Proportionate Deterrence

Proportionate Deterrence: A Model Nuclear Posture Review

George Perkovich, Pranay Vaddi

Executive Summary

Ever since the election of Bill Clinton in 1992, every U.S. presidential administration has published a Nuclear Posture Review (NPR) that explains the rationales behind its nuclear strategy, doctrine, and requested forces. These reviews have helped inform U.S. government personnel, citizens, allies, and adversaries of the country’s intentions and planned capabilities for conducting nuclear deterrence and, if necessary, war. The administration that takes office in January 2021 may or may not conduct a new NPR, but it will assess and update nuclear policies as part of its overall recalibration of national security strategy and policies.

Nongovernmental analysts can contribute to sound policymaking by being less constrained than officials often are in exploring the difficulties of achieving nuclear deterrence with prudently tolerable risks. Accordingly, the review envisioned and summarized here explicitly elucidates the dilemmas, uncertainties, and tradeoffs that come with current and possible alternative nuclear policies and forces. In the body of this review, we analyze extant declaratory policy, unclassified employment policy, and plans for offensive and defensive force postures, and then propose changes to several of them. We also will emphasize the need for innovative approaches to arms control.


The best nuclear doctrine and force posture would be one that—

  • is credible enough to deter adversaries and reassure allies and partners;
  • is least likely to provoke escalation if deterrence fails but could survive adversary escalation if it occurred; and
  • would not cause more destruction than necessary in the event of nuclear war, bearing in mind the law of armed conflict, and would engender deescalation.

That said, the best nuclear policy is one that encourages stable deterrence relationships among political adversaries, helps to preserve and strengthen international cooperation in preventing nuclear weapons proliferation and possible use, all while promoting the reduction of threats and arsenals. Recognizing that best outcomes are rarely achievable in the real world, this review highlights some of the challenges that must be overcome to bring U.S. nuclear policy closer to the ideal. In all of this, premium is placed on striving for proportionality between the threats that the United States and its allies face and the ends and means they pursue to deter or defeat them.


Nuclear weapons should be reserved for deterring threats of a scale and type that cannot be deterred or defeated by other means. Russia, China, and the Democratic People’s Republic of Korea (DPRK; North Korea) are the only potential sources of such threats today.1 These countries possess nuclear weapons and growing non-nuclear capabilities, including perhaps biological weapons in some cases, and have antagonistic relations with the United States and their U.S.-allied neighbors. The challenge is to dissuade Russian, Chinese, and North Korean leaders from believing that their nuclear and non-nuclear capabilities could enable them to successfully prosecute regional conflicts while deterring the United States from escalating as necessary to defend its allies.

Russia drives most U.S. nuclear requirements to the extent that its nuclear arsenal threatens the survivability of the U.S. nuclear deterrent and comprises the largest set of targets for U.S. forces. Russia seeks to weaken its adversaries through the lowest level of violence necessary and has developed a range of means to do so, including information warfare and cyber attacks, political subversion, and economic coercion. As chapter 2 describes, Russia also has deployed or is developing sophisticated conventional strike weapons and new nuclear systems for theater and intercontinental missions. Russia has not acquired these new capabilities in a vacuum, but rather as part of an action-reaction dynamic with the United States and other NATO (North Atlantic Treaty Organization) states.

To deter or defeat Russian threats below the level of armed conflict, NATO and the United States must bolster their resilience, unify their polities, and enhance conventional military and other coercive capabilities. And, because Russia derives coercive value from nuclear weapons, U.S. and NATO policymakers must deploy nuclear capabilities and defenses sufficient to credibly counter and thereby deter potential Russian attempts to prevail in escalatory armed conflict. This can be done while making clear that mutual threat reduction would be more beneficial to all concerned.

China also poses numerous and growing challenges to the United States and its allies and partners. Many of these challenges are economic and diplomatic, and therefore not central to nuclear deterrence. More pertinently, China continues to acquire a wide range of kinetic and nonkinetic capabilities to prevail in conflicts around its periphery while deterring the United States from escalating in defense of its allies and partners, particularly Japan and Taiwan. China also is increasing the sophistication, number, and survivability of its relatively small nuclear force, though compared with Russia (and the United States) it has not placed nuclear weapons in the forefront of its rhetoric, doctrine, and threat projection.

