Good morning.  It’s a pleasure to take part in the first workshop event organized by the new Sandia Security Studies Institute (S3I).  I’m delighted to see the laboratory doing more to build out its work where policy and strategy intersect with the many extraordinary technical things Sandia does.  If I might borrow a turn of phrase from Dean Acheson, it’s an honor to be, as it were, “present at the Creation.”

I started my professional engagement with weapons of mass destruction (WMD) issues in dealing with issues related to other countries’ WMD rather than our own: specifically, in working on matters foreign governments’ compliance with arms control and nonproliferation agreements.  My first engagement with U.S. nuclear weapons policy was more indirect, through the prism of defending it – when I served as U.S. Special Representative for Nuclear Nonproliferation – against some governments’ incessant claims that despite the extraordinary reductions we had made since the end of the Cold War, the United States still needed to reduce our capabilities more, and faster.

It’s amazing how long ago that now feels, and how far removed from present-day circumstances and needs, for in light of a radically changed security environment, we’re now here to talk about requirements and challenges for new U.S. theater nuclear capabilities.  This is a really important discussion to have, so kudos to Pranay [Vaddi], Jason [Reinhardt], and the rest of you at S3I and at Sandia for organizing this event, and thank you for inviting me.

I can offer nothing more than my own personal opinions, of course, but I’m happy to provide some thoughts.  To that end, I’ll divide my comments into three baskets, first talking about the broad structural circumstances of today’s international security environment that I feel create a “demand signal” for new capabilities, then about some of the capability requirements that I think grow out of this signal, and finally about some of the challenges we face in meeting those requirements.

The Demand Signal

This audience of deep technical and policy experts certainly doesn’t need me to tell you about the demand signal, so I’ll be brief here.  There are at least four facets to the structural problem. 

The first is that we have more near-peer theater adversaries than before – specifically, two rather than just one.  When it comes to high-end conflict and especially its nuclear aspects, the world is indeed quite multipolar now, and this raises tremendous problems for us. 

Second, this multipolarity is particularly problematic because it isn’t symmetrical.  In terms of the “sides” arrayed against each other in the geopolitical arena, our two near-peers both posture in ways quite adversarial to the United States, while yet remaining on quite good terms between themselves.  (The phrasing China and Russia have sometimes used about a “no-limits partnership” overstates the case – for there clearly are some limits – but it still makes the basic point: they view themselves as being together against us.)  Today’s new environment of “multipolarity,” in other words, isn’t actually very multipolar; it’s closer to a “two-on-one” problem than it is to a truly three-way game.  And, unfortunately, we’re the “one.”

A third facet is our lost capacity, for in the nuclear sense, we’d still be behind the game even if this all were a more symmetrically three-body geopolitical interaction.  Pleased that the waning of the Cold War seemed to make such things irrelevant, we dismantled most of our own theater-class capabilities – along with our shorter-range “battlefield”-type systems – decades ago.  To be sure, they’re not all gone, for we have a small number left, some of which are forward deployed in Europe so as to be closer to the potential targets they would need to reach via manned tactical aircraft in a Russia scenario.  Nonetheless, we have precious few of these, and they’re gravity bombs, for Goodness’ sake. 

What little we have left is thus a capability that probably doesn’t frighten (and hence deter) the bad guys too much, whereas they have either retained or built sizeable arsenals of theater-class systems that give them a wide and diverse range of capabilities, as well as a significant advantage in raw numbers.  We may perhaps be able to rely upon sophisticated conventional capabilities to partially make up for some of this gap, and cross-domain deterrence can indeed be valuable. 

Nevertheless, conventional-nuclear substitution only goes so far, especially when our traditional post-Cold War advantages in conventional warfighting capacity are being so rapidly narrowed, especially by the Chinese.  Moreover, even if we were to succeed in besting our adversaries on the battlefield with conventional might alone, the prospect of such a defeat might itself drive them into the limited use of nuclear weapons there to turn the tables against us – hence presenting us with escalation management challenges that we might be unable to meet without serious theater-class nuclear capabilities of our own.

