emment to eau e intimidation or fear amon~ a taraet audience. Although one: Z-ould broaden the definition of terrorism so as to include the action of a national govemment to c:ause terror among an opposing_ population. adopting s~ch a broad definition would wstract attention from what poucy makers would most lJke to kl1ow: how to combat the threat posed by subnational groups to state security. Flllthcr, it c ;ould also create analytic confusion. Terrorist organiz..~~ions and state govemments have different levels of resources, [ace different kinds of incenti\'eS, and are susceptible to different t)pes of pressures. Accordingly, the determinants of their behavior are not likely to be the same and. thus. require separate theoretical investigations. In gen~raJ, terrorism has two pwposes-to gain . upp01ters and to coerce opponento;. .\lfost terrorism seeks both goals to some extent, often ailning to affect enemv ' c;alculations while sim ultaneously mobilizing support for the terrorists cause and, in some cases, even gaining cU1 edge over rival groups in the same social movement. However, there are tracle-offs between these objectives and terr01ists can strike various balances between them. These choices represent different fonns of terrorism, the most important of which are demonstrative, destructive, and suicide terrorism. Denwn.s1ratit;e terrorism is directed mainly at gaining publicity, for any or all oftluee reasons: to recruit more activists, to gain atten tion to grievances from soft~ Jjners on the other side, and to gain attention from third parties who might exert pressure on the other side. Croups that emphasize ordinary, demonstrative terrorism include the Orange Volunteers (Northern Ireland), National Liberation Arm)' (Columbia), and Red Brigades (Italy). Hostage talong. airline hijacking, and explosions announced in advance are generally intended to use the possibility of harm to bring issues to the atten tion of the target audience. In these cases, terrorists often avoid doing serious harm so as not to undermine sympathy for the political cause. Brian Jenkins captures the essence of demonstrative terrorism with his well-known remark, 'Terrorists want a lot of people watching, not a lot of people dead. Destructive terrorism is more aggressive, seeking to coerce opponents as well as mobilize support for the cause. Destructive terrorists seek to inflict real harm on members of the target audience at the risk of losing sympathy for tl1eir c:ause. Exa(:tly huw groups strike the balance between harm and sympathy depends on ~e nature ~f the political goal. For instance, the Baader-Meinhoft group selecj tive)y assassmated rich German industrialists which alienated certain segments 0 German society but not others. Palestinian t~rrorists in the 1970s often sough~ to kiJJ ac; many lsrctelis as possible, fully alienating Jewish society but still evoking

cl SUf\1\'e a m ison an often emlJio,s a method of attack that requ,res tbe attack d ' . . 51 eed (such as plantmg a <:ar born b. wearing a su cl ers eath m ortler to succ . . JCI e vest or ram . . into a building). ln essence, a suicide terrorist kill h ' nuna an cur. . d pIane iciJls himse lf. 1n pnnc1pI s e. suici e terrorists coultl b ot ers rat the sarn c ti me d th the a 1, . cl . ses or could be UJnttc to targeted assassinations 1e use . 10r demonstrati\'e . . Purpo f' k . I kill . n practice. howe\'e . cide terrorists o ten see Simp y to the largest number of o le r. su~. ximizes the coerci"e Je,erage that c:an be gained from terr ~ P . .does:\lthough tlu ma; I b . f r on lt arP.atest cost tot 1e a.s1 o support 10r the terrorist cause \taxi m. . h o at the s . ~; --1= h . I . . mlZlna t e number . f enemy killed auenates t lose m t 1e target audience who might be svrnpathetic to o I il . . . ... tbe teJ TOJists cause, w 1 e t 1e act of su1c1de creates a debate and oft~u loss of su _ -fals .p port among moderate1segments of the terrorists' communitv even 1 o attracting 1 . . support among radica e ements. Thus, while coercion is an element in all terr . . . onsm. coercion is the paramount objective o f suicide terrorism.

" . 0 C loll'l . -;de terron slll :s ti1e most acrgressive form arC"tw;ts. . S c f u1 . o erronsm . 1 . t the expense ol CJSillg suppo rt among the terra . , . purs,un~ c.-oercion e,e~ a 15 hes il suicide terrorist is that the attacker doe nst own tornmunitv. \\11at d'st'ngu s not expect t0 . 1

, from ~l u'h .1 <:CJn muni tie . Other g10 In ilt ups t 1 e 1 . 1al 11 P ~'~ include the Jr:o;h Hepuhlic:an Arm\' the R . . mp tastze destn1cti . rronsJTl ' I . ,, e\ o1 utJona ,\ '~ te b1 (FARC . at.'l t 11! umcteen th-centurv :\n I . ~armed Forc-e of



The Coercive Logic of Suicide Terrorism

At its core, suicide terrorism is a strategy of coercion, a means to compel a target government to change policy. The central logic of this strategy is simple: Suicide terrorism attempts to inflict enough pain on the opposing society to oYetwbelm their interest in resisting the terrorists demands and, so, to cause either the uo,emm to concede or the population to revolt against the ao,emment. The common ent feature of all suicide terrorist campaigns is that they inflict punishment on the opposing society, either directly by killing civilians or indirectly by killing militruy personnel in circumstances that cannot lead to meanjnaful battlefield \ictory. As we shall see, suicide terrorism is rarely a one time e\ent but often occurs in a series of suicide attacks. As such, suicide terrorism generates coercive leveraae both from the immediate panic associated vvith each attack and from the risk of chilian punishment in the future. Suicide terrorism does not occur in the same circumstances as military ~er cion used by states, and these stmctural differences help to e:\'}Jiain the logic of th_e strategy. In virtually all instances of international military coercion, the coercer 15 the stronger state and the target is the weaker state; th e eoercer would . 1 l'ke1 be deterred or simply unable erations 1 Y to execute the th reateneu nn'liteary 0 ? In tll ese ctrcumstances c:oercers . . h have a cl1mce bel...reen hl~''0 main coerCl\'e strate ' or gies, punishment and denial. Punishment seeks to coerce bYnust~g the costslis . risks t0 th 1. ,.. ' e target society to a level that overwtle 1 the 'cuue of the mterests U1 t ms . . 1. pute D . erual seeks to coerce bv demonstrating to tbe target state that 1t s1mp \('nish . t a can61 not . Win the dispute regardless"of its level ofefJiort. . d therefore g 1ting oth n .I=san IS po I l th bilitr to conquer e w mt ess-for example because the coercer 1as e a b t the" puted t . , . 'ti'all , rely on pums men . . erntory. Hence, although coercers ma~ U1l )

