Postwar American security policy was built on a foundation of deterrence. In the early Cold War period, American leaders relied on nuclear deterrence to discourage Soviet or Chinese attacks against American allies in Western Europe and the Far East. When these countries developed the means to launch intercontinental nuclear attacks of their own, the USA counted on deterrence to prevent an attack against itself. Over the years, successive American administrations have also attempted to use deterrence to moderate the policies of Third-World states with which the USA or its allies have come into conflict. Partisans of deterrence assert that it has kept the peace between the superpowers and has been useful in managing lesser conflicts. This chapter disputes both claims. 1

When discussing deterrence, it is important to distinguish between the theory of deterrence and the strategy of deterrence. The former pertains to the logical postulates of deterrence and the assumptions on which they are based. Put succinctly, deterrence is an attempt to influence another actor’s assessment of its interests. It seeks to prevent an undesired behavior by convincing the party who may be contemplating it that the cost will exceed any possible gain. Deterrence presupposes that decisions are made in response to some kind of rational cost-benefit calculus, that this calculus can be successfully manipulated from the outside, and that the best way to do this is to increase the cost side of the ledger. Different scholars have developed their own variants of deterrence theory . All of them, however, are based on these assumptions.

Deterrence strategy is concerned with applying the theory of deterrence to real world conflicts. It has given rise to its own body of theory about how this is best accomplished. The first wave of this theory, almost entirely deductive in nature, was developed in the 1950s and 1960s by scholars such as Bernard Brodie (1959), William Kaufmann (1954), and Thomas Schelling (1966) . Most of these works stressed the importance of imparting credibility to commitments and explored various mechanisms leaders could exploit toward this end. The literature of this period is often referred to as classical deterrence theory (Jervis 1979).

Classical deterrence spawned a number of critiques. For our purposes, the most interesting were those that attempted to evaluate deterrence strategy in light of empirical evidence from historical cases. The works of Milburn (1959), Russett (1967), Snyder and Diesing (1977), and George and Smoke (1974) are representative of this wave of theorizing. These scholars sought to refine the strategy of deterrence in order to make it more useful to statesmen. Milburn, Russett, and George and Smoke argued that deterrence might be made more efficacious if threats of punishment were accompanied by promises of reward for acceptable behavior. George and Smoke and Snyder and Diesing sought to divorce deterrence from its Cold War context and root it in a less politically specific theory of initiation.

Empirical analyses of deterrence had implications for the postulates of deterrence theory . On the basis of their case studies, George and Smoke (1974) argued for a broader formulation of rational choice. They hoped that this would enable the theory to incorporate domestic political concerns and other factors affecting foreign policy behavior that deterrence theory had not previously taken into account.

This essay incorporates and expands upon elements of previous critiques to develop a more far-reaching critique of deterrence. The scholars I have cited argue that deterrence sometimes fails because it is implemented poorly or applied in circumstances in which it is inappropriate. Their criticisms, and those of George Quester in Chap. 3, are directed primarily at the strategy of deterrence. I argue that deterrence is by its very nature a seriously flawed strategy and theory of conflict management. I do not believe that attempts to improve and reformulate the theory will produce a better fit between its expectations and observable behavior across cases.

The critique of deterrence that Janice Gross Stein and I have developed (Lebow and Stein 1987a) has three interlocking components: the political, psychological, and practical. Each exposes a different set of problems with the theory and strategy of deterrence. In practice, these problems are often linked; political and practical factors interact with psychological processes to multiply the obstacles to successful prediction of state behavior and successful conflict management.

The political component examines the motivations behind foreign policy challenges. Deterrence is unabashedly a theory of ‘opportunity’. It asserts that adversaries seek opportunities to make gains and pounce when they find them. Case studies of actual conflicts point to an alternative explanation for resorts to force, which we term a theory of ‘need’. The evidence indicates that strategic vulnerabilities and domestic political needs often constitute incentives to use force. When leaders become desperate, they may resort to force even when the military balance is unfavorable and there are no grounds for doubting adversarial resolve. Deterrence may be an inappropriate and even dangerous strategy in these circumstances. For if leaders are driven less by the prospect of gain than they are by the fear of loss, deterrent policies can provoke the very behavior they are designed to forestall by intensifying the pressures on the challenger to act.

The psychological component is also related to the motivation behind deterrence challenges. To the extent that policymakers believe in the necessity of challenging commitments of their adversaries, they become predisposed to see their objectives as attainable. When this happens, motivated errors can be pronounced and identifiable. They can take the form of distorted threat assessments and insensitivity to warnings that the policies to which our leaders are committed are likely to end in disaster. Policymakers can convince themselves, despite evidence to the contrary, that they can challenge an important adversarial commitment without provoking war. Because they know the extent to which they are powerless to back down, they expect their adversaries to accommodate them by doing so. Policymakers may also seek comfort in the illusion that their country will emerge victorious at little cost to itself if the crisis gets out of hand and leads to war. Deterrence can thus be defeated by wishful thinking.

The practical component of the critique describes some of the most important obstacles to implementing deterrence. These derive from the distorting effects of cognitive biases and heuristics, political and cultural barriers to empathy, and the differing cognitive contexts the deterrer and would-be challengers are apt to use to frame and interpret signals. Problems of this kind are not unique to deterrence; they are embedded in the very structure of international relations. They nevertheless constitute particularly severe impediments to deterrence because of the deterrer’s need to understand the world as it appears to the leaders of a would-be challenger in order to manipulate effectively their cost-benefit calculus. Failure to do this in the right way can result in deterrent policies that actually succeed in making the proscribed behavior more attractive to a challenger.

The first two components of this critique challenge core assumptions of deterrence theory . The third component is directed at the strategy of deterrence. But it also has implications for deterrence theory . If the strategy of deterrence is so often unsuccessful because of all of the practical difficulties associated with its implementation, then the theory of deterrence must be judged a poor guide to action.

In assessing deterrence theory it is imperative to distinguish between the motives and opportunity to carry out a military challenge. Classical deterrence theory takes as a given a high level of hostility on the part of the adversary and assumes that a challenge will be made if the opportunity exists. In the absence of opportunity, no challenge will occur even though hostility remains high. Because it sees aggression as opportunity driven , deterrence theory prescribes defensible, credible commitments as the best way to prevent military challenges.