Fourth – and this unfortunately has to be said – we are also making these lopsided multipolarity problems much worse for ourselves by weakening the alliance relationships upon which we would ordinarily be able to rely in meeting adversary threats through the aggregated power of U.S. friends and allies acting in concert with us.  Too many people in Washington seem today to regard our allies and partners as nothing more than millstones around our neck and grifting free-riders who hide behind our military skirts while we do the heavy lifting of deterrence, but this is neither accurate nor fair. 

Yes, alliance management can be a colossal pain, and no, even our closest friends have agency and won’t always simply do everything we urge them to do.  (They are sovereign states, after all, not captive vassals.)  And it’s true that our allies rather badly need to do more – much more! – in their own defense than they have hitherto done.  Fundamentally, however, we still share powerful security interests with them that include not seeing Russian domination of Europe and Chinese domination of the Indo-Pacific.  This makes their power a powerful potential complement to our own, and makes it imperative that we work together. 

We do not live in an “America can handle it alone” world anymore, and despite the many frustrations they can and do cause us, our friends are, on the whole great assets to our geopolitical power – at least, that is, if we don’t screw things up by blundering about destroying their longstanding trust in our leadership and even making them see us as a threat.  But even this would still create a need for more capable U.S. theater nuclear forces, for as long as we care about deterring Chinese and Russian aggression at all, our eroding ability to work together well with our allies would force us to lean even more heavily upon nuclear tools.

These factors all add up for a rather significant demand signal for more diverse and more numerous U.S. theater nuclear forces. 

To be sure, I can’t imagine that we need to meet our adversaries’ capabilities on a one-for-one basis.  Far from it, most likely, for the uses for which they have sized their theater arsenals are different from the reasons for which we ourselves need such tools.  They seek nuclear leverage with which to facilitate attacking their neighbors without American intervention, while we seek merely to make such invasions and nuclear coercive bargaining unfeasibly problematic.  The kind of “damage limitation” needs that present themselves at the strategic level – that is, the need in a potential full-scale exchange to hit as many adversary nuclear assets as possible before they can be employed against us – do not so compellingly apply in theater, and I don’t think we need as many theater weapons as they think they do. 

That said, however, we now clearly do need more theater-class nuclear weapons than we have today, and we need a considerably broader range of such capabilities than we currently possess.  And that’s why we’re all here at this event today.

What We Need 

So what do we need in response to this demand signal for theater forces?  To put it bluntly, we need to put more “rungs” on our proverbial “escalation ladder” – that is, we need more capacity to answer adversary nuclear coercive bargaining and escalation manipulation (including limited nuclear use) with a flexible and diverse range of responsive options of our own.  We need more diverse and flexible forces to avoid being caught in the deterrence-erosive dilemma of having to choose between theater-level nuclear impotence, on the one hand, and potentially world-ending strategic overreaction on the other: we seek to avoid putting a future U.S. president in the position of having to choose between “suicide and surrender.”

            The Imperative of More Nuclear Options 

Such flexibility need not rely exclusively on nuclear tools, of course, and new well-chosen conventional capabilities can surely help too.  Some of the answer, however, simply has to be more nuclear capacity.  After urging this for some time, I’m thus gratified that the United States now seems to be genuinely interested in expanding its theater-class nuclear toolkit, and that the U.S. Congress has even passed a law requiring the creation of a “rapid capabilities program” for purposes of “develop[ing] new nuclear weapons or modified nuclear weapons that meet military requirements.”

Doing something useful is generally better than doing nothing, so I’m open to a variety of possible answers to this need for more capability.  My own favored approach is one I described in a recent War on the Rocks article, in which I urged the development of small nuclear warheads that could be optionally swapped into existing or planned long-range conventional precision strike missile systems of the sort that are already being developed and deployed. 

