Good afternoon. I want to thank the CSIS Project on Nuclear Issues and the Los Alamos National Laboratory for having me here today. New Mexico is a marvelous break from the Washington, D.C. humidity at this time of year, and it’s a pleasure to be back up here on the mesa. This is an expert and well-informed audience, and I’ve been encouraged to be thought-provoking and to stimulate discussion, so let me take a swing at exploring the analytical terrain surrounding a notion that I call “proliferation as policy.”
For those of us who spend a lot of our time fretting about such issues, as I have, both in and out of government, there is often a certain __obviousness__ about the value of nonproliferation a certain axiomatic quality that doesn’t really require explanation because it just __is__. And I think this is so for some pretty good reasons. One might come at the question as an activist who sees nuclear weapons as a __per se __evil and longs for a future world of “zero.” Alternatively, one might approach it as a realist power-maximizer in an existing nuclear weapon state hoping to keep additional countries from acquiring such tools. Or one might be a game theoretician keen to avoid the growing crisis instability risks that are likely to be inherent in an enlarged system of nuclear “players.” Or one could simply be eager to minimize the changes of ever seeing the human catastrophe of nuclear weapons use. From any of these perspectives, however, having more nuclear players in the world is a very bad thing indeed. And on that you’ll get no argument from me.
I sometimes wonder, however, whether the seeming irresistibility of the case for nonproliferation may sometimes get in the way of our analytical acuity as we look at the geopolitical environment. It is not uncommon in our diplomatic relations, for instance, to hear it declared with great assurance that “Country Such-and-Such shares our interest in preventing nuclear weapons proliferation” and to have it be assumed, in effect, that if we just __remind__ its leaders of this shared interest, they will see the light and come around to our point of view. If there is a problem in obtaining someone’s cooperation on nonproliferation matters, we tend to see this as being merely due to disagreements over “tactics,” or perhaps just some lack of__capacity__ to do be helpful despite genuinely good intentions.
At worst, we suspect merely that others are holding out in order to bargain for as high a price as possible in return for giving us the cooperation they really __do__, in their hearts, agree is important anyway. This may be unwise on their part or perhaps on ours for commoditizing such cooperation by trying to purchase it through concessionary inducements but we assume that such bargaining doesn’t __really__ bespeak a significant difference of opinion about the value of nonproliferation. Only the would-be proliferator regimes themselves, we might think, actually __want__ nuclear weapons to spread and even then one usually doesn’t have to look far to find some analyst who feels their pursuit of such devices is an unfortunate but understandable choice, taken only grudgingly in the face of real or perceived foreign threats. Almost no one, we sometimes seem to assume, __really__ supports proliferation.
But perhaps we should take a step back from the obviousness of such conclusions, and consider the possibility that, “proliferation as policy” is not always felt to be an inherently irrational strategy. It is a strategy that it remains powerfully in our interest to prevent others from adopting, of course. We probably miss something important, however, if we see proliferation as no more than some kind of aberrance or confusion.
“Proliferation-as-policy” is actually something that seems to have appealed to a number of real-world decision-makers in the past. Many of you doubtless knows these stories at least as well as I do, but let me offer some examples:
- The Soviets gave Beijing a great deal of help in its weapons development, though by all accounts Khrushchev balked just before fulfilling the final clauses of his 1957 cooperation agreement with the Chinese, and stopped before providing Mao Zedong with an actual weapon prototype.
- French scientists provided a great deal of sensitive dual-use technology to Israel, including the reactor at Dimona and a plutonium-separation capability that many observers believe to form the core of Israel’s weapons program even today.
- China helped Pakistan develop its nuclear weapons program, to the point it has been reported that Beijing provided actual weapons designs.
- Even in the midst of the ongoing nuclear crisis over Iran’s previously-secret nuclear program, Russia provided Tehran with the nuclear reactor at Bushehr. Russian entities apparently also designed the plutonium-production reactor that Iran is constructing at Arak.
- China is reported to have provided uranium hexafluoride (UF6) feedstock to Iran’s secret enrichment program in the early 1990s, and to have provided the blueprints for Iran’s uranium conversion facility.
- North Korea is reported to have provided Libya with UF6, and more infamously to have constructed a plutonium-production reactor for Bashar al-Assad in Syria, though the Israelis bombed it in 2007.
- The official story about Abdul Qadeer Khan’s notorious nuclear smuggling ring is that it was some kind of a rogue operation, without official support or encouragement, but few analysts today seem really to believe this. It is widely believed that the Pakistani government, or a significant portion of it, was indeed complicit in Khan’s operations.
- Finally, Saudi Arabia has long been rumored to have helped finance Pakistan’s nuclear weapons program.
The evidence thus suggests that proliferation isn’t just about those who want to acquire nuclear weapons: at one point or another, a number of countries have apparently concluded that __supporting__ proliferation is in their interest.
But let me bring the issue closer to home, and talk a little about American policy. To my knowledge, the United States has never __not__ cared about nonproliferation: the spread of nuclear weaponry to additional players has never seemed a particularly good idea to us. But it is also true that how __much__ we cared about nonproliferation has varied over time. The fact that nonproliferation is a critical foreign and national security policy priority for Washington today shouldn’t blind us to the fact that things seemed at least somewhat different in the past.