Our case material points to the importance of motive; hostility cannot be treated as a constant. In practice, it waxes and wanes as a function of specific foreign and domestic circumstances. There are, moreover, few states or leaders who are driven by pure hostility toward their adversaries. Hitler is the exception, not the rule. This is not to deny the existence of opportunity-based challenges . Postwar examples include India’s invasion of Pakistan in 1971, Iraq’s invasion of Iran in 1980, and Israel’s invasion of Lebanon in 1982.

In most adversarial relationships, leaders resort to military challenges only in extraordinary circumstances. Our cases suggest that this is most likely to occur when leaders confront acute political and strategic vulnerabilities. In these circumstances, military challenges may be carried out even when there is no apparent opportunity to do so. Leaders may convince themselves, quite without objective reason, that such opportunity exists. When leaders do not feel impelled by political and strategic needs, they are unlikely to carry out challenges even when they perceive the opportunity to do so.

The matrices in Tables 3.1 and 3.2 summarize some of the most important differences between the classical theory of deterrence and our findings about military challenges. These differences are explained by the political and psychological components of our critique of deterrence. The third component of that critique, the practical difficulties of implementing deterrence, pertains primarily to deterrence as a strategy. However, to the extent that it indicates the pervasive presence of serious obstacles in the way of applying deterrence, it suggests that deterrence theory is not a good guide for formulating strategy.

Table 3.1 Deterrence matrix
Full size table
Table 3.2 Lebow-Stein matrix
Full size table

In the real world, there can be no truly dichotomous distinction between opportunity and need as motives for military challenges. Many, if not most challenges, contain elements of both motives. In the case of Iran –Iraq , many analysts (Heller 1984; Tripp 1986) argue that Iraq attacked because of a complex mixture of motives. It saw the opportunity to take advantage of Iran ’s internal disarray—clearly a motive of opportunity—but also acted out of fear that the Ayatollah Ruhollah Khomeini would attempt to export Iran’s revolution in order to overthrow Iraq’s regime. Most of the cases we examined were nevertheless skewed toward one or the other of the extremes. For the purposes of analysis we have classified these cases accordingly.

Data and Method

Most of the evidence on which this analysis is based comes from historical case studies Janice Stein and I have published in Between Peace and War (Lebow 1981), Psychology and Deterrence (Jervis et al. 1985) and various articles. Both books analyzed deterrence encounters from the perspective of both sides; they examined the calculations, expectations, and actions of the challenger as well as those of the would-be deterrer. As the key to understanding deterrence successes and failures lies in the nature of the interactions between the adversaries, case studies of this kind shed more light on these phenomena than analyses of the deliberations and policies of only one of the involved parties.

Most of these cases are deterrence failures. Janice Stein and I have chosen to work with failures because they are more readily identifiable, thereby facilitating the construction of a valid universe of cases. Deterrence successes can result in inaction. Failures, by contrast, lead to serious crises and often to wars. Events of this kind are not only highly visible but almost always prompt memoirs, official inquiries, and other investigations that provide the data essential for scholarly analyses. Deterrence failures are also more revealing than deterrence successes of the complexities of international relations. Understanding why deterrence fails may lead to insights into the nature of conflict as well as to a more general understanding of the circumstances in which deterrence is likely to succeed or fail and the reasons why this is so. Nevertheless, restricting the analysis to deterrence failures imposes costs. Explanations of the causes of failure can only be tentative, because some of the factors that appear to account for failure may also be at work when deterrence succeeds. Hypotheses derived from a controlled comparison of cases of deterrence failure will ultimately have to be validated against identified instances of deterrence success.

What is a deterrence failure? The goal of deterrence is to dissuade another actor from carrying out a proscribed behavior. In the context of international relations, the most important objective of deterrence is prevention of a use of force. To do this, the theory stipulates that the deterrer must carefully define the unacceptable behavior, make a reasonable attempt to communicate a commitment to punish transgressors (or deny them their objectives), possess the means to do this, and demonstrate the resolve to carry through on its threat (Kaufmann 1954; Kissinger 1960; Brodie 1959; Kaplan 1958; Milburn 1959; Quester 1966; Schelling 1966: 374). 2 When these conditions are met and the behavior still occurs, we can speak of a deterrence failure. 3

Researchers can and do disagree among themselves about the extent to which any or all of these conditions were met in a specific instance. These disagreements usually concern the credibility of the threat, something deterrence theorists consider to be the quintessential condition of the strategy’s success. Unfortunately, it is also the most difficult to assess. This can be a serious problem, as it hinders a determination of whether a deterrence failure was due to the inadequacy of the strategy or merely to the failure of the country in question to implement it adequately. Deterrence supporters invariably argue the latter when critics make the case for the former (Orme 1987; Lebow 1987b).

The ongoing debate about the efficacy of deterrence is fueled by the inherent subjectivity of all interpretations of historical events. One way to cope with this problem is to use a sample large enough to minimize the significance of disagreements about individual cases. My arguments are therefore based on 10 examples of deterrence failure. 4 These cases are interesting not only because they document a pattern of deterrence failure, but because they illustrate diverse reasons why failures occur. Evidence from conflicts in which leaders used other kinds of strategies of conflict management will also be introduced where it is relevant.

There is a further difficulty that arises from presenting arguments based on case studies. In contrast to experimental or survey research, it is impossible to summarize data of this kind in a succinct manner. Nor would such a summary establish the validity of the findings even if it can be demonstrated that the nature of the database and the data analysis conformed to accepted research practice. As I have already observed, the reader must be convinced of the correctness of our interpretation of individual cases. Consequently, it is important to convey something of the flavor of the cases and the basis for our interpretation of the evidence. For this reason, I have chosen to incorporate as much case material as space permits. Readers who are interested in the data are referred to Psychology and Deterrence, Between Peace and War, and the several journal articles cited for a fuller exposition of the cases.

