Home : News : Display
March 7, 2024

Rethinking the Relevance of Self-Deterrence

Jeffrey H. Michaels
©2024 Jeffrey H. Michaels

ABSTRACT: Self-deterrence is critically understudied in deterrence theory. Similarly, deterrence practitioners prefer to focus on adversaries’ threats rather than seeking to account for the full scope of fears influencing the decision calculus of policymakers. Through historical case studies, this article identifies where self-deterrence has occurred, highlights the benefits of incorporating the concept in future strategic planning and intelligence assessments, and recommends that policymakers, strategists, and analysts acknowledge self-deterrence as an important factor when preparing for future wars.

Keywords: deterrence, self-deterrence, chemical weapons, nuclear weapons, decision making

Policymakers, strategists, and scholars tend naturally to discuss deterrence in terms of deterring others (how does A deter B). In contrast, they devote hardly any attention to self-deterrence (in what ways does B deter B). Many analyses fail to provide a comprehensive examination of the consequences the adversary’s leadership fears and avoid identifying their self-imposed limitations on the use of force. They also overlook how political and military strategies can exploit an adversary’s self-deterrence (how can A’s strategy take advantage of B deterring B).

States refrain from taking military action for various reasons, including fear of consequences other than those threatened by their adversaries.1 Sometimes, the consequences an adversary could impose are mental phantoms for policymakers and strategists rather than approximate reflections of reality. There are also countless examples of leaders refraining from initiating wars due to the anticipated consequences imposed by third parties, domestic elites, and their constituents. In cases involving territorial conquest and occupation, leaders may fear the costs of long-term occupation more than the costs of the initial conquest. Thus, we should not restrict our understanding of the consequences that adversaries fear to the consequences that we have the means to impose directly on them, particularly as other types of fears may have a greater effect on an adversary’s decision calculus.

Normally associated with the tradition or taboo on the nonuse of nuclear weapons, scholars who refer to self-deterrence typically ignore its wider relevance and utility for other dimensions of military affairs and statecraft.2 This article draws attention to this understudied topic and highlights its relevance for crafting deterrence strategies. It first reviews the concept and its traditional definitions, then demonstrates how self-deterrence can explain war-related decisions and restrictions (especially on the weapons employed) as well as influence decisions about military interventions and covert actions.

Defining Self-Deterrence

In contrast to the then-burgeoning deterrence literature that emerged in the 1950s amidst the Golden Age of strategic studies, references to “self-deterrence” only began to appear in the late 1970s and early 1980s.3 At the time, self-deterrence was effectively synonymous with self-restraint, usually regarding the prospect of US or NATO unwillingness to use nuclear weapons. In some contexts, self-deterrence was misused, such as when referring to the fear of Soviet retaliation for a nuclear strike on the Soviet Union. Here, a simple reference to deterrence would have sufficed. In other contexts, it seemed more appropriate, such as when expressing concern that NATO would be unwilling to use nuclear weapons on member states’ territory—particularly West Germany— to repulse a Warsaw Pact invasion.

Robert Jervis provided the first meaningful scholarly treatment of self-deterrence in the early 1980s, noting: “States can successfully deter others unintentionally or unknowingly. Because actors can perceive things that are not there, they can be deterred by figments of their imagination—‘self-deterrence’ if you will.” Although Jervis’s definition represented an advance, the distinction between deterrence and self-deterrence remained unclear. Jervis’s case study of British fears of German air superiority in the 1930s illustrates this conceptual overlap. He observes that British fears of the damage the Germans might theoretically inflict—as opposed to the damage the Germans could or intended to inflict—played a key role inhibiting a more robust British policy. Thus, although the existence of a powerful German air force was a fact, the British significantly exaggerated its power, which had a corresponding deterrent effect on their policy calculations.4

Shortly after Jervis’s contribution, Richard Ned Lebow expanded the concept. Instead of focusing on self-generated and misperception-based fears, Lebow spoke about the “anticipated domestic political, moral, and psychological costs of a foreign policy” that he believed “may be at least as important a motive for moderation as deterrence in the traditional sense, brought about that is by fear of external punishment.”5 Lebow also complained that deterrence theorists “have generally failed to take these considerations into account.”6 The identification of “domestic political, moral, and psychological costs” expanded upon Jervis’s more limited focus on military consequences. After that, self-deterrence was increasingly associated with a taboo. Although conceptually useful, its strategic implications remained largely unaddressed.

Two decades later, in his discussion of “internalized deterrence,” Lawrence Freedman observed that an actor might avoid actions because of “mythical fears about the possible consequences” and the “thought of how the target might respond.”7 Unlike other scholars who referred to self-deterrence to explain inaction or sought to overcome its constraints, Freedman highlighted how to use it as part of a strategy. After all, deterrence means inducing fear. Determining what available military and other means A had to deter B normally characterized A’s strategy. This approach unintentionally emphasized only B’s fears relevant to A’s means.8 Instead, a strategic mindset stressing self-deterrence would concentrate on B’s fears, irrespective of A’s means. Even if these fears included things A could not directly threaten, they might nevertheless include fears that A could exacerbate, for instance, through psychological manipulation, or, as Freedman put it, “no more than a hint there and a quiet word there.”9 At a minimum, exacerbating B’s fears in this way could improve A’s chances of successful deterrence.

