Threading the Needle: The Logic of Conventional Coercion in Nuclear Crises

US military and political leaders are once again worried about the potential for crises and conflicts with nuclear rivals. That concern has led to much recent literature examining the factors that lead to coercive success in nuclear crises, most of it focused on the role of nuclear superiority and nuclear threats.1 However, policymakers also pay close attention to the coercive potential of conventional military capabilities in nuclear crises. For instance, Elbridge Colby, current undersecretary of defense for policy, wrote in 2021 that China could use local conventional superiority in the area around Taiwan to seize the island and present Taiwan’s “allies and partners with new facts on the ground that, Beijing would reckon, they would consider some combination of too difficult, costly, and risky to reverse.”2 Partly out of that fear, Secretary of Defense Pete Hegseth has stated that he wants the United States and its allies to strengthen conventional warfighting capabilities in East Asia to prepare to “reestablish deterrence” and “fight and win, decisively” should a crisis erupt with China over Taiwan.3

Both Colby’s and Hegseth’s comments imply that nuclear powers can use their conventional military forces to coerce nuclear rivals. Nuclear powers have assumed the same, and have often issued conventional military threats to achieve coercive success in nuclear crises. In the Berlin Crises, Soviet Premier Nikita Khrushchev warned John McCloy, then chairman of the Council on Foreign Relations, that “if you deploy one division in Germany, we will respond with two divisions, if you declare mobilization, we will do the same. If you mobilize such and such numbers, we will put out 150–200 divisions, as many as necessary.”4 In the Cuban Missile Crisis, the Kennedy administration implemented a naval quarantine of Cuba and threatened an airstrike plus an invasion to take out the Soviets’ ballistic missile sites. During the Kargil War and Operation Parakram (2001–2002), India relied primarily on the threat and use of conventional forces to coerce Pakistan. Did any of these attempts at conventional coercion succeed? If so, why and how?

The utility of conventional coercive threats is a long-standing question in the nuclear politics literature. Some scholars argue that nuclear powers can use conventional force to achieve limited objectives, because their nuclear weapons make it too costly or risky for the adversary to respond to the defeat.5 Scholars aligned with the theory of the nuclear revolution, however, argue that threats of inflicting a conventional defeat should have little coercive utility.6 Escalation is hard to control and nuclear war could occur despite leaders’ best efforts to avoid it.7 Because of that, conventional capabilities coerce adversaries in a nuclear crisis by acting as “tripwires” or making “threats that leave something to chance,” which can manipulate the risk of uncontrolled nuclear escalation.8

I intervene in this debate by developing what I call “conventional options theory,” which identifies when and how threats of achieving limited conventional success will coerce nuclear-armed adversaries in a crisis. I argue that those threats succeed if the challenger meets two conditions. First, the challenger must possess an effective limited conventional option, or a conventional option that can achieve operational success while staying within desired escalation thresholds. Such options help a challenger manage the power-risk tradeoff: the dilemma that actions made to impose pressure on a target also increase the risk that the target escalates.9 Effective limited conventional options keep the risk of escalation within tolerable levels for the challenger, and they also impose a choice on the target between escalating or making concessions.

“Conventional options theory shows that both sides of the debate over the utility of conventional coercion have merit.”

Second, to encourage the target to choose the latter, the challenger must possess a strong nuclear retaliatory capability, also known as a “second-strike capability.” This capability makes escalation in response to a limited conventional defeat less rational for the target and makes any threats of nuclear escalation in response to a limited conventional defeat less credible. If the challenger has an effective limited conventional option and a strong second-strike capability, then the target chooses to back down rather than escalate to some option for the use of force that it finds too costly or risky.

Conventional options theory shows that both sides of the debate over the utility of conventional coercion have merit. There are conditions under which nuclear powers can use the threat of achieving limited conventional success to coerce nuclear rivals. Conversely, conventional coercion requires a challenger to thread a fine needle. That state needs to achieve victory while also restraining itself, but staying within escalation thresholds often acts as a constraint on military forces that makes achieving operational success more difficult. Potential targets can also adopt “asymmetric escalation” nuclear postures that increase the risk of escalation inherent in a conventional operation and impose more stringent thresholds on challengers, who nonetheless still try to keep conventional conflict limited. The opportunities to engage in successful conventional coercion should thus be rare.

Conventional options theory also shows that nuclear superiority should play a circumscribed role in nuclear crisis bargaining. A bargaining advantage is conferred to targets of coercion that are conventionally weak and have a high degree of nuclear superiority over the challenger. Even then, the effect of this superiority operates through a different mechanism than previous scholarship proposes.10 Rather than being emboldened by nuclear superiority, states with nuclear superiority often issue escalatory threats as a last resort. The adversary’s feeling of nuclear inferiority is then what makes these escalatory threats credible.

The Debate over Conventional Coercion in a Nuclear World

One major question in the field of nuclear politics is whether nuclear powers can use conventional military success, or the threat of it, to coerce a nuclear-armed rival. “Conventional success” encompasses operations using non-nuclear forces that defeat an adversary’s non-nuclear forces, seize territory, or achieve some other battlefield objective. Scholars usually regard such warfighting operations as following the logic of brute force—they directly attain some objective or deny the adversary the ability to achieve its objective. These operations, however, can also be used to coerce an adversary by convincing it that its military strategy for achieving victory will not succeed, and therefore it should concede to the state’s demands. Robert Pape’s classic work on coercion calls this strategy “coercion by denial.”11

In a nuclear context, conventional success—or the threat of it—coerces a target through a different logic. The challenger achieves some conventional success that is limited in geographic scope or in the number of the target’s forces that it defeats. To respond to that successful limited operation or campaign, the target would need to escalate, either vertically (by putting more forces into the fight) or horizontally (by starting a crisis or initiating operations in another theater).12 The challenger’s nuclear weapons, however, make such escalation carry a risk of nuclear war. The target is unwilling to run that risk, and it chooses to back down and concede to the challenger’s demands. This outcome is how nuclear weapons can provide a “shield” that enables nuclear powers to engage in faits accompli or “salami-slice” tactics with their conventional forces.13

The logic of this coercive strategy relies on the stability-instability paradox. This paradox argues that if both sides have a secure second-strike capability, “and both know it, they will be less inhibited about initiating conventional war . . . than if the strategic balance were unstable.”14 The mutual possession of nuclear weapons makes nuclear escalation between both countries irrational. Both sides should recognize that, and this situation puts a notional cap on escalation that makes engaging in conventional conflict (hypothetically) safer. Under this logic, threats of achieving limited conventional success should be useful coercive tools because the mutual possession of nuclear weapons makes the losing party reluctant to go to higher rungs on the escalation ladder to reverse an early conventional defeat.15

Scholars aligned with the school of thought known as the theory of the nuclear revolution argue that this coercive strategy, while logically sound, does not bear out in practice, and that as a result, successful conventional warfighting operations have little coercive value. Nuclear escalation is hard to control, and nuclear war can occur despite leaders’ best efforts to prevent it, which means that “the ability to win a conventional war is not synonymous with the ability to keep the war conventional.”16 One may expect to achieve a conventional victory, but the fighting is unlikely to end there. If that is true, as Kenneth Waltz points out, “fighting beyond the trip wire level serves no useful purpose,” and inflicting a conventional defeat should not generate much coercive leverage.17

Instead of coercing an adversary by threatening to achieve a limited conventional success, conventional forces coerce in a nuclear crisis by manipulating nuclear risk and making “threats that leave something to chance.”18 These threats use the possibility of engaging in a conventional escalatory action to raise uncertainty around the escalation process, and this serves to make credible otherwise noncredible nuclear threats.19 The classic example of such a threat, from the work of Thomas Schelling, is positioning conventional forces as a “tripwire” or a “plate-glass window,” and attacks on those forces obligate the state to react violently.20 Another example is deliberately entangling conventional and nuclear forces, either by employing lots of dual-use systems, co-locating conventional and nuclear weapons platforms, or sharing the same command-and-control (C2) systems.21 States can also use non-nuclear capabilities as “strategic substitutes” that increase the risk of nuclear escalation in a conflict.22

Other conventional coercive strategies are available as well. Nuclear states could use their conventional capabilities to inflict punishment on an adversary’s economy or civilian infrastructure. States could also engage in “conventional counterforce,” using conventional weapons to attack an adversary’s nuclear forces.23 While conventional coercive strategies utilizing such operations are possible, most of the debate on conventional coercion in nuclear crises focuses on the utility of achieving limited conventional success.