Political Failings

Deterrence theory assumes that utility, defined in terms of the political and material well-being of leaders and their states, can readily be measured. But political and national interests are subjective concepts. They are perceived different by different leaders, making it extremely difficult for outsiders to determine, let alone measure. It is even more difficult, if not impossible, to weigh the relative importance of emotional, intangible, unquantifiable concerns that history reveals to be at least as important for most peoples, Americans included, as narrow calculations of political interest. Why, for example, did the South challenge the North, which was clearly superior in military power and potential? Why did the Confederacy continue the struggle at tremendous human and economic cost long after leaders and soldiers alike recognized it to be a lost cause? Other examples can be cited where people wittingly began or continued a struggle against great or even impossible odds in the face of prior and even convincing efforts by the superior military power to portray the certain and disastrous consequences of a military challenge or continued resistance. From the Jewish revolts against the Romans to the Irish Easter Rising and the resistance of the beleaguered Finns in 1940, history records countless stories of peoples who began or continued costly struggles with little or no expectation of success. Honor, anger, or national self-respect proved more compelling motives for action than pragmatic calculations of material loss and gain were reasons for acquiescence or passivity.

Both the theory and strategy of deterrence mistake the symptoms of aggressive behavior for its causes. Specifically, it ignores the political and strategic vulnerabilities that can interact with cognitive and motivational processes to compel leaders to choose force.

In a previous study, I analyzed a class of acute international crisis, brinkmanship , whose defining characteristic was the challenger’s expectation that its adversary would back away from its commitment in preference to war (Lebow 1981). I found that, much more often than not, brinkmanship challenges were initiated without good evidence that the adversary lacked either the capability or resolve to defend its commitment; on the contrary, in most instances the evidence available at the time pointed to the opposite conclusion. The commitments in question appeared to meet the four necessary conditions of deterrence: they were clearly defined, repeatedly publicized, and defensible, and the defending states gave every indication of their resolve to use force in defense of them. Not surprisingly, most of these challenges resulted in setbacks for the initiators, who were themselves compelled to back down or go to war.

Faulty judgment by challengers could most often be attributed to their perceived need to carry out a brinkmanship challenge in response to pressing foreign and domestic threats. The policymakers involved believed that these threats could be overcome only by means of a successful challenge of an adversary’s commitment. Brinkmanship was conceived of as a necessary and forceful response to danger, as a means of protecting national strategic or domestic political interests before time ran out. Whether or not their assessment of international and domestic constraints was correct is a separate question for research. What is relevant is that leaders perceived acute domestic pressure, international danger, or both.

The extent to which policymakers contemplating challenges of their adversaries are inner-directed and inwardly focused is also a central theme of Janice Gross Stein’s two contributions to Psychology and Deterrence (1985a, b). In her analysis of the five occasions between 1969 and 1973 when Egyptian leaders seriously contemplated the use of force against Israel, Stein argues that decision making in all of these instances departed significantly from the core postulates of deterrence theory . All five decisions revealed a consistent and almost exclusive concentration by Egyptian leaders on their own purposes, political needs, and constraints. They spoke in almost apocalyptic terms of Egypt’s need to liberate the Sinai before the superpower detente progressed to the stage where Egyptian military action became impossible. They alluded again and again to the escalating domestic crisis that could be arrested only if the humiliation of 1967 were erased by a successful military campaign. By contrast, Israel’s interests, and the imperatives for action that could be expected to flow from these interests, were not at all salient for Egyptian leaders. They thought instead of the growing domestic and international constraints and of the intolerable costs of inaction.

In 1969, in the War of Attrition, the Egyptian failure to consider the relative interests of both sides resulted in a serious error. Egyptian leaders did not miscalculate Israel’s credibility but rather the scope of Israel’s military response. They attached a very low probability to the possibility that Israel would extend the war and carry out deep penetration bombing attacks against Egypt and escalate its war objective to the overthrow of Nasser . This was a miscalculation of major proportions given the magnitude of the punishment Israel in fact inflicted upon Egypt .

Egypt’s inability to understand that Israel’s leaders believed that defense of the Sinai was important not only for the strategic depth and warning time it provided but also as an indicator of resolve was merely one cause of its miscalculation in 1969. Egyptian leaders overestimated their own capacity to lay down favorable ground rules for a war of attrition and underestimated that of Israel. They also developed a strategy to fight the war, to culminate in a crossing of the canal, that was predicated on a fatal inconsistency: the belief that Egypt could inflict numerous casualties on Israel in the course of a war of attrition, but that Israel would refrain from escalating that conflict in order to reduce its casualties.

These faulty assessments and strategic contradictions are best explained as a motivated response to the strategic dilemma faced by Egyptian planners in 1969. Egypt could neither accept the status quo nor sustain the kind of military effort that would have been necessary to recapture the Sinai. Instead, Egypt embarked upon a poorly conceived limited military action. The wishful thinking and biased estimates were a form of bolstering; this was the way Egyptian leaders convinced themselves that their strategy would succeed. Israel’s deterrent failed, not because of any lack of capability or resolve, but because Egypt’s calculations were so flawed that they defeated deterrence.

Egyptian decision making in 1969 provides an example of what may be the most frequent cause of serious miscalculation in international crisis: the inability of leaders to find a satisfactory way to reconcile two competing kinds of threats. Our cases indicated that the psychological stress that arises from this decisional dilemma is usually resolved by the adoption of defensive avoidance as a coping strategy. Leaders commit themselves to a course of action and deny information that indicates that their policy might not succeed (Janis and Mann 1977). In the Egyptian case, the decisional dilemma that prompted defensive avoidance was the result of incompatibility between domestic imperatives and foreign realities. The domestic threat, the political and economic losses, was the overriding consideration for Egyptian policymakers. Their estimates of their vulnerability motivated error and miscalculation and culminated in the failure of deterrence.

The Egyptian decision to use force in 1973 was even more damaging to the logic of deterrence than the motivated miscalculation in 1969. Egyptian leaders chose to use force in 1973 not because they miscalculated Israel’s resolve or response but because they felt so intolerably vulnerable and constrained. If Egyptian leaders had miscalculated, proponents of deterrence might argue that human error accounted for its failure. Economists advance similar kinds of arguments: The strategy is not flawed, only the people who use it. Egypt’s leaders decided to challenge deterrence not because they erred but because they considered the domestic and foreign costs of inaction unbearably high. They anticipated correctly a major military response by Israel and expected to suffer significant casualties and losses. Nevertheless, they planned a limited military action to disrupt the status quo and hoped for an internationally imposed cease-fire before their limited gains could be reversed. In 1973, Egyptian leaders considered their military capabilities inferior to those of Israel but chose to use force because they anticipated grave domestic and strategic consequences from continuing inaction.

