Reducing Nuclear Danger

McGeorge Bundy, William J Crowe Jr, Sidney Drell. Foreign Affairs. Volume 72, Issue 2. Spring 1993.

A Dramatically New Situation


Two enormous events of recent years have opened the way for effective worldwide action against the danger of nuclear weapons. The first is the end of the Cold War and Soviet communism. The second is the sharp double lesson of the case of Saddam Hussein: that a rich and aggressive tyrant could get close to building a bomb of his own, but also that his effort could be blocked by effective international action. There is now a real prospect that almost all countries—those with many warheads, those with few and those with none—can come together in a worldwide program to reduce the existing nuclear arsenals and to prevent their further proliferation.

There are three immediate tasks: to execute the large bilateral reductions in U.S. and Russian forces that have been announced in recent years, to assure that Russia remains the only nuclear weapon state among the successor states of the old Soviet Union, and to reform and reinforce the worldwide effort to reduce the spread of nuclear weapons by applying the lessons learned in the case of Saddam Hussein.

In a dramatically new situation we must ask ourselves afresh the question that belongs to us all by the basic rules of our society: What should we ask of Washington in its policy and its behavior with respect to nuclear danger? What do present worldwide conditions tell us about what our government's policy should be, and what it should try to achieve, in the rest of this century? We have been led to the following propositions.

The United States should respect the reality that as long as there are bombs, there will be a nuclear danger that is unique in all history. The bomb is uniquely dangerous because it is uniquely destructive—so destructive that in all the conflict and tumult of the Cold War no one chose to use it or to provoke its use by others. That reality must permeate American policy choices even more thoroughly in the future than in the past. It requires intense and continuous effort to reduce the danger that comes from the very fact that the bomb is technically possible. It requires respect and support for those who have responsible control over such weapons. It also requires, at the other end of policymaking, a higher priority than ever for choices that can help prevent the possession of this weapon by international adventurers like Saddam Hussein.

From Opponents to Friends


What underlies the great new Russian-American arms control measures of the last two years, and gives promise of further progress, is a set of interlocking political decisions and actions that have turned the Russians and the Americans from opponents to friends. Their nuclear competition now has no cause but itself, and both sides, in their actions of the last few years, have shown their understanding that all they need from these weapons, with respect to each other, is clear assurance that neither side will ever be able to achieve, without ample notice, any nuclear advantage that might be considered usable against the other. There is no short road to complete nuclear disarmament, but the road away from superpower confrontation has indeed been opened; the United States and Russia can walk that road together, to their common advantage, as they learn the art of cooperative nuclear moderation under the first and second Strategic Arms Reduction treaties (START) and their exchanged announcements of major tactical reductions.

What is also important for the long run, and less often remembered, is that Americans have a parallel absence of underlying political hostility toward all other announced nuclear weapons states. The United States does not think of any of them as potential enemies in war-not the British or the French, self-evidently; not the Chinese because both countries have learned better over the forty years since Korea. The absence of direct hostility between the United States and China does not make that relationship the same as the American relation to England or France, or even Russia, but it is enough to mean that the nuclear danger between them that both have felt in the past no longer exists. Other differences with China persist, but it makes no sense for either side to let such differences prevent cooperation against nuclear danger.

Americans can expect differences with all four of the other permanent members of the Security Council, but not warfare with any. The political changes required to produce a change in this expectation are very large. Before they could have a direct impact on nuclear balance or nuclear danger, there would be plenty of time for any necessary American response in either conventional or nuclear rearmament. The prudent judgment for the present is that it should be both the American purpose and the American expectation that in the 1990s the five permanent Security Council members should continue on their new course of cooperation against nuclear danger. In these matters they are not in hot competition, and there is no underlying political animosity that need prevent them from working together. Pride is another matter; so is greed; so is misunderstanding. Any one of the three can hamper or even prevent cooperation. But the common real interest is to work together against nuclear danger.

The same general rule applies to our nuclear-danger relations with all other countries. Except for a few—perhaps Iraq, Iran, North Korea and Libya-all the world's governments now favor an end to nuclear proliferation. What is true of the five announced nuclear weapon states is true of the three other states that must be judged to have nuclear warheads of their own: Israel, India and Pakistan. Their formal status is a question of some difficulty, but neither their neighbors nor the world can doubt that, in terms of what they could do in desperation, they are nuclear weapon states. Still they are not eager for nuclear competition or open threat, let alone for aggressive use of what they have. They can join in reducing their own contribution to nuclear danger.

