Jeffrey LewisHeads I Win, Tails You Lose

I was reading this tendentious little op-ed by Keith Payne when I thought — “What a hypocrite!”

It took me about ten minutes to find Payne making the same argument that he’s now lambasting.

Well, that was easy.

Payne criticizes General James Cartwright (UMCS, ret.) and other authors of the Global Zero Nuclear Policy Report for making the argument that the security benefits of nuclear weapons — deterrence and assurance — are more psychological than material.  Here is what Payne writes:

For example, while the report calls for a realistic understanding of the post-Cold War security situation, it begins with, “Security is mainly a state of mind, not a physical condition.” Why this fatuous statement? Because if security is just a state of mind, old-fashioned security concerns can be banished easily by new thinking. But security is not mainly a state of mind; it often is predominantly a physical condition. Nations usually feel insecure because they are under threat or attack. Just ask the survivors of invasions, various genocidal campaigns and aerial bombardment or the folks in Syria who must dodge government attacks to survive. Real threats often underlie fears, and they require real solutions. Those who chalk this all up to “mainly a state of mind” and resist real solutions to real security problems often later are called “victims.”

Now, obviously what Payne really objects to are Cartwright’s recommendations for nuclear weapons strategy, forces and posture — a force of 900 warheads, with half deployed on a form of modified alert.

But Payne chooses to dispute Cartwright’s framing, which emphasizes the psychological aspects of deterrence.  The problem is that Payne makes exactly the same argument when its suits him.  Here is Payne, circa 2009:

Deterrence involves exploiting opponents’ fears and sensitivities and may have little or no connection to US preferences for the wartime employment of force for combat missions. Assurance, in turn, requires the easing of allies’ fears and sensitivities, which again may have little or nothing to do with how the United States might prefer to terminate a conflict. Whether US nuclear capabilities are regarded as useful or not “to fight or terminate a conventional conflict” may tell us nothing about their potential value for the political/psychological purposes of assurance and punitive deterrence. Deterrence, assurance, and war fighting are different functions with possibly diverse and separate standards for force requirements. The potentially different force standards for these different goals should not be confused.

It’s a nice trick. We cannot further reduce our nuclear stockpile because security is a function of material factors like the number and type of nuclear weapons.  Also, in the event that we don’t need all of these weapons, I would still like to keep them because security is not merely a function of material factors like the number and type of nuclear weapons.  Heads I win, tails you lose.

In the short-term, Payne has done rather well by such sophistry.  He sites on the STRATCOM SAG and his (awful) book is recommended reading in Omaha.  In the long run, though, his failure to articulate a coherent view of nuclear policy beyond “more, please” probably means he’ll ultimately be remembered mostly as the guy who was willing to put call 20 million dead US citizens a “Win.” (I once teased Keir Lieber and Daryl Press, both of whom I like tremendously, by asking if they dressed up as Keith Payne and Colin Gray for Halloween. If you think that’s funny, it’s because in contrast to their scholarly, if controversial, views, Keith Payne’s ideas are basically a costume that one tries on for effect.)

Too bad, really.  As it turns out, I believe this debate is the core question about nuclear weapons policy: How much do technical details matter for achieving the security benefits conferred by nuclear weapons? As I keep saying:

How many weapons are enough to ensure deterrence? How difficult is it to achieve and maintain deterrence? How important are the technical details of a country’s nuclear forces, such as the size, configuration, and readiness, to the goal of maintaining deterrence?…

One view, I would say the dominant view in U.S. defense planning, is that deterrence can be achieved only through difficult choices, sustained with intelligent effort, and will depend very much on the technical details. This is the view expressed in Albert Wohlstetter’s 1958 Rand monograph, The Delicate Balance of Terror, which helped to shape the dominant Cold War attitudes about deterrence.

A different view is that, beyond a certain point, all of this is crazy talk, and the technical details don’t matter very much at all. The balance of terror is anything but delicate. An enemy who can be deterred, will be deterred by the prospect of a counterattack, even if it consists of only a few nuclear weapons. Beyond that minimum threshold, nuclear weapons provide little additional deterrent benefit.

