1 Deterrence: An Overview
How should we think about deterrence in the new strategic environment? With the demise of the "Soviet threat," we have begun to focus much of our attention on deterring aggression by countries using conventional means. And yet, nuclear weapons and nuclear deterrence still loom large, because of the weapons' vast and instantaneous destructive power. In addition, the slow but apparently inexorable spread of those and other weapons of mass destruction raises issues that cannot be ignored. The different weaponsnuclear, chemical, biologicaland the actions that countries and sub- or transnational groups may take to obtain and possibly to use them, despite treaty constraints, must be separated and treated differently from each other because the time scales and physical attributes of the weapons' actions are different.1 How to deter the use of any of them, when we2 are unable, on moral and treaty grounds, to threaten to reply in kind in any but the nuclear area, and when even that response may be deemed morally insupportable in many situations, is a matter of deep concern. In addition, under the obligations taken by the nuclear weapons states, including the United States, under the nuclear Non-Proliferation Treaty (NPT) as extended for an indefinite period one year ago, the United States is obligated to decrease its nuclear arsenals and to decrease the leverage that nuclear weapons exert in the international arena. It is through this decrease in leverage that the
1 Passive defense against chemical and biological weapons, in the form of protective clothing, antidotes, and vaccines, is easier than passive defense against nuclear weapons, although we have given far less attention to such protection against biological weapons than against chemical weapons. In an unprotected environment, chemical weapons may tend to have more localizedalthough still deadlyeffects, and they are far less lethal per pound of agent delivered than are the other weapons. Biological weapons may take time to make their effects felt, and their source may be more difficult to identify if delivery is clandestine, but they may carry a greater risk of backfiring against the user. Biological weapons will remain highly dangerous to civilian populations unless passive protections can be much enhanced by intelligence, detection and evaluation methods and installations, and widespread availability of vaccines and antidotes for all known agents.
2 Throughout, the editorial "we" is used to refer to the U.S. policy makers and decision makers who must devise and decide on deterrence actions in any particular case.
discriminatory aspects of the nuclear Non-Proliferation Treaty are to be erased over time. This ultimate result remains highly uncertain under current world conditions.
These developments associated with weapons of mass destruction, together with the issues involved in deterring the use of conventional force for purposes inimical to our interests, further complicate the already intricate environment into which we would extend the concept of deterrence that served so well during the Cold War period. At the same time, changing attitudes in the United States, deriving from the subsidence of the most extreme dangers to U.S. security and that of our long-term allies, have altered the way we can respond to threats against our interests and against other nations whose security is related to ours. The tools available to us to respond have also changed, with the advent of conventional weaponry of unprecedented precision and power. All of these factors require new answers to the question of how to think about deterrence in today's world.
It is not easy to disengage from thinking about matters affecting the potential for peace, war, and survival that evolved for half a century. The discussions and the collection of papers (see the appendixes) that form the foundation of this overview of the subject represent the best thinking about the problem that could be generated at this stage of development of the new world outlook. Some of the points of view presented will solidify in much the form presented; others are not yet ready to do so. Important differences remain within the special group of study participants and in the nation; they are highlighted in this chapter. They occur especially in areas having to do with:
• The potential uses and value of nuclear weapons in deterring attacks on close U.S. allies and vital U.S. interests by states using powerful conventional forces or chemical and biological weapons;
• Active defense against ballistic missiles attacking the United States; and
• Assessments of the extent to which less-than-vital U.S. interests might justify the use of military force, with attendant casualties, in the eyes of the American public.
To answer the lead question, this chapter first examines what deterrence means in the new environment. The group of study participants found that some principles endure; these are reviewed in the new context. The group next examined ways to extend these principles into the post-Cold War world situation. Finally, this chapter examines some approaches to analyzing deterrence policies that can shed light on how such policies may function in the new environment and that can help the United States and the U.S. naval forces prepare for action in future situations.
THE MEANING OF DETERRENCE
The concept of deterrence applied to international affairs is generally well understood. However, it can become extremely complex in application.
In its simplest form, to deter means to inhibit or prevent someone from doing something. Military force used in some form and to some degree underpins all types of deterrence. In the context of an overall policy, however, military force is likely to be only one tool among many diplomatic, economic, political, and military responses or anticipatory actions designed to guide development of an international interaction in directions that will prevent an outcome inimical to our interests. If deterrence of an undesirable action on the international scene fails, we may use military force for "compellance," and if we judge that the threat of force may trigger a preemptive attack we may use forms of "reassurance,"3beyond simple declarations to convince an adversary that an attack on our part is not planned. Each of these concepts, and the other-thanmilitary tools associated with prevention of hostile acts or those threatening our interests, has its own qualities and modes of operation. None is independent of any of the others in actual policy implementation. In all of them, the issue is to devise means to affect other nations' behavior, and to recognize that our own behavior will be affected by their responses. To avoid further complicating an already extremely complex set of concepts, in this chapter the term "deterrence" is applied to the entire field of activities that may be involved in averting actions on the international scene that can harm U.S. interests and those of our close allies.4
Deterrence represents an early stage in a broad protection of our interests, analogous to a military defense in depth. The protection starts with avoiding war by many means, and ends with fighting it in both defensive and offensive phases. During the Cold War, it was an attractive approach for the United States when it was not feasible to defend everywhere against all conceivable threats. For example, defending the small, surrounded Western garrison in Berlin was impractical, but an attack could be deterred by the threat of certain NATO responses elsewhere. Also during this period, deterrence became closely associated with nuclear weapons and their unique ability to threaten immediate and massive destruction. In the NATO context, the policy of deterrence was elaborated over the years to include a number of steps ("graduated response," "flexible response," "selective release") involving a sequence of graded
3 The term "reassurance" was initially proposed by Sir Michael Howard to refer to the climate of reassurance that U.S. participation in European and global security arrangements during the Cold War conveyed to our allies (see Sir Michael Howard, "Lessons of the Cold War," Survival, Vol. 36, No. 4, Winter 1994-95, pp. 161-166). It has been extended by John D. Steinbruner of the Brookings Institution to "reassurance" that violence in settling their affairs is not planned among nations (see Appendix A of this report).
4 Richard Garwin notes, "I believe that it is desirable to retain the concepts of 'reassurance,' 'compellance,' 'defense,' and 'deterrence'; then there ought to be a pure 'deterrence' as a subcategory itself."
responses with conventional and tactical nuclear forces before an all-out nuclear response would be invoked. This approach provided an opportunity to convey the will to respond to an attack in a manner suited to the provocation, without immediately escalating to a full nuclear exchange.
During this period, deterrence was commonly thought about in terms of convincing our major opponents that a particular action would elicit a response resulting in unacceptable damage that would outweigh any likely benefit. These concepts of deterrence were extended to other potential flashpoints, such as Korea. We sought to threaten "unacceptable damage" in an attempt to make the Vietnamese Communists desist from pursuing their war in South Vietnam. It was clear from the outcome that the United States generally underestimated the level of damage, and perhaps misunderstood the kind of damage, that the North Vietnamese would consider "unacceptable"a result that has relevance to that concept today.
