Causes of War: Power and the Roots of Conflict [1 ed.] 9780801482953, 080148295X

Citation preview

Causes of War

A volume in the series

Corne/1 Studies ill Security Affairs edited by RobertJ. Art Robert )ervis Stephen M. Wall A list of titles in this series is available at

www.comellpress.comeU.edu.

Causes of War POWB:R AND nrn ROOTS OF CoNFLICT

STBPHEN VAN EVBRA

Cornell University Press ITRACA AND LONDON

Copyright C 1999 by Comcll UniverSity

All tights tt"Setvt!d. Except for briel quotatio1u i.n a review, this book. or p6tU th•...ol, must not be roprodu«9S·J (pbk. : olk. P•!"'r). 1. WM.

;t.

l)all)n--mestic economic or political systems of states, or the distribution of power among states). Many hypotheses that identify manipulable causes have not been tested, leaving skeptics f - to reject them. Accordingly, these writings are largely ignored by opinion leaders, policymakers, and peace groups.• Meanwhile, failed peace ideas-for example, disarmament, pacifism, and large reliance on international institutions to resolve conflict-remain popular for lack of better alternatives. ln short, our stock of hypotheses on the causes of war is large but unuseful. Those culling it for tools to prevent war will find slim pickings. Accordingly, my main purpose here is to propose prescriptively useful hypotheses on wars causes-that is, hypotheses that point to war causes that can feasibly be reduced or addressed by countermeasures. Toward this goal I elabt991. p. :zu). other swvey$ ol hypotheses~\ the causes of wttt are few, and r\01\e ace really comprehensive. Other useful surveys i.ndude Gug C~ What Ollllt$ Wnr7 An lntroduc. lion ID Tlrt:oriaofltll An important hypothesis points 'to a strong cause- that Is, one that has large effects. The greater the effect produced, the greater the explanatory power. ExplatUJiory ra ..gc: How many classes or phenomena does the cause affect? The wider the range of affected phenomena, the greater the explanatory range of the hypothe6is. Most social science theories have narrow range,buta few gems explain manydivcrsedomains.1 Appllc#billly: Is the caUS31 phenomenon that the hypothesis identifies common in the real world? How common are the conditions that it needs to operate?The more prevalent thecausearul its conditions, the greater the explanatory power. Even important (that is, strong) causes have tittle explanatory power if they are rare in the real world or if they require rare hothouse conditions to operate. A weaker cause can have more explanatory power if the cause and its required conditions are common. The past prevalence of the causes and conditions of a hypothesis governs its ability to explain history; their current and future prevalence governs its ability to explain present and future events. 2. Prescriplivt richness. Can the war cause that the hypothesis identifies be manipulated? U not, can its effects be mitigated by feasible countermeasures? Prescrlptively rich hypotheses point to causes thot can be manipulated or whose effects can be mitigated. J. Degree of sat!sfactliJn. How well does the hypothesis satisfy our curiosIty? The further the cause stands from Its proposed effect. the greater the

131

Cause. of War

satisfaction it offers. A politician once explained her election loss: "I didn't get enough votes!" This is true but unsatisfying. A good explanation would reveal why she did not get enough votes. I ask five questions of the hypotheses I examine: (1) How can they be tested? What predictions can be inferred from each, and what tests of these predictions might be devised? (z) Do they pass preliminary tests? I test three hypotheses using single case studies (in Chapters ), 4, 6, and 7). (3) How great is their explanatory power? That is, how large is their importance, how wide is their explanatory range. and how well can they explain recent and current international poUtics? (4) Are they satisfying? (5) What policy prescriptions follow from them?