The idea that the security benefits from nuclear weapons are subject to sharply diminishing returns beyond a basic survivable retaliatory capability is hardly an unreasonable notion. If Payne has some reason to think otherwise, he should spit it out.

Cartwright et al. clearly view deterrence in this way, arguing that current US and Russian stockpiles “vastly exceed what is needed to satisfy reasonable requirements of deterrence between the two countries ….”  One might complain that the report spends too little time outlining those reasonable requirements, but their discussion of force size and targeting makes clear their implicit view.

Of course, Payne might have some reason that a few hundred warheads isn’t enough damage to deter Russia.  I’d certainly like to know what use he envisions for, say, warhead number 301.

There are, of course, other serious objections to such a posture.  The two most important, from my perspective, are whether we can develop plausible operational concepts for a minimum deterrent on which a credible threat depends and whether we can extend deterrence to our allies.

Payne has arguments he could make in favor of both objections to the minimum deterrent that General Cartwright advocates.  But he largely avoids doing so, I suspect because the answers to both are straightforward. The credibility problems with nuclear weapons are political and not easily subject  to alterations in technical detail. As for our allies, we provide for their security largely through conventional deterrence. It would be better to be frank about what role nuclear weapons do, and do not, play in their security.

In the end, it’s always seemed simple to me. What deterrence there is to be had, is achieved by a basic survivable capability to retaliate against a small number of high value targets. Beyond that, more is wasteful — and I can’t see how spending money to get less defense should be worrisome to our enemies or reassuring to our allies.


  1. Stephen Young (History)

    I was similarly unimpressed by Payne’s piece. In particular, I was dumbfounded his statement that unilateral U.S. force reductions are “the oldest and most discredited promise in the arms-control playbook.” That is an interesting statement for Payne to make, given that he served in the DOD in the Bush the younger administration, which made deep unilateral cuts in forces which led directly to the Moscow Treaty. It also ignores the record of Bush the elder, who unilaterally eliminated thousands of U.S. warheads in 1991, leading the Soviet Union to follow suit. I was even started tracking down Payne’s email just to point this out to him, but now I guess I’ll just post it here. Maybe he’ll respond.

    • anon (History)

      I’m not sure about Keith’s role in the PNIs (he really wasn’t centrally located at the time), but he played a critical role in Bush II’s NPR, and, if you remember, the recommendations out of that review were for unilateral reductions. The U.S. announced that it planned to reduce to between 1,700 and 2,200 warheads, and that it would do so regardless of whether Russia came along (this is all available in the transcripts from the NPR release and the Crawford summit). The Russians objected because they didn’t want to be treated dismissively, and the Senate objected, for the same reason. So we signed a treaty with Russia, which happened to look a lot like the press release from the Crawford summit. Keith Payne was the intellectual godfather of the unilateral cuts. He loved them, and sold them, based on the fact that they were reversible and allowed the United States all the flexibility it would need in the future.

  2. Kingston (History)

    I always enjoyed it when Payne would rag on New START despite the fact that the Strategic Posture Commission, on which he was a member, endorsed just such a modest first step.

    Isn’t the old saying about extended deterrence something like what’s needed to deter adversaries is less than what’s needed to assure allies?

    And since you mention them, Lieber/Press would say that the credibility problems with nuclear weapons are in fact easily subject to alternations in technical detail; namely by devising a force that places greater emphasis on low-yield weapons for use against counterforce targets. As you hint, I don’t believe a reduction in yield solves the political problem. And where’s the evidence that current US adversaries scoff at the W88 but would be terrified if the Air Force resumed work on the PLYWD? Not to mention the nonproliferation implications of trying to make nuclear weapons more “useable” against targets we can already attack.

    • Daryl Press (History)


      Reasonable questions, and a reasonable critique you raise.

      I’d say — and I think Keir and I are on the same page on this point — that in our view there is no quick, simple technical solution to the credibility or broader nuclear deterrence challenges that the US may face in the coming years.

      Specifically, if adversaries use NW coercively against us in the course of a conventional war to get us to cease operations (or if Pakistan does this to get India to stop) — then having capabilities to respond in a fashion we’d find less-objectionable than burning down cities or raining fallout across regions (the W88 options) would give us better retaliatory options. And that fact should enhance somewhat our deterrent, in the sense that if our arsenal only allows us to do things we REALLY wouldn’t want to do, reasonable observers (both adversaries and allies) might doubt our willingness to do it.

