Chapter 4
A Defense Transition?
SDI and Strategic Stability
In evaluating the technical prospects for early deployment of a partial spacebased defense, the Defense Science Board's Strategic Milestone Panel raised the key question: Even if they should prove technically feasible, are strategic defenses desirable?[1] Plainly, this question must be answered before the United States proceeds beyond fundamental research. Would deployment meet the objectives of a sound military strategy? "Strategy" has been defined by one of its most astute modern students, Basil Liddell Hart, as "the art of distributing and applying military means to fulfill ends of policy."[2] The most fundamental end of U.S. foreign policy is to promote security and in particular to deter aggression against the United States and its allies. Therefore, the case for deploying defenses must rest on a positive assessment of their ability to serve that objective as compared with other military and diplomatic options.
All discussion of the strategic implications of a defensive system must remain speculative so long as SDI's technological prospects and the adversary's likely reactions remain uncertain. Even when the technological issues are relatively well understood, uncertainty about the adversary's responses—or what in technical jargon is called the "responsive threat"—clouds the issue. For example, the technologies needed for terminal ground-based defenses are understood well enough for deployment to be practical, but their cost-effectiveness and utility cannot be determined without taking into account the changes they provoke in the configuration of the responsive threat. Even in this instance, then, any discussion of the strategic implications of deployment is bound, as we
have noted, to be speculative. Benjamin S. Lambeth, in examining the Soviet view of SDI, has described the problem well:
Assuming that the SDI does lead to a deployed first-generation system, the Soviets will be driven to respond within the limits of their technical and budgetary resources. Any effort to anticipate this response must start with a cataloguing of options that are technically feasible, intuitively reasonable, and consistent with past Soviet practice. But that is fairly straightforward compared to the far more daunting task of predicting what they will do. The latter calls for a forecast in the presence of compound uncertainty regarding not only the concerns, motivations, and intentions of the Soviet Union but also the ultimate accomplishments of the United States in the SDI realm.[3]
Opinion is sharply divided among those who have ventured into this speculative realm. Those who favor SDI do not all envision a long-term transition to a nuclear-free world. Either because they agree that the prospects for a comprehensive defense are still uncertain, or because they favor continued reliance on nuclear weapons as the basis of deterrence, some SDI proponents, such as Zbigniew Brzezinski and Henry Kissinger,[4] maintain that defenses are a desirable way to protect offensive weapons because they make deterrence more credible. Other supporters of SDI, like Edward Teller, argue that defenses could provide some protection for population and military targets and that some such defense is better than none.[5] Others agree with Reagan that as defenses are deployed, stage by stage, the existing imbalance between offense and defense will be corrected. After a period of "defense dominance," the threat posed by nuclear weapons would effectively be eliminated, even though some of the weapons themselves might remain in existence.[6]
On the other side, opinion tends to be more unified. Those strongly opposed to SDI on strategic grounds argue that deployment would inevitably be destabilizing and therefore counterproductive. In their view, the current condition of mutual deterrence rests on the assurance that both sides are able to threaten the other with unacceptable destruction in the event of an attack. To the extent that defenses would diminish the credibility of the retaliatory threat, they can produce only instability.[7] Some in this camp concede that terminal deployments to defend offensive weapons could be stabilizing, because they would help to bolster confidence in the efficacy of the retaliatory threat. They are also likely to point out, however, that the same objectives can be achieved in other, less provocative, ways and that even the deployment of limited defenses by one side is bound to result in efforts to improve offensive measures as well as to match the other side's defenses. The result, in this view, can
only be an unrestricted, destabilizing arms race in which either side may be tempted to suppose that it has achieved or can attain superiority, in which case it can attempt to threaten the other for political gains. Another destabilizing effect of SDI is that either side might fear it is losing its ability to retaliate, in which case it might be tempted to risk a preemptive attack. (In technical terms, the first sort of instability is called "arms race instability," the second, "crisis instability.") If space systems are deployed, then both sides are also likely to deploy anti-satellite systems. In a climate of hostility and suspicion, such deployments could lead to crises that would prove unmanageable. Richard Ned Lebow has envisioned one such scenario:
The United States is committed to developing a space-based defense against ballistic missiles, something the Soviets strongly oppose. Moscow has already dropped hints that it is prepared to interfere with the deployment of such a weapons system. Suppose the United States, deeply committed to the project, disregards Soviet warnings and at some point begins to put important components of a missile defense in space. To show their displeasure, the Soviets orbit space mines in the vicinity. The United States, in turn, sends up a shuttle mission to remove or disarm the threat. But the mines, having been salvage fused (set, that is, to go off if tampered with), explode and kill six astronauts.[8]
In view of such risks, the critics of SDI argue that the current prohibition of BMD is preferable to the deployment of defenses because it promotes stability and allows for efforts to reduce nuclear arsenals on both sides. Although few of the critics hold that disarmament measures can succeed in eliminating nuclear weapons entirely, they generally contend that the best approach is, first, to eliminate weapons that could tempt either side to contemplate a first strike and, second, to continue to rely on mutual deterrence, but at sharply lower levels. With further steps to improve confidence and communication, potential crises could be better managed and the way would be open to a broad range of agreements designed to curtail the arms race and to lessen the danger of war.
Supporters of SDI rebut this view by pointing out that nuclear deterrence is dangerous in view of the risks that accompany crises when the threat of retaliation may fail. In any event, they argue, the strategy is becoming less and less credible. Thus Colin S. Gray cites with approval Fred C. Iklé's observation (in 1973) that it is probably unreasoanble to expect nuclear deterrence to work indefinitely.[9] As the U.S.S.R. has increased and diversified its strategic offensive arsenal, supporters argue, neither the United States' adversaries nor its allies will consider a threat
of nuclear retaliation credible. "A unilateral Soviet deployment of … advanced defenses," the Reagan administration has argued, "in concert with the Soviet Union's massive offensive forces and its already impressive air and passive defense capabilities, would destroy the foundation on which deterrence has rested for twenty years."[10] SDI advocates also argue, as Secretary Weinberger did, that the strategic situation is asymmetrical because if defenses were deployed by the United States, they would be used only for that purpose, whereas the Soviets, presumably because they are inherently bellicose, could not be trusted to refrain from using defenses in conjunction with offenses for aggressive purposes.[11] In any case, SDI proponents argue, if the Soviets should wish to avoid confrontation, they will cooperate with the United States in achieving a managed transition to greater mutual reliance on defenses, as both sides simultaneously reduce offenses and introduce defenses. The United States' pursuit of strategic defenses, they claim, could encourage the Soviets to follow suit; a defensive arms race would be preferable to an offensive one. Whatever the Soviets should decide, however, it would be in the United States' interest to develop whatever defenses prove possible in order to counteract Soviet offensive improvements and move toward "a defense-dominated strategic environment in which the United States and its allies can deny to any aggressor the military utility of ballistic missile attack."[12] In the first phase, a deployment of defenses would be designed to reinforce the credibility of retaliatory deterrence ("deterrence by uncertainty"), and in the later stages it would achieve deterrence by protecting against a missile threat ("deterrence by denial"). In the best of all possible worlds, the condition of mutual assured destruction would be replaced by that of mutual assured survival or security.
Even those who advocate such a defense transition do not generally suppose it will be easy to achieve or that its principles will prove readily acceptable. Paul H. Nitze, for example, has emphasized the need to engage the Soviets in elaborating the details of the transition. Otherwise, he argues, it would be more expensive and harder to achieve, although he readily admits that making this cooperative transition "could be tricky."[13]
In its most radical form, a defense transition would entail a fundamental reversal of the strategy of nuclear deterrence developed over forty years. Such a reversal requires careful justification. If the Soviets were on the verge of achieving some defensive breakthrough, no such justification would be needed. As we have noted, however, the Soviets
appear to be doing no better than the West in this effort. Although some level of research can be justified as a hedge against a possible breakout by the Soviets, a commitment to the deployment of comprehensive defenses is warranted only if they are desirable whatever the Soviets may do.
The opponents and advocates of SDI have such divergent assumptions and predictions that the issue cannot be resolved merely by appeals to logic and evidence. We cannot know enough about the future results of research or about the future of the East-West conflict to foreclose the debate or resolve it to universal satisfaction. Is there perhaps a middle course? We think so. This course would call for continued pursuit of research without premature commitment to deployment and in the context of an arms control regime designed to reassure both sides that neither can achieve some unilateral breakthrough that will confer strategic superiority.
