Table of Contents

As the 2018 NPR affirmed, U.S. nuclear policymakers can and should identify policies that could help reduce instabilities and the potential damage of a catastrophic war with Russia and China. Such policies could be adopted unilaterally through nuclear force posture changes, or—as arms control—through bilateral or multilateral negotiations.

Adversaries “pursue” arms control when they recognize mutual interests in reducing the costs and risks of destabilizing 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. By improving predictability for years at a time, arms control also helps participants manage national budgets and defense planning.

The old arms control agreements that helped manage and end the Cold War were hard to make. The task of reinventing arms control in the twenty-first century will prove harder. For one thing, there are new players—China the most important among them, from the U.S. perspective. China’s perceived military requirements derive not only from competition with the United States, American allies, and Russia; it factors India into its calculations, too. China’s capabilities then weigh heavily on India’s perceived requirements for deterrence, which Pakistan in turn seeks to match or surpass. All these states compete with one another in various ways.

New escalation risks also are shaping the arms control landscape. Whereas earlier nuclear arms control focused on nuclear-capable ballistic and cruise missiles, delivery systems today and in the future carry both conventional and nuclear warheads and move with varying speeds and trajectories from multiple launch platforms. These and other weapons that are harder to monitor, including electronic and cyber variants, may threaten targets ranging from civilian infrastructure and populations to early warning and command and control systems, as well as nuclear and conventional forces. These new weapons may be more tempting to use in ways that could be entangled with nuclear systems, which severely complicates the challenge of deterring conflict and its escalation. Negotiating acceptable and stabilizing balances among such protean weapons and their potential uses will be extremely difficult.

Rather than being guided by deterrence logic alone, the organizing principles and goals of arms control should be to reduce the probability of escalatory warfare and to physically bound the potential damage that would occur if nuclear deterrence fails.

Rather than being guided by deterrence logic alone, the organizing principles and goals of arms control should be to reduce the probability of escalatory warfare and to physically bound the potential damage that would occur if nuclear deterrence fails. No two antagonists should wield weapons whose number and explosive power could not only destroy their own nations but also cast innocent bystander societies into catastrophe.

Deterrence theory posits that the United States should hold at risk enough of what adversary leaders value that they will decide not to undertake actions that would cause U.S. leaders to strike these targets. There can be no certainty as to what type and number of targets suffice 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 would decrease to some degree 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 environmental catastrophe needs to be considered 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 arguably less than the risks of both countries’ retaining arsenals larger than these minimums. By offering to match Russia’s reductions—with corresponding adjustments by China—the United States would benefit in international politics by shifting the burden of debate on nuclear arms control and disarmament to Russia.

Some policymakers may assert that much less destructive arsenals could fail to deter Russian aggression. However, any nuclear war, beyond very limited attacks on remote targets, would portend devastation of a scale and pace that would be unprecedentedly catastrophic. The United States is now deterred from initiating conflict with North Korea, a country with perhaps a few dozen nuclear weapons. There is no reason to think that the United States and Russia would no longer be mutually deterred if they each had “only” a few hundred survivable nuclear weapons and much less prompt launch postures. The challenge in this scenario is to determine what combination of non-nuclear and nuclear capabilities and arms control constraints would make both sides (and China) confident that their deterrents were survivable.

Treaties cannot be expected to address all these dynamics in any foreseeable future. Political dynamics in Washington make it exceedingly difficult to ratify treaties. China’s skepticism about arms control forestalls even the beginning of formal negotiations with the United States. Additional forms of agreements and confidence-building measures will need to be created. The most promising formats for doing this will be in bilateral dialogues between the United States and Russia and the United States and China. China and Russia are now averse to trilateral negotiations. It is difficult to imagine them changing their positions unless they were going to band together against U.S. preferences. (Bringing the United Kingdom and France into the process could rebalance political dynamics, but attempting to do so in the near term would add a host of other complications.)

Progress in any of these dimensions of arms will require adjustments in everyone’s approaches. Most fundamentally, Americans will need to recognize that Russia and China will not accept agreements that they perceive to be unfair to them. This may be obvious: U.S. leaders and voters would not endorse agreements that are unfair toward the United States. Yet ever since the end of the Cold War, the United States as the world’s most powerful state has frequently sought one-sided advantages. This should not be expected to work with Russia and China in the future.

Why U.S.-Russia Arms Control Still Matters

The Trump administration argued that bilateral U.S.-Russian arms control is outdated, and it prioritized negotiating a trilateral treaty that limits China’s nuclear forces alongside those of the United States and Russia. Engaging China is a worthwhile goal that deserves real diplomatic efforts, as discussed below. It would, however, be a significant mistake to allow Beijing to veto U.S.-Russian arms control efforts. Bilateral arms control with Russia remains an important tool for enhancing the security of the United States and its allies. The United States should pursue it alongside efforts to engage China.

The United States and Russia size and posture their nuclear forces to compete against each other. Without arms control, this interdependence could intensify an expensive and dangerous competition in strategic forces—a particularly acute risk, given Russia’s development of new “exotic” nuclear weapons. Russia’s opaque nonstrategic nuclear forces are of particular concern to the United States and NATO. Russia has many such weapons, deploys them close to the territories of NATO’s easternmost members, and maintains options for using them early in a conflict. The United States is moving to counter Russian capabilities with new air- and sea-based capabilities of its own. Arms control could provide much-needed transparency in this domain and eventually help to cap and roll back the emerging nonstrategic nuclear forces competition.