Rather than a simple cost/benefits calculation, deterrence is more usefully thought of in terms of a dynamic process with provisions for continuous feedback. The process initially involves determining who shall attempt to deter whom from doing what, and by what means. The who may involve building a robust multinational "coalition of the willing" with a wide range of capabilities and overwhelming resources with which to threaten an adversary. This combination of will and resources must be credible not only in terms of an ability to inflict unacceptable damage and/or to deny success, but also in terms of the willingness to pay the necessary costs and bear the pain that may be associated with executing threats. Positive inducements also should be considered as part of this range of capabilities. Of course these capabilities must be communicated in a clear and authoritative manner to the parties to be deterred, including demonstration of the capabilities where warranted by the occasion.
The target of deterrencethe whom-needs to be explicitly defined so that the necessary analyses can be undertaken to understand the adversary's objectives, the importance he attaches to the action that is to be deterred, key vulnerabilities, propensity to take risks, bases of power, most valued assets, and other factors likely to influence key decisions. Not least in importance is the adversary's value system, in which the same set of facts and prospects may be viewed quite differently from the way the United States would view them. Within that value system, the prejudices, misperceptions, and calculations of a leader who may be an absolute ruler accountable to no one can loom large in commanding an understanding of the situation to be deterred as we view it. Creating this understanding is an essential underpinning for U.S. and allied deterrence policy and action, and yet it is sometimes neglected to the detriment of those policies.
The action to be preventedthe whatalso needs to be clearly defined. In principle, we might want to prevent all acts that violate the principles of international law, especially when such acts are not consistent with U.S. interests. As a practical matter, however, we usually focus on those that may
have the most important consequences for the United States. And, in some cases, the more serious threats are not those with a high probability of taking place. For example, even during the Cold War, the probability of nuclear attack generally was quite low; however, the consequences of such an attack would be so severe that deterring such action demanded a high priority.
Finally, the means matter. An attack by conventional forces calls for one kind of response, and permits graded responses by various opposing forces. The possibility of attack by weapons designed to produce mass casualties might be an entirely different matter. As discussion below makes clear, it is not always apparent what the promised means of response should be, and deterrence of such attacks may be deemed to require means, such as theater missile defense, that entail both tangible and intangible costs that we have yet to face squarely.
Several important assumptions underlie most thinking about deterrence. Practitioners tend to assume, for example, that states are unitary actors, and logical according to our concepts of rationality. Deterrence also assumes that we can adequately understand the calculations of an opponent. One of the most important assumptions during the Cold War was that nuclear weapons were the most effective deterrent to war between the states of the East and the West. This assumption, carried into the post-Cold War era, however, may promote nuclear proliferation. Indeed, some authors suggest that the spread of nuclear weapons would deter more states from going to war against one another. The weapons would, it is argued, provide weaker states with more security against attacks by stronger neighbors. Of course, this view is also predicated on the assumption that every state actor's rationality will work against the use of such weapons, and that nuclear arms races will therefore not end in nuclear warfare.
In addition to threats of unacceptable damage, a strategy involving deterrence may also include more positive efforts to gain the desired behavior through an approach that extends the idea of "reassurance" to the post-Cold War conditions. The United States may undertake a variety of measures designed to induce rather than compel desired behavior, from various political and economic incentives to agreements on non-threatening force postures. Thus the notion of “deterrence" in its fullest sense may involve a complete range of tools to influence international behavior, including the use of diplomacy, trade, aid, military force, arms control agreements, and other means at the disposal of the deterring power. One way of enticing a potential adversary to forego aggressive behavior is to help the adversary develop a stake in cooperative arrangements among nations to achieve mutually agreed ends. As part of this participation, and to help bring it about, the adversary needs "reassurance," based on observation, that the opposing military posture is not immediately threatening-that it serves merely as an "existential" deterrent. Conversely, changes in such a posture can be made to reinforce deterrence if necessary, with due attention to avoidance of misperception on the part of the adversary. In thinking of deterrence as a process, it is useful to envision a succession of related, sometimes parallel steps, and to anticipate some options if deterrence fails. Such steps could include the following: Detect the onset or plans of
threatening moves; dissuade the potential aggressor from making such moves (perhaps through reassurance, carrots, and implied sticks); deter by putting the sticks in place; defeat the move if it is made; and destroy the capability to make such moves in the future. Such a series of steps, however, does not imply a rigid, linear process. Continuing, timely feedback is crucial in each stage; parallelism between deterrence actions promising military reactions and those using measures of reassurance can provide powerful inducements; and, if deterrence fails, appropriate responses should be adapted to the dynamics of the situation, and designed to deny success to the adversary, rather than automatically following plans made in the abstract.
THE NEW CONTEXT
The context for deterrence has changed dramatically since the end of the Cold War. Deterrence of actions inimical to U.S. interests must now take place in an unstable world in which the power of nations and of transnational groups is more diverse and spreading more widely. Devising relationships with many of these power centers remains a dynamic and changeable process. The problem of deterrence is thus more complex than it was, and the approaches to situations requiring deterrent actions must be even more measured and flexible. The likelihood that the United States may become involved in warfare using conventional forces, initiated by states other than the major powers, in matters involving less than national survival, has increased, while the relative roles of nuclear and conventional weapons have changed.
In connection with the nuclear threat we now must place more emphasis on issues of proliferation, while conventional forces are becoming increasingly important for deterrence in the strategic sense. The other weapons that produce mass casualties-chemical and especially biological weaponsare coming to be of increasing concern, because they may be wielded by rogue states or terrorists who do not subscribe to the commonly accepted tenets of international law.
In the bipolar world, the West developed a generalized notion of deterrence that sought to prevent the Soviet Union from using force to further foreign policy goals, mainly by coupling nuclear and conventional forces so that any use of force between the superpowers raised the prospect of escalation and nuclear war. This approach to deterrence was not effective outside the main NATO-Warsaw pact confrontation in Europe, because it became clear that we would view it as neither appropriate nor necessary to go to the extreme of unleashing nuclear weapons against perceived Soviet proxies in limited conflicts like those in Korea and Vietnam. More specific invocations of the mere existence of deterrent forces were more successful. For example, in the Cuban missile crisis the U.S. blockade of Cuba risked escalation that could have culminated in the use of nuclear weapons. The willingness of the United States to take that risk, together with overwhelming U.S. conventional force superiority in the theater, persuaded the Soviet Union to draw back from its
deploy and sustain a nuclear force in Cuba.5 In that case, U.S. security was directly involved, lending credence to any implied military threat, including potential use of nuclear weapons if it came to that.