states and some have been strong enough to wage guernll a nllll~<:ll y campaigns as well as terrOJism, none have been strong enough to have senous prospects of achieving their poutical goals by conquest. The suicide len:olist group with the most significant militaJ} capacity l~a'i been the LITE.' hut_ 1t h_as not had a real prospect of controU ing the whole o( the homeland that tt clan11s , mcluding Eastern <Uld Northern Provinces of Sri Lanka. As a result, the only coercive strategy available to suicide terrorists is punishment. Although the element of''suicide'' is novel and the pain inflicte d on civilians is often spectacular and gruesome, the heart of' the strategy of suicide terrorism is the same as the coercive logic used by states when they employ ai1 power or economic sanctions to punish an adversary: to cause mounting civilian costs to overwhelm the target state's interest in the issue in dispute and so to cause it to concede the terrmists' political demands. What creates the coercive leverage is not so much actual damage as the expectation of future damage. Targets may be economic or political, miJitruy or civilian , but in aUcases the main task is less to destroy the spe cific targets than to convince the opposing society that they are vulnerable to more attacks in the future. These features also make suicide terrorism convenient for retaliation, a tit-for-tat interaction that generally occurs between terrorists and the defending government (Crenshaw 1981)... . Suicide terrorists' willingness to die magnifies the coercive effects of punishment in three ways. First, suicide attacks rue generally more destru ctive than other terrorist attacks. An attacker who is willing to die is much more likely to accomplish the mission and to cause maximum damage to the target. Suicide attackers can conceal w~pons on their own bodies and make last-minute adjustments more easily than ordinary terrorists. They are also better able to infiltrate heavily guarded targets because they do not need escape plans or rescue teams. Suicide attackers are also able t~ use ~rtain especially destructive tactics such as wearing "suicide vests'' and r~mm g veludes into targets. The 188 suicide terrmist attacks fiom 1980 to 200l ~edan average of 13 people each, not counting the unusuaU large numbe r of fataly Ities on ~eptember 11 and also not counting the attackers themselves. DuriJlg UH:l same penod, there were about 4,155 total terrorist incide nts worldwide, which killed 3 207 ~le (also excluding September 11), or less than one person per incident. Overall , from l980 to 2001, suicide attacks amount to 3% of all terrorist attacks but accoun t for 48% of total cl th d . ea s ue to terronsm, agam exclud ing Septem bet 11

t.'tlt.n h~t\ e lhe resour ces to create <1 (onnidabk !! treat to clctt tht oppon nt . in b~lttlt"' :md, if necessaJY. to achicvc a brute force milihu; ' 1 i.c ~, if lhC:' larg~~c:~ory t.'rnme nt ~''-'fuse s to thu~ge its hehador. Tlte Allied boJHhi 11 L. uf CC'rlllany in w ~v0 . . . ' . l"t' \\.u ll AmelicanbombingofNorth Vletnamln 1979 Gm< I ( 'l.llto n atlac:ks t " 11d -. ' 'gcltnst lruq in 1991 all Rt this pattem. Suicide terrorisn1 (and terrorism in genera l). occurs Hnder the revc rse stn, .. tuntl conditions. In suicide terrorism, tbe c~er~e.'."' w~~tk~~. actor and the targ~t is the stronger. Al though s?me _element~.?J th ~. sltuab~n ~ ( 111 _<~m. tl~~ ~arne, flipping the stronger and weake r stdes m a coeJclve_c.hspute has 1 <_lr,unatJc change on the relative feasibility of pu nishme nt and dental. f 11 th ese Cl rct~ 11 lstanc:es. denial io,; imi)OSsible because miJHaJ" conquest is ruled out by relat1ve weakn ess. Eve ' I . I) though some groups using suicide tenorism IHtvc receive~t1mp~>~ant su pport froll)

- '-'\J\UKJSM


suiddt : t l ,MC an especially c:onVJnc: secoOd, . .. . . mg way to si 1 h . . ain to wnw ' ~,. '>lll<:tde 1tst:lf 1s a costly l>. a] gna t e hkelil1ood of m~rek~rs coukl not !t..\ 1)ctn tleterrcd by a threat or':stl' ~.:; t~a~ suggtsts that the attac onsor suicidt' ai:ta('k~ can also deliberate[)' hy aliation. Organizations 0 re estratc . . the : tl1at sp the death o f' '' sw<.:HJe attacker to increase furth d ~lrcumstances ilfO\Jnks This can he callccl the " art of martyrdom." The er e~ctations of future: rtac . f .. . . a . ..;005 on lhe basrs o rchgtous or ideolooical,noti more swc1de t erron-;ts JUShfv tl tl eJr acu . b' ves lat match tl b I. r: ' . 1 ader national cmnm t mity, tire more the status of terrorist ma r . 1e e u~rs of a bro ore plausible it Le<.:nmcs that others wiU foUow in their f, trty 5 IS e~~ated. and t1 m 1e .. . nizations common 1 <.:u 1 . y hvate ..sacnficial myths" that.oo steps. Smctde terror. ist orga. k . . . me1 1 e1 Uue aborate t f bols and rituals to mar < mdiv1dual attacker's death m 'b . se so sym k ' r "1" 1 r a . . n Suicide altac.: ers am11 a so o ten receive mateJiaas e <.:ontnb ution to the es lI cl th f ntl tJO . . . .1 f war S 0 rist orgamzat1ons anu rom other supporters. As a result tl1e rt f rom the terro th . . a o martvrdom elicits popular supp01~ from e t~rronsts community, reducing the moral b~klash that suicide attacks rmght otherw1se produce, and so establishes the foundation for credible signals of more atta<.:ks to come. Third, suicide terrorist organizations are better positioned than other terrorists to increase e~ectations about escala~ng future_costs by deliberately ,iolating nonns in the use of VJolen~e. Th~~ can do tlus by crossmg thresholds of damage, by breaching taboos wncerrung legJhmate targets , and by broadeningrecmitment to confound ex'Pectations about limits on the number of possible terrorists. The element of suicide itself helps increase tl1e credibility of future attacks , because it suggests that attackers cannot be deterred. Altl1ough tl1e capture aJ'ld conviction of Timothy McVeigh crave reason for some confidence that others with similar political views micrbt be deterred. the deaths of the Septem ber 11 hijacke rs did not, because .Americans "vvould have to ex'Pec1: that future Al Qaeda attackers would be equally willing to die.