      On this point, our argument is very much what I think Jeffrey was alluding to when he wrote in passing about the importance of ensuring that whatever force one retains allows for “plausible operational concepts” — that’s exactly why Keir and I favor keeping in the arsenal forces that would allow US leaders to carry out the deterrent threats we’ve made — a condition we’re skeptical that W88s often provide.

      Last, I don’t think there would need to be costly effects on our counterproliferation policy — but that’s another post and debate.



  3. krepon (History)

    It’s elementary, my dear Watson:
    Accuse your opponent of the crimes you commit.

    • Cameron (History)

      We have always been at war with Eastasia.

  4. ajax151 (History)

    Isn’t the Washington Times a paper started by moonies?

  5. Daryl Press (History)


    Thanks for the kind words. Now, onto substance.

    Can’t one defend the Payne arguments by arguing — as I’m sure he would — that a robust nuclear deterrent aught to be able to clear two bars:

    a) Can the proposed force carry out the various missions which constitute the “or else” part of the deterrence equation? (i.e., the material calculation)

    b) Does the proposed deterrent posture — including the force structure — create the desired psychological state of mind among friends and foes (i.e., assurance and deterrence)? (i.e., the psychological calculation)

    If you satisfy (b) but not (a), then you run the risk that someone will figure this out and both deterrence and assurance may evaporate overnight.

    If you satisfy (a) but not (b), then you’ll have the right forces perhaps to wage war and conduct operations, but not to assure and deter.

    Regardless of whether Payne clearly articulated that we should strive to retain an arsenal that clears both the material and psychological hurdles, I think that’s his view, and I don’t disagree or see the problem with his argument.

    Of course, smart people will disagree about what’s needed to clear the bar on both metrics…particularly about (b) — namely what targets does one need to hold at risk to create the desired political effect (**in peacetime and in war**) on both adversaries and allies? I know that you hold the view that a “minimum deterrent” force can be postured and explained in a way that will clear both bars. Payne worries that our numbers may drop so low, and our nuclear capabilities may become so narrow, that we may start to lack the material capabilities to carry out the sort of strikes that our deterrence posture rests upon. To be clear, I worry about that too. But regardless, I don’t see how Payne’s view is internally contradictory.

    Thanks for the provocative post and kind words. (And BTW, last Halloween my 5-year-old son and I both dressed as frogs; it worked out well, but I’ll see if he’s interested in the costumes you suggested come next October!)


    • Jeffrey (History)


      I’ve just written one long comment, so perhaps I’ll come back to this in more detail.

      On the issue of (b) it just seems to me there are no objective ways to measure this that aren’t (a).

      If you look at recent accounts of Saddam’s decision-making, I think you’ll at least see why I despair of getting inside an adversaries head (or even those of our allies.) In practice, I think most policymakers do too, settling instead for satisfying themselves that they’ve done what they can. (Sam Nunn, incidentally, nicely expresses this outlook when he says, of the day after a nuclear attack, “What more will we wish we had done?” I think that’s the process for most policymakers.)

      So, let’s cut out the middle-man and just ask: What ought to be enough? It seems best to simply say we need to optimize our defense to provide the greatest chance of deterring an attack and prevailing in the event one should occur. If that’s not reassuring to our allies, I simply don’t see how a less optimal allocation would be. Keeping nuclear weapons we don’t need means forgoing other military capabilities we do.

      I am very uncomfortable with the idea of saying to Japan, in effect, “Look, we’ve cut our F-35 buy to make the remaining ones nuclear capable. Now, that’s completely useless, but such an extravagant waste ought to tell you how much we care. You are definitely not a lesser included case!”

  6. bureaucrat (History)

    You can disagree with Payne,I often do on numbers and the merits of A/C, but you mislead when you chose to conflate security and deterrence. Payne criticizes the notion that security is subjective not deterrence which he knows full well is.