Additional studies also framed nuclear non-use in terms of self-deterrence. For instance, Scott Sagan focused on the nonuse of nuclear weapons against nonnuclear states, noting that self-deterrence can result from “pressures from important allies or public opinion.”10 Similarly, T. V. Paul defined self-deterrence as “self-imposed reputational concerns arising from moral, legal, and other normative considerations.”11 While reputational consequences inhibit nuclear weapon use, particularly against nonnuclear states and non-state actors, and therefore constitute an important element of self-deterrence, the exclusive focus of Paul and other scholars on reputational concerns unduly restricts the concept’s wider applicability.

Before considering a less restrictive definition of self-deterrence, we must understand deterrence. Alexander L. George and George Smoke offer this commonly used definition: “Deterrence is simply the persuasion of one’s opponent that the costs and/or risks of a given course of action he might take outweigh its benefits.”12 Many similar definitions of deterrence contain two elements: two actors (A and B) and the communication of consequences intended to raise the costs of a contemplated action. These definitions often include but fail to distinguish between reference to deterrence as a process and an effect. The former alludes to the threat of consequences (as an input into the decision calculus), whereas the latter refers to success or failure (whether one has deterred or not deterred).

The same basic principles apply to self-deterrence; only the actor changes (B deters B instead of A deterring B). Rather than external adversaries threatening consequences (as in deterrence), self-deterrence consists of a distinct set of consequences articulated internally by policymakers that negatively affects their cost-benefit calculus. More precisely, whereas an adversary might threaten military punishment to deter an unwanted action, many other plausible or fanciful consequences that will affect the decision to take that action fall outside the purview of whatever the adversary threatens. These consequences can inc lude reputational damage from a hostile domestic or world audience, third-party intervention, and the long-term political, economic, social, and military problems that would arise regardless of any short-term military success. Policymakers might also have erroneous beliefs about an adversary’s capabilities and intentions. For example, policymakers may base expectation of an adversary ’s behavior or the consequences of pursuing a course of action on little more than an extrapolation from a historical analogy.13 So much depends, therefore, on policymakers’ conceptions of war and the consequences they associate with them.

There is also considerable overlap in terms of deterrence as an effect. Consider this basic definition of deterrence success: “A potential attacker’s restraint from using force because of the deterrer’s threats.”14 Other definitions also stress the importance of deterrent threats determining the potential attacker’s decision not to use force.15 Similarly, success in the context of self-deterrence can be defined as a potential attacker’s restraint due to a fear of consequences other than those threatened by the adversary. One can often find traces of self-deterrence in policymakers’ discourse when the prospect of military action is being contemplated. In some cases, policymakers will reject military action. In others, limited military action will go ahead. During these deliberations, reasons not to take military action or to restrict it will typically arise.

Crucially, the consequences policymakers fear will often not be threatened but may still be assumed. For example, throughout the Cold War, NATO did not threaten the Soviet Union with a large-scale resistance movement. Instead, NATO based its approach to deterrence on conventional defense and nuclear use. To threaten guerrilla warfare meant conceding that NATO territory might be seized. In stark contrast, the deterrence policies of non-NATO countries such as Yugoslavia and Switzerland relied heavily on the threat of guerrilla warfare. Nevertheless, leaders contemplating a foreign invasion may still fear a costly occupation, even if not threatened by the adversary. On the other hand, when threatened, it is unlikely that fears of occupation will affect policymakers if they believe the population will welcome the invaders as liberators. So much of the outcome depends on their conceptions of war, values, and fears, not what we think they should fear.

Precursor to the Nonuse of Nuclear Weapons

While scholars often associate self-deterrence with the nonuse of nuclear weapons, many of the concept’s features appeared long before 1945. Prior to World War I, there was significant opposition to using chemical weapons in war. This inhibition could be traced to the long-standing norm against using poison, later codified in various legal conventions, most notably at the 1899 and 1907 Hague Conventions.16 At the root of the matter was the belief that certain weapons were too horrific to use in combat or against civilian targets. The emergence of international norms led to fears associated with violating these norms. To engage in proscribed practices would damage one’s reputation, with all the attendant political and diplomatic consequences. This fear led governments to institute special controls to ensure that only senior leaders could authorize use of these weapons.

Amidst the carnage of World War I—often characterized as a “total war”—both sides breached the norm on chemical weapons use but restricted it to the battlefield. Governments rejected calls to extend its use to enemy civilian targets. Although both sides grew desperate to break the battlefield stalemate, two main arguments persuaded leaders to oppose the use of chemical weapons on enemy cities. The first was the fear of retaliation, a classic deterrence argument. The second was the anticipated reputational damage and associated consequences resulting from engaging in this type of warfare, or at least engaging in it first. One fundamental reason norms were taken seriously was the broader fear of alienating public opinion in neutral countries, which raised the risk of bringing them into the war on the wrong side.17 Which of the two arguments was more decisive in preventing the expanded use of chemical weapons is less noteworthy than the fact that both arguments strongly influenced the decision calculus.

