Greetings from sultry and steamy London, where it reached 94 degrees Fahrenheit yesterday, but where air conditioning seems to be a little too modern to have been particularly well assimilated.  I never imagined I would miss anything about the DC summer, but in this semiquincentennial year, God bless America, and God bless American air conditioning.

But more importantly, congratulations to all of you for being the 2026 Class of CSIS PONI scholars.  The Project on Nuclear Issues is a great program, and one with which I’ve very much enjoyed interacting for many years – going all the way back to my first tour in the U.S. Department of State, not all that long after PONI was first established in 2003. 

I know your program took rather a body blow when the Second Trump Administration foolishly cut its Pentagon funding last year, but it sounds like PONI has pulled through well, and it’s great to see all of you here.  This is a time of geopolitical challenges – including China’s massive nuclear build-up and our “two-peer” problem of having to maintain deterrence against Russia and China simultaneously – and the U.S. policy community badly needs a higher “nuclear IQ.”  And you are part of the effort to step up our collective game in that respect, and I’m delighted to meet you.  (Actually, one of you is actually a student of mine at Missouri State University’s School of Defense and Strategic Studies.  I’m delighted to see crossover between our two programs.)

I’ve been asked to say a few words today about how I see the future of nuclear arms control, and though my comments will just be my own personal opinions and won’t necessarily represent the views of anyone else, I’m happy to oblige.  I fear, however, that I don’t have a very optimistic message to convey, and I’ll even be suggesting – in effect – that the contemporary arms control community is at risk of getting in the way of arms control progress.  But I will walk through some ideas, and then offer you my perspective on where smart young folks like you should be devoting your energies in order to maximize the chance of a productive return to arms control.

Numerical Limits

Let me start by outlining two somewhat different types of arms control.  For many people, the most intuitively attractive and obvious form of arms control is the type that imposes numerical limits – that is, structured limitations on what types of nuclear systems we and our adversaries are permitted to retain and how many of them we are allowed to have.  Such numerical limits, however, don’t seem to me to be a very likely form of arms control at the moment.

            Russia Prospects

To be sure, Russia has proposed that it and the United States agree still to abide by the central limits of the now-expired New START agreement.  Nevertheless, my own feeling is that this would be a bad idea, since the longer we conform to that limit the longer we would be unable to start taking the steps we need to take in order to at least somewhat augment our nuclear posture in response to the “two-peer” deterrence problem. 

It’s worth reminding ourselves, in this respect, that Moscow has a long history of weaponizing arms control diplomacy by proposing things that would have the practical effect of destabilizingly locking in place strategic advantages for them. This was certainly the case, for instance, when the Soviets proposed a “nuclear freeze” at the end of the 1970s, after they had deployed their SS-20 missiles against NATO but before we had counter-deployed Pershing II ballistic missiles and Tomahawk ground-launched cruise missiles.  This was also the case when the Soviets proposed a “convention on the non-increase of military expenditures” after their own conventional arms buildup in the 1970s but when the Reagan-era American counter-buildup was just starting to get underway.  And it was the case, more recently, when Russia proposed a convention banning weapons in outer space only after developing an on-orbit anti-satellite system.  In all those examples, the purpose of Kremlin arms control diplomacy wasn’t to prevent an arms race but rather to prevent us from responding to their destabilizing advances.

So I don’t think it’s at all a coincidence that Putin proposed extending New START limits – and without that treaty’s verification provisions, by the way – only after it had become clear that the U.S. nuclear policy community was coalescing around the idea that we need to do at least some “uploading” of warheads onto our strategic systems in response to the “two-peer” deterrence challenge.  (Such uploading would not be permitted under New START limits.)  Accordingly, I’m glad that the Second Trump Administration didn’t take Putin’s bait.

As for some different approach to numerical limits with Russia, I guess that isn’t unimaginable.  But there is what I think of as a “China shadow” hanging over U.S.-Russian arms control, inasmuch as rigid limits with Moscow – as Putin’s disingenuous New START offer illustrates – would likely have the effect of preventing us from doing what we need to do vis-à-vis China as Beijing’s enormous nuclear build-up brings it closer and closer to strategic parity.

To be sure, I have speculated publicly about at least the theoretical possibility of a U.S.-Russian numerical limits deal were it to contain so-called “escalator clause” provisions that would help protect us against being outflanked by Beijing while in an agreement with Moscow.  One might, for instance, provide in a bilateral deal that if one of the parties determined that the nuclear arsenal or other capabilities of a third-party required it to increase its nuclear force beyond the numerical limit set by the treaty, then the cap could be revised upward at that party’s discretion, on the condition that its counterparty would also be permitted to increase its arsenal correspondingly.  That way U.S.-Russian arms competition could remain constrained by the treaty instrument – and the two of us kept at strategic parity – even while that instrument avoided locking either party in place vis-à-vis an aggressive challenger such as China.  This wouldn’t entirely crazy, though I still see little appetite for negotiations right now.

 China Prospects

More importantly, the prospects for numerical limits with that aggressive Chinese challenger itself unfortunately seem exceedingly unlikely.  It is hard to picture Beijing being interested in talking about numerical limits, if at all, until it has reached at least the dangerous strategic parity toward which it now appears to be sprinting.