Still more clearly on the side of reducing nuclear danger are the scores of states, large and small, that have definitely rejected the course of obtaining their own nuclear weapons. The basic division in the world on the subject of nuclear proliferation is not between those with and without nuclear weapons. It is between almost all nations and the very few who currently seek weapons to reinforce their expansive ambition.

A Doctrine of Defensive Last Resort


Any American first use of nuclear weapons should be governed by a stringent doctrine of defensive last resort, which can also be adopted by all other nuclear weapon states.

As matters stand at the beginning of 1993, there is no vital interest of the United States, except the deterrence of nuclear attack, that cannot be met by prudent conventional readiness—there is no visible case where the United States could be forced to choose between defeat and the first use of nuclear weapons. This is one crucially important meaning of the 1989-91 transformation of Europe.

In the foreseeable future, the American president will be able to accompany support for a doctrine of defensive last resort with the assurance that the president and commander in chief sees no present likelihood of any confrontation that would require this dreadful choice. Indeed the U.S. government has already gone a long way in this direction. A doctrine of last resort was adopted by NATO in 1991, and in the NATO context the word "defensive" can fairly be taken for granted—by its own language the North Atlantic Treaty is a defensive alliance. This new NATO doctrine gives reassurance to all members, both of support and of prudence. It can do the same thing more broadly as a basic doctrine of the United States.

Both nuclear history and present nuclear politics would support a doctrine one step more rigid-a doctrine that nuclear weapons should be used only in response to their use by others—commonly called no first use. This posture was urged by some for NATO when the dominant opinion was that only a readiness for first use would deter communist aggression. President Carter moved toward this posture when he declared in 1978 that the United States would never use nuclear weapons against a non-nuclear weapon state unless that state were acting as the ally of a nuclear power. The last clause was aimed at the possible case of an aggressive ally of the Soviet Union; the case lively in memory was that of North Korea in 1950-53. In 1993 the clear priority for American military planners is that they should aim to keep their present capability to deter or defeat a conventional aggressor by conventional means. There is no early danger that this capability will be lost, and accordingly the present leaders can reaffirm the 1978 assurance. Indeed, taking account of the end of the Cold War, they can strengthen it. There is no present obstacle to an American declaration against any use at all of nuclear weapons against a non-nuclear weapon state.

In recognizing the possibility of a future case in which there might be justification for a use of nuclear weapons in defensive last resort, we are simply resisting the notion that our country can be certain, a priori, that there will never be a case when such use might be the least bad choice. The steady policy of the United States should be to do its part, by arms control, conventional capability and the general worldwide reduction of nuclear danger, to ensure that no such terrible choice ever becomes necessary. All of these lines of action can and should be pursued under a doctrine of defensive last resort.

The doctrine of defensive last resort has one further advantage. It corresponds to the nuclear policy of other nuclear weapon states much more closely than a flat policy of no first use. At present Russian possession of nuclear weapons helps to give the Russians confidence that they do not endanger their homeland when they accept the independence of Eastern Europe and the breakup of the old Soviet Union. As long as the Russians have the bomb, there is little danger of large-scale attack by others on Russia for the decisive reason that no one could expect them to accept any large defeat without resort to the bomb. There is a parallel protection for France in its nuclear capability; without it Paris might be deeply uncomfortable with its larger but bomb-free German neighbor. The position of Britain is less clear-cut, because the surrounding seas protect it from those nearby and because the possible threat from Russia is not only distant but more nuclear than conventional—requiring defense by deterrence, not by a last resort to first use.

But it is not surprising that among the five permanent members of the Security Council only China has declared an unqualified policy of no first use. The Chinese have become confident, after conflicts with Soviets, Americans, Indians and Vietnamese, that they can hold their own against all possible opponents without resort to nuclear weapons except as a deterrent to the nuclear weapons of others. If Americans had only themselves to defend they could make a parallel declaration; no one is going to threaten the United States conventionally on its own territory. But as it is, with contingent obligations overseas and friends who still rely in some measure on its nuclear protection, and with other nuclear powers sharing its caution for their own reasons, it makes sense for the United States to have a doctrine one step broader—a doctrine of defensive last resort.