In that article, I explored this from the perspective of what Europeans might do if they conclude that they can no longer rely upon U.S. provision of “extended” nuclear deterrence.  Nevertheless, the same dual-capability idea would work at least as well as an American undertaking, and I can think of no better way to provide more theater nuclear capability more quickly, and less expensively.  (Such an approach might also form the basis of a “Nuclear Sharing 2.0” system in which we make U.S. nuclear warheads available for potential deployment on a broad range of hitherto conventional allied missile systems on terms much like those under which we prepare for the wartime possibility of letting NATO pilots deliver our B61 gravity bombs.) 

Tailored Deterrence vs. “Satisficing” 

In terms of the weapons themselves we might thus want to build, I suppose one would ideally seek to “tailor” our posture to the particular needs of deterrence, escalation management, and (potential) warfighting against the specific adversaries we face.  This might mean, for example, assessing what your adversary most wants to protect and then building tools designed to hold those things reliably at risk – thus facilitating a strategy of “deterrence by punishment” to intimidate him into avoiding aggression.  Alternatively or additionally, one could identify targets the destruction of which would simply physically preclude the adversary from accomplishing his objectives in a theater conflict, thereafter building tools optimized for hitting them – thereby facilitating a strategy of “deterrence by punishment,” as well as acquiring capabilities that would be of benefit in actual warfighting should deterrence fail.

In practice, however – as I outlined in a recent article on deterrence theory in my graduate school’s journal Defense and Strategic Studies Online – it may be difficult to do such “tailoring” with great precision or reliability, with the result that it might not be wise for us to put all of our deterrence eggs, if you will, in the basket of high-fidelity “tailoring.”  The adversary’s thinking might be too unclear to us to permit good tailoring, for instance, or we might simply misunderstand him.  Or the needs of tailoring vis-à-vis Russia and its requirements vis-à-vis China might point in different directions, making it hard to accomplish both sets of goals optimally with a single force posture.  Or the criteria for what would constitute good tailoring might change over time, particularly whenever new leaders take over in Moscow and Beijing. You can see the conceptual problem.

For this reason – and in an approach I tried to make memorable by terming it Bargaining and Deterrent Arsenal Super-Sufficiency (or “BaDASS”) – it seems wiser to acquire nuclear tools flexible and numerous enough to permit the accomplishment of as great a range of different missions as possible, to some extent even beyond what we may think we currently need.  If there’s a “tailoring” metaphor here, it may thus be in abandoning the idea of a nice bespoke suit in favor of something more loose-fitting and durable: more Carhartt workwear, if you will, than Saville Row finery.  Such a non-tailored answer might not be elegant or attractive, but it would likely be adequate as clothing in a much greater range of unforeseen and potentially trying circumstances. 

Reversing some of the implied logic of nuclear “tailoring,” this would not be working from an identified need to a specific use concept, and thence to the deployment of a particular capability fine-tuned to meet that need.  Rather, it would develop a maximally flexible capability, and then find ways to make some of the various things it’s capable of doing support a strategy that is as coherent as possible in light of the challenges we face any given time.  The basic idea of tailoring is not wrong, in other words, but we need to do it more at the level of strategy, declaratory policy, and behavior than at the capability level.

To my eye, as applied to the challenge of theater weaponry, I’d think we need systems with variable explosive yield, high accuracy, stealthy or otherwise interception-resistant standoff delivery, a weight and size small enough to allow deployment on a broad range of delivery systems, and perhaps even an enhanced neutron flux (a.k.a. “neutron bomb”-type) option in case one wants to maximize crew-killing impact against advancing Russian units or Chinese ships trying to fight their way across the Taiwan Strait.   To be sure, of course, this is an idealized list of characteristics, and in keeping with my more general theme of what Herbert Simon called “satisficing,” we may have to make trade-offs when balancing against the also-important values of cost, speed, and scale of production.  But we do need to be thinking about concrete capabilities with an eye to the feasible maximization of our range of available responses in crisis or conflict.