The same domestic considerations that compelled Egyptian leaders to challenge Israel also provided the incentives for Egyptian military planners to devise a strategy that compensated for their military weakness. Human ingenuity and careful organization succeeded in exploiting the flexibility of multipurpose conventional weaponry to circumvent many of the constraints of military inferiority. Egyptian officers strove to achieve defensive superiority in what they planned to keep a limited battle zone (Stein 1985a).

The Japanese decision to attack the USA in December 1941 seems analogous to the Egyptian decision of 1973. Like the Egyptians, the Japanese fully recognized the military superiority of their adversary, particularly the greater naval power and vastly superior economic base of the USA. The Japanese, nevertheless, felt compelled to attack the USA in the illusory hope that a limited victory would facilitate a favorable settlement of their festering and costly conflict with China .

As the Egyptians were to do more than 30 years later, the Japanese military devised an ingenious and daring strategy to compensate for their adversary’s advantages; they relied on air power and surprise to neutralize US naval power in the Pacific. They too deluded themselves that their foe would accept the political consequences of a disastrous initial defeat instead of fighting to regain the initiative. The Japanese strategy was an act of desperation. Japan’s leaders opted for war only after they were persuaded that the military balance between themselves and their adversaries would never again be as favorable as it was in 1941; time was working against them. They were also convinced that they could not attain their objectives by diplomacy (Butow 1961; Borg and Okamoto 1973; Ienaga 1978; Ike 1967; Russett 1967; Hosoya 1968).

The Japanese case highlights the importance of an uncongenial strategic environment as an incentive for a challenge. Leaders who anticipate an unfavorable decline in the relative balance of power may see no alternative to military action. President Sadat, for example, estimated that the longer he postponed war, the stronger Israel would become. This assumption helped to create a mood of desperation in Cairo, so much so that Sadat repeatedly purged the Egyptian military command until he found generals who were confident that they could design around Israel’s air and armored capability.

The Egyptian and Japanese cases indicate that a defender’s capability and resolve are only some of the factors challengers consider when they contemplate war. They are also influenced by domestic political pressures that push them toward action and their judgments about future trends in the military balance. A pessimistic estimate of the probability of achieving important goals by peaceful means can also create frustration and constitute an incentive to act. This was very much so in Egypt in 1973 and in Japan in 1941. Both these cases illustrate how frustration, pessimism, and a sense of weakness in response to an unfavorable domestic and strategic environment can outweigh considerations of military inferiority.

How Deterrence Can Backfire

When challengers are vulnerable or feel themselves vulnerable, a deterrer’s effort to make important commitments more defensible and credible will have uncertain and unpredictable effects. At best, deterrence will be benign; it will simply have no effect. But it can also be malignant by intensifying precisely those pressures that are pushing leaders toward a choice of force. Japan offers an example.

The USA and other Western powers imposed first an asset freeze and then an oil embargo upon Japan in July–August 1941 in the hope of moderating Tokyo’s policies. These actions were in fact the catalysts for Japan’s decision to go to war. Her leaders feared that the embargo would deprive them of the means of continuing their struggle against China and would ultimately put them at the mercy of their adversaries. It accordingly fostered a mood of desperation in Tokyo, an essential precondition for the attack on Pearl Harbor that followed.

In his contribution to Psychology and Deterrence, Jack Snyder (1985: 153–179) explores security dilemmas and their role in the outbreak of war in 1914. The distinguishing characteristic of a security dilemma is that behavior perceived by adversaries as threatening and aggressive is actually a defensive response to an inhospitable strategic environment. A perceptual security dilemma develops, Snyder argues, when strategic and psychological factors interact and strategic assessments are exaggerated or distorted by perceptual biases. In effect, leaders overrate the advantages of the offensive, the magnitude of power shifts, and the hostility of their adversaries.

In 1914, the major continental powers confronted elements of a security dilemma. As French fortifications improved in the 1880s, German security required the vulnerability of Russian forces in Poland ; without this vulnerability, the German general staff feared that Russia and France could mobilize to full strength and then attack jointly. Russian security, however, excluded precisely such a weakness: Russia could not tolerate a decisive German advantage in a short war and so planned to increase her standing forces 40% by 1917. With French financial assistance, Russia also constructed new railways to transport these forces more rapidly to her western borders. Defensive preparations by Russia constituted an offensive threat to Germany , and conversely, a defensive strategy by Germany seemed to require offensives directed against France and Russia. Offense and defense thus became virtually indistinguishable.

Although the strategic environment was inhospitable and dangerous, Germany’s military leaders greatly exaggerated the dangers and, as Snyder (1985: 170) demonstrates, reasoned inside out. They overrated the hostility of their adversaries and consequently assumed the inevitability of a two-front war. Once they did, the attractiveness of a preventive war-fighting strategy became overwhelming; German military leaders saw preventive war as the only alternative to their vulnerability. Indeed, the general staff gave no serious consideration after 1890 to the possibility of a defensive strategy against Russia and France. From then until 1914, the German military did not overestimate their offensive capabilities and then choose force; on the contrary, they exaggerated the hostility of their adversaries in ways that psychological theories expect and then argued that an offensive capability was the least unsatisfactory option. Because of this choice, Germany ’s neighbors confronted a real security dilemma.

In this kind of strategic environment, the attempt to deter Germany was counterproductive. Threats of retaliation and shows of force by Russia and France only fueled German fears and, in so doing, further destabilized an already unstable environment. The Russian mobilization designed to deter, for example, could not help but alarm German military leaders committed to an offensive preemptive strategy. In 1914, when Germany’s leaders chose to use force, they did so not because they saw an ‘opportunity’ for gain but because they believed the strategic consequences of inaction would be catastrophic. In an environment where already unfavorable strategic assessments were overlain by exaggerated fear and a sense of vulnerability, deterrence could only provoke the use of force it was designed to prevent.

Psychological Problems

The psychological component of this critique is also related to the motivation behind deterrence challenges. To the extent that policymakers believe in the necessity of challenging commitments of their adversaries, they become predisposed to see their objectives as attainable. Motivated error in the form of distorted threat assessments can result in the unrealistic expectation that an adversary will back down when challenged or, alternatively, that it will fight precisely the kind of war the challenger plans for. Once committed to a challenge, policymakers may also become insensitive to warnings that their chosen course of action is likely to result in disaster. In these circumstances, deterrence, no matter how well it is practiced, can be defeated by a challenger’s wishful thinking.