This doctrine also fits the requirements of the three unannounced nuclear weapon states—Israel, India and Pakistan. The Israeli case is the clearest: the underlying motivation for the Israeli bomb is clearly the reality of vastly outnumbering unfriendly neighbors. The fact that the Israeli bomb is not for casual use is evident both in the desperate conventional battles that have been fought without its use and in the intense Israeli commitment to conventional strength. Evidence on Indian and Pakistani policy is thinner, but the general weight of informed judgment is that each country is governed more by fear than by ambition, and that each feels its nuclear need primarily because of perceived danger from the other. In this situation the immediately urgent requirement is to reduce the level of tension between the two; the first contribution that nuclear policy can make is that of mutually recognized moderation in weapons development and deployment. Sharing in the worldwide adoption of a doctrine of defensive last resort could be helpful here.

The American interest in a worldwide effort against any use of nuclear weapons is clear and strong, and it is equally clear that it will take visible American leadership to ensure such an effort.

Reducing Nuclear Reliance


We support the new budgetary priorities that reflect the end of the Cold War and the need to attend to domestic concerns, but we also believe it makes sense to keep room, as a matter of the most elemental national defense priority, for a financially solid program of reducing nuclear danger.

American budgetary priorities in action against nuclear danger should be shifted away from U.S. nuclear deployments to support cooperative reduction with Russia, encourage full acceptance of non-nuclear status by other successor states like Ukraine, and reinforce international action to prevent further proliferation all around the world.

The United States should think about all its national actions and expenditures affecting nuclear danger with a recognition that there is no a priori reason to prefer arms to arms control. What is essential is to understand the particular advantages and disadvantages of particular choices and act accordingly.

During the Cold War Americans thought the best way to reduce the Soviet nuclear threat was by a fully matching set of nuclear developments and deployments. Now the best way is cooperative reduction. The last Congress and the Bush administration both took remarkable steps in this direction. Congress, with bipartisan support, assigned $400 million a year, in the Defense budgets for 1992 and 1993, to the task of assisting the nuclear weapons cleanup in the old Soviet Union. The executive branch in 1992 agreed in principle to purchase up to 500 tons of formerly Soviet weapons-grade uranium. These actions had general bipartisan support, demonstrating that there is wide understanding of their contributions not only to arms reductions but to the national security of the United States. They should also be seen as the first steps toward further cooperation, by all the nuclear-capable states, to limit the supply and control the diffusion of fissionable material.

Need for action of this sort will continue and expand. American readiness to help will strengthen the hands of all those in Russia and other successor states who prefer the turn against nuclear danger to any more adventurous course. Moreover, actions of this kind, when they provide American dollars for activities in the common interest of reducing nuclear danger, can also be a general reinforcement to the economy of Russia and other affected states, and therefore serve the broader American interest in the political success of democratic and moderate governments.

A quite parallel value attaches to the strengthening of international arrangements for limiting nuclear proliferation. If agencies like the International Atomic Energy Agency are to have wider responsibilities and powers in this field, American reinforcement will be essential, and the more so because both habit and interest will produce strong demands from other states that if the IAEA is to be stronger in enforcement it must also be stronger in its work for cooperation in peaceful nuclear uses. If the IAEA is to get a sufficient flow of information about trade in relevant materials and deices, there must also be reinforcement of the most important single source of information, American intelligence—and indeed such reinforcement has begun.

There is a complex question about the possible future role of a separate agency of the Security Council, functioning like the special commission that has been necessary in the case of Iraq. It is not clear that the IAEA, with its inherited ethos of trust in those it monitors, can do all the hard work alone.