Despite the lack of chemical weapons use on cities during World War I, the negative experiences and images that arose from their battlefield use led to the 1925 Geneva Protocol, which reaffirmed the earlier nonuse norm. By the start of World War II, chemical weapons use against both civilian and military targets was viewed as reprehensible. Nevertheless, during the interwar period, military theorists contemplated the prospect of widespread chemical weapons use at the outset of a future European war. Some governments maintained arsenals of these weapons as an existential deterrent and insurance policy if the norms were not adhered to and deterrence failed. Furthermore, they took extensive civil defense precautions, including distributing gas masks to their populations. These visions of future war proved ill-founded, however. It was not until May 1942 that concerns about German chemical weapons use against the Soviets led to the first extended deterrence threat to retaliate if these weapons were employed. That month, UK Prime Minister Winston Churchill publicly stated:

I wish now to make it plain that we shall treat the unprovoked use of poison gas against our Russian ally exactly as if it were used against ourselves and if we are satisfied that this new outrage has been committed by Hitler, we will use our great and growing air superiority in the West to carry gas warfare on the largest possible scale far and wide against military objectives in Germany.18

Why did it take until May 1942 for such a threat to be issued? Why did the Germans not use poison gas earlier in the war, for instance, against Poland in 1939, Belgium, the Netherlands, and Norway in 1940, or Greece and Yugoslavia in 1941? Given the virtually nonexistent prospect of retaliatory use of poison gas by these countries, deterrence seems an unlikely explanation, nor did Britain proclaim it would extend its chemical weapons deterrent to them. There are at least three plausible explanations:

1. There was no military advantage to using these weapons;

2. Adolf Hitler and German generals who experienced gas warfare in World War I had bad memories of it;

3. The Germans feared alienating international opinion, provoking a chemical weapons arms race, and opening a Pandora’s box of use by the British or another adversary.

The third explanation, self-deterrence, seems at least as plausible as the others, if not more so. Unlike most other weapons that could inflict mass destruction if used against urban targets or simply on an isolated battlefield, chemical weapons were deemed politically sensitive due to the prevailing international norms against their use and, hence, would incur greater reputational damage than the use of conventional explosives. Consequently, instead of delegating use decisions to lower-level commanders, special controls were established to ensure they were not used unless authorized at the highest level.19

Concerns about violating international norms were also evident with the advent of the atomic bomb and the political controls on its use. Although it was well understood prior to Hiroshima that the use of such a powerful weapon required high-level approval within the US government—and ostensibl y the approval of the UK government—the precise arrangements remained somewhat informal.20 Presidential control over decisions to use atomic bombs was only formalized several years later in NSC 30 (US Policy on Atomic Warfare).21 As other countries obtained nuclear capabilities, they, too, instituted political control systems.22 Like other WMD controls, deeming this control necessary reflected a recognition that nuclear weapons generated far more political controversy than nonnuclear weapons and incurred greater reputational risks.

War and the Conduct of War

Analysts are likely to recognize the reasons policymakers avoid wars and place limits on them as important, but they are highly unlikely to examine them, at least explicitly, using a self-deterrence framework. For example, analysts almost certainly underappreciate how much speculation about what an adversary might do—as opposed to actual threats or realistic capabilities— can deter action. Likewise, analysts are unlikely to pay sufficient attention to the fact it is not just adversaries that can impose costs. In practice, many of the costs that arise in policymakers’ discourse when debating the prospect of going to war exist in their heads rather than reflecting an adversary’s threats. These concerns often include allies’ negative reactions, domestic unrest, opportunity costs, jeopardizing a wider set of interests, and fear of creating a quagmire. The following historical anecdotes illustrate the types of issues deserving more analysis.

An examination of Israeli policy making across several Arab-Israeli conflicts during the Cold War highlights how a heavy reliance on external military and diplomatic support ensured Israel would have to restrain its behavior to avoid jeopardizing this support. Avi Kober observed that the “shadow of superpower intervention hovered over all the Arab-Israeli wars, and Israeli leaders often feared that such intervention might undo Israeli achievements on the battlefield.”23 Potential costs Israeli leaders recognized included the United States withholding support.24 As one CIA estimate noted, “A major factor tending to inhibit Israeli aggressiveness is that Tel Aviv presumably anticipates that an Israeli-initiated war would seriously damage relations with the US and jeopardize the flow of US military and economic aid.”25 As a practical matter, this fear translated into Israel feeling obliged to restrict the scope and intensity of its battlefield activities to ensure they were not counterproductive to the country’s longer-term interests and a military-planning assumption that its freedom of action to conduct a military campaign was limited to a handful of days.26

In the late 1970s and early 1980s, US policymakers were concerned about the possibility of a Soviet invasion of Iran and tried to deter it. Their actions led to the development of the Carter Doctrine and the formation of the Rapid Deployment Joint Task Force. American analysts often framed the issue as one in which the prospect of US military intervention would be the key factor deterring Soviet aggression and advocated strategies to maximize US deterrence in that contingency.27 Still, leaving the potential deterrent effect of any US military intervention aside, there were many local reasons a Soviet invasion of Iran was highly improbable.

