We Need More Off-Ramps for Nuclear Crises

Commentary by Heather Williams and Nicholas Adamopoulos

Published May 12, 2025

Nuclear risks are on the rise, and crises involving nuclear-armed actors seem to be increasing in frequency, with the potential to become a defining feature of future regional conflicts. India and Pakistan, two states with nuclear weapons, are again reciprocally escalating tensions across the Line of Control following a terrorist attack in Jammu and Kashmir in April. Pakistani officials called India's missile strikes a "serious provocation." Recent studies also suggest that China could rely on nuclear saber-rattling or use in a crisis over Taiwan. Putin's announcement of the full-scale invasion of Ukraine in February 2022 included an implicit nuclear threat, with Putin warning that those that interfered would face "consequences . . . as you have never seen in your entire history." The threat of nuclear escalation grew in the fall of 2022 with the emergence of a Russian narrative that Ukraine was preparing to conduct a false flag operation with a dirty bomb, a move some observers believed to be a pretext for Russian nuclear use in Ukraine. As states increasingly rely on nuclear weapons to achieve political goals and backstop military operations, the risk that crises around the world could escalate to nuclear use is likely to grow.

Some well-established tools to avoid escalation and nuclear use in crises include deterrence, disarmament, and arms control. A comparatively underexplored idea is the use of off-ramps-ways for both sides in a conflict to back away from escalation while still saving face. While off-ramps were occasionally used during the Cold War, like other tools to manage nuclear crises and mitigate escalation risks, they must be adapted to the changing nature of conflict to account for the unpredictable nature of modern crisis escalation in the information age. Given the speed of conflict, including disinformation and narratives around conflict, perhaps the most important role of off-ramps in the current nuclear landscape is to provide a pause, but the challenge for many leaders in implementing off-ramps may be domestic politics rather than international ones.

Rising Nuclear Risks

Russia's nuclear saber-rattling, India and Pakistan's border skirmishes, and fears of Chinese nuclear coercion in a Taiwan contingency are all individual data points characterizing a larger trend of rising nuclear risks. The reasons for this rise are many and varied. The bilateral arms control regime that governed much of the U.S.-Soviet and U.S.-Russian nuclear relationship has crumbled, increasing the opacity of Russia's nuclear modernization and intentions. China is undergoing a rapid expansion of its nuclear arsenal while refusing dialogue with competitors on their nuclear doctrine or intentions. Meanwhile, many nuclear-armed states are directly or indirectly involved in conflicts around the globe, and most of the tools used to manage such crises have fallen victim to growing political tensions.

Russia has increasingly relied on nuclear coercion to bolster its conventional war in Ukraine. Initial nuclear signals from the Kremlin were targeted at preventing NATO intervention in the invasion, with Putin arguing that "top officials of the leading NATO countries are indulging in aggressive statements directed at our country. Therefore, I order the Defence Minister and Chief of the General Staff to put [the] Russian Army's deterrence forces on high combat alert." In the first year of the conflict, however, Russia also used external channels to sow confusion and fear, with prominent nongovernment actors calling for nuclear use in retaliation to Western aid to Ukraine. Russian political scientist Sergei Karaganov's 2023 argument for nuclear preemption and TV presenter Dmitry Kiselyov's recent threats on Russian TV, for example, go against the narrower set of scenarios in which Russia's official nuclear doctrine justifies nuclear use. These statements flood the zone with contradictory information in an attempt to manipulate risk in the minds of Ukrainian and NATO decisionmakers and their publics.

Legacy tools to manage crises are under increasing strain. Risk-reduction measures based on transparency serve little purpose when one state sees value in manipulating escalation. Throughout the war in Ukraine, Russia has found value in playing spoiler in multilateral forums for arms control and disarmament dialogue and has repeatedly refused to seriously engage in bilateral arms control discussions with the United States. Since Russia's invasion of Ukraine, experts and policymakers alike have agonized over how to incentivize Russia to de-escalate its war in Ukraine, come to the table on arms control issues, and refrain from aggressive saber-rattling. In short, what are the incentives for Russia, or any actor in the future who uses nuclear weapons to backstop conventional aggression, to de-escalate? On the one hand, increasingly forceful demonstrations from the West that Russia will not achieve its aims in Ukraine-and will suffer more the longer the war continues-may eventually force Russia to negotiate an end to the war before it slips further from its original aims. On the other hand, too forceful a demonstration may force Putin into believing that he is backed into a corner and has no option left but to escalate massively and rapidly. Does the West have leverage to offer Russia another way out of the conflict?

