The recent crisis over North Korea's nuclear program is merely the latest evidence that the global nonproliferation regime, symbolized by the nuclear Nonproliferation Treaty (NPT), is inexorably breaking down. Although U.S. concessions may ultimately induce Pyongyang once again to allow international inspections, that will be a meager accomplishment. It will hardly offer reliable guarantees that a regime as secretive and politically opaque as North Korea's cannot evade International Atomic Energy Agency scrutiny while pursuing a clandestine nuclear weapons program. Iraq, for one, was certainly able to do so while complying with all IAEA inspection requirements.

North Korea is only one of several states with nuclear ambitions. India and Pakistan have also emerged as threshold aspirants to, if not already full-fledged members of, the once-exclusive global nuclear club. Persistent reports surface of Iranian and Libyan efforts to exploit the political chaos in the former Soviet Union to purchase their own small arsenals. Even Ukraine's agreement with the United States and Russia to turn over its nuclear warheads is far from certain, given the foot-dragging and obstructionist tilt of the Ukrainian parliament as well as the widespread public sentiment for retaining the weapons. These worrisome trends more than offset any positive developments, such as France's and China's decisions to adhere to NPT provisions or South Africa's announcement that it has given up the arsenal it had developed surreptitiously in the 1980s.

It is time for U.S. leaders to reassess Cold War policies on nonproliferation, security commitments and extended deterrence and to adapt them to changed international circumstances. These commitments may once have made sense, given the need to thwart the Soviet Union's expansionist agenda. But they are highly dubious in the absence of the superpower rivalry. They now threaten to embroil the United States in regional conflicts where nuclear weapons have already proliferated or will inevitably proliferate soon. Washington should give up its fruitless obsession with preserving the NPT and the unraveling nonproliferation system that it represents.


Proliferation is frequently occurring in areas where bitter regional rivalries, ethnic or religious tensions, and raging border disputes could potentially involve the United States. Stalinist North Korea's hostility toward South Korea has already produced one major war, and the ironically named demilitarized zone between the rival states remains the most heavily armed area on the planet. Relations between India and Pakistan are only slightly less acrimonious. The two countries have fought three major wars since independence in 1947, and border skirmishes over the disputed province of Kashmir are commonplace. Iran has ambitions to become the preeminent power in the Persian Gulf region, a prospect that alarms Saudi Arabia and other neighbors. Ukraine jealously guards its newly won independence from Moscow and worries that Russia will someday attempt to reassert its imperial prerogatives. Those fears are not unfounded, as Russian politicians across the political spectrum openly express irredentist sentiments toward Ukraine. A simmering dispute involving the Crimea, a Ukrainian territory inhabited predominantly by Russians, is an especially dangerous flashpoint.

The prospect that such virulent rivals are already armed or could soon be armed with nuclear weapons is alarming enough. Even more unsettling is the existence of security commitments that could embroil the United States. In the case of the two Koreas, the United States is pledged to defend the south from attack and currently deploys more than 36,000 U.S. troops on the peninsula. Washington's exposure in the Pakistan-India confrontation is less severe, primarily because no American forces are stationed as a tripwire on the subcontinent, but a 1959 defense agreement nonetheless obligates the United States to assist Pakistan if it becomes the victim of aggression. The U.S. commitment to protect Saudi Arabia's security may be informal but, as the Persian Gulf War confirmed, it is quite real. As of yet Washington has no known obligations that might involve it in a confrontation between Russia and Ukraine. But the "security assurances" that U.S. officials may have offered to entice Ukraine to relinquish nuclear weapons are worrisome, especially given the ongoing campaign among many in the U.S. foreign policy community to extend NATO's security jurisdiction to Eastern Europe. If the enlargement of NATO ultimately includes Ukraine, which some enthusiastic proponents have suggested, then the United States would be obligated by Article 5 of the North Atlantic Treaty to help defend that country should it come under attack.


American leaders hope to evade the dilemma posed by this proliferation of weapons and security obligations by redoubling efforts to prevent new nuclear-weapon states from emerging in the first place. Although a noble objective, it is also an impractical one. Diplomatic or economic inducements are unlikely to dissuade regimes that have expansionist aims or that fear for their own survival from acquiring independent deterrents. As the case of North Korea suggests, such a strategy will produce (at best) paper promises. Indeed, offering concessions if a nuclear aspirant agrees to forgo weapons programs creates a perverse incentive. Other nations may conclude that playing the nuclear card is an extremely effective way to get Washington's attention and extort concessions.