Basic Strategy

If the United States wants Russia to deal with U.S. concerns, it must be prepared to discuss and help address Russian concerns—in particular, about the survivability of its nuclear forces. These concerns have grown as a result of advances in non-nuclear weaponry, including high-precision conventional weapons and ballistic missile defenses, as well as projected modernization of U.S. nuclear forces.

Before addressing concerns about future arms racing and instability, the United States and Russia will need to deal with each other’s deep dissatisfaction over past performance in upholding earlier arms control treaties and international agreements. The United States, with evidence and reason, emphasizes Russia’s violations of the INF and Conventional Forces in Europe treaties, as well as problems with Russian practices related to the Open Skies Treaty. It also points to Russia’s recent violations of Ukrainian sovereignty and territorial integrity. Russia, for its part, emphasizes the U.S. withdrawal from the ABM Treaty, the INF Treaty, and the Joint Comprehensive Plan of Action with Iran, as well as the U.S. decision to withdraw from Open Skies. The result is that many Americans believe Russia will violate arms control agreements while Russian leaders believe the United States will withdraw from them when an administration believes it is unilaterally advantageous to do so. The moral and legal “superiority” of withdrawing from agreements compared to violating them is clearer to Americans than it is to many others, including U.S. allies that rely on the same accords for their security.

The United States and Russia will be unlikely to undertake new agreements if they do not explicitly agree on measures to reassure each other that their pattern of unilateral noncompliance and withdrawal will not be repeated. Updating verification measures in new agreements is one way to address these concerns.

This report is not the place to suggest detailed terms that negotiators on either or both sides should or likely would pursue in negotiations That said, an obvious and necessary first step would be to extend New START for up to five years, which Russia has offered to do and with which the United States should agree. Nothing would be gained by refusing this step. Without it, any further progress will be more difficult to achieve.

Assuming New START is extended, the next shared U.S.-Russia objectives should be:

  • Broad-based discussions of strategic stability and escalation risks
  • Negotiation of a follow-on agreement to New START
  • Negotiation of certain non–legally binding confidence- and security-building measures
  • Reciprocal inspections of empty nonstrategic warhead facilities
  • NATO commitment to refrain from modifying Aegis Ashore missile launchers deployed in Europe to contain offensive missiles
  • U.S. transparency regarding the technical capabilities of SM-3 interceptors based in Europe to demonstrate inability to intercept Russian ICBMs

Discussions of Strategic Stability

Strategic stability discussions help build security. Understanding an adversary’s concerns and redlines could help restrain future conflicts and reduce risks of escalation if conflict does occur. Stability discussions also can map a path toward concrete arms control measures.

Recently, each government has issued nuclear policy documents whose meanings and implications have been interpreted in various, often divergent ways.1 Sustained dialogue would provide opportunities for each side to better understand the other or to draw more informed conclusions about the sources of nonunderstanding. For example, diplomats and military officials from both countries could describe and explore the risks that they think would follow from limited nuclear use or launching nonnuclear attacks on dual-use command and control systems. Dialogues on these topics, which the Trump administration emphasized, should be continued.

Negotiation of a New START Follow-on

Even if New START is extended, the United States and Russia will be pressed for time to negotiate a successor agreement. It is difficult to imagine that they will be doing so in a decidedly better political relationship. Therefore, the most feasible objectives would be to:

Use the framework and verification approach of New START to further reduce deployed strategic delivery systems, launchers, and warheads, and to apply limits to new technologies that are strategically significant but not technically covered by New START. The aim should be to enhance stability and, as much as possible, lower the scale of global catastrophe if deterrence fails.

Specifically, the agenda for follow-on negotiations should address Russia’s intercontinental ground-launched boost-glide weapons, nuclear-powered torpedoes, nuclear-powered cruise missiles, and air-launched ballistic and boost-glide missiles.2 The United States’ air- and sea-launched boost-glide missiles now under development also would need to be included. In addition to limiting U.S. ballistic missile defenses, Russian negotiators obviously will posit additional priorities. These would likely include more specifically limiting the number of nuclear warheads that covered aircraft may deploy. (Under New START counting rules, each bomber plane counts as one warhead against the overall limit of 1,550, although each bomber may carry 6 to 20 nuclear weapons when deployed.)3 Russia also would at the very least demand that the United States redress its concerns over SLBM and B-52H conversions.4

Strategists often argue that targeting and operational objectives should be decisive in determining numerical limits for weapons, and that lowering numbers for their own sake is astrategic. Yet somehow, for decades, the target-based logic produced overkill in the quantity of deployed nuclear weapons, and successive new administrations repeatedly sought to correct this issue.5 Indeed, throughout much of the Cold War, military planners searched for targets to match the growing number of weapons on hand or in the pipeline. In this sense, the number of available weapons set a “budget” for targeting. If the number and explosive yields of weapons that that United States and Russia wield today could produce global climatic (and fallout) catastrophe in an all-out nuclear war, the strategic and legal case can be made to set a “destructiveness budget” to limit the number and yields of their arsenals. The numbers of weapons that the United States and Russia possess and deploy in toto today belie claims that targeting “requirements” are so precise. All treaties to date adopt limits in increments of fifties of weapons, not tens or ones.6