As with Russia, the priority of the United States and its allies and partners must be to strengthen their non-nuclear deterrence and defense capabilities in ways that do not exacerbate risks of inadvertent nuclear escalation with China, and to deploy nuclear weapons in ways that discourage destabilizing arms racing and potential escalation of war that neither side can plausibly win.

North Korea, too, poses conventional and nuclear threats to South Korea and by extension the United States. However, these threats do not require nuclear capabilities beyond those which the United States would deploy to deter or defeat escalatory conflicts with Russia or China, in part because efforts to acquire new capabilities to threaten the DPRK’s mobile nuclear weapons likely would exacerbate instabilities in U.S.-China relations.

This review highlights that the central overall challenge for U.S. nuclear policy is how to deter or counter adversary escalation of regional conflict and avoid catastrophe for all. Escalation can occur through calculation and/or inadvertence, especially as new cyber and kinetic technologies become entangled with nuclear force operations. Beyond the prospect of tens or hundreds of millions of people dying in a nuclear war, some scenarios of nuclear war produce the real possibility of an extinction-class event caused by the climatic and environmental harm of the atmospheric particulates produced by a nuclear exchange. Policymakers in all nuclear-armed countries have neglected this danger in recent decades even as recent modeling indicates it is irresponsible to ignore.


States generally put more stock in each other’s capabilities and actions than their declared intentions. At the same time, a state’s nuclear policies and forces require rationales to guide them. Declaratory policy articulates such rationales and intentions to one’s population and defense establishment, and to allies and adversaries, reflecting when the government thinks it could be prudent and justifiable to use nuclear weapons. Even if decisionmaking on capabilities sometimes has a logic of its own, declaratory policy should guide the acquisition and posturing of forces, as well as efforts to reduce unnecessary or destabilizing capabilities.

There is no perfect or nonproblematic declaratory policy. It may be tempting to issue blustery or vague threats of nuclear war in hopes of deterring all forms of aggression. Yet, because deterrence could fail, it would be folly to make threats that would be self-defeating to carry out, just as it would be imprudent to promise not to use nuclear weapons when there might be no better alternative to doing so.

U.S. declaratory policy since 2010 posits that the United States “would only consider the employment of nuclear weapons in extreme circumstances to defend the vital interests of the United States, its allies and partners.”2 This formulation—by not defining extreme circumstances or vital interests—does not adequately convey the importance of proportionality. As the Department of Defense Law of War Manual declares, “the overall goal of the State in resorting to war should not be outweighed by the harm that the war is expected to produce.”3

One policy alternative favored by many is no first use (NFU), in which the United States would pledge to never use nuclear weapons first in a conflict. However, some U.S. allies in Europe and in East Asia would perceive a declaration of NFU as a weakening of U.S. resolve to defend them. Meanwhile, Russia and China would not trust or rely on an NFU declaration if the United States did not remove or significantly reduce the nuclear and conventional forces and missile defenses that they perceive to be part of U.S. plans to preemptively strike their nuclear deterrents. Yet the political capital that a president would expend to instate NFU as a central policy in the face of objections from domestic opponents and key allied governments could leave little left to overcome traditional resistance to altering the offensive and defensive force posture (as we recommend).

Another alternative is to declare that the “sole purpose” of U.S. nuclear forces is to deter or defeat adversaries’ uses of nuclear weapons. This would be well advised if nuclear attack were the only adversarial threat that could not be defeated by non-nuclear means. However, if Russia or China were defeating U.S. and allied non-nuclear forces and threatening to inflict massive harm on their populations, then it would be imprudent to rule out proportionate use of nuclear weapons. It would be especially imprudent to do so if the United States, NATO, and U.S. allies and partners in Asia were not significantly improving their conventional military capabilities, the resilience of their military forces and societies, and their overall cooperation and cohesion.

Thus, we recommend that the United States adopt an existential threat policy (ETP), declaring that it would “use nuclear weapons only when no viable alternative exists to stop an existential attack against the United States, its allies, or partners.” No one knows whether and how the use of nuclear weapons against another nuclear-armed state would be kept limited and would not escalate. It would not make sense to use nuclear weapons unless the immediate threat was more dangerous than the likely consequences of nuclear war would be. The proportionality of an existential threat policy would uphold the United States’ commitment to comport with the law of armed conflict and demonstrate a more realistic appreciation of the risks and consequences of escalatory nuclear war.

