Our experts explain what US policymakers should know about deterring Russia’s and China’s nuclear threats

In June, the fellows from the Nuclear Strategy Project in the Scowcroft Center for Strategy and Security’s Forward Defense program met at a roundtable, supported by Atlantic Council Board Director Walter Slocombe, to discuss how to deter nuclear powers Russia and China from expanding their capabilities or incorporating them in their defense strategies. They narrowed in on Cold War deterrence methods, exploring whether those theories need to be completely changed for today’s strategic context. Below, edited for length and clarity, is their conversation moderated by Robert Soofer, senior fellow with Forward Defense.

Our expert analysts

  • Robert Soofer: Senior fellow at the Scowcroft Center for Strategy and Security’s Forward Defense program and former US deputy assistant secretary of defense for nuclear and missile defense policy
  • Keir Lieber: Nonresident senior fellow at the Scowcroft Center for Strategy and Security’s Forward Defense program and professor at Georgetown University’s Center for Security Studies
  • Amy Woolf: Nonresident senior fellow at the Scowcroft Center for Strategy and Security’s Forward Defense program and former specialist in nuclear weapons policy at the Congressional Research Service of the US Library of Congress
  • General Kevin Chilton: Distinguished fellow at the Scowcroft Center for Strategy and Security’s Forward Defense program and former commander of US Strategic Command
  • Franklin D. Kramer: Distinguished fellow at the Scowcroft Center for Strategy and Security and former US assistant secretary of defense for international security affairs
  • Hans Binnendijk: Distinguished fellow at the Scowcroft Center for Strategy and Security’s Transatlantic Security Initiative and former National Security Council official
  • Daryl Press: Nonresident senior fellow at the Scowcroft Center for Strategy and Security’s Forward Defense program and professor of government and director of the Institute for Global Security at Dartmouth University
  • Walter Slocombe: Board director, Atlantic Council
  • Rachel Whitlark: Nonresident senior fellow at the Scowcroft Center for Strategy and Security’s Forward Defense program and associate professor at the Sam Nunn School of International Affairs at the Georgia Institute of Technology

Lessons learned for deterrence theory

ROBERT SOOFER: China’s nuclear buildup, combined with Russia’s increased reliance on nuclear weapons, means that the United States will now be facing a two-peer nuclear environment. When it comes to lessons of the Cold War, are they applicable to this new tripolar context?. . .

Let’s start with deterrence theory: Does the Cold War strategy suffice, or does the United States need a new strategy? How do we think about deterrence during the Cold War? And is there something unique about this new era that requires us to rethink?

KEIR LIEBER: I’m going to focus on three key Cold War deterrence theories that I think are most relevant.

The three main deterrence challenges facing the United States were 1) how to deter a nuclear attack against the US homeland, 2) how to extend deterrence to cover allies, and 3) how to manage the risk of nuclear escalation in any conflict. The United States grappled with all three of these challenges during the Cold War, mostly over the best way to make nuclear deterrence credible.

In terms of the first challenge, the United States sought to deter Soviet nuclear attack against the US homeland by deploying a robust nuclear triad of forces that could survive any Soviet first strike and still be able to carry out overwhelming nuclear retaliation.

For the second challenge, the United States sought to deter a Soviet conventional invasion of its NATO allies in Western Europe by threatening to use nuclear weapons first. That is, the United States planned to escalate to nuclear war in response to a Soviet conventional attack. The US built the forces for this, and the United States forward deployed these forces in Western Europe to make this threat as credible, real, and convincing as possible.

Third, the United States sought to avoid escalation to all-out nuclear war in any conflict with the Soviet Union by seeking superiority at every rung of the escalation ladder. That is, the United States sought to convince the Soviet Union that the United States could, and would, respond in kind to any additional escalation. In other words, the United States pursued dominance for the purpose of deterrence; and if deterrence failed, the United States planned to rely on its dominance to coerce a halt to the conflict before things got totally out of hand.

Only the third deterrence theory requires major theoretical rethinking by US policymakers and defense planners today. I don’t think the United States needs a fundamentally new theory to understand how to deter an attack on the US homeland. Of course, the United States may need to build up its nuclear forces in response to buildups by China and Russia. This is the tripolar context. But it’s still hard to imagine a surprise nuclear strike that wipes out or disables the entire US arsenal, leaving nothing with which Washington can retaliate.

Extended deterrence to allies is getting more difficult but not necessarily because of the new dynamics of tripolarity. I don’t think the United States needs a new theory of extended deterrence in the age of tripolarity. Yes, the United States needs to do more to make the nuclear umbrella over Japan and South Korea more credible, but the real challenge simply comes from the growth of North Korea’s nuclear forces, their ability to hit the US homeland, and expected challenges like that.

However, when it comes to deterring and managing escalation in a conventional war against a nuclear power, I think Cold War theories, ideas, and explanations are hindering the US ability to think clearly and realistically about the threats. The key difference is that whereas in the Cold War it was the United States that sought to make nuclear escalation credible—that is, to make plausible in the mind of the United States’ key adversary that it would escalate to nuclear use—today, the shoe is on the other foot. Today it’s the United States’ potential adversaries—including China, Russia, North Korea, and someday Iran—that might need to threaten nuclear escalation to deter the United States in a conventional war.

Today the United States seeks to avoid nuclear escalation, not make it more likely. Yet, as seen in Ukraine and in the face of Russian President Vladimir Putin’s credible threats to escalate with nuclear weapons, most of the policy and analytic community is falling back on Cold War ideas about deterrence. Many seem to have little grasp of the reasons why an adversary losing a conventional war will seek to escalate, and what should be done about it. That’s where new theories are needed to understand and prepare for nuclear escalation in the conventional conflicts of today and tomorrow.

US leaders and the US analytical community need to better understand the new risks, dangers, and challenges of nuclear deterrence today. That’s why I see it as a theoretical challenge as much as anything else.

AMY WOOLF: The theory of deterrence and where there is a problem is not with the actual theory, but with the way that people shorthand the theory.

So throughout the Cold War, the shorthand—we would say “deterrence” and people would hear or read “deterrence by the threat of offensive nuclear retaliation.” But the theory of deterrence is that you ensure that the adversary making the decision—and this is a mind game—recognizes that the costs of taking action will far exceed the benefits of taking action. So the United States operationalized that during the Cold War through the threat of offensive nuclear retaliation.

In the circumstance of escalation from a conventional conflict to a nuclear conflict, deterrence—ensuring that the costs exceed the benefits of being the first one to use nuclear weapons—means a lot more than just threatening nuclear retaliation. You need to emphasize costs all along the escalatory ladder. If there is a regional crisis and you don’t want it to escalate, you can walk the crisis through in the Russia-Ukraine context.

But at the theory level, the idea is that through some integrated response to an adversary, you make clear to them that, at all steps of the escalation ladder, if they keep moving forward, their benefits will be swamped by the costs and dampened by the United States’ reaction. So it’s the cost-benefit mind game—which it was during the Cold War, but many shorthand it to nuclear deterrence in the Cold War.

KEVIN CHILTON: Perhaps assurance and escalation control are more closely linked than we thought of in the past. When thinking about what to field and how to posture forces, perhaps the associated considerations should not just be about escalation control but also about the benefits to assurance. One may enhance assurance or one may not, depending on how you think through both of these elements.

FRANKLIN D. KRAMER: The United States went from an early Cold War doctrine of threatening heavy use early to a doctrine of flexible response. And that basically meant that the United States was going to do its best to succeed at the conventional level. It didn’t take first use off the table, that’s for sure. But it certainly didn’t make first use predominant either.