Turning Conventional Success into Political Success

I developed conventional options theory to answer the question at the heart of this article. The theory predicts that threats of achieving limited conventional success will only coerce nuclear adversaries under the aforementioned two conditions. First, the challenger must possess conventional options that can achieve operational objectives while respecting desired thresholds on escalation; these are effective limited conventional options, or “Goldilocks options” for the use of force. Second, the challenger must possess a strong second-strike capability vis-à-vis the target. If both these conditions are met, the target faces a decision between backing down or escalating to some action that it finds too costly or risky, and it will choose the former. The decision tree in figure 1 illustrates the logic of the theory.

Figure 1. The logic of conventional options theory. Figure by the author.

Condition 1: Challenger’s Effective Limited Conventional Options

Conventional options theory starts with a focus on the challenger who sends a coercive demand to a target, and this demand is backed by the threat of achieving a limited conventional success. Coercive demands aim to convince the target to change a policy, accept a change to the status quo, or to stop doing something it is currently doing. Other scholars use the term “compellence.”24 This aim is separate from deterrence, which seeks to dissuade an adversary from taking some unwanted action.25 In this scenario, both the challenger and the target possess nuclear weapons and some capacity to launch them. The challenger and the target could be in dispute over some direct issue like an unresolved border, or over the fate of a non-nuclear third party. In the latter case, I focus only on how the challenger’s coercive actions affect the nuclear target’s willingness to defend or assist the non-nuclear third party.

When sending a coercive threat, the challenger faces a dilemma. It wants to put pressure on the target to concede to its demands, but it also wants to avoid a nuclear war. The problem is that actions to maximize one goal could jeopardize the other. Taking steps to reduce the risk of nuclear war could ease coercive pressure on the target. Actions meant to convince an adversary to back down could increase the risk of a nuclear conflagration. Fiona Cunningham calls this the “limited war dilemma.”26 Robert Powell calls it the “power-risk tradeoff.”27

Effective limited conventional options help the challenger address this dilemma. By limiting the use of conventional force, these options reduce the risk of nuclear escalation to tolerable levels. Conventional options are limited to the extent that they respect certain escalation thresholds. These thresholds could be placed with respect to geography, targeting, scope, or time. Geographic thresholds refer to efforts to keep a conflict within certain physical boundaries, or what Herman Kahn called an “agreed area of battle.”28 Targeting thresholds refer to restrictions on attacking certain adversary platforms, bases, or C2 infrastructures. For instance, a challenger may wish to refrain from attacking a target’s submarine bases or its mobile missiles because the target relies on those assets to have a survivable nuclear arsenal.29 Scope thresholds refer to the number and types of forces to be employed. A larger scope with more forces and more types of platforms employed not only increases the assessed material costs of using force, but also policymakers tend to perceive larger operations as coming with more risk of escalation.30 Time thresholds refer to the time it takes to complete a threatened operation. If a threatened military operation is expected to last a long period of time, there may be more chances for things to go wrong and for emotions to become more deeply engaged, generating a greater risk of nuclear escalation.31

Confining conventional military operations within geographic, targeting, scope, and time thresholds helps the challenger keep the risk of nuclear escalation within tolerable levels for three reasons. First, the limited aspect reduces the risk of escalation via accidental or unauthorized nuclear use. One reason why the risk of nuclear war grows as the intensity of a conflict grows is that a more intense conflict increases the “fog” and “friction” inherent in the use of force.32 There are more chances for misperception or miscommunication to occur. That situation increases the chances that some lower-level commander fires a nuclear weapon without the knowledge or authorization of central leadership, or that a senior leader decides to launch a nuclear strike based on an inaccurate understanding of battlefield conditions.

“One reason why the risk of nuclear war grows as the intensity of a conflict grows is that a more intense conflict increases the 'fog' and 'friction' inherent in the use of force.”

Second, keeping a conventional military operation within certain thresholds communicates limited aims to the target. If a conventional operation is limited in geographic scope, it is less likely that the challenger is engaging in operations that would overthrow the target’s regime or take its territory. That perception reduces the risk that the target’s leaders make a deliberate choice to employ nuclear weapons. Previous literature on coercion notes that challengers often fail to compel targets because they ask for too much or they fail to offer assurances.33 In my theoretical framework, communicating limited aims is important because it makes the target less likely to choose to escalate to the nuclear level.

Third, respecting escalation thresholds also makes targets less likely to take other provocative actions that raise the risk of nuclear war. Targets can respond to a conventional operation that transgresses some geographic or targeting threshold by increasing the number of troops and weapons they are committing to the fight or by initiating conflict in a new theater, instead of nuclear use. These steps, however, also increase the intensity of the conflict and thus the risk of nuclear war occurring at some point in the conflict. Going beyond certain escalation thresholds generates a higher risk of nuclear war even if doing so does not cause the target to immediately employ nuclear weapons.

To constitute a useful coercive tool, limited conventional options also need to be effective. By effective, I mean that the use of force achieves the challenger’s operational objectives. The prospect of conventional defeat is necessary to impose a choice on the target between conceding to the challenger or escalating to avoid making a concession.34 The target can no longer merely “stand firm” and resist the challenger’s demands. If the threatened conventional operation was unlikely to result in defeat for the target, then it would not face this choice, and coercion would be less likely to succeed. Effective limited conventional options thus constitute Goldilocks options for the use of force. These actions are strong enough to exert coercive leverage over the target, but they are not so strong that they provoke an escalatory spiral.35

The top half of figure 1 illustrates how possessing an effective limited conventional option affects the challenger’s calculus. For the challenger to issue a coercive threat in the first place, it needs to assume that it has a Goldilocks option: the ability to use conventional force with some probability of success while running a tolerable risk of nuclear escalation. After the threat is issued, there is usually a period in a crisis where the challenger and the target mobilize their conventional military forces.36 During this process, the more that the challenger continues to think that it has an effective limited conventional option, the more likely it is to use force and to continue to press its coercive demands.37 Once the challenger continues applying coercive pressure, the focus of decision-making moves to the target state, and the challenger’s nuclear retaliatory capability comes into play.

Condition 2: The Challenger’s Nuclear Retaliatory Capability

Possessing an effective limited conventional option imposes a choice on the target between either escalating the crisis, or conceding to the challenger’s demands and accepting an objectionable outcome. To avoid the latter, the target could threaten to do the former. As a concept, the power-risk trade-off primarily models the risk of nuclear escalation that results from a crisis going out of control, when one side employs nuclear weapons unintentionally or out of fear for its survival.38 Nuclear states, however, could employ tactical nuclear weapons (or nuclear weapons with lower yields) when their survival or territorial integrity is not at stake simply to avoid making a costly concession, and nuclear powers have often threatened to do this.39 If effective, these threats to employ tactical nuclear weapons by the target could induce a “substrategic paralysis” in the challenger that deters it from using conventional force.40

The challenger possessing a strong second-strike capability makes the target’s escalatory threats less credible and thus less effective. When the challenger has a strong nuclear retaliatory capability, the target has a low probability of destroying enough of the challenger’s nuclear weapons in a first strike to meaningfully limit the damage that the challenger’s nuclear weapons could cause. A weak second-strike capability means that the target has a high probability of conducting a first strike against the challenger’s nuclear forces and keeping the damage inflicted by a retaliatory strike below some meaningful threshold.41 Typically, nuclear scholars use the term “secure” to describe a state’s second-strike capability. Either the state can assure its adversary that it can inflict unacceptable damage in a retaliatory nuclear strike, or it cannot. I use the terms “strong” and “weak” to indicate that the strength of a state’s nuclear retaliatory capability is likely more continuous than the traditional terminology implies. For instance, Riqiang Wu’s analysis of the evolution of China’s second-strike capability reports different probabilities of China’s ability to launch a second-strike of various sizes in different scenarios. A stronger second-strike capability can be thought of as “more secure” and is associated with China having a higher probability of being able to retaliate and inflict unacceptable levels of damage after receiving a US first strike.42

Using tactical nuclear weapons generates a high risk of further escalation to a large-scale nuclear exchange.43 If the challenger has a strong second-strike capability, it can signal that such an exchange is likely to end in unmitigated catastrophe; there is no prospect of keeping damage within some meaningful level. That probability makes a limited nuclear strike in response to experiencing a limited conventional defeat a less rational response for the target, and makes any of the target’s nuclear threats to avoid making concessions less credible. Other scholars have argued that for conventional success to be useful in a nuclear crisis, the challenger needs to communicate that escalation in response to a conventional defeat is irrational and that sending an escalatory threat will be ineffective.44 Possessing a strong second-strike capability makes this task easier.