    This is an important distinction in terms of unpacking his argument. With deterrence you are trying to shape the adversary’s actions and perceptions with security you are trying to persuade yourself (and possibly allies)that one has purchased a sufficient amount of capability to assure one’s safety. Subjective both and exposed to cognitive impediments for sure but one far more than the other.

    Also your consistent assertion that allies in East Asia are assured by our conventional deterrent ignores some empirical facts. The extended deterrence dialogues with Japan and South Korea were both begun post NPR and NST. That should tell you something about how the political leadership viewed these developments. Dig around and you’ll find where the most recent meetings were held. That should tell you something about how much they care about the technical characteristics of our deterrent.

    • Jeffrey (History)

      You’re making Payne sound smarter than he is, but I should be grateful. I’d rather have this argument with you. But I think you may be guilty of assuming you know what I think on a number of issues. In some cases, you are clearly mistaken about my views.

      (Payne’s oped is clearly intended to argue that “security” is the same as “deterrence” and “reassurance”, regardless of what he has written in the past. For example, he argues that allies “do not believe that their security problems are mainly in their minds. They confront real external threats and want the assurance of security that resides in the U.S. nuclear extended deterrent.” What is that, other than a comment about reassurance needing to be based in material factors in direct contradiction to his 2009 view?)

      In any event, I have a specific view on this issue which is rather closer to Payne’s oped than his other work. I referred to the “security benefits” of nuclear weapons, of which deterrence and reassurance are the most important. I am, in real life as well as work, a materialist on most important matters. So I tend to take a simple view: optimize your defense based on material factors, fudge the rest with consultation. I happen to agree with Cartwright et al, but largely because I _don’t_ think security is a state of mind. If it were, I’d view their reductions are deeply destabilizing. (Or, at the very least, a mixed bag..)

      My view is that allies ought to be (that is, should be) reassured by the actual capabilities that we deploy to defend them. If, as an empirical matter, our allies are not, then I can imagine only two sensible remedies — buy better capabilities (see my much maligned advocacy of a new casing for the B83) or hold “better” (more frequent, more detailed, etc.) consultations to explain how current levels provide for a sufficient defense.

      I have not said, as you assert, that our allies are, in fact, reassured by our conventional deterrent. They are not reassured in some important ways, something I blame on Keith Payne and others who have mishandled consultations about nuclear weapons and the role they play in allies security. I would not, for example, continue the practice of spending scarce defense resources on obsolete systems that make only pyschological, rather than material, contributions. We do not have a limitless defense budget and most strategic systems has external costs beyond funds expended.

      (By the way, you’ve said something very important — that much of what passes for thinking about sufficiency is really “to persuade yourself … that one has purchased a sufficient amount of capability to assure one’s safety.” I think that is perhaps the most profound pathology in how we approach deterrence, which dates at least as far back as McNamara looking for the knee-bend in the damage curve from dropping hypothetical 1 MT warheads on the Soviet Union. Its not a permissable part of our discourse to examine, but it is the most revealing.)

      As for the specific cases of our allies, I did plenty of digging, thank you very much, especially in Tokyo, right after the NPR was concluded. (For all I know, bureaucrat, you helped us set up our meetings. Some of your colleagues did.) I took Linton Brooks, Mort Halperin and Walt Slocombe to Tokyo for a week of consultations inside and outside of the government. That’s a pretty fine group that was able to meet with pretty much anyone you’d care to name. Correlation is not causation — the consultations in Japan began because of the NPR’s existence, not because it was deeply unsettling. (Let’s hold aside the Koreans for a moment, as your argument there is rather more accurate.) As I have argued before, my interactions in Japan continue to suggest to me that Japanese policymakers do not have a very clear or accurate view of how the United States provides for Japan’s defense. Instead of viewing capabilities in terms of their military contribution, Japanese policymakers tend to treat them as proxies for some vague judgement about American commitment and resolve. I suspect that owes in equal parts to Japanese reluctance to discuss sensitive issues involving nuclear weapons, American concerns about Japanese treatment of sensitive information and the fact that it is just plain easier to point to hardware as a sign of commitment. Does anyone remember the asinine commitment to keep 100,000 troops in East Asia? So, my view, to be clear, is that Japan is unsettled because we have avoided a frank discussion with them about the material factors that provide for their security. We did rather better with the NPR, a process which I hope continues.