The Middle Course
Even among those skeptical about SDI, there is a recognition that defenses may, in principle, be desirable. Thus, former secretary of defense Harold Brown has observed that "if technology promised the certainty of a successful defense of the American population against strategic nuclear weapons, or even a high probability of overall success, it is likely that policymakers and defense analysts would agree that the United States should seek that objective." Along with many others experienced in military technology, however, Brown strongly doubts the objective is attainable: "Technology does not offer even a reasonable prospect of a successful population defense."[14]
The existing situation is fearful to contemplate. Although the threat of retaliation has inhibited armed conflict, it has sometimes perilously teetered on failure, as in the Cuban Missile Crisis of 1962. So long as the superpowers maintain their nuclear arsenals, war could start because of an accident, a miscalculation, the escalation of a conventional war (perhaps one between client states), or because either side might become convinced that the other side was about to strike and conclude that going first is preferable somehow to mere retaliation, however awful the result of attacking first might be. Should deterrence fail, for whatever reason, the resulting war could be catastrophic and not easily brought to a halt before much of the fabric of civilization is destroyed. An effective defense, even one that is not completely leakproof, could, in principle at least, offer some protection in the event that deterrence failed.
The trouble with this reasoning is, of course, that there can be no assurance at this stage that defenses will succeed against future offensive developments or that they would actually perform as designed. We can be assured, however, that an unrestricted competition to develop such defenses would cause political leaders on both sides to wonder whether their retaliatory capabilities were still reliable and whether their adversary was planning to take advantage of some presumed weakness. The fear that both sides might achieve some fundamental breakthrough would spur both sides to intensify their own efforts, making the total militarization of space a self-fulfilling prophecy. Given the extremely short time-lines involved, space-based defenses take political authority out of the loop entirely, opening the possibility for war to result because of a technical malfunction.
In view of the difficulties involved in strategic defenses, we must answer two questions: would the world be better off if defenses replaced offenses, and would the effort itself merely to develop and deploy them be destabilizing? Provided both sides recognize the risks and take the necessary steps to minimize them, research aimed at resolving SDI's extremely complex technological questions can proceed, at least for the immediate future, without causing such instability. In order to show that a postponement of deployment is the best alternative, we will review the argument in greater detail and present the options that now seem possible.
The Case For A Defense Transition
The argument for a defense transition assumes that the superpowers are in effect engaged in a technological race that will eventually produce something useful. Supporters of SDI often argue that the West has no choice but to pursue strategic defenses unless it is willing to accept Soviet domination. They contend that a determined Soviet offensive modernization and enlargement program is making the U.S. retaliatory force increasingly vulnerable to a surprise attack and that an undeclared Soviet SDI is designed to enable a breakout from the present mutual restraint on the deployment of BMD. If the U.S.S.R. should achieve a significant BMD capability, and if it should be in a position to threaten an attack that would degrade the West's retaliatory capacity (with the assurance that its defenses could absorb most of a U.S. retaliatory launch), it would also be in a position to contemplate a first strike, or at least to try to extract political concessions in a crisis. If the United States developed
defenses first, it would be in a similar position, although U.S. authorities insist that its superiority would not be used to threaten a first strike. In any case, supporters of SDI argue, the United States has offered to negotiate an orderly transition and even to share Star Wars technology with the Soviet Union so that neither side would be threatened and humanity could escape the peril of nuclear war.
The problem with this argument is twofold: SDI supporters greatly exaggerate the possibility of the erosion of the present situation, and they cannot demonstrate that a transition could be stable. The Soviet offensive buildup has not rendered the U.S. retaliatory force so vulnerable that the Soviets could conceivably have any incentive to launch a first strike. The Scowcroft Commission put to rest the fear that the Soviets might exploit a theoretically conceivable window of vulnerability.[15] Offensive developments in the United States, most of them initiated before the Reagan administration came into office, have kept the U.S. retaliatory force versatile and capable of surviving any first strike with more than enough residual power to inflict a fearful response. Even if the Soviets could count on destroying all or most of the U.S. land-based missile force, they would still face not only U.S. bombers, equipped with penetration aids and standoff weapons, but also a submarine-launched attack, which will soon employ missiles fully as useful as ICBMs for attacking hard targets. And, in any event, if any or all major components of U.S. strategic forces became vulnerable to preemption, the United States could always go to a launch-on-warning strategy, however undesirable that might otherwise be.
Critics have properly complained that SDI could lead in two very different directions. The original version, or "Star Wars I," would seek to realize Reagan's vision of a comprehensive defense of population and would not be primarily devoted to protecting retaliatory capacity—reasoning that such limited defenses would only perpetuate mutual assured destruction. It would almost certainly be necessary, therefore, to mount a full-scale layered defense, including space-based or pop-up systems, supplemented by an equally effective air defense capable of dealing not only with bombers and cruise missiles but also with low-trajectory ballistic missiles launched by submarines lying off the U.S. coast. Until such a system could be developed and put in place, at whatever considerable cost, the United States would still have to make substantial investments, at least at the prevailing level, in order to maintain and improve its offensive forces.
"Star Wars II" would be designed to provide a more limited defense,
one that would provide at least partial protection for elements of the retaliatory force and C3 facilities, while perhaps also providing some marginal population defense. Such a defense might enhance the existing form of deterrence but it would not meet Reagan's grand objective of making nuclear weapons "impotent and obsolete."
"These two visions of Star Wars," McNamara argues, "have diametrically opposite results."[16] Star Wars I would mean the elimination, or virtual elimination, of reliance on nuclear weapons to defend the United States and its allies. Star Wars II would mean, at best, a somewhat more reliable nuclear deterrent. It would not allow any abandonment of the doctrine of assured destruction but, instead, would complicate the commitment to flexible response should the Soviets deploy a comparable system of silo defenses. It is also not clear that such deployments would preserve whatever stability can be said to exist because both sides can inflict unacceptable damage in retaliation. As we have already noted, to the extent that one side perceived its defenses as significantly better, that side would have an incentive to launch a preemptive attack in a crisis. In this sense, crisis stability would be adversely affected. Again, as argued earlier, arms-race stability would also be jeopardized because the credibility of each side's nuclear deterrent would be threatened by any development or deployment of defenses. Neither side would have any incentive to negotiate agreements calling for equivalent levels of offense or even to abide by existing treaties. Even surveillance by national technical means might no longer be accepted once the strategic objective is to raise uncertainty about the effectiveness of any offensive deployments. In any case, a full-scale militarization of space would obscure the differences among types of satellites. Moreover, in a world in which both sides deployed "Star Wars" systems, those satellites originally deployed for defensive purposes would have the added mission of destroying the other side's space-based defenses in order to clear the way for retaliation. This is the so-called counterdefensive use of space-based strategic defense systems, clearly simpler (the ASAT mission is easier to achieve than missile defense) and more dangerous (the time lines in the counterdefensive mode are even shorter than those in the missile-defense mode).
Supporters of SDI might respond to such criticism by arguing that the effort to distinguish between Star Wars I and Star Wars II is an attempt to sow confusion. What is called Star Wars II should be understood, in their view, as a first step in a larger defense transition, the ultimate goal of which is Star Wars I. In the short run, a limited defense
may be advisable to guarantee the survivability of the ground-based leg of the triad as well as to protect C3 facilities so as to assure the continued credibility of "assured retaliation" before a population defense can be deployed. Even a limited defense could provide some population defense as well as protection against a small-scale accidental launch of nuclear missiles or an attack by a lesser nuclear power or terrorists. Indeed, the collective "footprint" of the projected force of ERIS missiles would cover most of North America. Those who advocate SDI argue that such initial deployments would provide experience with defensive systems and would persuade the Soviets of the seriousness of the U.S. commitment to a defense transition. By demonstrating the economic logic of defenses, moreover, an incremental approach to deployment would help persuade the Soviets to agree to a jointly managed transition.
But supporters of SDI would surely agree with Nitze that the transition to a comprehensive defense would be tricky. Even if offenses could be constrained at their present levels, defenses would be both expensive and hard to make survivable. To the extent that defenses would be considered effective, they might enhance the credibility of deterrence. But they might also lead to reckless action, by either side, such as mounting (and even using) counterdefensive systems. Efforts to promote cooperation in other respects might also become more difficult because neither side could be confident of its ability to protect itself either directly or by threatening a devastating counterattack.
Four Options
In order to come to grips with the strategic implications of SDI, it is essential to consider the four major alternatives and their implications.
|
|
All but the fourth of these options would permit continued research on strategic defenses without greatly increasing the risk of instability. The second would provide enhanced security only if it could be achieved in conditions of offensive and defensive parity. Both sides would deploy roughly comparable defenses and accept constraints on offensive deployments of an equivalent character. This option resembles the proposal for a defense-protected build-down, to be discussed below. Although we are sympathetic with the aim of this proposal, we believe it would be difficult, if not impossible, to implement. The first option has the virtue of being familiar and of having "worked" so far—i.e., having inhibited the two sides from becoming involved in a direct military conflict—but it is also open to criticism on various scores, not the least important of which is that if it should fail, the consequences would likely be catastrophic. The last option would be the least desirable, entailing large expenditures that would probably result only in greater instability. To show why, of all the options, we think the third is clearly preferable, we proceed to examine the implications in more detail.
