Reconsidering Arms Control Orthodoxy

4141208986_50fa2274ee_k (1)

The recent extension of New START brought with it a collective sigh of relief, bolstering optimism and leading some to extol arms control’s far-reaching potential to improve security conditions. But with optimism comes pessimism. Other analysts proclaim an imminent crisis and herald the demise of arms control. Both camps are falling prey to flawed assumptions. The optimists seem to forget that cooperation also comes with costs and risks, while the pessimists overstate obstacles and provide an overly narrow view of what is available for arms control. Of course, many experts find themselves in between those two camps. Yet clarity is missing even in this middle ground, where debates rage on about when arms control is possible, why states pursue it, and the nature of the benefits and drawbacks of various proposals.

In order to set the path to sound policy, we challenge several trends in the current discussion on the future of arms control. We identify five problematic arguments: first, that arms control is only possible during good relations; second, that a lack of mutual interest is the key impediment to creating agreements; third, that past cheating is an obstacle that needs to be addressed before future cooperation; fourth, that only bilateral (or conversely, only multilateral) approaches will be fruitful in the current security environment; and fifth, that issues that are non-starters for the United States should be avoided, while parity and symmetry should be preserved. Our analysis highlights a need to reconsider assumptions that inhibit more creative thinking about future arms control options.



While the U.S.-Russian relationship continues to be a useful model for thinking about the future, looking at a broader range of experiences shows that arms control is possible both when geopolitics are tense and when conditions are better, and that the effects of agreement violations are often misunderstood. Yet the simplicity of so-called mutual interests can be insufficient. Arms control requires a careful balance between the needs for creating transparency and preserving secrecy. In terms of agreement forms, multilateral and bilateral approaches may each be optimal under different conditions. Finally, the United States should also reassess what it is willing to offer in negotiations, including moving away from treating certain issues like missile defense as sacrosanct and seriously exploring the possibilities of deals with asymmetric terms.

Thomas Schelling and Morton Halperin defined arms control as “all the forms of military cooperation between potential enemies in the interest of reducing the likelihood of war, its scope and violence if it occurs, and the political and economic costs of being prepared for it.” Many analysts readily quote this definition, but then fail to apply it in their search for models and lessons. Taking this definition seriously reveals that arms control is a robust and long-standing foreign policy tool, involving a wide range of actors and types of activities. As such, it is time for policymakers and other stakeholders to reconsider and widen the scope of arms control, reevaluate common wisdom about obstacles to cooperation, and address the constraints and opportunities for future cooperation.

Arms Control Is Possible in Both Good Times and Bad

Analysts and policymakers often decry the current state of U.S.-Russian relations, identifying bad relations as a key hurdle for cooperation on arms control. At the same time, others note that the need for arms control further increases when risks of war, crisis, or arms racing are heightened. Taken together, these views suggest a misleading conclusion: that arms control is only possible during détente, but only truly valuable when state relations are poor. Both positions are overstated and combining them leads to excessively pessimistic conclusions about when arms control agreements can successfully be negotiated and what they can achieve. In reality, states have signed agreements in times of good relations as well as bad.

The 1987 Intermediate-Range Nuclear Forces treaty is a key example of an arms control agreement that was negotiated in the midst of tense U.S.-Soviet relations. At the same time, we have seen states sign agreements in periods of détente: SALT I in 1972 and the Moscow Treaty in 2002 are examples in the U.S.-Russian context. Some confidence-building agreements between India and Pakistan have a similar pattern, coming out of periods of relative calm. Arms control negotiations also often take place over many years — spanning changes in political leadership and geopolitical tensions. To conclude that starting arms control is impossible or unrealistic because of the current state of U.S.-Russian relations effectively forecloses opportunities to lay the foundation for future achievements.

Analysts argue that good relations foster great trust, which is a precondition for arms control. But the concept of “trust” is often misused. Arms control has never been based on blind trust. Trust in the genuine efforts of individual interlocutors, or a belief that the other side is serious about negotiations as a foreign policy process, can be beneficial to negotiations. However, states can and often do negotiate without trusting that the other side will ultimately comply with agreement terms. Arms control, of course, typically takes place between adversarial states rather than friendly ones. Rather than trust, states need an assessment that the other side perceives benefits from cooperation. If all parties do not see benefits in arms control, there is little basis for a negotiation.