“One could say that there should be plenty of military options before the target crosses the nuclear threshold.”

One can see this logic at work in discussions of the strategic implications of China’s nuclear buildup. Abraham Denmark and Caitlin Talmadge argue that “Chinese strategists have had to worry that in extremis, the United States might still resort to nuclear threats to defend its interests.”45 By expanding its nuclear arsenal, China is getting into a “deeper condition of stalemate” and gaining a more secure second-strike capability vis-à-vis the United States. That development makes nuclear escalation in response to a conventional attack on Taiwan much less rational. As a result, the United States will become less likely to issue a nuclear escalatory threat if China attacked Taiwan, and even if it did, such a threat would be less believable. Thus, by making itself less vulnerable to US nuclear threats, China (the potential challenger) is enabling itself to use its regional conventional superiority in East Asia to coerce the United States (the potential target) to accept a seizure of Taiwan.

One could say that there should be plenty of military options before the target crosses the nuclear threshold. This is true, but the logic of possessing a strong nuclear retaliatory capability for the challenger still holds. Escalating to a more intense conventional option still increases the risk of nuclear war, even if to a lesser extent than using a tactical nuclear weapon. The target should be less willing to run this nuclear risk when the challenger has a strong second-strike capability.

The bottom of figure 1 shows how the challenger’s nuclear retaliatory capability affects the target’s willingness to concede to the challenger’s demands. If the challenger has a weak second-strike capability, then the target chooses to threaten escalation in response to a conventional defeat. The challenger finds this threat credible and chooses to back down, resulting in coercive failure. If the challenger has a strong second-strike capability, then the target will choose to back down because its escalatory options carry an intolerably high risk of escalation to a catastrophic nuclear war. That scenario results in coercive success.

Conventional options theory contributes to the nuclear politics literature in three ways. First, it builds on previous arguments about the potential role of conventional superiority in a nuclear world. Barry Posen argued that if Iraq had possessed nuclear weapons during the Gulf War, the United States could still have defeated Saddam Hussein, but the US military would have needed to restrict the scope of its operations.46 Essentially, Posen is arguing that the United States would still have won the war because it had effective limited conventional options. I extend that logic here and apply it to actual crises between nuclear powers. Mark Bell and Julia Macdonald contend that the conventional balance matters more in nuclear crises when escalation is more “controllable.”47 I point out that conventionally superior challengers not only need to stay within certain escalation thresholds, which keeps escalation more controllable, but they also need to translate their conventional superiority into the ability to achieve conventional success within those thresholds.

Second, the theory shows that both sides of the debate over the utility of achieving conventional success in a nuclear crisis have merit. It is possible for nuclear states to use limited conventional success, or the threat of achieving it, to coerce nuclear rivals. This coercion also works through the mechanism identified by Colby above, where the target could respond but finds that its response options are “some combination of too difficult, costly, or risky” to implement. However, challengers need to thread a fine needle to turn conventional success into coercive success. They must use force effectively while staying within certain thresholds. Thresholds on escalation, however, often act as constraints that tie military leaders’ hands and make it more difficult for military forces to achieve their operational goals. In addition, targeted nations attempt to deny challengers the ability to achieve quick or limited victories.

Targets could also adopt an asymmetric escalation nuclear posture. Such a nuclear posture disperses nuclear forces into the field and delegates the authority to use them to lower-level commanders early in a crisis.48 This measure harms the challenger’s ability to have an effective limited conventional option. It is more difficult to keep the use of force limited because the risk of accidental or unauthorized nuclear escalation is higher against a target with an asymmetric escalation posture.49 An asymmetric escalation nuclear posture could also serve to make the use of conventional force less effective. For the challenger to still try to keep the use of force limited, it needs to respect more stringent thresholds and take more precautions to avoid hitting the target’s nuclear forces. Those more restrictive escalation thresholds are likely to decrease the probability of conventional military success.

Third, the theory shows that nuclear superiority can play a role in a nuclear crisis, but a circumscribed one. This scenario applies to conventionally weak targets of coercion, and only confers a benefit at high levels of superiority, when that target has a meaningful damage-limitation capability over the challenger. Even then, the mechanism through how nuclear superiority confers a bargaining advantage is different than that proposed by previous scholarship.50 Rather than being emboldened, targets with nuclear superiority issue escalatory threats within the framework as their last resort to avoid making a concession. The challenger’s acute nuclear inferiority makes this threat credible and deters it from using conventional force.51

Predictions, Patterns, and Cases

To test the theory’s validity, I first generate predictions that take the form of a “causal generalization” appropriate in a “large-n qualitative analysis” (LNQA) design.52 I predict that when the challenger has an effective limited conventional option and a strong nuclear retaliatory capability, then a coercive threat of limited conventional success will work. Otherwise, coercion will fail.

In addition to this prediction on coercive outcomes, conventional options theory also generates predictions about the causal process observations that one should see in different types of cases if the theory is true.53 If the challenger lacks an effective limited conventional option, then the target should stand firm and coercion should fail because the target does not face a true choice between escalating or backing down. If the challenger possesses a Goldilocks option for the use of force but has a weak second-strike capability, the target faces a choice, but coercion should still fail because the target’s escalatory threat is credible, which deters the challenger from using conventional force at all. Finally, if the challenger has an effective limited conventional option and a strong second-strike capability, then coercion should succeed because the target chooses to make concessions rather than escalate to some military option that it finds too costly or risky. Table 1 summarizes these predictions.

Table 1. The predictions of conventional options theory

Condition 1: Challenger has Goldilocks option	Condition 2: Challenger has strong second-strike capability	Predicted outcome	Predicted causal process observation
No	Yes or No	Coercive failure	Challenger and target grow skeptical of the ability of challenger’s conventional forces to achieve success within certain thresholds
Yes	No	Coercive failure	Challenger finds target’s escalatory threats credible
Yes	Yes	Coercive successive	Target backs down rather than escalate to an option that is too costly or risky

To test these predictions, I identify all cases of coercion backed by the threat of achieving limited conventional success in a crisis. I assess whether the outcomes of those cases accord with the predictions of conventional options theory. To conduct this part of the research design, I first take all cases of a crisis between two nuclear powers that featured at least one side making a coercive demand of the other backed by some threat of conventional force. I list these cases and discuss how I arrived at them in online appendix A.1.54 I then identify the cases of conventional coercion that relied, at least in part, on the threat of achieving limited conventional success. Those cases appear in table 2.