The Record of Suicide Terrorism, 1980 to 2001

To characterize the nature of suicide terrOJism, tllis study identified eve':suicide terrorist attack from 1980 to 2001 that could be found in Lexis l\exis' on-line database of world news media. Examination of the universe shows that suicide terrorism hac; three properties that are consistent with the above strategic logic but not with irrational or fanatical behavior: (1) timing -nearl y all suicide attacks occur in organized, coh~r~nt campaigns, not as isolated or randomly timed incidents; (2) nationalist goals- stuctd.e terroris t campaigns are directe d at gaining conhol of what the ter,rorists see as .theJr national homeland territOJ y, specifically at ejecting foreign forces tram that temtOt)': and (3) target selection-all suicide terrorist campaigns in the last t\\o decades h_a,~ been aimed at democracies which make more suitable trugets from the terronsts ' pomt of' VJew. Nationalist movements that face non clemocrabc opponents have not resoJted to suicide attack as a mecms of coercion.

Timing. As 1: bl 'd a e l indkates there have been 188 separate sutci e t erro Jist attacks bet\veen 1980 and 2001. or' tl1ese 179 or 95%, were parts of orgru1ized. coheredisn' camt h paig , , ns, w ile only nine were isolated or random events. Seven separate putes




I" V~''-'-

TABLE I SUICIDE TERRORJST CAMPAIGNS, I 98G-200 I t's No cff-~ --::--- ..... N~o~. Terrorist Tenons Allad.s Date Ki lled Target


--~------- ----.:.._ completed Campaigns

1. Apr-Dec




Hezboliah Hezbollah Hezbollah

1983 2. Nov 1983Apr 1985 3. June 1985June 1986

4. July 199GNov 1994 S.Apr 1995Oct2000 6. Apr 1994


u.s.!France out of Lebanon Israel out of Lebanon Israel out of Lebanon security zone Sri Lanka accept Tamil state Sri Lanka accept Tamil state Israel out of Palestine Israel out of Palestine



Complete Withdrawal P artial Withdrawal No change




Negotiations No change Partial Withdrawal Gaza Panial withdrawal from West Bank No change Hamas leader released No change No change


629 15

7. Od 1994Aug 1995



8. Feb-Mar 1996 9. Mar-Sept 1997 10. June-Od 1996 11 . Mar-Aug 1999

Hamas Hamas PKK PKK

Retaliation for Israeli assassination Israel out of Palestine Turkey accept Kurd autonomy Turkey release jailed leader





Ongojng Campaigns, as of
December 200 I
12. 199613.2000Rebels 14. 2000Rebels 15. 200116. 2000-

AI Qaeda Chechnen Kashmir LTIE Several

Total incidents No. of


U.S. out of Saudi Peninsula Russia out of Chechnya India out ef Kashmir Sri Lanka accept Tamil state Israel out of Palestine




53 45

3 6

d Ie tO suicidl 1 t !Sl C~u npaiJ;ns: thc t)resen. 0 f Allltn. I1<1,e Lehanon -.1 't t c><.:<.: 11pal1on o f \Vest Bank ,, <:e (' c.:an and French r cc>S I 11 k . nnt1 .>Ct7.<t. the . d ,or .1 regions of ~_ 1. J ..111 a, the: tnucpcndent:e of tl , K ' . tn ependc:nce: of 1e Tatn' . ll . occupatt.OII (l r c1 1mya, lntltan O<.:<.:upatiOJl KK llrd,sh regHm 0 rTllrk(:y we r as I1mt.r . , tl . d 1 ' 1ussranrican forces on l IH' Satl th. Aral>ian Peninsula o ll h an tc presence rAl11e distinct campatgns. IJccausc in certain disp 0tV<.:ra ' owever. tI1erc have .. o 16 b en . . u 1C: . e d operation ~ Oltt or more lllll<:s either in .re" es t1 terro nstc; eIccted to uspen f I .ponc;c: to C:On <;c , r 5 ., 50 ns. Eleven o t tC' campaigns have ended and fi . ther rec... \ e were on sston~ orf tor . 0 d 1 2001 The atta<.:ks comprising each campaign '""' : gorng as o the o "'re organized b t en. st group (or, somctinrcs. a set of e:ooperating group . . same ten on ) l . ) s ,ts m t1e onf1otna 1 .t ntifada" in l snte 1 a estme . <.:1 !I 11stcred in time publicall . . d. :> '"' sec:onu 1 'J' . .1 1 1di cl . ' YJUSti 11e m terms of a specified politi:e:u goa , a1 . ree:te agrunst targets related to that goal. :c. The most 1mportant mdicator of the strategic orientation of ... d . . . . f' .. .. . e itning of the suspens1 o campatgns, whJch most oft<::n SUJC:t . terronsts rs on occ 1 d tl1e t . . . I "' 1" . . 1 . . .. urs a mc dectswn by edc e1s o the terronst organ 17_at 1ons that r rtl )ase on ks stf<ateo . . u 1er attac be counteqJrotluctJve to the1r e:oer<.:ive purposes -for in~n"e . wOuld . .. m response to full or par~1al c~nccs~10ns b~ the. ta.rget state t~ the terrorists' political ~oals. Such suspensions ar: often ~cc~mpa n1ed by public explanations that justify tlle decision to opt for a cease-fire. Further, the terrorist organizations discipline is usually fairly good: althou~h there are exceptions, sue:h announced cease-fires usually do stick for a period of months at least, normally until the terrorist leaders take a new strategic decision to resume in pursuit of goals not achieYed in the earUer campaign. This pattern indicates that both te1Torist leaders and their recruits are sensitive to the coercive value of the attacks. As an example of' a suicide campaign , consider Hamas's suicide attacks in l99.5 to compel Israel to withdraw from tovms in the West Bank. Harnas leaders tleliberately withheld attacking during the spring and early summer in order to give PLO negotiations with Israel an opportunity to finalize a withdrawal. However, when in early July, Hamas leaders came to believe that Israel was backsliding and dela~incr withdrawal, Hamas launched a series of suicide attacks. Israel accelerated the pace of its withdrawal, after which Hamas ended the campaign .... If suicide terrorism were mainly irrational or even disorganized. we wOtJd expect a much diffe rent pattern in which either political goals were not art~<.:ulated (e.g., references in news reports to "rogue" attacks) or the stated goals.'a.ned considerably even within the same conflict. We would also expect the timmg ~o be either random or, perhaps, event-driven, in response to particularly pro,ocath:e or infuriating actions by the other side, but little if at all relat~d to the progress of negotiatio ns over issues in dispute that the terrotists want to mfluence.
J..Jl<:l " ,


