New Objectives for Nuclear Weapons Policy
Causes for conflict are not going to disappear and neither will the potential for disaster inherent in the existence of nuclear weapons. Given the terrible consequences of modern war and the likely continued presence of nuclear weapons, security considerations should impel the developed countries, particularly the United States and the Soviet Union, toward resolving differences cooperatively despite differences in policies and interests. This is particularly true when the conflicts have a potential nuclear dimension.
We believe that the creation and maintenance of effective cooperative security institutions and arrangements, particularly in the nuclear area, should replace containment of the Soviet Union as the United States' major security goal. This goal is not new. It was introduced in 1945 and has played an understated but significant role in U.S. policy ever since. Placing that goal at the center of policy will require the United States to take a different approach to nuclear deterrence and nuclear deployments. We address these two related matters in turn.
U.S. DETERRENCE POLICIES
The two primitive fission bombs detonated over Japan caused approximately 150,000 immediate casualties. Since the end of World War II, the
Page 15world's stockpile of nuclear weapons has grown to some 50,000 warheads, most of which have greater explosive power than those used on Japan. The use of only a small fraction of today's nuclear weapons would be an unprecedented catastrophe. Over almost the entire postwar period, minimizing the risk of such an event has therefore been a major goal of U.S. policy.
At the same time, the adversarial relationship between the United States and the Soviet Union has led to a U.S. nuclear deterrence policy with the twin, if sometimes conflicting, goals of minimizing the risk of nuclear war and deterring certain classes of conduct by the Soviet Union and, at times, other nations as well. The objectives of deterrence have been the prevention of nuclear attacks on U.S. or allied territories and forces abroad, as well as some kinds of aggression not involving nuclear forces, such as the invasion of Western Europe.
We cannot know the extent to which either nuclear deterrence or the present vast numbers of nuclear weapons have deterred war. The existence of nuclear weapons has certainly added caution to the conduct of U.S.-Soviet relations. The risk of regional conflict escalating into unlimited warfare has constrained the behavior of both countries. Beyond these qualitative observations, we have no way to assess how the number and type of nuclear weapons, their deployment, or the declaratory policy governing their mission have determined the two nations' conduct during the Cold War. The end of the Cold War gives us both incentive and opportunity to examine afresh the objectives of nuclear weapons policy and the mission of nuclear weapons.
The principal U.S. objective remains the prevention of nuclear war, which perhaps alone of all external threats could threaten the continued existence of the United States. Preventing nuclear war means first of all preventing nuclear attack. It also means preventing the occurrence of situations so dangerous that they might lead to the use of nuclear weapons against the United States. In the past, these postulated situations included some massive nonnuclear attacks, and the United States sought to extend nuclear deterrence to prevent them.
Any concept of “extended” deterrence, to deter massive nonnuclear attack on one's own country or its allies, suffers from a basic tension of values. The credibility of a U.S. threat of nuclear retaliation against a nonnuclear attack, however dire its consequences, is impaired if the U.S. homeland would then itself be subject to a nuclear counterstrike (as symbolized by the remark that the United States would not have traded New York for Paris). Therefore, an aggressor planning to initiate a massive nonnuclear attack may or may not be willing to accept the risk of nuclear escalation.
So long as war is possible and nuclear weapons exist, this tension cannot be eliminated; for many years the West decided to live with it. But the
Page 16tension could be reduced. Reduced risk of conventional invasion, such as has now occurred in Europe, can considerably diminish the tension, and has in fact already led to a major decrease in the U.S. nuclear presence in Europe. The tension can be further reduced by shrinking the size of the nuclear forces on both sides. Lessening the need for time-urgent steps in major crises, which now also seem less likely, would help as well.
The size and nature of the present nuclear forces were reached on the basis of Cold War conditions that no longer hold. Reducing the forces bilaterally would provide both sides' political and defense establishments with needed confirmation that these conditions have changed. Retaining the forces as they are would convey the opposite message.
We note that, at any level of force, the success of deterrence can never be fully assured. This is one reason for limiting the world's inventory of nuclear weapons, thus limiting the worst case consequences of the complete failure of deterrence. Unfortunately, so long as cities are subject to nuclear attack, either as deliberate targets or through collateral damage, the consequences of failure would still be very large even at the reduced levels discussed in this report. We do not believe that the present world situation will lead to relief from this danger, although we believe the steps recommended here can move us toward such relief.