What is clear is that if there is to be timely international action against suspect states, there must be readiness to act in the U.N. Security Council, and here American support will be crucial. When there is also need for a U.N. blue-helmet force, it will help greatly if there is a new American readiness to participate. Similarly, if there is to be effective regional action that encourages restraint and penalizes its opposite, there must in most cases be American readiness for a share in such action. More generally, as the case of Saddam Hussein illustrates almost too clearly, the power that backs up any insistence on unwelcome inspection is likely to be insufficient unless Americans are visibly able and ready to help. Fortunately it is probable that very few cases will be as sharp as that of Iraq. What is largely missing so far on the American scene, and greatly needed, is a major effort, always including leaders from both parties, to make it clear to the American people that U.S. assistance of this kind is as much a matter of attending to nuclear danger as any direct payment for American nuclear weapons systems. The funds for such work should somehow be considered a part of a national bud-et against nuclear danger, because treating them in this way will make clear not only what they are for, but how moderate they are in comparison with the tens of billions the United States will still be spending on all aspects of American nuclear weapons even as such forces are reduced.

Phasing Out Nuclear Testing


The United States can strengthen its own effort against nuclear danger by durable changes from its past policy on such matters as testing and strategic defense.

The problem of nuclear testing presents both an obvious opportunity and a somewhat less obvious difficulty for the United States. The comprehensive test ban has been debated for more than thirty years. In principle the United States has supported it, but in practice the Cold War produced nuclear fear and competitiveness that prevented agreement. Now, with the end of the Cold War and the beginning of nuclear moderation, the United States does not need tests for new warheads with either larger or smaller capabilities than those it now has. America's present large-yield warheads are quite destructive enough, and at the lower end of the scale it would do much better to seek conventional weapons with improved accuracy and penetration.

It is excellent that the Bush administration in 1992 ruled out nuclear weapons tests for new designs, and it is wrong for anyone to encourage the nuclear weapons laboratories to seek support for such designs, except for safety improvements. Programs of this kind, indeed, would sharply weaken the American effort to reduce nuclear danger. They would make it plain to the rest of the world that the U.S. policy of reduced reliance on nuclear forces was for export only.

In this situation the most important past arguments against a comprehensive treaty fade, and the historic arguments in its favor come to the fore: it will help limit the spread of weapons; it will help dampen the competition among those who already have warheads.

More generally, the end of testing has been an asserted objective of all members of the Nonproliferation Treaty for almost twenty-five years. In that time the United States, which has always found reasons of one sort or another for resisting a ban, has come to be perceived as deliberately hostile to an objective it formally accepted as part of the bargain struck between states with and without nuclear weapons in that treaty. There is therefore a considerable political cost in any continued American opposition to a comprehensive test ban.

For all these reasons we support the emerging U.S. policy of 1992: a phased approach to an internationally agreed end of testing during this decade, an approach that permits a limited number of tests in the 1990s, with the primary objective of safety. In this period overall safety will also be improved if the deep reductions of START II are carried out as scheduled, because many present concerns about safety center on warheads that will be retired in these reductions; in this sense nuclear arms reduction is itself a contributor to weapon safety. For remaining nuclear weapons the most important safety goals can be achieved by the cut-off date, September 30, 1996, that was enacted by Congress late in 1992.

It may well be true, as partisans of testing often argue, that American testing is not a cause of nuclear ambition in a leader like Saddam Hussein; such leaders will want the bomb, if they can get it, with or without American underground tests. But the comment misses the real point: a great many people, in many countries, who are strong believers in a common effort against nuclear danger, have learned over a long period to put a test ban high on their list of priorities, and to measure the seriousness of the nuclear weapon states by their readiness to support such a ban. The United States does not need new nuclear capabilities, and it would strengthen its position among. people whose help it needs by supporting a nuclear test ban.

Still a limited and temporary exception for safety testing makes sense. The safety of nuclear weapons is a common interest of all states, and since we cannot expect an early end to the current stockpiles—only their steady and sober reduction—limited safety testing is justified to reduce real concerns about warheads remaining-in the stockpile. A limited and temporary exception for underground safety testing will not seriously affect the prospects for a comprehensive test ban treaty, especially if the safety testing is conducted as openly as possible so that the international community of nuclear experts can be assured that safety is indeed the object of such testing.

The need for safety tests is not good news for those who have both a deep commitment to early achievement of a complete test ban and a long experience of delays defended on the ground that there was this or that continued justification for testing. But the reality is that some American weapons are not as safe as they could and should be. They can be made properly safe, but not without some time and some tests. While the 1996 cut off in the new law does not hold if a foreign state tests after that date, it is a reasonable deadline for American safety testing. The passage of this conditional test-ban legislation is a landmark showing the new post-Cold War priorities of the American government.