This is not to say that underlying conditions do not matter. Indeed, they may affect outcomes well beyond the immediate security challenge. When relations between states are relatively good (or at least less tense), arms control can lay the groundwork to forestall arms racing and reduce the risks of conflict if (or when) relations take a negative turn. For example, arms control confidence-building agreements in the Organization for Security Cooperation in Europe established at the end of the Cold War (e.g., the Vienna Document) have provided a framework for reducing military risks in the region that have endured even when relations between the United States and Russia have soured. In a different context, the Treaty of Tlatelolco (which established a nuclear weapon-free zone in Latin America) was also an effort to transform geopolitics more broadly. Agreements negotiated during high tensions can also affect broader competition, including by forestalling new technologies that could undermine strategic stability.

Even a relatively quick glance at historical patterns suggests that there is no perfect moment for arms control. Waiting for geopolitical relations to improve risks missing moments of opportunity when both sides see benefits in avoiding a certain costly or risky aspect of their ongoing competition. Neglecting arms control when relations are good suggests a failure to anticipate future threats and changes, not to mention forgoing opportunities to save resources. As long as the United States and other parties can identify a mutual interest in arms control, the United States should pursue negotiations.

Mutual Interest Is Not Enough 

Yet agreements are still difficult even when both sides see benefits. Arms control can create benefits for states by reducing the costs of competition. This can come from reducing the material costs of arms racing or engaging in a war, reducing the risk of conflict starting or escalating, or reducing the damages sustained if war does happen. Yet there are also costs associated with arms control agreements. Importantly, arms control agreements that rely on intrusive verification and monitoring raise concerns that these provisions will provide the opponent with increased intelligence about the state’s activities beyond the contours of the agreement. For example, inspections to confirm that a military base does not contain prohibited weapons may also reveal information about other military capabilities or make it easier to attack the base in the future. States often find themselves unable to resolve the tradeoff between having enough transparency to make sure that each party is complying with the agreement and enough secrecy to protect each state’s security, leading to few possibilities for even those arms control agreements that reduce risks or costs.

Yet this tradeoff challenge is based on each side’s understandings of what it needs for secrecy, and these perceptions can change. In early Cold War arms control negotiations, the Soviet Union was deeply concerned about maintaining secrecy around its military sites, not only because of the military threats from the United States but also because they believed that transparency to Western audiences would create domestic doubt in the authority of the Soviet system. During the mid-1980s, the Soviet leadership’s assessment of the need to maintain such secrecy shifted, and this contributed in part to changes in their positions on intrusive verification. Parties’ assessments of how much information they need to verify compliance may also shift, as was the case when joint verification experiments allowed the United States and the Soviet Union to draw new conclusions about the capabilities of measurement tools for monitoring nuclear testing under the Partial Test Ban Treaty. Finally, some technological innovations can create new possibilities for verifying compliance without creating new security risks due to an agreement. The introduction of satellite surveillance in the 1970s allowed the United States and the Soviet Union to rely on “national technical means” rather than inspections for some agreements, and advances in satellite technology today continue to enhance treaty monitoring from afar. Another option to proactively address this dilemma is to empower third parties, like the International Atomic Energy Agency, to handle monitoring. However, all parties need to have confidence that the third party has the capacity not only to detect agreement violations but also to hold each state’s secrets secret.

The dilemma of the transparency-security tradeoff has a number of implications for next steps. First, cognizance of this dynamic is important for setting expectations — adversaries often have security reasons for avoiding transparency in arms control, even when they are, in principle, interested in cooperation. Addressing those concerns might need to happen both within the specific provisions set by an agreement, or in the context of security relations well beyond the issues at the agreement negotiating table. Second, there should be continued investment in research and development of verification technologies, including in collaboration with other countries and international third parties. Finally, the discussion of arms control should not idealize greater transparency as an all-around positive. Indeed, domestic political pressure for transparency that goes beyond the compliance needs of an agreement may undermine the deal as a whole.