In 2013, the Pentagon concluded that the United States could fulfill its strategic deterrent requirements with 1,000 weapons under New START counting rules. In other words, the United States could unilaterally reduce by more than 500 deployed strategic warheads below New START warhead limits. The Obama administration made a political judgment not to pursue this course, and instead to reduce further only with reciprocity from Russia.7 For their part, unofficial Russian experts also have spoken and written of making a deployed warhead limit of 1,000 an objective for a follow-on agreement to New START.8

One reason for making 1,000 an objective is that it would still leave the United States and Russia with strategic weapons numbering in four digits, signifying their strategic superiority compared with the three digits of the next largest nuclear arsenals, France (300) and China (low 200s and growing). The attraction of this symbolism is understandable from several perspectives, but it is a liability from others. The rest of the world, particularly most of the 185 non-nuclear-weapon states under the NPT, are so frustrated with the lack of progress in nuclear disarmament that they support the 2017 UN Treaty on the Prohibition of Nuclear Weapons. To the extent that the nuclear nonproliferation regime remains important for international security, keeping these states invested in that regime is an important objective. A U.S.-Russia agreement that brought both countries’ arsenals below the 1,000-warhead threshold could be useful in this regard. It is ridiculous to argue that either the United States or Russia would be less secure with, say, 999 operationally deployed strategic weapons than they would be with 1,000.

Whatever number negotiators would seek, another key strategic objective should be to reduce the weapons that are the greatest source of instability and environmental damage in a potentially escalatory conflict. Those are silo-based ICBMs, especially ones with multiple warheads. (See chapter 4 for the problems associated with these weapons.)

Whatever number negotiators would seek, another key strategic objective should be to reduce the weapons that are the greatest source of instability and environmental damage in a potentially escalatory conflict.

There are roadblocks to negotiating the eventual elimination of silo-based ICBMs. Russia continues to depend on multiple-warhead silo-based ICBMs as a cost-efficient way to deploy large numbers of warheads.9 The warheads on these large ICBMs ostensibly pose the greatest risk to the geographically vast U.S. ICBM launch facilities. As U.S. missile defenses advance, the Kremlin places more importance on this objective.10 Even as Russia has shifted more of its deployed strategic nuclear forces to mobile ICBMs over the past ten years, it is replacing the aging SS-18 heavy-silo ICBM with the even larger Sarmat, which will be deployed in the 2020s. The Sarmat and a variant of the SS-19 silo-ICBM also serve as planned delivery systems for the Avangard boost-glide vehicle. Russia chose to advance the Avangard-carrying missiles over certain mobile ICBM systems in its ten-year armament plan.11 This decision may reflect a priority on countering U.S. homeland missile defenses. In that case, Avangard—if retained while Russia reduces other multiple warhead ICBMs—could help reassure Russia that U.S. homeland missile defenses will not negate Russia’s deterrent.

Both countries’ armed forces and military-industrial complexes are attached to silo-based systems.12 Nevertheless, the United States and Russia also have a shared interest in reducing them. Each side’s silo-based ICBM force largely justifies the modernization of the other’s. The strategic interactions between the two create incentives to prepare for and possibly use ICBMs in preemptive strikes. This dynamic, while meant to strengthen deterrence, can also weaken crisis stability and create opportunities for inadvertent escalation. Stability could be enhanced instead by increasing the ratio of highly survivable delivery systems in each country’s nuclear force and discarding less survivable systems, recognizing that survivability requires a weapon to be able to penetrate adversary defenses.

The timing is right within each country’s modernization programs to halt or limit new (or replacement) silo-ICBM deployments. Each country’s domestic military budgeting, development, and operational planning for future silo-ICBMs is not finished. Russia’s Sarmat is nearing the end of its development cycle, but not deployed yet. The GBSD is still on the drawing board. The development of each missile will advance rapidly in the next two to three years. After deployment, Washington and Moscow (and their respective silo-ICBM stakeholders) will be less inclined to eliminate the new missiles. Before deployment, GBSD and Sarmat represent good starting points for the traditional horse-trading that accompanies preparation for arms control negotiations. An agreement to not deploy, or to reduce newly deployed silo-ICBMs (by replacing older ones on a less than one-to-one basis), could be an early confidence-building measure by the parties while the details of a New START follow-on are jointly pursued.13

Finally, the United States and Russia still share an interest in limiting MIRVed silo-based ICBMs. Both countries committed to these limitations in order to enhance strategic stability under the START II Treaty, even though the treaty never entered into force. This goal is still important. Reducing silo-based ICBMs would reduce the most destabilizing type of MIRVs—those that are fixed and therefore prone to strategic instability. Although Russia’s interest in MIRVs remains strong—and is well explained by Russian Ministry of Foreign Affairs official Vladimir Leontiev—the possible deployment of a new U.S. ICBM and other modern delivery systems could create incentives to negotiate.14

The specifics of an agreement which would reduce silo-ICBMs obviously must be left to arms control negotiators with access to confidential information on U.S. and Russian force structures. The U.S.-Russian proclivity to seek numerical parity across kinds of delivery systems as well as overall warheads will require creative force-structure planning and negotiating. If either or both governments move beyond demands for parity of numbers and focus instead on stability, more options for agreement would arise.