ARCUM11NTS ADVANCED, ANSWilRS OFFERED

This book concentrates on war causes related to the character and distributi.o n of national power. Power factors deserve attention because they strongly influence the probability and intensity of war, they are relatively malleable, and they remain understudied and underappreciated, despite a growing )jterature that addresses them .' Thomas ScheUing notes that there is "something we might call the 'inherent propensity toward peace o r war' embodied in weaponry, the geography, and the military organization of the tirne." 10 This volume develops SchelUng's idea by exploring five speci6c hypotheses: H a. War is more lilrIU:. "OrigiM ol WM." p. ,o; Mearohelmhe\mer, "Back to the Futu.re/" pp. t8-t9,; and a.rguiJ:\g: the opposite, 81ai.ney, CDu.stS of War, pp. 109-14· Some Type Urealists pay some altention lo misperceptl..,. of tho gross 5trucrur. o/ po~r u a war cause. See. for ); Finn!Jh&vlfWar One Mexican officer boasted lhat his cavolry could brook American infantry squares wllh lhe lasso." Before its crushing 1866 loss to Prussia, Austria exaggerated its own strenglh and underestimated every category of Prussian military strenglhquantity of trained troops and quality of leadership, tactics, military technology, and logistics.•> Pour years later, France was so certain of victory against Prussia lhat lhe French Army issued maps of Prussia but not Prance to its officers." French Minister of Foreign Affairs Due de Gramont boasted lhat "in a few days we will be on lhe Rhine" shortly before lhe Prussian Army routed French forces and besieged Parls1 7 French military advisors assured Empress Eug~nie lhat "our offensive across lhe Rhine will be so shattering lhat it will cut Germany in two and we will swallow Prussia in onegulp.n " Russian leaders approached their 1904 drubbing by Japan wilh high confidence. Minister of War A. N. Kuropatldn assured lhe Czar lhat his armles would quickly expel Japan from Manchuria and Korea and end lhe war with a "landing in Japan, annlhllation of the Japanese territorial army ... and capture of the Mikado." 19 Hence he dismissed Japan's prewar threats as bluff: ''TheJapanesewlll not dare to light, they are unprepared, they are only putting on airs, thinking that we shall be frightened and believe them.""' One member of the Russian general staff forecast that "we will only have to throw our caps at them and they will run away:•>! Russian Navy Chief of Staff E. I. Alekseiev declared that "I personally cannot admit the possibility of the destruction of the Russian Beet by the Japanese" shortly before Russian flotillas were beaten at Port Arthur and crushed at Tsushima.» " BaUey, Oiplon1Atk History of u,. Antuio2n PtOplt, p. >S9· as Richard Smoke, Wa-r. Umlrolling £scalgJion {Cambridge: H.arvard Univet1iity Pre551 1977). p. 98· " B!ainn. 1'1rl fa9: Saburo lenaga, 1'1rl Pacific War, 1931-1945 (New York: Pantheon, 1978), p. Ss. ~ ICNga, P"cific WQr, p. 85; ~d EcfwU\ P. Hoyt, Jopc~n"s WQr: T1re Cmll Alc:ific Conflict, t8$J lo1951 (New York: Mmp~y CO>rry the war into the Cennan cmplt. ). M. Madcintosh, "The Development or Soviet Mllltlry Doctrine sbn 1918,"in MlcNel Howa.rd, ed., Tl,. 7'11rhard Welnberg notes that Hitler thought the United Sta,.. was •deprived by Its radal compooit:in of th. .blllty to produce an effective mllltary lo~• Wd td. (Now Yorlc Oodord Uniwr-56 (on the l>tet950S and ..rty t!)6o&); and JohnJ. Mea.rsheimet, ..Why the Soviets Can't Win Quickly in Central Europe," lnl.mtttti(tltit./ Stcur· ity7 (Summer •98>): )-)9 (on the late 19705 and euly tgSos). 61 On these points see Chapter.) at notes of7- S3,a.nd Oupter6oat notn Unijia~tion (London: Longman, 1991), p. •89n Thylor, Rumours o{W11r, p. 1,56. " See Chapter 7 at notes 20, 11, 40. and t}1. " Mooes, Politi=749· The Poles opposih!ly oxpected more holp than they got from Britain and France ln '939· The French assured the Poles they would launch a major offensive ogainst C.,rmany ln the third W72n; Chinese ca.su.altie& are &om J. Oa>~id Sitlge:r and Melvin Small, Tht W~tgu of W.:rr, t8t6196J:A Sta/isliall Handbook (New Yotk: Wilcy, 1912), p. 117. Compamon to other wars is with duta from Sivard. W(Jr/d Milillny and Sodtt1 ExptndllurtS 1991. pp. u-zs. Sivil!d estimates t~ w1d..ths ln Korea al three million. 21 Sch.lesinger first estimated US. tasuaJties from va_ rious Soviet countedorce attacks at fifteen to tw(!nty-6ve tholl8And, and later at one to two miUiol'\, i.n conpsslor\ll "ppea.rance:s in •974·1ohn Ouflield and Fnonk von Hlppe~ "TheShort·TennConseq..,.,.. of Nuclear War for Civilians." inJulius London and Cilbcrt F. White, eels., 7M Environmmtal Effsting esllmote see Wllli•ro O..ugherty, Oarb.ua Levi, and Frw von Hippel, "The Cons.St t)-)4 million U.S. d .. ths. ut Major Alexander dl' Seversky. quoted in Scott D. Sagan. Motring TttTBds: Nu.claAr Stndtgyalld Nalio••IS«urily (Princ«on: Princ«on Univet'$ity Pr» A euphoric Belgian officer exclaimed, "War, what an exalting thing!" when Germany presented its ultimatum.'""

To l'REv!!NT WAR, PltOMOTl! l'RANSPARI!NCY The historica.l record suggests that false optimism is a potent and pervasive cause of war. False expectations of victory widely coincide with the outbreak of war."" This suggests that false optimism is a strong and common cause of war. If so, the control or abatement of false optimism would prevent many wars. It follows that the risk of war is reduced when relative power is dear and the nature and costs of war are known. States seldom fight when both sides know the winner and the price. Hence anything that makes the world more transparent will reduce the risk of war. Policies of deception and secrecy are bad; policies of open disclosure are good. ocomed the notion of petP!tual peace u ""not only impoosible but immoral u wcll" Quoted in BlaMing, Origin~ of lh< Frtrrd> Rects victoty oJter • brief 01\d triwnplulnl compes ( - -) .... strC)IIg proposed "'""'' paths Ught llnes (--)are v.ry w...k p~ ausal ~"'""'

