Richard K. Betts A Review Essay - Stanford Universityweb.stanford.edu/class/polisci211z/2.1/Betts IS...

Richard K Betts

A Review Essay

Stephen Van Evera Causes of WarPower and the Roots of ConictIthaca NY Cornell University Press 1999

War is like love it always nds a waymdashBertolt Brecht Mother Courage

tephen Van Everarsquosbook revises half of a fteen-year-old dissertation that must be among the mostcited in history This volume is a major entry in academic security studies andfor some time it will stand beside only a few other modern works on causesof war that aspiring international relations theorists are expected to digestGiven that political science syllabi seldom assign works more than a generationold it is even possible that for a while this book may edge ahead of the moregeneral modern classics on the subject such as EH Carrrsquos masterful polemicThe Twenty Yearsrsquo Crisis and Kenneth Waltzrsquos Man the State and War1

166

International Security Vol 24 No 2 (Fall 1999) pp 166ndash198copy 1999 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology

Richard K Betts is Leo A Shifrin Professor of War and Peace Studies at Columbia University Director ofNational Security Studies at the Council on Foreign Relations and editor of Conict after the Cold WarArguments on Causes of War and Peace (New York Longman 1994)

For comments on a previous draft the author thanks Stephen Biddle Robert Jervis and JackSnyder

1 EH Carr The Twenty Yearsrsquo Crisis 2d ed (New York Macmillan 1946) and Kenneth N WaltzMan the State and War (New York Columbia University Press 1959) See also Waltzrsquos more generalwork Theory of International Politics (Reading Mass Addison-Wesley 1979) and Hans J Morgen-thau Politics among Nations The Struggle for Power and Peace 5th ed (New York Knopf 1973) Afterthese standard works it is difcult to select the most important from recent decades Among themwould be Quincy Wright A Study of War 2 volumes (Chicago University of Chicago Press 1942)and Robert Gilpin War and Change in World Politics (New York Cambridge University Press 1981)Most intense examination of the causes of war has come from scholars in the realist tradition Fora provocative attack on the main arguments in this tradition see John Mueller Retreat fromDoomsday The Obsolescence of Major War (New York Basic Books 1989) For a liberal perspectivesee Michael W Doyle Ways of War and Peace Realism Liberalism and Socialism (New York WWNorton 1997) For the best overall review of contemporary literature see Jack S Levy ldquoThe Causesof War A Review of Theories and Evidencerdquo in Philip E Tetlock Jo L Husbands Robert JervisPaul C Stern and Charles Tilly eds Behavior Society and Nuclear War vol 1 (New York OxfordUniversity Press 1989)

There is one particular book to which Van Everarsquos especially begs compari-son Apparently unconcerned about having order clerks in college bookstoresconfuse his own book with a contemporary classic he chose the same title ashistorian Geoffrey Blainey In this generation Blaineyrsquos is the book most similarin scope although different in approach and style and it is one that willultimately last at least as well Blainey examined and debunked more than adozen popular notions about why wars happen and by process of eliminationsettled on one main conclusion ldquoWars usually begin when two nations dis-agree on their relative strengthrdquo At least one of the nations in conict mustmiscalculate who would succeed in a test of arms or the weaker would yieldwithout a ght Thus a clear pecking order in international relations may notnecessarily produce justice but it promotes peace A roughly even balance ofpower in contrast makes miscalculation easier The key to peace is clarityabout the distribution of power2 Blainey arrived at these spare and powerfulconclusions inductively and his book is a tapestry of unconventional questionsanalytical excursions and examples and ironic observations that social scien-tists would consider literate and lively but unsystematic

Van Evera travels a different route one that proceeds more methodically anddeductively and aims to honor the canons of social science He arrives at aplace apparently different from Blaineyrsquos but actually similar Van Evera ar-gues for ve hypotheses but concentrates on one ldquoWar is more likely whenconquest is easyrdquo3 This seems to contradict Blaineyrsquos view that imbalancefavors peace because weakness encourages compliance with the strongerpower In the course of the book however Van Evera makes clear his convic-tion that the main problem is quite different from the one stated in the initialhypothesis it is not that conquest actually is easy but that most often it is notyet is mistakenly perceived to be easy This amendment is quite compatible withBlaineyrsquos bottom line

Van Evera also has an ambitious aim Where Blainey remains strictly empiri-cal Van Evera seeks to provide prescriptive analysis that can show policymak-ers how to manipulate causes of war and deploy countermeasures againstthem As a good realist he does not claim that he can make war obsolete buthe does imply that if statesmen understood the analysis in the book and actedon it searches for security could be relaxed and more wars could be foregone

2 Geoffrey Blainey The Causes of War 3d ed (New York Free Press 1988) p 293 see alsopp 109ndash1143 Stephen Van Evera Causes of War Power and the Roots of Conict (Ithaca NY Cornell UniversityPress 1999) p 4 Subsequent references to this work appear parenthetically in the text

Must War Find a Way 167

Because it is mainly mistaken beliefs about threats to security that account forwar educationmdashreading this bookmdashshould contribute to a solution

This is a good book It covers a lot of territory in a clear and organizedmanner It focuses on a coherent set of issues and schools of thought andevaluates them systematically It distills much of the conceptual and theoreticalapparatus that evolved in strategic studies during the Cold War and appliesthe ideas broadly It makes strong arguments The book takes off from theessential premise of structural realismmdashthat international anarchy is the per-missive cause of warmdashand investigates more specic ways in which beliefsabout the options provided by certain kinds of power may yield sufcientcauses Van Evera does not demonstrate that any single cause of war issufcient nor does he claim that any except perhaps onemdashthe ease of con-questmdashis likely to be His analysis does provide reasonable grounds to believethat combinations of some of the causes he examines may well be sufcient inmany cases The bookrsquos reach turns out to exceed its grasp but it is better toaim high and fall short than to fulll a trivial mission

Remember all this as you read on A review essay cannot justify article lengthif it simply lauds its subject It must do one of two things It can use theassignment as a pretext for an excursion of the reviewerrsquos own and ignore thework that is supposed to be examined This common ploy is irresponsible Thealternative is to dwell on the limitations of the work under review That is whatthis essay does This emphasis would be unfair to Van Evera if readers forgetthe generally favorable regard within which my criticisms are wrapped

Social science has more inherent limitations than natural science and allimportant works have important limitations Because it is easier for academicsto score points by nding such limitations than by developing unassailabletheories of their own every major book becomes a target This one goes outof its way to draw re however by claiming to offer a ldquomaster theoryrdquo Theclaim overreaches for four main reasons

First the volume aims for great prescriptive utility but its conceptual appa-ratus is deeply rooted in two events of one century (our own) World War Iand the maturation of nuclear deterrence between the United States and theSoviet Union This may not limit the lessons for the twenty-rst century if weare concerned only about interstate wars among great powers but it is lesscertain that this frame of reference is the best guide to understanding themajority of contemporary wars intrastate struggles among groups vying todetermine the scope of communities and the character of regimes

Second the book does too much The master theory (offense-defense theory)turns out to roll so many things into the denition of one master variable (the

International Security 242 168

offense-defense balance) that the distinction between the authorrsquos theory andthose he seeks to surpass erodes The improvement over theories that focus onpower in general such as Blaineyrsquos becomes hard to discern Van Everarsquos bookdoes not make the concept of offense-defense balance identical to relativepowermdashhis denition does leave out some signicant aspects of the lattermdashbutit comes too close for comfort

Third the book does too little Like most realist theories it avoids consider-ing the substantive stakes of political conict Such issues may be secondaryto the main causes that lie in the distribution of power but they are still tooimportant to be excluded from a master theory especially one that deals withcost-benet calculations of attackers and defenders Van Evera piles his mostimportant arguments onto the question of when and why states believe thatconquest is ldquoeasyrdquo This is a vague adjective on which to rest a master theoryIt would be more specic and relevant to say instead when states believe thatconquest of a desired objective is achievable at acceptable cost That in turndepends not only on the odds of military success but on the value that thegovernment or group places on the objective As with structural realists ofother stripes there is an apolitical avor to some of the assumptions thatanimate the analysis

The fourth signicant limitation is the bookrsquos normative bias in favor of thestatus quo and the prevention of war Perhaps all civilized people in theenlightened and satised West can agree on these aims They contaminateanalysis however when they spill over into an empirical assumption thatstates normally seek security and cause problems for others because of mis-guided notions that security requires expansion This premise is really a hy-pothesis that remains to be demonstrated and contradicts the traditional realistview that statesrsquo ambitions grow as their power expands4 It is all the moresignicant because Van Everarsquos fundamental diagnosis is like Pogorsquos (ldquoWe havemet the enemy and he is usrdquo) Given that real insecurity is rare misperceptionto the contrary becomes ldquoa self-fullling prophecy fostering bellicose policiesrdquo

4 See Fareed Zakaria ldquoRealism and Domestic Politicsrdquo International Security Vol 17 No 1(Summer 1992) pp 191ndash192 196 and Zakaria From Wealth to Power The Unusual Origins ofAmericarsquos World Role (Princeton NJ Princeton University Press 1998) Van Evera cites the UnitedStates as a good test of his theoryrsquos prediction that states with high security will be less aggressiveHe dismisses Zakariarsquos alternative explanation (weakness of the state apparatus) by saying that itdoes not explain the ldquodecline of US activism after 1815 or after 1991rdquo (pp 182ndash184 184 n 224)US activism however did not decline in these periods In the decades after 1815 the United Statesdoubled in size taking half of Mexico Washington has been remarkably active since disposing ofthe Soviet empire ghting two medium-sized wars in less than a decade since the Cold War endedengaging in several smaller interventions on three different continents and sporting the mantle ofworld orderer with alacrity


