New Paradigms Forum Proliferation Issues and Much More …


The Catch-22 of NFU

In the wake of a conference I attended a couple of months ago on nuclear deterrence and nuclear disarmament, I ended up having dinner and a glass of wine with a couple of friends of mine, and we ended up spending a fair amount of time arguing over the issue of nuclear weapons declaratory policy.

My two friends disagreed vehemently about the issue, but the discussion was a good deal more interesting and thought-provoking one than most of the fairly sterile discussions of “no first use” (NFU) declaratory policy I used to hear during my U.S. Government service doing nonproliferation diplomacy.  In particular, one of my friends offered an argument for NFU that it seems worth exploring here on New Paradigms Forum. I don’t agree with his position, but it deserves consideration because it’s the analytically sharpest NFU pitch I’ve yet heard.

I.        Debating the “Sole Purpose” Thesis

Our dinner conversation was sparked by a comment we had heard earlier that day from a senior U.S. official that he could think of no circumstances under which the United States would want or need to use nuclear weapons.  Perhaps one should not read too much into this, for this remark could have signified merely that the official had been caught off guard by an unexpected question and couldn’t – when put on the spot – quite remember the odd and complicated formula that constitutes the Obama Administration’s new nuclear declaratory policy as spelled out in the 2010 Nuclear Posture Review Report. (After all, not being able to think of such circumstances at the moment is not necessarily the same thing as there existing none.)  As one might expect, however, his comment had set off a flurry of predictable discussion among conference attendees along the lines of “well, if you wouldn’t ever use them why do you have them?”

One of my friends at dinner started from this point too, but with an interesting variation.  In most discussions of NFU, including the points at which this issue had come up during our conference, proponents of NFU argue the “sole purpose” thesis – namely, that the only purpose nuclear weaponry has is to deter the use of other nuclear weapons.  Advanced as a blanket statement about all nuclear weaponry, this argument is important to the disarmament narrative, because if one accepts it, what usually follows is an argument that all nuclear weapons possession is entirely self-canceling, making nuclear weapons abolition analytically indistinguishable, from a deterrence perspective, from our present situation.  From this, it is often reasoned that there is thus no reason for all possessors not to move quickly to “zero.”

When stated in general categorical form, the “sole purpose” thesis sounds like nonsense to me.  Clearly, nuclear weapons possessors do feel that there are potential uses for nuclear weapons beyond just deterring the use of other such weapons.  Even the Obama Administration took the trouble to draw attention to the possibility that it might need to threaten or use nuclear weapons in response to biological weapons (BW) threats – a position quite consistent, by the way, with Nixon-era reassurances that we didn’t need to worry about deterring BW use by others as we scrapped our own BW program, because it would still be possible to rely upon our nuclear weapons to deter adversary states’ use of BW.  Several other possessors have, at various points, taken the view that nuclear weapons can in effect “cross-deter” against other types of weapons of mass destruction.

More importantly, many nuclear weapons possessors have taken the position – at one point or another, if not in all cases today – that nuclear weapons have utility in deterring large-scale conventional attack.  We and our NATO allies relied upon that notion for years during the Cold War, and the Russians believe so today.  India began its weapons program after a large-scale conventional invasion by China, South Africa feared revolutionary onslaught from its post-colonial neighbors, and the UK and France feared Warsaw Pact occupation as much as (or more than) Soviet nuclear weaponry.  Pakistan feels it needs nuclear weaponry to ensure against an Indian invasion, while the presumed possessor state of Israel is believed to have nuclear weapons for fear of the non-nuclear numerical superiority of its many hostile neighbors.  Russian and Chinese officials sometimes claim to fear – or that they might come to fear – a disarming first-strike by U.S. conventional weapons, and North Korea is generally thought to fear conventionally-armed “regime change” more than it fears U.S. nuclear weapons.  And even the International Court of Justice left a notable loophole in its 1996 advisory opinion on nuclear weapons with respect to “an extreme circumstance of self-defence, in which the very survival of a State would be at stake.”  (The Court certainly did not tie the legality of nuclear weapons use exclusively to nuclear threats.)