Option I. Deterrence As Usual
For forty years deterrence of nuclear war by the threat of retaliation has been the guiding principle of U.S. strategy, despite early discussions of nuclear preventive war and later considerations of the possible first use of nuclear weapons in Europe, Korea, and Vietnam. There is evidence, reviewed in chapter 2, that although Soviet military strategists were slower to accept this principle, preferring to think of nuclear weapons as war-fighting elements, not only the Soviet political leadership but also the military elite have come around to the same view. As Michael MccGwire has observed, "the evidence now suggests that the Soviets will wait to launch their strategic forces until they have warning of a U.S. attack. In other words, they are braced for NATO and the United States to launch first in the course of a war."[17]
This form of deterrence is often said by students of strategy to be stable in the two senses of crisis stability and arms race stability, but
neither is foolproof. Both benefit from such factors as diversity and redundancy (as in the case of the U.S. triad of land-based, sea-based, and air-based weapons), hardening, dispersal, and security of communications. Both are impaired to the extent that each side relies on weapons with rapid hard-target kill capacity, because these weapons are hypothetically most useful in a first strike. The degree of stability that may realistically be expected under the best of circumstances is inevitably relative and uncertain, because the forces deployed by the two sides are different in character, because the propensity to use such weapons is affected by all the factors that may influence the judgment of political leaders in a crisis, and because accidents could trigger an unintended train of consequences. By its nature, moreover, this form of deterrence implies the possibility of nuclear war. The weapons are there, ready to be used at short notice, in case deterrence fails. If it fails even once, the consequences would be catastrophic. Realistic estimates suggest that a full-scale exchange of thermonuclear warheads, given the size of the arsenals available to both sides, could result in the deaths of hundreds of millions of people, with casualties possibly amounting to a billion or more, and in the contamination of whole continents, even if not the global "nuclear winter" that some analysts also fear.[18]
To the practical deficiencies of this form of deterrence have been added criticisms on the level of moral principle. The Roman Catholic bishops in the United States have criticized it because it violates the just war maxim, according to which an act of self-defense must not be directed against innocents. "Specifically," the bishops noted, "it is not morally acceptable to intend to kill the innocent as part of a strategy of deterring nuclear war." Because even counterforce strikes would kill innocent civilians, and since a limited nuclear strike would probably lead to a wider exchange, the bishops expressed serious reservations about the morality of current U.S. policy.[19] Fred C. Iklé has described mutual assured destruction as "assured genocide," though he does so in order to defend a policy of relying on counterforce targeting by accurate missiles with relatively low-yield warheads.[20]
Although the moral criticism of nuclear deterrence raises serious issues, it is open to several responses. One of these is that in the absence of international law and a supranational government capable of enforcing such law, states cannot be expected to observe the same moral restrictions that individuals ordinarily respect in domestic environments in which adherence to accepted norms is enforced by collectively authorized sanctions. States customarily put the survival of their societies ahead of other moral obligations. If survival requires inflicting or threatening
to inflict incidental harm on innocents, they will do so when necessary to national security, as governments have done through history. Such behavior may be deplorable, but moral indignation alone will not change it.
Just war principles, moreover, were formulated and applied, however haphazardly, during periods when a fairly sharp distinction could still be made between war among combatants on the battlefield and the extension of warfare to civilians who had little if anything to do with the conduct of war. In the era of total war, that distinction has been obliterated. Particularly in World War II, political and military leaders on both the Allied and Axis sides undertook campaigns of "strategic bombing" on the principle that a nation's ability to carry out a war depended critically on communications and transportation networks, manufacturing capacity, access to raw materials, and even the morale of the civilian populations. All of these became "legitimate" targets. In view of this precedent, and of the collocation of civilian populations and such targets, as well as the large lethal range of nuclear weapons, it is hard to imagine that a future war involving superpowers will be confined solely to the front-line combatants. As a result, the actual conduct of warfare will inevitably blur the distinction between belligerents and civilians.
Though the issue is debated among philosophers, there is a significant moral difference between posing a threat to use force in self-defense and actually using that force.[21] To be sure, if a threat is to be meaningful, it must be credible, in the sense that it reflects a readiness to make good on the threat. Even Leon Wieseltier, who defends this form of deterrence, admits it is immoral inasmuch as it is "a promise of murder." As he observes: "If the deed cannot be called moral, the threat cannot be called moral."[22] For this reason, he, like the Catholic bishops, finds a strategy of retaliatory deterrence provisionally defensible only if it is accompanied by serious efforts to achieve nuclear disarmament. Our views, while in general parallel to those of the Catholic bishops, have a different basis. We assume that the overarching moral imperative of our time is the avoidance of a nuclear holocaust. Arms control and disarmament activities that serve to support that goal are good; actions that work against it are bad. Nuclear force postures, policy declarations, and military deployments are to be judged in the same way. In the long run, we believe, human society must find some way to introduce a functioning legal order among states, and thus to eliminate both war and the nuclear threat. In the meantime, the goal must be to avoid nuclear war for as long as it may take to work out a better arrangement.
On moral grounds, we believe there is a significant difference between
threatening to retaliate and actually carrying out a retaliation. The threat and the overt act are not equally blameworthy. Most people would not criticize a property owner for posting a sign warning that his home is protected by a vicious dog, even though they might criticize him, and the law might punish him, for actually loosing the dog on a trespasser, especially if the trespasser enters the property by accident, unarmed, and without intent to commit harm. When nuclear weapons are designed to serve as instruments of deterrence, and when they are actually integrated into strategic planning in this form, rather than as war-fighting elements, we believe they can be said to serve a moral purpose—the preservation of peace. This is all the more the case when such weapons are maintained with the clear understanding that their actual use could well amount to an act of suicide. Under these circumstances, we do not agree that the threat to use nuclear weapons is no different morally from actual use. Just as the threat of lawful punishment, when used by a state authority to deter domestic crime, should not be equated morally with the commission of an act of violence in violation of the law, so the threat of retaliation in self-defense should not be considered as blameworthy as an act of aggression.
Nor should morality ignore practical considerations. On practical grounds, it can be argued plausibly that the existence of credible nuclear retaliatory capabilities, more than any other factor, has kept the two contending alliances from going to war with each other in times of crisis. Prior to the development of nuclear weapons, there was no such inhibition. Possession of less lethal means of destruction has sometimes only encouraged nations to make war, in the hope of achieving objectives that would outweigh the costs. In the nuclear era, leaders on both sides have acknowledged that a nuclear war would be an insane and futile undertaking, with no conceivable victor. Reagan has said repeatedly that a nuclear war cannot be won and should not be fought. Peace between the superpowers has indeed become, as Churchill prophesied it would, the sturdy child of terror.[23] Even in situations where they might otherwise have been tempted to make use of conventional forces, both superpowers have been inhibited by fear of escalation to nuclear conflict. In this important sense, nuclear weapons may be said to have served a moral purpose far better than previous declarations of peaceful intent or treaties like the Washington Naval Disarmament Agreement of 1922.
In practice, whatever their status in moral theory, nuclear weapons have been significant restraining influences in at least two well-known cases: the U.S. decision not to respond to the entry of China into the
Korean War by carrying the war above the Yalu River, and the Soviet decision to back down during the Cuban Missile Crisis. In the first case, President Truman decided that even though the war against China would be fought with conventional force against a nation that was not a nuclear power, it could provoke intervention by China's then ally, the Soviet Union, which was already a nuclear power, though one without an arsenal equivalent to that of the United States. Similarly, the Soviets backed down in the Cuban confrontation even though the first stages of the conflict that threatened to ensue would probably have been fought with conventional forces. The fear that any such confrontation would escalate to the nuclear level may well have been an inhibition. It is important to bear in mind, however, that the Soviets made the decision to withdraw knowing that they were not yet the equal of the United States in strategic nuclear power. In some future confrontation, having achieved the parity they subsequently set out to acquire, they would presumably feel less pressure to make concessions. Nevertheless, neither the Soviets nor the United States could have any incentive to engage even in a conventional conflict just because it could escalate to the use of nuclear weapons.
The lesson of these examples is that the advent of nuclear weapons and intercontinental ballistic missiles has compelled political leaders and generals to recognize that the Clausewitzian formula, whereby war is an extension of policy by other means, is obsolete. The behavioral restraint produced by this recognition has been more impressive than any verbal testimony. Military leaders who at first regarded nuclear weapons as simply very large conventional weapons are now sometimes among the first to decry the waste of resources in building nuclear "overkill." Many recognize that because nuclear weapons are not likely to be used in a war, it is useless to acquire more of them than are necessary for purposes of deterrence. As retired Adm. Noel Gayler (U.S. Navy, ret.), former commander-in-chief of U.S. Pacific forces, has noted: "There is no sensible use of any of our nuclear forces. Their only reasonable use is to deter our opponent from using his nuclear forces."[24] These military leaders would prefer to see resources channeled into conventional weapons more likely to be needed in an actual combat situation—combat, that is, not involving the other superpower, but third countries, and therefore less likely to escalate into nuclear war.