In its assessment of this scenario, the CIA identified basic problems that would self-deter the Soviet leadership. These problems did not include battlefield defeats that Iran could inflict on the Soviets. The CIA expected it would take 14 weeks—at the most—for the Russians to overcome strong Iranian resistance; with limited resistance, it would take roughly six weeks.28 Instead, due to Iran’s size, terrain, and large population, the CIA found the invasion requirements to be on a scale far surpassing “anything the Soviets have attempted since World War II” and estimated the Soviets would “need to commit a large occupation force, probably 300,000 to 500,000 men, to contend with an anticipated Iranian guerilla movement.”29 Analysts also assessed the action would generate adverse political consequences outside the region and assumed the Soviets’ recent experience of getting bogged down in Afghanistan was “likely to make them extremely chary of sending troops into either Iran or Pakistan and thus multiplying their problems with guerrilla resistance.”30 Despite an awareness that local problems would make a Soviet invasion of Iran a highly unattractive endeavor, US policymakers and analysts did not address these factors in policy debates. Instead, they overwhelmingly discussed the prospect of a Soviet invasion of Iran in terms of American deterrence rather than Soviet self-deterrence.31

Self-deterrence also significantly influenced American political and military leaders when they were contemplating an invasion of North Vietnam in the late 1960s. Importantly, little evidence exists that US leaders feared a military defeat by the North Vietnamese. Moreover, to limit the possibility of a Chinese intervention—an assumed threat extrapolated from the Korean War experience rather than a formal threat issued by the Chinese government—US military planners specifically designed campaign plans of limited duration and geographic scope. In this sense, the campaign design already included self-deterrence. These self-imposed limits had a knock-on effect. In terms of the utility of an invasion, US leaders saw little point initiating a large military campaign in which only limited North Vietnamese territory was seized prior to the United States withdrawing from it shortly thereafter. In 1967, when US military planners proposed an invasion of North Vietnam with combined airborne, amphibious, and ground forces consisting of six divisions, the costs involved in generating sufficient forces and providing logistical support far outweighed any conceivable strategic benefit.32

The Kennedy administration’s assumptions about the worldwide consequences if it attacked Cuba in October 1962—especially its fear that the Soviets would retaliate by seizing West Berlin— reflect the shortcomings of deterrence relative to self-deterrence in explaining military inaction. The following sequence illustrates the basic problem US policymakers faced: the leadership of A assumes that if it attacks B, then B’s ally C will attack A’s ally D. Senior policymakers repeatedly invoked the analogy of the 1956 Soviet intervention in Hungary (being facilitated by the Anglo-French invasion of Egypt) despite its questionable historical accuracy. As President John F. Kennedy warned: “If we attack Cuba . . . then it gives them a clear line to take Berlin, as they were able to do in Hungary under the Anglo war in Egypt. . . . We would be regarded as the trigger-happy Americans who lost Berlin. We would have no support among our allies.”33 Secretary of State Dean Rusk similarly noted the United States was “ . . . so intimately involved with 42 allies and confrontation in so many places that any action that we take will greatly increase the risks of a direct action involving our other alliances and our other forces in other parts of the world.”34 Chairman of the Joint Chiefs of Staff General Maxwell Davenport Taylor concurred: “The worldwide problem has certainly been before us, Mr. President. . . . It may have been a deterrent to my enthusiasm for an invasion of Cuba.”35

The fears Kennedy, Rusk, and Taylor expressed can hardly be classed as successful deterrence by Cuba and the Soviet Union. To deter a US invasion of Cuba, the Soviets issued a public statement on September 11, 1962, that contained a vague threat of nuclear retaliation and no mention of attacking West Berlin.36 Although the Soviets had the military capabilities to seize West Berlin, they also had other available retaliatory options in Europe and elsewhere, including nuclear use of some kind. Kennedy’s focus on West Berlin was entirely speculative, rather than based on an explicit Soviet threat. As such, assumptions about actions the Soviets might conceivably take based on a prominent historical analogy or pure guesswork, and concern about the damage to America’s relationships with allies, should more appropriately be understood in terms of self-deterrence.

The Cuban missile crisis also highlights the perceived reputational consequences of taking military actions widely deemed to be unethical despite being militarily prudent. Several key US policymakers were concerned about conducting a sneak attack on Cuba, an action analogized to Pearl Harbor “in reverse.”37 For instance, Under Secretary of State George Ball warned that by attacking Cuba without first issuing a warning, the United States would “alienate a great part of the civilized world.”38 It was largely due to this negative prospect that Kennedy rejected the option of an air strike on Cuba without a warning.39 Indeed, America’s leadership chose the blockade option, despite its questionable legality, due to concerns about the detrimental impact on US allies and the propaganda defeat Washington would suffer if it took military action contravening international law.40

Military Intervention and Covert Action

Decisions on whether to undertake military interventions or rely on covert action often involve complex calculations about whether to use force (as opposed to some nonviolent policy tool), what level of force to use, and the problems created by using a state’s armed forces. In most cases involving powerful states acting against weaker states, there is little fear of military defeat or the inability to achieve an immediate policy objective. Instead, the principal risks are quite different.