So it seems to me that the challenges are pretty much the same. That is to say the United States always had to think about, if it started to succeed at the conventional level, the possibility that the adversary would go nuclear. That’s certainly the United States’ significant problem. It’s a problem in Europe. It’s a problem in the Indo-Pacific. But it’s not different, in my opinion, than the problem that the United States had once it moved to the flexible response approach.

HANS BINNENDIJK: But there is a distinction between deterring a rational actor and deterring a decisionmaker who is more emotionally engaged. During the Cold War, the United States was dealing with a Kremlin that, despite everything else, was comprised of actors that were still rational. I’m not sure Putin today is a rational actor. I’m not sure the United States is dealing with a rational actor in North Korea. And so as the United States thinks about deterrence theory, it might want to shape that theory with those two kinds of decision makers in mind: a rational actor and an emotional actor—someone deciding based on his or her own personal survival, on history, on culture, and all the things seen coming out of Putin’s mind.

DARYL PRESS: During the Cold War, with respect to the intersection between conventional war and escalation, there was a theoretical question the United States was trying to solve and a theoretical question that the Soviets were trying to solve.

This theoretical question links directly back to Thomas Schelling: How does the United States make credible threats that it will use nuclear weapons if its conventional forces are failing, when nuclear escalation at its highest level would lead to mutual disaster? And that’s a difficult theoretical question that gave birth to Schelling’s Nobel Prize-winning work about managing risks in a world of mutual destruction. So it was a theoretical problem that the United States faced in the Cold War, and it made a lot of progress on that problem, both theoretically and operationally in terms of deterrence strategy.

The Soviets faced a different theoretical problem, which is how do you beat in conventional war an enemy who has the ability to escalate to the nuclear level to avoid defeat? And I think they never solved that problem. The reason that’s relevant now is that’s the problem the United States faces. If it faces war on the Korean Peninsula tomorrow, the problem is how does the United States convince the North Koreans to allow it to beat them without them employing stalemated weapons? How does the United States get Russia to allow it, with its allies, to impose a really bad defeat on Russia without incentivizing an escalation? And it’s the problem the United States might face in Taiwan.

So, yes, at the most abstract level, make sure that costs exceed benefits? Check. But the analytic community for thirty years was focused on one problem in the Cold War. It is focused on a different problem today. And the fact that it is not crystal clear about that is, itself, a problem.

Cold War nuclear strategy

ROBERT SOOFER: How did the United States operationalize deterrence during the Cold War? What are the implications for the current situation?

WALTER SLOCOMBE: It seems to me there are distinct phases. The first was in the immediate post-Cold War period, where the United States had actually very few weapons, only fifty Hiroshima-size weapons, and the targets were all cities. This was a holdover from the Army/Air Force fascination with strategic bombing. And its problem is demonstrated by the question an Army guy at one of these briefings supposedly asked: I know you can destroy the fifty biggest cities in Russia, the Soviet Union. But how will that stop the Red Army? In some sense, that’s a question we’ve never answered.

The second phase is 1950 to 1955, roughly. The United States had a lot more weapons, a very big, stacked plan summarized by a naval officer at this time who said that it is a plan for reducing the Soviet Union to a smoldering, radiating ruin in about two hours. Holding back the Soviet military was added as an objective, but there was still a very heavy emphasis on urban-industrial targets. Next phase—and I’m sticking strictly to the strategic part of this problem, leaving aside the tactical—was the period as the Soviets began to develop significant forces of their own. Meanwhile, the US concept had not changed, but the target list was growing.

And just as a sideline, this was the period when the Air Force was finally coerced into fighting on the same side as the Navy—or, more accurately, letting the Navy fight on the same side as they did. Then the Kennedy administration, perhaps the most important development there was when former US President John F. Kennedy and then US Secretary of Defense Robert McNamara were briefed on the Single Integrated Operational Plan, their immediate answer was it was too big. And when they heard General Lyman Lemnitzer say, quite candidly to his credit, under any circumstances, even during a preemptive attack by the United States, it would be expected that some portion of the Soviet long-range nuclear force would strike the United States—this was a period of very large increases in both sides’ forces.

The next phase, roughly the Nixon years, former US Secretary of State Henry Kissinger—who had, after all, made his name by a book that was substantially about limited nuclear war—made the point that the president needed to have options other than Armageddon or inaction, some way to be effective but also to constrain escalation. And the directives were to create smaller options. Some options were developed—that was a period, at least part of it, when I was on the National Security Council staff. There were lots more numbers on the briefing slides, but they were all pretty big, unless they were extremely little.

Then the Carter administration is, of course, most famous for Presidential Directive 59 (PD-59), which, partly because I was involved, I think was the first serious effort to ask what would deter the Soviets. And the conclusion was that the risk of losing the things the leader managed most (Communist Party control, conventional military power, and, importantly, ethnic Russian hegemony in the Soviet Union) would deter the Soviets. And this was the theory of articulation of the point that Amy made, that the objective is to make sure the enemy thinks that the costs and risks are graded in turn.

Over this history, I think there are a number of common themes. One, and I think it’s very important, is what I call presidential revulsion. I would assert that every US president, including Harry Truman, and even Donald Trump too, has just been appalled when they learn about nuclear weapons and wants to figure out some way out of this bomb. And so the search is to have credibility for meaningful attacks, but ones that also don’t force escalation.

The second common theme is the very weak link between theory and practice. It is remarkable the degree to which different doctrines, written up very elegantly, had essentially minimal impact on the actual planning. Now, this was partly due to massive nontransparency in the process. The most extreme version was the early days, when the Navy and the Air Force literally didn’t know each other’s target. Thus, the search for options with limited results, partly due to military resistance was, to quote Kissinger, “always short of being insubordinate but also short of being useful.”

Another theme is rapid early response. That is to avoid preemption. But it also requires preplanned options, which are very inflexible. And it is a capabilities-based approach to cover targets that are identified only in the most general terms and that the military-urban distinction is quite artificial.

Not totally artificial, but even the military can see that the co-location of facilities, particularly industrial facilities, is such that it is hard to distinguish an attack on military and industrial targets from an attack on cities as such. One important point was the absolute assertion of the proposition that only the president can reduce and authorize the use of nuclear weapons. And pretty firm pushing back at any effort to change that.

There’s a short epilogue post-1990. Very different. End of the Cold War. But it was the time when for the first time, under Franklin C. Miller’s leadership, there was real civilian access and detailed review, drastic cuts in the weapons allocation, and drastic cuts in the number of weapons.

On the United States’ nuclear strategy, mutually assured destruction (MAD) was accepted as a test of sufficiency. It was not used—to put it mildly—either for planning purposes or procurement. There is a mark of continued aspiration, fascination with the idea of going first. So if you’re going first, you’re going very quickly, so that you can reduce damage. None of the schemes, I think, had a whole lot of value.

On the United States’ current problem with tripolarity, I think there is an interesting analogy to something that runs all the way through the history. That is to have what was called in PD-59 a secure reserve force. And in many ways, a secure reserve force as a way of deterring, still by the threat of retaliation, but hoping that the enemy will avoid further escalation because of the prospect that we can still trump his ace. And I think that is in some sense simply a general case of what the United States has faced now in terms of deterring either Russia or China, if the United States is engaged in a war.

KEIR LIEBER: I would just dissent from the view that one strategic approach characterized US planning in the Cold War. In the first phase of the Cold War, US strategy entailed massive nuclear use at the outset of a Soviet invasion of Western Europe. And that persisted for quite a while. It was only until the early 1960s that the United States really got serious about a more flexible response, even though it’s still not clear whether a more flexible path would have been possible.