Ambiguity is unavoidable in any declaratory policy, including the current “extreme circumstances” formulation. This review goes further than official U.S. or other states’ policy documents in discussing threats that could rightly justify nuclear use. We believe international debate over these issues is educational to all concerned and international pressure should be mobilized to push other governments, particularly Russia and China, to clarify whether and how national and international law will guide their conduct.


Employment policy directs how U.S. nuclear forces should be used in the event that deterrence fails and an adversary undertakes military action—most obviously nuclear attack—that cannot be stopped by non-nuclear means.

The primary challenge in today’s security environment is to disabuse Russian, Chinese, and North Korean leaders from thinking that they could severely threaten U.S. allies and partners and then deter or prevent the United States from deploying and using forces necessary to defeat them. In such contests, the United States needs to be able to deter or defeat adversary plans to use nuclear weapons in ways that would compel the United States to stop fighting and accept defeat.

One way to do this is to attack adversary nuclear forces before they can be used—preemptive damage limitation. During the early years of the Cold War, the United States and Russia could plausibly attempt to accomplish this only with nuclear weapons. Over time, both sides came to accept the reality of mutual assured destruction, though this realization did not cause them to stop preparing for counterforce strikes. Today, they also develop and deploy non-nuclear precision-strike weapons and perhaps cyber capabilities that could be employed for this purpose. Yet the quest for preemptive capabilities not only drives arms races and the procurement of excessive arsenals; it also increases pressures on adversaries to launch nuclear weapons preemptively or on warning of incoming attack. This increases the risks of mistaken or inadvertent nuclear use. In addition, large-scale nuclear counterforce attacks themselves could cause fires sufficient to produce the catastrophic worldwide climatic effects associated with nuclear winter, along with widespread radioactive fallout. Thus, the strategic imperative to prevent self-destructive escalation of war requires the pursuit of alternative force postures, policies, and plans.

The potential global destructiveness of nuclear war can be reduced by mutually lowering the number and explosive yields of U.S. and Russian nuclear weapons, especially silo-based ones that are the most feasible targets for preemptive damage-limiting strikes (as discussed in chapters 4 and 6). More immediately, the United States could abandon plans for preemptive strikes on Russian (and Chinese) nuclear forces, and instead focus U.S. nuclear attacks on targets necessary to deny Russia and China the prospect of winning a regional conflict or escaping unacceptable damage in a general nuclear war with the United States.

In all of this, the United States requires upgraded nuclear forces and command, control, and communication systems (NC3) that could survive adversary attacks and/or (partial) technical malfunction and still provide adequate confidence that presidential nuclear employment orders would be executed. The Defense Department has long prepared to give the president the option to launch within minutes the intercontinental ballistic missiles (ICBMs) that sit vulnerably in silos in the Midwest, so that a detected incoming attack would not destroy them. This practice is known as launch under attack (LUA). However, if the sensors and systems intended to detect a potential incoming attack on these land-based installations erred in their calculations, or inaccurately assessed the magnitude of the incoming attack, the United States would risk starting or escalating a nuclear war by mistake.

Several options exist for dealing with these challenges. The top priority, which must be pursued vigorously for many reasons, is to strengthen NC3 survivability. If U.S. leaders are confident in the survivability of submarine and air forces and command and control links to them, they could then exercise the option to more reliably assess a detected attack on vulnerable land-based forces before ordering U.S. counterattacks. This could mitigate risks of mistaken warning and assessments of incoming attacks.

To redress risks of launching vulnerable ICBMs on mistaken or false warning of a large, incoming attack, former vice chairman of the Joint Chiefs and NORAD commander James Winnefeld has suggested developing plans and capabilities to decide under attack (DUA). Under DUA, a president could transmit preplanned orders for U.S. strikes with a time delay on their execution.4 This delay would allow an authorized U.S. strike to be canceled or adjusted, and also would enable surviving forces (likely bombers and ballistic missile submarines [SSBNs]) to be positioned to carry out orders at the appropriate time. Unlike the immediate response programmed into a LUA scenario, if the detected attack were proven to be false or mistakenly assessed, under DUA the president could cancel or adjust the preplanned orders. To be sure, under wartime conditions there is no guarantee that the president or a successor would survive and be successful in canceling or modifying a delayed launch order. However, if LUA were the order of the day, there would no possibility of doing so.