So while it’s marvelous that we in the United States presently have to think so hard in order to come up with plausible nuclear use scenarios for our own weapons, we shouldn’t pretend that this is not, in historical terms, quite a luxury.  The “sole purpose” thesis, if it says anything truthful at all (and I doubt it is true even for us), seems only to be a statement about nuclear weapons in the United States under the present set of geopolitical circumstances.

But, to its credit, that’s where my friend’s argument began: he advanced the “sole purpose” thesis, but posited that it applied only to the United States at this time. Because of this, however, the argumentative logic of “zero” did not necessarily flow automatically from his assumption.  (The idea that one can radically scale back and then eliminate nuclear weapons because they only deter other nuclear weapons is a coherent thought only when “sole purpose” thinking applies to all such weapons.  If some country wants or keeps nuclear weapons in existence for a different reason, then others may have to keep them in order to deter that country’s use ... and so forth, as the cycle begins again.)  Instead, my friend used his U.S.-focused “sole purpose” conclusion as a jumping-off point for proposing only that we adopt an NFU policy.

II.        Crisis Stability and NFU

I’ve certainly heard many disarmament activists argue for an American NFU pledge before, but this suggestion was more novel in that it relied heavily upon the argument that a U.S. NFU pledge would contribute to crisis stability.  (Crisis stability arguments are ordinarily more common from opponents of NFU, who tend to think that there is deterrent value in ambiguity – to keep a potential aggressor guessing – and worry that NFU, if believed, could tempt non-nuclear aggression of some sort.  Admirably, my friend was trying to “speak disarmament in the language of security” by supporting an element of the disarmament community’s agenda with an argument that made sense to people who think in national security terms.)  As he saw it, an American NFU pledge would help remove any fears other nuclear weapons possessors might still entertain of a U.S. nuclear first strike, thus reducing their temptations to adopt stability-endangering positions such as launch-on-warning or even nuclear preemption.  Our nuclear forces are large and competent enough, he suggested, that fear of our undertaking a disarming first strike could contribute to unwanted escalation in a crisis.  An NFU promise, however, would help keep others from worrying about this.  This crisis stability idea, he argued, was the “most important” reason to adopt an NFU declaratory policy.

My friend argued that an American NFU pledge would be believed, moreover, because in a legalistic democracy such as ours, such a promise would create powerful “audience costs” – to use a sociologists’ term – constraining our ability to break it.  In our democratic system, he contended, violating an NFU promise would entail severe enough domestic political costs, not to mention international diplomatic ones, to make a violation unlikely.

His argument is as good a bit of NFU advocacy as I’ve seen, but I think it still comes up short.  Leaving aside my skepticism that “sole purpose” thesis applies even for the United States – a position which I share with the Obama Administration, which in its Nuclear Posture Review explicitly declined to adopt this formulation – I think the alleged strategic benefits of NFU are illusory.  Recall that my friend bases his “most important” argument for NFU on the supposition that an American pledge would be believed with such certainty that our potential adversaries will place reliance upon its sanctity even in the throes of a crisis so severe that they might otherwise be tempted to contemplate nuclear preemption.  I would be the first to agree that we Americans are people of our word, but I see no way that a mere NFU declaration could carry this extraordinary persuasive burden.

III.        The Unbelievability of No-First-Use

My general critique of NFU has for years been simply that such promises are not particularly credible.  I have never fully trusted anyone else’s NFU, and I’m hard pressed to explain why they should be particularly confident in ours.  The Soviet Union constantly beat the NFU drum in its nuclear disarmament diplomacy after making a much-publicized pledge in 1982, but after this position was formally abandoned in 1993 – by a post-communist Russia that had become very weak in conventional military terms and decided that it needed nuclear weapons to make up for this – it was admitted that Moscow’s NFU pledge had been propagandistic hooey all along.