Table 2. Threats of limited conventional success in nuclear crises

Crisis	Challenger-Target	Demand	Coercive strategies
Berlin II, 1958–59	USSR to US	Leave West Berlin; sign peace treaty with GDR	Limited conventional success
Berlin III, 1961	USSR to US	Leave West Berlin; sign peace treaty with GDR	Limited conventional success
Cuban Missile, 1962	US to USSR	Withdraw missiles from Cuba	Limited conventional success; conventional counterforce
Sino-Soviet War, 1969	USSR to China	Come to negotiating table over border disputes	Limited conventional success; nuclear risk; conventional counterforce
Cienfuegos, 1970	US to USSR	Remove submarine base	Limited conventional success
Yom Kippur War Nuclear Alert, 1973	USSR to US	Get Israel to abide by ceasefire	Limited conventional success
Kargil War, 1999	Pakistan to India	Recognize change to territorial status quo in Kashmir	Limited conventional success; nuclear risk
Kargil War, 1999	India to Pakistan	Withdraw behind Line of Control (LoC) in Kashmir	Limited conventional success
Operation Parakram, 2001–02	India to Pakistan	Cease support for jihadist insurgents; stem their flow into Kashmir; extradite twenty criminals	Limited conventional success
Operation Parakram, 2001–02	Pakistan to India	Stop coercive campaign	Limited conventional success; nuclear risk
US–Democratic People's Republic of Korea (DPRK), 2013	US to DPRK	Make concessions on nuclear arsenal	Limited conventional success
Russia-Ukraine War	Russia to US	Stop assistance to Ukraine; accept ceasefire on Russian terms	Limited conventional success; nuclear risk
Russia-Ukraine War	US to Russia	Stop war effort in Ukraine; cede captured territory to Ukraine	Limited conventional success (through Ukrainian forces)

For each coercive threat, I assess the outcome based on whether the target conceded to the challenger’s main demand(s). I also code the presence of each factor. The two factors can be combined to form a 2 x 2 grid, which appears in figure 2. Cases in bold indicate coercive success. I put US and Russian attempts to coerce each other in the Ukraine War in italics because that conflict is still ongoing and the final outcome is yet to be determined, though US policymakers, as of the time of this writing, appear willing to make concessions and accept a ceasefire closer to Russia’s terms.55 Explanations of coding for that case and others not examined in detail in this text appear in appendix A.3 in the online appendix.

Figure 2. Empirical patterns in coercive threats of conventional defeat. Figure by the author.

If conventional options theory is correct, then the top left cell should contain instances of coercive success. The other cells should contain examples of coercive failure. Figure 2 shows that successful coercion via threats of achieving limited conventional success in a nuclear crisis is rare. Of the thirteen coercive demands I identify as being backed, at least in part, by the threat of achieving limited conventional success, only five worked. As predicted by conventional options theory, these coercive successes are located in the top left cell. In the other cells, coercive threats of conventional defeat failed to compel the target.

The overall pattern of coercive outcomes therefore accords with the predictions of conventional options theory. These findings do not say, however, whether the causal process observations predicted by conventional options theory were present in each case, which matters because challengers often issue multiple threats in a nuclear crisis.56

This limitation leads to the second part of the research design, which is a focused comparison of two pairs of cases. These case studies examine whether the causal process observations predicted by conventional options theory were present and if the factors specified by the theory produced the observed coercive outcome. For the first pair of cases, I compare India’s attempts to coerce Pakistan in, first, the 1999 Kargil War, and second, in Operation Parakram (2001–2002), the latter of which involved India mobilizing troops to try to compel Pakistan to cease its support for Kashmiri jihadist insurgents after a terrorist attack on the Indian Parliament building in New Delhi. This pair of cases offers a good test of the theory for three reasons. First, the balance in resolve was similar because the issue of Kashmir, which featured in both crises, was important to both sides, since it is key to each countries’ sense of national identity.57 Second, the conventional balance did not change much between the two crises, allowing me to home in on the role of Goldilocks options in making the ability to defeat an adversary’s non-nuclear forces yield a coercive success.

Third, the nuclear balance and Pakistan’s and India’s nuclear postures remained consistent between the two crises. Matthew Kroenig codes India as having nuclear superiority across the two crises and argues that this imbalance helped India coerce Pakistan in the Kargil War.58 If nuclear superiority helped India succeed in the Kargil War, then presumably, it should also do so in Operation Parakram. Pakistan had a nuclear posture of asymmetric escalation across the two crises,59 which did make Indian leaders more wary about the use of conventional military force in both crises and imposed tighter constraints on the use of conventional force. A focus on Pakistan’s nuclear posture alone would struggle to explain why the coercive outcomes of both crises are different.

Conventional options theory, however, would predict a different coercive outcome. As I will show, India had a strong second-strike capability in both cases. In the Kargil War, India had an effective limited conventional option. In Operation Parakram, by contrast, India lacked an effective limited conventional option. Conventional options theory would predict a coercive success in the Kargil War and coercive failure in Operation Parakram. These parameters allow me to test the role of the theory’s first main independent variable—the possession of an effective limited conventional option—in producing coercive success in nuclear crises against two prominent alternative explanations.

For the second pair of cases, I compare Soviet attempts to coerce the United States in the Berlin Crises of 1958–59 and 1961 against US attempts to coerce the Soviet Union in the 1962 Cuban Missile Crisis. This comparison is helpful for testing conventional options theory because both target states in each crisis recognized that the challenger had effective limited conventional options. However, the strength in each challenger’s nuclear retaliatory capability was different. In the Berlin Crises, the Soviet Union had a weak second-strike capability vis-à-vis the United States. In the Cuban Missile Crisis, the United States had a strong second-strike capability. Conventional options theory thus predicts that the United States should be able to translate its conventional military advantage into a coercive success in the Cuban Missile Crisis, but not the Soviet Union in the Berlin Crises. This pair of cases is ideal to see whether the challenger’s possession of a strong second-strike capability affects its ability to send successful conventional coercive threats to a nuclear rival. This pair of cases also allows me to assess the strength of conventional options theory versus a theory based solely on the role of nuclear superiority.

A War Won, A War Unfinished: The Kargil War and Operation Parakram

During the partition of India in 1947, Pakistan and India disputed the status of Jammu and Kashmir, a Muslim-majority territory controlled by a Hindu ruler, Hari Singh. The Indian government claimed that Singh had signed an Instrument of Accession with India, but Pakistan challenged this, and Pakistani troops began trying to seize the province by force. India retaliated by dispatching troops of its own, leading to a bloody conflict that ended with a ceasefire along a Line of Control (LoC) in 1948. That line separated Pakistan-administered Kashmir from India-administered Kashmir. India and Pakistan fought wars over the territory in 1965 and 1971 and engaged in crises over the territory in 1984 (when India occupied the Siachen Glacier), 1986–87 (“Operation Brasstacks”), and 1990.60

Another crisis erupted again when Pakistan began infiltrating forces from the Northern Light Infantry (NLI) into unoccupied outposts on the Indian side of the LoC in the Kargil-Dras sector of India-administered Kashmir sometime in February and March 1999.61 By May 1999, Pakistani forces had penetrated 8–12 kilometers inside India-administered territory and took up positions overlooking the Srinagar-Leh Highway, which was critical to supplying Indian troops in the region, particularly those occupying the Siachen Glacier.62 The Pakistani military’s “theory of victory” is unclear, but it seemed predicated on Indian forces being unable to retake the high mountain posts. At that point, India would be reluctant to escalate, and the international community would be forced to intervene to end the conflict.63

Pakistani leaders would have been correct had India not possessed a Goldilocks option for the use of force. India could oust Pakistani troops from the high mountain posts without crossing the LoC. At first, Indian leaders imposed both the geographic threshold of staying within its side of the LoC and a scope threshold to only use infantry troops to retake the Kargil heights. In a May 18, 1999, meeting of the Cabinet Committee on Security (CCS), India’s Minister of External Affairs Jaswant Singh “expressed reluctance to authorize employment of air force assets owing to concerns about escalation.”64 After pleas from the Indian Army to involve the Indian Air Force (IAF), however, Indian Prime Minister Atal Vajpayee authorized the use of IAF missions to supplement the infantry attacks in a May 25 meeting, and the IAF was restricted from crossing into Pakistan-administered Kashmir (PaK).65

After initial setbacks, both the Army and Air Force adapted to the high-altitude environment and began to turn the tide of the battle. The most important battlefield within the Kargil sector was then occurring at Tololing Ridge. On May 22–23, the initial assault on the ridge and Point 5140 had failed, but Indian forces were able to retake it on June 20. Another important feature, Tiger Hill, was captured in the early hours of July 4.66 By the time Indian forces captured Tiger Hill, it was clear that the Indian Army “would slowly but surely overcome the defenders, barring significant reinforcement [by Pakistan] from north of the LoC.”67

India also had a strong second-strike capability vis-à-vis Pakistan despite its nascent nuclear arsenal. In 1999, analysts estimate that India had 8 nuclear warheads.68 India could have employed these weapons using Prithvi short-range ballistic missiles or as gravity bombs dropped by various aircraft serving in a dual nuclear-conventional role, such as the Jaguar, Mirage 2000, MiG-27, or Su-30MK.69 It is not clear how these nuclear weapons would have been targeted against Pakistan at this time, but given their limited accuracy, most targeting schemes involved striking Pakistani cities and civilian economic infrastructure.70 This outcome could be devastating to a country like Pakistan, which is a densely populated country where many of its people are concentrated in a few large cities.