U.S. and Russian nuclear forces remain much larger and more destructive than those of any other country. They are still excessively driven by Cold War notions of counterforce nuclear warfighting and arms racing, made more ominous now by the unraveling of arms control. The advent of precision-strike non-nuclear weapons further complicates deterrence and arms control diplomacy between these two countries and prospectively China. This review analyzes arguments for and against each of the main weapons systems in current and planned elements of the U.S. nuclear triad of air, sea, and land-deployed forces. This summary focuses on the four most controversial systems.

B-61 nuclear bomb. This nuclear bomb deployed in Europe is militarily unnecessary and will be even more superfluous if the Long-Range Standoff cruise missile is deployed, and/or the Low-Yield Trident D5 (LYD5) remains deployed. But until NATO requests this weapon’s removal, the political and deterrence consequences of withdrawing it to the continental United States would be more costly than the disarmament gain, unless Russia reciprocated in some meaningful way.

Low-Yield W76 warhead for Trident D5 submarine-launched ballistic missiles (SLBM). The United States recently replaced twenty 90-kiloton (kt) W76-1 warheads with 5–7 kt variants, called W76-2.5 (For comparative perspective, a 5–7 kt warhead is approximately ten times more powerful than various estimates of the yield of the chemical explosion that destroyed the port of Beirut in August 2020.6) The Defense Department switched warheads without adequate congressional briefing and debate that could answer the important questions explored in this chapter. Nevertheless, we do not recommend removing these warheads from service now if they would be replaced with their much higher-yield predecessors. Nuclear weapons should have yields no larger than necessary to destroy legitimate targets—both for legal and strategic reasons and for the purposes of reducing potential climatic effects of nuclear war.

Nuclear-armed sea-launched cruise missile (SLCM). Former president Donald Trump’s administration sought to develop this weapon to provide a “non-strategic regional presence, an assured response capability, and an INF-Treaty compliant response” to Russia’s violation of the now-defunct 1987 Intermediate-Range Nuclear Forces (INF) Treaty. However, this weapon could detract from the vital conventional war-fighting missions of the attack submarines that would carry it, especially in Northeast Asia. U.S. naval forces should retain their conventional focus and the nuclear-armed SLCM should be canceled.

Ground-Based Strategic Deterrent (GBSD) ICBM. Silo-based nuclear-armed ICBMs are the most vulnerable element of the U.S. strategic nuclear deterrent because their locations are fixed and well known. The United States has redressed the stability problem in part by limiting its ICBMs to carry only one warhead, rather than several, in order to require Russia (or any other adversary) to disadvantageously expend more than one weapon to target each U.S. ICBM warhead. Still, the vulnerability of silo-based ICBMs creates pressures on leaders to launch these weapons in the minutes before an incoming Russian attack could destroy them, with attendant risks as described above.

The Defense Department now proposes to spend an estimated $264 billion (in lifetime costs) to develop a new ICBM, the Ground-Based Strategic Deterrent, with the Air Force awarding an initial $13.3 billion contract to Northrup Grumman in October 2020. We recommend pausing this currently unnecessary program (with its assuredly underestimated costs). Instead, the United States should extend the lifetime of the current Minuteman force, which is feasible if their numbers are reduced and certain aging components undergo improvements. If efforts to negotiate bilateral strategic force reductions with Russia fail, then the United States could reconsider procuring a new ICBM.


U.S. missile defense policy and deployments should be considered in the context of deterring adversaries from escalating conflicts, reassuring allies, and avoiding destabilizing and excessively costly arms races.

U.S. missile defenses come in various forms, and have different capabilities, technical reliability, objectives, and costs. They also produce different reactions from allies and adversaries. Forward-deployed missile defenses play a role in regional deterrence. If they perform as intended, they will protect allies, U.S. forces, and critical military and civilian installations on allied territory. Missile defenses on and near U.S. territory are supposed to defend the U.S. homeland from ballistic missile attacks of the scale that North Korea might be able to launch. The key policy question is whether the deployment of such defenses can be done without provoking destabilizing arms racing and escalatory pressures with Russia and/or China that would leave the United States and its allies and partners less secure.