During the 1980–81 Polish crisis, the CIA assessed the prospects of a Soviet military invasion of Poland and identified challenges the Soviets would face. As with its militarily unopposed 1968 intervention in Czechoslovakia, the Soviets did not expect significant resistance from the Polish armed forces. Instead, Moscow assumed that because of “widespread Polish resistance by civilians and possibly by some military forces,” the Soviets would still require a large force to overcome immediate resistance and to engage in occupation and policing tasks over the longer term.41 There also seemed little chance for the Soviets to establish a “viable indigenous vassal regime.”42 Consequently, Moscow would be obliged to administer Poland directly and indefinitely. In these circumstances, Poland’s economic productivity would drop, forcing the Soviets to provide subsidies. The Soviets would also have to anticipate a global propaganda defeat. Having recently invaded Afghanistan, this action would reinforce the world’s view of the Soviets as aggressors, undermining Soviet efforts to weaken NATO.43 In short, the Soviets would have to contend with a wide range of longer-term, nonmilitary, and international consequences resulting from a successful intervention.

Concerning Cuba, US leaders shared many of the Soviet misgivings about military interventions in Eastern Europe. The Bay of Pigs in April 1961 was conducted as a covert operation due to a deliberate rejection of an overt military intervention. American policymakers offered numerous reasons why a military intervention would be detrimental to US interests, none of which involved fears of the Cuban military inflicting substantial casualties.44

Admiral Arleigh Burke, Chief of Naval Operations, drew attention to the legal straitjacket preventing military action. He noted that America’s treaty commitments obliged it to refrain from military interventions unless sanctioned by other Latin American countries, meaning that if the United States wanted to overthrow Fidel Castro, it had three options: “(1) multilateral intervention through the OAS (Organization of American States), (2) unilateral overt intervention, and (3) unilateral covert action.”45 The problem with the first option was the United States was unlikely to receive support from other Latin American countries. Moreover, even if the OAS sanctioned it, the matter would still likely require UN Security Council approval, but this was impossible due to a Soviet veto. A unilateral overt intervention was expected to be disastrous for the US global image. By not going through the OAS, the United States would face charges of aggression at the UN, alienate its regional allies, and demonstrate disregard for international law. As for unilateral covert action, despite “no certainty of success,” its key advantage was that the United States would “not be charged with aggression.”46

When President Kennedy came to office, many political assumptions about the disadvantages of a military operation carried over from the previous administration. Concerns were also expressed about the risks of a covert operation. Arthur Schlesinger Jr. articulated these risks most cogently, warning that a covert operation would undermine the “reawakening world faith in America,” specifically the new administration’s image. Although officially deniable, the United States would still “be held accountable for it before the bar of world opinion.” For instance, Schlesinger argued that the Communists would score a massive propaganda victory in the third world by portraying the Kennedy administration “as a gang of capitalist imperialists maddened by the loss of profits and driven to aggression and war.” They would also portray Castro as “the defender of the colored races against white imperialism.” Other consequences would likely include attacks on American embassies and isolation of the United States at the United Nations. Furthermore, should the covert operation appear to be failing, Schlesinger warned that domestic pressure would build for an overt military intervention, which would have “almost irreparable” international political consequences equivalent to an “American Hungary.”47

After the Bay of Pigs invasion floundered and Kennedy refused to intervene overtly with the US military, the administration’s Cuba policy remained torn between conducting additional covert operations and preparing for an invasion. Prior to the October 1962 Cuban missile crisis, US officials continued to debate the merits and drawbacks of these options. As before, policymakers raised the same arguments, and the option of military intervention remained unattractive.48 The same was also true during the missile crisis when the option of a full-scale invasion (OPLAN 316-62) was debated again and rejected. One of the great ironies of Nikita Khrushchev’s decision to send nuclear weapons to Cuba—ostensibly to deter an American invasion— is that it was probably unnecessary. Had the Soviets appreciated the extent to which Kennedy feared a seizure of West Berlin in retaliation for an invasion of Cuba, threatening this latter option might have sufficed as a deterrent.


Are policymakers deterred from taking military action, not by fear of what an adversary threatens, but due to a range of consequences that do not fit neatly within traditional understandings of deterrence? As this article shows, self-deterrence is often a crucial factor. On the other hand, the spectrum of explanations for military inaction extends beyond deterrence and self-deterrence. There are many reasons why states are not constantly at war with one another and why small wars do not expand into larger wars. Nevertheless, the examples provided here highlight that self-deterrence frequently affects major policy decisions about war and its conduct, military interventions, and covert action and is therefore a concept worthy of future scholarly investigation and practitioner interest. American officials responsible for crafting deterrence strategies and communicating warning messages will be more successful if they avoid restricting their focus to consequences that can be directly threatened with the means at hand, particularly the military means.

Given the numerous deterrence failures traced to narrow assumptions about an adversary’s cost-benefit calculus, policymakers, strategists, and analysts tasked with warning about and preparing for future wars should utilize self-deterrence to account for and exploit a broader range of adversary fears than they are traditionally used to doing. As the objective of a deterrence strategy is to produce a deterrent effect upon an adversary, an appreciation of the full range of consequences the adversary fears can offer a greater number of opportunities to achieve that goal. When assessing the prospect of China, Iran, North Korea, Russia, or another adversary waging war in the future, analysts should devote more attention to the self-imposed restrictions these states place on using military force. It is also crucial to be aware that attempting to manipulate or take advantage of an adversary’s fears is something the adversary will almost certainly be doing as well. It would be extremely shortsighted to assume that the adversaries who are intent on deterring the United States are not seeking to identify and exploit self-imposed limitations on American power.