Pakistan had a limited capability to destroy India’s nuclear forces in a first strike. It too had an estimated 8 warheads in its arsenal. This number would not have been enough to destroy all the airfields and associated hardened aircraft shelters that could have staged Indian dual-capable aircraft in a crisis.71 Pakistan also did not have a satellite in orbit that could track any Indian mobile missiles in real time, which meant that India’s nuclear weapons could cause a lot of damage to Pakistan, and Pakistan had little way to mitigate that damage. As India’s Defense Minister George Fernandes recalled about the crisis when speaking in 2004, if India launched a nuclear strike, “Pakistan may have been completely wiped out.”72

“The events in the Kargil War align with the predictions of conventional options theory.”

With an effective limited conventional option and a strong second-strike capability, conventional options theory predicts that India should achieve coercive success. Pakistan should back down because it chooses to make a concession rather than escalate.

The events in the Kargil War align with the predictions of conventional options theory. On the Indian side, the possession of an effective limited conventional option allowed Indian leaders to keep applying coercive pressure via military operations. In late May, as V. P. Malik noted, the time of conventional defeats for India was also “the closest we came to enlarging the conflict area.”73 One analyst of the war notes that had “India’s 56 Mountain Brigade not won the hand-to-hand contest for Tololing. . . . India would have been forced either to acquiesce to Pakistan’s territorial gains across the LoC or escalate the fighting and run the risk of nuclear war and international opprobrium.”74 India’s conventional victories from mid-June onward allowed it to avoid having to make this choice, and it continued to use military force.

On the Pakistani side, India’s conventional success imposed a choice on Pakistan’s leaders. In early July of that year, Pakistani Prime Minister Nawaz Sharif, “fearing wider escalation and eventual defeat, sought assistance from the United States” in ending the conflict and flew to Washington, DC, to meet with President Bill Clinton.75 At this meeting, Bruce Reidel, then a member of the National Security Council (NSC), recalls that “Sharif had a choice, withdraw behind the LoC and the moral compass would tilt back toward Pakistan or stay and fight a wider war with India without American sympathy.”76 The decision to keep fighting and escalate the war would have also increased the risk of nuclear escalation. The result of the summit was a Pakistani announcement on July 4 that it would withdraw its remaining forces in India-administered Kashmir behind the LoC, restoring the status quo ante. Faced with a choice between escalating or backing down, Pakistan chose the latter.

In contrast to the Kargil War, India lacked an effective limited conventional option vis-à-vis Pakistan during Operation Parakram in 2001–2002. On December 13, 2001, members of militant groups Lashkar-e-Taiba (LeT) and Jaish-e-Mohammed (JeM) attacked the Indian Parliament building in a shootout.77 After the attack, India made a series of coercive demands of Pakistan, including that it extradite twenty criminals, check jihadi infiltration into Jammu and Kashmir, renounce terrorism, and shut down terrorist training camps in Pakistani territory.78 To back up these threats, Operation Parakram mobilized over 500,000 troops for a period of eight months.79 The plan for these forces in January 2002 was to make limited thrusts across the LoC into Pakistan-administered Kashmir. After the terrorist attack on the Kaluchak barracks in May 2002, India modified its plan to take its three strike corps into Rajasthan and inflict a massive defeat on Pakistani ground forces.80

India could not successfully execute its threatened military operations while also staying within desired escalation thresholds. Consider the first plan for limited strikes across the LoC in Kashmir. Kanti Bajpai points out that India had three strategies for operations across the LoC: hot pursuit, surgical strikes, and destroying military camps and holding territory in Pakistan-administered Kashmir.81 Hot pursuit represented an option in which India used airstrikes alone to destroy terrorist camps in Pakistan-administered Kashmir, while surgical strikes included the use of special operations forces in addition to air forces to clear out terrorist camps. Admiral Raja Menon, a former submarine commander and now defense analyst in India, assessed in 2002 that India did not have the capability to do either.82 Menon wrote that “the army has its own army aviation corps but no helicopter that can carry more than four men. . . . Between the three services, we cannot muster one helicopter that can fly low over rough terrain on a dark night.”83 Other Indian military professionals threw cold water on the idea of hot pursuit or surgical strikes in Pakistan-administrated Kashmir. Col. P. N. Khera, then editor of Asian Defense News International, declared that “people who talk of hot pursuit don’t know the terrain. It would be like walking into a trap. The Pakistanis are, after all, not sitting on their side waiting to be shot.”84

The inability of Indian forces to conduct limited attacks into Pakistan-administered Kashmir led India to plan to attack Pakistan itself. This plan clearly outstripped geographic and scope escalation thresholds, but it also would have required a long war to achieve any operational success. At this time, India still operated according to the Sundarji Doctrine, in which the forces near the border were organized into “Holding Corps” while much of the military’s offensive power was stationed in three “Strike Corps” in the middle of the country.85 The “Strike Corps” contained more capabilities for striking into Pakistan, but they took time to mobilize (hence the order for mobilization that initiated Operation Parakram).86 Even once these forces did mobilize and move in place, their ability to conduct strikes into Pakistan in a short time frame was questionable, because Pakistan had invested in defensive forces and infrastructure close to the international border and the LoC. These forces were built up along a 2,000-km long vertical line of “canal defenses” with concrete bunkers, pill boxes, and minefields up to a depth of 1000–2000 meters.87

The result was “operational-level parity” in which “India did not have any qualitative or quantitative advantage over Pakistan” in a short war.88 India therefore lacked the ability to take small swaths of territory in Pakistan in a period of less than four to six weeks.89 India may have been able to take Pakistani territory and bring its aggregate conventional superiority to bear, but only in a longer, more intense war that was less constrained in time, geography, and scope.

The nuclear balance remained the same in 2001–2002 as in 1999, meaning India still retained a strong second-strike capability. The lack of an effective limited conventional option, however, meant that its coercive attempts failed. Pakistani leaders never truly faced a choice between backing down or escalation and could therefore stand firm against India’s coercive demands. President Pervez Musharraf gave a speech on January 12, 2002, denouncing terrorism and banning JeM and LeT. US Secretary of State Colin Powell visited the region and assured India that Pakistan was serious about curbing terrorist activity and was considering the extradition of the twenty criminals.90

Pakistan, however, did not extradite the twenty criminals, and insurgent infiltration into Kashmir continued. In June 2002, the United States sent Deputy Secretary of State Richard Armitage to Pakistan. Armitage extracted a promise from Musharaff to curb cross-border infiltration and shut down training camps; Powell conveyed this commitment to Indian leaders and the international community on June 10.91 Pakistan’s compliance, however, was mixed, and most US policymakers viewed them as “expedient” rather than “substantive.” Both Powell and Armitage “knew that the Government of India knew that it could not bank on Musharraf’s promises,” which provided a convenient exit strategy for India rather than a firm set of assurances.92 This situation led Indian military analysts to think that “none of the other Indian demands [in addition to the check on cross-border infiltration] were met because of this so-called coercive diplomacy.”93

“India’s conventional coercion likely failed because Pakistani leaders did not fear a conventional defeat once their forces were in place near the border with India.”