campaigns No. isolated

Source: Robert Pape '11le u . . mOlicago, l}'pescrjj>t. mverse of SUicide Terrorist Attacks worldwide, 198D-200 1.. universitY


Nationalist Goals. . SW<:tde terrorism is a high-cos t strategy. one tl1at ':vou11 011 1. make stratemc sense for l ) o . agroup when h1gh mterests are at stake and , even tllen, . . a last re ort. The reason lS . . . as U YnPnse lat sutC'ide terrmism maximizes coercive le"en1ge at the e,.r- . of support among . h th e the tenonsts own commun 1 . ity and so can be sust,Uile<I over. time on ~\ w en Tbe 'ts er alread the potential poo1o1 recru1 YeXISts a high degree of commitment among d f ts J1 omeiand Jnost tmportant goal that a commun ity can have lS tl1e mdepen ence o 1








Homeland Region Dispu te

Lebanon. 1983-86 West Bank! Gaza. 1994Tamils in Sri Lanka. 1990Kurds in Turkey, 1990s Chechnya, 2000Kashmir, 2000Saudi Peninsula. 1996-


-2oo ,

U.S./F/IDF military presence IDF military presence SL military presence Turkey military presence Russia military presence Indian military presence u.s. military presence

Terrorist Goal
U.S./F/IDF w1thdravval IDF withdrawal SL withdrawal Turkey withdrawal Russian withdrawal Indian withdrawal U.S. withdrawal

iar gct a Democracy..,

Ye> Yes Yes (1950) Yes (1 983) Yes (1993) Yes Yes

(population , property. and way oflife_) from foreign inAue n~e or co~1trol: As a result, a strc~.tegy of suicide terrorism is most I~ kely to he us~d to c~dne ,e nationalist g~als. su<.:h as gaining control of what the te:ronsts see a<;. the1r national homeland ternto ry and expelling foreign military forces from that terntory. In fac.t, every suicide campaign from 1980 to 2001 has had as a major objectiveor as its central objective c-oercing a foreign govem ment that has military forces in what they see as their homeland to take those forces out. Table 2 summalizes the disputes that have engendered suicide terrmist campaigns. Since 1980, there has not been a suicide terrorist campaign directed mainly against domestic opponents or against foreign opponents who did not have military forces .in the terrmists homeland. Although attacks against civilians are often the most salient to Western observers, actually every suicide terrorist campaign in the past two decades has included attacks directly against the foreign military forces in d1e count ry, and most have been waged by guerrilla organizations that also use more conventional methods of attack against those forces. Even A1 Qaeda fits this pattem. Although Saudi Arabia is not unde r Ameiican military occupation per se and the terrorists have political objectives against the Saudi regime and others, one major objective of Al Qaeda is the expulsion of U.S. troops from the Saudi Peninsula and there have been attacks by terrorists loyal to Osama Bin Laden against American troops in Saudi Arabi a. To be sure, tl1ere is a major debate among Islamists over the morality of suicide attacks, but within Sauru Arabia there is little de?ate.over AI Qaeda's objection to American forces in the region and over 95% of Saudi soc1ety reportedly agrees with Bin Laden on this matte r. . Still, ~ven if s~icid e terrorism follows a strategic logic, could some suicide ter:o~ lSt camprugns be trrational in the sense that they are being waged for unrealis~c goals? .The an~er is tha~ some suicide terrorist groups have not been .real.istic 10 ~xpe<.~mg the full concesstons demanded of the targe t, but this is normal for disputes mvolvmg overlapping nationalist claims and even for coercive attempts in general

the arnbitiora~ ,r noris t leaders art: realic;ti .. {i<~ther, . two oJists, IJO Jl,, l' < r n:,. .f not therr methods <.: rn r. other senses. First . . I t de ten arc 01ten , sutct rs realize; thLJ g~ nerally reflect quite tom .more mainstream th bserve I f" t1 . mon strrught f, an o f d tennination c .L rn o ICtr comm unity. Seco d th - orward nationalist s~l - et suprJoJ for tlwir policy goals versus the h n t. ese groups often havt: si(Tt ruficafl . < e same as t IJU'>~ oI" otl1er nationalists .vitl rgetl state ' goals that are t\nic:alo . lv rnueh th . d " . . , ' the terron sts an more moderate" leadenn 1etr corn muntty. DtiTerences . all between ain level of violence and-s omet imes -the rs usu y coneern th e usefulness legt 0 f a cert< tmacyof . ets besides foreign troops in the country, suth as' att ks. 1ttaclG ngadditiona] ' tar~ st third parties and civiLans. Thus, it is not that thac 10 ~ther <.:ountries or aga~n I and then seek ot ters , support. Rather, the terrorist e terronsts pursue radical . goals th vthem 1 ;,. societies who are e most optimistic about thesarestmp1 embers of. ' u1e.u use fulness of v 1 r achieving goals ~hat many, and .often most, support. 'IO ence lO r The behav1or of Hamas tllustrates the point. Hamas t . h raeli retaliation t hat 1 been tostly for Palestinians erronsm as .provoked 1as whl [s . . 1 f b t . ' Ie appare ntly unrealJStJC-goa o a o 1shmg the state of Israel. Althoupursmng etheh ro state establishlnC1 an Arab b d in all of "b.istoric Palestine" may be poo g PPal P .~c; of ~ Id . bl . . . r, most estimans agree that 1t wou e esrra e tf poss1b l~. Hamas's terrorist violence was in fact car~~y calculated ~nd ~ntro~ed. In 1994, as its first suicide c-ampaign was begmrung, Ham~ leade~s e:'Plamed that martyrdom operations" would be used to achjeve intermediate objectives, such as Israeli withdrawal from the 'Nest Bank and Gaza, while the final objective of creating an Islamic state from the Jordan lliver to the Mediterranean may require other forms of armed resistance.