      The Koreans are a more interesting case, where I have done a bit less digging. They are, it seems, unsettled by reductions in US nuclear forces as you imply. As best I can tell, this reflects a real awareness that North Korea is willing, perhaps more so given improvements in the DPRK’s nuclear weapons capabilities, to conduct limited offensive actions against South Korea. We have yet, I fear, to find the appropriate mix of military capabilities to prevent such provocations. Where perhaps we differ is that I see little value in new nuclear weapons for materially addressing this very specific problem. I am willing to do a sled test here or there, but the salient factor — as I said in the post — is that we have political constraints on the use of nuclear weapons, constraints that amount to a no-use policy. These constraints are not, I would argue, subject to being resolved with weapons of lower yield or other sorts of technical innovations.

      So, there we are. I still think Payne is internally inconsistent, though perhaps the oped doesn’t deserve equal weight with everything else he’s written. I can’t bear the thought of a comprehensive survey. I take a materialist line on deterrence and reassurance; I am not sure what you think. I admire your insight into the role of self-assurance in discussions of sufficiency. I disagree with how to interpret the anxiety expressed in Japan, though perhaps that’s right with regard to Korea. I don’t think forgoing further nuclear weapons help solve that problem.

  7. archjr (History)

    Terrific post and comments. I also happen to think Cartwright,, with whom I also agree, have nothing to fear from Payne’s perceptions, which is all his comments really are.

    Jeffrey, do you really mean our situation in Korea amounts to a “no-use” policy? Would be interested in an expansion of your thoughts on this one.

  8. jeannick (History)

    I’m not quite following the debate on the subjectivity of security ,
    certainly there is a spectrum between objective capabilities
    and subjective appreciations
    the United kingdom has the capability to destroy Washington , the risk appreciation is rather remote .
    of course the Brits are satisfied they will not be targets either.

    This debate is pretty much about Russia ,
    to make Latvia and Poland feel safe about Russia would take the total destruction of the whole country and occupation up to Vladivostok by a Baltic armored corps
    some allied are not quite rationals ,

    they also happen to be expendable , their fate wouldn’t justifies throwing nukes over the Arctic .
    the only rational use of nuclear weapons are for supreme U.S. interests ,
    protecting the homeland is the Alpha and Omega
    the rest is tacticals ,some glorified nuclear artillery

  9. Bureaucrat (History)

    Jeffrey, was going to leave this as it is, after all it’s your blog so you should have the last word, but some of what you’ve written begs a response.

    I’ll stipulate that I may have overstated your emphasis on conventional replacing nuclear by reading too much into the Post above (“As for our allies, we provide for their security largely through conventional deterrence”)and the 2010 piece that you did for Nautilus. You indeed caveat the above with “largely” and the Nautilus piece on second read doesn’t quite make the conventional for nuclear argument. I also agree that Payne, and Bridge Colby for that matter, both seem to have an appetite for new capabilities that provide only marginal, or niche advantage, and even then only at some substantial cost – both financial and political.

    Nonetheless, I continue to disagree with your characterization of Payne’s argument. You indeed found the one place in the piece that he uses the deterrence. Everywhere else he criticizes the Global Zero report for describing security in the same way that deterrence is typically described (“security is a state of mind not a physical condition”). This Global Zero argument essentially replaces deterrence with security – this is what Payne attacks not the metaphysical/mumbo jumbo nature of deterrence. Regardless of where else he may be misguided Payne is right to criticize describing security in this manner.

    Re: correlation is not causation. This is an analytic pathology that I am well aware of but that you misues here. Let me be clear, the NPR , as with the QDR, included an active plan for consultation with allies and key states. While I did not use the word unsettling to describe NPR consulatations both Japan and South Korea responded with concern about the functioning of extended deterrence in an environment where we emphasized the reduced role of nuclear weapons. I’ll avoid the materialism argument you make becuase I largely agree with it and because it lends little to the point I’m making in that these partners were indeed concerned (i.e., and we did not want them to be less assured than prior to the NPR).

    The formal extended deterrence dialogues that have been established with both Japan and South Korea follow directly from what we heard in the NPR consultations – concern – and are frank and specific. Indeed, in addition to the usual convening of meetings in capitols some of the sessions are held in locations that allow us to illustrate capabilities in detail.