India’s conventional coercion likely failed because Pakistani leaders did not fear a conventional defeat once their forces were in place near the border with India. Pakistan moved two infantry divisions of the 10 (Rawalpindi) Corps and an independent armored brigade close to the international border in Punjab and Rajasthan in late December.94 As these and other forces got into position, “the atmosphere at GHQ [General Headquarters] and ISID [Inter-Services Intelligence Directorate] was far more relaxed” by January 2002.95 This sentiment carried over into the second phase of the crisis in May 2002, with some Pakistani military officers saying that “India doesn’t have the military capability to bring Pakistan to its knees. That is our assessment.”96

Meanwhile, the lack of effective limited conventional options may have deterred Indian leaders from authorizing military operations. Peter Lavoy writes that “it is widely speculated that Indian Prime Minister Atal Bihari Vajpayee decided against a military attack when his troops had moved into their strike positions by the middle of January because Pakistani troop deployments indicated that Islamabad was well-prepared to counterstrike at locations of its choosing.”97 That development led India to ease off its coercive pressure, resulting in coercive failure.

Conventional options theory provides a better explanation of the difference in coercive outcomes between the Kargil War and Operation Parakram than a theory based on nuclear superiority alone. If nuclear superiority helped India in the Kargil War, then why would it not have done so in Operation Parakram? In addition, little evidence exists that any supposed nuclear superiority emboldened Indian leaders to take more risks in either crisis. Pakistan’s asymmetric escalation posture made Indian leaders more cautious about the use of conventional force in both crises, but one still needs to explain why the coercive outcome of each crisis was different for India. A focus on the role of effective limited conventional options can do that.

The Berlin Crises and the Cuban Missile Crisis

In November 1958 and June 1961, Soviet Premier Nikita Khrushchev sent ultimatums to the United States, along with France and Great Britain, pressuring them to withdraw from West Berlin, effectively turn it into what he called a “free city,” and recognize East Germany. As noted in the introduction to this article, Khrushchev backed up this demand with the threat of defeating allied forces in Germany. He issued such a coercive threat because he thought that Soviet conventional superiority in the area around Berlin gave him coercive leverage. “These days with regard to conventional weapons,” Khrushchev told his colleagues in the Presidium (the Soviet Communist Party’s main decision-making body), “these considerations do not concern Berlin. This is a matter of consideration for Laos, for Cuba, for the Congo, even, perhaps for Iran. But [in Berlin] meanwhile we are stronger than they are, and they say, ‘The Russians have the advantage. . . .’ This means, they will agree.”98

Indeed, the United States recognized this threat of conventional defeat as credible and knew that it would force them to choose between escalating or backing down. US leaders could not stand firm if the Soviets were to attack West Berlin. An NSC document from before the crisis stated that “if determined Soviet armed opposition should develop when US units attempt to force their way into or out of Berlin, no additional units would be committed, but resort would have to be made to general war.”99 During the crisis, the US ambassador to the Soviet Union, Llewellyn Thompson, told President John F. Kennedy that Khrushchev has “always felt that he had us over a barrel in Berlin.” Kennedy responded with a laugh: “Yeah. I think he does.”100

At the same time, the Soviet Union lacked a strong second-strike capability vis-à-vis the United States. In the late 1950s and early 1960s, the Soviet Union was struggling to deploy much long-range nuclear strike capability that could hit the US homeland. By 1961, the Soviet Union only had four operational ICBM launch sites.101 That number has led some nuclear scholars to conclude that the Soviet Union had a weak second-strike capability vis-à-vis the United States during the period of the Berlin Crises.102

The Soviet Union had a Goldilocks option for the use of force but a weak second-strike capability against the United States during the Berlin Crises. Conventional options theory predicts that, contrary to Khrushchev’s bravado, coercion should fail. This failure should come because the United States, which had nuclear superiority, could choose to escalate or send a credible escalatory threat, and deter Moscow from using conventional force.

Evidence from the Berlin Crises accords with these predictions. On the American side, US leaders decided to rely on nuclear threats to dissuade the Soviets from following through on their conventional coercive threat. In an NSC meeting on December 11, 1958, President Eisenhower told his advisors that the United States needed to “make it clear that we consider this no minor affair. In order to avoid beginning with the white chips and working our way up to the blue, we should place them on notice that our whole stack is in play.”103 Secretary of State John Foster Dulles communicated this determination to resist the Soviet ultimatum to Anastas Mikoyan (first deputy chairman of the Council of Ministers and one of Khrushchev’s main advisers) during his visit to the White House on January 5, 1959: “We did not want war over Berlin, nor for that matter, over anything; but we were not prepared to avoid war by retreating wherever we were under pressure.”104

Kennedy displayed a similar willingness to rely on nuclear threats. He wrote to General Lucius Clay, named head of NATO forces in West Berlin during the 1961 crisis, that “it is central to our policy that we shall have to use nuclear weapons in the end, if all else fails, in order to save Berlin.”105 In a speech on July 25, 1961, Kennedy declared that the US commitment to West Berlin would “not be broken in time of danger.”106 Unlike Pakistan in the Kargil War, when faced with a choice between escalating or backing down, the United States chose to threaten escalation.

Khrushchev eventually found American nuclear threats credible. At first, he did not think that US leaders, even with nuclear superiority, had the resolve to run nuclear risks for the sake of West Berlin. He told columnist Walter Lippmann in April 1961: “In my opinion there are no such stupid statesmen in the West to unleash a war in which hundreds of million would perish just because we would sign a peace treaty with the GDR that would stipulate a special status of ‘free city’ for West Berlin with its 2.5 million population. . . . There are no such idiots or they have not yet been born.”107 After Kennedy’s speech, however, Khrushchev changed his assessment. On August 3, 1961, he told a group of Warsaw Pact leaders: “Anything is possible in the United States . . . War is possible. They can unleash it.”108 This change prompted him to drop his ultimatum strategy and initiate a “fallback” option of building the Berlin Wall to at least stem the flow of emigrants from East Germany. US nuclear threats thus deterred Khrushchev from following through on his ultimatums, which led his coercive threats to fail.

“US nuclear threats thus deterred Khrushchev from following through on his ultimatums, which led his coercive threats to fail.”

Yet Khrushchev still wanted to achieve his ultimate goals on Berlin, and he decided to send nuclear medium-range and intermediate-range ballistic missiles (MRBMs and IRBMs) to Cuba to put pressure on the United States, provoking the Cuban Missile Crisis of October 1962.109 In doing so, he was trying to install the nuclear-tipped missiles secretly, as opposed to coercing the United States to accepting the installation of nuclear missiles in Cuba.

When the United States discovered the missiles, Kennedy and his advisers set about getting the Soviet missiles out of Cuba through a strategy of conventional coercion. Thus, although Khrushchev’s plan to deploy missiles to Cuba instigated the crisis, it was the United States that acted as the challenger, or the actor sending the coercive threat. In the Cuban Missile Crisis, the United States had conventional options that were somewhat limited and a strong second-strike capability, and these circumstances contributed to American coercive success, though not without plenty of moments of danger.

At first, many of Kennedy’s military and political advisors advocated outright destroying the missile sites. Kennedy closed an ExComm meeting on October 16 saying: “We’re certainly going to do number one. We’re going to take out these missiles.”110 The administration eventually settled on trying to convince Khrushchev to remove the missiles through coercive diplomacy. The United States would institute a naval blockade, or quarantine, and threaten a subsequent airstrike plus invasion of Cuba should Khrushchev fail to remove the missiles. Secretary of Defense Robert McNamara called this strategy “a minimum military action, a blockade approach, with a buildup to subsequent action.”111 This move was a “blockade and negotiate” strategy, with the threat of an airstrike plus invasion acting as the main coercive tool.

Did the threatened airstrike plus invasion constitute an effective limited conventional option for the United States? On the one hand, an attack on the missile sites would transgress targeting thresholds, which made US leaders worry about the risk of escalation to nuclear war. The Soviets could escalate by employing nuclear weapons within the theater or respond by attacking Berlin or Turkey. On the other hand, a US airstrike and invasion likely would have succeeded in taking out the Soviet MRBMs and IRBMs given US conventional superiority in the area around Cuba. Such an attack could be kept limited to the theater and respect geographic thresholds by not attacking Soviet forces elsewhere. Therefore, the United States had an effective conventional option that was somewhat limited.