To date, the United States has insisted that regional defenses intended to deter or block North Korean and/or Iranian attacks do not pose threats to Russian and Chinese strategic nuclear deterrents. Moscow and Beijing, however, profess not to believe these statements. To the extent that Russia and/or China add strategic offensive capabilities to counter such defenses, would the benefits of defenses against regional Iranian and/or North Korean missiles outweigh the costs?

The United States also seeks increased capabilities to defend its regional forces and allies and partners from shorter-range Russian and Chinese conventionally armed and nuclear-tipped missiles. This is especially important in East Asia where China’s military power projection capabilities continue to grow. Yet it is easier and cheaper for China to add missiles of this range than it is for the United States to add defenses to feasibly counter them.

Another conundrum involves U.S. homeland defenses against ballistic missiles. Today, these defenses are scaled to defeat and thereby deter North Korean launches of nuclear weapons against U.S. territory. But if Chinese leaders genuinely perceive such defenses to threaten the viability of their second-strike long-range deterrent force of around 180 missiles after that force has been attacked by U.S. conventional or nuclear weapons, they would have an incentive to build up or hasten the launching of China’s nuclear force. Russia is already developing and deploying additional long-range nuclear delivery systems to defeat current and expected future U.S. missile defense capabilities.

Current U.S. missile defense policy will suffice if policymakers believe that an unconstrained competition in offensive and defensive weapons is preferable to potential agreements that would limit some elements of U.S. missile defenses in exchange for Russian and Chinese concessions. However, U.S. interests—and those of allies and the rest of the world—would be better served by exploring what possible trade-offs could be negotiated between transparency and potential limitations on some U.S. missile defense capabilities, on one hand, and Russian and Chinese reductions and/or constraints on some of their current and prospective offensive capabilities, on the other. The most promising way to assess these possibilities would be to explore whether and how regional and homeland missile defenses could be designed and deployed to protect against the missile threats posed by Iran and North Korea, without creating the realistic prospect of the United States successfully negating Russia’s and China’s deterrence of disarming first strikes, which would perpetuate arms racing.


Adversaries “pursue” arms control when they recognize mutual interests in reducing the costs and risks of competition in building and deploying weapons, especially those that exacerbate risks of inadvertent or accidental escalation. Arms reductions can also lower the level of damage that could be done if deterrence failed. Additionally, by improving predictability for years at a time, arms control helps participants reduce the costs and risks of arms racing and misperceptions about forces that can increase risk in crises.

The old arms control agreements that helped contain and end the Cold War were hard to make. The task of reinventing arms control in the twenty-first century will prove harder. There are new players—most importantly China—all of whom compete with and respond to one another. Escalation risks no longer come exclusively from familiar missiles but also from new technologies with multiple uses that are harder to count and monitor from afar. These new technologies, which may be more tempting to use, will be entangled with nuclear systems in ways that severely complicate the challenge of deterring conflict and its escalation to and through nuclear war.

Rather than be guided by deterrence logic alone, the organizing principle of arms control should be to reduce the probability of escalatory warfare and, with Russia and eventually China, to physically bound the potential collateral damage and long-term consequences that would occur if deterrence fails. No two antagonists should wield weapons whose number and explosive power could not only destroy their own nations but also have catastrophic effects on innocent bystander societies.

Deterrence theory posits that the United States should threaten to destroy enough of what adversary leaders value that they will choose not to take actions that could cause U.S. leaders to strike these targets. No one knows for certain what number of targets suffices to deter Russia and China; in any case, the United States should plan to use nuclear weapons only against targets that cannot be destroyed or disabled by non-nuclear means. The number of such targets likely would decrease depending on how many nuclear weapons Russia, and subsequently China, were willing to eliminate through negotiation. Moreover, the global security gain from reducing the probability that nuclear war would produce climatic catastrophe needs to be factored along with deterrence theory in deciding “how much is enough.” The overall risk of negotiating reductions to the minimal level Russia would accept—with parallel limits by China—is less than the risks in both countries’ retaining larger arsenals. By making an offer to pursue such reductions, the United States would benefit in international politics by shifting the burden of debate on nuclear arms control and disarmament to Russia.