    I’ll not comment further on your informed view of Japanese policymakers (lack a clear view of how we provide security) other than to say this is not what I have found to be the case but reasonable people can disagree. I don’t think new nuclear weapons are required, at least in terms of capabilities, to achieve our security objectives on the Peninsula. Anyway, thanks for responding to an anonymous poster – am sure you can understand my need to do so.

    • Jeffrey (History)

      On Payne, I object a little to the characterization that I “found the one place” where Payne conflates the two. My argument about Payne’s sudden criticisms of Global Zero being hypocritical are based on a decent familiarity with his work and limited interactions with the man. As I have understood his contribution (such that it is) to our thinking about deterrence, he represents a kind of off-brand Herman Kahn-ism. In On Thermonuclear War (and in some of Wohlstetter’s writing, too), Kahn makes this argument in favor of the arms race based on the possibility of surprise, particularly of the technical sort (See especially Chapter 5, “Stresses and Strains.” Perhaps I make too much of this in Kahn, but he provides the earliest use of the ubiquitous “hedging” concept that I find.) Minimum deterrence might be fine now, he is willing to stipulate for the purpose of argument, but we cannot be sure it will be fine in the future. What Payne seemed to do, from my perspective, is to take that argument about the possibility of technological or geopolitical surprise rooted in a specific fear of the now defunct Soviet Union and modify it — very effectively — to take advantage of the general unease with a uncertainty of the immediate post-Cold War period. (See here, especially, Payne’s The Fallacies of Cold War Deterrence and a New Direction.) It was a nice trick because one might have suspected the arguments of Kahn and Wohlstetter to lose appeal without a plausible peer adversary to breathe life into them; Payne succeeded in disembodying that fear from a specific actor, and rooting it in a generalized paranoia about the international security situation. (Arnie Kanter used to have this lovely, derisive description of geopolitical hedging that captured this notion — “Joe Stalin comes back to life” he would say, with a grin.)

      I have only interacted with Payne a few times, but each time he has told the same story about how he was surprised to “learn” that the Sovs were really deterred not by the Reagan defense buildup, but rather Reagan’s decision to fire the air traffic controllers. He used this story, in the perhaps three or four times I’ve been in a room with him, as a homily about our inability to determine what capability might deter, largely as a reason for keeping some capability that others are arguing has no apparent material value. Youneverknow, as Joaquin Andujar would say.

      So, I really do judge Payne’s recent criticism of Global Zero to be hypocritical, based on how much he embraced non-material factors in his rationale for (what I would regard as) excessive nuclear weapons capabilities. Perhaps he is consistent now, but I would argue the body of his work is of the same tradition as the Global Zero “security as state of mind.” The nicest thing I can say about Payne is that his reversal is part of a larger shift where opponents of nuclear weapons, who used to dismiss the military utility of the weapons, now deride their political utility; while proponents do the opposite. Perhaps he is no more hypocritical than the typical policy type.

      In this way, however, I distinguish Payne from Bridge Colby, who is a friend. Bridge and I disagree about many things, but I never feel like he is trying to “win” an argument or using anecdotes tactically. In our interactions, he is usually being honest about his views as he thinks through a problem. That is one reason that I find his recent scholarship, for example on limited nuclear options, worth reading — even if I usually disagree. My complaint about Payne is that he is outcome-oriented, fixing the arguments to always result in the same outcome. My guess is that Bridge, on the other hand, would happily accept the retirement of a particular nuclear weapons system or the ratification of an arms control treaty provided that it fit in his broader analytic framework. It doesn’t happen so often, of course, but he’s not ideological of close-minded.

      On Japan, rather than us trade comments, you might just invite Mort, Linton and Walt in for a chat some morning. Perhaps they will object to my characterization of our meetings. Or they might reinforce some of my points. Probably a mixture of both, I would guess. In any event, it might be interesting for you and your colleagues. It would certainly be pleasant, since they are such nice fellows. I know Mort has been back to Tokyo and Seoul since we went as a group, and I wouldn’t be surprised if either Linton or Walt have had additional interactions.