As a result ldquothe prime threat to the security of modern great powers is themselvesrdquo (pp 191ndash192 ellipsis in original)

ldquoSecurityrdquo is a polymorphic value as hopelessly slippery as ldquonational inter-estrdquo Van Evera does not dene ldquosecurityrdquo in detail but he implicitly identiesit with the political status quo especially in terms of what states control whatterritory Dispossessed groups with grievances however have no reason toidentify security that way and no more reason to see resort to force for revisingan unjust status quo as illegimate than Van Evera has to see use of force fordefending the status quo as legitimate Or greedy groups may want to takewhat others hold irrespective of whether what they hold themselves is secure

The Cold War hard-wired the value of stability in the minds of Westernstrategists but stability is as normatively charged as any value Sanctifyingstability as EH Carr argued represents the international morality of theldquohavesrdquo against the ldquohave-notsrdquo5 If war is always a greater evil than injusticeand if prevention of war is the only value at issue the solution is simplepacism In an important sense as Clausewitz notes wars are begun not byinvaders but by defenders who resist them6 (As George Quester once pointedout to me this is why we say World War II began in September 1939 whenPoland resisted the German invasion rather than in March 1939 when Czecho-slovakia did not) If the aim of avoiding war is subordinate to some othernormative value and if a book on causes of war is to be one for all seasons itis a leap of faith to accept that the uniform answer is the one Van Evera preferssecuring stability through recognition that defense is easier than attack

The book proceeds by examining ve hypotheses (A handy appendixpp 259ndash262 lists the numerous corollaries that go with the main hypotheses)Each hypothesis gets a chapter which reviews the arguments relates VanEverarsquos impressions of what his years of research have suggested to him aboutthe bulk of evidence that bears on the hypothesis and presents a few casestudies to test the hypothesis Four of the ve hypotheses are dealt with in therst half of the book The core of the book is the second halfmdashthe explorationof the fth hypothesis and the body of offense-defense theory associated withit This essay comments on the arguments in the book in the same order andproportion

5 Carr Twenty Yearsrsquo Crisis pp 53ndash55 79ndash846 Carl von Clausewitz On War ed and trans Michael Howard and Peter Paret (Princeton NJPrinceton University Press 1976) p 377


The First Four Hypotheses

The early chapters present a good survey of how several aspects of power andperceptionsmdashespecially what Van Evera calls the ldquone-grainedrdquo as opposed togross structure of powermdashcreate incentives for war The author strives toencase his arguments in social scientic rigor but he does not shrink fromrecognizing some of the limitations of evidence that make historians skepticalof sleek generalizations Indeed Van Everarsquos forthrightness in laying out thelogic and the illustrations that he believes test it eases the criticrsquos task7

The rst four hypotheses (p 4) are that war is more likely when

ldquostates fall prey to false optimism about its outcomerdquoldquothe advantage lies with the rst side to mobilize or attackrdquo (This hypothesisencompasses ldquostability theoryrdquo and concerns about how incentives forpreemptive attack can be inherent in certain congurations of capability)ldquothe relative power of states uctuates sharplymdashthat is when windows ofopportunity and vulnerability are largerdquo (This encompasses concerns abouthegemonic transitions and impulses to preventive war)ldquoresources are cumulativemdashthat is when the control of resources enables astate to protect or acquire other resourcesrdquo

The analysis of the fourth hypothesis on cumulativity of resources is mostconvincing It engages the question of economic costs and benets of war andthe logic of imperialism and rests on the notion that if the cost of conqueringterritory to extract resources exceeds the utility of the resources ldquocumulativityis negativerdquo (p 106) and incentives for war are reduced In agricultural erascumulativity was high but in recent times cumulativity has declined becauseknowledge-based economies are harder to loot than are smokestack economies

7 There are questions about Van Everarsquos method that this essay does not have room to explorein detail He sets out criteria for case selection and theory testing that are somewhat confusing asstated in the book (eg saying that the case studies ldquoproceed by comparing the case to normalconditionsrdquo without indicating what a normal condition is supposed to be) notes that orthodoxsocial science methodology would count his tests as weak because they do not adhere to certainrules and dismisses the criticism by saying that he has ldquonever found these rules usefulrdquo butwithout saying why (p 12) He also reports that he surveyed thirty wars (mostly modern ones)in background research (p 13 n 24) but does not indicate why he chose those particular thirty(Why for example the FalklandsMalvinas War but not the much larger Chaco War or theIndochina War but not the more consequential Russian Spanish or Chinese Civil Wars) Forclarication of Van Everarsquos views on methodology see his Guide to Methods for Students of PoliticalScience (Ithaca NY Cornell University Press 1997)


(pp 114ndash115) Van Evera presents a balanced and reasonable argument thatconquest still pays sometimes but less often than in the past

The rst three hypotheses are more problematic They are compromised byunderwhelming evidence or prove less convincing as explanations of war thanof strategy and timingmdashexplanations of ldquohowrdquo and ldquowhenrdquo more than ofldquowhetherrdquo to ght

unsurprising optimismVan Everarsquos rst hypothesis says that governments are more likely to ght ifthey believe they will win handily and that they are often mistaken in thisbelief This is convincing but not surprising especially for one who has readBlainey Yes false optimism ldquopreceded every major war since 1740rdquo (p 16) Itmust also have preceded most wars before 1740 but so did valid optimismWhen a war has a winner and a loser Van Everarsquos proposition and its oppositeare both true by denition the victorrsquos optimism proves correct and the loserrsquosmistaken If neither side is optimistic about its chances on the other hand waris less probable because neither side sees much to gain from starting it

The hypothesis is more relevant for understanding pyrrhic victories orstalemates that expose false optimism on both sides or cases in which attackersstart wars that they expect to win but wind up losing The hypothesis isimportant if such cases are more typical of war than cases in which one sidewins at acceptable cost that is if most cases are ones in which optimism byeither of the contenders is invalid Determining that would require a substantialempirical project World War I is a case of that sort and it is omnipresent inVan Everarsquos theoretical apparatus and empirical analysis but being an impor-tant case does not make it typical

How much does it help to cite a case like Frederick the Greatrsquos invasion ofAustria (p 17) where the defender rsquos optimism proves wrong and the at-tackerrsquos right That combination conrms common sense Or what about caseswhere indications of false optimism are awash in other evidence of pessimismsuch as US decisions on Vietnam (pp 17 23)8 A problem that aficts analysesof government decisions in many cases is that one is likely to nd ample

8 Van Evera says that ldquoUS ofcials recurrently underestimated their opponents in Vietnamrdquo butthe record in general does not support this See Leslie H Gelb with Richard K Betts The Irony ofVietnam (Washington DC Brookings 1979) and HR McMaster Dereliction of Duty (New YorkHarperCollins 1997) Van Evera cites a Joint Chiefs of Staff estimate that 205000 US troops wouldbe needed to achieve goals in 1961 but does not mention that US troops were not committed inforce until four years later when communist forces were stronger and army and marine corpsestimates of needed US manpower ranged between 500000 and 700000


evidence of both optimism and pessimism as policymakers confronting a clashof incentives and constraints are racked by ambivalence When historiansknow that whatever sample of available evidence is incomplete how rm alesson can we draw from the many cases in which say ve statements reect-ing optimism can be stacked against three reecting pessimism

stability theoryExploring the second hypothesis Van Evera amends and salvages ldquostabilitytheoryrdquo and its emphasis on ldquorst-move advantagesrdquo He argues that incen-tives to mobilize or strike rst do not cause war in the manner of ThomasSchellingrsquos image of ldquoreciprocal fear of surprise attackrdquo9 but that rst-moveadvantages cripple peace-seeking diplomacy by leading countries to concealtheir grievances capabilities and plans (p 39)

Van Evera asserts that real rst-move advantages are rare while the illusionof such advantage is common (p 71) This could be true but the point is notdemonstrated by the analysis In one paragraph he gives a list of cases hebelieves support his assertion (pp 71ndash72) but does not present the bases forthat conclusion Instead he presents three case studies in detail that togetherdo as much to question as to conrm his arguments that rst-move advantagesare rare and that belief in them prompts war World War I Chinarsquos entry intothe Korean War and the 1967 Arab-Israeli War In these three cases there werereal rst-move advantages Van Evera reports that on the eve of World War IFrench General Joseph Joffre warned ldquothat France would lose fteen to twentykilometers of French territory for each day mobilization was delayedrdquo (p 71n 31) as if Joffre was wrong If a mobilization lag would make no differencein where the lines wound up Van Evera does not say why As it was bySeptember 1914 the Germans penetrated close to 100 kilometers into Francemdashabout 20 kilometers from Parismdashand they managed to ght the rest of the waron French territory That seems to be a signicant advantage from moving rstand fast Similarly in the other cases used to explore the hypothesis theChinese army rocked the 8th Army back into the longest retreat in US militaryhistory in 1950 and Israel clobbered Egypt and Syria in 1967

Perhaps Van Everarsquos point should be a different onemdashthat rst-move advan-tages rarely are great enough to win a war If so it would be clearer to focuson decisive rst-move advantages The Germans did lose World War I four

9 ldquoIf surprise carries an advantage it is worthwhile to avert it by striking rst Fear that the othermay be about to strike in the mistaken belief that we are about to strike gives us a motive forstriking and so justies the otherrsquos motiverdquo Thomas C Schelling The Strategy of Conict (NewYork Oxford University Press 1963) p 207


years after they overran northern France But the Israelis won decisively in1967 and the Chinese succeeded in settling the war in 1953 on terms thatsecured their gains on the battleeld and ratied the expulsion of UnitedNations (UN) forces from North Korea (which they had almost entirely liber-ated before the Chinese threw them back) To his credit Van Evera recognizesthat rst-move advantages were real in at least the 1950 and 1967 cases buthis argument still claims more than he demonstrates Additional data mightsupport his views but the evidence selected for presentation in the chapterdoes not demonstrate that even decisive rst-move advantages are rare