Democracies as the Targets.

Suicide terrotism is more likely to be employed against states \\itl1 democratic political systems than authoritarian govemments for several reasons. First, democracies are often thought to be especially vulnerable to coercive punishment. Domestic critics and international rivals , as well as terrorists, often view democracies as "soft,- usually on the grounds that their publics have low thresholds of cost tolerance and high ability to affect state policy. Even if there is little evidence that democracies are easier to coerce than otl1er regime types , tllis image of democracy matters. Sinte terrorists can inflict only moderate damage in comparison to even small interstate \\h terrorars, ism can be expected to coerce only if the target state is viewed as especially vulnerable to punishment. Second, suicide tenorism is a tool of the weak, which means that, regardless of how much punishment the terrorists inflict . the target state almost always hac; the capacity to retaliate \vith far more extreme punishment or e,en by exterminating the terrOtists' comm unity. Accordingly, suic.ide teJTOrists must ~ot onl)' have high interests at stake , they must also be confident that their opponent \~rill be ~t least somewhat restrained. \1\th.ile there are infamous exceptions, democracles ~a"e - n geneldUY been more restrained in their use of force agru.nst Cl\ilians at least .smce . . World War II. Final ly, suicitle attacks may also be harder to organize or pub~cwtha~n authontarian poli "hili. 'ce states, although tl1ese poss1 ties are"eakened b, tbe fact t we-ak auth ontan an states are also not targets. h b d mocIn f:act, the target state clern SUIClCle campaign as d a .e .. een of evei)' mo racv Th U . . r ,.. ' e ruted States, Fran ce, Israel , India, Sn. ~.J<U1ka Turke, an Russia were






'I 1 1 1 "' (> < 15 :u (. t'lllOt'nll'l('$ W lel tll")' ,vtJ 1ttackcd b) suicide ll'JTo! i~.;l c.l'llJ1 a;rr1 ' ('V(;Il 1-l thowrh the three became dcmocracres more rccerlll) I h 111 t 11 ' <llwr-; .. .. ... t trtddlc Turkev and Iraq. illustrate tl1 l)cJiltl t1 1 .-11 .. I( at IC:Jt T l1' 1:'\.UJ'dS. W 1 ]1 :S < liC tt' lTOJi. t carnpaigns are more likely to be targ~ted. ~tgal!l~t d(~IIIOer:n<.:ies lh<tn . . .w ll1011'tanan reg'ul1es. A1LI1<)llgh rraq has been far 11101 ( bn 11 .tl [I)\\ md Its Kill'( ['1s I1 . J popu Iat 1011 tl l<ll1 Ilas -nl11-ke)', vi<)lent Kunlish 'gr<)llps l1aV<' mPd sni<.:idc 'lttat ks . . y -~anst 'c . excIusl\re 1 an< 1 c]e 1110c1-.(1t1 Turkey and not al!<11r1St the aut hontana11 rc1, 1111 <.' 11 1 . ~-> lraq. There are plenty of national groups Ltvmg llll~N < lllthontanan regirnPs With ~rie,ances that could possibly inspire suicide terron sm. bill non<: ha,e. Thus. tlw fact that rebels have resorted to this strategy on I)' when they fa<.:<' the mo re suitabl< type of target counts against arguments th~t st~icidc tcrroric; m is a nonstrat<gic: response, motivated mainly by fanaticism or trrabonal hatreds.
.1 '



The main ree:Lson that suicide terrorism is growing is that terrorists have learned that it works. Even more troubling, the encouraging lessons that te rrorists have leanwd from the experience of 1980s and 1990s are not, for tJ1e most part, products of wilueyed interpretations or wishful thinking. They are, rather, quite reac;onable assessments of the outcomes of suicide terrorist campaigns during lhis period. T<> understand bow terrorists groups have ac;sessecl the effectiveness of suic:iclr terrorism requires tJ1ree tasks: (J ) explanation of appropriate standards for evaluating the effectiveness of coercion from the standpoint of coercers; (2) analysis of' the 11 suicide terrorist campaigns that have ended as of 2001 to determine how frequently target states made concessions that were, or at least could have been, interpreted as due to suicide attack; and (3) dose analysis of terrorists' learning from particular campaigns. Because some analysts see suicide terrorism as fundamentally irrationaL it is important to assess whether tJ1e lessons that the ten-01ists drew were reasonable conclusions from the record. The crucial cases are the Ham as and Islamic: Jihad campaigns against Israel during the 1990s, because they are most frequently cited ac; aimed at unreallstic goals and therefore as basically irrational.