Flawed Assessments

I have already described Egypt’s flawed assessments in 1969 and 1973. In 1969, the Egyptians convinced themselves that Israel would engage in a costly war of attrition along the canal, despite the well-known fact that Israeli strategy had always been premised on the need to avoid this kind of conflict. In 1973, the Egyptians assumed that Israel would accept the loss of her positions along the east bank of the Suez Canal as a fait accompli despite Egypt’s own inability to do this for much the same reasons. Both expectations flew in the face of obvious political realities. They led to costly wars that very nearly ended in political disaster. Both resorts to force were a response to Egypt’s leaders’ need to reassert their strength at home and abroad, a need that prompted grossly distorted estimates of Israel’s likely responses to a challenge.

The Japanese decision to attack Pearl Harbor is another example of a strategic decision based on wishful thinking. The Japanese military settled on a limited war strategy because they knew that it was the only kind of war they could hope to win against the USA, given the latter’s superior economic and military power. They convinced themselves that a successful counterforce strike against US naval units in the Pacific would convince Washington to withdraw from the Western Pacific and give Japan a free hand in the region. The American reaction was, of course, nothing of the kind. Public opinion in the USA was enraged by Japan’s ‘sneak attack’ and intent on waging war against her a Voutrance. President Roosevelt and Chairman of the Joint Chiefs of Staff George C. Marshall had a difficult time throughout the war in directing America’s principal military effort against Germany , which they rightly concluded constituted the more serious threat, because public opinion was more interested in punishing Japan.

The origins of World War I offer a third example of how wishful thinking can defeat deterrence. German policy in the July crisis was based on a series of erroneous assumptions on the probable Russian, French, and British reaction to the destruction of Serbia by Austria-Hungary . German leaders were on the whole confident of their ability to localize an Austro-Serbian war despite all the indications to the contrary and, one fleeting moment of hesitation by the German chancellor aside, urged Vienna throughout the crisis to ignore all pleas for moderation.

Germany’s strategy was remarkably shortsighted. Even if the unrealistic assumptions on which it was based had proved correct, it still would have been self-defeating. Serbia’s destruction would only have aggravated Russo-German hostility, making Russia even more dependent on France and Britain and setting the stage for a renewed and more intense clash between the two blocs. This outcome aside, all of the assumptions on which German policy was based proved ill-founded; Austria’s declaration of war on Serbia triggered a series of responses that embroiled Germany in a war with Russia, France , Belgium , and Great Britain .

The German strategy only makes sense when it is understood as a response to the contradictions between the country’s perceived strategic needs and perceived strategic realities. The former dictated support of Austria, Germany’s principal ally, as a means of shoring up her self-confidence and maintaining the all-important alliance. The latter dictated caution because Germany’s politicians shied away from responsibility for a European war while her generals were uncertain of their ability to win one. These contradictions were reconciled in a strategy premised on the illusion that Austria, with German support, could wage a limited war in the Balkans without provoking the intervention of the other great powers. German leaders were only disabused of their illusion after it was too late to alter the course of events (Lebow 1981: 26–29, 119–124; 1984).

Challenger’s Insensitivity to Warnings

Motivated errors can play a major role in blocking receptivity to signals. Once leaders have committed themselves to a challenge, efforts by defenders to impart credibility to their commitments will at best have a marginal impact on their adversaries’ behavior. Even the most elaborate efforts to demonstrate prowess and resolve may prove insufficient to discourage a challenger who is convinced that a use of force is necessary to preserve vital strategic and political interests.

Irving Janis and Leon Mann (1977) , in their analysis of decision making, argue that policymakers who contemplate a course of action, but recognize that their initiative entails serious risk, will experience psychological stress. They will become emotionally upset and preoccupied with finding a less-risky alternative. If, after further investigation, they conclude that it is unrealistic to hope for a better strategy, they will terminate their search despite their continuing dissatisfaction with available options. The result is a pattern of ‘defensive avoidance’, characterized by efforts to avoid, dismiss, and deny warnings that increase anxiety and fear.

One of the three forms of defensive avoidance identified by Janis and Mann (1977: 57–58, 107–133) is bolstering. It refers to a set of psychological tactics that policymakers may resort to make a decision they are about to make, or have already made, more acceptable to themselves. Bolstering occurs when policymakers have lost hope of finding an altogether satisfactory policy option and are unable to postpone a decision or shift responsibility to others. Instead, they commit themselves to the least objectionable alternative and proceed to ‘spread the alternatives’, that is, to exaggerate its positive consequences or minimize its costs. They may also deny the existence of aversive feelings, emphasize the remoteness of the consequence, or attempt to minimize personal responsibility for the decision once it is made. Policymakers continue to think about the problem but ward off anxiety by practicing selective attention and other forms of distorted information processing.

Bolstering can serve a useful purpose. It helps a policymaker forced to settle for a less than the optimal course of action to overcome residual conflict second move more confidently toward decision and action. Bolstering can occur before and after a decision is made. When it takes place before, it discourages leaders from making a careful search of the alternatives. It subsequently lulls them into believing that they have made a good decision, when in fact they have avoided making a vigilant appraisal of the possible alternatives in order to escape from the conflict that would ensue. When leaders resort to bolstering after a decision, it tends to blind to warnings that the course of action to which they are committed may prove unsatisfactory or even disastrous.

Janis and Mann (1977: 74–79) consider insensitivity to warnings a hallmark of defensive avoidance. When this becomes the dominant pattern of coping, ‘the person tries to keep himself from being exposed to communications that might reveal the shortcomings of the course of action he has chosen’. When actually confronted with disturbing information, leaders will alter its implications through a process of wishful thinking; they rationalize and deny the prospect of serious loss. Extraordinary circumstances with irrefutable negative feedback may be required to overcome such defenses.

Selective attention, denial, or almost any other psychological tactic used by policymakers to cope with critical information can be institutionalized. Merely by making their expectations or preferences known, policymakers encourage their subordinates to report or emphasize information supportive of those expectations and preferences. Policymakers can also purposely rig their intelligence networks and bureaucracies to achieve the same effect. Perspectives thus confirmed and reconfirmed over time become more and more resistant to discrepant information and more difficult to refute.