Cheating Is Not Always an Obstacle

The common wisdom on arms control is that states only sign treaties they plan to comply with anyway and cheating that does occur should both be met with clear punishment and is also a sign of agreement failure. When Donald Trump’s administration withdrew from the Open Skies Treaty, it claimed that it did so in response to Russia’s violations of the treaty. Today, some analysts argue, the United States ought to somehow address Russian cheating before negotiating a new arms control agreement, even though the United States has already withdrawn from the 1987 Intermediate-Range Nuclear Forces Treaty where the violations occurred. Yet the history of arms control experience suggests that cheating does not necessarily derail prospects for future agreements, and there have actually been innovations in the design of agreements that allow states to address the consequences of cheating while still maintaining the overall architecture of a negotiated deal.

Cheating on arms control agreements is not a new phenomenon. For example, the Soviet Union violated the SALT II agreement, both by developing the SS-25 missile and by encrypting missile test data that was supposed to be shared for verification purposes. The Soviet Union also violated the 1972 Anti-Ballistic Missile Treaty by developing the Krasnoyarsk radar system in the late 1970s and early 1980s. At the time, many claimed that the United States needed to address Soviet cheating and respond to these violations by withdrawing from these agreements before committing to any future deal. However, neither of these violations led to the collapse of either agreement. Moreover, the United States and the Soviet Union began negotiations on what would become the 1987 Intermediate-Range Nuclear Forces Treaty and the 1990 START I agreement despite outstanding violations of prior agreements and the tensions they produced. Instead, the United States and the Soviet Union were able to reach agreement on the Intermediate-Range Nuclear Forces Treaty — a significant bargain that banned all nuclear and conventional ground-launched ballistic and cruise missiles that had a range of between 500 and 5,500 kilometers — in 1987. Soviet treaty violations were only finally addressed in 1989.

The implications of cheating and the options available for addressing it vary, but there is reason for optimism. Despite a tendency to focus on the negative effects of violations, the consequences of cheating depend partially on how the other parties respond. The Assad regime’s violations of the Chemical Weapons Convention in Syria, for example, did not lead to a flood of states violating the agreement and using chemical weapons in war. Nor did North Korea’s withdrawal from the Treaty on the Non-Proliferation of Nuclear Weapons lead to a rush of states pursuing nuclear weapons. Although instances of cheating on agreements are never desirable, the public identification of such violations combined with serious efforts to enforce the rules of the agreements demonstrate that the regime is working as intended and provides value in policing threats that arise.

Even when parties violate agreements, it may still be worthwhile to maintain them. This is the case if an agreement can still offer information that is more difficult to obtain or reveal otherwise. Despite concerns over Russian compliance with the Open Skies Treaty dating back to at least 2010, the United States has still gained transparency and predictability regarding Russian military forces that it now lacks after having withdrawn from the treaty. Specifically, information gained via Open Skies was critical for the United States in addressing Russian activities in Ukraine in 2014. Information gained through a treaty monitoring process is also easier to share with both allies and adversaries, allowing Washington to raise concerns or make claims in international disputes without revealing its own intelligence sources and methods. The implication here is not that cheating should be allowable, but rather that is not as clear cut of a problem as is often portrayed.

The design of arms control agreements has also evolved to allow more flexible responses to cheating. In most cases, a reciprocal violation, either secretively or more commonly through a formal treaty withdrawal, is the fundamental punishment for another state’s noncompliance. But this response scuttles the agreement as a whole and reactivates broader obstacles to cooperation, including long negotiations, a possible lack of political will, and domestic politics that may imperil treaty ratification. Violation resolution and punishment mechanisms built into the very terms of a treaty have the potential of improving agreement longevity. While such mechanisms were relatively rare in early arms control efforts, notable modern agreements include innovations addressing violations. The Treaty on the Non-Proliferation of Nuclear Weapons itself does not include dispute resolution or interim punishment, but the U.N. Security Council has placed sanctions on Iran for its noncompliance with the treaty without any parties withdrawing from it. Returning to compliance with the treaty was the requirement for lifting the sanctions and avoiding further punishment. Likewise, the Iran nuclear deal allows parties to punish Iran with “snapback” sanctions if Iran fails to satisfy other parties’ accusations of noncompliance, but it does not require any parties to withdraw, and Iran can receive relief from sanctions if it reverses its violations of the agreement. Beginning with the 1972 Anti-Ballistic Missile Treaty, bilateral arms control agreements have also included joint consultative commissions to work through irregularities in compliance.