JfCSSIII'( '.rdlcss of whether that . coerClve { PIE:ssur<! is L the no..nic hard' l or otlwr hnJes of costs H . . k eco ". . . n . . owever fr m lne form ofJ, os ' rcer the ke' q I''>Lon IS whether a partkula .. OJn the perl>pt<:tive of "' coe effective t-1 . st . olf~ 1<lll ,1It 1rnat1ve methods of inA r e:ocrc1ve r.tte~ promises to . . . I,e morencreased ) effort . Th1s 1s cspe<.:ially true ~ uenc:e and so, warrants c:ontin( . 0r t erronsts wh0 h d or J I cl . u~ ed to a particular goa a11 so W illing to exhaust viJtuaJI are 1 e:omghly m1tt . bandoning it. Tn tlds search for an effec:tive st t y any alternativt rather t}lan a ' . . . ra egy, CO<!r . . . likely to be lar~ely a f undwn of est1mates of the succ:ess of c:ers assessments a:de terrorists, th1s means assess ments of whether p t ..~,ast efforts; for sui. as S UICJuc c:a1n cl e significant concessiOns. patgns prod duce . .. lance at the b e hav10r or sutc:tdc terrorists re . 1 h Ag 'ea s t at s ch t . rr l10 ds are_ i~portant ~n their calculations. AJI11of th:cle-on_s between alternative met ons that have resmted to suJc:H.le terronsm began the' . f orgamtl r..a . . . . tr e:oercJ\'C e forts ,.._ith re conventional guet n 11a operattons, nonsuicide terrorism 0 b th b O 11 ez ollah m . '} 1 l p rJ mas Jslamtc f11ac, t 1e KK , t I1e LITE and AJ Qaetla all, r clocl . na . .. . . .' use cmonstrative and destrucbve ~1eans ol vJOienc_e long before resorting to suicide attack. lndeed. looking at the_ tra~ectOJy of te~Tonst groups ov~r time, there is a tlistinct element of eA-perunentatwn m th e technt~ues and strategtes used by thesE:: groups and distinct movement toward those techn1ques and strategies that produce the mo t effect. Al Qaeda actuall~.prides_ itself for a.~ommitmen t to even tactical learning over timethe infamous terronst .~11anual stresses at numerous points the importanc:e of wtiting "lessons learned memoranda tJ1at can be shared with other members to improve the effectiveness of future attacks ....



The Apparent Success of Suicide Terrorism

Perhaps the most striking aspect of recent suicide terrorist campaigns is that they are associated with gains for the terrorists' political cause about half the time. As Table 1 shows, of the 11 suicide terrorist campaigns that were completed du.rincr 1980-2001, six closely correlate with significant poucy changes by the target state toward the te rrorists' major political goals. In one case, the terrorists territorial goals were fully achieved (Hezbollah v. US/F, 1983); in tJ1ree cases, the terro1ists territorial aims were partly achieved (Hezbollah ,.. Israel. 1983-85: Hmnas ,.. Israel, 1994; and Hamas v. Israel, 1994-95); in one case. the target crovernment entered into sovereignty negotiations with the terrorists (LTIE ,._ Sri Lanka, 1993-94); and in one case, the terrorist organization's top leader was released from prison (Hamas v. Israel, 1997). F ive campaigns did not lead to noticeable conces. SJons (Hezbollah ,s second effmt against Israel m Le banon, 19::: "6 a Hamas camo.:.r-o , paign in 1996 retaliatinO' for an Israeli assassination: the LTf E ,._ Sri ~ka. 1995-2002; and both PKK campaigns). Coercive success is so rare that _even a ~0% sur . "'cess rate is signific<m t because internatJOna1 mili'tat) . and economtc coercJOn. . us h ' ks 1 tl1. 11 a third of the t1me. ' mg t e same standards as above, generally wor ess " 980s and Th . cl to "'run m the 1 ere were limits to what suicide terronsm appeare eo . 'n<t interests 1 11 0 990s. Most of the gains for the terrorists' cause were modest. not m'o ' .all . _ <'ent a] .it] t r tothe target countries secuJity or welll l. an cl mos were potenti :' re' oca . c 1 poliC\ ble. F tJ . . . . elati\'elr nunor 10re gn , . or l e Umteu States and France, Lebanon was a~ [ rv) 1 to 1997 were mterest 1srae l's apparent concessions to t1 P I sti mans rom 1-.,on 1e a e

Standards of Assessment
Terrorists, like other people, learn from experience. Since the main purpose of suicide terrorism is coercion, the learning that is Hkely to have the greatest impact on terrorists' future behavior is the lessons that they have drawn from past campaigns about the coercive effectiveness of suicide attack. Most analyses of coercion focus on the decision making of target states, largd)' to determine their vulnerability to various coercive pressures. The analysis here, how~~er, seek.s to determine why terrorist coercers are increasingly attracte? W a specfic c.-oerciVe strategy. For this purpose, we must develop a new set of standards. because assessing the value of coercive pressure for the coercer is not the same problem as assessing its impact on the target. f From the perspective of a target state, the key question is whether the value 0 UJe L'Oncession that the coercer is demanding is greater than the costs imposed by


_,. .



lor('( s


securitv bufTer zone along the soutJ1em edg~ of the c;~)IIJ1t.r: 101 <llll~L hcr 15 years f rom w'1 1 a second HeziJoiJaJ1 stLicide terronst <:cllllJ><llgn fmiPc.l to dsloclue it 1.1 uc 1 . . t'1 1e Sri Lmkan uove mment did conduct appa rentl y senou. ncgotiallc with the 1:TTF ms from Nove J~be r 1994 to Ap1ill995, but did not <:Oncede th: Taut il's main dernanci c . uove . 10r md epen denC"", . ... and since 1995, the r: mment has prefe necl to prosecute tl e war rathe r than consider permitting Ta.rml secesson. . Still, these six concessions, or at least apparent conc<.ss1ons. lwlp to explain whv suici de terrorism is on the rise. In three of the ~ases , the target gove rnment po~cy changes are clearly due to coercive pressure from the. terrorist ~roup. The American and French withdrawal was perhaps the most clea r-cut cocrc1ve su<:<.:es~ for suicide terrorism. In his memoirs, Pres ident Honald Heagan explained the U.S. decision to with draw from Lebanon:
The price we

fr0111 ) of Caz.1 and the \\'es.t Ba~k and released Sheikh Yas in dt,::- 1c same P<;~~ Israelis ttlemcnt in the occupied territoties aJn~ost dou blt d :tllt! r <t'lll C\e nts have sl1own that L racl is not cletened from sending for<:c h~t~k m wlw 11 rlt>ce ~say. 1n two disputes. the tenorists a<:hieved initial su<:ccss l ~nt frult:d ~~ ~\...11 greater goals Alth ough Israel \\ithdrew from mu<:h of Lebanon m June l\-J~:), t rl'lained a six- rlil~ 1