Nonbinding Confidence- and Security-Building Measures

The United States and NATO are deeply concerned by Russia’s large and technically improving stockpile of NSNW. Russia is agitated by the possibility that the United States could and would modify and use Aegis Ashore launchers to conduct offensive missile strikes against Russia, notwithstanding these launchers’ stated purpose of defending Europe from missile threats emerging from the Middle East (for instance, Iran). Russia also is alarmed by the possibility of the United States deploying B61 nuclear bombs in Poland.

Treaties are unlikely to address these concerns, for both technical reasons and the general political difficulty of the ratification process in Washington. This does not mean that nothing can be done. Transparency and other confidence-building measures can be imagined that would redress the concerns of the United States, NATO, and Russia. Given that each side has different concerns regarding different weapons systems, the most feasible way to find mutual satisfaction would be to negotiate a basket of arrangements that would in total and on balance make everyone better off than they are today. Three possible examples of such arrangements are worth a closer examination.

The United States and Russia should consider reciprocal inspections of empty nonstrategic warhead facilities.

To date, the United States and Russia have made little progress in increasing mutual transparency regarding NSNW. An intrusive agreement to inspect active warhead storage facilities is not now politically feasible. However, inspecting empty formerly active warhead storage facilities on NATO and Russian territory would serve several important purposes.

First, it would demonstrate that an inspection regime for NSNW is feasible. Each country would gain valuable information on the types of NSNW storage practices and facilities that the other side possesses.

Second, such inspections could reduce fears that either side has secretly located nondeployed breakout potential in nonstrategic warheads. NATO suspicions that nuclear warheads may be stored in Kaliningrad—creating near nuclear-capable missile forces that could strike Alliance territory with little warning—could increase incentives to strike Russian forces in the enclave early in a conflict. If NATO were able to confirm that Kaliningrad storage facilities did not have these types of warheads, the knowledge could attenuate its worst-case scenario planning for a regional nuclear war. A reciprocal arrangement regarding storage facilities in NATO states could similarly reduce Russian incentives to strike first.

Third, negotiators could adapt an inspection regime for empty warhead facilities into an agreement for inspecting active storage facilities, rather than attempting to negotiate a comprehensive active-warhead storage facility inspection regime from scratch.15 Such an arrangement could help develop additional transparency measures regarding Russia’s NSNW in Europe, which are of particular concern to some NATO members.

NATO should issue a public commitment to refrain from modifying Aegis Ashore launchers deployed in Deveselu, Romania, to contain offensive missiles, such as the Tomahawk land attack cruise missile.

Following U.S. and Russian termination of the INF Treaty and subsequent U.S. land-based missile tests, Russia has concerns that Aegis Ashore could be modified to launch nuclear-armed offensive missiles. The United States regularly notes that loading offensive missiles into stationary, easy-to-target Aegis Ashore launchers would be foolish from an operational perspective; however, it has not forsworn the possibility. Yet the United States did confirm in its bilateral agreement with Romania that the Aegis Ashore site can be used only for SM-3 interceptors for defensive purposes.16 To assuage Russian concerns, the United States, Romania, and NATO could issue a joint statement that the sites will only be used for defensive purposes, and make a commitment to refrain from bringing offensive missiles to the site.

Following the political commitment, NATO should provide transparency into the technical capabilities of SM-3 interceptors based in Europe, to confirm they are technically unable to intercept Russian ICBMs.

U.S. missile defense experts continue to argue that SM-3 interceptors do not have a technical capability to intercept Russian ICBMs from Europe. Currently, the SM-3 Block IB interceptors based in Romania have a burnout speed of 3 kilometers per second (kps) and a range of a few hundred kilometers. The SM-3 Block IIA, when deployed, may have a higher burnout speed of 4.5 kps with a range of thousands of kilometers. The United States tested the new missile against an ICBM target in November 2020. This test reinforced Russian suspicions of U.S. intentions regarding missile defense.

For the reasons discussed in chapter 5, increasing confidence that U.S. missile defenses cannot negate Russia’s nuclear deterrent is a prudent step to lessen bilateral arms racing and first-strike pressures. In 2011, the United States offered to allow Russia to observe the test of an SM-3 Block IB interceptor and use its own equipment to measure that interceptor’s burnout speed. Although Russia declined then, the United States could revisit this proposal for Block IB and, prior to their deployment, Block IIA interceptors. Initiating the proposal is important for European political reasons regardless whether Russia accepts the offer. The United States should share its deployment plans and schedule for new SM-3 missiles on land and at sea as part of a new strategic arms control negotiation and explore additional transparency steps.17

Bringing China into Arms Control

American and allied interests would be served by meaningful U.S.-China security dialogue. Such dialogue could explore how confidence-building measures and arms control could help avoid a destabilizing and costly open-ended arms competition that could significantly increase catastrophic consequences if deterrence fails. Serious, sustained dialogue leading up to potential confidence-building measures and arms control could lessen the chances that skirmishes could escalate into conflict and armed conflict could escalate to nuclear war.

The United States and its allies and partners also share interests in encouraging China to affirm in words and deeds its long-standing comparatively restrained approach to nuclear weaponry and potential use. For the past five decades, China has maintained a relatively small nuclear stockpile and has committed not to use nuclear weapons first. Furthermore, Chinese eschewal of nuclear arms racing (if credibly continued) would reduce the imperative India feels to build up its nuclear and missile-defense capabilities. This, in turn, should reduce Pakistan’s incentives to augment its nuclear capabilities against India.