Deterrence relying on retaliation is significantly supported by arms control agreements. Even the agreements that now exist between the superpowers, which are by no means as thoroughgoing as advocates of
arms control would like them to be, serve a useful purpose in promoting both crisis stability and arms control stability. By preventing either side from deploying territorial missile defenses, they reinforce the credibility of the retaliatory threat and minimize the importance of marginal advantages that may be gained by offensive modernization. By permitting surveillance by "national technical means" (NTMs), they promote assurance that arms limitations are being observed, as well as provide information about weapons systems the other side has under development. NTMs also contribute to crisis stability by keeping the other side informed of what its adversary is actually doing during crises, thus preventing false rumors or worst-case analyses from precipitating decisions to act preemptively.
As is also now well understood, the success of this form of deterrence depends on a country's ability to maintain retaliatory forces in a state so that enough of them are adequately invulnerable to a surprise attack. The United States has chosen to emphasize diversity in its deployments and to put special value on submarine-launched missiles, because these are considered to be the least vulnerable. The Soviets have chosen to put somewhat less emphasis on diversity and to rely mostly on land-based ICBMs, presumably because they are the most accurate systems and the easiest to control and because Soviet competence in long-range aircraft and submarines has historically been less advanced. The Soviets clearly recognize the need to maintain invulnerability, especially in the face of the increasing accuracy of U.S. land- and submarine-based missiles, and are therefore shifting toward greater reliance on mobile missiles that will be harder to locate and target. In principle, there is no reason that both countries should not be able to continue to deploy, for the indefinite future, relatively invulnerable retaliatory forces—forces, that is, which cannot be so heavily degraded by a surprise attack as to be incapable of inflicting unacceptable damage in retaliation.
Proponents of strategic defenses often dispute this contention. As we note in chapter 1, they make several arguments. The Soviets, they contend, are engaging in a massive buildup of offensive forces that will soon outstrip those of the United States, despite the existing arms-control agreements (or indeed, as they would say, thanks to those agreements, which have had a more inhibiting effect on the United States than on the Soviets). The Soviets are also investing far more money than the United States in building and deploying strategic defenses. Some claim the Soviets are building toward a breakout from the ABM Treaty; others claim they have already broken out of the treaty and are in the process of erecting a nationwide territorial defense. Once this defense is fully in place,
the Soviets would be in a position to use their offensive superiority more effectively against the U.S. retaliatory force. Even as of now, the DOD estimates that the highly accurate Soviet ground-based ICBMs, notably the SS-18 launchers, could successfully destroy as much as 80 percent of the U.S. ground-based missile force.[25] Such an attack could also destroy a significant portion of the U.S. bomber force, which is not on active alert and is thus incapable of escaping attack, as well as perhaps 50 percent of the U.S. submarine force, which could be destroyed in port. If the attack also included a decapitating strike against the national command authorities, it is not clear that the United States would be in a position to organize, much less to order, a retaliatory attack, or that it would have much incentive to do so if Soviet defenses could absorb much of what was left of the U.S. retaliatory arsenal. This threat will only grow larger as the gap between U.S. and U.S.S.R. offenses widens and as and if the Soviets improve their ASW capabilities so as to be able to target and destroy the currently least vulnerable branch of the triad.
This scenario makes a great many convenient assumptions that turn out to be doubtful on closer examination. The possibility that a Soviet strike could eliminate 80 percent of the U.S. Minuteman arsenal depends wholly on theoretical calculations. The Soviets have never actually fired their missiles over the North Pole and do not therefore know with certainty how the gravitation fields of the polar region might affect ballistic trajectories. They would also have to have great confidence that all the missiles would work as expected. U.S. missile silos have been hardened to withstand all but very close hits. It is at least conceivable, moreover, that some of the U.S. missiles might be launched while under attack. Indeed, such a policy has often been advocated as a way of putting to rest concerns over the supposed window of vulnerability.[26] Another assumption is that the Soviets are making significant progress in the development of effective strategic defenses against air attack, an assumption we argued against in chapter 2. If the Soviet Union were to break out or even creep out of the ABM Treaty by deploying a nationwide territorial defense, the United States would have no reason to remain within the bounds of the SALT agreements. The United States could and almost certainly would respond to such a Soviet breakout by building up its offensive forces so as to saturate any Soviet defenses. If these are only ground-based defenses (which at present seems likely) modeled on the Moscow ABM, there is no reason to suppose they would be effective against a saturation attack, or that they would represent a cost-effective strategy on the part of the Soviets.
The notion that the Soviet Union is doggedly committed to an unrelenting
buildup of strategic forces ignores evidence that its leaders have changed course since they began to build up their forces after the Cuban Missile Crisis and, having achieved parity with the United States, are now more interested in improving their conventional war fighting capability. It also ignores the evidence of the Soviet offers to reduce offensive forces initially by as much as 50 percent, and even more radically in the future. Even if these can be regarded as propagandistic gestures, they do not support the view that the Soviets are determined to outstrip the United States in offensive forces; at most, they may be said to reflect an effort on the part of the Soviets to maintain a status quo in which they may suppose they have an edge in some respects important to their strategic doctrine. In general, the prospects for a successful surprise attack can hardly arouse great enthusiasm in the minds of Soviet political or military leaders. As McNamara has aptly observed, "only madmen would contemplate such a gamble. Whatever else they may be, the leaders of the Soviet Union are not madmen."[27]
For the foreseeable future, then, SDI is unnecessary as a means for reinforcing deterrence if the objective is to meet any conceivable new Soviet threat. Its only value is to provide a hedge against the possibility that Soviet research efforts will produce greatly improved techniques for BMD. So long as both sides have adequate and sufficiently invulnerable nuclear retaliatory forces, neither side needs to deploy BMD simply for the sake of maintaining deterrence. This consideration was precisely the one that served as the foundation of the ABM Treaty and also provided a reasonable basis to expect reductions in offensive forces and agreements to move away from reliance on weapons with first-strike potential to those useful primarily for retaliation.
A major commitment to SDI, especially one that involves early deployment, can only serve to undermine the status quo. Even if it does not eventually succeed in producing effective comprehensive defenses, it would create a climate of apprehension in which both superpowers would be reluctant to reduce their offensive arsenals. The willingness of the United States to contemplate reductions would be read by the Soviets to mean that deployments of defenses are being planned that would negate the value of the remaining Soviet offenses. Far from necessarily standing as an earnest of an intention to achieve a defense transition, this offer on the part of the United States (to the extent that it is coupled with a strong commitment to SDI) may well suggest to the Soviets an effort to upset the prevailing equilibrium in favor of U.S. superiority. The Soviets can have little incentive to agree to major reductions in offensive forces if they expect such an outcome. They may think the
likelihood of SDI succeeding is doubtful, and they may believe they can deal with any impending defensive threat by countermeasures, but all the same they would be foolish to commit themselves to reducing their offensive forces if such a commitment might remove the possibility of using superior numbers to defeat a defense—especially an "intermediate" defense that would rely heavily on terminal ABMs. Nor would they have an incentive to abide by the ABM Treaty restrictions on the deployment of ground-based systems and the testing of space-based weapons. Having called for the extension of the ban on space-based weapons of mass destruction to include all weapons, the U.S.S.R. would surely read U.S. insistence on pursuing the SDI (by testing in space and by making deployments that would patently violate the ABM Treaty) as an indication that it was no longer interested in maintaining parity with the Soviet Union but had become committed to a no-holds-barred arms race.
As things now stand, it is difficult enough for the United States and the Soviet Union to find common ground in the search for mutually acceptable measures of arms control. When missile defenses were eliminated from the equation, as seemed to have been done in 1972, at least the agenda for maintaining equilibrium was narrowed. Even this relatively narrow agenda has proved difficult to manage because of the asymmetry in deployments and the other pressures that prevent agreement. The pursuit of strategic defenses in ways that are certain to go beyond the restrictions of the ABM Treaty forces the issue of defense back onto the agenda in such a way as to crowd and complicate it. As a result, the status quo can be maintained only if the competition to develop strategic defenses can be contained within agreed boundaries; the entire issue could then be ignored.
Option 2.
The Case For A Limited Deployment
The case for a limited deployment of defensive systems to protect retaliatory forces and C3 centers rests on at least three considerations. The first is that improvements in offenses may make important elements of the retaliatory force more vulnerable unless they are protected by active defenses. The second is that active defenses may be needed to protect command-and-control authorities from a decapitating attack. The third argument is that limited defenses could provide some population protection against accidental or terrorist attacks. The first two are more compelling than the third.
One of the impulses behind both the High Frontier program and
Reagan's SDI was the claim that U.S. ground-emplaced missiles were vulnerable to a Soviet first strike. If so, an entire leg of the triad was at risk. As previously indicated, the same fear extended to the long-range bombers and submarines that might be caught at their bases in surprise attack. Only if a defensive layer could be added could the United States be sure of producing enough uncertainty in the mind of a potential attacker to make it desist from attempting such an attack. Such defenses, however, need not be comprehensive or space-based. Assured protection of a fraction of the land-based missile force can be achieved by preferential defense—defense concentrated on protecting a certain proportion of the assets the enemy would have to target. Preferential defenses force the adversary into an impossible guessing game and make it cost-ineffective to target everything with equal force. As a result, they virtually guarantee the survival of at least a fraction of the retaliatory force. A ground-based defense is probably more feasible now than it was in 1972, or in 1975, when the Grand Forks ABM system was deactivated, because radars can be made harder and mobile and can be supplemented with airborne detectors.[28] In addition, the nonnuclear interceptors now under testing could provide a more effective two-layer terminal defense with better homing ability than was possible in 1972. Because of these advances, even some critics of SDI support the construction of terminal defenses to protect missile silos.