And then later in the Cold War there was a major effort to develop counterforce capabilities to target Soviet nuclear delivery systems, and an interest in first strike: that’s a reality. Diving deeper into the Cold War, it is becoming increasingly clear that counterforce efforts pervaded many different parts of the US government and military. It’s important to understand that US goals, doctrines, and postures were variable, not constant.

DARYL PRESS: The United States may be making kind of an analytic and theoretical mistake, which is, if the United States’ principal wartime task toady was the same as it was during the Cold War—meaning stopping somebody with conventional superiority from defeating us—then the question worth asking is does the United States still have the ability to employ nuclear weapons in some sort of limited, managed way, to try to reestablish deterrence, the Cold War objective? But if the objective now is to be able to impose a defeat of some fashion on nuclear-armed adversaries—whether that’s Russia in Ukraine, or China in Taiwan, or North Korea in some war—then simply being able to continue to impose costs upon them as they’re escalating might not be enough, from a theoretical standpoint, because the costs to them of accepting defeat are so very high.

So that’s why at least it raises the possibility that the questions the United States should be asking today are different from the ones it was asking in the 1970s and 1980s. Whereas then it was about can the United States use and manage, now it’s what can the United States do to stop them from using as the United States is winning the conventional war?

WALTER SLOCOMBE: The right question to ask is how does the United States make them believe that they have more to lose in the first place by starting at all? And that’s very important. But in the second place, that they have more to lose after they use nuclear weapons? One place that I think it would be interesting to do more research is tactical nuclear weapons. For various reasons, maybe it’s just the National Security Archive hasn’t gotten around to doing it, I was able to find a great deal less information about doctrine in theater in the Cold War period than strategic.

AMY WOOLF: Nuclear strategy and employment policy are very, very, very different things. Ninety percent of this conversation is about what the United States says about how it would employ nuclear weapons, and what it would seek to achieve by employing them. But what came out of Joint Strategic Target Planning Staff (JSTPS) and STRATCOM and the actual planning had very little variation to it over time.

Where anytime you go and talk to JSTPS, even now probably going out to STRATCOM, they do their job systematically and systemically, but the doctrinal reports and papers that are sent out don’t always match. So in the two-peer environment what the United States says about what its goals are may be extraordinarily unlikely to change, because its goals are general enough.

The United States wants to restore deterrence at the minimal level of damage and keep the adversary from controlling escalation. There should be no newness to those goals. But how the United States would actually plan and employ the use of nuclear weapons may change. And that gets to the question of whether you’d see those changes if you went to the Targeteers and saw how they implemented change. Because I once asked General Lee Butler why he became an abolitionist. And he said it was because nothing ever changes. The Targeteers will go straight to all-out nuclear war no matter where you go. So it’s two separate questions.

WALTER SLOCOMBE: I think the idea in PD-59 was to have some options for use that were clearly less than all-out. My favorite was always attack the Soviet forces defending the Chinese border. Something that they would really not like very much. It wouldn’t be fatal, but it would divert their attention. And at least conceptually, that’s the sort of thing that it seems to me what the United States should be looking for.

It’s a very hard problem, in the case of Ukraine. I’m not sure it’s a harder problem in the case of having a post-exchange or a post-crisis, which doesn’t have to use nuclear weapons, with a peer when you’re still concerned with deterring the other peer. And, of course, they’re not independent variables.

Today’s nuclear force posture

ROBERT SOOFER: Today the United States has options that are extremely flexible. And whatever is not on the books, the president can order up, and the country can make happen. So the United States can use nuclear weapons in the most limited and flexible manner possible. So I would suggest that US operational policy, engagement policy, and capabilities are very much in concert now with a strategy of limited controlled use of nuclear weapons.

KEVIN CHILTON: First, I’m not sure our options are as flexible as they need to be in the current or pending environment. They certainly aren’t as flexible as they were in the European theater in the 1970s and 1980s. As we look forward, suicide cannot be the president’s only deterrent option. By that I mean it would likely be intolerable (or just as bad, unbelievable) to attempt to deter a regional conflict with threats of a strike on the homeland of a near-peer adversary who would likely respond in a similar, if not an escalated, fashion. And so not enough time, in my view, is being spent on examining the role of theater weapons, designed to be employed on targets other than those on an adversary’s mainland, in deterring a conventional conflict that the United States thinks it may otherwise lose. If, for example, the US combatant commander in the Indo-Pacific region assesses that US conventional capabilities cannot deter or defeat an invasion of Taiwan, then the United States should consider what theater nuclear threat might sow doubt in a Chinese assessment of their ability to achieve unification by force without provoking a strike on the US homeland. Additionally, we should consider what role theater nuclear weapons might have in deterring the first use of nuclear weapons by China in a case where China begins to lose a conventional conflict it initiates against Taiwan. Also, theater-based nuclear weapons are not just escalation control weapons and deterrent weapons; they can also be meaningful assurance weapons.

Secondly, the United States strategic deterrent force is essentially a dyad today. The bomber force remains flexible but difficult to put on alert. Indeed, we should take a serious look at what it would take to put bombers on alert in adequate numbers to credibly signal resolve or, in the more stressing case, backfill the loss of one of the other legs of the triad, whether it be for technical reasons in their warheads or delivery vehicles. This should include examining tanker requirements as well as security force, maintenance, and alert accommodations for crews and weapon storage area requirements.

Today, our nuclear weapon enterprise cannot manufacture our way out of a geopolitical surprise like the rise of a second near-peer competitor, which I would argue is no longer a surprise. It’s a reality, as we see China growing its strategic forces. There are those who say that there is no need to change the declaratory policies, force structure, and posture of US nuclear forces as China becomes a near peer. I think that is a sign that the United States hasn’t thoughtfully considered the potential impacts of this important change. We likely will need more flexible capabilities and employment options than what we have today. Examining what it would take to reinvigorate the readiness of the bomber leg is one thing I think the United States should be considering in this regard.

ROBERT SOOFER: But with respect to US strategy, is there any reason to change from this notion that the United States should have controlled, limited, flexible use of nuclear weapons with the purpose of limiting escalation short of strategic nuclear exchange? Is there any reason to change that strategy?

WALTER SLOCOMBE: There really is a formula for making any nonsuicidal adversary realize that the costs to them will be greater than the gains. But I’m in favor of what you call a more tailored, more limited response. Also, it’s important to remember on the tactical front, where I think some of the answers will lie, many tend to think of tactical nuclear weapons as platform-based. The issue really is not where the weapons come from, but where they go. And I would have thought one of the arguments for keeping the bombers, even if they’re limited use, in a large-scale option is that they have some very clear advantages.

DARYL PRESS: I’d just say again, in a different version, if the United States is trying to deter an adversary for reaching for nuclear weapons while the United States inflicts devastating defeats on them in a war, then I think the ability merely to respond in kind and inflict additional pain on them doesn’t make sense analytically. If you’re the North Korean government and the consequence you see coming down the path from this war is that South Korean forces are going to end up in Pyongyang, and you’re either going to get hung in Pyongyang or across the border in Beijing, the fact that United States might respond with nuclear weapons if you escalate shouldn’t, logically, give you any additional fear.

At least for the nonpeer competitor cases, if you want a coercive ability to try to convince them to keep their nuclear forces on the table while they’re suffering terrible defeat, you should have response options that do more than punish, but probably disarm. And that’s the argument for having at least good counterforce capabilities in those cases. Whether that’s feasible or desirable against China or feasible or desirable against Russia is a different can of worms.