China has some understandable reasons for being reluctant to engage the United States in dialogue on strategic stability and (potentially) confidence-building and arms control. Chinese leaders note that the U.S. and Russian nuclear arsenals each are roughly 10 to 15 times larger than China’s, depending on the means of calculation. For years, Beijing has said that it would consider joining nuclear arms control only when the United States and Russia have reduced their nuclear arsenals to levels much closer to its own.18 Washington and Moscow have categorically refused even to discuss this prospect.

The United States has affirmed that “it is not our intent to negate Russia’s strategic nuclear deterrent, or to destabilize the strategic military relationship with Russia.” Chinese analysts note that Washington has not articulated a similar policy toward China.19 Instead, they perceive that the United States maintains a preemptive counterforce damage limitation strategy, based on expansive conventional and nuclear weapons and missile defense capabilities, to target China’s smaller nuclear arsenal. The 2019 Missile Defense Review states that “The United States relies on nuclear deterrence to address the large and more sophisticated Russian and Chinese intercontinental ballistic missile capabilities,” but Chinese analysts retort that the 2018 NPR and other statements by U.S. officials do not indicate that U.S. policy is based on mutual vulnerability.20 How would arms control serve any state’s interests if the counterpart pursues it to maintain or seek unilateral advantage instead of stabilizing mutual deterrence and mitigating the potential for conflict escalation?

In order to motivate China even to explore the possibilities of arms control discussions, Washington must do something that no prior administration has yet publicly done: demonstratewillingnesstoaddressChina’s concerns about certain U.S. offensive and defensive military capabilities and intentions that may be inferred from them.

U.S. officials explain privately that the United States has not publicly acknowledged that mutual nuclear vulnerability is inescapable with China because doing so would complicate the United States’ security reassurance of Japan. This consideration is understandable but it does not address the destabilizing consequences with regard to China, which are not in the interest of the United States, Japan, and other allies and partners. Publicly basing policy toward China on mutual nuclear vulnerability, rather than preemptive nuclear counterforce, could motivate Japan and other U.S. allies and partners to join the United States in enhancing non-nuclear capabilities to deter or defeat Chinese offensive military operations. However, domestic politics in Washington impede this prospect. Officials from prior Democratic administrations note privately that explicitly basing nuclear policy toward China on mutual vulnerability would elicit such condemnation from Republicans in Congress that the gains of doing so would not be worth the political cost. Yet this political calculus merely reinforces Chinese skepticism that the United States would be willing to pursue arms control or other equitable arrangements.

Chinese officials and analysts are also chary of the intrusive verification requirements that the United States (and Russia) traditionally demand with arms control. In their view, the United States would use transparency and verification measures to enhance targeting of China’s smaller arsenal. As the weaker party—one whose government tightly controls information—China feels that opacity enhances the survivability of its deterrent. This tendency may be exacerbated by the lack of institutional knowledge and experience with verification within the Chinese government, especially at the highest levels. Unlike the United States and Russia, China lacks large cadres of diplomats, military officers, and scientists deeply versed in arms control and verification.

For these and other reasons, it will be extremely difficult to draw China into sustained dialogue and, eventually, negotiations to verifiably limit the growth of its nuclear and other relevant military capabilities. However, as China modernizes and grows its arsenal, it will have less basis for claiming that its forces are too comparatively imbalanced to allow a fair negotiation.

In order to motivate China even to explore the possibilities of arms control discussions, Washington must do something that no prior administration has yet publicly done: demonstrate willingness to address China’s concerns about certain U.S. offensive and defensive military capabilities and intentions that may be inferred from them. Willingness to discuss these issues, informed by sustained consultation with U.S. allies and partners, need not imply commitments to limit or reduce them. However, refusal to address them will confirm Chinese leaders’ views that dialogue and arms control are ruses to perpetuate U.S. military superiority.

The United States’ willingness to limit its military competition with China, in turn, will depend heavily on China’s willingness to demonstrate its understanding that strategic stability requires not using force or physical actions to change the territorial status quo, which could lead to crises with escalatory potential. The United States could facilitate a strategic stability conversation with China that includes these factors along with a discussion of military capabilities. If China is not prepared to reassure the United States and others—particularly Taiwan—of its willingness to refrain from territorial expansion in deeds and not merely words, then the United States should be expected not to constrain deployments of additional non-nuclear capabilities to defend its allies and U.S. forces in Asia.

To reverse the recent downward spiraling in U.S.-China relations, the two governments might begin by encouraging track 1.5 dialogues on these issues. Such dialogues have occurred for years, but often lacked imprimaturs and directives from the highest civilian and military leaderships. As a result, they generally have not led anywhere. To be productive, civilian and military leaders at the highest levels must encourage such dialogue, meet with participants from their own country to discuss and establish objectives, and then ask to be briefed on results. This would entail little cost or risk to either side.

The following three topics of substantive value would enable American and Chinese counterparts to assess prospects for further beneficial work.

Limit aggregate numbers of launchers for delivery systems with ranges greater than 500 kilometers.

Seeking mutually acceptable balances in nuclear forces alone—delivery systems and warheads—would work only if both the United States and Russia considered reducing to levels close to China’s (or, conversely, would agree to “allow” China to build up close to theirs). Neither scenario is likely, though U.S. and Russian agreement to further reduce their deployed strategic warheads to 1,000 or less may provide incentive to China to negotiate in other areas.