Whether imagined or decisive rst-move advantages are hard to cite ascauses of two of the three wars Van Evera examines under this hypothesisRather they were causes of strategy The Chinese would not have refrainedfrom intervening in Korea if General Douglas MacArthur had not made iteasier for them by splitting his forces in the advance to the Yalu Van Everaoffers a convoluted rationale for why Beijing could have decided to stay outif US signals had been different but notes nevertheless that Mao Zedongdecided to intervene as soon as US forces entered North Korea before theyapproached the Yalu (pp 56ndash61) Similarly he says ldquoEgypt and Israel hadmany reasons for war in June 1967 and probably would have come to blowsabsent a rst-strike advantagerdquo (p 68)

windowsThe third hypothesis is that uctuations in power cause war by creatingwindows of opportunity and vulnerability and thus incentives for preventiveor preemptive action Declining states are tempted to launch preventive warsand take bigger risks diplomatic alternatives to war are weakened becausedeclining states cannot make credible offers and rising states conceal theirgrievances to avoid triggering preventive war and if the balance of powershifts peacefully the new disequilibrium may produce war if the fallen staterefuses to yield privileges This chapter produces a generally sensible discus-sion

Again however the three cases Van Evera chose to test his hypothesisprovide very mixed evidence two of the three do not provide good supportThe best of the three and a quite powerful one is Japanrsquos 1941 attack on theUnited States (pp 89ndash94) Without the windows created by Washingtonrsquos com-bination of escalating demands and lagging mobilization Japan might neverhave struck rst In regard to Germany in 1939 however ldquowindow theoryrdquoproves convincing as an explanation of the timing rather than of the decisionfor war The book provides no reason to disbelieve the widely held view that


Hitler wanted war from the beginning If Hitler was determined to have a warhis strategic decisions were about when not whether Van Evera argues(pp 94ndash99) that this was true only of German designs in the East that Hitlerrsquosaims did not require war against France and Britain The Alliesrsquo guarantee toPoland however meant that Hitler could not take Poland without a war inthe West and could not move against the Soviet Union without rst defeatingFrance Alliances account for his decisions to attack in the West windows mayaccount for when he did so A cause of strategy is not the same as a cause ofwar

The third case is the US military buildup in the early 1950s which VanEvera discusses as a cause of increased American belligerence This confusescause and effect The United States did act more cautiously when it was weakin 1950ndash51 and more toughly when a tripled defense budget gave it morepower a few years later It did not however become more belligerent becauseit became more powerful but the reverse The Korean War prompted themilitarization of containment The desire for better strategic options to supportdeterrence produced the buildup which enabled prior policy to be pursuedwith more condence Actions could catch up with aims This example is nota case where a window caused war or even one where it caused belligerencebut where the buildup was caused by the desire to avoid warmdashand suc-ceeded10 This example belongs in a different volume Causes of Peace

Master Theory Overmilking Nuclear Deterrence and 1914

The heart of Van Everarsquos book is the promulgation of ldquooffense-defense theoryrdquo(henceforth ODT) as a ldquomaster key to the cause of international conictrdquo aldquomaster theoryrdquo that ldquohelps explain other important causesrdquo and provides ldquothemost powerful and useful Realist theory on the causes of warrdquo (pp 190 117emphasis added) So much for Thucydides Machiavelli and Hobbes or CarrMorgenthau and Waltz This is no mean claim The claim is all the more boldbecause it feeds on a theory conceptually rooted in very recent times ODT hasbeen overwhelmingly inuenced by the frame of reference that developedaround nuclear weapons and the application of deterrence theorems to theconventional case of World War I

10 Van Evera suggests a more dangerous result when he cites evidence that US leaders thoughtabout preventive war in this period (p 100) On this see especially Marc Trachtenberg History andStrategy (Princeton NJ Princeton University Press 1991) pp 103ndash107 Indications that ofcialsruminated about preventive war does not mean that it reached the point of being considered aserious option for deliberation and decision at the highest levels


nuclear thinkingMilitary historians and budding arms controllers in the interwar period havewritten on the import of shifting advantages of attack and defense The recentevolution of theory on the subject however reects a systematic and sustainedfocus on the idea of strategic stability which crystallized only in the nucleardeterrence theory of the Cold War American deterrence theorists becamepreoccupied with how specic congurations of weaponry and targeting op-tions might autonomously induce war or peace The key was how combina-tions of technology could provoke or constrain decisionmakers by the prospectof how a nuclear engagement would turn out depending on who red rst

The nuclear framework provided a simple model of mutual deterrence basedon a few variablesmdashtypes of targets types of forces and their relative vulner-abilitymdashand clear distinctions between offensive and defensive operationsNone of the weapons that gured in standard scenarios and calculations weredual-purpose For example intercontinental ballistic missiles (ICBMs) could bered only to attack the adversaryrsquos territory not to intercept an adversaryrsquosattack on onersquos own territory Attacking and defending forces could supporteach other (counterforce rst strikes could boost the effectiveness of antiballis-tic missile (ABM) defenses by drawing down the number of the victimrsquosretaliating forces that had to be intercepted) but the scenarios of targeting andforce interaction that dominated nuclear theorizing were few and uncompli-cated compared to the historical variety of conventional warfare The literatureon nuclear strategy drew analysts toward a xation on autonomous militarycauses of war and peace and away from attention to the substantive politicalissues and motives in international disputes that had traditionally preoccupieddiplomatic historians and political scientists If the seedbed of deterrencetheory does not account for the production of recent ODT it certainly accountsfor the intellectual receptivity of its consumers in the peak period of attentionin the 1980s

Common to academic analysis of the nuclear era were convictions thatstrategic stalemate was inherent in the nature of military technologies and forcestructures of the times resort to war (at least nuclear war) was irrationalbecause it was tantamount to suicide recognition of this reality should preventwar the prime danger to peace was misperception and miscalculation andstability might be enhanced through measures of arms control that reinforcedcondence in mutual deterrence Bits and pieces of such ideas can be un-earthed from earlier writing about military matters but coherent applicationof them to conventional strategy became prominent only in the surge of debate


about the balance between the North Atlantic Treaty Organization (NATO) andthe Warsaw Pact during the Cold War

Van Evera and other contemporary exemplars of ODTmdashGeorge QuesterRobert Jervis Jack Snydermdashwere all steeped in nuclear deterrence theorybefore venturing into analysis of conventional military strategy11 The compan-ion image of the ldquosecurity dilemmardquo also gured heavily in ODT (it is evenevident in postndashCold War applications to unconventional civil wars)12 TheCold War agenda made ODT especially appealing Most obviously in the formpresented by Jervis the theory offered a way out of the security dilemma13 Ifcapabilities could be congured to the advantage of defense and Washingtonand Moscow could recognize and codify the situation they would be lesssusceptible to inadvertent war or expensive arms competitions arising from aspiral of misperceptions If only the conventional balance could be made likethe nuclear balance competition could be stabilized

Given the strength of mutual nuclear deterrence the main debates aboutstability in the Cold War devolved onto the conventional military balancebetween NATO and the Warsaw Pact Doves wanted to nd an excuse for lessreliance on the threat of deliberate nuclear rst use to deter Soviet conventionalattack and warmed to analyses suggesting that NATOrsquos conventional forceswere less hopelessly incapable of holding the line than typically assumedhawks wanted stronger military capabilities across the board and denied thedefensive adequacy of Western forces on the central front ODT reinforced one

11 George Quester Offense and Defense in the International System 2d ed (New Brunswick NJTransaction 1988) Robert Jervis ldquoCooperation under the Security Dilemma World Politics Vol 30No 2 (January 1978) pp 167ndash214 and Jack L Snyder The Ideology of the Offensive Military DecisionMaking and the Disasters of 1914 (Ithaca NY Cornell University Press 1984) These versions aremore articulated than other forms of offense-defense theory that predated the nuclear era Forexamples of how frameworks redolent of deterrence theory were superimposed on World War Isee the section of articles on ldquoThe Great War and the Nuclear Agerdquo in International Security Vol9 No 1 (Summer 1984) Scott Sagan also argued within the terms of reference of deterrence theoryto criticize interpretations by Van Evera and others who idealized defense dominance and to arguefor benets of offense dominance Sagan ldquo1914 Revisitedrdquo International Security Vol 11 No 2 (Fall1986) pp 151ndash176 It is also important to note that contributors to ODT do not all agree aboutmany things When this essay makes general statements about ODT implications it is (unlessotherwise specied) referring to implications consistent with Van Everarsquos12 Barry R Posen ldquoThe Security Dilemma and Ethnic Conictrdquo Survival Vol 35 No 1 (Spring1993) pp 27ndash57 and Barbara F Walter and Jack Snyder eds Civil Wars Insecurity and Intervention(New York Columbia University Press forthcoming)13 Jervis ldquoCooperation under the Security Dilemmardquo pp 186ndash214 In principle ODT is inde-pendent of normative preferences for stability and defense dominance Its association with thesepreferences may have been stronger during the Cold War given the prospective costs of warbetween the superpowers than it will be in the future


side of the debate by offering ideas for affordable ways to make a Sovietarmored offensive as unpromising as a nuclear strike

Events of 1914ndash17 seemed to illustrate a conguration of capabilities thatshould have done this had the defense-dominant nature of that congurationbeen recognized Failure of the Schlieffen Plan years of inconclusive carnageon the western front and costs far greater than the belligerents would havechosen to bear had they foreseen them showed that understanding of theoffense-defense balance should have led to restraint in the crisis of 1914 If thechances of successful conventional offensives on the central front in Europe inthe Cold War could be reduced to what they had been in 1914 and this couldbe demonstrated to political leaders East and West could compete in peaceImplicit in the promise of ODT was the hope for a technical x to the ColdWar military competition

The fascination of many realists with ODT was not shared by all especiallyReaganite idealists who saw the Cold War problem as Soviet aggressive mo-tives rather than unfounded mutual fear To them seeing the security dilemmaas the essence of the problem implied moral equivalence The strategic taskwas not to reassure the Russians but to press them These skeptics rejected thecanonical wisdom about the benign nature of mutual assured destruction(MAD) and promoted nuclear counterforce capabilities and across-the-boardincreases in conventional forces This view carried the day in the policy arenabut was a minority position in academia where satisfaction with stable nucleardeterrence prevailed