But why protect these silos when the alternative exists of moving to mobile or deceptive basing? This alternative has been proposed for the MX missiles and for the smaller, single-warhead Midgetman mobile missile. Although the MX proposal proved politically unacceptable, the Midgetman is still under active consideration. Deployment of these missiles on military bases, either on trucks or railroad cars, would provide a cost-effective alternative to sole reliance on silo-based ICBMs. Barrage attacks against the bases would not be sufficient to assure the destruction of such mobile missiles. The Soviets are already moving in the same direction, which may account for the willingness they have expressed in arms-control talks to consider reducing their strength in stationary ICBMs. Strategic defenses could conceivably prove to be more cost-effective, but that has yet to be demonstrated, and their deployment could incur other costs that can be avoided by relying on mobile missiles.
The argument that terminal (or other) defenses are advisable to protect against attacks from smaller nuclear powers or terrorists or against accidental launches is weak, inasmuch as other delivery methods can readily be used by terrorists. Small countries and terrorist groups, "if
they did manage to acquire a nuclear weapon," as McNamara points out, "are far more likely to deliver it by the smuggled suitcase bomb than by the long-range ballistic missile."[29] The case for intercepting accidentally launched missiles is more cogent, except that a limited deployment might not have a large enough footprint to intercept more than a small fraction of accidentally launched missiles. In addition, it is unlikely that an ABM system would be designed with a "hair trigger" sensitive enough to make it react to a single missile launch, or even a small handful of such launches.
In general, then, the case for deployment of terminal defenses to protect missile silos and national-command authorities can be appealing under certain circumstances, and does not vitiate the case for relying primarily on deterrence by retaliation. On the contrary, terminal defenses could make that strategy more credible by further reducing the chances for destroying even one leg of the triad in a first strike. The Soviets have already deployed substantial ground-based air defenses, but since the ABM Treaty does not prohibit defenses against bombers, this is entirely legitimate, even though it creates an asymmetry between the United States and the U.S.S.R. in this respect. This asymmetry could, perhaps, be partially offset by limited deployment of terminal ABM defenses.
The problem with this approach, of course, is that a deployment beyond the allowable one hundred launchers would violate the ABM Treaty as it now stands. One way to seek a remedy would therefore be to renegotiate the Treaty so as to allow for a higher level of ABM deployment, at a minimum the two hundred launchers allowed under the treaty as it was originally conceived. (In the original version, half the launchers would be deployed to protect a missile field, the rest around the national capital to protect national-command authorities. They might all be in one place in a renegotiated arrangement.) The Soviets, who are so strongly committed to defense, might not object to such a renegotiation, especially given that they have production lines available for such missiles, which are already deployed around Moscow. For supporters of SDI, initial deployment of terminal defenses could be regarded as a downpayment on a fuller system. The net immediate effect, however, would be to shore up the prevailing strategy of deterrence by retaliation, and to do so within the terms of the ABM Treaty.
In more general terms, partial deployments of BMD could pose serious complications. As an OTA study points out, certain BMD systems can be conceived as operating in a "randomly subtractive" mode. Those
deployed in space or on the ground with very large footprints all fall in this category. These would be designed to shoot at as many enemy RVs as possible, without attempting to distinguish among them. In that case, BMD would not be limited to hard-point defense, and the utility of the defense would depend upon the efficiency of the kill ratio. A defender is likely to consider this option only if he can mount a defense in which the kill ratio per shot or per available quantity of interceptors and RVs is high enough to make defense cost-effective.
Even if the kill ratio were well below 100 percent, preferential defenses could assure the survival of at least some high-priority assets. A completely reliable preferential defense of unique or uncommon targets would be difficult, if not impossible, to achieve because the adversary could shoot enough RVs at such targets to exhaust the defenses. A "semi-preferential defense" is a more attractive possibility. Such a system would be based on a preexisting plan designed to protect at least a fraction of a specified set of targets. Some of these targets might be over-defended while others would be underdefended. A guessing game would inevitably ensue between the defender and the attacker. The attacker would allocate RVs to targets based on an estimate of their importance to the defender. If the attacker chose to devote enough RVs against the preferred targets to exhaust their defenses, it could defeat the intention. In practice, the need to allocate multiple warheads to all targets would make it likely that some fraction of the target set would survive. If the U.S. strategy were to be completely preferential, the U.S.S.R.'s best tactic would be to attack all the silos with the same number of RVs. If the defense could destroy a fraction of the attackers, it would preferentially destroy all the RVs directed at some portion of silos and save that portion. Thus, if the Soviet Union attacked the 1,000 undefended U.S. ICBM silos with 5,000 RVs, they could hope to destroy all of them. If the United States had BMD capable of destroying 1,250 RVs, 250 silos with 700 RVs could be saved from destruction. A semipreferential defense, because it concentrates defenses, could be more effective unless the Soviets greatly increase their volley of RVs or guessed correctly which silos were being protected preferentially.[30]
If the Soviets deploy equally effective limited preferential defenses, fewer U.S. than Soviet RVs would survive an attack because the current Soviet ICBM inventory of five thousand RVs is twice that of the United States. Defensive deployments alone, even if they are equal, do not guarantee equal results.[31]
On balance, the OTA study concludes that a limited defense designed
to protect missile silos and national-command authorities could be beneficial by contributing to the credibility of the retaliatory capacity, but that a higher level of defense could be destabilizing unless it promised virtually complete protection of all assets. The OTA study is pessimistic, however, that such a total defense could possibly be achieved: "Assured survival would probably be impossible to achieve if the Soviets were determined to deny it to us"[32] —and vice versa, without doubt. As we have previously noted, there can be no successful "last move" in the contest between offense and defense, except in the case in which one side simply abandons the contest.
Limited defenses designed to protect miliary assets would not impair crisis stability, and might enhance it by providing additional insurance that retaliatory forces could be used, if not as counterforce weapons, then as retaliatory weapons aimed at civil assets. If they are deployed on the basis of negotiated agreement, and especially if they can be deployed in tandem with reductions in offense, or at least with controlled levels of offense, they could contribute to stability. But the OTA study cautions that if the Soviet Union is able to use more warheads to attack U.S. missiles than the United States has warheads on those missiles, the net effect of symmetrical defense on both sides would be to reduce the total size of the potential U.S. retaliation. The study also cautions that a defense transition would take many years and could involve instabilities if either side perceived its defenses to be either more or less effective than the other's. Just how complex the situation would be during the transition is evident in the reasoning of the OTA analysts:
While we would not necessarily fare worse in a nuclear exchange under these circumstances than if there were no defenses—indeed we might fare considerably better—we might fare worse. The Soviets might be able to strike first and defend completely against our retaliatory strike. In one view, the expanded uncertainty in the minds of the Soviets regarding the outcome of a nuclear exchange would aid deterrence. In another view, the possibility that the Soviets could strike first and suffer no damage, a possibility that does not exist if the offenses dominate, would undermine deterrence. [Some believe that while the United States currently has the weapons to retaliate for a first strike, the Soviets may believe that the United States lacks the will to retaliate.] The knowledge by each side that the other might be able to strike and suffer no retaliation has important implications for stability.[33]
In estimating these strategic possibilities it is important to bear in mind that effective large-scale defenses capable of successfully handling foreseeable Soviet reactions are, at best, very far off in the future—so far, in fact, that it is idle to make predictions about such major factors
as the details of the technology of that time, or the U.S.-Soviet military balance, or even the general political configuration of the world as a whole. In such circumstances it is clearly impossible to make the kind of system-analytic calculation that would be necessary to determine whether or not the strategic relationship would be stable or unstable.