In the current world situation, with one exception to be noted, U.S. security is not directly threatened, but our interests remain global. They have taken on an underlying economic character, which interacts with our desire to see democratic regimes spread and flourish based on the premise, articulated in the President's National Security Strategy, that "[t]he more that democracy and political and economic liberalization take hold in the world . . . the safer our nation is likely to be . . . . " In addition, as the developing countries of the world sort out their internal and external relationships in the post-Cold War environment, this sorting out process has generated ethnic conflict, large refugee migrations as people flee violence and war, and the potential for mass starvation and disease that the United States and its Western allies may feel obliged to alleviate on humanitarian grounds. Steps in this direction often involve military forces because they are well organized, they are disciplined, they can act quickly and with unique ability to focus resources, and the activities often need protection. After Somalia, Haiti, Rwanda, and Bosnia, the humanitarian mission for the U.S. military is becoming a more formally recognized element of the mission spectrum than was the case earlier.
The waning of direct threats to our national security and the rise of more general threats to international stability that are nevertheless inimical to our interests over the long term require thought about the nature of our strategic interests and about appropriate responses to such "less-than-vital" threats. Our strategic interests must now be seen as those involving the fundamental elements of our political cohesion, our economic well-being, and our ability to use our military power to defend them, wherever they are seriously challenged, rather than involving only an attack on our homeland. Deterrence in the face of such threats, which have reduced immediacy but nevertheless retain strategic importance, will probably require U.S. leadership to develop a sustained pattern of international responses to low levels of aggression that makes it clear that such behavior will not be tolerated by the international community.
Unless Russia, with its remaining large stock of nuclear warheads and intercontinental delivery systems, again turns hostile, we no longer face a serious military threat to our immediate survival. The prospect of war among the major powers is lower than at any time in centuries. Large-scale aggression is not a major temptation for the major powers; seizing territory by force is not worth the risk of major power conflict with the human and material expense that would be involved, although there may be a few marginal situations where the chance for a response is thought to be low, as in the Spratly Islands, or where the stakes are perceived by those involved to embrace issues of high political interest, as between China and Taiwan.
5 See GEN Anatoly I. Gribkov and GEN William Y. Smith, USA, Operation ANADYR: U.S. and Soviet Generals Recount the Cuban Missile Crisis, edition q, inc., Chicago, 1994.
There is some evidence that Russia's substantial numbers of nuclear weapons, representing the only serious threat to U.S. survival for the foreseeable future, may take on increasing political significance for Moscow as one of the few residual military supports for its status as a major power, after the serious deterioration of other Russian military capabilities. A more immediate danger is posed by the possible loss to outside forces of Russian nuclear warheads, or nuclear materials and expertise that would enable others to make nuclear weapons in a relatively short time. And the illicit "nuclear weapons" need not be nuclear explosives; if the bomb that exploded in the New York World Trade Center had included large amounts of highly radioactive material, for example, thousands of deaths could have resulted over time. These possibilities present the most serious security problem facing the United States and, accordingly, must continue to receive priority attention. In this case, an effective deterrent strategy will need to include substantial reassurance efforts vis-à-vis the Russians, involving cooperative relationships in many areas, and helping to provide assurance with high levels of certainty that all Russian weapons and fissile material are secure and fully accounted for.
The implications of the fundamental change in context are most significant for the nuclear force postures of the United States and Russia. Their operational safety is now more important than operational readiness. During the Cold War these forces were designed and organized to generate massive strikes from widely dispersed locations on short notice. But large numbers are no longer necessary, and time is no longer critical. As a result, the capabilities that contributed to a high state of readiness during the Cold War could now pose risks of unauthorized launch. Moreover, we are likely to have ample warning should Russia again turn hostile toward the United States or NATO allies. In addition, the conventional military balance has shifted dramatically in favor of NATO. Therefore, the primary objective now is to achieve a stable and enduring posture of deterrence that is not threatening to either side.
Another hostile superpower with a large nuclear arsenal of size comparable to our own under the START agreement could emerge at some future time, an event that would also heighten our concern about the risks of a major nuclear standoff. Thus far, such an event appears unlikely, but in any case we should have ample warning of its development. Consequently, we have turned increasing attention to the risks attending proliferation of nuclear weapons and other weapons of mass destruction. The dangers of proliferation are exacerbated by the existence of a few rogue regimes seeking such weapons, who appear to want them to establish regional dominance and to spoil perceived influence by the United States and its allies, and who we fear may use them in what we view as reckless disregard of the costs and dangers involved. These developments pose a new kind of problem for the United States: how to deal with a small, intensely hostile nuclear power.
First, we must decide how we define our interests where they interact with the presence of such opponents. Actual or potential military conflict with a rogue regime at some future time could, if the rogue regime possesses nuclear
weapons, place our troops and our allies at high enough risk to deter us from acting as forthrightly as we might if only conventional forces were involved. In effect, the possession of nuclear weapons by a rogue regime would lead us to view the importance of certain interests differently from how we might view them otherwise. Nor are we certain whether we would or should use nuclear weapons in response to attacks involving biological or chemical weapons.6These problems contribute to continuing issues in defining the role of our nuclear deterrent if non-proliferation efforts fail. There is as yet no agreement among the experts on these issues (including those who participated in this study). They are examined again, from different perspectives, at several points below in this chapter.
A second problem, which the experience of the Gulf War has led us to associate closely with the so-called rogue regimes, is the role of active defense against delivery of weapons of mass destruction by long-range ballistic missiles. Through a series of arms control agreements, the United States and the Soviet Union limited their defenses against ballistic missiles, and as a mutually agreed result both remained vulnerable in the interest of deterring a nuclear war. But, given the change in threat, a debate has emerged (and has also been reflected in differences of view, to be highlighted in due course, among the participants in the current study) over the wisdom of developing a theater missile defense (TMD) system whose performance characteristics could impact our Antiballistic Missile (ABM) Treaty obligations.
Some argue that such a capability is necessary to defend against proliferators and that it would help to deter such strikes. The defenses are believed to be especially needed to protect our allies against hostile actions by rogue regimes. Others argue that TMD systems can be easily and inexpensively overcome by alternate delivery means for weapons especially aimed at civilian targets, and that development of antimissile defenses threatens to undermine the fundamental stability among the major nuclear powers that is reflected in the ABM treaty and, by extension, associated treaties like START II. These arguments extend to protection of the U.S. homeland from attack by regional powers who may have small numbers of nuclear weapons and intercontinental-range delivery means derived by extending the capabilities of medium-range ballistic missiles or space launch capabilities (and who may have the ability to deliver biological weapons by the same means).
This controversy accentuates several issues in crafting a new set of deterrence policies. They include important linkages between offensive and defensive systems; the key role of arms control in constraining both; the implications of technical boundaries between the theater and strategic-level
6 Richard Garwin notes, ''If the treaties against possession of chemical and biological weapons are universalized, then an international coalition could act even before the use of such weapons, and even against a nation that had not accepted the biological or chemical warfare treaty. (That is my meaning of the term 'universalized.')" With the use of nuclear weapons to respond to biological warfare, the additional problem is what to destroy. To destroy only the biological warfare production capability may not be a significant deterrent.
systems, and the viability of those boundaries as guarantors of adherence to the ABM treaty; and the important role of the rogue regimes in setting the stage for the arguments. Examples of the last include the Iraqi Scud missile attacks against Tel Aviv during the Gulf War, and the significance for Japan of the prospect of North Korean long-range missiles that could carry nuclear warheads.