We also note that retaining nuclear weapons in excess of what is needed to deter the Soviet Union is a liability for other reasons. At a time when both the defense and the overall federal budgets face severe constraints, maintaining excessive numbers of nuclear weapons would come at the expense of improvements in their quality, safety, and survivability. The potential for accidents and diversion, although mainly determined by qualitative factors, nevertheless increases with greater numbers of weapons.
Finally, nuclear weapons require the maintenance of an infrastructure for safety and security involving continuously alert and disciplined personnel, transportation, the control and remediation of the environmental effects of the military fuel cycle, and the replenishment of tritium, all of which are burdens that increase with the size of military nuclear stockpiles.
In judging future requirements for nuclear weapons, one must avoid both the conclusion that acquiring more weapons is always more “conservative” and the conclusion that reducing the number of weapons is always more “stabilizing.” Instead, we seek an appropriate balance between the positive and adverse effects of nuclear weapons in the face of many uncertainties. A departure from that balance in the direction of excessive nuclear weapons would increase the financial and environmental burdens and could even
Page 17increase the likelihood of nuclear war. On the other hand, reducing nuclear weapons without appropriate attention to survivability could lead to dangerous instability in the event of a crisis.
Our assessment of the optimal balance in the post-Cold War era leads us to conclude that numbers of weapons much below present stockpiles could provide adequate strategic deterrence. The extent of the reductions, however, cannot be rigorously derived for several reasons:
The international security framework under which nuclear weapons will exist cannot be predicted, although alternative developments and cooperative security arrangements can be identified.
The fundamental tension between “deterrence” and “credible use” can be reduced but cannot be eliminated. Thus a sharp delineation between what hostile conduct is to be deterred by nuclear weapons and what conduct will be unaffected is unlikely to emerge.
Deterrence describes a state of mind of national leaders, not a more concrete reality. Thus a quantitative basis for specific “requirements” of nuclear forces to support specified deterrent missions cannot be established.
With these caveats, we conclude that both the Soviet Union and the United States can be more secure with fewer, more survivable, nuclear systems.
At the same time, smaller deployments of nuclear weapons must remain secure from attack, as must the warning, communication, and command and control systems that are needed to operate them. Without such survivability, reduced deployments could contribute to instability rather than to deterrence. Furthermore, the United States cannot gain its security at the expense of increased insecurity for other nations, in particular the Soviet Union. U.S. nuclear forces of all classes should be designed for maximum survivability and the United States should not attempt to lessen the ability of the Soviet Union to develop survivable forces. The transition from the current large arsenals to smaller forces, either through independent action or negotiated arms control, must be managed to preserve, and where possible increase, the survivability of nuclear forces throughout the process.
The minimum number of nuclear weapons needed to deter a nuclear attack depends on the country to be deterred and the circumstances under which deterrence is to be effective. The number needed to deter a nuclear attack on the United States from Great Britain or France is zero. This may become the case with the Soviet Union. We are not there yet. For our work in this report, rather than attempt to prescribe the minimum number of weapons needed to deter at various stages between the Cold War and some future era, we describe cuts that in our judgment would leave ample forces to cover any likely target for nuclear weapons and at the same time would
Page 18be politically feasible under certain likely specified future situations. At each stage, we assume that overall parity between the United States and the Soviet Union would be maintained, if only for the sake of political stability, although parity among the individual components of the force (such as submarines or mobile missiles) would not be necessary.
These deterrence policies would continue the dependence we have had on offense-dominating defense, that is, on “offense-dominated stability.” In principle, there could be “defense-dominated stability” if technology permitted nations subject to nuclear attack to prevent the arrival of nuclear weapons through adequate defenses. This is not the place to discuss critically the long-range technical prospects for converting the current offense dominance to a defense-dominated world or to examine whether a transition to such a world could be executed stably. The advent of nuclear weapons has strongly tilted the traditional offense-defense competition in favor of the offense since each delivered nuclear weapon possesses such great explosive power. This fact, combined with the wide variety of means to deliver nuclear weapons and to ensure that they reach their targets, has led to the present situation. No technical developments are expected to upset this condition in the foreseeable future.