With an eye to thus doing our “tailoring” primarily at the policy level, moreover, I’d also like to see us do more on the conceptual side of integrating theater and strategic effects to counter theater-class nuclear coercive bargaining or escalation games by our adversaries.  This might mean, for instance, more work on cross-domain operational concepts, more serious nuclear-related coordination planning with more allies – including with the hitherto conceptually rigid and too often generally unhelpful French – and more wargaming and scenario-explorations that run notably “right of ‘Boom’” into limited-nuclear-use territory.  This is also an area where the unclassified, outside-of-government world of think tanks and academia – including my own School of Defense and Strategic Studies – can perhaps play a role in helping increase the West’s collective “deterrence IQ.” 

Addressing the Challenges

So what are the main challenges of making such a vision come true?  Put simply, they relate to money, time, technical design, and politics.

            Expense

First, cost.  While there are surely less expensive and more expensive approaches one could take, the development of any  new theater nuclear capabilities, even if prudently chosen, will have real expense.  And we shouldn’t pretend that actually paying such a cost will be easily done at a time in which federal spending seems to be more out of control than ever. 

Our national debt, for instance, has doubled under Presidents Biden and Trump to an appalling $40 trillion, creating an absolute monument to fiscal irresponsibility and a whole coop of ugly chickens, as it were, that will surely come home to roost before too long.  This horrifying debt burden isn’t primarily due to defense spending, of course – not by a long shot – for the federal government today spends more on Social Security, more on Medicare, and more merely in paying interest on the exploding national debt than it does in defending the American people.  Nevertheless, thanks to long decades of federal fiscal foolishness, making the case for spending more money on just about anything  may become harder and harder in the years ahead. 

The Need for Speed

Another constraint is time.  We were, I think, quite stupid years ago to hard-wire our nuclear weapons production infrastructure in place at a post-Cold War low that would quite predictably make it staggeringly difficult to build more weapons or develop new ones at any meaningful scale if the international security environment worsened, as it is now doing.  This is, in principle, a fixable problem if we’re willing to invest the money – and I’m pleased that we started to do so in the First Trump Administration – but it isn’t fixable quickly. 

In the short term, our ability to respond to new strategic challenges with more or new nuclear tools thus lies only in the “hedge” we have maintained in the form of mothballed weapons that could be taken out of storage and uploaded onto those U.S. strategic delivery systems with spare capacity.  That’s not nothing – and I’m also pleased that we made a forthright call for such uploading in last year’s U.S. Institute of Peace report on strategic stability.  Nevertheless, uploading is only a “Band-Aid” sort of a fix as long as our strategic adversaries remain convinced that they can ultimately win an ongoing arms race against us because we are unable actually to build anything more or anything new at any meaningful rate.

This means that as we move forward to try to address the problem of theater overmatch, we should place a premium upon designs that are capable of being produced not just relatively cheaply but also rapidly.  If this entails foregoing “exquisite” weapon capabilities in favor of ones that are simply good enough – buying Hondas over Ferraris, if you will – so be it.

Finding the Right Design(s)

From outside the U.S. national security bureaucracy where I now sit, it’s not clear exactly what actual weapons remain sitting in the long queue of devices awaiting dismantlement.  Which of these weapons might most easily, quickly, and usefully be dusted off, tuned up, and repurposed for the kinds of roles we’re talking about here?  Or what designs derived from the long legacy of our weapons program might be suitable for resurrection? 