The United States also had a strong nuclear retaliatory capability vis-à-vis the Soviet Union. During the crisis, William Kaufmann, a political science professor at MIT who was also serving as a special assistant to Secretary McNamara, assessed that in an all-out Soviet attack, even including the missiles fired from Cuba, the United States “would be left with at least 483 nuclear weapons” to fire in retaliation at the Soviet Union.112 As Deputy Secretary of Defense Roswell Gilpatric declared in a speech the previous October in 1961: “We have a second strike capability which is at least as extensive as what the Soviets can deliver by striking first.”113

Conventional options theory predicts that with a somewhat limited, effective conventional option, US leaders should be concerned about following through on their conventional threat, but also should be willing to press their coercive demands. And when Soviet leaders received continued American pressure, they should choose to back down rather than escalate.

The events of October 27 that led to the resolution of the crisis support these predictions. On that day, Khrushchev sent a second, public letter to President Kennedy offering a missile swap—the Soviet Union would remove its missiles from Cuba, and the United States would remove its Jupiter MRBMs from Turkey.114 This followed a private, first letter on October 26 that had laid out a narrower offer to remove the missiles in exchange for a US pledge to not invade Cuba. The public letter caused Kennedy to think that the window for coercing the Soviets had closed; his choice for resolving the crisis was now between taking out the missiles by force or publicly accepting the missile swap. Many of Kennedy’s advisers found the latter unacceptable. National Security Adviser McGeorge Bundy said that “if we accept the notion of a trade at this stage, our position will come apart very fast.”115

Kennedy, however, feared an airstrike plus invasion would create nuclear escalation. The Soviet Union could horizontally escalate by striking Berlin or Turkey, greatly increasing the risk of nuclear war. In an earlier meeting with the Joint Chiefs of Staff on October 19, Kennedy said that if he attacked Cuba, “there’s bound to be a reprisal from the Soviet Union, there always is—[of] their just going in and taking Berlin by force.” Due to US conventional weakness in Berlin, described above, Kennedy felt that such a move would “leave me only one alternative, which is to fire nuclear weapons—which is a hell of an alternative.”116 By October 27, this dilemma made Kennedy willing to accept the public missile swap, which meant convincing NATO allies to go along with it by telling them “that in two or three days we may have a military strike which would bring perhaps the seizure of Berlin or a strike on Turkey, and then they’ll say ‘My God, we should have taken it.’”117 Doubts about the ability to keep a conventional option limited made Kennedy less willing to execute an airstrike plus invasion against Cuba and more willing—if there were no other alternative courses of action available—to accept the public missile swap.

Kennedy did not, however, totally abandon efforts to keep coercive pressure on the Soviet Union to reach a settlement more favorable to American interests. Former ambassador to the Soviet Union, Llewellyn “Tommy” Thompson, argued that there was still room for coercive diplomacy to work—the United States could still get Khrushchev to “refocus on Cuba,” because “the important thing for Khrushchev, it seems to me, is to be able to say, ‘I saved Cuba, I stopped an invasion.’”118 Refocusing on Cuba was attractive to the members of ExComm (the committee of top-level national security decision-makers in the Kennedy administration) because it would help the United States avoid the alliance cohesion difficulties involved in publicly accepting a missile trade. Thompson’s intervention led Kennedy and his advisers to favor a “Trollope ploy.”119 Before authorizing any airstrike plus invasion or accepting the public missile swap, the United States would publicly respond to the first letter that only mentioned a non-invasion pledge towards Cuba, and would give private assurances that the Jupiters would be removed in a few months (Kennedy wanted these missiles removed anyway).120 To get the Soviets to agree, the United States would continue to threaten to attack Cuba.

On October 27, the Joint Chiefs of Staff recommended carrying out an airstrike on the missile sites (OPLAN 312) by October 29 and an invasion (OPLAN 316) within a week afterwards.121 Preparations for these operations had been ongoing throughout the crisis. McNamara said on the evening of October 27 that the United States should continue these preparations to “keep some kind of pressure on.”122 Kennedy agreed, saying, “I think we ought to do it.”123 He also dispatched his brother, Attorney General Robert F. Kennedy, to meet with Anatoly Dobrynin, the Soviet ambassador to the United States. In that meeting, Kennedy told Dobrynin that the United States would pledge to not invade Cuba in exchange for a Soviet missile withdrawal, and gave Dobrynin a private assurance that the Jupiters would be withdrawn. To convince the Soviets to accept the American offer, Robert Kennedy also stressed that preparations for an attack against Cuba were proceeding rapidly and that US generals were “itching for a fight.”124 In other words, despite trepidation about carrying out an airstrike plus invasion, the United States continued to press conventional coercive threats against the Soviets.

“Despite US fears of an escalatory Soviet reprisal against Berlin, Khrushchev never considered this a serious option.”

When Soviet leaders received intelligence on continued American preparations for an invasion coupled with Dobrynin’s notes on his meeting with Robert Kennedy, they found the American threats credible. This information made them eager to accept the American deal to end the crisis. In a marathon Presidium meeting on October 28 at Novo-Ogaryovo (late October 27/early morning of October 28 in Washington, DC), Oleg Troyanovsky, a top Kremlin aide, “began to read his notes on Dobrynin’s report [to the Presidium]. They [Khrushchev and the others] asked me to read the notes again. It goes without saying that the contents of the dispatch increased the nervousness in the hall by some degrees.”125 Soviet military intelligence also picked up word that Kennedy would make a major speech at 9 a.m. on October 28, and Soviet leaders thought this might be a speech signifying the start of conventional operations.126 After Khrushchev received these two pieces of information, he rushed to agree to withdraw the missiles, conceding to US demands.

Soviet leaders felt nervous at the prospect of an American airstrike plus invasion because if the Americans did attack Cuba, the Soviet Union would face a choice between withdrawing the missiles or escalating the crisis. Despite US fears of an escalatory Soviet reprisal against Berlin, Khrushchev never considered this a serious option. On October 23, Vasily Kuznetsov, a prominent Soviet Foreign Ministry official, proposed challenging Berlin in response to an American attack on Cuba. Khrushchev immediately shut that line of thinking down. “Keep that kind of advice to yourself,” he told Kuznetsov. “We don’t know how to get out of one predicament but you drag us into another.”127 On October 25, Khrushchev may have been referencing Berlin when he vaguely said that “we have succeeded in some things and not in others,” after which, as Fursenko and Naftali note, “discussion of the Berlin tie-in immediately became taboo.”128 Indeed, in the notes by Vladimir Malin (chief of the General Department of the Central Committee) during the Presidium meeting on October 28, when the prospect of an American attack was at its highest, there was no mention of any discussion of conducting a retaliatory conventional attack against West Berlin or any other target outside the Caribbean.129

These developments meant that Soviet decision-makers, from their perspective, faced the choice on October 28 between conceding to the latest US terms offered by Kennedy or escalating to the use of tactical nuclear weapons to resist the impending American attack on Cuba. As Lieutenant General Nikolai Beloborodov, commander of the Soviet ballistic missile forces sent to Cuba, recalled: “It was clear that in the conditions of the existing balance of forces in conventional arms, which was ten to one against us, there was only one way we could repel a massive assault—by using tactical nuclear weapons against the invaders.”130 Khrushchev was unwilling to take that step. Before the crisis, Khrushchev had sent LUNA rockets that could fire tactical nuclear warheads, along with four submarines each armed with a tactical nuclear missile, to Cuba and authorized local Soviet commanders to use them. Within the crisis, however, he tried to draw back some of this authority. Khrushchev cabled General Issa Pliyev, the head of Soviet forces in Cuba, to “use all means except those controlled by Statsenko [Major General Igor D. Statsenko, the commander of the rocket units] and Beloborodov.”131

Khrushchev had no good symmetrical response to the threatened American attack on Cuba, but he was unwilling to escalate. This dilemma put heavy pressure on him to back down. In the words of Anastas Mikoyan, American conventional military pressure “made it more and more clear that we would have to give in.”132 Mikoyan’s son, Sergo Mikoyan, later assessed that the threatened airstrike plus invasion hung like a “sword of Damocles” over Soviet leaders, and “it was precisely this sword which in the end played the deciding role in accelerating Khrushchev’s proposal for a compromise.”133 By October 28, Sergei Khrushchev wrote that indications of an imminent American invasion meant that Moscow “had to accept Kennedy’s proposal now, before a war started,” and that a public trade “would have to be given up.”134 Conventional options theory can explain why US threats of conventional attack, despite worrying American leaders about escalation risks, convinced the Soviets to remove their missiles from Cuba.