To be sure, I think there’s nothing wrong, in principle, with a multilateral treaty that imposes numerical limits that differ between the parties.  After all, we shouldn’t assume that numerical arms control always has to specify strict parity. 

There is precedent in the Washington Naval Treaties of 1922, for instance, for a multi-party agreement which sets different force levels for different players.  There, capital ship limits for the U.S. Navy and the Royal Navy were set at parity, but for other signatories – including Japan – they were set at differing, lower levels.  A separate portion of the treaty package helped accommodate Japan’s rise as a major power by keeping Britain and America from building new bases in the Pacific or fortifying existing ones; this basically ceded the Western Pacific to Japan, which could enjoy predominance there because both the British and the Americans had to divide their forces between the Atlantic and the Pacific – both of them having global navies, whereas Japan’s security concerns were only regional.

This combination of provisions in the 1922 treaties made the differing overall levels acceptable, even to Japan at the time.  The package accommodated America’s rise as a first-rate naval power by basically recognizing that status by through the establishment of formal parity with the Royal Navy, while protecting the British from being out-built by the rapidly rising and increasingly wealthy Americans.  Japan, meanwhile, got the regional predominance it wanted in the Western Pacific, and wasn’t much bothered – for some years thereafter, at any rate – by having a lower overall naval limit because unlike the Brits and Americans it didn’t need to deploy in the Atlantic as well.

This makes clear that it is at least sometimes possible to have successful multi-tier multilateral arms control, and indeed that 1922 package deal lasted into the 1930s.  Things went rather badly off the rails thereafter, of course, but a decade’s worth of preventing an arms race in capital ships isn’t nothing. 

On the basis of that 1922 precedent, therefore, I could at least imagine a “trilateral” U.S.-Russia-China deal that was structured analogously, with a higher limit for the United States because we have to deter the two of them simultaneously, while they claim not to worry about each other’s nuclear arsenals at all, and indeed boast of their so-called “no limits” partnership against us.  In that sense, it’s only the United States that has a two-adversary-peer problem, and this extra challenge would need to be accommodated in the overall force limits of a 1922-style treaty.  (As for the precedent of recognizing a sphere of influence for a rising revisionist power in the Western Pacific, that makes less sense today than in 1922; the United States has more alliance structures and commitments there today than in those years.)

Nevertheless, it’s almost impossible to imagine China accepting this logic, and the complaints from Beijing (and Moscow, for that matter) about so-called “second class citizen” status would all but write themselves.  On the whole, therefore, I’m pretty pessimistic about the odds of numerical limits at the strategic level; I just don’t think China will negotiate.

But what about non-strategic forces?  Well, you sometimes hear speculation about the possibility of a numerical limits on theater-class nuclear systems.  Yet I think that would be quite unwise at the moment from a U.S. national security perspective, because those systems represent an area in which our adversaries currently have a huge and destabilizing lead – one that it would be a terrible idea to lock in place. 

If anything, in fact, we presently need more theater-class systems even more urgently than we need more strategic systems, since the greatest risk of war with China or Russia probably now lies in escalation from a regional conflict (e.g., over Taiwan or the Baltic States), and such a regional conflict is itself made more likely by the theater-class overmatch our adversaries enjoy, which gives them advantages in coercive bargaining and escalation management in a crisis.  Accordingly, we shouldn’t lock ourselves into this overmatch problem by agreeing to theater-level numerical limits. 

We should expect, of course, that if it starts to look like we’re actually getting serious about deploying new theater capabilities, Moscow and/or Beijing will propose some kind of a theater-class freeze – this being attractive to them precisely because such limits would undermine our own security by locking in our adversaries’ advantages.   But our answer to such proposals should be “no.” 

Risk-Reduction Measures

On the whole, I’m thus pretty pessimistic about numerical limits.  But what about risk-reduction measures – a second aspect of arms control that for a while took rather a back seat in our collective consciousness compared to numerical limits, but that can be very valuable? 

I think agreement upon risk reduction measures is easier to imagine than for numerical limits at the moment, but I still don’t see this as very likely.  In principle, risk reduction arms control might make a lot of sense, and there are a number of ideas that would be hard to gainsay.  One might imagine a multilateral launch notification treaty, for instance, to reduce the potential tensions and crisis escalation risks associated with ballistic or other forms of missile testing.  It would also be hard for Russia or China to object to this idea because they already have an agreement of this sort between themselves. 

I could also imagine a good case being made for other transparency and confidence-building measures (TCBMs) such as reciprocal nuclear test-site visits, improved crisis communications such as through the expansion of the existing Nuclear Risk Reduction Center (NRRC) network, routinized strategic stability diplomacy, regular structured data exchanges, and code-of-conduct understandings or “agreements – analogous to the Incidents at Sea (INCSEA) Agreement of 1972 – to control provocative behaviors that could lead to escalatory incidents.  Things like this would probably be valuable, and indeed U.S. officials have promoted such concepts for some years.