Option 3. Arms Reduction And A Moratorium On Deployment
Before SDI was announced, the United States and the Soviet Union had embarked on a process that was supposed to lead to negotiated arms reductions. In SALT I and II the process produced accords designed to take advantage of the agreement to rule out more than a minimal deployment of BMD. SALT I did put limits and sublimits on launchers but did not prevent the enlargement of warhead inventories through MIRVing. Nor did it or SALT II prevent modernization of existing weapons and the development of certain new ones. Nevertheless, these agreements were understood by both sides to be tentative steps along the road to more restrictive agreements. Such future agreements, however, would have to satisfy the basic military needs of the two sides, as each perceived them to be. Asymmetries in force structure inevitably led to suspicions on both sides that the existing treaties left the other with advantages. The Reagan administration came into office convinced that SALT II was "fatally flawed" because, among other things, it left the Soviets with an advantage in heavy missiles capable of prompt hard-target kill. This advantage was said to have created a "window of vulnerability" by putting the entire land-based leg of the U.S. triad at risk in a preemptive attack. The Reagan administration reluctantly agreed to abide by the treaty limits informally so long as the Soviets would, too, and to discuss further arms-control agreements with the Soviets. But there was no great enthusiasm for these negotiations in the administration, and as a result a coherent policy was not adopted during Reagan's first term.[34]
In the second term, however, after the president declared SDI, the negotiations picked up steam for reasons to be examined in chapter 5. Reagan appeared to have undergone a change of heart, and U.S. policymakers may also have been eager to counter the propaganda gains that the Soviets were threatening to achieve from Gorbachev's peace offensive. The Reagan administration was hardly of one mind, however. There were those high in its councils who clearly would have preferred
that there be no arms-control agreements with the Soviets. They did not expect the Soviets to abide by any such agreements, and they feared that such agreements would freeze a situation in which the Soviets were in the stronger position. The U.S. edge in technology would not be exploited, and the peace forces in Congress and elsewhere would use an agreement to clamp a lid on defense spending. The net result would be to weaken rather than strengthen U.S. defenses. The cross pressures within the administration were scarcely resolved. (A good illustration is the administration's decision to call for a ban on mobile missiles, which was in direct contradiction to the recommendation of the Scowcroft report, and which the administration may not have intended to honor in negotiations.)
The signing and ratification of the INF Treaty left the larger question of strategic offenses and defenses unresolved. If the Bush administration elects to continue to disregard the SALT II limits and to call for early deployment of space-based strategic defenses, any momentum afforded by the INF Treaty could well be lost. It is possible, however, that the INF agreement will "break the ice" and lead to further negotiations, all the more because this would be an agreement signed by a president who came into office convinced that the U.S.S.R. was "an evil empire" given to lying and cheating on all its agreements, and who seemed more interested in challenging the Soviets to a renewed arms race than in controlling it. The president's defenders have argued, however, that Reagan was concerned to get serious arms reduction, and that his approach of making it clear to the Soviets that he would not settle for just any agreement paid off in the case of the zero option and will eventually also pay off in the other areas. If this interpretation proves correct, an INF accord should lead to strategic arms reduction. How it can do so in the absence of an agreement to restrict SDI to what is permitted under the ABM Treaty is hardly obvious.
In domestic political terms, even the achievement of a broad framework agreement (modeled on that adopted by Ford and Leonid Brezhnev in Vladivostok in 1974) on offensive arms and SDI would have important consequences in forging a bipartisan consensus and assuring the continuation of the SDI program in some form. That program could otherwise be in serious difficulty in Congress and with the defense establishment. But the larger question that the Bush administration must resolve is how an arms-control agreement can be compatible with the desire for strategic modernization and SDI. In other words, the issue is what strategic commitment the administration intends to make—is
there any coherent strategic commitment, or is there merely a ragged patchwork of ad hoc agreements and concerns?
That there could be a coherent strategy is indicated by Nitze, who has argued that the U.S.S.R. had become complacent about negotiations with the West because it enjoyed a certain strategic superiority. In his view, it now enjoys a strategic superiority in conventional forces and a geographic advantage in Europe. The U.S.S.R. has a particular advantage in prompt hard-target kill capability because of its large land-based ICBMs. In addition, "they were quite satisfied with the pre-SDI imbalance in strategic defense activities between the U.S. and the U.S.S.R. They see little advantage in moving cooperatively to a more defense-reliant regime under which their current advantages in both offense and defense would be reduced or balanced."[35] Nitze therefore argues that SDI played a significant role in compelling the Soviets to think more seriously about negotiations with the United States. They are "genuinely nervous," he suggests, about a concerted effort by the West to develop exotic technologies and apply them to defenses. They may well fear that "the marriage of Western technological genius and American space expertise can lead to U.S. dominance in the military uses of space."[36] Nitze doubts the Soviet Union seriously fears that the United States has a hidden agenda in SDI that aims to develop offensive space weapons because, he says, they know that an SDI system would be optimized for defensive purposes, whereas the strategic modernization program aims to develop new systems to attack ground targets. SDI gives the Soviets an incentive to agree to or at least to recognize the need to allay Western discontent with Soviet advantages. Having traded away their SS-20s, the Soviets now appear prepared for serious negotiations aimed at reducing conventional forces in Europe, but these moves would not affect the competition with the United States in strategic forces. Something more would be required on both sides.
Nitze argues that both sides would benefit if they could agree to the concept of a defense transition to be managed cooperatively. He admits that the Soviet negotiators have expressed little interest in the concept, but he sees this reluctance as largely tactical, because the Soviets have committed themselves to the public argument that SDI is designed to produce "space strike arms" and to militarize space, and they are reluctant to acknowledge that what is really at issue is the extent to which it may be possible to rely on a more defense-dominated strategy—one that would permit the United States to equalize what Nitze sees as the current Soviet advantage in combined offensive and defensive forces.
But Nitze admits that the Soviets may be skeptical that the United States would deliberately "introduce future strategic defenses in such a way that neither side would gain unilateral advantage."[37] To remove these suspicions, U.S. negotiators would need to engage with their Soviet counterparts in serious, detailed discussions about how such deployments might be made.
The Reagan administration balked at entering into such discussions, because the Nitze view was not shared by others influential in the administration. The discussion within the administration—originally led by Weinberger—of an early deployment of space-based defenses, ran counter to Nitze's idea of a cooperative transition. It reflected a willingness to proceed unilaterally to achieve a transition, on the theory that in the short run a deployment of defenses would reduce or eliminate the advantage the Soviets have in land-based missiles. Eventually, the thinking went, such a deployment would create so much uncertainty in the minds of the Soviets as to negate the value of such weapons and lead, either through negotiations or otherwise, to the elimination of such weapons as a concern. The Weinberger approach, in other words, seems to deny any need for Soviet cooperation in a defense transition. It seems to be based instead on the assumption that an affordable U.S. program can be pursued that would install intermediate defenses—based on relatively conventional technology—in the short run, and replace or supplement this system with more exotic technologies as they become available. In the process, the United States would at first maintain its efforts at strategic modernization, so as not to rely entirely on the space shield, but would gradually increase the role of defenses in comparison to that of offensive forces; the Soviets would simply be compelled to follow suit because they would not be able to rely on their offenses.
It is not clear whether future administrations will feel a need to choose between these two approaches, but it is clear that they are very different. Nitze's approach does not anticipate early deployment of space-based systems unless and until it can be shown that they meet the stiff criteria he set, and Nitze himself has said that as of now the systems being proposed do not meet those criteria. Weinberger, in effect, introduced new criteria—of affordability and, even more important, of demonstrating to the Soviets and to the American people that the United States intends not just to do research on strategic defenses but also to make use of that research to deploy actual systems. Such deployments, if they are even partially effective, would force the U.S.S.R. to make expenditures it would prefer to avoid to modernize its offensive forces, and to impose
penalties on those forces, which would make them far less threatening to U.S. military assets and perhaps also to the U.S. population. If in the process the existing arms control agreements were scrapped, this would presumably be all to the good, in that it would remove the illusion that the security of the West can be based on such agreements rather than on the will to resist Soviet expansionism by maintaining the strongest possible defensive capabilities.
So long as the U.S. administration is of two minds on the issue, U.S. negotiators can present no coherent proposal for a defense transition to the Soviet Union. Even if they could, there is no indication that the Soviets would be willing to go along with it. They might have good reasons for rejecting any such proposal. They may calculate that they benefit from their present advantages in strategic defenses and may be unwilling to give up that advantage until the United States actually makes a commitment to deploy space-based defenses. Just as they resisted proposals to consider the zero-zero option until the West began to deploy its responses, the Soviets may elect to wait until the United States makes a decision to deploy a space-based defense. They are well aware that the proposal has strong domestic opposition and would be costly. They also have the alternative of concentrating on countermeasures to defeat the system, on the theory that countermeasures might be cheaper and more reliable than any kinetic space-based system the United States could orbit in the near future. They may well also believe that one of the best countermeasures to such a system would be not only to maintain their current missiles and warheads but also to build still more missiles, more modern than present missiles (especially those with faster-burning boosters).
At Rejkyavík, however, the Soviets gave evidence of a willingness to agree to reductions in offensive missiles, along the lines previously agreed to at Geneva, so long as the United States would agree to abide by the ABM Treaty for at least ten more years. But that extension would have included a commitment not only to refrain from deploying strategic defenses not permitted under the treaty but also to restrict research so that there would be no testing in space. If the United States is willing to wait until SDI research produces results that would meet the Nitze criteria—results that would certainly take at least ten years, in the opinion of such qualified technical observers as the APS review panel—then an agreement to reduce nuclear weapons is within reach. How significant that reduction would be is another question. A 50 percent reduction would still leave both sides with more than enough accurate offensive capability
to inflict devastating damage. But if the sublimits can be negotiated to the satisfaction of the United States, then the Reagan administration's concern over the threat of the Soviet SS-18s could be considerably alleviated.