mort ' modest than the, micrbt appc;u. Altho11gh Isrcwl with drc" .ts

.\ ' d 1994, Ham as began a . lsra~ J to the J Jt'b~o \1l' Afte r two atta<:ks lsenels of suic:ide hornbincr . latton , srae dedd d 't/> m 1m: h wac, regu r\hdrawal fro m G.a ' ired under the Oslo A(J e to a<:celerate its ,v:~ b n delayed. Ha rn:1~ l hen suspended attacks ~ c, . reement but whj h Iad ee 9- I I ( cl or nve month F c I 4 to Augu st 19 .?. amas an Islamic.- Jihad ) earned out s. rorn Oc:toher 199 '"' gainst Israel. h ~cptember 1995 Israel ag d a total of se\'en suic:ide ttacr--' " ' ree a. t Bank town s that Dece mbe r, whkh it earlier had to withdraw from certain 1 . d could \\es Apri l 1996 at 1 c r t 1c ~oonest. Hamas then suspend d rume not be done beore I gn durin g the ast week of. February and first e attac:ks until . retaliation k f Its caJ11PaJ . . M atch 1997, Hamas began a suicide attack campwee o ~are; h 19(\6 Fmally. ai th~t . , 111 bo t every two months until September 1997 In res gn m~luded a u . h ponse 1sraeh Pri an attack . . tan)1ahu autl1 onzed t e assassination of a Hamas d Tl me\['tn1ster \ 1 er. le attem .e r .. ,' ] d h 1 I ea man. Jord an. wue o an t e srae 1 agents were captu d , Pl , .m .Am I Sh ikh Ah re . t O ~et them b k Israel agreed to re e~e e m.ed Y assin. spiritual leader of \\a~. 'd . rule tl1.iS was .not a conceSSIOn to the terro nsts' territorial goals , the . d b. . , . . re lS no amas mterprete t IS m an) way different from the standard evl . ence that . . H . vle" that tlus release was the prod uct o f Amencan and Jordanian pressure ....

r concessio n~.


* "';..





had to pay in Beirut w~ so great; the tracredy at the barrac~ was so enormous.... \-\le had to puJJ out . . .. \Ve couldn t stay there and 11111 the nsk of another


Despite suicide terrorists' reasons for confidence in the coerche effectiveness of t!Us strategy, there. are sharp limits to what suicide terrorism is Likely to accomplish iJ1 the future. Dun ng the 1980s and 1990s. terrorist leaders learned that moderate punishme nt often leads to moderate concessions and so concluded that more ambitious suicide campaigns wou ld lead to greater political gains. Howe,er, toda,-'s more ambitious suicide te rrorist campaigns are likely to fail. Although suicide t;rrorism is somewhat more effective than ordinary coerche punishment usina air power or econ omic sanctions, it is not drastically so.

suicide attack on tJ1e Marines.

The IDF withdrawal from most of southern Leb anon in 1985 and the Sri Lankan governme nt decision to hold nego tiations with the LTIE were also widely understood to be a direct result of the coercive punishment imposed by HezboUah and LITE respecti\'ely. 1n both cases , the concessions follO\.ved perio ds in which d1e terrorists had turned more and more to suicide attacks , but since Hezbollal1 and the LITE employed a c:ombination of suicide attack and conventional attack on their opponents, one can question the relative weight of suici de attack in coercing these target states. However, there is little question in eithe r case that punisbmen t pressures inflicted by these terro rist organizations were decisive in the outcomes. For instance. as a candidate in the Novem ber 9, 1994 , pres idential election of Sri Lanka, Mrs. Chandrika Kumaratw1ga explicitly asked for a mandate to redraw boundaries so as to appease the Tamils in their de mand for a separate homeland in the island's north east provil1ces, often saying, "vVe defi nitely hope to begin discu ssions \\ith the TamiJ people, with their represen tativ es-i ncluding the Tigers- and offer them political solutions to end the war ... (invo lvi ng] extensi,'e devolution." This would, Kumaratunga said, "cre ate an enviJonment in which people could live without fear. The other three concessions, or argu able concessions, are less clear-cut. All three involve llamas campaigns against Israel. Not counting the ongoing seco~d intifada, Harnas waged four separate suicide attac k cam paigns agai nst Israel. 10 1994, 1995, 1996, and 1997. One, in 1996, did not correspond with Israeli conces 51005 'T'L: _ of 1 J.Jw; campaign was announce d as retaliation for Israe l's assas sination ' Hamas leader; no particular coercive goal was annou need , and it was suspended .b~ Hamas after four attacks in two weeks. The othe r thre e all do correspond w,th

Suicide Terrorism Is Unlikely to Achieve Ambitious Goals

In intern ation al military coercion, threats to inflict military defeat often generate more coercive leve rage than punishment. Punishment, u ing anything short of nuclear weapons, is a relativelv weak coercive strateg)' because modem nation st-ates generally will acce pt lugh costs rather than abandon important national goals, while mod em administ rative techniques and economic adju tments on'r time often allow state s to minimize civilian costs. The most punishing air attacks ''ith conventional munitions in history were the American B-29 rai~ ag~t Japan 's 62 largest cities from March to August HMS. Although the e nuds killed nearly 800,000 Japa nese ci\il ians -almost 10% died on the first da~ the ~tare~~ 9 1945 flre-bombing ofTokvo whic h killed o,er 85,()()()-the cou,entional bom mg ~d not.~mpel the Japa nese to su rrender. diffi ult than for Sutcide terrorism makes adjustment to reduce damage more . c -ect states fa"".J 'th . ,L "'"-u V.'l militruv coercion or econon11c saneti.ons Howe,er it does not aill . uae target state s intere~ts in the issues at stake. As a result. suicide terro nsm can
J '

\I ,_

l l I 1..