China has maintained an NFU position since first acquiring nuclear weapons in 1964, but many experts wonder what this actually means, coming as it does from a regime that in the past has been quite good at linguistic and definitional contortions in order to paint its own behavior as virtuous and non-self-interested no matter what that behavior actually is.   (Maybe they mean it; maybe they don’t.  Or maybe “first use” doesn’t really mean what we think it does – or they wish us to think it does – in the first place.)  The literature is divided, and with good reason.  For decades, China has prized ambiguity, secrecy, and deception in its nuclear force posture, feeling that these contribute to deterrence by sowing uncertainty in a potential adversary.  Might it be too much to see Beijing’s NFU pledge as being a deliberate contribution to such uncertainties?  If so, the whole point is that we don’t know whether to believe it or not.  That make may NFU useful from a Chinese perspective, but it certainly doesn’t help my friend’s case for a notion of crisis stability grounded in the de facto inviolability of an American NFU declaration.

For its part, India also makes much of its NFU policy.  As suggested above, however, it’s hard to forget that New Delhi began its nuclear program in response to a large-scale Chinese invasion in 1962, before either China or Pakistan had nuclear weaponry.

The more important problem, however, isn’t possible insincerity ab initio.  It’s simply that it seems inherently unbelievable that an NFU pledge would be followed in all imaginable circumstances.  Even if the promise had been sincerely offered and resolutely intended, one might wonder whether a country with nuclear weapons would be willing to place such stock in NFU that it would choose to lose a major war or countenance the emergence of a dramatic new threat without employing the one tool that might be able to turn things around.

I thus tend to think that all NFU promises have implicit caveats that kick in when things really get bad.  “Audience costs” are not nothing, especially for a legalistic democracy like our own, but they are unlikely to trump the danger of impending national catastrophe.  I’m willing to believe that the U.S. political system would help make us very scrupulous in sticking to an NFU pledge in most circumstances, but most of those cases would not really present serious “use” incentives for us anyway.  As we have already seen, America is in the enviable position of having non-nuclear options for handling all but the gravest of crises.

Since there is little reason for others to fear our nuclear weapons in anything but the most outlandishly calamitous of situations, and no persuasive reason to trust in our self-sacrificial restraint in an extreme case, I’m unpersuaded that an American NFU would add anything meaningful to anyone else’s strategic equation.  One might call this the “Catch-22 of NFU”: when others could most trust it, we wouldn’t need it – and when we might most need it, they wouldn’t trust it.

IV.        NFU as Propaganda?

To be sure, even with all these believability problems, an NFU pledge might have considerable propaganda value.  (My friend’s argument also notes the considerable diplomatic benefits we would gain in disarmament circles from such a promise, though he stresses that these are secondary to its presumed impact upon crisis stability.)   I could imagine an argument for making an NFU promise without actually meaning it, so that the United States might be able, as the saying goes, to “have our cake and eat it too.”

By this logic, anyone who fundamentally dislikes and distrusts us might still be relied upon usefully to fear our nuclear weaponry, because they would never believe our NFU sincerity in the first place.  At the same time, however, the fact that an American NFU pledge might be believed by at least some other countries could win us diplomatic kudos and improve our nuclear disarmament “street cred.”  This argument, in effect, seeks to put us in the position of the Soviet Union in the days when it sought to reap diplomatic benefits by merely pretending to have an NFU policy – or, arguably, in the position of China or India today.

Yet even if one were willing to endorse an approach of such cynical dishonesty when it comes to grave matters of public policy concerning the potential incineration of millions of persons – which I am not – I worry that such a too-clever-by-half approach could have undesirable side effects.  The propagandistic approach to NFU already presumes that potential adversaries wouldn’t credit a declaratory policy of unqualified restraint in the first place.  But how credulous would third parties really be, or remain?  How long could we maintain the fiction that one were truly serious about NFU?  And what might it do to deterrence, in a more general sense, to adopt a policy built on the implicit assumption that what we say cannot always be trusted?