An alternative to an agreement to reduce nuclear weapons, coupled with a moratorium on development and deployment—the defense-protected build-down—has been proposed by Weinberg and Barkenbus.[38] This proposal would couple reductions in offense to deployment of defenses. Initially, these defenses would be ground-based. The reductions might either be arranged by negotiations or by tacit bargaining. Each time one side substituted a unit of defense for a unit of offense, the other side would reciprocate, until virtually all offenses were replaced by defenses.
There are a number of difficulties with this proposal.[39] One is that it does not fit the current national strategy of extended deterrence. In that strategy, there must be enough offensive weapons to make an attack possible on the enemy's military forces, perhaps even in a first-strike mode. The number of such weapons is much larger than the number needed for threatening cities in a "minimum deterrence," or purely retaliatory, mode. It is therefore not possible simply to trade off offensive weapons for the kind of defenses that would guarantee the survival of a few ICBMs. In addition, defenses cannot be tested as effectively as offenses and must meet more stringent criteria. An offensive weapon can cause great damage even if it explodes at some distance from its intended target. A defensive weapon must be able to intercept a target with a high degree of accuracy in a very short time. Whether military officers would be willing to substitute defensive for offensive weapons is questionable, in view of such inherent disadvantages to the defensive systems. Parity would be hard to assure in view of the different characteristics and effectiveness of the various components that would constitute the defensive systems. This would become even more of a problem when the components in question would be not only the conventional nuclear-tipped BMD missiles but exotic technologies, which might be either ground-based or space-based and would depend on a variety of sensors. Would a space-based sensor be considered the equivalent of a ground-based radar? Would all types of defensive missiles or laser weapons be considered interchangeable units for purposes of achieving parity? Could such deployments possibly be made subject to adequate verification? In principle, the idea of achieving a defense transition by means of a defense-protected build-down is attractive, but in practice it would
be very hard to implement. As the OTA study noted: "Nobody has suggested how the problems of measuring, comparing, and monitoring disparate strategic forces—problems which have plagued past arms control negotiations—could be satisfactorily resolved in the far more difficult situation where both offensive and defensive forces must be included."[40]
That difficulty could be postponed and perhaps avoided altogether if both sides could agree to reaffirm the ABM Treaty, under its traditional interpretation, while reducing nuclear weapons by half. Such an agreement need not be thought a dead end. Many experts agree that further reductions could be envisioned that would reduce the stockpiles to a level no higher than several thousand warheads each without affecting each side's ability to continue to deter the other by the threat of retaliation. If these warheads were all on systems that would not threaten a first strike, crisis stability would be enhanced. Neither side would be inhibited from continuing research on more effective defensive systems or from modernizing offensive weapons to enable them to penetrate existing defenses or defenses that might be contemplated. But a climate of cooperation would have been established that could lead to cooperation in the deployment of defenses. Freeman Dyson has argued cogently that the effort to develop defenses will prove impossible if both sides are free to build countermeasures and offensive forces to defeat them. It is not impossible, however, as he has also noted, that if the offensive forces can be reduced significantly by mutual agreement, relatively inexpensive defenses could be introduced that would be effective enough against small stockpiles to serve as insurance against third-party attacks and enable virtual nuclear disarmament to take place.[41]
"General and complete" nuclear disarmament is probably too much to hope for. Nuclear weapons cannot be uninvented. Small stockpiles could be hidden, although sophisticated delivery systems might be harder to hide. But the inspection required to make certain that no weapons at all were secreted somewhere would be so intrusive as to be virtually intolerable to sovereign states. In a time of crisis, moreover, any state capable of producing such weapons could do so on fairly short notice if the facilities and materials had not been destroyed. And there would always be the paradox that in a world in which nuclear weapons have been reduced to very low levels, a state possessing even a small number could become a superpower by virtue of those few weapons.
The big advantage in accepting such an agreement would be, first of all, that the present system of deterrence would not be rendered uncertain, as it would be by a decision to develop and deploy a substantial
defense of uncertain effectiveness. Both sides could continue to count on the efficacy of their deterrent forces and would not need to restructure those forces or reorient their production facilities immediately to take account of impending defensive systems. Such modifications to offenses could be introduced gradually. This would be of obvious advantage to the Soviets, but that advantage would translate into greater assurance for U.S. allies that the Soviet Union would not seek to compensate for inferiority vis-à-vis the United States by raising the threat to them. The political gains of accommodation could be enormous in maintaining or restoring the momentum of the process of arms control and possibly promoting the broader form of détente that eluded Nixon and Kissinger in the 1970s, but that the present Soviet regime shows signs of being interested in. Nor would such an agreement stymie SDI. On the contrary, the president or his successor could properly claim that SDI had helped to bring about the agreement, and that only its ultimate success would make possible further reductions that would lead to the substitution of defenses for nuclear weapons. This would be a potent formula to assure support of a regular, if modest, level of expenditures prior to the decision to move to development and deployment—a decision that could be affected for the better by Soviet agreement to modify the terms of the ABM Treaty to allow for specific forms of testing and deployment.
Option 4. Unrestricted Competition
The most troubling alternative to some form of agreed transition or to a postponement of any final decision about such a transition is unrestricted competition in both offensive and defensive weapons. The consequences, as McNamara points out, would be very serious:
A unilateral shift to Star Wars would have disastrous consequences. The ABM Treaty, the most important arms control agreement of the nuclear age, would be the first casualty. Because of the resulting race in strategic defense weaponry, each side would begin to build up its offenses to ensure deterrent capability. Thus, the existing SALT agreements would be the next casualty. Future accords to restrain the offensive arms competition would be impossible until the Star Wars race was halted and reversed.[42]
In strategic terms, any decision by either side to proceed unilaterally to violate the ABM Treaty (not in some specific way—as the Soviets did in the case of the Krasnoyarsk radar—to take advantage of a loophole without affecting the general aim of the treaty, but in a way that would undermine the treaty in an essential way by producing a territorial defense
or even a major defense of missile silos) would certainly be read by the other side as a commitment to achieve military superiority, either to prepare for a preemptive strike or simply to be in a position to harvest political benefits. Such an effort would be met by a concerted effort on the part of the other side to match the buildup or, possibly, to defeat the defenses. During such a competition, both sides would have reason to fear that their retaliatory deterrents could be put in jeopardy. Under the circumstances, there would be both crisis instability and arms race instability.
If both sides were to deploy space-based defenses using speed-of-light weapons, the temptation to strike first, in a counter-defensive mode, would be greater than it is in a condition of offensive mutual deterrence. Under such circumstances the stage would be set for what a Senate staff report called "a 'high noon' shoot-out in space."[43] Each side would have to recognize that the first to shoot would win, unless—as is highly unlikely—weapons and sensors can be made undetectable or invulnerable. Directed-energy weapons will very likely be even more effective against satellites, including those that would carry strategic defense sensors and weapons, than they would be against ballistic missiles. Although such a conflict would not itself involve the use of nuclear weapons, it could easily lead to their first use by the side whose defenses were under attack, if that side should become convinced that it was being put into a situation in which it must either "use them or lose them."
Those in favor of SDI nevertheless argue that whatever response the Soviet Union might make, the United States would have an advantage in military technology and could exploit this advantage to develop defenses that would be so cost-effective that the Soviets would have no choice but to recognize that their offenses had indeed become "impotent and obsolete." Meanwhile, of course, the United States would continue to rely on its offensive deterrent to forestall any Soviet preemptive attack. But this argument falters on the issue of cost. It is easy enough to postulate that the United States can develop defenses while maintaining a potent retaliatory force, but given that an effective defense would have to include a layered screen capable not only of intercepting virtually all ballistic missiles fired at the United States and its allies—whether from long-range or short-range weapons—but also as air-launched cruise missiles, bombs, and standoff weapons, such a defense is bound to be extremely costly; vast sums must meanwhile be spent on keeping offensive forces up to date. The possibility that the Soviets would shift to a policy of emphasizing conventional war, perhaps conducted by proxies,
against vital U.S. interests, or that other threats would develop (like that posed by Khomeini's Iran in the Persian Gulf) would force the United States to spend possibly even more than it now spends on conventional military forces. That a defense transition would be feasible under these circumstances altogether ignores the pressures for trade-offs among military options as well as the pressure exerted during peacetime on the domestic budget. A strategy that is not affordable and is not likely to be sustained politically is one that cannot be maintained in either strategic or economic terms.
A strategy must also take into account the psychological and other effects on the structure of alliance in which the United States plays a leading role. To the extent that SDI has already aroused misgivings among U.S. allies, because it seems to endanger the process of arms control and détente, a decision to proceed beyond the present research stage into development and deployment would risk gravely exacerbating these fears, quite possibly arousing such political opposition as to bring about the collapse of the alliance. The alliance has been held together for four decades on the assumption that the nuclear shield would serve as a deterrent against Soviet aggression. The abandonment of that nuclear shield in favor of a defensive system in which the allies have no confidence, and which might trigger Soviet counterefforts against them could strengthen the current trend to regard the Soviets as less of a threat to Europe than the United States. A strategic revolution of the kind implied by the idea of a defense transition can be promoted unilaterally only at great peril to the alliance. This is especially the case inasmuch as West Europeans are likely to fear that if both superpowers achieve even reasonably effective strategic defense not only would the independent deterrents of France and Britain be nullified but also the way would be open for the U.S.S.R. to exploit what is perceived to be its advantage in conventional force in Europe.