Acknowledgments: The research presented here is largely drawn from a report prepared for the Office of Net Assessment, Office of the Secretary of Defense.


Jeffrey H. Michaels

Dr. Jeffrey H. Michaels is the Institut d’Estudis Nord-americans (IEN) Senior Fellow in American Foreign Policy and International Security at the Institut Barcelona d’Estudis Internacionals (IBEI). His earlier experience includes working as a senior lecturer in defense studies at King’s College London, as well as serving as an official with NATO and the US Department of Defense. He holds visiting fellowships with the Department of War Studies at King’s College, the Oxford University Changing Character of War Centre, and was recently the Senior Eisenhower Defense Fellow at the NATO Defense College, Rome. He is the coauthor, with Sir Lawrence Freedman, of The Evolution of Nuclear Strategy: New, Updated and Completely Revised (4th edition, Palgrave Macmillan, 2019).




  1. See, for instance, Michael Howard, ed., Restraints on War: Studies in the Limitation of Armed Conflict (Oxford, UK: Oxford University Press, 1979); and Barry Buzan, “Internal Restraints on the Use of Force,” in The Use of Force in International Relations, ed. F. S. Northedge (London: Faber and Faber, 1974), 166–93. Return to text.
  2. Nina Tannenwald, “The Nuclear Taboo: The United States and the Normative Basis of Nuclear Non-Use,” International Organization 53, no. 3 (Summer 1999): 433–68, https://www.jstor.org/stable/2601286; and T. V. Paul, The Tradition of Non-Use of Nuclear Weapons (Redwood City, CA: Stanford University Press, 2009). Return to text.
  3. For example, see Herman Kahn, “Issues of Thermonuclear War Termination,” Annals of the American Academy of Political and Social Science 392, no. 1 (November 1970): 133–72, https://doi.org/10.1177/000271627039200113; Colin S. Gray, “Theater Nuclear Weapons: Doctrines and Postures,” World Politics 28, no. 2 (January 1976): 310, https://doi.org/10.2307/2009894; and Colin S. Gray, “Nuclear Strategy: The Case for a Theory of Victory,” International Security 4, no. 1 (Summer 1979): 54–87, https://doi.org/10.2307/2626784. Return to text.
  4. Robert Jervis, “Deterrence and Perception,” International Security 7, no. 3 (Winter 1982–83): 14–19, https://doi.org/10.2307/2538549. Return to text.
  5. Richard Ned Lebow, “Windows of Opportunity: Do States Jump through Them?,” International Security 9, no. 1 (Summer 1984): 181, https://doi.org/10.2307/2538638. Return to text.
  6. Lebow, “Windows of Opportunity.” Return to text.
  7. Lawrence Freedman, Deterrence (Cambridge, UK: Polity Press, 2004), 30. Return to text.
  8. Freedman contrasts self-deterrence with “the sort that captures the most attention from policymakers and academics, when a determined effort is made to dissuade another party from taking action you judge harmful to your interests,” Freedman, Deterrence, 31. Return to text.
  9. Freedman, Deterrence, 31. Return to text.
  10. Scott D. Sagan, “Realist Perspectives on Ethical Norms and Weapons of Mass Destruction,” in Ethics and Weapons of Mass Destruction: Religious and Secular Perspectives, ed. Sohail H. Hashmi and Steven P. Lee (Cambridge, UK: Cambridge University Press, 2004), 80–81. Return to text.
  11. T. V. Paul, “Self-Deterrence: Nuclear Weapons and the Enduring Credibility Challenge,” International Journal 71, no. 1 (March 2016): 20, https://doi.org/10.1177/0020702015618682. Return to text.
  12. Alexander L. George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974), 11. Return to text.
  13. Yuen Foong Khong, Analogies at War: Korea, Munich, Dien Bien Phu, and the Vietnam Decisions of 1965 (Princeton, NJ: Princeton University Press, 1992). Return to text.
  14. Vesna Danilovic, “Conceptual and Selection Bias Issues in Deterrence,” Journal of Conflict Resolution 45, no. 1 (February 2001): 102, https://doi.org/10.1177/0022002701045001005. Return to text.
  15. For instance, Paul Huth and Bruce Russett, “Testing Deterrence Theory: Rigor Makes a Difference,” World Politics 42, no. 4 (July 1990): 466–501, https://doi.org/10.2307/2010511. Return to text.
  16. See Article 23 in the “Convention with Respect to the Laws and Customs of War on Land (Hague, II) (29 July 1899)”: “Besides the prohibitions provided by special Conventions, it is especially prohibited: To employ poison or poisoned arms.” Department of State (DoS), Treaties and Other International Agreements of the United States of America 1776–1949, vol. 1, Multilateral 1776–1917, DoS publication no. 8407 (Washington, DC: US Government Printing Office [GPO], 1968), as cited in “Convention with Respect to the Laws and Customs of War on Land (Hague, II) (29 July 1899) [Translation],” 32 Stat. 1803, Treaty Series 403, Avalon Project, Yale Law School (website), https://avalon.law.yale.edu/19th_century/hague02.asp#art23. Article 23 reflects a similar reference in the Declaration of the 1874 Brussels Conference: “Art. 13. According to this principle are especially ‘forbidden’: (a) Employment of poison or poisoned weapons.” See D. Schindler and J. Toma, The Laws of Armed Conflicts (Boston: Martinus Nijhoff Publishers, 1988), 22–34, as cited in “Article 13,” Project of an International Declaration Concerning the Laws and Customs of War. Brussels, 27 August 1874” (historical treaty), International Humanitarian Law Databases (website), https://ihl-databases.icrc.org/en/ihl-treaties/brussels-decl-1874/article-13?activeTab=undefined. Return to text.