Legally binding treaties are unlikely to be the only modality for arms control in the foreseeable future. Beyond the complicating need to account for new technologies, political dynamics in Washington make it difficult to ratify treaties. China’s skepticism about American intentions in arms control forestalls even the beginning of treaty negotiations. Absent major political change in Washington and Beijing, sustained strategic dialogues, executive agreements, and reciprocal confidence-building measures all will be required to make progress on arms control.


For the United States and Russia, the most feasible way to serve mutual interests in arms control is to extend New START (Strategic Arms Reduction Treaty), which is set to expire in February 2021. For all its criticisms of what New START does not cover, the Trump administration failed to persuasively explain how the United States would be better off without it.

Assuming New START is extended, the next shared objectives for the United States and Russia should be

  • broad-based discussions of strategic stability and escalation risks;
  • negotiation of a follow-on agreement to New START; and
  • negotiation of non–legally binding confidence- and security-building measures (see three options discussed in this chapter).


For years, Chinese leaders have resisted arms-control-inflected dialogue, let alone negotiations. Beyond political, bureaucratic, and perhaps cultural factors, a central problem is that Chinese leaders doubt that the United States is prepared to accept and articulate that its strategic competition with China is predicated on mutual vulnerability to nuclear retaliation, which cannot be escaped through first strikes or missile defenses. The United States should acknowledge this fact and test whether Beijing will reciprocate by engaging in sustained, productive strategic dialogue that generates greater transparency regarding China’s future development and deployment of nuclear forces and other capabilities that could challenge strategic stability.

U.S. allies and partners will fret that clarifying a relationship of mutual U.S.-China vulnerability would weaken deterrence and increase Chinese assertiveness. Yet the benefits of drawing China into a process that could lead to mutually beneficial transparency, confidence-building, and eventually arms control are greater than allowing current trends to continue. If attempted engagement with China fails, then the United States and its allies and partners will be on firmer political ground to bolster the range of capabilities needed to counter Chinese coercion.

The following three topics are potential foundations for dialogue on U.S.-China stability:

  • exploring the feasibility of bilateral or regional limits on aggregate numbers of missile delivery systems with ranges greater than 500 kilometers;
  • demarcating regional missile defenses; and
  • exploring risks of cyber operations against nuclear command and control systems.


As a co-creator and longtime champion of the Treaty on the Non-Proliferation of Nuclear Weapons (NPT), the United States (along with other nuclear-weapon states) must take seriously its obligation to pursue good-faith negotiations on effective measures relating to the cessation of the nuclear arms race and to nuclear disarmament.

To demonstrate its commitment to the disarmament principles expressed in Article VI of the NPT, the United States should design a prototype nuclear disarmament regime that would encompass all states and should invite international discussion and debate over it. Designing effective and sustainable nuclear disarmament of any nuclear-armed state requires much more than dismantling warheads and controlling fissile material stocks. Specialists from all relevant U.S. government agencies should contribute to this effort. Having demonstrated its thinking on the potential requirements for implementing and enforcing verifiable nuclear disarmament, the United States should invite all other nuclear-armed states to do the same if they do not concur with the U.S. model.

Finally, the new administration of Joe Biden and Kamala Harris should commission the National Academy of Sciences to evaluate extant studies on the possible climatic effects of nuclear war. The national security imperative here is to enable U.S. decisionmakers and citizens to better understand the potential consequences of nuclear weapons use, as well as the likelihood that reductions in weapon numbers and yields, and changes in target selection, would reduce prospects of unnecessary suffering if deterrence fails. The United States should call on other nuclear-armed states to conduct and publish similar studies or critiques of the U.S. study.


1 If other states violate their obligations under the Treaty on the Non-Proliferation of Nuclear Weapons (Non-Proliferation Treaty; NPT), they could then become objects of U.S. threats to use nuclear weapons to defend itself and its allies and partners, as noted in the 2010 and 2018 NPRs.

2 U.S. Department of Defense, Nuclear Posture Review (Washington, DC: U.S. Department of Defense, 2010), 17,

Courtesy: (

Posted in