As social science theories go nuclear deterrence was exceptionally elegantso the ODT about conventional military competition that fed on it evolved intoa hot genre in strategic studies As such it naturally spawned criticism Oneskeptical article by Jack Levy appeared at the same time as Van Everarsquos initialanalysis of World War I14 Levy faulted the genre for the variety of terms inwhich the offense-defense balance had been dened ldquothe defeat of enemyarmed forces territorial conquest protection of population tactical mobilitythe characteristics of armaments attackdefense ratios the relative resourcesexpended on the offense and defense and the incentive to strike rstrdquo He alsofaulted ODT for tautological hypotheses circular denitions confused con-cepts and other aws and concluded that ldquothe concept of the offensivedefensive balance is too vague and encompassing to be useful in theoretical

14 Jack S Levy ldquoThe OffensiveDefensive Balance of Military Technologyrdquo International StudiesQuarterly Vol 28 No 2 (June 1984) pp 219ndash238 and Stephen Van Evera ldquoThe Cult of theOffensive and the Origins of the First World Warrdquo International Security Vol 9 No 1 (Summer1984) pp 58ndash107


analysisrdquo Critics chipped away at the technological formulations of the of-fense-defense balance especially challenging the notion that particular weap-ons supported either attack or defense as consistently as they did in nucleartheory15 Others counterattacked on behalf of ODT16 Van Everarsquos book engagesonly some of the critiques and usually in brief asides

One problem in this body of theory is confusion of dimensions of analysisin the denition of the offense-defense balance Before the nuclear era mostconceptions of the balance were in terms of military outcomesmdashwhich sidewould have the operational advantage when attackers crashed into defendersin a combat engagement Offense was identied as attack by ground naval orair forces against an adversaryrsquos territory (either to seize ground or to destroyassets behind the lines) and defense as blocking the attacking forces (holdingground or intercepting bombers or missiles before they can destroy theirtargets) A balance was considered ldquooffense dominantrdquo if other things beingequal attacking forces were more likely to succeed in advancing or destroyingtheir target than defending forces were to keep them from doing so

The newer variants of ODT confuse the question by sometimes shifting thedenition of offense-defense balance to the level of political outcomesmdashtheeffects of operational advantages on national decisions about whether to initi-ate war Confusion is most obvious in the inversion of terms when nuclearweapons are considered Oxymoronically most contemporary exponents ofODT cite the nuclear offense-defense balance as defense dominantmdasheventhough no effective defenses against nuclear attack yet existmdashbecause mutualvulnerability to the operational dominance of offensive forces prevents politi-cal leaders from deciding to start an engagement (The traditional approachcalls the nuclear balance offense dominant because when red in an engage-ment attacking missiles would accomplish their mission and helpless defend-ers would be destroyed) Even if the oxymoronic denition were reasonablefor the nuclear case considered alone it muddies the conceptual waters It isinconsistent to characterize both the 1914 conventional balance and the ColdWar nuclear balance as defense dominantmdashwhich most offense-defense theo-rists domdashbecause that is true in terms of combat interactions for the rst but

15 Levy ldquoOffensiveDefensive Balancerdquo pp 222ndash223 226ndash227 22916 Sean M Lynn-Jones ldquoOffense-Defense Theory and Its Criticsrdquo Security Studies Vol 4 No 4(Summer 1995) pp 660ndash691 Some versions of ODT especially recent ones that have absorbedearlier critiques do not assume that particular weapons can be categorized as inherently eitheroffensive or defensive For the most recent critiques and responses see James W Davis Jr BernardI Finel Stacie E Goddard Stephen Van Evera and Charles L Glaser and Chaim KaufmannldquoCorrespondence Taking Offense at Offense-Defense Theoryrdquo International Security Vol 23 No 3(Winter 199899) pp 179ndash206


not the second or in terms of political effect for the second but not the rst17

It would be more accurate to call the nuclear situation ldquopeace dominantrdquo if thecriterion is decisions in peacetime rather than results in wartime Moreoverthe inverted conception has other limitations

None of those who purveyed the inverted denition in the 1980s approvedof its purest application to conventional warfare Samuel Huntingtonrsquos pro-posal to shift the logic of NATOrsquos conventional deterrent from denial topunishment from a defensive blocking strategy to a retaliatory counteroffen-sive Huntington transposed the scorpions-in-a-bottle logic of nuclear deter-rence directly to conventional strategy arguing that a threat to retaliate againstSoviet assets in Eastern Europe was a more potent deterrent to Soviet adven-turism than an uncertain capability to hold the inner-German border againstSoviet tanks18 Proponents of ODT were uniformly aghast at Huntingtonrsquosproposal and insisted on a NATO conventional strategy that was literallyrather than oxymoronically defensive

ODTrsquos benign view of the effects of nuclear weapons on the danger of warrests on an old belief that MAD guarantees military restraint a belief that wasnever universally shared among analysts or policymakers Those who lackedcondence in MADrsquos pacifying effect focused on the logical tension betweencondence in mutual deterrence between the superpowers at the nuclear leveland condence in mutual deterrence at the conventional level This debateabout the credibility of threats to escalate bedeviled NATO strategy for morethan three decades and was never resolved If mutual nuclear deterrence wasso solid that neither side would ever strike rst then the world was safe forconventional war the side with conventional superiority could win while bothsidesrsquo nuclear forces held each other in check To threaten deliberate escalationeven unto intercontinental strikes in order to deter conventional attackmdashNATOrsquos ofcial doctrine for decadesmdashmutual nuclear deterrence must havean exception Either mutual fear of nuclear annihilation is so strong that neitherside would ever strike rst or it is not Theorists and policy analysts spilledgallons of ink in arcane and contorted attempts to square this circle and never

17 The balance in 1914 was defense dominant only in operational terms because decisionmakerswere not constrained by accurate anticipation of what would happen when the forces collidedThe reality became evident only with the test of war One might then say that the offense-defensebalance should be dened in terms of perceptionmdashwhat decisionmakers believe would happenmdashwhich is indeed the basis for characterizing the nuclear balance given that there has been no testof war for this case If belief rather than demonstrated reality is the criterion however the 1914balance must be cited as offense dominant18 Samuel P Huntington ldquoConventional Deterrence and Conventional Retaliation in EuroperdquoInternational Security Vol 8 No 3 (Winter 198384) pp 32ndash56


succeeded without resorting to more oxymoronic solutions like ldquothe rationalityof irrationalityrdquo or simple assertions that the ldquobalance of resolverdquo would favorthe side defending the status quo19

ODT glosses over the fact that the view of MAD as pacifying or defensedominant prevailed only among academics and diplomats and lost where itcountedmdashwithin policy circles that determined procurement force postureand war plans20 (This will be a problem for students reading Van Everarsquos bookA decade past the Cold War few of them now know any more about nuclearstrategy than they read in the works of theorists and believe that MADcaptures all they need to understand about the subject) The glossing mayreect Van Everarsquos barely concealed contempt for the foolishness of those whohave sought alternatives to reliance on MAD a contempt that occasionallyleads him to obiter dicta or errorsmdashfor example his charge that Air Force Chiefof Staff Curtis LeMay ldquourged general nuclear war during the missile crisisrdquo(p 141 n 90)21

The biggest disappointment of this volume is the brevity and pugnacity ofVan Everarsquos chapter on nuclear weapons which is far less scrupulous inargumentation than the rest of the book It reads like a polemical bit of armwaving written for some other purpose and tacked on to the book This chapterignores all the old unresolved debates about nuclear strategy and conventionalstability says next to nothing about real dilemmas and simply asserts theineluctability of MAD It adds nothing to the literature on nuclear strategy orto the arguments in the rest of the book For example Van Evera proclaimswithout any consideration of counterarguments that ldquoconquest among greatpowers is almost impossible in a MAD world States also can better defendthird parties against aggressors under MAD the United States deployedtroops to Germany as a tripwire during the Cold Warrdquo (p 246) It is as if historystopped in 1957 before the long and intense debates over extended deterrence

19 For criticism of this assertion see Richard K Betts Nuclear Blackmail and Nuclear Balance(Washington DC Brookings 1987) pp 133ndash14420 Aaron Friedberg ldquoA History of US Strategic lsquoDoctrinersquomdash1945ndash1980rdquo Journal of StrategicStudies Vol 3 No 3 (December 1980) pp 37ndash71 Desmond Ball ldquoUS Strategic Forces How WouldThey Be Usedrdquo International Security Vol 7 No 3 (Winter 198283) pp 31ndash60 Fred Kaplan TheWizards of Armageddon (New York Simon and Schuster 1983) Desmond Ball and Jeffrey Richelsoneds Strategic Nuclear Targeting (Ithaca NY Cornell University Press 1986) and Scott D SaganMoving Targets Nuclear Strategy and National Security (Princeton NJ Princeton University Press1989) The benign view dominant among academics was far from a consensus among analysts ingeneral (especially ones involved in government) Among those dissatised with MAD were FredIkleacute Paul Nitze Albert Wohlstetter Colin Gray Donald Brennan and William Van Cleave21 He cites Raymond Garthoff Reections on the Cuban Missile Crisis rev ed (Washington DCBrookings 1989) p 94n This passage actually refers to LeMayrsquos support for the conventionalair-strike option against bases in Cuba


exible response the credibility of NATO doctrine and prolonged conven-tional defensemdashwhich the author does not even mention let alone analyze

offense-defense theory opposes defenseIronically Van Evera is most adamant against the idea of real nuclear defensedominancemdashthat is a world in which defense dominates at the operationallevel of combat effect He coins the cute acronym BAD (for ldquoBoth Are De-fendedrdquo) to contrast this sort of world to MAD and opines that ldquoif BAD isever achieved it will not last long One side will soon crack the other rsquosdefensesrdquo Perhaps but the same could as easily be said of any defensedominance constructed at the level of conventional combat Van Evera thenrails that ldquothe lives of states in BAD will be brutal and short They will conquerand be conquered at a fast pacerdquo (p 251) As to why he gives not a clue Thecavalier quality that diminishes his offensive against nuclear defenses is par-ticularly unfortunate because academic conventional wisdom opposing ballis-tic missile defense is now on the losing end of policy debate it dominatedpolicy in the 1970s and fought a successful delaying action against RonaldReaganrsquos Strategic Defense Initiative in the 1980s but has lost its inuence inthe 1990s Even Bill Clintonmdashwho populated his foreign policy team withliberals and arms controllers who would fall on their swords to save the ABMtreatymdashhas accepted that the United States should deploy antimissile defenses