  17. For instance, in the German debate about unrestricted submarine warfare, concerns about American entry into the war were initially a major consideration, albeit by the end of 1916 the belief existed that this previously unthinkable scenario could be tolerated. See Dirk Steffen, “Document of Note: The Holtzendorff Memorandum of 22 December 1916 and Germany’s Declaration of Unrestricted U-Boat Warfare,” Journal of Military History 68, no.1 (January 2004): 215–24, https://doi.org/10.1353/jmh.2003.0412. Return to text.
  18. As cited in Frank A. von Hippel, “ ‘This Odious Weapon’: Winston Churchill and Chemical Warfare,” Finest Hour 195, no. 1 (October 2022): 26ff, https://winstonchurchill.org/publications/finest-hour/finest-hour-195/this-odious-weaponwinston-churchill-and-chemical-warfare/. Return to text.
  19. John Ellis van Courtland Moon, “Chemical Weapons and Deterrence: The World War II Experience,” International Security 8, no. 4 (Spring 1984): 12–13, https://doi.org/10.2307/2538560. This important study of the topic mostly skips over the pre-May 1942 period, especially the German decision calculus. Return to text.
  20. The August 1943 Quebec Agreement stipulated that neither the United States nor the United Kingdom would use the bomb “against third parties without each other’s consent.” See Franklin D. Roosevelt and Winston S. Churchill, “Quebec Agreement,” August 19, 1943, atomicarchive.com (website), https://www.atomicarchive.com/resources/documents/manhattan-project/quebec-agreement.html; and Alex Wellerstein, “The President and the Bomb,” Restricted Data: The Nuclear Secrecy Blog (website), November 18, 2016, Meditations, https://blog.nuclearsecrecy.com/2016/11/18/the-president-and-the-bomb. Return to text.
  21. The document states: “The decision as to the employment of atomic weapons in the event of war is to be made by the Chief Executive when he considers such decision to be required.” As cited in: Staff Study Prepared by Representatives of the Special Committee of the National Security Council on Atomic Energy (Washington, DC, June 11, 1952), document no. 37, in Foreign Relations of the United States (FRUS), 1952–1954, vol. 2, National Security Affairs, part 2, ed. John P. Glennon (Washington, DC: GPO, 1991), https://history.state.gov/historicaldocuments/frus1952-54v02p2/d37. Return to text.
  22. Jeffrey G. Lewis and Bruno Tertrais, The Finger on the Button: The Authority to Use Nuclear Weapons in Nuclear-Armed States, CNS Occasional Paper no. 45 (Monterey, CA: Middlebury Institute of International Studies at Monterey/James Martin Center for Nonproliferation Studies, February 2019), https://nonproliferation.org/wp-content/uploads/2019/02/Finger-on-the-Nuclear-Button.pdf. Return to text.
  23. Avi Kober, “Great-Power Involvement and Israeli Battlefield Success in the Arab-Israeli Wars, 1948–1982,” Journal of Cold War Studies 8, no. 1 (Winter 2005): 20, https://doi.org/10.1162/152039706775212058. Return to text.
  24. Yaacov Bar-Siman-Tov, The Israeli-Egyptian War of Attrition, 1969 to 1970: A Case-Study of Limited Local War (New York: Columbia University Press, 1980). Return to text.
  25. CIA, memorandum, “Factors Affecting Arab and Israeli Politico-Military Decisions,” May 12, 1975, https://www.cia.gov/readingroom/docs/CIA-RDP79T00865A002600020001-2.pdf. Return to text.
  26. Kober, “Great-Power Involvement,” 2006. Return to text.
  27. For example, see Joshua M. Epstein, “Soviet Vulnerabilities in Iran and the RDF Deterrent,” International Security 6, no. 2 (Fall 1981): 126–58, https://doi.org/10.2307/2538649; and Kenneth N. Waltz, “A Strategy for the Rapid Deployment Force,” International Security 5, no. 4 (Spring 1981): 49–73, https://doi.org/10.2307/2538713. Return to text.
  28. Director of Central Intelligence, “Prospects for Iran: Interagency Intelligence Memorandum,” CIA, July 29, 1982, https://www.cia.gov/readingroom/docs/CIA-RDP85T00176R001500190014-6.pdf. Return to text.
  29. CIA, “Prospects for Iranian Resistance to a Soviet Invasion: An Intelligence Assessment,” CIA, December 1986, https://www.cia.gov/readingroom/docs/CIA-RDP88T00096R000400440003-8.pdf. Return to text.