The issue of strategic defense also runs into the problem posed by the possible resort to conventional force. Although both sides have maintained considerable conventional forces and both have brought these forces into play in military engagements—the United States notably in Korea and Vietnam, the Soviet Union in Czechoslovakia, Hungary, and Afghanistan—they have been inhibited in the use of conventional force in situations that could lead to direct confrontation and escalation to the use of nuclear weapons. In that sense, nuclear weapons have had an inhibiting effect on the actual or threatened use of conventional force. To the extent that that option or danger is removed, would the world be
made safe once again for conventional war? The U.S. guarantee of extended deterrence for its allies would be rendered meaningless, because it rests not on its ability to help Europe to deter a Soviet attack by providing conventional support but on its "nuclear umbrella."
Even now, Western Europeans doubt whether they can rely on the U.S. nuclear guarantee. They wonder whether, despite formal commitments, a U.S. president would risk the safety of the United States by ordering a nuclear strike against the Soviet Union to defend Europe. Those who express this doubt often argue that NATO should be better prepared to check any Soviet aggression by conventional means in order to minimize the need to rely on the threat of U.S. nuclear forces. For the same reason, two European states, Great Britain and France, have felt compelled to develop independent nuclear deterrents. If these were nullified, the conventional threat would indeed become more serious.
If the conventional threat became more serious, however, the West Europeans could be expected to take the case for conventional armament more seriously. Until now, they could avoid taking the hard decision to increase spending and to devote more manpower to the defense of Europe by relying on the U.S. nuclear guarantee. If they could no longer rely on such a guarantee, they might simply have to respond by building up conventional defenses. This will become more and more difficult as the Soviets acquire more accurate conventional weapons with which to strike at NATO airbases and command facilities, but it is certainly not impossible for the West to build up its conventional forces to an extent that could deter the Soviets or to negotiate asymmetrical reductions.
Outside Europe, the U.S. nuclear shield is even more dubious. The United States may formally pledge to defend the vital interests of the West by any military means necessary, but the unwillingness to rely on nuclear weapons led to the Carter administration's creation of the Rapid Deployment Force, a move reinforced by the Reagan administration's decision to create a new command for mobile and swift intervention. Although the defense of Japan and South Korea also involves the nuclear option, here, too, conventional force deployment is seen as a vital complement to a credible defense. It is not obvious that the elimination of the nuclear umbrella would necessarily place these allies at greater risk, or that the risk could not be kept minimal by the deployment of greater conventional force.
In Europe and other theaters, significant expansion of conventional forces would entail costs, but these could be shared with allies who can
now well afford to contribute more to their own defense. The costs are therefore not a significant deterrent to a shift to greater reliance on conventional forces. Nevertheless, it would remain possible that the deterrent effect of conventional forces would be far smaller than that created by nuclear weapons. There have been major wars not involving the superpowers—in the Middle East, for example—in which the sides were not inhibited from going to war because they did not fear nuclear attack. It is at least conceivable that in Europe or the Far East, either side might be prepared to enter a conventional conflict to achieve some limited objective. War, in other words, might become more "thinkable" among the superpowers than it has been since both acquired the means to deliver great numbers of nuclear weapons.
For this reason, the abandonment of the strategy of deterrence by the threat of retaliation in favor of the development of assured defense—always assuming such a defense is possible—could indeed increase the danger of conventional war. This danger is compounded, as previously noted, if the atmosphere is characterized by mutual mistrust and there are no constraints on the development of military forces of all types. The effort to develop defenses against conventional attack may succeed in allaying some of the difficulty, but there is little reason to suppose that such defenses could be good enough to forestall conventional wars altogether, under favorable circumstances.
Is There No Better Alternative?
If defenses against nuclear attack are not built, is there no way to escape the danger that nuclear weapons might be used in the event deterrence fails? There are ways (through arms-control measures of various sorts, including confidence-building measures) to lessen the likelihood of a failure of deterrence. It is also possible to remove by agreement weapons that are especially valuable in a first strike or those that are so vulnerable to preemption that they are usable only in a first strike, and instead to leave both sides with weapons more suitable for purposes of retaliation. But the existence of large arsenals of nuclear weapons would mean that the danger of atomic war would persist. The danger of the effects of such a war can be reduced by reducing the size and lethality of the stockpiles. If both sides come to see nuclear weapons as deterrents rather than as war-fighting weapons, they should be willing to reduce them to minimally acceptable levels by agreements, particularly if lesser powers are willing to reduce their nuclear stockpiles as well. Such agreements
could seriously inhibit further proliferation of nuclear weapons and lead to more agreements patterned after Latin America's agreement to achieve further regional bans on the deployment of nuclear weapons.
So long as the weapons exist, so long as there are no effective defenses against them, and so long as war remains a "normal" way of resolving conflict, the danger would remain that they might be used. To minimize this danger further, some forms of defense—both passive and active—may be desirable and could be put in place by both sides without threatening the other or the credibility of its deterrent. As such steps are taken (quieting fear and antagonism), institutional changes might be made in the form of recognized boundaries and spheres of influence such as have already been achieved in Europe. These institutional changes could be coupled with measures to increase trade and communication, in line with recent developments. In a climate of cooperation, political changes could strengthen integration and make international agreements more enforceable.
In view of the potential for stability, it seems an act of folly to commit the United States to a unilateral pursuit of strategic defense on the ground that this will somehow lead to a cooperative defense transition, whether the other side agrees or not. The decision to rely on military force alone, and not on political bargaining and negotiation, is defensible only if it would necessarily redound to national security. And, as we have seen, such an outcome is hardly inevitable. The vigorous pursuit of an illusory perfect defense hardly promises to offset the risks and costs that would certainly be involved.
A better approach would be to maintain the present strategy of deterrence while working to reduce the size of the nuclear arsenals presently considered necessary for this purpose by both sides and, in addition, to attempt to substitute conventional weapons for nuclear ones. The arsenals can be greatly reduced without jeopardizing security. In view of improvements in accuracy, U.S. ICBMs, SLBMs, and air-borne standoff weapons can destroy Soviet targets with less redundancy. Shorter-range military missions can be accomplished by using recently developed, highly accurate short-range missiles with conventional warheads.
Neither the United States nor the Soviet Union would risk any loss of international power or prestige by agreeing to such a mutual reduction in nuclear forces. Their arsenals would remain significantly larger than those of other states, and their agreement to reduce arsenals would be a powerful incentive for them to prevent further nuclear proliferation and
the buildup of nuclear arsenals by states that already belong to the nuclear club but are still dependent on the superpowers for military and economic assistance. Superpower cooperation would provide an incentive for the settlement of regional conflicts, especially those that may set the superpowers against each other.
An agreement to restrict the military use of space to present practices would have other significant strategic benefits. Both sides could be confident that the satellites they rely on for early warning and treaty verification would not be readily attacked. Although there could be no perfect certainty (ground-based lasers, for example, could be used in a surprise attack to blind satellites), unilateral measures could be taken by each side to protect satellites and to harden them against the possibility of breaches. Civil cooperation could be undertaken without compromising military objectives. Technology transfer would be less restricted under this approach. Otherwise, the rules would have to take account of potential benefits for space warfare. Another major benefit would be that both sides would not have to make hard trade-offs between space weapons and other forms of military modernization. The military services on both sides would probably welcome a decision to avoid space warfare (with the exception of those with a vested interest in making war in space). It would, in addition, relax pressure on military budgets and thus make it easier to gain support for modernization efforts in other areas.
Nor should the positive opportunities be ignored for transforming the cold war. To some extent, at least, the persistence of hostilities between the superpowers is a function of mutual mistrust and fear. To the extent that the superpowers can maintain the arms-control process and achieve greater strategic stability, the result will be worth more than many a new weapons system. Reliance on weapons systems alone does nothing to alleviate mistrust and uncertainty, however much it induces restraint on the part of potential aggressors. So long as fundamental issues regarding values and interests divide East and West and the international system remains anarchic, it would be too much to expect that cooperation in arms control will by itself change the cold war into a warm peace. But arms-control measures can contribute greatly to the prospects of détente and to cooperation in other respects by eliminating some of the grounds for suspicion. The introduction of techniques for verifying nuclear explosions and adherence to treaties by on-site inspection and monitoring will also help to diminish tensions. Although
hard experience has taught us that linkage between arms control and other forms of détente is by no means automatic, extensive superpower collaboration on arms control could lead to cooperation in managing other conflicts. We therefore turn next to the implications of SDI for arms control.