The United States might lose a conventional war over Taiwan. It might. But it also might win overwhelmingly. It might. And if the Chinese Communist Party is faced with a catastrophic, embarrassing, humiliating defeat, leading to the loss of a big portion of their territory, China might have incentives to escalate. And simply the ability to inflict suffering in response might not be the right coercive mechanism.

To me, the case is clear, with regard to little countries, like North Korea and, in the future, Iran. It’s less clear, but the United States just can’t fall into these same ways of thinking that says coercing them from escalating is the same thing we were doing in the Cold War, because it’s not.

I think China sees the nuclear balance very differently than a lot of people in the United States do. There is a reason why China is adapting its nuclear forces in a hurry, the way they are. And I believe they perceive serious threats to their nuclear arsenal. So there is a tradeoff here. Building additional disarming capabilities against China solves one problem and leans into a different one. And the different problem it leans into is it generates the kind of responses being seen from them right now. And it’s true.

On the other hand, accepting a condition of MAD—not a policy of MAD, a condition of MAD with China—leans into a situation where they do have escalatory options to avert the worst sorts of conventional defeats. At the end of the day, at some point the United States has to give up the ability to have, I think, superior options over countries when they’re powerful, rich, and technically capable enough. But how far up that adversary continuum do you want to go in terms of building disarming capabilities, such that you can not only deter conventional war but you have a chance to deter adversary escalation in such a war and to mitigate costs to US allies and forces if that happens?

And simply saying it’s all one thing, imposing costs, it mixes all the nuance and it leads you into this situation which is giving leaders options that are no greater than different forms of defeat and disaster. And that would be a failure.

KEIR LIEBER: Can the United States do a counterforce disarming strike? That’s a calculation it needs to think about for a country like North Korea. But for Russia and China, even if that’s not possible, which is worth debating, that’s going to be an ongoing conversation. But is a limited nuclear response to, say, Russian nuclear escalation in Ukraine the right doctrine, the right approach, the right policy?

Even a limited nuclear response—what does that accomplish? If you believe that you are not willing to go up the escalation ladder all the way to all-out nuclear war, and if you believe that Russia has a more vital stake in the conflict in Ukraine than the United States does, then I’m not sure whether even a limited response makes any sense in that case.

KEVIN CHILTON: I think there is value in being able to respond in kind, and perhaps even a little bit larger than in kind, in such a circumstance. And at least, if nothing else, having the ability to do that, and threaten that, I think serves a purpose for not only escalation control, but for fundamental deterrence.

And if all you have in your response options are “sledgehammers,” then the credibility of your deterrent is reduced. Because oftentimes a bully will think, surely they’re not going to pull the sledgehammer out if I just kick them in the shins. But, surely the bully may be wrong, which is dangerous for all sides. So I think there is value both in deterring first use, but in also escalation control, by fielding those capabilities.

In the Korean scenario, I agree if Kim Jong-Un thinks he’s going to die no matter what, then he’s probably more likely to do something radical. But there’s an assumption there that the United States would drive all the way to Pyongyang and drag him to the gallows to stop an invasion of South Korea. I think that’s a bad assumption.

If what the United States is trying to do is preserve South Korea, there are choices that can be made in a conventional conflict that perhaps do not put him in that corner. And the United States needs to think about that when it considers how to posture, structure, and consider its policies with regard to nuclear use.

WALTER SLOCOMBE: It would be a mistake to assume that all wars, even all serious wars, are wars about literally national existence. I mean, for Ukraine it is, but it’s not for Russia. Taiwan is important to China, but the Chinese Communist Party has been ruling China quite successfully since 1949 with no control over Taiwan. It’s not that it’s impossible to create a situation in which the Chinese, however reluctantly, come to the conclusion that pressing an attack on Taiwan—or even starting down that path—is not very much in their interest.

Extended deterrence and assurance

ROBERT SOOFER: Making extended deterrence and assurance credible was the toughest job during the Cold War. So how did the United States do this during the Cold War? How should the United States use this moving forward?

HANS BINNENDIJK: Extended deterrence is harder than trying to deter an adversary from attacking your own territory; not just because it’s someone else’s territory, but you are also trying to deter both conventional and nuclear attacks. Having said that, it worked. It worked really quite well during the Cold War—no attack on a NATO ally or other US allies in Asia, no proliferation on the part of US allies, which would be a demonstration of lack of assurance. Contrast this with the Israeli case where the United States did not provide extended deterrence, and Israel proliferated to defend themselves and deter.

What were the characteristics of what I think was a very successful policy during the Cold War of extended deterrence? I would say there are four Ts. The United States had treaty arrangements with all of these nations. It had troops in the region, both to defend conventionally but also as—call them nuclear hostages. The United States had theater nuclear weapons there, so thinking back to the 1980 debate over Intermediate-Range Nuclear Forces (INF) deployments. But the United States had the nuclear weapons there, and it had trust. It had trust with the allies. And it worked hard on building that trust. It had all sorts of consultative mechanisms. It shared plans, policies, etc. So these four Ts together allowed the United States to get through the Cold War with extended deterrence working.

Then you had the next thirty years after the end of the Cold War, and what happened to extended deterrence there? It continued to be successful. There was a dramatic change in the threat from the Soviet Union to a greater focus on the Korean Peninsula. The United States had treaties remaining in place, by and large, troop deployments there—declining numbers, but still there. Trust became a little more difficult, especially during the Trump years, and the theater nuclear weapons were largely removed except for those delivered by dual-capable aircraft in Europe. So you had a shift there in that next thirty-year period, but extended deterrence continued to work.

And now there’s a new era where both the Russians and the North Koreans are threatening use of nuclear weapons, and there is a growing INF gap, both in Europe and in Asia. So we’re beginning to lose some of those four Ts, as I call them.

I’d say there are two questions. One is extended deterrence for treaty allies—does the United States need to make changes there to continue the successful extended deterrence as seen over the last seven or eight decades? I would say the United States doesn’t need a change of theory there. But it does need, perhaps, to implement things a bit differently and strengthen several of those four Ts, and maybe look at missile defense as another element.

And the second question is, what about extended deterrence for nontreaty allies? Ukraine, Taiwan, the obvious two. The United States doesn’t really have policies there. All of the Ts, if you will, are missing. And can you have any degree of extended deterrence without treaty arrangements, without troops, without trust, and without weapons in theater?

FRANKLIN D. KRAMER: I’m not sure that I would agree that extended deterrence worked. I would say it didn’t fail. And the reason I make that point is that there are a lot of things that come into play to determine whether or not a war starts, and it’s not just the nuclear weapons, so the United States’ conventional capabilities, the geostrategic set of arrangements, etc. etc. And so that may affect how we think about things.

The second thing I think is also important to say is when we talk about so-called theater nuclear weapons, whether a weapon is strategic or not I think depends on where it lands. So if we have a nuclear-armed sea-launched cruise missile, for example, that hits the Chinese mainland, from their perspective, that’s a strategic weapon. That doesn’t mean we should or should not have it. It just means that we need to understand that calling something theater doesn’t necessarily reduce the risk of escalation. It really depends on what you do with it.

So with respect to hitting their fleet in the water might be significantly less escalatory than hitting some appropriate—from our perspective—military target on land, and I would think, pretty certainly in my opinion, if we hit the Chinese mainland, they’re going to hit the US mainland.