There are still convincing reasons for continuing to oppose investment inactive defenses (although there is no excuse for opposing passive civil de-fense)22 For example no one has yet demonstrated how any system designedto date will overcome efforts of a major power to saturate it with attackingforces or of a minor power to circumvent it by using alternate means ofdelivery The case is not nearly as clear however as it was during the ColdWar For example the United States and Russia might hypothetically agree toeld limited and comparatively cheap active defenses that would not besufciently effective to negate their deterrents against each other (given thateach could retain 1000 or more deliverable weapons) but that could bluntattacks by third countries with small numbers of missiles North Korea IraqIran or Libya will not have the resources to eld more than a few ICBMs EvenChina in 1999 has fewer than twenty ICBMs to deploy enough to saturate aUS or Russian defense might require a tenfold increase a daunting expense

22 Richard K Betts ldquoThe New Threat of Mass Destructionrdquo Foreign Affairs Vol 77 No 1(JanuaryFebruary 1998) pp 36ndash40


for a country with other priorities This sort of speculation may not justifymissile defense programs but Van Evera does not consider any of the argu-ments for them in the chapter that condemns them out of hand

If old-fashioned mainstream academic deterrence theory is all we needmaybe we can rest easily with the oxymoronic notion that lack of defensesconstitutes defense dominance Such faith smacks of hubris The strength ofnuclear deterrence theory came from its parsimony That in turn reected thetheoryrsquos blindness to big problems revealed by a focus on different variablesBruce Blair John Steinbruner Desmond Ball and Paul Bracken showed howfundamental assumptions about stability in mainstream deterrence theorywere thrown into serious doubt if analytical focus was shifted from the vul-nerability of forces to the vulnerability of command systems23 These concep-tual revisionists took a fundamental assumption of standard deterrence theory(the physical capacity to effect retaliation if the victimrsquos forces survived attack)subjected it to empirical analysis and showed it to be shaky

Finally the combination of political and military criteria exemplied indenitions of the nuclear offense-defense balance raises questions about con-sistency in operationalization of the concept Van Evera writes ldquoI use lsquooffense-defense balancersquo to denote the relative ease of aggression and defense againstaggressionrdquo (p 118 n 2) ldquoAggressionrdquo is a value-laden term implying unpro-voked attack which does not cover all cases of rst-strike or invasion Whichones it does cover depends on evaluation of motives and attributions ofillegitimacy to particular attacks Unless Van Evera really meant to say justldquoinitial uses of forcerdquo rather than aggression this denition of the offense-defense balance mixes normative political and empirical military criteria inambiguous ways Was an offense-defense balance that enabled a defensivelymotivated Israel to save itself from destruction in 1967 by initiating a devas-tating conventional offensive one that was offense dominant or defense domi-nant All in all it is better for analytical accuracy clarity and simplicity toreturn to strictly military operational criteria for categorizing the offense-defense balance and to keep separate the purposes to which the operationaladvantages may be put or the decisions they may prompt

23 Bruce G Blair Strategic Command and Control (Washington DC Brookings 1985) Blair TheLogic of Accidental Nuclear War (Washington DC Brookings 1993) John D Steinbruner ldquoNationalSecurity and the Concept of Strategic Stabilityrdquo Journal of Conict Resolution Vol 22 No 3(September 1978) pp 411ndash428 Desmond Ball Can Nuclear War Be Controlled Adelphi Paper 169(London International Institute for Strategic Studies 1981) and Paul Bracken The Command andControl of Nuclear Forces (New Haven Conn Yale University Press 1983)


how much does world war I proveThe utility of offense-defense theory for stabilizing the Cold War was under-written by another wave of analysis of the dynamics of the 1914 crisis Haddecisionmakers in European capitals then understood that the military equa-tion of the time favored the defense rather than assuming the reverse peacemight have endured The diplomatic events of 1914 and the military events of1914ndash17 inspired theoretical excursions that were applied more generally toconventional military strategy From extensive reading and thinking Van Everahas concluded that ldquoperceived offense dominance is pervasive and it plays amajor role in causing most wars Knowing exactly which wars would still haveerupted in its absence requires a close analysis of each casemdashimpossible hereBut the evidence does indicate that it had a vast rolerdquo (p 185 emphasis added)

It is frustrating to hear the author say that he cannot show the evidence butthen hear him report what the evidence shows Apart from one case the bookpresents scant evidence of instances in which policymakers decided to initiatewar because of perceived offense dominancemdashthat is where it is clear that theperception was a principal cause rather than one among a dozen on a laundrylist of reasons The one exception is World War I which gets a whole chapterof its own as well as space in other chapters Although he notes ve otheranalysts of World War I who doubt that a ldquocult of the offensiverdquo caused it VanEvera reports that the ldquotestrdquo case of World War I conrms twenty-four oftwenty-seven predictions from ODT and concludes ldquoWe have a verdict waris markedly more likely when conquest is easy or is believed easyrdquo (p 233see also p 238 n 180)

Eyebrows might be raised when we see the case that inspired much of recentODT being presented as the test of the theory especially when two pages laterVan Evera mentions offhandedly that some other cases show offense domi-nance as conducive to peace Similarly the book has a short section onldquoQualications When Offensive Doctrines and Capabilities Cause Peacerdquo(pp 152ndash160) but gives little indication of how heavily these qualicationsshould weigh Implicitly they are minor exceptions that do not vitiate ODTrsquospresumption in favor of defense World War II however is a rather bigexception As Van Evera recognizes ldquoThe Western allies of the 1930s neededoffensive capabilities and might have deterred the war had they adopted moreoffensive strategiesrdquo belief in defense dominance contributed to war (pp 154235) Why should we be sure that we have more to learn from 1914 than from1939


Fine Grain or Gross A Bloated Concept of Offense-Defense Balance

The most powerful concepts are ones that can be dened simply and opera-tionalized clearly Most attempts to promulgate simple and powerful conceptsget battered if not discredited by the complexity of real life This did nothappen with nuclear deterrence theory because reality never provided any testcases (nuclear wars in which the conditions of deterrence failure and the resultsof combat engagements could be studied) Because of the volume of evidenceon conventional warfare in contrast the simplest versions of ODT ranaground Recent debate highlights the trade-off between accuracy and elegancein the theory

conceptual enlargementRening conventional offense-defense theory proved to be trickier than thethought experiments and exchange calculations of nuclear strategy where onlya few scenarios were at issue and weapons could be used operationally onlyfor attack or defense rather than for both Critics raised questions about howfar the simplicity of nuclear terms of reference could be transferred to conven-tional war where attack and defense are advantages that shift in the course ofa war the same weapons can be used for both purposes multiple weaponscan be reorganized and combined with doctrines in different integrated sys-tems with different implications the tactical offensive and strategic defensive(or the reverse combination) can be combined in campaigns or even tacticaloffensive and defensive actions can reinforce each other within the sameengagement For example in May 1940 the Germans expanded bridgeheadsby using ldquointerior lines and the ability to concentrate against one face of theirsalients while standing on the defensive on the other twordquo24 in 1973 theEgyptians used air defense missiles offensively to screen the advance of ground

24 Archer Jones The Art of War in the Western World (Urbana University of Illinois Press 1987)p 531 ODT promoters also sometimes cite the authority of Clausewitz when he proclaims thatldquothe defensive form of warfare is intrinsically stronger than the offensiverdquo (emphasis in original) butneglect to mention that Clausewitz says this to endorse tactical exploitation of defense in theservice of a strategic offensive Given that defense has a negative object it ldquomust be abandonedas soon as we are strong enough to pursue a positive object When one has used defensivemeasures successfully a more favorable balance of strength is usually created thus the naturalcourse of war is to begin defensively and end by attacking It would therefore contradict the very ideaof war to regard defense as its nal purposerdquo (emphasis added) Clausewitz On War p 358 see alsopp 370 600


forces and the Israelis used attack aircraft and tanks defensively to hold theirline and so forth

ODT supporters have generally responded by introducing qualications thatreduce the emphasis on particular military technological determinants andtake the theory further from parsimony Some buttress the theory by addingand subtracting components to cope with anomalies and making the theoryrsquosclaims more limited In an otherwise impressive article Charles Glaser andChaim Kaufmann complicate the specications of the theory so much that areader would need to keep a cue card just to be able to think about how toapply it They also introduce simplifying assumptions that ease theorizationbut do not comport with important aspects of the empirical record of modernconventional war and that essentially eliminate strategy from the strategicequation They insist that ldquothe offense-defense balance should be assessedassuming optimalitymdashthat is countries choose the best possible strategies andforce posturesrdquo and they eliminate differences in ldquomilitary skillrdquo from consid-eration25 By this restricted denition the offense-defense balance must havefavored France over Germany in 1940 Egypt over Israel in 1967 Argentinaover England in 1982 and Iraq over the United States in 1991 The authorsappropriately keep their claims restrained saying that ODT ldquodoes not claimthat the offense-defense balance is in general a more important determinant ofmilitary capabilities than is power or skillrdquo26