A Short Thought Experiment on Off-Ramps

In a 1991 analysis of escalation, British historian and strategist Lawrence Freedman observed, "Escalators can go down as well as up; they rarely go straight from the bottom to the top floor, but there are intermediate stages at which one can get off and consider where one is going." Those "intermediate stages," when one can stop, get off the escalator, and "consider where one is going," are essentially off-ramps. For Americans, the concept of off-ramps will be most familiar in the context of highways. When you want to take an exit, you pull off the highway and go up a ramp leading you to somewhere else.

To take the analogy further, all off-ramps lead to slowing down. Some might lead to an ever-faster highway, cross a Rubicon, or go onto a dirt road. But rarely does an off-ramp lead to nowhere. The concept of "off-ramps" was rarely, if ever, explicitly used during the Cold War. Instead, the majority of Cold War-era scholarship focused on linear, and potentially automatic, escalation. Deterrence theorist Herman Kahn's escalation ladder, for example, included 44 steps, with nuclear first use happening at approximately the midway point on that ladder.

Aside from a limited number of uses in the 1990s and early 2000s, the term gained prominence when used by U.S. Admiral Mike Mullen in 2008: "These on-ramps and off-ramps provide a vital measure of control in conflict and give both sides a chance to solve problems more carefully." To be clear, Mullen was hardly talking about off-ramps in the context of arms control; rather, he framed them explicitly as mechanisms for enhancing deterrence:

Flexibility in our deterrence construct hedges against the possibility that adversaries might incorrectly perceive their actions as "below the threshold" of U.S. resolve and response. We must manage that threshold by looking at ways to limit the pain an adversary can cause through advanced defensive measures. Adversaries must know that they have a limited ability to hurt us.

But the speed, complexity, and uncertainty of crisis escalation has only gotten worse since Mullen first used the term. Russia's war in Ukraine has been a manifestation of "wormhole escalation," whereby conflicts escalate and de-escalate erratically rather than following the stepwise ladder envisioned by Kahn. Additionally, crises in North Korea and South Asia have demonstrated the increasing role of nontraditional signaling mechanisms and disinformation in shaping crisis dynamics.

Toward a Typology of Off-Ramps

Recent history offers at least three broad types of off-ramps worth considering. First, communication off-ramps allow for leaders to engage during a crisis and send a signal of de-escalation or enhanced deterrence. The classic example of this, of course, is the Cuban Missile Crisis, when U.S. President John F. Kennedy and Soviet Premier Nikita Khrushchev brought the world back from the brink of catastrophe through a series of exchanged letters. The slow pace with which Moscow and Washington were forced to communicate led to mutual interest in concluding the 1963 Hotline Agreement, as diplomatic cables during the crisis took hours to receive. At the height of tensions, Khruschev pleaded, "if there is no intention . . . to doom the world to the catastrophe of thermonuclear war, then let us not only relax the forces pulling on the ends of the rope, let us take measures to untie that knot." On this off-ramp, the assumption is that escalation has occurred inadvertently or by accident, and leaders just need a moment to grasp the other sides' perspective to realize they were on a highway they did not way to be on, and to get off. Crisis communication channels, therefore, are a popular recommendation for de-escalation and risk reduction, such as direct communication links and formalized military-to-military communication protocols.

A second type of off-ramp is through a third party, which forces protagonists off their chosen course and onto the off-ramp. The most prominent examples of this are in the context of South Asia, where the United States has repeatedly acted as a mediator. For example, in the 1999 Kargil War, the United States played a crucial mediation role and President Clinton was able to persuade Pakistani Prime Minister Nawaz Sharif to withdraw his forces from the Line of Control; later reports suggest that Sharif was not aware of just how escalatory the crisis had become. In the India-Pakistan crisis over Kashmir, President Trump again appears to be trying to position the United States as a mediator; but, of course, it is ultimately up to leaders in New Delhi and Islamabad to decide whether to take an off-ramp.

Finally, narrative manipulation is a face-saving way of ending a conflict. This appears to have been the off-ramp of choice in the U.S.-Iran crisis in the early days of 2020. Five days after the assassination of General Qasem Soleimani on January 3, Iran attacked a U.S. base in Iraq. In addition to direct U.S.-Iran confrontations, Iran accidentally shot down a Ukrainian passenger jet during the height of the crisis. This tragedy likely contributed to de-escalation, but both sides embarked on ambitious social media campaigns to frame the narrative around the crisis and its conclusion. For example, the Iranian Ayatollah posted a photoshopped photo of Trump on Twitter featuring slap marks across his face and the caption "just a slap, revenge is another matter." Trump, meanwhile, quickly responded, "All is well!... So far, so good!" The manipulation allowed both sides to avoid further tit-for-tat escalation while saving face domestically.