More realistically, the United States, Russia, and China (bilaterally or trilaterally) could explore how to aggregate heavy bombers and launchers for missiles with ranges greater than 500 kilometers into a common “basket,” whether they carry nuclear or conventional warheads. On that more equitable basis, they could negotiate numerical limits. But even if this innovation were accepted, asymmetries in numbers of land-based versus sea- and air-based capabilities, and attendant verification challenges, make this an exceptionally complicated prospect. Missile defenses—especially those of the United States and its allies—add to the challenge.

However, with the demise of the INF Treaty, the potential for an intermediate-range missile race in the Asia-Pacific regions needs to be addressed. Some U.S. and allied experts believe that China has already been “racing” unilaterally. It was never a party to the INF Treaty and therefore has been free to produce and deploy land-based missiles with ranges between 500 and 5,500 kilometers, which the United States—until recently—was prohibited from doing.21 Ever since the INF Treaty was terminated in 2019, the United States has conducted flight tests for two new land-based missile systems. China recently debuted an intermediate-range hypersonic boost-glide system that may be dual-capable.22 Russia, for its part, has developed the SSC-8 ground-launched cruise missile and is deploying the dual-capable Kinzhal ALBM.23

Though the potential increased deployment of dual-capable (nuclear and conventionally armed) missiles of this range is alarming, it also could create a basis for future limitations. Unlike the huge disparities in both countries’ long-range missile arsenals, their numbers of land-, air-, and sea-based missile launchers (including heavy bombers) with ranges greater than 500 kilometers are roughly similar.24 This rough parity could create an equitable political and a military basis for negotiations, which could take into account systems that otherwise would be accountable under New START and the now defunct INF Treaty.

Heuristically, an agreement could establish a single ceiling for U.S., Russian, and Chinese submarine-based ballistic missile launchers, land-based silo and mobile launchers for ballistic and cruise missiles, and heavy bombers based on their definitions in the INF and New START treaties.25 Each country could be permitted the “freedom to mix” nuclear and conventionally armed missiles on delivery platforms as necessary to satisfy military requirements, so long as the number of total launchers and heavy bombers remains under numerical ceilings. Such an agreement could help curtail an accelerating area of competition and create a basis for further multilateral arms reductions. Even if the complexity of the challenge, especially regarding verifiability, proved insurmountable to negotiators, the process of exploring these issues could point to other ways to mitigate instability.

Demarcate regional missile defenses.

The United States says that its interest in missile defense is primarily to defend against North Korean nuclear-armed missiles that could threaten U.S. and allied bases and forces in the Asia-Pacific region as well as U.S. and allied homeland population centers. China has little interest in North Korea being able to wage nuclear war in the region. Washington and Beijing, then, could explore the desirability and feasibility of conducting a joint technical study of a potential U.S. missile defense system that could defend against North Korean missiles but would not undermine China’s second-strike nuclear deterrent by intercepting its long-range missiles.26 As with Russia, albeit on a smaller scale, China’s development of MIRVed and maneuverable boost-glide intercontinental delivery systems to evade U.S. homeland defenses could ameliorate China’s concerns, even as they aggravate American worries. In such discussions, all systems can be on the agenda.

Understand risks of cyber operations against nuclear command and control systems.

A third topic for U.S.-Chinese dialogue could be both countries’ concerns over potential cyber threats to their NC3 systems. The authors have devoted more attention to this exceedingly complex and sensitive topic elsewhere.27 Defense leaders in both countries would benefit from sharing views on which types of cyber operations and countermeasures they would find particularly escalatory. Both sides also could benefit from exchanges of information on their internal processes for overseeing cyber capabilities. Cyber operators may not know as much about nuclear dynamics as they should, and political leaders may not know as much about cyber dynamics as they should. Discussing how militaries may plan to respond to cyber threats to one another’s NC3 systems can help foster stability.

Neither government’s leadership would lose much, if anything, if dialogue on any of these topics (or others) indicated little progress toward developing an agenda for productive further work. Conversely, if the United States and China could agree on an agenda for useful further work, experts in and out of both governments could begin to develop problem-solving approaches that have been absent heretofore. Debate and refinement of such approaches could then prepare the ground for official dialogue and confidence-building measures when broader political conditions allow. There does not appear to be another way to bring China into arms control.28

Advancing the Broader Nuclear Disarmament Agenda

As a co-creator and longtime champion of the 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. Remarkably, neither the United States nor any other nuclear-armed state has bothered even to model how they would define nuclear disarmament of their state and others, and how they would expect to verify and enforce such disarmament. Effective, sustainable nuclear disarmament of any nuclear-armed state requires much more than dismantling warheads and controlling fissile material stocks.29 Disarming states would need to agree (presumably with each other) what types and numbers of delivery systems—especially missiles—would be permitted or forbidden, and how monitoring to ensure compliance with such terms would not encroach on monitored states’ legitimate security interests. Both nuclear-armed and non-nuclear-weapon states would need to determine what peaceful nuclear or space activities would remain during and after nuclear disarmament, and under what reassurance and monitoring conditions. At least some states likely will press for monitorable limits on research and development activities that could be vital to reconstituting nuclear arsenals.