Van Evera reacts to criticsrsquo challenges about how much ODT can explain ina different way and in the bargain sidesteps debates about what elementsshould be included in the concept Instead of rening the denition of theoffense-defense balance he broadens it27 instead of resolving debates aboutwhat should be included he includes almost everything that has ever beensuggested and instead of modifying claims he expands them Although hecriticizes other realists for misdirected attention to ldquogrossrdquo power rather thanthe ldquone-grainedrdquo structure of power he crams so many things into thedenition of the offense-defense balance that it becomes a gross megavariableHis offense-defense balance includes ldquomilitary technology and doctrine geog-raphy national social structure and diplomatic arrangements (specically

25 Charles L Glaser and Chaim Kaufmann ldquoWhat Is the Offense-Defense Balance and Can WeMeasure Itrdquo International Security Vol 22 No 4 (Spring 1998) pp 46 48 5526 Ibid pp 48ndash4927 This point leaps out It is also made by all three critics responding independently to a previewof Van Everarsquos book See Davis Finel and Goddard ldquoCorrespondence Taking Offense at Offense-Defense Theoryrdquo pp 180 187 190ndash191


defensive alliances and balancing behavior by offshore powers)rdquo as well asmanpower policy (the leveacutee en masse) strategic choices in campaigns (themaldeployment of UN forces in the advance to the Yalu) popularity of re-gimes and the breakdown of collective security organizations (the League ofNations) (pp 122 160ndash166) Phew If we could get him to throw in populationand nancial resources we could almost call the offense-defense balance ldquorela-tive powerrdquo and be back to basic realism Van Evera does not eliminate thedistinction between offense-defense balance and power but he blurs it enoughto make the difference in their implications unclear28

Van Evera would have none of this because he sees his approach as a basicvariant of structural realism differentiated from othersrsquo focus on polarity Hedistinguishes four schools (p 10) Type I classical realism Type II structuralrealism which focuses on polarity his own preferred Type III (ldquone-grainedstructural realismrdquo) and Type IV (ldquomisperceptive ne-grained structural real-ismrdquo) He does not believe that herding a horde of variables into the offense-defense balance washes out the ne grain Indeed he insists that his expansivedenition is very parsimonious because it ldquoachieves simplicity binding to-gether a number of war causes into a single rubric Many causes are reducedto one cause with many effectsrdquo (p 190) In a response to critics he argues ldquoAtheory is not shown to lack parsimony simply by demonstrating that itsconcepts include a diverse range of lesser-included concepts All conceptsare aggregations of lesser conceptsrdquo29 But why is it more useful to consider allthe variables lumped in the denition above as ldquolesser-includedrdquo elements ofan offense-defense balance than it is to consider a more coherently speciedbalance as a lesser-included concept of ldquopowerrdquo It might be if we had aformula that converts all the ldquone-grainedrdquo elements of Van Everarsquos balanceto a metric that is more illuminating than relative power as a whole But thereis no indication of how all these elements in the concept should be aggregatedand thus no clear way to measure the offense-defense balance Van Evera isreduced to relying on ldquoauthorrsquos estimatesrdquo when he comes to categorizingwhether the balance in actual periods of history favored offense or defense

28 There is also a problem of virtual circularity in the theory A big advantage for defenders isthat other status quo powers often align with them Keeping alignment out of the equation makesthe offense-defense balance a much weaker predictor Putting it in the equation however makesstrategic behavior both a cause and an outcome See Davis ibid p 180 and Goddard ibid p 191On whether he uses the behavior of states to explain the behavior of states Van Evera says ldquoIplead guilty I donrsquot see a problemrdquo Ibid p 19929 Ibid p 196


How then can other analysts apply this concept Without a clear formula it ishard to see how far we are left from Blaineyrsquos conclusion that war hinges ondisagreements (miscalculations) about relative power30

distinguishing the offense-defense balance from the balance of powerThere is a difference between the offense-defense balance and power as awhole that is clearer in more limited formulations of ODT Blainey views parityof power as destabilizing because he does not see decisionmakers differentiat-ing the convertibility of resource inputs for combat outputsmdashthat is he doesnot consider that governments might take fewer risks if they understood thatequal resources would produce unequal effectiveness in attack and defenseRough parity lets governments miscalculate easily in Blaineyrsquos world becauseeither side can convince itself that a modest edge in the balance of power givesit an opening to gain victory Power parity promotes peace in a recognizeddefense-dominant world in contrast because both sides know they wouldneed a much bigger edge than they have in aggregate resources to overcomethe other rsquos defense

Van Evera preserves the crucial criterion that the balance should be somesort of pound-for-pound ratio rather than a measure of aggregate probabilityof combat success when he formulates the concept as ldquothe probability that adetermined aggressor could conquer and subjugate a target state with compa-rable resourcesrdquo (p 118 n 2 emphasis added) But in the next breath he subvertsit by admitting alliance shifts to the calculus because realignment can overturnthe balance of resourcesmdashindeed that is its purpose This is why the focus onldquogrossrdquo power (polarity) of the Type II structural realists from whom Van Everawants to differentiate himself is likely to remain most important If not oneshould expect that Van Everarsquos Type III focus on the ne grain should yielddifferent or more useful explanations and predictions about who would winwars But does it Would his framework have told statesmen more than ldquogrossrdquopower analysis about what they would have wanted to know before going towar in the Persian Gulf or Kosovo Perhaps but it is hard to know without aformula that converts the offense-defense megabalance into a sort of ldquounit-pricerdquo measure of relative combat efciency

30 One problem is uncertainty about what aggregate ldquopowerrdquo means or whether there is such athing Realists tend to see power as a conglomeration of material capabilities mainly economicand military that is fungible Others who do not see power as fungible insist that one must askpower to do whatmdashwhich is what an offense-defense balance is about See David A BaldwinldquoPower Analysis and World Politics New Trends vs Old Tendenciesrdquo World Politics Vol 31 No2 (January 1979) pp 161ndash194


One reason Van Evera focuses on the ne-grained structure of power is thathe considers it ldquomore malleable than the gross structure hence hypotheses thatpoint to the ne-grained structure yield more policy prescriptionsrdquo (p 8)Others such as Quester also see the benet of focusing on the offense-defensebalance as its comparative malleability (although he keeps the concept moremanageable by restricting it to the level of military capabilities and tech-niques)31 It is indeed sensible to concentrate on causes that can be manipu-lated and a central mission of strategic studies is to direct attention to thecausal signicance of military force structures and strategies in internationalrelations There are two problems however in touting the malleability of thebalance

First if we keep the concept more coherent than Van Evera does and wefocus on the military elements of the offense-defense balance we can see thatthey are more manipulable than some factors such as geography but notnecessarily more manipulable than others such as alliances or diplomaticstrategies For example what was crucial in setting off World War II in Europewas not any notion of a military offense-defense balance or changes that theGermans could make in it Rather it was the stunning German-Soviet nonag-gression pact which solved Hitlerrsquos problem of a potential war with greatpowers on two fronts ldquoFine-grainedrdquo aspects of power (such as innovationsin doctrine for armored warfare) may explain why the Germans did so wellin early phases of the war but the gross distribution of power (poles and theiralignment) tells us more about how the war got started32

Second if we accept Van Everarsquos sweepingly inclusive denition (whichsimply incorporates the Hitler-Stalin pact in the offense-defense balance) itbecomes more important to remember that manipulability works two waysVan Evera is interested in presenting a menu of variables that can be manipu-lated to reinforce peace but there is nothing that ordains that status quopowers will shape the balance to their purposes more successfully than revi-sionist powers will shape it to theirs Manipulability is necessarily conduciveto peace only if we assume that the problem lies entirely in the securitydilemma that countries in conict will welcome a structured stabilization ofcapabilities that appears to make conquest hard for either of them

Packing many elements into the offense-defense balance concept does notalter the fact that the elements are very different things Whether ldquolesser-

31 Quester Offense and Defense in the International System p 1032 See Randall L Schweller Deadly Imbalances Tripolarity and Hitlerrsquos Strategy of World Conquest(New York Columbia University Press 1998) chaps 4ndash5


includedrdquo or not this variety means that there are numerous points of entryfor a revisionist strategist to try to overturn a benign balance This does notmean that it can always be done In some conventional military equationsstructural constraints may prove irremediably favorable to the defense But themore complicated the equation the more variables and the more potentialcombinations a revisionist strategist has to work with Organizations anddoctrines may be developed to change the effects of combined arms operationsand force structures may be adapted to apply existing technologies in novelcombinations or innovation and research may be channeled to develop tech-nologies to counter tactical defensive advantages or strategic plans may befound to capitalize on tactical advantages of defense for strategically offensivepurposes or allies may be acquired or detached from adversaries and so forthIndeed this is what strategy is supposed to be all about Jonathan Shimshoniillustrates how successful militaries do this how they strive to make balancesdynamic and evolving rather than static and conning to give them an edge33

An assumption that countries will seek stability if only its foundations can bemade clear to them is an assumption about motives not capabilities

Beyond Power

Over time observers have cited many motives for war God gold and gloryor fear honor and interest34 Like most realists Van Evera has little to say aboutany of these but fear and interest35 Indeed because his premise is that mostwars result from security seeking he focuses primarily on fear and how torelieve it by adjusting the structure of power and suppressing misperceptionsof it Van Evera shares with Blainey a lack of interest in the political substanceof issues at stake in conicts among statesmdashterritorial claims religious orideological disputes economic interests and so forth To Blainey under-standing aims and motives may provide a theory of rivalry but not of war36

If rivalry is a prerequisite for war however reducing causes of rivalry will tendto reduce the probability of war