The problems entailed in stopping or even slowing proliferation will tax our ability to craft consistent and effective deterrence policies. The diversity of the actors in proliferation, and the complexity of our relationships with them, mean that they must be treated case by case. The ambitious Iraqi nuclear weapons program that had made significant progress, and Iraqi development of chemical and biological weapons for delivery by theater ballistic missiles, highlight the difficulty of learning the details of, much less deterring, such developments. The approach of involving the "target" nation in cooperative relationships with other nations, together with providing reassurance that an attack on that nation is not planned, is being tried as part of our approach to prevent acquisition of nuclear weapons by North Korea, with as yet uncertain prospects of success. The argument with Russia over reactor sales to Iran shows how difficult it is to establish and maintain situation-specific flexibilitythe Russians raise the question of why, if we are willing to give reactors to North Korea, we do not want them sold to Iran, subject to strict non-proliferation safeguards. Nuances of differences about the depth and reliability of inspection and control regimes to guard against potential diversion of weapons-grade materials become lost in the first-order political disagreements among the parties in the exchanges. All such interactions move our policy challenges far from the straightforward alignment of capabilities to counter opposing capabilities that characterized deterrence during the Cold War.
As noted in connection with deterrence of chemical and biological weapons use by rogue regimes, we must come to grips with still-unresolved questions about the appropriateness of invoking the prospect of a nuclear response to deter the use of those weapons, and if that is not appropriate, what such a deterrent should be. All such weapons are delivered in a political context. If the United States and our major allies are not directly threatened, the use of nuclear weapons against populations who are viewed as distinct from their "evil" leaders would likely be deemed inappropriate, no matter what weapons atrocities the leaders perpetrate. U.S. and allied holdings of chemical and biological weapons are prohibited by treaty obligations, precluding a response in kind; such a response would in any case be deemed inappropriate, for the same reasons. Thus, there is a growing body of opinion, still in contention, that nuclear weapons should be used by the United States and the industrialized nations solely to deter the use of nuclear weapons, while the use of other weapons of mass destruction should elicit conventional-weapons responses in addition to a wide array of active and passive countermeasures.
The coalition response to Iraq's invasion of Kuwait provides a model to deter such efforts in the futurea determined and decisive U.S.-led multinational force operating under U.N. authority. The Gulf War outcome
demonstrated that high-technology conventional weapons can strike devastating blows against the political, economic, and military power base of a country without the need to resort to the use of nuclear weapons. This demonstration has value in deterring major attacks with conventional forces, but as noted above that value might become less certain should another opponent, or Iraq in a future action, have nuclear weapons at the outset. Thus, one is led to the conclusion that both are necessary: powerful conventional weapons that we are willing to use in a strategic sense, and a nuclear deterrent that nobody would challenge by using nuclear weapons against us. In addition, U.S. nuclear forces must be sufficient to deter any combination of attackers who may have such weapons from using them against us or our closest allies.
The widely held perception, growing since the Gulf War, that the United States is unwilling to take casualties in conflicts that do not directly affect important U.S. interests also has important implications for deterrence in the new strategic environment. The rapid withdrawal of U.S. forces from Somalia after 18 soldiers were killed, and the ongoing uneasiness (before and even during their deployment) about casualties that might be suffered by U.S. ground forces in the former Yugoslavia, foster this perception.7Resolution of the issue of whether the United States has the will to use its powerful deterrent forces in any specific situation thus remains ambiguous, and detracts from our deterrent credibility.
Many also believe that we are at the dawn of a new era in the use of military forces. A military-technical revolution is said to be occurring that will fundamentally change the way forces are used. It is based in part on steadily advancing military technology, especially in the information-based areas of knowing the dispositions and activities of an opponent's forces and critical systems and facilities in detail, and being able to strike at them from long distances in a short time with great precision. Although it is not yet clear how we can integrate the new technology into new operational concepts and suitable organizations, some of the more important components are likely to include information warfare, precision strike, and decisive maneuveri.e., maneuver that overwhelms an opponent before he can respond. The challenge is to apply these new capabilities and concepts credibly, in fact or in prospect, to specific problems that we seek to deter, while preventing an opponent from using them against us, or mitigating the effects of such use.
The trends and developments that have reduced the chances of war between the major powers also suggest that international security arrangements may increasingly emphasize more use of positive, cooperative relations in place of the threats of punishment that have dominated deterrence over the past 50 years.
7 Some of the study participants who contributed to this report felt that the deleterious effects of U.S. reluctance to use force have been overstated. They believe that our prior actions, in the Gulf War, and during the Cold War before that, show that we will use our military when we deem the provocation to be sufficiently serious.
Thus we might expect to see measures involving inducement and reassurance play a much larger role in future strategies of deterrence.
The Dynamic Quality of National Interests
Defining national interests entails stating what is important to the nation and why. National interests are difficult to define outside a specific context, and thus it is difficult to prioritize them for purposes of allocating economic and military resources. We enunciate policies, argue them, and publish documents such as the President's National Security Strategy that tell the nation and the world what they are. But these statements have an abstract quality that is easily upset by world events. In some cases, such as the Korean Peninsula in 1950 and Kuwait in 1990, the context in which we evaluated our interests changed once we saw the facts of invasion and realized their full implications.
The implications of rapid geopolitical change for perceptions of national interest are not as commonly recognized as the presence of "stable" threats to obvious and enduring interests. We can tolerate, even encourage, change in national and international alignments without war, but violent change has its own threatening quality. Thus, in Korea, while a divided Korean peninsula that might evolve toward a different political condition was not of vital concern, the action taken to unify it by force under hostile Communist rule, especially when that action was viewed as a continuation of Communist expansion that had just "captured" China, could not be accepted. Forty years later, we faced a similar situation in the Middle East. While we felt we could watch without intervention while Kuwait and Iraq argued about the exploitation of oil fields that straddled their border, we could not tolerate the threat to the Western world's oil supply posed by an Iraq that moved militarily to sit astride some of the main sources of supply and to threaten major additional sources in Saudi Arabia. Elsewhere, it was argued by some, before our entry into Bosnia, that the war there was of no direct concern to the United States. But others noted that the war could expand and involve Greece, Turkey, and Russia, and threaten the unity of NATO. Had that happened, the same area would clearly have come to be of vital concern to the United States as well as to the others involved, and it was that risk that induced the United States to lead a strong effort, including military action, to terminate the ongoing war there. It was also argued that the need, on purely humanitarian grounds, to prevent mass murder of genocidal proportions offered sufficient reason for intervention.
Thus, perceptions of national interest appear to depend very much on the geopolitical context revealed by the dynamics of rapidly moving events. Deterrence policy cannot be formulated on the premise of a static world in which today's view and policy endure indefinitely. Change and conflict have always characterized the world. When the prospect of adverse change, potentially involving severe conflict, posed a direct threat to our survival during the Cold War, we acted to prevent the most threatening kind of violent change.