A comprehensive test ban extending over generations would eventually bring a gradual erosion of confidence in the reliability of remaining nuclear warheads. (Reliability in this context means dependable explosion on command; it is not the same as warhead safety.) This would present new issues, and the proper response will depend on the degree of progress that has been made by then in reducing both nuclear danger and the traditional reliance on nuclear deterrence of high technical precision.

The congressional action of 1992 has laid the basis for U.S. leadership in reaching a comprehensive test-ban treaty while also attending to real problems of safety. Such a treaty, much more than a fragile moratorium, will weigh against proliferation, and indeed against enlargement of nuclear arsenals.

Strategic Defense


The case of strategic defense is one of the best examples of the requirement that we respect the realities of nuclear warheads. There is a recurrent tendency to ask technical experts to do more than nature permits—to make us safe, by science and technology, from what science and technology have made possible. Sometimes technological enthusiasts contribute to the confusion by advertising more than they can deliver. The case that is currently relevant is that of strategic defense. There really is no present prospect that all-out defense can outrun all-out offense in nuclear warfare, because of one simple reality: the overwhelming destructiveness of every single nuclear warhead. A kill-rate of even ten percent will defend you against ordinary aircraft with ordinary bombs—the largest bomber force will fade in a month of such continued attrition. But with nuclear warheads a kill-rate of 90 percent—extraordinarily hard to achieve—is wholly inadequate. Even 95 percent is not enough: for the smallest American and Russian forces in sight at the end of the century, it could mean some 150 warheads on homeland targets. And no one is now asserting that a defensive force of this quality is possible by any technology at all.

Once overall strategic defense is put aside as currently impossible—which is in reality what the U.S. government has already done in the last two years—one can address more realistically the question of what should be done.

The United States can keep a sharp eye out for some real possibility of a technical revolution between strategic offense and defense; prudence requires no less, and the United States should not allow its bad experience with the Strategic Defense Initiative, as originally advertised, to make it imprudently inattentive. It can also sustain an active and flexible program consistent with the Antiballistic Missile Treaty of 1972 on the less difficult but still demanding mission of defense against limited strikes. In regional conflicts such defenses can have both practical and psychological value against short-range missiles with less-than-nuclear warheads, and the development and production of such systems can proceed in accord with the ABM treaty. There are still many outstanding questions about the potential effectiveness of nationwide defenses against only one or a few nuclear-armed missiles of intercontinental range launched, perhaps, accidentally. These can be addressed by a treaty-consistent research program. But effective defense against a large force of nuclear-armed missiles would require a wholly different level of effectiveness simply because of the requirement of a kill-rate near 100 percent. A gulf of ignorance and uncertainty still separates the United States from any prospect of a defensive capability that would make it useful to seek a change in the ABM treaty. It is important to remember that a basic principle undergirding the strategic reductions of recent years is that any nation remains inescapably vulnerable to a large-scale attack. It is hard to accept this reality, but not so hard to see that all one can get from imperfect defense is the same situation with multiplied costs and fears.

Lowering Weapons Deployments


The U.S.-Russian reductions now under consideration should not be seen as the end of the road; in the long run Americans should work for still further reductions and improvements.

There is no higher priority for the reduction of nuclear danger than the orderly execution of all that is promised in START I and II and in the 1991 exchange of unilateral assurances on tactical warheads. Over the next few years the execution of these agreements and assurances should have operational priority, but the resulting. level of 3,000-3,500 strategic warheads on each side should not be accepted as permanent. Any such acceptance, over time, could pin both sides into levels of expense that would be undesirable in themselves and destructive in their impact on the political relations between the two powers.

If we should disregard the present size of the two great arsenals and simply ask ourselves what is the lowest level of strategic strength, measured in warheads, that would give both sides ample assurance, both against nuclear attack and against the other side's breakout to any significant strategic superiority, we would not come out with the currently agreed range. That range reflects an understandable respect for existing habits of targeting, indeed for existing targets, that will not survive the changes already in prospect.