There is a role for signaling in deterrence, and in fact, most of deterrence could be thought of as an exercise in signaling intentions and capabilities in such as way as to convince a potential adversary that aggression would be unprofitable and dangerous.  Deliberate ambiguity can play a role in this, as can efforts to convey clarity and resolution.  Indeed, dishonesty can even be an important part of the process, in the sense that one might claim to be – and want an opponent to believe one is – more resolute than one actually feels.  But dishonest positions, in this regard, are useful only to the extent that they are in fact given credence.

This is one danger of a blatantly propagandistic NFU strategy: being seen to be duplicitous may undermine one’s ability to use signaling (including sending false signals!) in other respects when one really needs to.  What if, for instance, one wanted to send a truthful signal of one’s own likely restraint – e.g., as part of an effort to tamp down the escalatory dynamics of an action-response cycle in a crisis – but had become known for sending false signals of likely restraint in the past, pursuant to NFU propaganda games?

I don’t want to overplay this “distrust” objection, for I think the strongest reason to eschew the disingenuous “have our cake and eat it too” approach to NFU is instead a moral one.  Potential nuclear war is too serious a business for us to make propagandistic gamesmanship the basis of our declaratory policy.

With regard to practical consequences, the sky doubtless would not immediately fall if we took such an approach.  The Soviets had a propaganda-based NFU posture for years, after all, and some observers today question the foundations of the Chinese and Indian pledges.  Deterrence did not fail for them just because they advanced a NFU policy that others only half believed.  (If we worried about NFU undercutting the “extended” nuclear deterrence we offer our allies in the face of regional and sometimes non-nuclear threats, I suppose we could privately  reassure our allies that we didn’t really mean it in the first place – and that we don’t think their adversaries believe us either.  If one has the stomach for such things, cynical NFU pronouncements no doubt provide endless opportunities for deviousness.)  Nevertheless, acquiring a reputation for dishonesty in deterrent signaling games could at some point be perilous, and I’m not sure one would get enough propaganda benefit from a duplicitous NFU posture to outweigh the potential risk from entering some future nuclear crisis with a credibility problem.

V.        Conclusion

But let’s return to my friend’s proposal for an American NFU policy.  Make no mistake: he does not advance NFU as the kind of cynical ploy I discuss above.  (Quite to the contrary, his scheme needs NFU to be believed by a potential adversary in a crisis if it is to make any sense.  He needs NFU to take any possibility of a U.S. first strike “off the table” in an adversary’s eyes.)  To some extent, I fear that his NFU approach could be mistaken for just such disingenuity.  But this is secondary.  The main reason for my skepticism of his NFU proposal, however, simply goes back to believability.  I am not convinced that his plan would in fact provide the advertised crisis stability benefits, because those countries whose behavior we might most wish to shape in a crisis (e.g., in dissuading launch-on-warning or preemption) would be those the least likely to believe our NFU promise in the first place.

This is, in a sense, the flip side of the cynics’ argument in favor of NFU duplicity.  The cynic argues that we could still implicitly threaten rogues with nuclear weapons even if we had an NFU declaratory policy.  I suggest that we cannot disabuse potential nuclear adversaries of their distrust in our intentions by a mere declaratory posture that we can revoke at will and that, in an extreme case, would ask them to believe that we will elevate promise-keeping over national survival.  In both cases, one sees the inherent unbelievably of NFU promises “in a crunch.”  The cynic’s approach depends on this unbelievability, but my friend’s approach is undermined by it.

So I don’t buy my friend’s argument on NFU.  It is nonetheless a valuable debate to have.  Readers interested in learning more about his position should consult the recent article he wrote on the subject: Michael S. Gerson, “No First Use: The Next Step for U.S. Nuclear Policy,” International Security, vol. 35, no. 2 (Fall 2010), at 7. I have only briefly summarized his points here, and may have done him some injustice in the abbreviation, so I’d encourage you to read it.  New Paradigms Forum would be interested in any comments and feedback you might have, so I invite you to send any to me via e-mail at

-- Christopher Ford


NPF Pages

Recent Additions to NPF

NPF Discussion Pages

Calendar of NPF Postings

January 2021
« Dec    

NPF Archives (by month)