33 rdquoA military entrepreneur must constantly ask two questions (1) what would war look likeif fought today and (2) how can I lsquoengineer rsquo the next war away from (1) so as to maximize myrelative advantages rdquo Jonathan Shimshoni ldquoTechnology Military Advantage and World War IA Case for Military Entrepreneurshiprdquo International Security Vol 15 No 3 (Winter 199091) p 19934 The latter three are emphasized by Thucydides See Donald Kagan On the Origins of War andthe Preservation of Peace (New York Anchor Books 1995) pp 7ndash8 and passim35 For systematic attempts to integrate the issues stakes or political motives of conict in theoriesof war see John A Vasquez The War Puzzle (New York Cambridge University Press 1993) andKalevi J Holsti Peace and War (New York Cambridge University Press 1991) chap 1036 Blainey The Causes of War pp 149ndash150


political stakes and objectivesTo Van Evera the beauty of a benign offense-defense balance is that it canprevent war irrespective of motives because states can see that attack cannotsucceed37 The common sense that appeals to Blainey is also implicit here if aprospective war seems unwinnable no one will start it One reason to payattention to political incentives as well as to military constraints is that thisstarkly utilitarian common sense does not always govern decisions on war andpeace For honor or other reasons that seem quaint to economists countriesmay ght in the face of probable or even certain defeat38

There is also reason to consider motives within the utilitarian calculusSometimes the problem is as Blainey argued that contenders of roughly equalpower do not see that they will not be able to use force effectively especiallyif they believe they can exploit novel stratagems A defense-dominant balancemight solve this problem but it would have to be lopsided Van Everarsquosconcept of the balance includes so much that it would be hard to ndmany situations in which some elements of the conglomeration did not seemto cancel out other defense-dominant ones leaving an opening for success-ful attackmdashfor example where geography or social structure or alliancechoices conducive to offensive action overrode military technology conduciveto defense

Where the answer to which side would prevail in combat is a matter ofprobability rather than certainty risk propensity and willingness to sacriceenter the equation These can depend in turn on the value of the stakes ldquoAplace in the sunrdquo may strike Van Evera a humane and modern man as nodecent reason to launch a war that courts catastrophe The value of that aimwas far different to the Kaiserrsquos Germany however when it was coming lateto the imperialist scramble at a time when the world was made of empires Adifference in the relative value of the stakes to two contestants may also affecttheir relative will to sacrice the amount of their power that they commit tothe contest and their chances of winning North Vietnam had far less powerthan the United States but the United States had far less stake in the outcomeof the war so committed a much smaller fraction of its power Van Everaexamines the costs of war intently but neglects the benets which requireexamination of stakes and motives

37 Van Evera ldquoCult of the Offensiverdquo p 10538 Robert Jervis ldquoWar and Misperceptionrdquo in Robert I Rotberg and Theodore K Rabb eds TheOrigin and Prevention of Major Wars (New York Cambridge University Press 1989) pp 103ndash104


Realists usually ignore the stakes in dispute for theoretical purposes becausethey assume that conict is more or less a constant Something always comesup to put countries in conict but many conicts persist without war Thequestion is what pushes conict over the brink to combat For a highly generaltheory of average outcomes over time in international politicsmdashwhat interestsKenneth Waltzmdashthis lack of attention to the substantive issues of disputesmakes sense For any particular case however the fact remains that there aretwo necessary conditions for war a distribution of power that makes warappear an effective option and a conict of interest The rst without thesecond will not produce war The United States has more than enough powerto take Canada but lacks sufcient political incentive to do so

War between countries in conict can be avoided through either of twoways constraining their power or resolving their conict ODT sees morepromise in the rst because it seems more manipulable But is this necessarilyso If the offense-defense balance includes geography social structure andother such weighty things as in Van Everarsquos formulation why is it moremanipulable than attitudes toward stakes Washington and Moscow dueledinconclusively for two decades over nuclear arms control and the Mutual andBalanced Force Reduction talks but the Cold War ended more because of apolitical decision to resolve the conict than because of any beliefs about anoffense-defense balance between East and West Van Evera argues that MikhailGorbachev was willing to make concessions because he adopted ODT (p 119)If we look at all that happened at the end of the 1980s however it is hard tobelieve that whatever thought Gorbachev gave to ODT was more than atactical rationalization for much grander political decisions Gorbachev did notbargain with the West to negotiate arms reductions or diplomatic deals Hesimply gave up the global ideological contest that was the subject of the ColdWar Realist ODT was less evident in his thinking than liberal institutionalistnostrums about a ldquocommon homerdquo The Soviets did not exchange concessionswith the West They made unilateral concessions which the West pocketedThe Cold War ended not because the Russians bought some academic ideaabout how much easier military defense of their interests could be but becausethey gave away those interests surrendered their empire and left the Westwith nothing to ght about

subconventional warMore attention to political stakes would also help if we want a theory aboutthe causes of war in general rather than the subset of wars that interests VanEvera and most international relations theorists big world-shaking interna-


tional wars among great powers Those sorts of wars are most important andthey could occur again They have not been representative of the phenomenonof war since 1945 however and with at least a temporary respite from seriousgreat power conict after the Cold War hot wars may be of other types forsome time to come

Most recent wars have been intranational internal civil wars Most interna-tional wars are about which countries control which territories most intrana-tional wars are about which group gets to constitute the state Van Everarsquosresearch surveyed thirty wars 90 percent of which were international and 10percent intranational (p 13 n 24) Since 1945 the proportions of war that haveactually occurred are closer to the reverse In civil wars the relative salience ofconcerns about political values as opposed to material power is usuallygreater than in international wars

Some wars could count in both categories because they involve interventionsby outside countries in civil wars and some categorizations could be disputed(eg the Korean War could be considered a civil war between northern andsouthern Koreans who considered their country to be temporarily divided oran international war between two Korean statesmdashas American white south-erners regarded what northerners called the Civil War) In terms of militarycharacter some civil wars are similar to international ones if the contendinggroups are geographically separated and they marshal conventional armiesMany contemporary internal wars differ signicantly however from the con-ventional terms of reference by which recent ODT has developed Whencontending groups are divided by ideology religion or ethnicity they are oftenintermingled and combat is not carried out on one front with conventionallines of operation Rather the front is everywhere as groups compete todominate villages provinces and loyalties throughout a country

Irregular or guerrilla warfare does not follow the operational logic of ODTgrounded in modern conventional war and high force-to-space ratios that arepossible on a linear front Irregular war is more similar to that of ancient andmedieval times where low force-to-space ratios made raids the usual form ofcombat The rule of thumb purveyed by counterinsurgency theorists duringthe Cold War was that a government defending against guerrilla attacksneeded a ten-to-one superiority in forces This is thirty times the legendaryone-to-three ratio in the rule of thumb for the tipping point between attackand defense in conventional warfare When guerrillas can be engaged in aconventional battle where they ght on the defense they can be annihilatedInsurgents who hide among the civilian population in contrast can concen-trate at will to raid specic objectives destroy pockets of defending forces and


retreat to avoid battle with the governmentrsquos main forces The governmentmust disperse its forces to defend everywhere The ability of the insurgents tohide in the population in turn is determined by political and social factors

Van Evera does not agree because he thinks about international war Inanswer to a critic he writes ldquoModern guerrilla war has defended many coun-tries and conquered none It is a fundamentally defensive form of warfarerdquo39 Butthe communist guerrillas in Greece Malaya and South Vietnam or the anti-communist mujaheddin in Afghanistan and Contras in Nicaragua were attack-ing native governments Foreign intervention on behalf of such governmentsinternationalized the wars precisely because of the guerrillasrsquo offensive success

A nal reason to look beyond power and especially beyond the offense-de-fense balance is the conservative presumption in Van Everarsquos theory The bookimplies that war is justied only to defend what one has not to conquer whatothers hold If we were to leave open the normative question of whetherconquest or coercion are ever desirable we would leave open the question ofwhether it is desirable to institutionalize defense dominance In principlehowever there is no automatic reason why it should be any more legitimateto use force on behalf of what is than on behalf of what should be Once whatshould be becomes admissible as a warrant for force there can be no consistentpreference for defense over attack because there is no automatic congruencebetween what should be and the status quo Revolution may be justied in anunjust society and international revisionism may be justied in an unjustinternational order If the norm is to be that a just cause can warrant resort toforce whether peace should be preferred to war is an issue not a premise Onthis matter Blaineyrsquos analysis is more detached and value free because heregards peace as no more natural than war40

Perhaps it sounds perverse or insidious to raise this issue But is it anyaccident that American or other Western analysts favor peace when the UnitedStates and other Western great powers are status quo states satised with theinternational pecking order as it is a bunch of prosperous righteous countriescalling the tune for others When an unjust status quo catches the eye ofWestern leaders however they do not hesitate to see the virtue in attacking itWitness the assault launched against Serbia to wrest an oppressed Kosovo fromits control Bill Clinton and Tony Blair were certainly glad to enjoy offense

39 Van Evera ldquoCorrespondence Taking Offense at Offense-Defense Theoryrdquo p 195 (emphasisadded)40 Blainey The Causes of War chaps 1 16


dominance as NATO air forces could range across Serbiarsquos skies unchallengedmethodically destroying the country with impunity

War Will Usually Find a Way

Offense-defense theory is in many respects quite hardheaded and realisticabout what military power can accomplish but some aspects of it are non-Clausewitzian ODT is most concerned with stability how to prevent war bymaking it ineffective as an instrument of policy Clausewitz thinks in terms ofstrategy how to make war serve policy Most devotees of the theory would liketo paralyze strategic options for using conventional forces as thoroughly asMAD seems to paralyze nuclear strategy The ideal offense-defense balance isimplicitly some benign form of military stasis Clausewitz in contrast payslittle attention to the causes of war He focuses on its nature and purpose Thestrategic orientation he represents is dynamic how to scan the environmentand history to nd a way around obstacles that protect the adversary how tond a way to make war work for onersquos purposes

Some carefully exclude strategy from ODT in contrast to Van Everarsquos exces-sive inclusiveness this narrows the concept too much Shimshonirsquos critiquethat strategic entrepreneurship can overcome a constraining offense-defensebalance is dismissed by Glaser and Kaufmann as awed because it ldquorelaxesthe assumption of optimality All of his examples hinge on states havingsignicant advantages in military skill over their opponentsrdquo41 This protectstheir conceptualization of an offense-defense balance but does not give anyreason to look to such a balance as a reliable bar to war As Levy noted theeighteenth century has generally been regarded as defense dominant but thecampaigns of Frederick the Great were an exception Analysts do not call theoffense-defense balance offense dominant in this period because Frederickrsquossuccess came from tactical and strategic innovation rather than technologyldquoStill it cannot be denied that Frederick demonstrated what was possible giventhe technology of the timerdquo42