H2, V...ton 1: Tndilional5btbility theory (Schtilling)

Expl. li2A:

r---- OpportwUstK lint atrlke (lnlf'HlA)

J)

J

1----•~ Wllr

ExpL 112B,: 1 - - - - - - - - - + Pre, and ptr«pdOM CltuP li2C)

Expl. tLlO:

1---+ Huty/trunated -------+1 diplomacy (lntP 1-120)

Expl 1-12£: ' - - - - • Offensive .Ovantage _ _ _ _ _ _ _., (lntPHlE)

H2: Hypothesis2 £,cpL: ~p1an.tion lniP: Intervening phenomenon

Jumping tho Gun !he costs of a prolonged confrontation." Finally, a mobillzation can spur another's attack If !he other thinks the mobillzation is a harbinger of war and that the side strildng fust will have an advanta.g e. These secondary effects ca.n ~ potent causes of war; ~ a first·mobiliz;ation advantage is oken nurly as d.tn3erous as a first-strike adva.ntage. I alao use "fi~t-move advantage• broadly in a second sense: to include first strikes and mobilia:ations against one's oppo11enl (for example, Israel's 1967 attac.k on Egypt) and 6:~1 strikes against11tulral lltlrd porlits launched to forestalla.n enemy attack on the third party (for example, Germany's 1940 attack on Norway). Both incentives raise the risk of conflict. The first gener· ates conflict between two prime adversaries, the second widens a confl.ict to engulf neutrals.

HYPOTHESES ON THl! fuFECI'S OF FtRST-MOYE ADVANTAGES

First-move advantages have five dangerous effects. Of these, the danger of preemptive war (explanation H28) is the best known. This danger ex· plains a significant amount of war. Its most famous variant, "reciprocal fear of surprise attack," is very weak. however. Reciprocal fear has been much discussed but rarely occurs. The strongest danger-the concealment of grievances, capabilities, plans, and perceptions (explanation H2C)-has been overlooked so far. It holds that first-move advantages cause war by crippling diplomacy. Critics have complained that stability theory posits a mechanistic, apolitical interaction. The "concealment" hypothesis frames a danger that stems from the corrosive effect of fi.rst-move advantages on the diplomatic process, and shows diplomacy shaped and confined by military factors. Opportunistic •lt•ck (expla110tion HlA) A first·move advantage improves an attacker's prospects for victory, since the attac.k er claims the fust move and enjoys the benefits that come with it. This causes war by tempting states to play the attacker. Starting wars carries smaller risks and pays larger rewards; hence states are quicker to start !hem. Such oppo.rtunistic logic has figured in many decisions for war. Japan at· " ~U otTattglsts h.a'"' long held 11\atIMael must altJCId I~ •Crisis and~ In nu.. Ar.ob-tmoll W•ra." In Knon and Mor· pn. SIHUflt MJ/Jbuy s.,....., pp. 111-46at nf>->7, I)}; Mooho O.y011, Sl«ytfMy /Jft! J\!1 A• tol>'"lfi.,.,., (New Yorlc WllmOf Boob, tiJ76), pp. 49- 1:0: Lullwak and tloiowilz.lmdi Army, pp. :us·>J;)&nlcoCJONSiein. "Th. " 15no.eU mllltaty lnt.Wgmce thought the Arab$ could be cnuhed as euily .. in •967· See ln$1ght T..m of !he London Sunday Times, The Yom Kippur Wor (Canden City: Doubleday,

of"''"""""''

1471

CauSilSofWar

son to make conce$Sions." Negotiations were probably doomed in any case, but Egypt's deception made certain they went nowhere. Du.r ing the fall of 1950, China likewise took elaborate steps to conceal the s trength of its growing Korean force from the United States. Chinese troops were given gross underestimates of their own force size, then ordered to surrender in hopes that their U S. interrogators would believe these under· estimates, as in fact they did.'"' On the battlefield the Chinese theater corn· mander pursued a strategy of purposely showing weakness, to encourage the United States to rush unwarily into China's military trap." Mao was overjoyed when his intelligence leamed that General Douglas MacArthur, the UN commander, had swallowed U>e deception." This made MacA.rthur easier to surpriSe,"' but it also fed the growth ofU.S. war aims: U.S.leaders soon embraced the goal of conquering North Korea, which seemed an easy prize. Hidden military plans In a world of first· move advantage, states more carefully conceal their military plans and capabilities to preserve their capacity to strike by sur· prise. Otherwise, targets are forewamed and tactical surprise is lost. But such secrecy raises the risk that leaders will issue orders that have unforeseen military effects, which raises the danger of inadvertent war. For ex· ample, in July 1914 Russian civilians authorized preliminary mobilization t974), p. 65. Reflecting this confidence, Israeli Deputy Prime Minister Yigocl AUon dedru\'d during the swnrncr oltm that -egypt has no military option at aU." DonaJd NeH,. Wgrrion for jtn~SPltm : nt. Six C/rlys 'IMt Ch•ngtd ""' MUidlt Eas/ (New York: Simon & Schusrer. 1ga.),