To demonstrate that it takes Article VI of the NPT seriously, the United States should design a prototype nuclear disarmament regime that would encompass all states and invite international discussion and debate.30 Specialists from all relevant agencies in the U.S. government should contribute to this effort—most obviously, the Department of Defense, the Department of Energy and the national laboratories, the Department of State, and the intelligence community. The United States should invite all other nuclear-armed states to do the same if they do not concur with the U.S. model.

The United States should also continue the multilateral dialogue initiative, Creating the Environment for Nuclear Disarmament, and the International Partnership for Nuclear Disarmament Verification. Both multilateral forums provide useful dialogue opportunities for nuclear-weapon states, umbrella states, nonaligned states, NPT parties, and non-NPT parties. These discussions are productive and substantive, and are designed to find solutions to security and technical challenges to nuclear disarmament.

Finally, the next administration should commission the National Academy of Sciences (including its Committee on International Security and Arms Control) to evaluate extant studies on the possible climatic effects of nuclear war. Some studies dating from the early 1980s through 2019 have concluded that nuclear war involving the strategic forces of the United States and Russia would likely produce fires that would loft smoke into the stratosphere sufficient to cause “nuclear winter,” with devastating consequences for food production in the United States and globally. Studies of much more limited exchanges of Indian and Pakistani nuclear weapons (a total of one hundred 15-kt weapons) also would produce severely disruptive global climatic effects.31 All of these studies can be and are disputed.32 The national security imperative here is to enable U.S. decisionmakers and citizens to better understand the potential consequences of nuclear-weapons use. It can also raise the question of whether reductions in the numbers and yields of weapons, and changes in target selection, would reduce the likelihood of unnecessary suffering in the event that deterrence fails.

If the academy concludes that a new study (or studies) is warranted, the U.S. government should fund it. Such a study should consider scenarios for U.S.-Russian and U.S.-Chinese nuclear exchanges, drawn from Defense Department (including USSTRATCOM) war games. An unclassified version should be made freely available to invite international discussion and debate. The United States should welcome similar studies by other nuclear-armed states and be prepared to engage in discussions with counterparts on them.


1 See 2018 NPR; 2019 MDR; Bureau of Arms Control, Verification and Compliance, U.S. Department of State, “Strengthening Deterrence and Reducing Nuclear Risks”; and Trenin, “Decoding Russia’s Official Nuclear Deterrence Paper.”

2 From Russia’s perspective, its new nuclear-armed delivery systems (publicized in March 2018) help reset the U.S.-Russia mutual deterrence relationship that was lost following U.S. withdrawal from the ABM Treaty. Not surprisingly, U.S. officials view the new capabilities as qualitative improvements to Russia’s already assured deterrent and presume that Russia’s intentions are to disrupt strategic stability.

3 See Alexei Arbatov, “A New Era of Arms Control: Myths, Realities, and Options,” Carnegie Moscow Center, October 24, 2019,

4 For background, see Michael R. Gordon, “Russia Warns US Moves Threaten 2011 Nuclear Pact,” Wall Street Journal, January 15, 2019,

5 This history is recounted most recently in Kaplan, The Bomb.

6 Of course, if sublimits were used in a future agreement, similar to START I, it is conceivable that certain types of delivery systems may be limited to smaller increments (e.g., START limited the number of heavy ICBMs to 154).

7 Kaplan, The Bomb, 270.

8 Alexey Arbatov and Vladimir Dvorkin, “The Great Strategic Triangle,” Carnegie Moscow Center, April 1, 2013,

9 New START data exchanges conducted twice per year demonstrate Russia’s continued preference for deploying fewer delivery systems capable of carrying multiple warheads. In doing so, Russia maintains a rough warhead parity, around 1,300 at last count. Bureau of Arms Control, Verification and Compliance, “Fact Sheet: New START Treaty Aggregate Numbers of Strategic Offensive Arms,” U.S. Department of State, July 1, 2020,

10 See Alexey Arbatov and Vladimir Dvorkin, “The Impact of MIRVs and Counterforce Targeting on the US-Soviet Strategic Relationship,” in The Lure & Pitfalls of MIRVs: From the First to the Second Nuclear Age, eds. Michael Krepon, Travis Wheeler, and Shane Mason (Washington, DC: Stimson Center, May 2016), 71,

11 “Avangard Hypersonic Missiles Replace Rubezh ICBMs in Russia’s Armament Plan Through 2027,” TASS, March 22, 2018,; and “Russia Halts Years of Work on Ballistic Missile to Pay for Hypersonic Weapons,” The Drive, March 23, 2018,

12 See Amy F. Woolf, “Russia’s Nuclear Weapons: Doctrine, Forces, and Modernization,” Congressional Research Service, July 20, 2020,; and Alexei Arbatov, “The Modern Arsenals of Nuclear States,” in Nuclear Reset: Arms Reduction and Nonproliferation, ed. Alexei Arbatov and Vladimir Dvorkin, English version ed. Natalia Bubnova (Moscow: Carnegie Moscow Center, 2012), 55,

13 If both countries go ahead with their respective new ICBM programs, other confidence-building options exist, such as numerical limits on siloed MIRV ICBMs; a phased, verifiable drawdown of “old” ICBMs as new ones are deployed; and delays in the deployment of new ICBMs.