Strategic schemes help war nd a way ODT looks for means to bar the wayto neutralize strategy or freeze it This proves harder in conventional warfarethan in the nuclear deterrence theory in which contemporary ODT germinated

41 Glaser and Kaufmann ldquoWhat Is the Offense-Defense Balancerdquo p 55 n 4042 Levy ldquoThe OffensiveDefensive Balancerdquo pp 231ndash232 See also Finel and Goddard ldquoCorre-spondence Taking Offense at Offense-Defense Theoryrdquo pp 183ndash184 193


When Waltz quoted the line from Brecht that appears at the beginning of thisessay he followed up with the point that ldquofor half a century nuclear war hasnot found a wayrdquo43 This led Waltz John Mearsheimer and a few others toview the spread of nuclear weapons as a welcome force for peace Whethervalid or not this idea appeals only to a handful of hyper-realists and has nofuture as a policy prescription44

If ODT is to help bar the way to war it has to do so at the conventional levelof combat In theory this might be accomplished by nding defensive tech-nologies as enduringly superior as offensive systems have been in nuclearforces or by freezing defense-dominant congurations in arms control dealsSo far however no formulation of defense dominance for conventional forceshas proved immune to strategy The stalemate of 1914ndash17 came closest to suchrobust defense dominance but it too was eventually overcome by innovationsGerman inltration tactics on the one hand and British deployment of tankson the other made attack more effective and unfroze the lines If strategic effortcan nd ways to overcome an inhibiting offense-defense balance and if politi-cal stakes and motives are strong enough to provide an impetus war will stillnd a way

That is not to say that victory will necessarily nd its way There is no reasonto expect that miscalculation or plain foolishness will become less commonNor do the various problems discussed in these pages mean that the idea ofan offense-defense balance is wrong or useless The question is still openhowever as to how best to conceptualize the balance and how much to claimfor it If it is to be analytically useful the conceptrsquos scope should be kept limitedto the dimension of military operations relative combat effectiveness on apound-for-pound basis when attackers crash into defenders in a tactical en-gagement or strategic campaign There is no consensus yet on a precise deni-tion in these termsmdashany one yet offered seems to run up against some questionor scenario that makes it problematicmdashbut we could do worse than to workfor particular purposes with some of those that have emerged from methodo-logical wrestling For example Levyrsquos ldquothe offensivedefensive balance isinversely proportional to the ratio of troops needed by an attacker to overcomean enemy defending xed positionsrdquo or Lynn-Jonesrsquos ldquothe amount of re-sources that a state must invest in offense to offset an adversaryrsquos investment

43 Waltz ldquoWaltz Responds to Saganrdquo in Scott D Sagan and Kenneth N Waltz The Spread ofNuclear Weapons A Debate (New York WW Norton 1995) p 93 (emphasis in original)44 As James Kurth says ldquoThere probably has not been a single foreign policy professional in theUS government who has found this notion to be helpfulrdquo Kurth ldquoInside the Cave The Banalityof IR Studiesrdquo National Interest No 53 (Fall 1998) p 33


in defense the offensedefense investment ratio required for the offensivestate to achieve victoryrdquo or Glaser and Kaufmannrsquos ldquothe ratio of the cost ofthe forces that the attacker requires to take territory to the cost of the de-fender rsquos forcesrdquo45

It will not do as Van Evera has done to swallow up contradictions andconfusions brought out in recent debates by rolling all manner of things thatmake conquest easy into the offense-defense balance Including factors thataffect the total resources available to contestants makes the balance almostcongruent with a comprehensive net assessment of which side would win awar and too close to just another term for how to parse ldquopowerrdquo To makegood use of the concept of the offense-defense balance it is necessary to returnto more limited and coherent formulations of it The last word on this subjectremains to be written

Van Evera makes an important point that is buried in a footnote early in hisbook ldquoMost of the important contributing causes of war are neither necessarynor sufcientrdquo (p 41 n 19) The works of Kenneth Waltz concentrate on oneof the main necessary or permissive causes of war international anarchyBlainey concentrates on another disagreements about relative power Butnecessary or permissive causes are not enough to understand specic wars asany statesman will insist that a theory aspiring to prescriptive utility must doFor that looking back toward classical realism (eg Carrrsquos emphasis on theinteractions among grievance satisfaction and power) helps more

More interest in efcient or sufcient causes will lead us to look more atmotives and stakes than Waltz Blainey or Van Evera has yet done The subjectsthat Van Evera announces he will deal with in a planned second volume getcloser to the efcient causes He plans to explore four explanations that (1)militaries cause war ldquoas an unintended side-effect of their efforts to protecttheir organizational welfarerdquo46 (2) states ldquoinfuse themselves with chauvinistmythsrdquo and ldquounderestimate their own role in provoking othersrsquo hostilityrdquo (3)

45 Levy ldquoOffensiveDefensive Balancerdquo p 234 Lynn-Jones ldquoOffense-Defense Theory and ItsCriticsrdquo p 665 and Glaser and Kaufmann ldquoWhat Is the Offense-Defense Balancerdquo p 4646 It is interesting however that in the vast region where militaries were most dominant inpoliticsmdashSouth America in much of the twentieth centurymdashthere was almost no interstate warFelix Martiacuten shows that neither main paradigm of international relations theorymdashrealism orliberalismmdashexplains this a much longer peace than the vaunted Long Peace in Europe during theCold War (Causes of war identied by realism existed in South America but war did not occurand causes of peace identied by liberalism did not exist yet peace endured) If anythingorganizational interests and transnational professional identicationmdashin effect an epistemic com-munity of military ofcers in the regionmdashdiscouraged a resort to force across borders FelixMartiacuten-Gonzalez ldquoThe Longer Peace in South America 1935ndash1995rdquo PhD dissertation ColumbiaUniversity 1997


states ldquooften misperceive because they lack strong evaluative institutionsrdquo and(4) states avoid dening their national strategies clearly so ldquoofcial thinkingdeterioratesrdquo states are less able to assess each othersrsquo intentions and theybecome ldquoblind to the otherrsquos concernsrdquo (pp 256ndash258)

Whether we are concerned with necessary permissive efcient or sufcientcauses there is not likely to be a breakthrough on the rst-order issues ofdebate about what causes war The debate has cycled back and forth forcenturies in many variations on a few themes in the realist and liberal tradi-tions of international relations theory This elemental debate is no more likelyto be resolved in the next century than in the last one if only because eachtradition has a purchase on different facets of the problem and the glass thatis half full for one will always be half empty for the other

At less cosmic levels of debate where most research takes place progress ismore plausible even if it is likely to be more halting and inconclusive thanardent social scientists expect There will be room for revisionist research andcreative revisitations of old issues as long as historians keep political scientistshonest by disciplining their oversimplications about what past cases proveThere will be plenty of work to do for a long time on the question that isamong the handful of fundamental issues in international relations Whatcauses wars Van Everarsquos book does some of that work imperfectly like allgood books but usefully It need not make good its claim to a master theoryto merit a place at the table


There is one particular book to which Van Everarsquos especially begs compari-son Apparently unconcerned about having order clerks in college bookstoresconfuse his own book with a contemporary classic he chose the same title ashistorian Geoffrey Blainey In this generation Blaineyrsquos is the book most similarin scope although different in approach and style and it is one that willultimately last at least as well Blainey examined and debunked more than adozen popular notions about why wars happen and by process of eliminationsettled on one main conclusion ldquoWars usually begin when two nations dis-agree on their relative strengthrdquo At least one of the nations in conict mustmiscalculate who would succeed in a test of arms or the weaker would yieldwithout a ght Thus a clear pecking order in international relations may notnecessarily produce justice but it promotes peace A roughly even balance ofpower in contrast makes miscalculation easier The key to peace is clarityabout the distribution of power2 Blainey arrived at these spare and powerfulconclusions inductively and his book is a tapestry of unconventional questionsanalytical excursions and examples and ironic observations that social scien-tists would consider literate and lively but unsystematic

Van Evera travels a different route one that proceeds more methodically anddeductively and aims to honor the canons of social science He arrives at aplace apparently different from Blaineyrsquos but actually similar Van Evera ar-gues for ve hypotheses but concentrates on one ldquoWar is more likely whenconquest is easyrdquo3 This seems to contradict Blaineyrsquos view that imbalancefavors peace because weakness encourages compliance with the strongerpower In the course of the book however Van Evera makes clear his convic-tion that the main problem is quite different from the one stated in the initialhypothesis it is not that conquest actually is easy but that most often it is notyet is mistakenly perceived to be easy This amendment is quite compatible withBlaineyrsquos bottom line

Van Evera also has an ambitious aim Where Blainey remains strictly empiri-cal Van Evera seeks to provide prescriptive analysis that can show policymak-ers how to manipulate causes of war and deploy countermeasures againstthem As a good realist he does not claim that he can make war obsolete buthe does imply that if statesmen understood the analysis in the book and actedon it searches for security could be relaxed and more wars could be foregone

2 Geoffrey Blainey The Causes of War 3d ed (New York Free Press 1988) p 293 see alsopp 109ndash1143 Stephen Van Evera Causes of War Power and the Roots of Conict (Ithaca NY Cornell UniversityPress 1999) p 4 Subsequent references to this work appear parenthetically in the text



























































































































Beyond Power








































































































































































Beyond Power















































Richard K. Betts A Review Essay - Stanford Universityweb.stanford.edu/class/polisci211z/2.1/Betts IS...

Documents

Transcript of Richard K. Betts A Review Essay - Stanford Universityweb.stanford.edu/class/polisci211z/2.1/Betts IS...