In an earlier study, I (1981: 101–228) described in detail how this process occurred in Germany in 1914, in the USA in 1950 with regard to the possibility of Chinese entry into the Korean War , and in India in 1962 during its border dispute with China . In all three instances, policymakers, responding to perceived domestic and strategic imperatives, became committed to risky military policies in the face of efforts by others to deter them. They resorted to defensive avoidance to insulate themselves from the stress triggered by these warnings. They subsequently allowed or encouraged their respective political-military bureaucracies to submit reports supportive of the policies to which the leadership was committed. Institutionalized in this manner, defensive avoidance succeeded in blinding the policymakers to repeated warnings of impending disaster.

Motivated bias is a response to personal needs or external pressures. Evidence drawn from these cases suggests that at least one mediating condition of motivated bias is a choice by policymakers of a course of action that they recognize could result in substantial loss. Once challengers become committed to such an action, even the most strenuous efforts by a deterrer to define a commitment and give it credibility may have little impact. Motivated bias, in the form of faulty assessment of an adversary s resolve, overconfidence, and insensitivity to warnings, can defeat even well-articulated and well-executed deterrence.

Problems in Applying Deterrence

Deterrence is beset by a host of practical problems. One of these is the difficulty of communicating capability and resolve to would-be challengers. Strategies of deterrence generally assume that everyone understands, so to speak, the meaning of barking guard dogs, barbed wire, and ‘No Trespassing’ signs. This is not so. Signals only take on meaning in terms of the context in which they are interpreted. When sender and recipient use quite different contexts to frame, communicate, or interpret signals, the opportunities for miscalculation and misjudgment multiply. This problem is endemic to international relations and is not limited to deterrence (Jervis 1979: 305–310; Lebow 1985: 204–211).

A second problem, and one that is more specific to deterrence, concerns the difficulty of reconstructing the cost-benefit calculus of another actor. Deterrence requires the party intent on forestalling a challenge to manipulate the cost-benefit calculus of a would-be challenger so that the expected costs of a challenge are judged to outweigh its expected benefits. If credible threats of punishment always increased the cost side of the ledger—something deterrence theory takes for granted—then it would be unnecessary for deterrers to understand the value hierarchy and outcome preferences of target states. This convenient assumption is not borne out in practice. Leaders may be driven not primarily by ‘opportunity’ but rather by ‘vulnerability’. When they are, increasing the costs of military action may have no effect on their unwillingness to tolerate the high costs of inaction.

Deterrent threats in these circumstances can also provoke the very behavior they are designed to prevent. This happens when, contrary to the deterrer’s expectations, they intensify the pressures on the challenger to act. Unfortunately, the kinds of considerations that determine how a threat will influence an adversary’s cost-benefit calculus are often invisible or not easily understood from the outside.

The Cuban Missile Crisis offers a striking example of this phenomenon. Scholars have advanced several hypotheses to explain why the Soviets placed missiles in Cuba in September and October of 1962. By far the most widely accepted is the perceived Soviet need to redress the strategic balance. The deployment was a reaction to American pronouncements of strategic superiority in the fall of 1961 (Horelick and Rush 1966: 141; Hilsman 1967: 200–202; Tatu 1968; Abel 1966; Allison 1971: 52–56). At that time the Soviets possessed a very small fleet of long-range bombers, a sizable number of medium-range ballistic missiles (MRBMs) and intermediate-range ballistic missiles (IRBMs) and a small number of intercontinental ballistic missiles (ICBMs). All of these weapons were based in the Soviet Union and were of limited use in any retaliatory strike against the USA. The bombers were slow and easy to detect; they could not be expected to penetrate American air defenses. The medium- and intermediate-range ballistic missiles were excellent weapons but incapable of reaching the continental USA, and the first-generation ICBMs, for which the Soviets had great hopes, proved too unreliable and vulnerable to serve as a practical weapon. Only a few of them were actually deployed.

American estimates of the size and effectiveness of the Soviet missile force had been highly speculative after May 1960 when U-2 overflights of the Soviet Union had been discontinued. This situation was rectified in the late summer of 1961 by the introduction of satellite reconnaissance, which gave American intelligence a more accurate assessment of the number of Soviet missiles. At that time, a far-reaching political decision was made to tell Moscow that Washington knew of its vulnerability.

The risk inherent in such a course of action was not fully appreciated by President Kennedy, who feared only that the Soviets would now speed up their ICBM program. The president and his advisers were more sensitive to the need to moderate Khrushchev’s bellicosity, alarmingly manifest in his several Berlin ultimatums, and thought this could be accomplished by communicating their awareness of American strategic superiority. The message was first conveyed by Roswell Gilpatric , deputy secretary of defense, in a speech delivered in October 1961, and was subsequently reinforced through other channels.

For Soviet leaders, the political implications of this message must have been staggering. Almost overnight, the Kremlin was confronted with the realization that its nuclear arsenal was not an effective deterrent. In words of Roger Hilsman (1967: 164):

It was not so much the fact that the Americans had military superiority—that was not news to the Soviets. What was bound to frighten them most was that the Americans knew that they had military superiority. For the Soviets quickly realized that to have reached this conclusion, the Americans must have made intelligence breakthrough and found a way to pinpoint the location of the Soviet missiles that had been deployed as well as to calculate their total numbers. A soft ICBM system with somewhat cumbersome launching techniques is an effective weapon for both a first strike… and a second, retaliatory strike so long as the location of the launching pads can be kept secret. However, if the enemy has a map with all the pads plotted, the system will retain some of its utility as a first-strike weapon, but almost none at all as a second-strike weapon. The whole Soviet ICBM system was suddenly obsolescent.

The Soviets were in a quandary. The missile gap could be closed by a crash program to develop more effective second-generation ICBMs and perhaps a submersible delivery system. Such an effort was extremely costly and likely to meet strong opposition within the Soviet hierarchy. More importantly, a crash program did nothing to solve the short-term but paralyzing Soviet strategic inferiority that could be exploited by American leaders. The deployment of missiles in Cuba can be viewed as a bold attempt to resolve this dilemma. If this interpretation is correct, the American warning had the paradoxical impact of provoking the action it was designed to deter.

For 25 years, all interpretations of Soviet motives and policies in the missile crisis were speculative. Existing Soviet commentaries, among them Khrushchev ’s memoirs (1970, 1974) and Anatoly Gromyko’s study of the crisis (1971), contained enough obvious falsehoods to make them highly suspect sources. In October 1987, an extraordinary meeting took place in Cambridge, Massachusetts. A small group of scholars—myself among them—and former Kennedy administration officials met with three Soviet officials to talk about the origins and politics of the missile crisis. The Soviet representatives were Georgi Shaknazarov , a member of the Central Committee, Fedor Burlatsky , a former Khrushchev speech writer, and Sergei Mikoyan, a foreign ministry official and son of former deputy prime minister, Anastas I. Mikoyan .