More hard-line policies, however, show no better promise of halting proliferation. A strategy of coercive nonproliferation, relying on economic sanctions or preemptive military strikes, is not a viable option in most cases. Sanctions have not had an impressive record of persuading target regimes to alter their policies on important issues. Embargoes are especially likely to prove futile in compelling countries such as North Korea and Iran to abandon nuclear aspirations, since those countries are already largely isolated from the global economic system.

Preemptive military strikes would theoretically be much more effective, but that approach also has serious drawbacks. Clandestine weapons development sites are hard to identify and attacks on operating reactors run the risk of spreading radioactive fallout, endangering civilians in the target country as well as those in neighboring nations. Moreover, the country that had been assaulted would have every incentive to seek revenge against the United States. At the very least, a preemptive strike could increase the prospect of terrorist incidents and attacks on U.S. military forces and regional clients. Bombing North Korea's nuclear facilities, for example, could trigger a general war on the Korean Peninsula, engulfing the American troops stationed directly astride the invasion routes leading from the demilitarized zone to Seoul.


The impracticality of attempting to stem the spread of nuclear weapons in the post-Cold War world means that Washington must face some hard choices concerning its security commitments to regional allies and clients. Greater proliferation calls into question the benefits to the United States of preserving the doctrine of extended deterrence, and thus the credibility of U.S. guarantees.

To discourage nuclear proliferation, U.S. policymakers have been willing to continue the Cold War era bargain: if Washington's allies and clients renounce ambitions to acquire independent deterrents, the United States will help protect their security, including in many cases extending the protection of the U.S. nuclear arsenal. But the nature of the risk has changed dramatically. Throughout the Cold War, the United States had to deter only one hostile nuclear-weapon state. (By the time China acquired a credible arsenal, the rapprochement between Beijing and Washington was already under way.) True, even in the bipolar context extended deterrence entailed some risk, but as time passed an implicit set of rules governed the superpower rivalry, and a reckless Soviet initiative that would lead to a nuclear conflagration seemed less and less plausible.

The benefits of extended deterrence have not been trivial. Washington's nuclear umbrella provides an incentive for its allies to forgo developing independent arsenals and may also deter an aggressive regional nuclear power from attacking its neighbors. These are the reasons why advocates of extended deterrence and nuclear nonproliferation cling so desperately to a policy of extensive U.S. political and military engagement. Washington Post columnist Stephen S. Rosenfeld epitomizes this position when he proposes "drawing would-be nuclear countries into a network of smothering security and other ties that at once provide those countries with political and military reassurances and restrict their nuclear options."1

But the "smothering" strategy has significant problems. Although extended deterrence might make attacks on U.S. protectorates less likely, even in a multipolar nuclear system, it also guarantees that the United States would be disastrously entangled in any conflict that did erupt. During the Cold War a persuasive case could be made that the United States had no choice but to accept that risk in order to prevent Soviet global hegemony. But the argument that America must continue accepting similar risks merely to discourage regional conflicts is considerably less compelling. Moreover, Washington's ability to dissuade nations from acquiring nuclear weapons is extremely uneven. A network of smothering ties may work with some Cold War era allies, but how much effect will it have on North Korea, Iran or Libya, or even on a nation such as Pakistan that sees a regional adversary threatening its very survival?

That latter problem underscores a crucial point: the nonproliferation system is not merely breaking down, it is breaking down asymmetrically. The regimes that seem most determined to develop their own arsenals are in many cases precisely the ones that are most likely to contemplate using them. They are also frequently regimes over which the United States has little or no influence. Washington may inadvertently create a situation in which the more peaceably inclined nations remain nonnuclear, and entirely dependent on U.S. protection, while unstable or revanchist states are armed to the teeth.

The credibility of extended deterrence is likely to be questioned by U.S. clients as well as their adversaries. The benefits to the United States of deterring an attack on a client would appear to be far less today than during the intense U.S.-Soviet rivalry. America's global dependents would have to ask themselves whether U.S. leaders would actually accept the consequences of a nuclear war if the commitment were challenged. For them, it is not an academic matter; their survival as independent nations could hinge on the answer. A similar gnawing uncertainty about the reliability of Washington's nuclear shield was one factor that impelled Gaullist France to adopt a hedging strategy by developing a small independent arsenal in the early 1960s. Similar considerations for other U.S. protectorates can only be more acute in the post-Cold War era.

The assumption that extended deterrence will today have the same stabilizing effect it did during the Cold War is fallacious. As Steven E. Miller points out, nuclear deterrence is not equally effective in all circumstances. In particular, "deterrence will not work well when dealing with ambiguous borders or disputed territories."2 Territorial ambiguity was never a salient factor in the U.S.-Soviet competition. But both the India-Pakistan and Russia-Ukraine rivalries involve precisely that problem. The North Korea-South Korea confrontation is even more serious: neither regime even accepts the legitimacy of the other. Although Miller's observation was intended to support the proposition that an independent Ukrainian arsenal might not effectively deter an attack by Russia, his point is even more applicable to deterrence guarantees extended by a distant power. Their ability to prevent conflicts in cases of virulent feuds between regional enemies is doubtful.