In the 1950s and 1960s, for example, American planners do not seem to have attached as high a priority to nonproliferation as we do today. When it came to nuclear matters, U.S. leaders’ top priority was clearly to prevail in or at least to survive their Cold War competition with the Soviet Union. Financial resources, diplomatic energy, and political attention being finite, Washington not unreasonably prioritized preventing Warsaw Pact aggression on the Central Front in Europe. Nonproliferation was important, but it paled by comparison to __that__ goal. If the United States had been forced to choose between preventing World War III and keeping some randomly-selected developing country from acquiring a handful of atomic weapons, the right answer would have been pretty clear.
At a time when the superpowers faced each other in a tense deterrent standoff with thousands and eventually __dozens__ of thousands of nuclear weapons on each side, moreover, the likely strategic __impact__ of some proliferator country acquiring a small, “entry-level” arsenal would not have been very great. (Indeed, under such circumstances, a few more weapons here or there might be considered little more than strategic “noise.”) This didn’t mean that U.S. leaders didn’t care about nonproliferation, and indeed over time they played an important role in dissuading several countries from pursuing nuclear weapons development. But nonproliferation was not the __foremost__ value; it was merely one among others.
And, significantly, the demands of maintaining deterrence in a Cold War context existed, to some degree, in __tension__ with nonproliferation goals. It came to be felt, for instance, that high-level strategic nuclear deterrence faced theoretical problems when it came to deterring conventional aggression in Europe. Everyone understood that it was __possible__ that Washington would initiate a general nuclear exchange with the Soviet Union if Warsaw Pact forces crossed the inter-German border, but questions came to be raised about the credibility of that deterrent threat. How plausible was it, for instance, that a U.S. President would risk the vaporization of New York in order to prevent the occupation of Bonn? There was thus for years an ongoing debate about the possibility of nuclear weaponization in additional European allies and some interest, not least in West Germany, in this nuclear “option.“ Under those circumstances of presumed conventional disparities and the credibility problem of strategic deterrence vis-à-vis regional conflicts, deterrence and nonproliferation values thus came to point in different directions. We didn’t want to contribute to proliferation if we could help it, but we also had a Cold War to manage. Did we have to choose between these goals?
The solution to this problem at the time was actually quite shrewd, however, as the members of this audience will surely recall. In effect, the solution developed by NATO allowed us to have our cake and eat it at the same time. This was the policy of “nuclear sharing,” whereby American nuclear weapons were stationed on European soil carefully in U.S. custody, but on the scene in theater, and assigned to be __made available__ to NATO allies in time of war. This was felt to strengthen deterrence by addressing the perceived “credibility gap” of strategic deterrence vis-à-vis more local conflicts. Whether or not __Washington__ would be willing to sacrifice New York to save Bonn, for instance, it was by no means implausible that __Bonn__ would be.
At the same time, however, this approach was intended to serve nonproliferation, by obviating the need for additional NATO allies to develop nuclear weapons of their own. __That__ box could be checked, as it were, by American deployments under conditions of anticipated wartime “sharing”: additional allies did not need their own devices in peacetime because in time of war, they’d theoretically have access to some of ours. As long as peace lasted, moreover, this was consistent with Article I of the Nuclear Nonproliferation Treaty for no such transfer would actually take place: NATO merely __planned for__ __it__ as a way of making a Soviet invasion seem especially ill advised. In this way, we attempted to resolve the tension between deterrence and nonproliferation where U.S. allies confronted a potentially overwhelming conventional military threat in circumstances in which questions existed about the credibility of a purely strategic deterrent.
Whatever you think about how useful the continuation of such arrangements is for NATO today, I think this “sharing” model was a good choice under the circumstances of the time and that it is today given too little credit as a step that helped to keep the peace in Europe and to prevent nuclear weaponry from spreading much more widely. But the main point here is that such cleverness was actually __needed__, because a real tension had developed between broader U.S. strategic objectives and nonproliferation goals. We resolved the conflict fairly well at the time, but we miss something important today if we forget that this tension __was__ very real.
I would like to suggest that this history can provide us with an analytical window into the problem of “proliferation as policy,” for it may well be that for some countries, under some circumstances, there is no approach capable of so easily resolving this kind of tension between nonproliferation and other geopolitical equities. As we think about proliferation challenges, therefore, we must remember the possibility that proliferation might sometimes be felt on balance to be desirable, from the perspective not merely of a would-be weapon-acquirer itself, but also from that of other international players, including perhaps one or more nuclear weapon states.