All three Soviets were remarkably forthcoming; they shared with us their personal memories and feelings about the crisis and also what they had learned from talking to other officials at the time and subsequently. They did not always agree among themselves about important aspects of the crisis and were careful to distinguish between fact and opinion and between what they had witnessed or learned about only secondhand. The American participants, some of whom, like Raymond L. Garthoff and Robert McNamara , had extensive prior experience with Soviet officials, came away convinced that the Soviets were telling us the truth as they understood it.

The Russians advanced three explanations for the Cuban missile deployment: the perceived need to deter an expected American invasion of Cuba, to overcome Soviet strategic inferiority, and to attain political-psychological equality with the USA. They disagreed among themselves about the relative importance of these objectives for Khrushchev and other top leaders.

Sergei Mikoyan (1987: 20, 40, 45–47) maintained that ‘there were only two thoughts: defend Cuba and repair the [strategic nuclear] imbalance’. ‘Our “pentagon”’, he reported, ‘thought the strategic balance was dangerous, and sought parity’. Marshall Rodion Malinovsky , Soviet defense minister, was adamant about the need to secure a more credible second-strike capability. ‘Khrushchev’, too, Mikoyan continued, ‘was very concerned about a possible American attack. He worried… that somebody in the United States might think that a 17-to ~ l superiority would mean that a first strike was possible’. Mikoyan insisted, however, that Khrushchev’s primary objective was to prevent an American assault on Cuba, something the Soviet leadership believed to be imminent.

Fedor Burlatsky and Georgi Shaknazarov agreed that Khrushchev wanted to protect Castro but maintained that he was even more concerned to do something to redress the strategic imbalance. They gave somewhat different reasons for why Khrushchev sought to do this.

Georgi Shaknazarov (1987: 17–18, 58, 75–76) emphasized the military consequences of American superiority. ‘The main idea’, as he saw it, ‘was to publicly attain parity’. This was critical ‘because there were circles in the United States who believed that war with the Soviet Union was possible and could be won’. The Cuban missile deployment was accordingly attractive to Khrushchev because it offered an immediate solution to the strategic vulnerability problem at very little cost. ‘It was an attempt by Khrushchev to get parity without spending resources we did not have.’

Burlatsky (1987a: 17–18, 30–31, 115–116; b: 22) agreed that Soviet leaders had a long sense of nuclear inferiority, especially at this time. Many Soviet officials, he reported, really feared an American first strike. But Khrushchev did not. He worried instead about American efforts to exploit its superiority politically. Khrushchev was particularly aggrieved by the Kennedy administration’s deployment of missiles in Turkey , missiles that, because of their vulnerability to air attack, could only be used a first strike or political intimidation. ‘Why do the Americans have the right to encircle us with nuclear missile bases on all sides’, he complain to Burlatsky, ‘yet we do not have that right?’ Burlatsky believed that the Jupiter deployment in Turkey had been the catalyst for Khrushchev’s decision to send missiles to Cuba. ‘These missiles were not needed for deterrence’, he explained. ‘Our 300 were already more than enough to destroy the United States—more than enough. So it was a psychological thing. From my point of view, it was the first step to strategic parity’.

The Kennedy administration officials at the Cambridge conference admitted in retrospect that the Jupiter deployment had been provocative and unwise. Kennedy, it is apparent, had gone ahead with the deployment in spite of considerable opposition to it within State and Defense where the missiles were viewed as obsolescent and provocative. He did so because he was afraid that Khrushchev would misinterpret cancellation of the proposed deployment as a sign of his weakness and lack of resolve and become more emboldened in his challenges of Western interests in Berlin and elsewhere (Lebow, forthcoming).

Robert McNamara , secretary of defense during the Cuban crisis, was surprised that Khrushchev worried about a Cuban invasion. McNamara (1987: 59) assured the Russians ‘that we had no plan to attack Cuba, and I would have opposed the idea strongly, if it ever came up’. But he acknowledged that he could understand why the Soviets could have concluded that an invasion was imminent. The covert military operations the administration was conducting against Castro’s regime, he agreed, conveyed the wrong impression about American intentions. They were ‘stupid but our intent was not to invade’. From the vantage point of a quarter-century, McNamara was struck by the irony of the situation. ‘We thought those covert operations were terribly ineffective’, he mused, ‘and you thought they were ominous’.

McNamara and former national security advisor, McGeorge Bundy , were also surprised that the Soviets could have worried about a first strike. They knew that the administration had no intention of carrying one out! McNamara (1987: 76) asked Shaknazarov: ‘Did your leaders actually believe that some of us thought it would be in our interest to launch a first strike?’ Shaknazarov: ‘Yes. That is why it seems to me Khrushchev decided to put missiles in Cuba’ (1987: 76). The Russians went on to explain that their fears of a first strike had been aroused by the Kennedy administration’s strategic buildup, its stationing of first-strike weapons in Turkey , and claims by its defense and military leaders that they could destroy the Soviet Union without losing more than 25% of its own population in a counterblow.

Although none of the Soviets mentioned the Gilpatric speech, they all emphasized the extent to which American military preparations and assertions of strategic superiority exacerbated Soviet strategic insecurities and pushed Khrushchev toward more confrontational policies. From this it is clear the Gilpatric speech and related American attempts to manipulate the cost-calculus of Soviet leaders backfired. The Kennedy administration’s efforts to increase the cost to the Soviet Union of any challenge had the real and undesired effect of making such a challenge more attractive to Soviet leaders.

The missile crisis indicates that deterrence, as practiced by both superpowers, was provocative instead of preventive. Khrushchev and other top Soviet leaders conceived of the Cuban missiles as a means of deterring American military and political threats to Cuba and the Soviet Union. The American actions that provoked Khrushchev had in turn been envisaged by President Kennedy as prudent, defensive measures against perceived Soviet threats. Both leaders, seeking to moderate the behavior of their adversary, helped to bring about the very kind of confrontation they were trying to prevent.