The issue that U.S. policymakers must face, and have thus far refused to face, is whether assuming the risk of nuclear war is justifiable absent the security challenge of a superpower rival. If proliferation trends continue, Washington could face the prospect of having to deter numerous nuclear-weapon states from attacking or intimidating U.S. allies and clients. That could be a more unwieldy and more dangerous mission than was deterring the U.S.S.R. Some new nuclear powers are likely to be governed by regimes considerably less predictable and "rational" than the Kremlin leadership proved to be. Such regimes might gamble that, while Washington was willing to court nuclear war to block Soviet global hegemony, it will not be merely to prevent a shift in a regional balance of power. Further, even a reasonably rational government might conclude that stifling a hated adversary is worth risking Washington's wrath. Only one deterrence failure would cause a catastrophe for the United States.


The recognition that Washington's hoary policies of nuclear nonproliferation and extended deterrence will not suffice in a post-Cold War world does not mean that the United States can do nothing to reduce the danger of regional nuclear wars. U.S. leaders can pursue a number of worthwhile, low-risk initiatives.

Further reductions in America's own oversized arsenal would foster a less threatening global environment and weaken the incentive for other states to counter perceived U.S. intimidation. Extending the moratorium on U.S. nuclear tests would also engender a less confrontational atmosphere. Washington should encourage the establishment of nuclear-free zones in regions such as South America and sub-Saharan Africa where no nation currently possesses such weapons or seems on the brink of doing so. In cases where two or more regional rivals already have arsenals, the United States can help them develop reliable command-and-control systems to prevent accidental launches or the theft of weapons by terrorist organizations. An active U.S. diplomacy can also assist such adversaries to articulate defensive nuclear doctrines to minimize the chances of miscalculation.

Such an approach would require a willingness to acknowledge that the number of nuclear-weapon powers is certain to grow regardless of America's wishes. As one Indian scholar has noted, the issue of nonproliferation is a moot point with respect to his region; rather, "The question is how to make a nuclearized South Asia stable and free of war."3 Such a policy focus is more realistic than the cliched recitation of the virtues of nonproliferation in places where proliferation is already a given. The relevant task now is not so much to prevent proliferation as to learn how to live with it.

The United States can take some steps to help make a multipolar nuclear world marginally safer. Under no circumstances, however, should Washington place this country at risk in purely regional disputes that have nuclear dangers. One can readily sympathize with countries that seek stability for their regions. But the reality is that conflicts between long-standing rivals are an ever-present danger. Some scholars, most notably Kenneth Waltz and John Mearsheimer, have argued that the possession of nuclear weapons can actually be a stabilizing factor since they tend to make their possessors more risk averse. Such arguments may be correct, but they are based on an extraordinarily thin set of experiences. The threat of mutual destruction did indeed make the United States and the Soviet Union cautious in conducting their Cold War rivalry, and Chinese officials did appear to become more responsible once their country acquired an arsenal. Nevertheless, it is a leap of faith to assume that the existence of nuclear weapons will produce similar restraint in much more volatile regional settings.

Rear Admiral Edward Scheafer, Jr., head of U.S. Naval Intelligence, warned in a recent interview that "somewhere, some time in this decade, someone is going to set off a nuclear weapon in deadly earnest."4 One cannot rule out the possibility that Scheafer will be proven correct, if not in this decade, then in the next. The first task of a prudent U.S. security strategy must be to keep America out of the nuclear crossfire.


1 Stephen S. Rosenfeld, "Bombs for Everyone," The Washington Post, June 25, 1993, p. A25.

2 Steven E. Miller, "The Case Against a Ukrainian Nuclear Deterrent," Foreign Affairs, Summer 1993, p. 71.

3 C. Raja Mohan, senior research associate with India's Institute for Defense Studies and Analysis, quoted in Peter Crier, "Asian Nations Pose Greatest Nuclear Risk," Christian Science Monitor, Aug. 3, 1993, p. 1.

4 Quoted in "Nuclear Survival Task Falls to U.S.," Atlanta Constitution, Aug. 5, 1993, p. 18.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Ted Galen Carpenter is Director of Foreign Policy Studies at the Cato Institute in Washington, D.C. His most recent book is “A Search for Enemies: America's Alliances after the Cold War.”
  • More By Ted Galen Carpenter