In sum, the United States should carefully assess the consequences of other states cheating on arms control agreements and develop a proportional response, which in most cases should not be a rejection of all further efforts at cooperation. The design of future arms control agreements should be an innovative enterprise, focused in particular on developing legal mechanisms that would allow states to address violations within the context of the arms control agreements and related institutions.

Is Three (or More) a Crowd? 

Recommendations for next steps on arms control have recently either highlighted the value of the U.S.-Russian bilateral model, alternatively claimed that a multilateral approach is necessary in the modern security environment, or suggested that both should be attempted. The problem is not in the proposals themselves, but rather in the lack of analysis devoted to why or when bilateral versus multilateral approaches may be preferable. At the same time, evidence from past cooperation suggests that the choice of approach is a consequential one both for basic progress in negotiations and in getting to effective agreements. In considering whether to pursue a bilateral or multilateral approach, the United States should assess the source of the challenge and the relevant parties needed to address it.

In multilateral agreements, reciprocity is diffuse: States do not engage in individual tit-for-tat bargains with one another but rather commit to general standards of behavior. Consequently, when the United States seeks to stave off threats that could potentially arise from anywhere (e.g., nuclear proliferation, terrorists acquiring nuclear weapons, or the placement of weapons of mass destruction in outer space), it has often pursued multilateral agreements rather than bilateral ones. The Australia Group — a multilateral export control regime to prevent the spread of chemical and biological weapons — has made important progress even though Russia does not participate. Other export control regimes (e.g., the Nuclear Suppliers Group or the Hague Code of Conduct on Ballistic Missile Proliferation) likewise aim to prevent any state that does not currently have these weapons and delivery vehicles from acquiring them. By making blanket rules for a large number of states, multilateral agreements may also allow the United States to preempt unforeseen threats or developments, and lock in favorable conditions. Moreover, when the costs of compliance are low for most states, multilateral agreements can formalize and institutionalize support for U.S. objectives.

Yet for many issues in arms control, there are only a few actors that the United States seeks to restrain, and the United States would do better to negotiate directly under conditions of specific reciprocity with those states whose behavior or capabilities threaten the United States. For example, it makes little sense at present for the United States to pursue multilateral negotiations to limit or ban nuclear-powered cruise missiles, given that Russia is the only state even considering developing such missiles. As many scholars have found, increasing the number of actors in a negotiation or bargaining scenario increases the difficulty of finding compromises and agreements that will be acceptable to all parties. Importantly, many states (even allies) do not always see eye-to-eye with the United States on both threats and solutions. Regional approaches to arms control highlight this difficulty: Japan and South Korea are both U.S. allies, but differ from each other and from the United States in their relationship with China and the threats they prioritize. These factors make a regional approach to arms control in East Asia more difficult and complicated than a bilateral Sino-American negotiation. Proposals for trilateral negotiations between Russia, China, and the United States are worth exploring because direct reciprocity could be established between this small set of parties. However, skepticism is also warranted: Russia and China differ substantially in both their capabilities and their relationships with the United States, and a deal between two parties may be unacceptable to the third. This could block progress that would otherwise be possible in a bilateral agreement.

Neither bilateralism nor multilateralism is inherently better for arms control. The U.S.-Russian bilateral model of arms control should not be taken as the only model, and the format for negotiations should be tailored to the source of the challenge. The United States should pursue multilateral initiatives when there is broad agreement on the problem, when risks and benefits are diffuse among states, and when the costs of compliance are low. Bilateral initiatives will be more fruitful when there are strong competing interests among both adversaries and allies, challenges and benefits are more concentrated and precise, and the United States is primarily concerned about the capabilities or behavior of one key state.

The American Offer 

Analysis of future arms control options needs to include a close reassessment of at least two common views about what the United States should seek to include in negotiations: first, that some U.S. capabilities are supposedly off the table, and second, that states need to be at rough parity on particular capabilities in order to want arms control. Both supposedly untouchable systems and asymmetric deals should be the subject of intensive internal review.