More rapid change, and instability, with counteraction much farther from influence by our direct intervention, will characterize the world of the future. Threats to our immediate survival will be fewer, at least for a time, but threats to our ultimate well-being as a nation will ebb and flow. Our deterrence policy will have to be formulated with these dynamics in view.
Deterrence can succeed only if the combination of threats and incentives is credible, and this requires both capabilities and political will. The United States, for example, can call on a wide range of political, economic, and military capabilities that would be overwhelming in most cases. However, several adversaries have not been deterred because they judged that the United States lacked the political will to incur casualties, sustain costs, take risks, and deepen its involvement when vital interests were not at stake. Also, many potential adversaries probably doubt that the United States will use nuclear weapons short of responding to a major nuclear attack on the United States or U.S. forces.
To persuade an opponent not to take proscribed actions, the capabilities and prospective outcomes invoked as a deterrent must convince the opponent that the costs, in terms of opportunities and value lost, judged by his own means of measuring them, will not be worth paying, and that in any case the deterring capabilities will prevent him from achieving his objectives. Furthermore, the opponent must be convinced that punishment will be forthcoming, and he must fear the punishment. Likewise, he must perceive that inducements offered will in fact be delivered. While the entire world understands the divisions between executive and legislature in the United States and various other parliamentary governments, doubt about whether the legislature will permit the executive to deliver on promised benefits can have as deleterious an effect on positive measures to induce desired behaviors as failure to punish can have on deterring undesirable behaviors. It is in areas such as these that uncertainties arising from vastly different political systemsU.S. and the potential opponents'-can contribute to the failure of deterrence.
Deterrence requires effective communications correctly perceived, so that the potential adversary knows that by undertaking the prohibited action he will incur substantial loss, or that by not undertaking it he can make a substantial gain. This can pose a dilemma for the deterring party in terms of the degree of specificity or ambiguity that should be communicated with regard to responses. In some cases, we may want a potential adversary to be uncertain about whether the United States will respond and in what manner. In other casesusually involving more important interestswe want potential adversaries to know very clearly that we will respond with overwhelming force. In the latter cases, the message can be communicated through exercises demonstrating the capability to
respond, as well as by direct and unambiguous communication. Indeed, coupling the two may make for the most powerful deterrent.
Similarly, the extent to which we should be transparent or secretive will vary with the specific situation. As a general rule, we may want to be more secretive when our capability or will may be inadequate, and we should be more transparent when we are more confident in our ability to act decisively.
When to communicate is also an important aspect of the message to be conveyed. It is difficult to know from historical situations what effect the timing of communications may have had on some action. Strategic games, such events as movements of British forces to Kuwait in 1961, and apparently the prelude to the Gulf War and the October 1994 U.S. deployment to the Gulf to counter threatening Iraqi troop movements, all suggest that movement of forces when a crisis appears imminent, perhaps together with verbal communications, has a more powerful effect than forces that remain in their precrisis posture, however powerful the latter may be. In the Cold War, heightening of alert status of forces had a similar effect. This observation simply conveys that communication need not be only verbal; the key is to make the message understandable, and to time it properly, and to make certain that it has been received and understood.
Our perceptions of what it takes to achieve effective deterrence may be different from the perceptions of those we are trying to deter. This possibility places a priority on understanding the other party, particularly in terms of vulnerabilities and needs. We also need to judge the propensity of opposition leaders to take risks. In any event, our calculations must explicitly identify assumptions and our level of confidence in the underlying estimates and assessments of alternatives for both sides. Calculating what constitutes unacceptable losses for a particular opponent is quite difficult. North Vietnam, for example, demonstrated an exceptional willingness and ability to sustain heavy losses; although deterrence was not an explicit part of U.S. strategy in Vietnam, dissuading the North Vietnamese from continuing the war by punishing them and thereby inhibiting their ability to pursue the war was a part of that strategy, even though it was largely unsuccessful.
Cultural and perceptual blind spots also present a danger in developing a strategy involving deterrence. Unfortunately they usually become apparent only after a disaster. Additionally, they may be peculiar to leaders, attendant on their political positions and the attention they must give to various internal constituencies. We must overcome cultural biases that color the views of our society, including our leaders, about others. As a minimum, we should assume that the rest of the world, including allies and potential adversaries, do not think as we do. They may be more willing to sacrifice human life to achieve certain goals; they may be willing to suffer more damage than we would in a similar situation; they may hold dear things that we would not, and vice versa.
Bridging such fundamental gaps requires concerted efforts to understand the perspectives of all the parties with stakes in a situation. One key question to ask is how they define the problem. Typically, it will be quite different from our own perceptions. Another key question is to ascertain what risks they perceive as being involved, and their willingness to undertake those risks. We must also be certain that we understand their own views of their strategic interests, the solidity of their power base, and other matters related to peace-or-war decisions.
Not all actions can be deterred. Some adversaries are willing to pay any price to achieve a vital national goal in an area that is of less than vital interest to the United States, so that threats of substantial damage or destruction are meaningless. Most terrorists fall into this category; indeed, we may not know the identity of the perpetrators until after the event, if then. Thus, this circumstance lacks the "deter whom" component. The possibility of failure, however, does not mean that we should not pursue a policy of deterrence. Some behavior, like crime, may not be totally preventable, but we nevertheless do and must continue to take steps to deter it. And, since history does not reveal its alternatives, we would have little idea how much worse the behavior might be without the deterrent actions.
In determining applicability, we should consider the extent to which the United States can influence a particular situation; the timing, including the domestic political calendar in the United States; and the most important factor, the resources we are willing to commit. If the deterrent effort will require a substantial commitment of resources over a sustained period-especially the lives of soldiersthen the stakes must be very high. In a related area, civilian casualties inflicted on an opponent also raise humanitarian concerns that affect the severity of punitive actions we will be willing to take, given the stakes. Anticipating the dynamic quality of our national interests, and acting against potential outcomes that are not immediately threatening but that may become so, must be part of the formulation.
There is a need for much enhanced intelligence and better means of interpreting intelligence data. The earlier a potential problem can be identified, the wider the range of options for action and the more likely that the problem may be deterred or deflected. Policy makers cannot avoid paying more attention to specific current events than to the distant future. Thus, there is a need, in association with an enhanced intelligence capability, for a recognized, high-level body of "strategic worriers," experts having a diversity of views and approaches, who can look at the more distant future and identify issues that need to be addressed as far upstream as possible. Once such problems are identified, timely and accurate intelligence, interpreted in light of the strategic issues related to U.S. interests, is required to support strategies of deterrence. Such
information will have to be developed on a worldwide basis and address the full range of capabilities, vulnerabilities, intentions, and likely perceptions of potential problem states in each region. Furthermore, intelligence must identify the most effective means of communicating with key individuals and states, both formally and informally.