14 “Under such conditions, there was no alternative to ground-based missiles. And it was the development of nuclear weapons delivery vehicles, such as Sergei Korolev’s R-7 ballistic missile or Lavochkin design Bureau’s Burya cruise missile, that made it possible to ensure parity and shape the basis of the nuclear relations that later became known as strategic stability. After all, the key pillar of the strategic stability in the traditional military sense of this term involves, first and foremost, the ability to reach the enemy. This is its central and determining factor.” See “Article by Deputy Director of the Department for Nonproliferation and Arms Control of the Ministry of Foreign Affairs of Russia Vladimir L. Leontiev ‘Message Received. Attempt at Distance Communication With the U.S. State Department’ Published in the Magazine ‘Independent Military Review,’ April 12, 2019,” Ministry of Foreign Affairs of the Russian Federation, April 12, 2019,

15 The inspection procedures used in the START and INF treaties to confirm that a former nuclear base is no longer accountable provide a good model for a future empty warhead storage facility agreement.

16 See “Technical Aspects of the United States Ballistic Missile Defense System in Romania,” Ballistic Missile Defense Agreement Between the United States of America and Romania, Office of the Spokesman, U.S. Department of State, September 13, 2011,

17 See Steven Pifer, “The Future of U.S.-Russian Arms Control,” Carnegie Endowment for International Peace and Chicago Council on Global Affairs, 2016,

18 See “Department of Arms Control and Disarmament Holds Briefing for International Arms Control and Disarmament Issues,” Ministry of Foreign Affairs of the People’s Republic of China, July 8, 2020,

19 “Report on Nuclear Employment Strategy of the United States,” 3.

20 “Beijing sees this decision [the development of low-yield nuclear weapons as advanced by the 2018 NPR] as the most recent evidence that the United States seeks not only to advance its nuclear warfighting capability but also to deliberately lower the threshold for nuclear use, a U.S. choice driven by the suspected intention of being able ‘to conduct a preemptive strike.’” See chapter 1 of Tong Zhao, “Narrowing the U.S.-China Gap on Missile Defense: How to Help Forestall a Nuclear Arms Race,” Carnegie Endowment for International Peace, June 29, 2020, pp. 15–16,

21 Elbridge Colby, “The INF Treaty Hamstrings the U.S. Trump Is Right to Leave It,” Washington Post, October 23, 2018,

22 Ankit Panda, “Questions About China’s DF-17 and a Nuclear Capability,” Diplomat, February 16, 2020,

23 Thomas Nilsen, “Russia’s Top General Indirectly Confirms Arctic Deployment of the Unstoppable Kinzhal Missile,” Barents Observer, December 19, 2019,

24 A “heavy bomber” as defined in Part One of the Protocol to the New START. See also Tong Zhao, “Opportunities for Nuclear Arms Control Engagement With China,” Arms Control Today, January/February 2020,

25 Russia and China may seek to limit U.S. SLCMs in any agreement that limits air-launched and land-based missiles. However, it will be exceptionally difficult for either country to convince the United States to accept numerical limitations on conventional SLCMs. There are also inherent difficulties in verifying those limits. These factors make any prospective conventional SLCM arms limitations unlikely, though presumably with the advent of more advanced conventional SLCMs and sea-launched boost glide systems, the three powers will debate this issue in future strategic dialogues. An upcoming Carnegie report on the future of arms control will discuss this issue in greater detail.

26 This follows a cooperative effort by the United States and Russia, beginning in 1998, to share information related to ballistic missile launches, to assuage Russian concerns regarding U.S. national missile defense plans, and to focus on the growing ballistic missile threat posed by Iran. For an overview, see John David Steinbruner, “The Significance of Joint Missile Surveillance,” American Academy of Arts & Sciences, January 2001,

27 James M. Acton, Thomas D. MacDonald, and Pranay Vaddi, “Revamping Nuclear Arms Control: Five Near-Term Proposals,” Carnegie Endowment for International Peace, December 2020.

28 A slew of recent ideas in these areas provides fruitful ground for arms control efforts. See Arbatov, “A New Era of Arms Control”; and Ulrich Kuhn, “Between a Rock and a Hard Place: Europe in a Post-INF World,” Nonproliferation Review 26, no. 1–2 (2019): 155–66, DOI: 10.1080/10736700.2019.1593677.

29 Toby Dalton and George Perkovich, “Thinking the Other Unthinkable: Disarmament in North Korea and Beyond,” Livermore Papers on Global Security No. 8 (Livermore, CA: Lawrence Livermore National Laboratory Center for Global Security Research, July 2020),

30 For further exploration, see George Perkovich and James Acton, “Abolishing Nuclear Weapons,” Adelphi Paper 396, International Institute for Strategic Studies, 2008; and Dalton and Perkovich, “Thinking the Other Unthinkable.”

31 See, for example, Kathryn Hansen, “How Would Nuclear War Affect the Climate?,” NASA Global Climate Change, February 22, 2011,; and Michael J. Mills, Owen B. Toon, Julia Lee‐Taylor, and Alan Robock, “Multidecadal Global Cooling and Unprecedented Ozone Loss Following a Regional Nuclear Conflict,” Earth’s Future 2, no. 4 (2014): 161–76, DOI: 10.1002/2013EF000205.

32 Jon Reisner et al., “Climate Impact of a Regional Nuclear Weapons Exchange: An Improved Assessment Based on Detailed Source Calculations,” Journal of Geophysical Research: Atmospheres, February 13, 2018, DOI: 10.1002/2017JD027331.