The Primacy of Self

Deterrence purports to describe an interactive process between the defender of a commitment and a would-be challenger. The defending state is expected to define and publicize its commitment and do its best to make that commitment credible in the eyes of its adversary. Would-be challengers are expected to assess accurately the defender’s capability and resolve. The repetitive cycle of test and challenge is expected to provide both sides with an increasingly sophisticated understanding of each other’s interests, propensity for risk taking, threshold of provocation, and style of foreign policy behavior.

My analysis of adversarial relationships indicates that the expectations that deterrence has about deterrer and challenger bear little relationship to reality. Challengers frequently focus on their own needs and do not consider, or distort if they do, the needs, interests, and capabilities of their adversaries. Moreover, at times they are motivated not by ‘opportunity’, as deterrence theory expects, but rather by ‘vulnerability’ and perceived weakness. Deterrers, in turn, may interpret the motives or objectives of a challenger in a manner consistent with their expectations, with little regard to the competing expectations of the challenger. Both sides may also prove insensitive to each other’s signals. Under these conditions, deterrence is likely to fail. Even recurrent deterrence episodes may not facilitate greater mutual understanding. On the contrary, experience may actually hinder learning to the extent that it encourages tautological confirmation of misleading or inappropriate lessons.

Implications for Deterrence Theory

Some empiricists (Achen and Snidal 1988; Tetlock 1987; Huth and Russett 1984) contend that Stein and I misunderstand the purpose of social science theory. They argue that its goal is to predict human behavior, not necessarily to explain why that behavior occurs. With respect to deterrence, they insist that our sample, based only on deterrence failures, significantly biases our results. If we looked at deterrence successes, they argue, we would discover that deterrence is more successful than not. This would confirm the validity of deterrence theory as a predictor of state behavior.

These criticisms are methodologically and conceptually misguided. No analyst has yet succeeded, or ever will, in identifying the relevant universe of cases. Some of the reasons for this were made clear in the introduction of the chapter (see also Lebow and Stein 1987b). Chief among these is the difficulty of identifying deterrence successes. The more successful deterrence is, the fewer the behavioral traces it leaves behind. The assertion that deterrence theory is a better-than-average predictor of state behavior cannot be demonstrated. The several empirical studies that attempt to validate this claim do not come to grips with this and other methodological obstacles.

Because the universe of relevant deterrence cases can never be identified, the significance of the cases examined becomes critical. But aggregate data analyses have made no attempt to weight their cases in favor of those they deem the most important or critical. Instead, they treat them as equivalent in every respect. Analysts who use the case study approach, by contrast, make strenuous efforts to identify critical cases and to justify their choices. The case study literature on deterrence failure has now identified an impressive number of important deterrence failures. These cases, which led to major crises and wars, stand as a sharp challenge to the expectations of deterrence as a theory and its practice as a strategy.

These cases of deterrence failure reveal important common features. Flawed information gathering, evaluation, attribution, and decision making on the part of initiators were most often responsible for the miscalculations that defeated deterrence. These all-important processes cannot be captured by aggregate analysis. This requires in-depth study of individual cases of deterrence failure and success. As the old French saying goes, God is to be found in the details.

These same empiricists acknowledge the power of our argument about the importance of domestic politics and strategic vulnerabilities in pushing states toward military challenges. But they—and even some deterrence critics (George and Smoke 1974)—believe that deterrence theory could accommodate these considerations. To do so, they propose expanding it to incorporate a much wider range of political factors. Utility estimates could then take domestic politics, strategic vulnerabilities, and other factors into account instead of being based solely on narrow, and admittedly misleading, calculations of the relative military balance. This is easier said than done.

The incorporation of new variables would require an entirely new set of propositions to guide their weighing. Just how much importance, for example, should be given to national honor in comparison with the domestic political interests of leaders, ideological goals, or allied obligations? Deterrence theory provides no guidance for discriminating among these variables in order to construct a weighted model. For this reason, the Lebow-Stein critique of deterrence strikes at the theory, not only at deterrence as a real-world strategy.

Even if this problem could somehow be solved, deterrence theory would confront another insuperable obstacle. The incorporation of additional, more political variables would not help deterrence theory come to grips with the evidence that leaders deviate significantly from the process of rational choice in making critical foreign policy decisions. It is these deviations that largely account for deterrence failures. Models based on rational choice cannot therefore predict strategic decisions with any impressive degree of success. Analysts interested in improving the predictive capability of their models must abandon rational choice or, at the very least, incorporate significant elements of non-rational processes into their models. They cannot do this and retain the core of deterrence theory .

Notes

This chapter is based on Part I of ‘Beyond Deterrence’, coauthored with Janice Gross Stein , Journal of Social Issues (Winter 1987) 43, no. 4, 5–71. Research and writing of the paper were supported by grants from the Carnegie Corporation of New York to Richard Ned Lebow and the Canadian Institute of Peace and Security to Janice Gross Stein.

  1. 1.

    For a fuller treatment of the detrimental effects of nuclear deterrence between the superpowers, see Lebow (1987a).

  2. 2.

    The definition of adequate communication and apparent resolve is difficult. Students of deterrence have traditionally assessed credibility with reference to bow a would-be challenger’s leaders perceived the commitment in question. There is a serious problem with this approach; it risks making determinations of credibility tautological. If a commitment is challenged, it is assumed not to have been credible. This ignores the possibility that the commitment should have been seen as credible but was not for any one of a number of reasons independent of the defender’s military capability or resolve. For this reason, the appropriate test of credibility must be the judgment of disinterested third parties and not that of the would-be challenger. As 1 will show, a challenger’s receptivity to communications and its judgment about a commitment’s credibility can be impaired by motivated biases. Thus, deterrent threats that appear credible to third parties can fail to be perceived as such by leaders intent on a challenge.

  3. 3.

    George and Smoke (1974: 519–520) argue that the outcome of a deterrence encounter can also be mixed. This occurs, in their opinion, when deterrence succeeds in dissuading a country’s leaders from choosing certain options as too risky but does not prevent them from embarking upon another, less-risky challenge of the status quo . Such cases undoubtedly occur but I am not persuaded by the examples George and Smoke cite.

  4. 4.

    These cases are Fashoda (1898), Korea (1903–1904), Agadir (1911), July 1914, the Chinese entry into the Korean War (1950), Cuba (1962), the Sino-Indian crisis of 1962, and the Arab-Israeli wars of 1967, 1969, and 1973.