DERIVATIVE POLICIES AND KEY ISSUES
Policies Involving Nuclear Weapons
Many of the deterrence issues that must be dealt with in the new context will involve changes in our policies involving nuclear weapons. Given the central role that nuclear weapons played in U.S. deterrence strategy during the Cold War, it is impossible to arrive at a new set of policies involving those weapons without extensive discussion, involving wide-ranging argument and sometimes strong differences of view. Many of the issues in which nuclear weapons play a role have yet to be articulated in all their complexity; many of them will appear to have one apparent solution in the abstract, only to have specific situations pose different choices from those that had been anticipated. In such cases, resolution of the issues will be in doubt until the nation is faced with the need for urgent decisions that pose the issues in concrete, situationspecific terms. The issues are presented in the following paragraphs without indication of firm solutions. For reasons similar to those expressed for the nation in general, differences in view within the special group of study participants reflected differences within the broader community concerned with matters of national security.
Coupling of Nuclear Weapons with Other Forces
The discussion about whether to separate or integrate nuclear and conventional forces to deter war takes on a different significance and orientation under current circumstances. During the Cold War, the issue was resolved in favor of integration in both Europe and the Pacific, mainly on the grounds that such integration helped offset the unfavorable conventional balance. The most recent Nuclear Posture Review8reportedly did not address basic missions for the nuclear forces. Given the new strategic environment, there is more reason to decouple nuclear forces from conventionally armed warfighting forces, and to limit the nuclear forces to deterring the use of nuclear weapons by others. This approach would be consistent with priority efforts to engage Russia in an overarching network of cooperative relationships as part of the new security framework for Europe. Establishing such a network of relationships involves clearly shifting U.S. and NATO relations with Russia from confrontation to cooperation, in a manner similar to what was done when the allies changed their
8 Secretary of Defense, Annual Report to the President and Congress, February 1995, pp. 8392.
relationships with Germany and Japan after World War II. The complications added by unfolding events in the former Yugoslavia, and the Russian reaction to proposals to admit Eastern European countries into NATO, show how difficult this approach will be. Such relationships must be built on a community of interests and understandings, which have yet to be fully established between the two powers.
Finally, it may be argued that decoupling nuclear weapons and limiting their use to deterring the use of others' nuclear weapons, and possibly others' use of chemical and biological weapons, may support non-proliferation efforts. Actions and policies to this effect may devalue their importance to potential proliferators and help reassure Russia that the United States does not intend to coerce Moscow and exploit Russian vulnerabilities.
Counterarguments to reserving nuclear weapons solely to deter the use of nuclear weapons hold that small nuclear weapons may have unique applicability in situations where it is important to take advantage of properties such as deep earth penetration, strong electromagnetic pulses, or enhanced radiation for rapid and widespread weapon effects that cannot be easily achieved by other means. In addition, as is indicated in the discussion titled "Extended Deterrence," below, situations might be foreseen in which it is important to retain a nuclear option to respond to an overwhelming conventional-force attack against a close ally who does not have nuclear weapons, unilaterally or as part of a coalition.
A further set of arguments must be reviewed in the context of a potential nuclear response to the use of chemical or biological weapons. As is indicated below in the discussion of targeting policy, our willingness to use nuclear weapons in situations where national survival is not at stake is problematic. Chemical weapons effects, deadly though they may be locally, would not be widespread enough to call forth a devastating nuclear response. Response to the use of biological weapons, if used against our own or allies' civilian populations, could face the difficulty of identifying the source, as well as the considerable time delay that might occur between the distribution of agents and the appearance of mass casualties. A nuclear response at that stage could be viewed as a response coldly calculated to create extensive casualties rather than a response in the heat and locale of battle, thus challenging our moral precepts. The situation could look very different in the event of an actual or potential biological attack leading to thousands or millions of civilian casualties by an identified opponent, against the U.S. homeland or a related vital interest, such as a similar attack on the NATO region. In that case, especially if nuclear weapons were the most ready response that could assure immediate devastation of the source of the attack, a U.S. president could be impelled by circumstances to make a nuclear response to attack. Use of chemical weapons would be more problematic, as their radius of action would be more localized, but the response could also be driven by the specifics of the events. This appears to be another situation in which the potential response would be conditioned on immediate circumstances at some future time. There is as yet no resolution of these issues in view; they have barely been discussed in public forums.
Even if nuclear weapons are for all practical purposes decoupled from battlefield forces and other weapons of mass destruction, and explicitly limited to nuclear deterrent roles, the implied potential for their usetheir very existencestill may make them effective deterrents in contexts other than the use of nuclear weapons by an opponent. For example, recent evidence from Iraq suggests that Saddam Hussein was deterred from using chemical and biological weapons that were prepared and delivered to missile sites and airfields because he believed that the United States would respond with nuclear weapons. According to the admittedly sketchy available data, such perceptions were said to be based mainly on Secretary Baker's threat of massive destruction of Iraqi industry if such weapons were used, with the implied threat of nuclear response. This is an illustration of the complexity of the issue in attempting to focus nuclear weapons on a tightly limited deterrence objective. Other complexities become apparent in this chapter's continuing discussion.
The new strategic environment suggests a need to review targeting policy for nuclear weapons. If the primary objective is to deter the use of nuclear, and possibly chemical or biological, weapons by others, then appropriate targets should include installations that are highly valued by potential opponents, including those that affect their ability to deploy and employ the weapons. Especially difficult would be the selection of appropriate targets for nuclear weapons that respond to the use of chemical or biological weapons against our forces in the field. Humanitarian concerns about opponents' civilian casualties will also figure in selection of targets, since if the weapons of mass destruction are used by a "rogue regime" that is seen as not representing the will of its people, there will be U.S. public resistance to inflicting severe punishment on civilians who might be viewed as the regime's additional innocent victims. Such regimesor, at least, their leadersrecognize these concerns on our part and often embed their military activities in their own civilian populations and infrastructures, making our decisions in this area especially difficult. Our response, again, could depend on specific circumstances and the level of casualties (and therefore public revulsion and anger) the attack might create. Part of our deterrence action must be to signal these factors and their possible consequences to the regime if the strategem does not work in their favor.
No First Use
There is disagreement about whether the United States should establish and, further, whether it should then announce a "no first use" policy. During the Cold War, the United States was under pressure from the Soviet Union and China to adopt a policy of "no first use" of nuclear weapons. Subsequently, however, Russia abandoned this public policy position, although China and other states still call for such official statements on the part of all nuclear powers. Many argue that a no-first-use policy is not credible, nor is it binding.
For example, nuclear weapons may be used in ''defense-of-last-resort" circumstances by weaker countries under mortal attack by powerful neighbors. While the United States does not now anticipate such circumstances, this cannot be said to be true for all our allies in all circumstances.