On the issues of assurance and extended deterrence, I think it’s useful to outline a few things. One, what we did during the Cold War didn’t necessarily assure the allies. The ROKs and the Taiwans each had so-called secret nuclear programs. We eventually stopped them, but they certainly weren’t assured for a considerable period of time. There were some people down in Australia who would have liked to have moved—even a few people in Japan—not their government. So assurance was not ironclad—let’s put it that way. And certainly, that was not the case in Europe when Schmidt—I think it was 1977 if I remember right—raised the issue of the then-Soviet weapons—the SS-20 intermediate-range nuclear missiles in particular which led to the dual-track kind of approach, deploying ground-launched cruise missiles and Pershings, but also the INF talks. And then those eventually did succeed until, of course, the Russians decided to violate it.

So the lesson I think that one wants to take is that it’s a really bad idea to think about extended deterrence in and of itself. It’s part of overall deterrence; that is to say both conventional and nuclear. There is a high degree of importance in consultations and engaging the allies so that—the dual-capable aircraft I don’t think are very important weapons from an employment approach, but they are very important weapons in terms of maintaining the alliance’s working together. The Nuclear Planning Group is sort of a talking shop, but then the same thing—it is importance for alliance management, and it’s important, I think, to modernize in a way that the allies feel most comfortable.

And one example of things changing where that’s important, of course, is the German decision late last year to go with the F-35s, which I think was not a foregone decision until the Russians decided to invade Ukraine, and then Germany’s policy very substantially changed.

In Asia, we’ve got the Japanese, the Australians, the Group of Seven for that matter, the ROKs all talking about the importance of stability in the Taiwan Strait. And so, again, they have a high degree of interest in what the United States does in this regard including nuclear weapons, but not only nuclear weapons.

And I haven’t figured out really what to do about the DPRK. I don’t think it’s accurate to say that [Kim Jong Un is] not a rational actor; it’s just that the way he analyzes things is just different. And it’s the same with Putin. Putin is rational but he’s got an entirely different set of motivations and factors that he considers.

So to go forward, I think we should do some of the things that we are doing, which is to say modernize, have an adequate amount of weapons. I think we need to look hard at, with at least three—and maybe with Iran coming—four potential nuclear adversaries. Just do some of the analysis that is done out in Omaha, main points, survivability, those kinds of issues.

We need to think through missile defense as part of that. I think we need to step up the set of consultations. The administration says it is doing that. I think that’s the right way to go. And then, we need to think through hard where we stand on Taiwan. The lesson that a lot of people have taken from Ukraine—and the administration is concerned about escalation with respect to Ukraine—is that it’s not so necessarily likely that we would utilize weapons to defend Taiwan, and does that make it more likely or not that the Chinese would attack?

ROBERT SOOFER: Is there anything else that changes in the two nuclear peer environment—two nuclear peer plus because we’ll include North Korea—and is there anything else that the United States should consider to shore up extended deterrence/assurance?

KEIR LIEBER: Two things come to mind: The first is about the issue of characterizing leaders as rational versus delusional. I think this is a red herring. It’s a distraction from the real issue and about how we should go about analyzing it.

What I always say is let’s just assume they’re a rational actor—just make that assumption, then figure out what we would do, and what he or she would do, and then we can adjust things based on whether one thinks they are crazy, recklessly aggressive, emotional, etc. When I think about Putin’s potential nuclear use, China’s potential nuclear use, North Korea’s potential nuclear use, I don’t have to go to the “they’re crazy or delusional” argument to get to nuclear use. I mean, I completely understand what the incentives would be for Putin to do this. And I don’t think we’re doing ourselves any favors by writing someone off as just a crazy actor who does or does not operate according to rational deterrence theory. Let’s first figure out what a rational actor might do and then take it from there. That might be worrisome enough.

The second thing is that there is a tendency to delink the capabilities that the US thinks it needs for actual deterrence from the capabilities it thinks it needs for assurance. I’ve always said that what US allies want it to deploy is what they think will deter the common adversary—not anything different than that. And so it’s not like when the Tomahawk land attack missile was retired, when everybody described a favorite ally as saying that they really care about, or don’t care about, this one capability—at the end of the day, I think all they care about is what are the capabilities that are going to be most likely to deter. Are the dual-capable aircraft armed with B61-12s, is that the key thing to deter a Russian nuclear escalation? I think not. I’d much rather have a new nuclear-armed sea-launched cruise missile or the low-yield ballistic missile around if deterrence is the goal.

But my point is that’s a discussion about what we think will deter. And the thing that will deter is the thing that’s going to assure the best. We should be careful not to delink these concepts as if assurance is a separate thing from deterrence—that is, just some other psychological thing or calculation.

WALTER SLOCOMBE: I think in many ways the platform issue is much less about the Russians than it is about the allies and partners. The British minister of defence, Denis Healey, said the problem with assurance is it’s a 5 percent chance that it will work—may well be enough to deter the Russians, but a 95 percent chance that it may not be enough to reassure the allies.

The dual-capable aircraft had some unique capabilities—they’re not a waste—but they are primarily there because they are a symbol of nuclear sharing, they’re a symbol of our commitment. They are—this can cut both ways—they are physically in Europe, in country, and there are a number of things that I think come in that category.

ROBERT SOOFER: But what does that mean—except for the fact that we may need additional capabilities, but we need to be able to convince our allies that we are willing to use nuclear weapons on their behalf. We have to have those capabilities. That was the case during the Cold War, and that’s the case today. So we need to think of those types of capabilities that will deter the adversaries, and by doing so, now we will reassure the allies. That’s what we did in the Cold War, and that’s what we’re doing today. So what’s different?

HANS BINNENDIJK: Well, one of the things that’s different is that the four Ts that worked during the Cold War and weren’t needed during the next thirty years are going to be more needed in the future. And getting to them is going to be difficult. For example, in the European context, I think we would probably strengthen deterrence if we had some more theater nuclear capability in Europe. Getting there would be politically very dangerous. We do not want to have another dual-track debate—not at this particular point in Europe, given the unity we’ve been able to craft around Ukraine.

So I think the best we can do in the NATO context is to call for a review of nuclear deterrence, and do that quietly on the side with our allies. The worst thing we could do is tell them that extended deterrence is no longer as credible as it used to be. So I think that has to be done very carefully.

Asia is probably different politically. We can probably introduce theater weapons there with less political baggage. So that’s the first thing—getting back to the four Ts, if you will, and how to do that carefully.

And then the second thing I would say, there really is a difference in trying to deter attacks on non-allies. There is no theory for that, and if we’re thinking about new theory, that’s where we need it.

FRANKLIN D. KRAMER: It seems to me that some of the discussion implies that the way the Chinese—to just use one example—think about deterrence, and the use of nuclear weapons, and the like, is the same as the way we do. I don’t think we should take that as a given. I think that we need to think—we need to do our best to try and have a much better understanding of what the Chinese value, what they see as most important, how they would act.

There’s been a big difference in general since Chinese leader Xi Jinping took over with respect to the way the Chinese go at a lot of things, so the Xi approach is quite different than the approaches taken by former leaders Jiang Zemin and Hu Jintao. That may or may not make a difference with respect to how they would decide whether to go into a fight or what they would do once they were in a fight. But it’s certainly something that requires a lot of thinking—in my opinion—and not simply saying that we understand what would deter the Chinese, and I don’t think that’s right.

They are human beings, they are smart—they are really smart—but how they sort out what would make a difference I think requires a lot of thought.