pp. 6--7. lmeU Prime Mi:nbter Colda Melr declared the Arab ottt.:.c.k ..suidd:aJ .. when sht an· nounced the attock to Israel's public. Nadov Salran. lsratl: Tht fml>atrltd Ally (Cambridge: Harvard University ~'reo5) and preliminary mobiliz.ation was already underwtty. Sc:hm.itt,. CDmir1g oflht War, 1 :505-6~ 2:9+- 96; 0. C. B. Lieven. RJISSiQ and thl Orig;ns oftht Fi1$t World War (New York: St. Martin'S,. 1983)1 pp. 144 - 4.S· This a-ror st l\onald H. Sp«tor, E«g~•g•il1$l lh< Sun: Th: 5)9· " jacl< S. t..evy, •f>referenc.ed that the United States would not help Egypt recover any lands it lost in a new war. He blithely expected the United States to force Israel to retum a.n y con· quesbl, M it had in 1956-57. As a result, he thought he could risk war safely. Israel struck before the U nited States could dispel his lllus ions.n

Offmsivt military foru postures and capabilities (rrpltmntum HlE) A first-s trike advantage invites states to adopt offensive force posture&. since striking first requires operating on the tactical offensivt'. As a result forces grow more offensive, raising the risks of offense dominance (see Chapter6). States fearing attaek are also slower to mobilize for self-defense in a aisis, since military moves are more likely to trigger preemption by the other side. In tryin8 to avoid provoking war, however, states may do too little to deter it, leaving themselves exposed to an opportunistic strike that mobilization might dctcr.7l The acuteness of this dilemma grows In direct proportion to the si2C of the advantage accruing to the side striking fint. Russia hns experienced both sides of the dilemma. Russia began preliminary n\obili?.ntion i.n 1914 partly to "take all the steps which would enable us to focc a.n attacl;," but in so doing triggered a perhaps avoidable war." Twenty-seven years later, as Hitler readied his Barbarossa attack, Stalin re"'On Rwolo~ 15J>Or.m«ol5• 1)9-40; M. F. Sclillllrlg. H..,lht Wor Bqon in 1914, IT&N. W. Cyprtan 8ridg:J>S-:19. ~ V. NomllE os required by the need to gain surprise or by the risk of being surprised. For example, officials might explain that "we must attack to forestall our opponents' attack" or •we must fool our adversaries into underestimating our military strength, to ripen them for surprise." 4· Decisions to use force, or that cause others to use force, should embody or respond to one or more of phenomena H2A-H.2E,or should reOect mlsperceptlons arising from these phenomena, when taken by states that perceive a first·move advantage. For example, we should observe states starting wars in the tilraU of false optimism that stems from an opponent's concealment of its capabilities, or unwittingly triggering war by harming interests that the other never shlted.

Tests These predictions are tested in three cases: the U.S...China conJlict of 1950, World War I, and the 1967 lsraei·Bgypt War. I chose these cases for three reasons. First, they are well recorded. Second, in each case large real or per· ceived first-move advantages obtained or emerged. This creates a good set·

OIUStS of War

ting for tests that contrast conditions in these cases to more average times and places. ntird, in two cases thesizeofthcreal or perceived litst-moveadvantagc va.r led sharply across time or space. in the 1950 U.S.-china case, the size of the fitst-move advantage grew markedly over time. This allows £ruitful within-Seph>mber>). Willi;un Stueddcrs wotc:hed the first-strike advantage appea.r and expand in October 1950 (prediction 2). China then

•s

lsSJ

]unrpi11g lht Gun

made s triking use of strategies of surprise attack Md concealment (pn!diction t). US leaders saw no first-strike advMtagc. Acoordingly, their ron· d uct showed no change during October (there was no US. consideration of prt!t!mptive war or greater policy concealment), a-eating a clear difference betw--. Chinese and US. conduct (also prediction 1 ) . Chinese records indicate that Chinese policymakers acted as they did to gain the advantage of surprise (pn!diction J). Evidence also indicates that the U.S. rush to the Yalll. which set the stage for war, was encouraged by these Chinese actions (prediction 4). The U.S..China case corroborates s tability theory, but how strongly? We judge the significance of a passed test by asking whether the successful predictions are unique." In this case, two of stability theory's successful predictions are so. No competing theory predicts either the Chinese switch to a lulling strategy in October 1950" or the Chinese testimony that China lulled to gain surprise. The U.S. rush to the Yalu has other possible explanations (for example, "McCarthyite U.S. domestic politics compelled aggressive risk taking by President Trwnan in Korea in disregard of China's warnings'}. The sharp difference observed between Chinese and U.S. behavior also has other plausible causes (for example, "China acted more secretly because totalitarian states arc reflexively more secretive than democracies'}. Hence these events only wealdy corroborate st??iUty theory. But China's switch to lulling and Its explanation for the switch strongly support stability theory because other explanations for these events arc hard to imagine. How much Importance does this evidence assign to stability theory? That is, how large is the indicated Impact of first-move advantages on states' ideas and behavlor? Thls depends on how greatly state behavlor varies with shifts In the size of the first-m ove advantage. Here we see large variance. China's conduct changed markedly as the first•move advantage appean!d in October t9SO, and Chinese conduct differed sharply from norniAI lntema· tional conduct and from U.S. conduct. FlnnUy, did China's October-November lulling strategy cause an otherwise avoidable war? This seems quite plausible. Re«nt research has shown that Mao's decision to enter the war in force was erystliiUzed not by the US. approoch to the Yalu River but by the earlier entry of U.S. forces into North " Unlqu< predlla. '"' 11 approved partial mobilization against Austria unaware that this was tedmi: ,.,- 8; Alb1 tells me thtt mWwy pr54 of Egypfs 385 rombal liltraft on il1e lint day, while losing 19 of ils 197 combat aircnft. l'l:S Ki.mc:be and &wlcy, Six·Drly Wll"r, p. 283.