The United States has offered some assurances to the effect that such weapons will not be used against non-nuclear weapons states that are Non-Proliferation Treaty (NPT) signatories other than in circumstances such as alliance with another nuclear power. Similarly, we have pledged to take positive steps, including calling in the U.N. Security Council with its nuclear member states, in response to any threat or use of nuclear weapons against a non-nuclear state. Thus, in our own strategy, given our preponderance of conventional strength, we convey the message that we do not currently visualize using nuclear weapons in response to a conventional attack on our interests. How the policy would play out, in terms of warning and response, against a conventional attack by a country that had nuclear weapons and might evince willingness to use them against a massive U.S. and allied conventional response, is a question in need of further consideration. As in other areas involving the use of nuclear weapons, decisions on these matters will likely be resolved in ways particular to specific situations. It would certainly be useful, however, to give some a priori thought to them, considering the circumstances.
Any reexamination of this issue should focus initially on exploring any post-Cold War changes in the policies of all five declared nuclear weapons states (United States, United Kingdom, France, Russia, and China), which are also the five permanent members of the U.N. Security Council. The examination must then be extended to explore the circumstances of our allies, and the allies of Russia and China, in different areas of the world, and how a policy of no first use on the part of the United States would affect all of our strategic positions vis-à-vis regional neighbors.
Weapons of Mass Destruction and Precision-Guided Munitions
At the core of the concept of deterrence is the known ability to inflict damage that the opponent will view as unacceptable. Therefore nuclear weapons have come to be closely associated with deterrence because of their well-known ability to cause mass destruction and casualties. Other means of producing mass casualties, such as chemical and biological weapons, are, for reasons of principle and treaty obligations, not available for the United States in a deterrent role, even for response in kind. As suggested above, we could be self-deterred from using nuclear weapons in any circumstances where they are not used first, by the prospect of unintended civilian casualties and destruction. Traditional conventional weapons, including incendiaries that were used extensively in World War II, can cause just as much damage to a nation's vital facilities, given enough time and delivery resources. Extensive damage to resident populations is also a byproduct of such attacks; as has been noted at several points above, inflicting such damage in matters involving less than
national survival is likely to be viewed as unacceptable to the U.S. public, and likely our close allies' societies.
The capabilities of conventional weapons have evolved dramatically, however, offering what many see as an exit from the mass-casualty dilemma. Highly accurate and lethal advanced conventional weapons with precision guidance can destroy specific targets with much less collateral damage, and they can be delivered at long ranges with relatively low risk of friendly casualties. To be effective, however, they require accurate and timely intelligence and full situational awareness about the status of targets and defenses. These new capabilities and attending support requirements are currently driving the evolutionary development of the U.S. armed forces. For the United States in its current world situation and posture, these weapons may be more credible than nuclear weapons in deterring any kind of aggression below the level of a nuclear strike. However, their action, while more precise and selective, is also much slower in a relative senseit takes days or weeks for their full effects to become apparent, as we saw in the Gulf War and Bosnia, compared with the time to deliver a single or a few nuclear weaponsand affords more ability to respond. A party that thinks it can respond successfully will be less likely to be deterred by the potential use of conventional weapons. Also, if nuclear proliferation creeps ahead, the relative postures of the United States and nations whose actions we wish to deter will change. At low levels of threat, conventional weapons may be a more effective deterrent than are nuclear weapons, but the contributions of conventional weapons in deterring a nuclear threat are another matter. Thus, the relative value of the two kinds of weapons for deterrence remains to be seen, probably through many trials, as in the Gulf and Bosnia examples. Again, the choice of weaponry for deterrence and response will probably depend at least as much on the specifics of situations and the intensity of threats to our interests as on the inherent capability of the weapons. In the long run, therefore, the nuclear and advanced conventional capabilities are complementary rather than interchangeable, and both sets of capabilities must be retained within our force posture.
"Extended deterrence" refers to the umbrella we extend over our allies to protect their homelands, as well as our own, from attack. In the Cold War, extended deterrence referred mainly to nuclear attack, although as this chapter's authors note, nuclear attack and conventional attack in NATO Europe were, by design, not decoupled from each other in deterrence policy. Although the threat of attack on our closest allies is low, the U.S. nuclear umbrella remains important. And because the threat is low, the relationships that form this important framework could atrophy unless they receive regular attention.
Extended deterrence also serves to obviate the need for the allies to develop nuclear weapons capabilities of their own. Germany and Japan, for example, could easily (in a technical sense) develop nuclear weapons but instead rely on a
close security relationship with the United States. However, the spread of nuclear weapons and other weapons of mass destruction to smaller states, and neighboring rogue states, coupled with perceptions about U.S. willingness to respond in local conflicts not involving deep U.S. interests in the static situation, could change those states' perceptions of need for their own weapons. In addition, power relationships in the world will change over time, and a currently unforeseen, credible threat of an overwhelming conventional attack against a close ally could arise in the future. Such a development could resurrect the Cold War era arguments in favor of using nuclear weapons in extremis to respond to massive conventional attacks. Thus, although the drift of events and world power structures appears to favor reserving nuclear weapons to be used only to deter the use of nuclear weapons, including their use in extended deterrence, their potential use as a deterrent against conventional attacks in some future circumstances cannot be totally ruled out as we maintain our extended deterrence posture. This is a matter of some controversy. It may be a case in which, by necessity or by design, ambiguity should be preserved until the need for a decision in a specific situation appears.
Applying Deterrence Policy
"Existential deterrence" simply means the existence of powerful forces that a potential challenger knows can be brought into action if the need arises. There is always hope that when the United States expresses a desire to influence the outcome of a situation in which the use of military force may be involved, the sheer power we bring to the table by virtue of the existence of the strongest military forces in the world will weigh heavily in leading to a resolution of the issues. However, in the new environment as in the old one, the fact of the existence of those forces, alone, will not be sufficient to prevent many conflicts or actions against U.S. interests. This is largely the result of perceptions about the willingness of the United States to actually commit its formidable military power in specific situations. Therefore, if a threat is perceived, it will probably require more direct efforts on the part of the United States to communicate our will and intention to act than has been the case in the past. The lessons of history about the dynamic quality of our interests, and the impact of force movement and visibility on adversaries' perceptions at critical times, will be especially important.
In connection with existential deterrence, however, nuclear weapons play a special role. They provide existential deterrence whether they are actually deployed or not. Nuclear weapons have been given credit by some for having produced the longest absence of all-out world war in recent history. This role of nuclear weapons might even continue in the form of "virtual extended deterrence," since most industrial nations could regenerate nuclear weapons in a period as short as 1 or 2 years if an extended worldwide conflict were to occur
again. Moreover, certain U.S. deterrent actions might be conditioned by knowledge of nuclear holdings by a prospective antagonist.
The United States can be self-deterred from acting by establishing overly strict criteria for the use of military force. The criteria put forth by former Secretary of Defense Caspar Weinberger, for example, which established conditions of mission specificity and the predictability of success, and specified conditions for withdrawal of forces before an expeditionary mission could be undertaken, could preclude U.S. intervention in a wide variety of conflicts that we would nevertheless like to deter. Similarly, we have implicitly established very high thresholds for the use of nuclear weapons so that their use is credible only when the most vital interestsperhaps limited to prior use of nuclear weapons against the United States, NATO allies, and Japanare at stake.