It certainly seems to me to be a very important approach. And the same thing actually applies to Putin in a difference as I said earlier—that his calculations are, from his point of view, quite rational. He thinks that Ukraine is and should be part of Russia—we can go into a long rap about that—what risks he’s willing to take given that analysis, that starting point needs to be thought through, and what would make a difference.

I simply do not think that calling something theater makes it less strategic from the point of view of the adversary, whose country it lands in.

DARYL PRESS: I see changes, dramatic and potentially very significant changes, across the nuclear deterrence landscape. One is that we’re trying to deter nuclear use by adversaries who, in many cases, may feel they have regime threat levels of incentives to employ, even if they are rational. And that’s a different deterrence challenge. But that’s one that we’ve talked about considerably.

The second one is we’re trying to do deterrence in an era in which it’s not an exaggeration to say it is just unprecedented rates of technological change, and it’s technological change across the board in fashions that challenge the nitty-gritty of the nuclear mission. Obviously, accuracy revolution has gone through the roof, but the other shoe has yet to fall on the accuracy revolution, which is a world in which adversaries—countries with nuclear weapons also have substantial numbers of deliverable conventional weapons that pose major threats to their enemies’ nuclear forces. When that shoe falls, everybody from the United States rethinking land-based ballistic missiles to other countries around the world, and that’s the accuracy revolution.

The sensing revolution—it’s hard to talk about it in unclassified spaces, but it just changes everything. And even while people can say, oh, open ocean submarine operations will always—God has told me they’ll always be secure, let me say we’re arming ballistic missile submarines in fashions that make me think we imagine using them in limited war situations, and that’s going to be really hard in a world of really new sensing.

So number one was these new escalatory environments. Number two is technology change. Number three is the one you keep asking us to engage with, and we’re barely doing it, which is tripolarity.

But then there is number four, which is there is an era of—we’re used to deterrence in an era of nuclear plenty. We’re not in an era of nuclear plenty anymore. We’re in an era of 1,500 weapons. We’re in an era of fewer ballistic missile submarines going to sea with smaller numbers of tubes. And you’re in an era in which—although we have an air-delivered deterrent—I was in a meeting just a couple of days ago, and I was talking about the lack of tankers that will be available in US conventional operations against China because of the STRATCOM requirements, and they laughed me out of the room and said STRATCOM won’t have any nuclear requirements for tankers in a major conventional war against China. And I think that’s crazy, and I think it’s a big hole in our thinking. And I would agree with General Kevin Chilton, which is there will be a lot of STRATCOM requirements for tankers if you are fighting a conventional war.

So I think it’s away from an era of nuclear plenty; there are real constraints in the nuclear force. Tripolarity creates new problems. It’s a whole world of technology, and the interaction between conventional, and nuclear, and sensing, as well.

And then lastly is we’re in a world where, for the first time, sensible, rational, shrewd people have rational interests in using nuclear weapons in a conventional war against us as a way of keeping themselves off the noose. And you put all those things together, and I think—I think the environment’s changing.

KEVIN CHILTON: Just a few points I wanted to bring up that I think may be important in the broader discussion. One is we should always be careful not to mirror image our adversaries, that is, we should not assume that they are deterred by the same things that we would be deterred by.

The other is a caution not to assume China and Russia can be deterred in the same manner—or North Korea or anybody else. One of the fundamental things in talking about deterrence theory—which I do not think has changed—is that the key questions remain: Who (that is, what decision maker) do you want to deter? From doing what? And in what circumstances? Are you already in a conventional war? Is one looming? What are the circumstances that an adversary may be weighing? I don’t think the need for this type of analysis has fundamentally changed.

On the assurance side, we must remind ourselves that we don’t get to decide if someone else is assured. No. They decide if they are assured or not, and if they are not, then we end up with the undesirable effect of an increased risk of proliferation

And the last point I’d make is that I don’t think we’ve adequately thought through what is motivating China to building up its nuclear arsenal. I think one option that I haven’t heard brought up today is they may want to put themselves in a position to coerce us into not intervening conventionally in a Taiwan Straits or South China Sea crisis. Their previous posture, which was more of a retaliatory posture, appears to be growing into a first-strike capability and the threat of that capability maturing gives a different set of considerations to a future president as to whether or not he or she will come to the aid of our friends and allies in a conventional fight in the Western Pacific. Again, I think it is important for the United States to consider what roles theater nuclear weapons might play in deterring or countering that type of strategy.

And to Frank’s point that where the weapon lands matters, I couldn’t agree more. I’ve always found it rather cavalier on our part to assume that we would strike the Chinese mainland with conventional weapons in a Taiwan Straits scenario as a going-in assumption because I kind of have a sense how we might respond should our homeland be attacked with conventional weapons in this scenario.

Arms control and nonproliferation

ROBERT SOOFER: The United States has tried to reduce nuclear risks during the Cold War through a combination of nuclear deterrence, arms control, and nonproliferation. How successful were we during the Cold War and how might we use these tools going forward, if at all?

AMY WOOLF: So I do this spiel on how we should stop calling arms control arms control because when you do, everybody hears sign legally binding treaties that limit and reduce nuclear weapons. Please, may I talk you out of thinking those two phrases are the same thing, arms control and legally binding limits?

We, during the Cold War, did sign eight treaties and agreements with the Soviet Union and Russia that limited or reduced deployed numbers of nuclear weapons. But that’s not the fulsome nature of our arms control engagement and it’s not even a cause-and-effect relationship.

Most of the treaties that we signed that did require reductions in nuclear weapons were signed after we’d already decided to reduce requirements for nuclear weapons. So they codified those numbers.

Arms control does not force us to reduce nuclear weapons, nor can we expect it to force adversaries to accept reductions in their nuclear weapons unless they’ve already decided that they can get by with fewer.

So I’d like to disabuse you at the get-go from using that definition of arms control to answer Rob’s questions because I won’t.

What did we use arms control for during the Cold War? I have this nice fancy definition but mostly the core of it is to manage nuclear competition, encourage restraint in the numbers of nuclear weapons, and mitigate the risk of nuclear war, and we did it using a process of thirty out of fifty or forty out of fifty years of ongoing discussions, negotiations, consultations, things we called arms control negotiations that occasionally produced legally binding treaties.

But it was the process of engagement that is the sort of concept of arms control that we’re going to want to take forward in the current security environment where we have more than one peer with whom we would want to hold those discussions.

So what am I getting at here? When we sat down with the Soviet Union in the mid-1960s, we had very different concepts of what we would do with nuclear weapons, what we meant by strategic stability, and what our goals and roles were; and we’ve heard about that already today, that our deterrence problem was different from their deterrence problem, and even if we didn’t change each other’s minds over the years of pursuing these negotiations, these discussions, these engagements, this cooperation, we got a better understanding of what each other was talking about.

And that may not sound like a lot but it actually is because deterrence is a mind game and if you don’t know what levers to pull to affect the mind game then you’re not getting anywhere.

So if you think of arms control as a process of managing the nuclear competition by communicating, transparency, predictability, rather than the exact numbers of weapons that are deployed, you can see that it was quite successful during the Cold War to have those engagements with the Soviet Union and then with Russia.

If you measure it by limits on nuclear weapons, you would be far less sanguine about its success. For a period of time, we all agreed to limit our numbers of nuclear weapons. But when requirements change, as they are doing now, the willingness to limit numbers is going away.

So what does that mean for arms control, going forward? First off, do not consider arms control something you would do at a table with Russia and China at the same time. We have gotten to a place with Russia and the Soviet Union of having done this for fifty years we have a toolkit of things that we do and things that we accept to manage our nuclear competition and mitigate the risk of nuclear war.