            In Praise of Dual-Capable Systems 

For my part, I’m intrigued by some of the smaller devices we used to have – nuclear weapons which would surely fit into all sorts of modern conventional delivery systems with room and weight to spare, thus both leaving space for the clever interface gear that would be needed for nuclear “plug-and-play” optionality, and perhaps also making possible extended missile range.  During the Cold War, the United States built and deployed a whole range of smaller nuclear weapons, even in the form of munitions for eight-inch (203-millimeter) and even six-inch (155-millimeter) artillery such as the W79 and the W82 with very low yields of between 100 tons to two kilotons and the option of an Enhanced Radiation Weapon (ERW) neutron flux capability of the sort I mentioned earlier.  (Nor did those particular devices depend on the same sort of plutonium “pit” that we are struggling to learn how to produce again: it’s been reported that the old artillery shells relied on the compression effects of “linear implosion” rather than starting with our now-standard spherical pits.  Perhaps such a simpler mechanism could be more “producible” at tolerable speed and scale?)  As for reliability, we explosively tested a whole raft of old designs decades ago, including these, so further testing of a “resurrected” design might not be needed. 

Regardless of which specific type of smaller, variable-yield device we pursued, however, a modular nuclear warhead freely “swappable” in or out of precision-strike missiles would provide much greater operational and long-term flexibility than the nuclear systems – the Submarine-launched Cruise Missile-Nuclear (SLCM-N) and the W76-2 lower-yield warhead for the Trident Submarine-Launched Ballistic Missile (SLBM) – the United States is known to be developing or fielding to meet theater needs.  This is true for at least six reasons.

• First, because it would be “delivery system agnostic,” a modular dual-capable approach would offer us greater ability to meet future deterrence or warfighting needs that differ from what we presently anticipate we will face.  Having variable nuclear yield and effects with a warhead that could be swapped into different delivery systems would keep us from being stuck with a delivery system that might turn out to be not precisely what we need for deterrence at some future point, but with which we’d nonetheless be stuck because such “bespoke” nuclear-only delivery systems take forever (and cost a king’s ransom) to replace.  The “radical optionality” of such a modular system would be thus a terrific way to “hedge” against future changes in the security environment affecting operational requirements for theater nuclear delivery.

• Second, a system based around such dual-capable optionality could reduce the burdens upon scarce U.S. submarine assets – e.g., the attack submarines that would be needed to deliver SLCM-N – that would also have value in a high-end conventional fight.  If we could swap nuclear warheads into a range of other delivery systems in ways that met our needs, we’d have less need to devote precious torpedo tubes or Vertical-Launch System (VLS) capacity on our submarines to SLCM-N.  (Modular warhead units might, for instance, be carried aboard submarines as small, separate, insertable units without encumbering magazine capacity for conventional munitions, only being “swapped in” at the last minute if needed.)  This would, again, conduce to overall flexibility and optionality over time.

• Third, having “swappable” warheads would also increase U.S. leaders’ choices by giving them the option of generating additional nuclear force capacity in the field in ways that were either visible or not visible to the adversary.  This would allow strategic signaling options – or, alternatively, quasi-covert force generation – in ways not feasible merely with SLCM-N and the W76-2. 

• Fourth, being able to swap small nuclear warheads into new strike systems such as hypersonic missiles would permit faster response times against fleeting or otherwise high-priority adversary targets than the old Tomahawk platform used for SLCM-N, which travels at significantly less than the speed of sound – an operational deficiency likely to be particularly problematic over the vast distances of the Western Pacific.  (The W76-2 can be delivered quickly since it rides on the submarine-launched Trident missile, but it has the drawback of requiring that a visibly strategic asset be used for delivering theater effects, with potentially escalatory implications.) 

• Fifth, being able to deploy warheads on faster-moving theater systems would presumably also provide advantages in penetrating modern air defenses.