Opportunities for Off-Ramps and the Challenge of Domestic Politics

In the coming year, Russia and Ukraine may agree to a peace settlement. Russia may take a few years to revive its conventional capabilities and potentially even take up President Trump's offer for arms control. Tensions between India and Pakistan seem to have subsided following the recent crisis. But the risk of crisis escalation between nuclear-armed adversaries is not going away. A recent CSIS study on strategic stability in Europe, for example, concluded that Russia may increase reliance on nuclear weapons in future conflicts following the war in Ukraine. The India-Pakistan crisis is the most recent in a decades-long conflict that epitomizes the stability-instability paradox. Potential conflict with China looms, and North Korea continues to pose a nuclear threat in the region. In short, now is the time to develop off-ramps for managing escalation, strengthening strategic stability, and reducing the risk of nuclear use.

What might off-ramps look like in practice, and how can the United States build them? To prevent misperception during a crisis, nuclear possessors can establish lines of communication and transparency now, in advance of a crisis, that may prove valuable in due course. The most immediate opportunity lies in the potential for China to join the Nuclear Risk Reduction Centers (NRRCs), a critical communication channel through which the United States and Russia share notifications related to treaty compliance. The NRRCs also allow the United States and Russia to exchange routine maintenance, test, and ad hoc messages. The value of the NRRCs as channels for ad hoc notifications was recently demonstrated when Russia used Oreshnik ballistic missiles against Ukraine, as Russia felt compelled to notify the United States of their use ahead of time using an NRRC.

Another opportunity would be to establish a crisis stability initiative within the P5 process in the Nuclear Nonproliferation Treaty (NPT), which could draw on lessons learned from recent crises, such as Ukraine, about how escalation happens and where states can press "pause" and take an off-ramp. Such a discussion would provide both a retrospective analysis of the ongoing war in Ukraine and a forward-looking exploration of new tools for de-escalation and off-ramping. Perhaps the most pressing need for off-ramps, however, is in South Asia, which for years has been the subject of recommendations on developing crisis communication channels and off-ramps. Stimson Center cofounder Michael Krepon, for example, has recommended expanding missile flight test notification agreements, establishing nuclear hotlines between India and Pakistan, and modernizing nuclear facility "non-attack" agreements.

Turning to off-ramps by third parties, while historically the United States has often played this role, European states might also step in as mediators and facilitators of off-ramping. European states such as France and the United Kingdom have traditionally been leaders in risk reduction efforts, such as in the P5 process. As Europe assumes greater strategic autonomy and a more prominent role in arms control, it might also expand on this role by offering off-ramps during crises. For example, European states were a key player in bringing Iran to the negotiating table leading to the Joint Comprehensive Plan of Action (JCPOA), including through measures such as the European Union's sanctions on specific sectors, which were critical in encouraging Iran to make significant concessions during the 2015 negotiations.

Finally, states will likely continue to use narratives to frame de-escalation in face-saving ways. On May 10, India and Pakistan agreed to a ceasefire, and each claimed the conflict as a military victory. Pakistani Prime Minister Shehbaz Sharif declared, "We have won. This is victory." Meanwhile, Indian Navy Captain Ravi Nair claimed, "Every misadventure by Pakistan has been met with strength, and every future escalation will invite a decisive response." The risk of this approach, of course, is that it could further whip up nationalist discourse in further conflicts.

Conclusion

States can always choose not to take off-ramps and stay on the fast track. Russia's behavior in Ukraine, for example, suggests that some escalation may be highly intentional rather than accidental or due to miscommunication. Russia withdrew from arms control agreements, which provided unique communication channels and opportunities for dialogue, amid the crisis. Nevertheless, the war in Ukraine has also demonstrated that it is better to have tools for crisis management and de-escalation than to not, even if parties to the conflict use them sparingly. As the war trends toward a negotiated settlement that no party may be wholly content with, off-ramps that allow for domestic political face-saving will be vital. Furthermore, the growing presence of Europe as a strategic actor may require its participation in and facilitation of off-ramps in crises around the world. As crises become more complex and interconnected, so too must the tools policymakers use to manage them. In the face of such complexity, off-ramps may serve as valuable pauses, forcing conflict parties to stop and take stock before deciding which escalator to take.

Heather Williams is the director of the Project on Nuclear Issues and a senior fellow in the Defense and Security Department at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Nicholas Smith Adamopoulos is an associate director and associate fellow with the Project on Nuclear Issues at CSIS.

Programs & Projects

Commentary by Lachlan MacKenzie and Catherine Murphy - May 1, 2025

Commentary by Heather Williams - May 8, 2025

Report by Heather Williams, Nicholas Adamopoulos, Lachlan MacKenzie, and Catherine Murphy - April 21, 2025