  30. 30. Directorate of Intelligence, “Soviet Military Forces opposite Iran and in Afghanistan,” CIA, November 18, 1986, https://www.cia.gov/readingroom/docs/CIA-RDP86T01017R000505390001-8.pdf. Return to text.
  31. Despite acknowledging a wider set of inhibiting factors, the CIA still stressed the American deterrent role in Soviet calculations. As one estimate noted, “The scope and nature of their actions would depend in large part on their assessment of the likely US response.... If Moscow judged the risk of a direct confrontation with the United States to be high . . . it probably would be deterred from undertaking massive military action to seize the oilfields in Khuzestan.” National Foreign Assessment Center, “Soviet Response to an Iranian-Iraqi Clash: An Intelligence Memorandum,” CIA, April 1980, https://www.cia.gov/readingroom/docs/CIA-RDP81B00401R000500050007-8.pdf. Return to text.
  32. The most in-depth account of US invasion plans can be found in: Mark A. Viney, Determined to Persist: General Earle Wheeler, the Joint Chiefs of Staff, and the Military’s Foiled Pursuit of Victory in Vietnam, vol. 2 (BookBaby, 2022); and Office of National Estimates, memorandum, “Communist Reactions to US Operations in the DMZ,” CIA, December 22, 1967, https://www.cia.gov/readingroom/docs/CIA-RDP79R00904A001400010026-0.pdf. Return to text.
  33. John F. Kennedy, as cited in Ernest R. May and Philip D. Zelikow, eds., The Kennedy Tapes: Inside the White House during the Cuban Missile Crisis, concise ed. (New York: W. W. Norton, 2002), 111. Return to text.
  34. May and Zelikow, Kennedy Tapes, 36. Return to text.
  35. As cited in Sheldon M. Stern, “Source Material: The 1997 Published Transcripts of the JFK Cuban Missile Crisis Tapes: Too Good to Be True?,” Presidential Studies Quarterly 30, no. 3 (September 2000): 590, https://doi.org/10.1111/j.0360-4918.2000.00131.x. Return to text.
  36. For text, see “Soviet Statement on U.S. Provocations,” CIA, September 11, 1962, https://www.cia.gov/readingroom/document/cia-rdp79t00428a000200010014-0. Return to text.
  37. Dominic Tierney, “ ‘Pearl Harbor in Reverse’: Moral Analogies in the Cuban Missile Crisis,” Journal of Cold War Studies 9, no. 3 (Summer 2007): 49–77, https://doi.org/10.1162/jcws.2007.9.3.49. Return to text.
  38. George Ball, as cited in Tierney, “Pearl Harbor,” 63. Return to text.
  39. Tierney, “Pearl Harbor.” Return to text.
  40. Jonathan Colman, “Toward ‘World Support’ and ‘The Ultimate Judgment of History’ :The U.S. Legal Case for the Blockade of Cuba during the Missile Crisis, October-November 1962,” Journal of Cold War Studies 21, no. 2 (2019): 150–73, https://doi.org/10.1162/jcws_a_00879. Return to text.
  41. Emphasis mine. CIA, “Implications of a Soviet Invasion of Poland,” July 9, 1981, https://www.cia.gov/readingroom/docs/CIA-RDP84B00049R000601580007-0.pdf. Return to text.
  42. CIA, “Soviet Invasion.” Return to text.
  43. CIA, “Soviet Invasion.” Return to text.
  44. Michael Poznansky, In the Shadow of International Law: Secrecy and Regime Change in the Postwar World (Oxford, UK: Oxford University Press, 2020), 78–110. Return to text.
  45. Arleigh Burke (Chief of Naval Operations) to Livingston Tallmadge Merchant (Under Secretary of State for Political Affairs), letter, Washington, DC, February 26, 1960, document no. 466, in FRUS, 1958–1960, vol. 6, Cuba, ed. John P. Glennon (Washington, DC: GPO 1991), https://history.state.gov/historicaldocuments/frus1958-60v06/d466. Return to text.
  46. Burke to Merchant. Return to text.
  47. Arthur M. Schlesinger (President’s Special Assistant) to John F. Kennedy, memorandum, Washington, DC, April 10, 1961, document no. 86, in FRUS, 1961–1963, vol. 10, Cuba, January 1961–September 1962, ed. Louis J. Smith (Washington, DC: GPO, 1997), https://history.state.gov/historicaldocuments/frus1961-63v10/d86. Return to text.
  48. See, for instance, Inter-Agency Staff Study, n.d., document no. 274, in FRUS, 1961–1963, vol. 10, https://history.state.gov/historicaldocuments/frus1961-63v10/d274; and Sherman Kent, memorandum for the Director, “Probable Reactions to a US Military Intervention in Cuba,” CIA, April 10, 1962, https://www.cia.gov/readingroom/docs/CIA-RDP79R00904A000800020022-0.pdf. Return to text.

Disclaimer: Articles, reviews and replies, and book reviews published in Parameters are unofficial expressions of opinion. The views and opinions expressed in Parameters are those of the authors and are not necessarily those of the Department of Defense, the Department of the Army, the US Army War College, or any other agency of the US government. The appearance of external hyperlinks does not constitute endorsement by the Department of Defense of the linked websites or the information, products, or services contained therein. The Department of Defense does not exercise any editorial, security, or other control over the information you may find at these locations.