• Sixth and finally, having the capacity to swap nuclear payloads into a wide range of conventional-strike tools that we have already deployed – especially were we clever enough to engage in “shell game” activity designed to obscure exactly where (if anywhere) our insertable nuclear munitions actually were at any given time – makes any such system a potential nuclear delivery system, thus presenting the adversary with a dispersed and non-characterizable targeting challenge so insurmountable as to be more likely to deter aggression than to provoke it.  (By contrast, since Russia surely knows full well exactly where we forward-deploy our B61 gravity bombs in Europe today, does anyone think that those bases would survive more than a few minutes after World War Three began, or that the ease of striking those bases themselves might not – in the wrong circumstances – tempt preemptive aggression?) 

 “Entanglement” Risks are Overrated

Nor, as I explained in my War on the Rocks piece, would such U.S. dual-capable systems entail the kind of destabilizing “entanglement” problems that some scholars have claimed.  To begin with, it is only arguable that “entanglement” is destabilizing at all, since nobody seems to fear that the ongoing rain of Russian conventionally armed but dual-capable missiles on Ukraine actually signals the start of a nuclear war.  At most, the risk of mistaking an incoming conventionally armed dual-capable missile for a nuclear attack is merely contextual, not axiomatic. 

But even if you feel that it would be better if nobody had dual-capable systems, that’s not the world in which we live.  In our world, our adversaries have fielded lots of them for a long time – and that’s even if you don’t count multiple generations of fighter-bombers on both sides that can carry nuclear weapons if they need to.  If “entanglement” is a problem at all, therefore, it’s a worse problem if only one side had dual-capable systems – in which case, as I have written, 

“our adversaries might expect us to be at least somewhat ‘self-deterred’ by fears of attacking their dual-capable systems, while they would feel much freer to attack as many of our own (currently non-dual-capable) systems as they can.” 

Deploying our own dual-capable systems would likely reduce the degree of destabilization by ensuring at least that its impact is not asymmetric in ways that favor Russia or China.

But I don’t want to get bogged down in such hardware questions here.  I’m not a technical expert, and for all I know these mere speculations of my own from outside the weapons infrastructure may, from the perspective of a real weapons designer, be more embarrassing than enlightening.  Nevertheless, whether or not these specific musings make any sense, I think my broader point is sound: if we are willing to be creative, there may be a great deal of past learning upon which we can very helpfully draw to help answer the contemporary theater-overmatch problem – if not precisely cheaply and quickly, then at least comparatively so.

Programmatic Sustainability and the Politics of our Deterrence Discourse 

If good plans for new theater-class systems can indeed be devised and implemented, we would then face one additional major challenge: that of sustainability.  Our problem of theater nuclear overmatch developed slowly, as a result of relentless Chinese and Russian building programs, as well as the Kremlin’s willingness to engage in arms control cheating by not fulfilling its “Presidential Nuclear Initiatives” promises from the early 1990s and not complying with the Intermediate-range Nuclear Forces treaty.  Any “fix” to this problem will also take some time to be implemented.  Even thereafter, moreover – absent significant arms control progress that might occur but of which we cannot be assured – we would still be faced with the challenge of sustaining that solution over some considerable period of time. 

Ultimately, this raises the question of the politics of deterrence – that is, of whether we can sustain support for a robust deterrent posture over time.  I remain confident that the American people are capable of that kind of enduring seriousness, both about national defense in general and about nuclear deterrence in particular. 

To ensure such sustainability, however, U.S. leaders at all levels – from the most modest operational officer, think tank scholar, or academic all the way up to the president himself – will need to do a lot more to recover the ability to speak persuasively about nuclear matters, basing both their strategies and their associated public messaging on clear concepts and compelling logics.  This is the kind of intellectual endeavor we’re working hard to support at the School of Defense and Strategic Studies, but after the U.S. national security community has spent a generation or a de facto “strategic holiday,” it is not a quick or easy process.

Conclusion

I’m glad to see S3I starting to play a role in thinking through a way out of this labyrinth.  If you ask me, this is exactly the sort of way in which S3I should be contributing to the enrichment of U.S. national security discourse, and it’s a pleasure to be here. 

Thanks for listening.

—Christopher Ford