One counterargument is to contend that nuclear superiority was decisive. The United States had far more nuclear weapons than the Soviet Union and achieved its goals in both crises.135 Further examination casts doubt on that conclusion and provides more support for conventional options theory for three reasons. First, Khrushchev issued coercive threats in both Berlin Crises even though he possessed inferior nuclear capabilities. If the nuclear balance were decisive, why would Khrushchev not recognize the Soviet Union’s nuclear inferiority and demur from issuing an ultimatum in the first place?

Second, in the Cuban Missile Crisis, Khrushchev likely heard about the United States’ DEFCON 2 alert sometime on October 25, but even after that, his willingness to withdraw the missiles rose and fell based on his assessment of the probability of a US conventional attack.136 On October 26, Soviet intelligence indicated that a US invasion of Cuba was imminent. In response, Khrushchev decided to send the non-invasion pledge for missile withdrawal offer.137 When no invasion came over the next 24 hours, Khrushchev became calmer and decided to push for the withdrawal of Jupiter missiles from Turkey in addition to the non-invasion pledge. Khrushchev stated, “Can they attack us now? I think that they won’t venture to do this.”138 On October 27, after hearing Robert Kennedy’s conventional military threat via Dobrynin, Khrushchev again became eager to accept a deal to end the crisis. If one thinks that nuclear coercive threats were the factor driving Khrushchev to back down, this process would be difficult to explain.

Third, Kennedy and other US leaders were not emboldened by nuclear superiority, especially during the Cuban Missile Crisis. In a meeting on October 19, the Air Force chief of staff, General Curtis LeMay, pushed Kennedy to do something more aggressive than the “blockade and negotiate” strategy, in part because the United States had nuclear superiority. Kennedy disagreed with LeMay’s premise. As noted above, he described nuclear war as “a hell of an alternative,” and told LeMay that “you’re talking about the destruction of a country.”139 Kennedy did not seem encouraged by American nuclear superiority to make coercive nuclear threats. This sentiment is perhaps why US leaders relied primarily on conventional threats to convince the Soviets to remove their missiles from Cuba.

Conclusion: Discussion and Implications

A long-standing debate within the field of nuclear politics is whether nuclear powers can turn conventional success into coercive success in a nuclear crisis. Scholars that support the stability-instability paradox argue that nuclear powers can engage in faits accompli or “salami-slice” tactics. These limited conventional successes can coerce a nuclear adversary to accept changes to the status quo because escalating to the nuclear level in response to them is less rational. Scholars that align with the theory of the nuclear revolution point out that because controlling escalation is difficult, nuclear war could occur even if no one wants it to. As a result, threats of achieving limited conventional success have less potency, and conventional forces mainly coerce nuclear adversaries by making “threats that leave something to chance.”

This article advances that debate by identifying when and how threats of achieving limited conventional success can coerce nuclear rivals. Conventional options theory argues that coercive demands backed by the threat of inflicting a conventional defeat succeed when the challenger has an effective limited conventional option and a strong nuclear retaliatory capability.

Conventional options theory shows that both perspectives on the coercive utility of limited conventional success in a nuclear crisis have merit. There are conditions under which states can use the prospect of conventional defeat to coerce their nuclear rivals, but the conditions to engage in successful conventional coercion are rare. Coercive threats of conventional defeat require challengers to simultaneously show restraint by respecting various escalation thresholds and bring enough force to bear to achieve relevant military objectives. As shown, accomplishing both at the same time is difficult. In addition, targets actively try to thwart a challenger’s ability to thread this needle: They build conventional capabilities to prevent the potential challenger from achieving a quick or limited victory, and can adopt asymmetric escalation nuclear postures, making it more difficult to keep conventional operations limited and imposing more stringent escalation thresholds on challengers.

“This article’s findings imply that successful coercion via the threat of limited conventional success may not just be rare but also a product of good fortune.”

Conventional options theory also shows that nuclear superiority matters in a nuclear crisis, but in a circumscribed way. The benefit of nuclear superiority applies to conventionally weak targets of coercion that have a meaningful damage-limitation capability, but possessing such a capability requires the target to have a high degree of nuclear superiority over the challenger. Even then, the target with nuclear superiority turns to threats of nuclear escalation as a last resort to avoid concessions, rather than due to emboldenment. It is the challenger’s acute nuclear inferiority that makes this escalatory threat credible.

This article’s findings imply that successful coercion via the threat of limited conventional success may not just be rare but also a product of good fortune. Even if a challenger can find an effective limited conventional option, its conventional coercive efforts may generate high risks of nuclear escalation that are hard for policymakers to foresee. There were several “close calls” in the Cuban Missile Crisis, such as an American U-2 spy plane straying into Soviet airspace or a test tape causing a radar facility in Moorestown, New Jersey, to pick up indications of a Soviet nuclear strike.140 In addition, Khrushchev tried to revoke the authority to use nuclear weapons from field commanders, but there was a limit on how much he could do to revoke the ability to give launch orders, especially for submarine commanders141—producing a tense situation where a Soviet submarine crew nearly decided to use its nuclear torpedo against American ships.142 American officials recognized a risk of nuclear escalation in the Cuban Missile Crisis, but likely not from these sources, raising the possibility that other crisis managers will not be so lucky.

Conventional options theory has relevance for policymakers trying to reorient US security policy in a world in which the US faces two major nuclear adversaries. The US military, which after the Cold War became used to operating in uncontested and nuclear-free environments, now must adapt to a context where multiple adversaries could cross the nuclear threshold, and could threaten to do so in response to US and allied conventional success.143 This situation is what General Mark Milley, former chairman of the Joint Chiefs of Staff, called, in the context of war between Russia and Ukraine, the “nuclear paradox.” In his words: “The more successful the Ukrainians are at ousting the Russian invasion, the more likely that Putin is to threaten to use a bomb—or reach for it.”144 This articulation of the power-risk tradeoff applies to US defense planning today vis-à-vis both Russia and China.

Conventional options theory at least provides a starting point for addressing this problem set. In a future crisis with a nuclear-armed adversary, if the United States wants to achieve conventional success and then translate that into coercive success, planners should work now to identify the thresholds and boundaries that, if respected, would generate a tolerable risk of nuclear escalation. An example could be withholding attacks on certain leadership targets or refraining from attacking targets in an adversary’s homeland. The United States should then develop operational plans that can succeed within those thresholds, and policymakers need to provide the resources to execute those plans. If those plans succeed, the United States should be prepared to issue deterrent threats to encourage the adversary to back down rather than escalate. Admittedly, success is not guaranteed; it could be difficult to craft operational plans that succeed within certain thresholds, and an adversary may choose to escalate despite the United States’ best efforts. Such an approach, however, presents the best chance of using conventional warfighting operations to achieve political success in today’s nuclear environment.

 

Tyler Bowen is an assistant professor in the Deterrence Studies Institute and Strategic and Operational Research Department at the United States Naval War College. His work focuses on issues related to nuclear deterrence, nuclear strategy, and US grand strategy. He previously worked as a visiting professor at the United States Air Force Academy and as a postdoctoral fellow at Johns Hopkins SAIS. He received his PhD in political science from Yale University in 2021.

Acknowledgments: I would like to thank two anonymous reviewers, David Allison, Mark Bell, Damon Colletta, Fiona Cunningham, Alexandre Debs, Michael Desch, Matthew Fuhrmann, Charles Glaser, Michael Goldfien, Phil Haun, Stephen Herzog, Fahd Humayun, Charlotte Hulme, Patrick Hulme, Kendrick Kuo, Alex Yu-Ting Lin, David Logan, Vipin Narang, Michael Poznansky, Erik Sand, Shane Smith, Caitlin Talmadge, and especially Nuno Monteiro for their feedback and/or mentorship throughout the process of researching and writing this paper. All errors are my own.

Image: The Command Post of the North American Air Defense (NORAD) Cheyenne Mountain Complex. Computer-generated images are projected on two large display screens145

 

© 2026 Tyler Bowen