It may be better to examine the long-run prospects for lower warhead ceilings by a fresh start upward from zero, with ample safety factors. Assume that the United States will want to be able to hit both military and industrial targets, while keeping substantial forces in reserve and having insurance against losses from a surprise attack. The table on the following page shows two alternative force levels for these purposes.

If the reductions now being considered are carried out successfully, the United States and Russia may well be ready by the year 2000—and perhaps sooner—to go to still lower levels. Numbers like these might raise a question, in Moscow or Washington or both, about the comparative strength of British or French or Chinese forces. All three in 1993 appear to be headed upward, toward numbers of strategic warheads in this same general range. But there is nothing immediately troubling here. If a continuing improvement in superpower nuclear relations permits the big two to agree on such reductions five or ten years from now, there may well emerge a parallel preference for reductions elsewhere, but it may also become necessary for Americans and Russians to accept forces in those countries that become relatively larger as time passes. Any resistance to such a relative change would be based more on habits of thought than on true national interest. Happily there is some evidence from the events of recent years that nuclear moderation can be as contagious as nuclear competition was in the Cold War years.

To some arms controllers, even these reductions will seem like slow work. We can only remark that as it took time to get ourselves burdened by these enormous forces, so it will take time—though much less-to cut them back. Nor is 1,000-1,500 warheads the lowest level obtainable by the early 21st century. Our concern here is simply to make it clear that there is no good reason to accept the much higher targets of 1992 as the best that can ever be done.

The true interest of both sides is that each step toward lower and less threatening deployments should be seen as a step forward by healthy majorities on both sides, so that nuclear moderation remains, for both, as it was in 1992, a broadly popular policy. In particular America should avoid the temptation to use a season of great Russian economic stress to drive for one-sided advantages, either by negotiation or appropriation. At any levels now in prospect such marginal advantages would have no real value, and their political damage could be serious. Complaints against such "advantages" supposedly gained by the other side have helped hard-liners to gain strength in both countries in the past. A time of general political success is not a time to risk that success by making the mistake of "piling on."

Finally, reductions in warhead numbers are not the only means of nuclear restraint. The same broad objectives can also be served by reducing vulnerability, improving controls, avoiding destabilizing surprises and making a broad attack on controlling and limiting weapons-grade material.

Immediate Steps and Long-Term Progress


Moderate and immediately practicable courses of action like those recommended here for the reduction of nuclear danger must also be held to the test that they should contribute to the prospect of still larger progress over the longer term.

The question could be raised, why not go further? Can we not roll back or even end the nuclear programs of some existing nuclear states? Can we not insist that there must be absolutely no first use of these weapons by anyone, especially ourselves Can we not get to numbers of warheads, even in the United States and Russia, that are well below the four-figure levels we have considered? Can we not, perhaps, revive the old notion of putting all nuclear weapons in the hands of an international agency? Could a properly arranged monopoly be operated by a selected government or governments on behalf of the United Nations? Should we not, in short, be more bold?

We have two answers to this set of questions. First, by the standards of the past and by what we know of the ways of nations, our proposals are more demanding than cautious. re are asking for a sustained and genuine partnership of the nuclear superpowers in a shared program of national reduction, and of proliferation avoided in the other countries of the old Soviet Union. We are also proposing that the United States call on all others to join in giving a new worldwide priority to the avoidance of further nuclear proliferation. We have argued that the nuclear moderation of every existing nuclear weapons state will be critical to any such program.

Second, these short-term efforts must always be consistent with further progress later toward the further reduction of worldwide nuclear danger. We have examined all our proposals by this test, and they all make further steps easier, not harder. In particular it is right to urge greater strength for international agencies, greater use of open formal agreements governing numbers, safeguards and disclosure, and lower reliance on the threat of nuclear warfare.

More generally it is right to begin to displace the inherited distinction between those with weapons and those without by a wider assertion that in reality all the nations should be on the same side, against nuclear danger, whatever their present degree of reliance on nuclear weapons. This last proposition is so clear and strong that a steady respect for it is the best single guide to action for any country, both now and in the long run.

In their actual behavior recent presidents and Congresses have done much to reduce the level of nuclear danger in the world, but they have done much less well in explaining and winning support for the new policy of partnership against that nuclear danger that they have joined with foreign leaders to begin to create. Effective future action will require a stronger policy of public explanation from American political leaders than ever before.