.:>I KA I t::GIC




dest goals. such <L'i ''~thdr, \'.,d fro 111 t<~r. cocr<'e :>tates to abandon limited ~r ' no eJ.s case in J994 ~utd 199.t:;. 1 tu nr)ort nt()ry . . . or as tn 1sra < 1)' a11d of 1 strategc unportance ow rtant area. However sus~,; Id<.' l<>rro . . a1 . cl ~ ~' fJ ., more nnpo p< uti \\1t1 Ja\\ dJ ro11 ~ l I al5 central to t h<'l.r we;;Jth nr "~'<:ruitv s nsn 1 is I .k 1 t to aban<. on go uc 1 as. un lJ e y to cause targc s d .-.' the economic pro peels of tJw , ,,\I( or strencrl) .u loss of territorv that woul " (::ii.J\en n I Lit . the rivals of the state. .. 1111 t more effective than in ill t<'rncttio 11 . . Suicide terrorism makes purus1 en . ., . 1 . . . . ' 11lntl. . . . . riJJiJ g to counten,tnC( htg 1 cost~ ror llnpori 1tan: coerc1on. Targets rematn " 1 .. cl' . <lt tl ' . . . omic or rmUtarv a jiiStmcnt:, to pt cvent suic'1 l . . . .k I . goals. but admmJstratJve, econ l < 11 e ) tu be deterred h .1 . . d attackers themseh es ,u c u n attack are harder. w h1e sU ICl e . . l'k I . ~ Y . . Accordingl)' su 1c1de attack ts 1 e ' to pH sent a thn~at 0f tl1e threat of ret 1anon. al . . . . . . 1 .ted c ,.,,....0 } . Junishment that the target go' em ment cannot co,n continumg LmJ 'J\ < ': . . p IeteIv eIumna te, . d tlle upJJer bound on what punishment can ge:un for c:oerccrs an . . 1 .. .s recogru'zably 1 . 1 . 1 1 terrorism than 111 mternationa lntlitcuv cocrc:11J1 ' 1 ug11er 11 su c dal . . . r. The data 0 11 suicide tenorism from 1980 to 2001 support th1s condu. ioJJ. 'While suicide ten orism has achieved modest or very limited goals, it has so far failed to compel target democracies to abandon goals central t~ national.wealtl. or securitv. When the United States withdre'v from Lebanon m 1984, 1t had no import~nt security, economic, or e,en i~eoJogical interests at stak~. Lebanon was J araelv a humanitarian mission and not v1 ewed as central to tl1e national welfare of 0 tJ 1 United States. Israel withdrew from most of Lebanon in June 19R.5 but e remained in a security buffer on the edge of southern Lebanon for more than a ' decade afterward, despite the fact tbat 17 of 22 suic:ide attacks occurred in J98.5 and 1986. Israel's withd rawals from Gaza and the 'Nest Bank in 1994 and 1995 occurred at the same time that settlements increased and did littJe to hinder the JDF's return. and so these concessions were more modest than they may appear. Sri Lanka has suffered more casualties from suicide attack than Israel but has not acc-eded to demands that it surrender part of its national territory. Thus, tJ1e logit of punishment and the record of suicide terrorism suggests tl1at, unless suicide terrorists acquire far more destructive technologies, suicide attacks for more ambi tious goals are likely to fail and will continue to provoke more aggressive military responses.


While the rise in suicide terrorism and the reasons behind it seem daunting, there are important policy lessons to learn. The current policy debate is misguided. Offensive military action or c:oncessions alone rarely work for long. For 0.'er 20 years, the g?vemments of Israel and other states targeted by suicide terrons.rn have engaged m extensive military efforts to kill, isolate, and jail suicide terr?n t leade~ ~d operati~~ sometimes with the help of quite good surveillanc.:e of th~ terro~. commumties. Thus far, they have met \\~th meager success. Althoug ~~n of ~de terrorist organizations can disrupt their operations t~m porarily, at rarely ~'lelds long-term gains. Of the 11 major suicide terrorist camp<Ugns

that 1 p decapilatic,n . lwn tire leader in Turki h ur <:y-<lid so as a result r I aders 11 cl I l . . , s <:ustoclv ask d h. o e So far, lea er-. ''P c <:e.apttatlon Ita<; also not ended' <.; ~ follower<; to stoJ~ogh the U nittd ~~ ,,tc~ suc:ce~sfully toppled the Talib~n ~aeda's <.:arnpaign. Al mber 200] ' A] Q ,teda laum:h~d scvtn SU(;(;essful . . :u In Afghanistan in 1 1 oec:e t December 2002. killing some 2.50 \Vesterr1 c ~~~ <: e terrori!>t attacks fro 1n l An o JVJ 1ans n . h . I~ before September 11 , 2001 , <.:ornbined. ore t an 1n the three 5 yearconcessions are also not a simpie answer. Con<.: . that are w1c1e1 beId t I)e terronsts, <.:ornrnunit esslonsdto natio nat1St gne'v m .. ances .~ k' . y can re uce pop 1 , ~ rther terrorism, ma mg 1t more dj fficult to re<.:ruit . . u ar support ~or uoving the standing of more moderate nationalist el'tnew t~UI<:Jd~ attack<;rs and 1<:::s wno are 1n c.:o ru-t' . 1mpr I I 0. t1 the terrorists. Sue 1 )ene11ts can be realiz~d howev:. m~---dtion 1 r t he <:once . ,,, 1 11 .F h ' .. r, on "J reallYdo substantia y satls y t e nationalist or self-detennination . . . SSions . aspirations of a large fraction of the communih'. . . Partial, incremental, or dehberately staggered con<:ession . th . d ut over a substantial period of time are likely to become thew~ t atflare raggt:u th o . . . ~ . rs o JO worlds. Incremental co~promdselm.ay app~ar-or eastly be portra~ed-to the terrorists' community as s1mply e <l)1ng tactics and, thus, mav fail to reduc . . . h h . . , e. or actuall, . net JJl crease the1r distrust t at t etr ma.Jn concerns will e,er be 1 . Furth er. . merementalism provides time and opportunity for the terrorists to intentionalh- pro,oke the target state in hopes of derailing the smooth progress of negotiated compromise in the short term, so tbat they can reradicalize their own communitY and actually escalate their efforts toward even greater gains in the long term. Thus. states that are willing to make concessions should do so in a single step if at all po ible. Advocates of concessions should also recognize that. e,en if they are successful in undermining the terrorist leaders' base of support. almost any con<:ession at all ,,;11 tend to encourage the terrorist leaders further about their O\\ll coerci\'e