So far, however, our adversaries just aren’t interested in such risk-reduction measures.  I suspect this is because they don’t really actually to want to reduce nuclear risks in the first place.  After all, their strategies in many ways center on risk manipulation for advantage against us – that is, creating risk – such as through Russian employment of nuclear saber-rattling for coercive bargaining vis-à-vis NATO, or China’s apparent effort to use its massive nuclear buildup to achieve strategic parity and thereby create a “Mutually Assured Destruction” (MAD) standoff in which Washington would be more likely to be deterred from intervening to stop a Chinese invasion of Taiwan.  Both Russia and China also engage in deliberately provocative tactical behavior in an effort to intimidate U.S. reconnaissance or freedom of navigation operations in international waters or airspace.  Those powers are thus not much interested in reducing nuclear risks; they are interested in manipulating such risks against us.

When we ask for nuclear risk reduction agreements, therefore, Moscow and Beijing probably think we are urging them to reduce their ability to gain asymmetric advantage vis-à-vis U.S. conventional military power – and they’re not having it. This thus really is not a good era for arms control of any sort, I fear.

Making the Adversary Want to Negotiate

All this probably means that the most important question for arms control right now is not what sort of arms control might work, but rather the conceptually antecedent question of how to interest our adversaries in having any kind of negotiation with us in the first place.  Coming up with decent ideas about what one might do in an arms control agreement isn’t actually all that hard, and there’s a whole arms control community out there that is presently essentially unemployed, and which is really quite desperate to find negotiating opportunities.  But when it comes to giving the other side incentives to come to the table at all, this is something that the Western arms control community isn’t all that good at, or even interested in, doing.

My own feeling is that if there is a pathway back to arms control, it is very likely to lie through more robust nuclear postures, more investment in nuclear capabilities, and more forward-leaning U.S. and other Western approaches to nuclear weapons policy.  For arms control to emerge, after all, the adversary has to feel it in his interest to talk with us about limiting armaments.  And this means that the alternative to negotiating limits needs to look – to him – worse than does the status quo of not negotiating.  (If we cannot confront him with the possibility of less desirable outcomes than an arms deal, why would he bother talking about one?)

This is why I think it’s so shameful, for instance, that we have let our nuclear weapons and delivery system production infrastructure atrophy into such an attenuated state that we are now struggling with plans to increase our productive capacity to a mere 80 plutonium “pits” a year – a tiny fraction of U.S. Cold War production, which at times could apparently reach 25 times that figure – even as programs to replace and modernize existing U.S. delivery systems on a slightly-less-than-one-for-one basis are grievously over budget and behind schedule.  Under these circumstances, why would our adversaries negotiate arms limits with us?  We’ve already effectively capped ourselves.  What adversary will want agreements to prevent an arms race they think they can win?   

So it might sound a bit paradoxical, but I think the best thing we could do for arms control right now is to build up our capabilities – as well as our ability to build up our capabilities further, if you see what I mean.  But this insight points toward a pretty hawkish nuclear weapons policy agenda, and this doesn’t sit well with the modern U.S. arms control community.

Many members of the arms control community these days tend to scorn, dismiss, or simply neglect “how-to-get-the-other-side-to-the-table” issues.  They tend to want arms limitations and risk reduction, and many of them just can’t bring themselves to endorse the additional weapons and deliberate risk manipulation that we probably actually need – in a two-peer-adversary environment – to deter aggression, to cope with adversary coercive bargaining, to manage escalation in a crisis or conflict, to handle warfighting needs if deterrence fails, and to give our adversaries reasons to reach arms control agreements with us. 

The instinctual assumption of many in the arms control community seems to be that arms control is so obviously and transcendently essential that if we merely explain well enough or loudly enough to our adversaries how it’s really in their interest to negotiate – if we merely urge them to negotiate fervently enough – all will be well.  It confuses and perplexes them that our adversaries don’t see things as they do, but the idea of incentivizing negotiation by confronting the bad guys with elevated nuclear risk if they don’t come to the table remains largely anathema. 

This leaves many contemporary arms controllers all but paralyzed in the face of contemporary challenges, and unable to say anything constructive about meeting such challenges except by doubling or tripling down on ever more desperate calls for arms limits and ever more hyperbolic narratives about how the sky will quickly fall without them.  Needless to say, such discourse does nothing to convince the autocrats in Beijing and Moscow to take us seriously, instead simply signaling weakness and our vulnerability to nuclear coercive bargaining, which just makes things worse.  This is why I fear the contemporary Western arms control community may actually be getting in the way of making arms control progress. 

Accordingly, if I had some advice to give to you sharp young PONI scholars, it would be to worry less – for the moment, anyway – about what specific forms future arms control agreements might take, and more about how to get the bad guys to start caring about any arms control at all.  Specifically, we need to confront them with the prospect of an arms control-free future that they would dislike more than they currently dislike the idea of negotiating with us.  And that, it seems to me, is more a job for hawkish statesmen and strategists – and weapons designers and military operators – than it is for arms controllers.

This probably isn’t quite what you expected me to say, but there it is.  I look forward to our discussion. 

Thank you.

—Christopher Ford