1671

C•umcfWor

sis of •908, the European crises of 19J3-39, the Pacilk crisis of •9-IO- ~t, the Taiwan StraiiS crissof 19.54- 55 and •958, and !he Berlin crises of •9411 and 1958-6~ The slu of this deviation-not enonnous but significant-is a measure of s tlbility lhrory"s importance. Was first-move advantlge a n~ry cause the •967 war? Egypt and Israel luld rru~ny reasons for war In June 1967 and probably would lulvc come 10 blows absent a first-strike advantage. With more time, however, Nasser mjght hl>vc realized the world's unwillingness to save him from defeat and might have backtracked accordingly. 11 alro seems remotely possible that with more time for rcOcction the United Stales would have Intervened more forcefully to prevent war.

Assessing stability thoory Stability theory passes a number of stro11g tests. Several unique predictions from stability theory are fulfilJed. State behavior and ellte ideas vary both regularly and markedly with the size of real and perttived lirst-move advantages. The regularity of this v~ gives s tability theory a good deal of credibility. The marked size of this ~-tlult is, lhe large shifts in government behavior and ellte ideas tlult accompany varian.:e in fust-move advantage-indkates tlult first-move advantage is an important (tlult is, strong) cause of war. U we grant that stability lhrory is valid and important, counterfactual speculation sutl&ests that it also explains a fair amount of history, especially if we frame it to cover the dange.r s of perceived as weU as actual first-move advantages. I argued above that first-move advantages, or perceptions of them, may have OOC!n necessary to the chemistry that brought on the three wars nnalyzed above''" In one of these cases (1914) this seems very likely; In another (1950) 11 seems quHe possible; In a third (1967) lt seems uniJIgic: "You do no< want to to. JUbfuaotodlllwn promptly promd 10 i""fws~'" tlw ~ aa 10118 as hil ~ olkr~ )'0\1 tlw occaaion; for il )'0\llcl lt pooo.fugltlvo opportunity will """"'ave1 into tl1o _,.,y's c•mp; on., Diplom«y, p. >159· you.• QuOifd in Vop. o.{r>lu ' Prevtntlve war l&aometlnw defined u dJ1vtn by two fun: of dcdlnln,s re,Latlve power, and o( later Att.lCk by the oppontnl Hence it iJ a war fought now to prtvent a war begu.n by others '-'ter under wor&e ('ircum.stances (thu$ the ttrm ·p~ven1lve'"), Sc.oe, for f'xample, Vagts, Otfm• rmd Dlplomnty, p. 263.1 use a broader definition: prevt_ontlve war bl driven by fear of dt'ClWng rclatil.'e power and fear of loner disputes with the opponent. 11.c auaeket's purpose

••d

l$to h•lt lttown rei•U~d«.Une ln ~·~ror toO'e•tell fAit accompllon th~IJAI.. tndll.pu~. ltl g011l il to avoid a w.u w•gtd t.ter under won~e c:ondHiON, or to •void ~'13 !..tu oom~ polfdcos provmtlvo: "lt Is .-...ry tobrtal)8, )p, 356; Michael Handel. "Crisis and Swprlse in Three Arab-JsmeU Wars; in KlaWi Knorr and Patrick Morgan. cds... Slrottgk Mflltary Surpr;se (New Brunswick: Transaction Books, 1983), pp. 111-46 at 115; Mlch.\el Hand•l, lsratl~ Polilim· bridge: Ha_rva_rd Unive1'5ity, Center for International Affair5, Occasional Paper$ no. )0, July 197)), p. )>: Em..t Stock, lsrad on thL Rood to Sinai, '949-1956 (lthaca: Cornell Univerolty p...,., •967), pp. 147·). » On July 6, 19J..4· quoted in Lebow,Bttwm1 P02Ctimd W,;zr, p.l8.Jagow likewiSe declared on July t8: "I have no wish for a preventive Wil..r# but if the fight offers itself, we dare not flinch." Luigi Albertini, Thr Originsofthr Wfr of t9t4.) vols.,tr.w. and ed. bobsey (•9s>-s7; Wcstpotl, CoM.: CrSs-86. 401->. 404-8. F..uon note& lhat •gt'fttnents are difficult when e:tates will later have i:n~tive& to renege on the te~and that