We can’t expect our relationship with China to be in the same place. That toolkit that we’ve developed over fifty years with Russia isn’t going to be the same toolkit that we would use if and when we sit down at the table with China in an effort to manage our nuclear competition and encourage restraint.

We are not going to sit down at the table with China and start with legally binding treaties that limit and reduce nuclear weapons. We’re just not going to be there. The Chinese have said they are unwilling to do that.

We wouldn’t even know what to put in the treaty, and as General Chilton said, we don’t even know where they’re going with this modernization or why they’re getting there.

The first question we would want to ask them is not how many silos they are building now but how long they are going to keep at this and how many they are going to get to in the long run and why. The why is a really big question.

Now, knowing why may not affect anything about what we do next with them, but it will help us understand what our deterrence job is and what our nuclear requirements are. So the first step in arms control with China now is the same as what the first few steps with arms control with the Soviet Union were fifty years ago.

Figure out what language we’re speaking—and I don’t mean English and Chinese, obviously; I mean, in deterrence theory what language we’re speaking—what issues are of concern, what the roles and requirements of nuclear weapons are, and see if we can somehow manage this next phase of adjusting our requirements so that we don’t have misperceptions, misunderstandings, and inadvertent escalation.

So the role of arms control as a tool to manage competition in the 1960s and 1970s is the same role of arms control to manage nuclear competition today. We now have to do it at two different tables. But if you’re starting with that level of what is the role of arms control, that’s the same role.

RACHEL WHITLARK: One key takeaway from thinking about the history of the Cold War is that over time, two enemies, two competitors, managed to figure out how to cooperate or at least discuss areas of mutual interest despite the ongoing competition and contestation. That discussion was especially important in the areas where we were able to come up with standard-setting rules of the road to help us mitigate some of the most dangerous possibilities or at least work to improve those areas where there was ample opportunity for miscalculation. This type of behavior is going to be even more important in this more complex realm of the tripolar problem, especially since we’re not in a world at the present where we can expect formal treaties to be forthcoming with Russia or China.

Especially important is keeping areas of communication open and we’ve seen the US leadership recently move to try to keep areas of communication open with the Chinese, and I think there was some important news this week about the Biden administration working or taking steps to foster an agreement on the role of artificial intelligence in launching nuclear weapons.

So I think this is a good sign that we are moving in the right direction to at least try to keep conversation open despite what’s going on in Ukraine or other areas of friction in the dyadic relationships.

Of course, there’s more to do in thinking about the need to focus on the possibility of having US and Chinese agreement to alert one another if there’s going to be an ICBM test or have something akin to the hotline that we eventually stood up between the United States and the Soviet Union.

There are many things that we could do in this kind of vein to help improve the environment overall, to help cut down on some of the worst risks that exist.

The other item I would note is that it’s important to remember that arms control, however you want to define it, worked in conjunction with both nonproliferation and counterproliferation tools, including extended deterrence tools that Hans and Frank put on the table for us earlier. So there’s this broader toolkit to keep in mind.

The United States worked diligently, and especially with friends, to bring states into the Treaty on the Nonproliferation of Nuclear Weapons framework or to keep states already adhering to the treaty adhering to their commitments.

But we also worked to prevent the spread of nuclear weapons to other countries, thinking about the broad tools of counter proliferation to roll back programs that were in development.

So this toolkit includes the diplomatic portfolio, formal and informal naming and shaming in international organizations or otherwise; economic sanctions, offering economic goodies or threatening to take them away, working unilaterally or multilaterally with the economic tools in the toolkit; threatening or using military force, using force alone or with other partners, against third parties that were seeking nuclear weapons.

And in the modern era, we’ve also introduced cyber tools into the counterproliferation toolkit to infiltrate states’ nuclear development in order to delay or forestall someone’s successful nuclear acquisition.

The United States has used all of these tools individually and in tandem over time, and I think it’s helpful to think about how they are relevant today; perhaps potentially in need of modification, but still relevant to our new environment.

That said, it is worth noting that these tools worked in some cases and others not. For example, in the North Korean context, we’re still dealing with the fact that some of our negotiations didn’t turn out the way that we had hoped.

But I guess in the end I would note that there were famous expectations, including by Kennedy in 1962, envisioning a world where we would have twenty or twenty-five or thirty states armed with nuclear weapons.

And we didn’t just get lucky, as I don’t need to remind people in the room, that Kennedy’s prediction wasn’t borne out. But, in fact, it was the significant story of US leadership, diverse leadership on arms control, nonproliferation, and counter proliferation. So today it is critically important to have thoughtful people leaning in to think about the complexity of this new environment. It is equally important for US leadership to continue in this more complex environment, moving forward, thinking through the totality of the arms control, nonproliferation, and counterproliferation tool kit.

HANS BINNENDIJK: I really think you’re right. It’s a question of engagement, of process, of trying to find the points that could lead to real instability such as lack of knowledge about tests, or their inability to distinguish tests that we’re doing.

But I have a question for you. It took several decades of work by Walter Slocombe and others to come to an understanding with the Soviet Union about the nature of and the theory of nuclear stability, and once that theory was agreed to, out of it came a whole array of arms control proposals—SALT, START, the ABM treaties, etc. It all flowed from this kind of common approach, which actually now we are losing with the Russians. My question, though, is what would you envision that theory of stability might be with the Chinese? What is the theory of stability with China that the United States might be working towards as it engages?

AMY WOOLF: First, I want to dispute your premise. I agree we spent years, decades, trying to come to an agreed definition of strategic stability with Russia—with the Soviet Union.

There’s even a joint statement from 1990 between then US President George H. W. Bush and then Soviet President Mikhail Gorbachev which defined strategic stability as the circumstance where both sides have a secure second strike and no incentive to launch first with lucrative targets.

I’m not sure the Soviet Union or Russia ever agreed to that definition of strategic stability. Even inside the United States, you can find different policy elites who don’t buy that definition of strategic stability.

It was the operational definition that got us to START and START II. But calling it an agreed-on definition is a bit of a stretch. Also, you can look back over the arms control history and see that definition in operation in the treaties, obviously.

The ABM Treaty limited the numbers of defensive interceptors and, therefore, you could come to a limit on the number of offenses. You can see it in the treaties. But if you think about arms control, going forward, start at the beginning and don’t look back in time. Put stability on it.

What arms control really did was it brought about limits and restraint in areas where the two sides didn’t think they needed the weapons anymore. Either they weren’t cost effective or they weren’t militarily effective.

Missile defenses, even with limits on the interceptors, were going to be totally ineffective in a world of multiple independently targetable reentry vehicles. So you’ve got something that’s not cost effective and the scientists were testifying to Congress that they weren’t going to be technologically effective. We continued to buy them for leverage but giving them up was really not that difficult.

And I can go treaty by treaty by treaty and do that and I actually have done it in a few things I’ve written in the last couple years. But the point remains that the operational definition of strategic stability—and I’m going to be a little cynical—is our stuff is stabilizing, your stuff is destabilizing. We want to keep ours and get rid of yours, and you can see that in the arms control history, too.

But mostly what we’ve done in arms control is cleaned out the attic and the basement. We’ve agreed to limit those weapons that aren’t central to our requirements, and we’ve done so to great effect.

We’ve had lots of weapons limited and lots of weapons eliminated but only after we already decided we didn’t really need them. So if you want to extend that to how do we talk to China about this, they don’t have anything in the attic and the basement yet. They’re still filling the living room.