It would not be hard, for instance, to imagine that many of the stories I recounted earlier about proliferation assistance can be understood through the prism of some such logic. Soviet help for China’s nuclear weapons program is likely to have made sense in Moscow for so long as Mao Zedong was an ally and junior partner against the imperialist powers, and it __stopped__making sense as tensions mounted between Moscow and Beijing within the international communist movement. China may have found aid for Pakistan’s program quite handy as part of a complex geopolitical chess game vis-à-vis a populous and developing India that had shown itself to possess nuclear capabilities in 1974. At least initially, both Russia and China may also have had some reason to support nuclearization in Iran and in North Korea as well in order to complicate American strategic planning and compensate for the loss of global “balance” against U.S. power previously provided by a communist Eastern Europe. And Saudi Arabia may have supported Pakistani weapons development simply in order to ensure the creation of an “Islamic Bomb“ __somewhere__, or perhaps with some eye to obtaining reciprocal assistance at some point down the road such as, for instance, if the hated Islamic revolutionary regime in Tehran preceded it down that road. My point is not to offer definitive accounts of the motivations involved in such cases, but rather merely to point out that nonproliferation is apparently sometimes considered a value that must fall by the wayside in light of some other concern.
So perhaps we should be less sanguine about being able to talk all other players in today’s world into acting on what we presume to be the belief they share with us in the crucial importance of nonproliferation. If they __don’t__ actually share that belief, enticing cooperation will be more difficult, requiring either that __we__ pay a higher price to elicit helpful steps from other players, or perhaps that __they__ be made pay a much higher price for being so uncooperative. The specific policy import will presumably vary with the circumstances, so I don’t want to offer any categorical prescriptions here except just to urge that we be especially careful to avoid mirror-imaging as we analyze our strategic environment.
And let me make one further point. The analytical prism that helps us understand at least the__possibility__ that one or more major players might decide that under certain circumstances nonproliferation might be only a secondary priority if indeed it is considered important at all should also help attune us to the existence of a broader range of potential alternative nuclear futures than might otherwise cross our minds. Not all of these futures are desirable ones, of course. But just as analytical honesty should force us to consider the possibility that nonproliferation values may not today be held as sacrosanct by all players as we hold them, so it might be useful to ponder what might lead U.S. leaders__ __to reconsider __their__ fidelity to them in the future. We have had to confront such choices in the past, after all, and it is not beyond imagining that they may arise again. Particularly if we __do__ wish to prevent the development of circumstances in which the United States will feel forced to view nonproliferation as being less important than some other value, such possibilities bear pondering now, lest by pretending that they don’t exist we make their arrival, paradoxically, more likely.
Imagine, for instance, a future in which U.S. allies in some part of the world face a significant regional power on military terms that are highly disadvantageous, and yet at a time when our__own__ ability to provide these allies with security in the face of such threats has been degraded either because that power has grown very strong, or because we have allowed our conventional military posture and global power-projection capabilities to erode and to fall behind the technological times, or both. This might be a Middle East facing the prospect of nuclear-empowered Iranian hegemony, or it might describe mid-century East Asia. But at what point might future U.S. leaders face a __new__ tension between nonproliferation values and the security imperatives of credibly deterring a would-be regional hegemon from opportunistic moves against our friends?
Might future U.S. leaders confronting the potential collapse of their alliance assurances especially were this to entail the potential extinction of one or more vibrant political democracies at some point be tempted to conclude that nonproliferation is not an __absolute__value after all? Could they, in good conscience, tell a threatened ally that nonproliferation is so compelling a value that it is the duty of that ally to __sacrifice itself__ on the nonproliferation altar rather than entertain thoughts of nuclear weapons development? Would it be possible to find some approach perhaps even an analogue to NATO’s “nuclear sharing” policy, now applied elsewhere that would enable us to thread the needle between nonproliferation and deterrent credibility in a regional context that might otherwise seem at risk of being “decoupled” from our strategic deterrence? Or might at some point future U.S. leaders discover what some other countries also seem from time to time to have concluded over the last few decades namely, that __some__ proliferation may actually be attractive as a strategic policy?
These are disturbing scenarios. As someone deeply devoted to nonproliferation, I think it is important to ponder such possibilities precisely in order to __prevent__ them from transpiring. Perhaps, for instance, we can help make such problems less likely by rededicating ourselves to the maintenance of U.S. conventional military power and global power-projection capabilities with more vigor than is currently fashionable in today’s cash-strapped Washington, with the objective of helping reduce the risk that a “gap” will occur between strategic deterrence and deterrence __in theater__. Being better aware of the dangers may also give us a chance to engage in more conceptual creativity, in the interest of being better prepared, and more able to find innovative approaches that avoid or at least limit the damage to nonproliferation equities, if and when such dilemmas actually do confront us. It may also be that making __other __countries and perhaps especially those would-be regional hegemons I hypothesized a moment ago better aware of these possibilities will help us persuade them to behave more constructively today, precisely in order to prevent __us__ from having to face such perilous choices in the future.
Perhaps, in other words, both we and others can glean some wisdom from the insight that nonproliferation is a good that could conceivably become devalued even in the eyes of its foremost present-day proponents. And perhaps such wisdom can even be used to help__prevent__ that devaluation. So I hope I’ve managed to problematize our familiarity with the seeming obviousness of nonproliferation values. It we are truly the friends of nonproliferation, you might say, we cannot just be its __flatterers__. It may be that nonproliferation will be best served if we recognize not just its fragility but in fact its __contingency__ as a desirable “good” in the first place.