[So)

POWI!T Shift$

Agreements are easy in worlds of windows only if they freeze the two sides' relative power, thus shutting the window. But windows are usually created by forces that diplomacy cannot control, so this is rarely an option. For example, a settlement between Palestinian Arabs and Zionists proved impossible in the 193os partly because Arab leaders feared the Zionists would break the accord after Jewish numbers and power in Palestine grew.» Japan's leaders likewise doubted the value of an agreement with the United States in 1941 because they feared the United States would break it after Japan's relative power waned.40 As Bismar. AJ

Hastened or lruncnled diplomacy (explanation H3F! Ukc first-move advantages, windows force states to hurry diplomacy or to end it before it bears fruit. Declining states rush to secure agreement before their bargaining power vanishes, or to use fore4 and :;q, a scant nine to twelve days after lnchon, and before China could convey a clear warning against crossing. Actions outran diplomacy partly because the post-lnchon U.S. window required rapid exploitation.,. In 1962 it took a tense week for the U.S. and Soviet governments to devise a settlement to the Cuban missile crisis, and a faster pace might have led the United States to use force before that settlement emerged. Robert Kennedy later wrote that if the United States had been compeUed to make decisions on the first day of the crisis, U.S. policies might have been "quite different and filled with far greater risks." The days of quiet planning were "essential in choosing our ultimate course."60 During the first three days of the crisis, October t6-t8, the case for a U.S. air attack on the Soviet missiles in Cuba seemed compeUlng to most U.S. policyrnakers. and the consensus swung away from it only on October 19-20.•1 The concept of a quarantine, which President Kennedy finally adopted, was not developed until October 19," and pressure to take faster action would have left it undiscovered. President Kennedy's advisors did not recogni2e the promise of the no-invasionpledge formula that resolved the crisis until October 26-twelve days into the crisis." • lke, fDpon~ D«Jskm, p. •)?iset' also 186, 168, 1g6. " Rlc:hatd Newtadt Prmdmlial PtnDC< ro Pml H.,bor: Th< Fa!lun oflhhips wider construction, to 900.000 tons Ior the United Stat-:zoJ •• •Ss.

[89)

Owseso[War

Atlantic in the spring of1941, leaving it weaker in the PaciJic!' Finally, the June 1941 German attack on the Soviet Union reduced the danger that the Soviets would attack Japan in the north if Japan moved south. These events gave Japan a golden, never-to-be-repeated c:hanoo to conquer an Asian empire.>5 Third, the United States imposed an oil embargo on Japan in July 1941. This embargo created a huge Japanese window of vulnerability. Japan would exhaust its oil stocks in two yea.rs at most under the embargo, leaving its oildependent navy helpless.'• Fourth, Japanese planne.rs believed that any strike to the south should be launched in the winter months, when the Soviet Union could not exploit Japan's distraction to striJ. "'7·

1901

PowrrShifts We should grasp the favorable opportunlty that now presents ilsclf.• " After 0.., US. """"'buildup began and 0.., United States Imposed its oil embargo, Japan's hawks had visions of vulnerability. Prime Minister Hldeld Tojo summarized 0.., common view on November 12. 1941: "Two years from now we will have no petroleum for military~. Ships will stop moving, When I think about the strengthening or AIYI(!rican dcfenses in the So~Od>y- lllldYllnt to the Bay ol ~. Ibid.. p. " n;.. ~htn Pm libwbo""""' tlvt J•pon~ poowac aclmirw ot""' c:cnltftdS•· ., lke, Japan's Da:ision, p. 242.. 111 lriye, Origins cf1Ju. Snxmd WOTld War, p. 162. • Other factort that propelled Japan's atliK'k on the United SU.te:s inc-lude JapAn~ rabid hy· pemationalism. its perception of a fliSt·mOYe advantage (as mentioned in Chapt12r )), and its undte or th< U.S. will to fight (.. menrionr." '"' Ten days later he launched his Ughtnlng war on Poland, triggering World Warn. • Pout Ke.'lneely, TJrt Rnt and Fall oftht CrNt ~wm: Eamomic Clttmgt and Military CmtjUct from 1500 tq 2000 (New York: RAndom House. 1987). p. 296. USing the Hillman estimate for 19)0 and averaging the HiJ1man and Correlates of War estimates for other years. These fig"""' ohow the loUowillg Cermany: Britain-pluo-France military spe buildup opened a sizable U.S. window of opportunity. As wE know, window theory's main predicted outcome-preventive war-failed to rnatcrlall7.e. Are ilq other predictions also unfulftlled? Marc Trachtenberg, a diplomatic historian, studied the Cold War win· dows of 1950-54 with this question in mind."' He found that U.S. officlah shaped pol.lcy to lit predicted shifts in the military balance, a.n d often justified their policies on window grounds. When US. officials thought thE United States was weak but gaining strength, they advised caution; wher they thought US. strength had peaked, they advised belligerent policies and considered preventive war. This pattern Is observed across periods, regions. and individuals. Thus window theory's predictions are largely fulfiUed. The US. buildup of 1950-54 was the largest peacetime military builduf in US. history. lt was spurred by NSC~. a secret 1950 State Department srudy that painted an ominous Soviet threat, and by the outbreak of the Korean War, wh.ich US. officials thought Stalin had ordered, and which they feared was a precursor to wider Soviet aggression. During the buildup, U.S. dcfcnse spending more than tripled, rising from $t J·' billion (4.6 percent ol U.S. GNP) l.n 1950 to $46.6 billion (u.S percent of GNP) In '954· The buildup hit full stride in 1952, when spending reached $44.0 bill.lon (u.7 percent ol GNP), almost double US. defense spending in 1951 (S.u.s billion, or 6.9 per· cent of GNP). After 1954, U.S. defense spending dropped back to an average level of $41.9 billion and an average of 9.8 percent of GNP in 1955-.58.117 ThJs massive buildup shifted the East-West military balance toward th•