So our job right now is to figure out what they consider necessary—not how they would define strategic stability but what they consider necessary to their national security goals to put in the living room, and once we know what that is, then we can start talking to them about restraint and excess.

And you may want to define strategic stability as a way to figure out what restraint and excess would be but I’m not sure that’s really necessary if we can get to an understanding of what their deterrence problem is. I mean, what’s the problem and then how do you solve it rather than what’s the definition of the problem and how do you solve it.

I know that’s a little wonky and I’m a lot cynical, but I’ve been thinking about this a while.

ROBERT SOOFER: Let me just try to focus us on the near future. So we have a New START treaty that’s about to expire in February of 2026. What should be our objective with Russia when New START expires?

So, for instance, one example that I’ve heard and I support is a new treaty—maybe it’s not a formal treaty. Maybe it’s a political framework. I don’t know what it is. But a new set of limits that are higher than what we have today, right, so it accommodates the additional forces that we need to address China in the future and our regional deterrence gap but still limits, nevertheless, and you now capture all nuclear weapons, tactical as well as strategic.

So I’ll put that as a marker and see what people think the next step needs to be in anticipation of the expiration of the New START treaty. Yes, Walt?

WALTER SLOCOMBE: First of all, I totally agree with Amy that the main impulse for Cold War arms control was not doing things the country had already pretty much decided they didn’t want to do unless the other side did something. And that element is critical to be able to get agreement.

I think if we’re going to talk about what we’re going to do in future arms control with the Russians, there is this little matter of the war in Ukraine, and if we could somehow just persuade them that it was in both of our interests to hold the status quo—that is, to comply maybe without even at least all of the verification and transparency measures—that’s a good thing.

I’m perfectly willing to hear arguments. I don’t think the China-Russia problem requires more numbers. It requires other things. Notably, it requires very secure command and control.

The second is we’re kidding ourselves if we say we’re going to count all nuclear weapons and not do anything about defense. The Russians are crazy paranoid about defense and the Chinese are worse.

And I’m not saying we shouldn’t have defenses. I am saying that to have a negotiation in which the Chinese and the Russians believe we’d go hell for leather for missile defense even if we don’t want to and they’re going to agree on limits. . . here might be a deal but it didn’t do a whole lot more than just kind of maintain compliance. . .

But if we’re going to go beyond that, we got to be more realistic about what the other side might propose, and particularly if you want to say anything meaningful to the Chinese.

RACHEL WHITLARK: I think it’s worth keeping in mind the interconnectedness of all of these issues.

If we are in a world where we’re talking about raising the New START limits our allies are going to have views on that, right. So it may affect other nonproliferation considerations outside of the New START framework.

Similarly, if we’re going to be talking about defenses with the Russians and the Chinese, those conversations have implications for our extended deterrence commitments. Thus circling us back to some of those other challenges that we’re discussing.

I note this fact simply as a call for eyes wide open if we’re going to pursue those kinds of conversations.

AMY WOOLF: I would strongly advise thinking in terms of what do we do after New START as something that has to involve numbers and the reason for that is yes, if we want to count things, we should count everything.

Well, before you can do that, you have to define what the everything is, and then after you define what the everything is, you have to figure out how you’re going to keep track of that everything—what’s your monitoring concept.

Once you say it’s about a specific number, you have to be able to follow on with all that other formal stuff. So I would strongly argue against proposing that the future is a next round of keeping to a specific number.

There are a number of proposals in the literature right now about what next looks like, and just in response to Jake Sullivan’s speech on Friday—I was there—he said, we will talk to the Russians about two things: first, risk reduction now before 2026 and a follow-on to New START after 2026.

He talked about transparency in that follow-on and he talked about risk reduction, and the reaction out of some of the people in the room was, I heard good things about transparency and risk reduction but he didn’t say anything about reducing numbers or moving towards a world free of nuclear weapons.

No one in the Biden administration in any speech in the last two and a half years has talked about the follow-on to New START bringing about deeper reductions or keeping numbers. So they recognize that that whole package of a future of arms control isn’t really on the table.

But the thinking in the analytic community has a lot of different options about the United States and Russia just agreeing without any framework or monitoring or verification to stay within the restraint environment that they are in now, however you define that. Is it only on strategic? Is it on everything?

But it’s a stay restrained or—and once they stay restrained stay transparent, and all of those words are undefined but they all come from a point of view that we are not getting a next round of limits under a numerical ceiling because, as Rachel said, if that numerical ceiling is higher than the one you have now, even that’s because you’re counting more things, it’s going to be politically untenable amongst some of our allies.

If it’s lower than the number you have now, even if it depends on what you’re counting, it’s going to be politically untenable in the United States. So starting from the perspective that what comes next has to be a number that you define and then you monitor is just a nonstarter.

So you have to think about what comes next in more generic terms. Restraint, transparency. What do you mean? I don’t know yet. I’ll know it when I see it. But it’s not going to have a number in it, and if it does have a number the number is not going to be meaningfully defined or monitored.

ROBERT SOOFER: Amy, I think I agree with you in principle but the politics of the situation is going to drive us to a number. There are forces that want to maintain the current New START force limits. There are others that want to go beyond it, right. It’s going to happen whether we think it’s a good idea or not.

Look, arms control, essentially, is very much a political exercise, right. The Trump administration proposed a nuclear freeze on all capabilities with very little verification. So there are Republican administrations that have agreed to reduce from START levels to the Moscow treaty level without even a treaty, without verification per se.

So anything, I think, is possible but I think things are going to be driven in that direction to come up with some sort of an agreement about what happens.

The arms control community will want to see limits. The hard-line Republicans will want to see additional nuclear forces. And the only way you’re going to get both of those together is the traditional approach where you have an arms control track and a nuclear modernization track.

AMY WOOLF: I’m going to beg to differ.

First, the Moscow treaty didn’t limit or reduce anything. It didn’t have any definitions in it. You can do whatever you want. You don’t have to report on what you’re doing. There’s no monitoring. There’s no verification. On the day the treaty is due to expire, you tell us what number you count for yourself and then a minute after you do that it doesn’t count anymore.

So. . . that’s not a treaty based on numbers. That’s a treaty based on do whatever we want.

Besides, to go to my original point, the number we put in that treaty even undefined was the number we already decided we’d be happy with. So we didn’t do numbers, and the Trump administration stuff was we’re going to have a one-year freeze on the size of total stockpiles when in one year that number is not going to vary very much anyway. So, again, you’re using your numerical agreement to ratify reality, which is different from saying we’re going to have a number that is going to change our reality.

Also, you like to say that the people on the arms control side are going to force us to have a numerical treaty. In the current political environment that is a striking minority of people.

Even in Congress amongst people who prefer arms control they recognize that in the current political environment, we are not going to numerically limit or reduce our numbers of nuclear weapons.

So unless you tell me in 2026 we’re not going to be mad at the Russians anymore and everybody’s going to be good friends again, which is usually not required for arms control, but unless you tell me the political environment is going to change by 2026, I’m going to tell you that the pressure for numerical arms control is not going to be sufficient to force us to have a numerical treaty.

I agree that the pressure is there but relatively speaking, in the current political environment, it isn’t going to drive the day.

Further reading

Related Experts: Robert Soofer, Hans Binnendijk, Kevin P. Chilton, Franklin D. Kramer, Keir Lieber, Daryl G. Press, Amy Woolf, and Rachel Whitlark

Image: Russian President Vladimir Putin and Chinese President Xi Jinping arrive for a signing ceremony at the Kremlin in Moscow, Russia.