In assessing the future of arms control, experts often emphasize what the United States should seek to limit in other states, such as Russia’s tactical nuclear weapons or new delivery systems, or a broader need to “bring China into arms control.” Yet the reverse — what the United States is willing to put on the table — is equally important. To successfully limit adversaries’ capabilities, the United States needs to consider how it can incentivize adversaries to agree to such limits.

One area where this is particularly clear is missile defense, which is often treated as a third rail in arms control given the strong bipartisan support in the United States for maintaining (and increasing) it. Since leaving the 1972 Anti-Ballistic Missile Treaty in 2002, the United States has worked hard in arms control negotiations to avoid limits on missile defenses, accepting only a broad mention of a connection between defensive and offensive systems in the New START preamble. Yet study after study shows the limited utility of missile defense in defending against both nuclear and nonnuclear attacks despite the massive investment in it. Many analysts argue that investing in missile defense actually drives arms racing rather than making the United States safer. Rather than treating missile defense as untouchable, Washington should reassess whether the gains from preserving missile defense are worth the tradeoffs of what could be obtained through limiting missile defense as part of an arms control agreement.

Second, evaluating asymmetric outcomes is difficult, but should be approached with less trepidation and more urgency. It is already clear that Russia and the United States have diverging views about which specific capabilities add to their security, so meeting an underlying goal of reducing spending and risks may require moving beyond the traditional model of like-for-like limitations. Indeed, we have already seen steps in this direction, such as in how New START allows each side flexibility on the mix of delivery systems under the central limits.

Beyond rejecting symmetry in specific weapons, analysts should reconsider the role of balance in agreement obligation or in states’ underlying interests in pursuing limits. Analysts sometimes claim that states will never accept arms control for capabilities where they have an advantage. Yet arms control can also be a tool to preserve advantage. Disadvantaged states may not like it, but accepting an agreement could still be better than the alternative. The Treaty of Versailles, for example, was a dramatically unequal arms control agreement that required German disarmament in exchange for Germany remaining a sovereign country. Proposals for addressing North Korea’s nuclear program often tacitly call for asymmetric arms control. And the 2015 Iran nuclear deal imposed a host of obligations on Iran to curtail its nuclear program in exchange for commitments from the P5+1 to lift sanctions. While experts might argue about which side got the “better end” of that deal, it is in its form an asymmetric bargain.

If Washington stops treating issues such as missile defense and asymmetry as arms control non-starters, then the bargaining space for future cooperation can substantially increase. Of course, few specific agreements will ultimately be viable, but more creative options may emerge in the process. We can imagine two routes for undergoing this reassessment — one where the United States focuses on an internal debate before coming to the negotiating table, and a second where the United States conducts its review while international negotiations are also ongoing. The first is the most intuitive, but the second has considerable merit that should not be overlooked. Learning more about other parties’ position, demands, or even possible bluffs in the course of negotiations can inform America’s self-assessment and internal planning process. It may uncover new options, including with respect to allies. For example, recent research suggests U.S. efforts to increase the credibility of extended deterrence commitments may be counterproductive. A reassessment of what allies really need for assurance may benefit arms control negotiations. It is also possible that pressure for negotiations would arise from other states before new U.S. internal assessments are complete. Being unprepared to take advantage of such moments could lead to missed opportunities for progress, especially when high-level political attention is a key ingredient for an agreement.


That the future of arms control poses many challenges is not news to most experts. Yet some of the barriers to arms control are overstated, while others have been overlooked. Unpacking assumptions about the form of arms control agreements and the conditions under which arms control can take place reveals that opportunities for arms control extend far beyond the model epitomized by New START. Arms control is not dead and the goals of arms control remain as important as ever. Arms control is a policy tool, not an end in itself — it may be appropriate in some cases and less applicable in others. Future progress requires a careful assessment of the consequences of technological and political developments, and a more creative, expansive understanding of what arms control involves.



Naomi Egel is a Ph.D. candidate in government at Cornell University and the Janne Nolan Nuclear Security Visiting Fellow at the Truman Center for National Policy. 

Jane Vaynman is an assistant professor of political science at Temple University.

Image: White House (Photo by Pete Souza)