1991

CRIIstS of War

United States and caused a marked shift in U.S. perceptions of the military balance. In 1950, U.S. officials saw the United States in a period of vulnerability and weakness. For example, in December 1950 the State Department policy planning staff spoke of "our present position of military weakness," and General Omar Bradley guessed that if global war occurred in November 1950, "we might be in danger of losing."'" Later official estimates are far more optimistic: in 1953, Secretary of State John Foster Dulles spoke of "our much greater power and the Soviet Union's much greater weakness currently."' 119 U.S. policy fits the contours of this power shift. The United States pursued a cautious policy in 1950-5e1 during the perceived valley of U.S. weakness, then a much tougher policy in 1952-54, after crossing the valley."" U.S. policymakers explained their policies as responses to impending power shifts. Officials who advised the cautious 1950-51 policy argued that the United States should avoid provoking war until the military balance improved. 121 Later advocates of toughness argued that the United States should exploit its fleeting superiority. For example, the joint chiefs of staff argued in June 1954 that the United States should press its demands on the Soviet Union while the United States still held atomic superiority,' 22 and President Eisenhower briefly considered preventive war, wondering to Dulles in September 1953 "whether or not our duty to future generations did not require us to initial< war at the most propitious moment we could designate." 123 llt fortign &liltitms t>f tM Um.ted States, 1950, vol. 1, Natilllme. Ibid. On US. ...train~ ,.. o.IJo PP• 6. "· >4- 46. "' Ibid., pp. 6, 28-)1, 46. "' Ibid., p. 32. •• Ibid., pp. 2.1- >s. IJJ lh1J cautica was embodied in Sovitt acqul~ with Wtst ~rmAA rearmament in 19)l and in the M"r'C'h 10, 1951., Soviet offer to settle the Cerm•n qul!6ti0fl on the ba..sia of reunJftcatlon and neutTalluHon.lbld.• pp. 47-49-

[101)

ClusnofWar

window of opportunity (predlctlon J). Observers who believed that the U.S. window was la11:e favored more belligennt policies than th~ who ~ lleved it smaller (prediction 4). The observed s.hilts in U.S. policy an substantiaL svggesting that window theory has substantial impo~. U.S. conduct changes markedly, although not overwhelmingly, with the opening of the US. window. The United States does not move aU the way from pe~ce to war, but acros5 the globe it does move by marked degrees towMd belligerence. Doei! the failure of preventive war to occur in 1951-54 count as a OWlked test? Yes, and it shows the limits of window theory. The window of 1952-54 was only middle-$i1ed, and the costs of war seemed large to U.S. leaders. They had little fear that U.S. sovereignty would Inter be threatened if they let Soviet power grow, since U.S. superiority would fade into parity, not inferiority. The post-5talin softening of Soviet policy reassured some that war was not inevitable. U.S. officials also thought a preventive war would be rosily, and U.S. allies would bedisapproving." 0 Thus U.S. Ieadensaw a fading window of opportunity, but the opportunity was not very attractive; it would not be foUowed by vulnerability; and it probably would not be followed by a war. The case indicates that windows lose some of their potency under such conditions. They can bring states to risk war, but not to start il ~lng

Window Tluory

Many of window theory's predlctlons are borne out. Official Ideas and national policies shift both regularly and morkedly with the appearance and disappearance of windows. The regularity of these shifts, and the weakn!!$S of competing explanations for them, give window theory subs tantial credlbillty. The large size of these shifts indicates that windows have a sizable impact. Window theory's success against Ideology In the Germany 1933- 45 case olso supports an inference that it has large Importance, although the U.S. 1950-54 case shows the limits of its strength. U we grant that window theory has validity and importance, counterfactual speculation suggests that it also explains a substantial amount of international history, especiaUy if we expand it to include the effects of perceived as weU as octual windows. A plausible argumttgic Policy, t890-1987 (New Yodc: St.• Martin's, 1988), pp. U)- 16. On others see John A. Thompson, "The llx3ggerati.on or American Vulnerability: The Anatomy or 3 Trndltlon," Diplomatic Hf&t"')' t6 (Winter 1992): 23-~ •t 35-37: And W•Jtor Uppmann, U.S.Fcrtlgn Policy: Shkld of the Rcpwbli< {Boorton: Uttle, BIOW11. t94J), pp. to8-tJ, 162-4 A general survey and assessment or U.S. goopotitical thought is Robert J. Art, "A DefensjbJe DeknJe: America~ Grand Strategy arter the Cold Wu,• lntn7111tion•IS«urity 15 (Spring 1991): s-s3 at 10-23. 10 Sec Melvyn P. t..ffier, "The American Conception of N•tional Security ond the Begiro· ningsor the Cold War," AmtrU>m Hf&loriW RnitrD89 (April 1984): 3~-81 at 356-8, )JO, 374, 377; and John Lewis Gaddia, Stmtegia ojQmtoinmoll; A Criticol Apprais.U ofPOosewho have the greatest industrial bose. . .. 11lose people who hove the illdustr;at power and the power of invention and science will be able to defeat all others/' Quoted ln Paul

Kenn«iy, StrnlS of lsr.acl~ situation lndude 0... Horowil2, /m possibility that no coalition •dequat< to confront the Kremlin with8J"•tcrstrengthcould be assembled." the operative word "ill\)"' suggesting that every resource and region essenli>tlo ll1A> defense of every o!l1A>r. Ibid., p. j86. 29 Kennan. MtntDt'n t9lS-J9SO, pp. 12.9- JO· Kennan conduded that "'no one people is great enough to establish a world hegemony,'' hence ...one must not be too frightened of tho$e who aspire to world domination." lbid.• p. 1)0. » Assessment& o( the domino theory include Jeromt Slater, "Dominos in Central America: Will They Fall? Does 11 Matter?" lnltnUJIIt>riAI $t)'S«\ "!$ War llon oiSoY..-cisnty,• in &n.t.otoo c..mp;.~ onc1 Tlt'WUont-toctical de/ense dominaJ1« w;ually bolsters strategjc deftnse domlnaJlce, and tactical offen>< dominance bolst.,. str•tegic of· fensedomlnance-butnotalwayo. For.,..mple, bctl~ 1!o< C.US. Creene. t9)S). p. sS; Ceneral Omar Brodloy, "This Way Ues Peace," S.t· urdlly E""'ing P..l, October 15. 19-19, pp. J), 168-10 at)), 168; John Herz, "'dealist Internationalism and the S8-88; Mutl8 10r Bors And Wlm Smlt, ocb.. Noo·Piwliw ~., • l'rlrtcfpkof A-IWwdJon (Atn5t;on (!nu> HSB)

Expl. H5D:

Mo...ing first is moft' rew.uding (lntP HSO) - -- -.1

Expl. H5J:

More Intense :tm'\$ r.\dng (tntP H5fl

Expl. H5K!

011.,.. grawo even Jtrong«(ln!P HSK) _ _ __......,

War

- - - - --+1

HS. Verslon 3: Van Evera variant. Ch.anga hom Venlon 2.: ExplanltioN H5C. HS~ H5P, H5G, HSH#HSI, are :tdd~.

Olfensedomin.\1'\C'e

ExpL tlSA:

OpportwUatk t'X'p4l'Sion (lnu> liSA)

Expl. tlSB:

0.!a'~Si'id.lcy. ""d John Hen. £ad> propoo) • ., 5r lack of strategic depth." Quoted in Arid Lcvtce, OjJalS< •nd Ckfm~ In ISratli Mililttry D«trim- (Boulder. Westvicw. •989). p. 6o. u General Israel Thl, ...Israel's Oefense Doctrine: Background and Dynamics/" MiliiDry RAtri1, also pp. >), "'~· t88; 8ruce J, AUyn, lames C. Bligllt, and Oavid A. Weldl. "E$sentt of Revision: MOSdMr, P"'P'dl"" on o.ltr· r" M~tion from the very~· Not even aU members of the Central Committee or the cabi~ net mlnlsters were l.nvolved. lt was only posstble to discus51t wJthln a drd~ of obout tm or maybe £iiteen people, not ""'"'·· Quored in Bemd Greiner, ""The Cuban Mlasile CriJls bconsidmd; The Soviet View: Art lnt.erview with Sergo Mikoya.n,"' Diplom#tic History 4 (Spring 1990): 205- 21 at 213. " The United Sta... faUed to wam Khrushchev against deploying missiles to Cuba until September, when they were already en route. finally issued wunlng statements on September 4 o.nd •3· S... Bundy, O.ngtr •IIA Su,.;.,., p. 393 for texts. US. policymakor-s thought the SovietS were unlil00->0t; -also pp. 172-7). >OJ. )}9· Even Ceorge Kenna.n. when head of the State Deparllnent poUcy planning stoJJ (1947-49) and hes1Ct' the

•s•-SJ·

senior advisor to the ~tary of State on the policy up«ts ol atomic weapons matters, never kTiew the size or power of the U.S. atomic ans.mal. lbid.1 p. 201. ~

Stephen £. AmbrQSe, £isnthowa~ 2 vols. (New York: Simon and Schuster, 1~4)~ 2!92Jl. " Nilcita Khrushchev, Khr1lshdlro Rmomtbm: The lAst Tm•m