A major problem for deterrence is dealing with incremental, threatening steps taken such that each one may not warrant a major response, but which cumulatively will result in a situation we want to deter. Taking a massive punitive action in anticipation of a possible but not certain outcome toward which only a first small step may have been taken, will raise public concerns and objections that can have undesirable domestic and international political consequences. A related problem involves a kind of counterdeterrence when an opponent raises the stakes to a level we find unacceptable. An example of these problems is our effort to deter North Korea from building nuclear weapons and North Korea's counter to the threat of sanctions, claiming that they would be an act of warsomething we clearly want to avoid. How to come to grips with such situations is a major factor in developing a post-Cold War deterrence strategy. Perceptions of will and willingness to take risks and suffer the consequences of failed deterrence will be determining factors in any such situations. Parallel actions of reassurance must clearly be part of the arsenal of tools at our disposal when military force alone will not resolve undesirable situations.
This discussion notes at several points (e.g., in the above section titled "Communications") the dilemma between enunciating clear deterrence policies and maintaining ambiguity in situations where such announcements may either exacerbate a situation or provide an opening for exploitation or miscalculation by a would-be aggressor. Yet clear policy declarations are often necessary for domestic political purposes and to reassure our allies, even as they may be viewed as stimulating undesirable reactions from prospective opponents. The circumstances in which declaratory policies are advisable will vary; often they
are not predictable, as we found in the cases of Korea and Kuwait. An intermediate position would eschew explicit declarations of policy in areas where the dynamics of a situation could change the policy rapidly, but would maintain a transparent diplomatic and military posture that could rapidly be translated into action, and thereby designed to give a potential aggressor pause in a situation of ambiguity. Given the many remaining uncertainties attending the formulation of policies regarding the potential use of nuclear weapons, that area may be one where it is most advisable to assume such a position. At the same time, for the sake of our own security and that of our allies, there must be no doubt about the fundamentals of where our interests and potential for action lie in areas of our own and our allies' vital interests.
Alliance Implications of Exercising Deterrence Policy
Unilateral and Multilateral Deterrence
Unilateral deterrent capabilities are attractive because they provide more freedom to act, they have simpler requirements than multilateral actions do and thus can be undertaken more quickly, and they are likely to be more secure, with preparations that can be undertaken in secret. However, "mutual deterrence"of one party by another has lost much of its meaning with the end of the Cold War. The objects of deterrencethe "deterrees"are less easily foreseen. Moreover, the United States has been increasingly less willing to use its military forces with inherent risks of large casualties without acting in cooperation with others. Therefore U.S. unilateral deterrence is becoming progressively less credible, and a trend toward multilateral deterrence is becoming more likely, although U.S. leadership is still sought.
Multilateral deterrence offers compensating advantages. Most importantly, a widely shared effort may be perceived as overwhelming. Also, with a shared burden, the cost of deterrence is lower for each party. Furthermore, a multilateral effort is more legitimate in international perceptions than are unilateral efforts, particularly when it includes states from different regions and cultures. A broadly based coalition also may be less vulnerable to attacks by the party to be deterred, and it would allow a wider variety of potential responses. Unilateral deterrence and multilateral deterrence are not always mutually exclusive; unilateral actions can be used to stimulate or to complement multilateral actions.
A difficulty with multilateral deterrence, illustrated in Bosnia, is that it would be more constraining. Achieving agreement within the deterring coalition as to when and how to react to provocation could be more difficult, thus potentially presenting many opportunities for the aggressor to play on individual coalition members' special interests, and so to divide the coalition or force it into inaction. In this respect, regional coalitions of nations that have a commonality of interest in a specific situation, such as NATO during the Cold War, will prove more powerful than generalized coalitions such as the United Nations.
Alliances and Coalitions
Almost every situation in which we will want to deter some action inimical to our interests and those of our closest allies will involve an international coalition, often assembled ad hoc. Preparatory steps to ready such coalitions for expeditious action will make it easier to activate them in crises, and will make the potential of the coalitions more compelling to would-be aggressors. Such steps may include discussions with governments that may be involved, review of forces that may be assembled under U.N. or NATO aegis or some other international agreement, and military visits, combined practice, and development of common command language and military force interoperability in peacetime.
To take such steps, we must anticipate where crises may arise. Although premature publicity about such anticipation can have adverse political impact both nationally and internationally, taking the preparatory steps during normal interactions between governments and forward-deployed U.S. forces, with appropriate timing linked to ongoing events, should mitigate the risk of adverse political impact and could in some circumstances add to deterrence. The particular circumstances of potential coalition members, their relationships with more powerful neighbors, the opportunities to benefit from U.S. extended deterrence and to influence U.S. regional policies and actions, and eased circumstances for U.S. presence in a region may all be inducements to enter coalitions despite perceived political risks in doing so.
Defense Against Ballistic Missile Attack
There is disagreement, nationally and reflected in some of the appended papers, about the value of strategic defense of the United States, whether such defense should be deployed, and the role of theater missile defense in the broader picture.
The Strategic Significance of Ballistic Missile Attacks
The capacity to launch ballistic missiles is spreading to over 40 nations, large and small, in different parts of the world, and some number of these nations will use such missiles for military and political purposes. The range of missiles that can be used even by small nations is growing from a few hundreds of kilometers to over 1,000; nations that will have significant space launch capabilities will be able to convert those capabilities to attack targets at distances from 3,000 kilometers to intercontinental range. Such missiles are viewed as an especially grave threat not only because of their capacity to attack tactical forces from long range, but also because of their ability to strike at vital centers of logistics, industry, and population with warheads of mass destruction. While the missiles would be vulnerable to destruction if not held in silos or caves, circumstances could be imagined in which they can be used in a first strike. Although the ballistic missiles used by Iraq in the Gulf War had only
conventional-explosive warheads and were highly inaccurate by modern standards, they could hit cities in Israel and Saudi Arabia. Thus, they carried important political implications, and their use diverted significant effort from the main military tasks of the coalition opposing Iraq in the war. In this sense of being able to strike at an enemy's heartland, and of being able to affect political alignments and war plans in a military conflict, the weapons must be considered to have "strategic" as well as tactical value, whether they are used against the United States, or our forces and bases overseas, or our allies.
Defense in Depth Against Ballistic Missiles
Active defense against ballistic missiles must be considered, in the strategic sense, as only the last stage of a defense in depth that begins with peacetime policies including deterrence of aggression in the first place. More specifically focused deterrence attempts to inhibit the spread of effective ballistic missile capability, through treaties and treaty-like arrangements such as the Missile Technology Control Regime to which many nations, including the United States, the nations of NATO Europe, Russia, and China, subscribe. Deterrence of military conflict of any kind is the next stage, and deterrence of the use of missiles in such conflict, and especially of missiles with warheads of mass destruction, follows. The defense against the missiles themselves is part of a broad defensive array that encompasses a spectrum of activities and systems. Included in that array, in addition to antiballistic missile defense, are attacks against missile launch sites, antiaircraft defense against attack by manned aircraft and cruise missiles, and widely distributed defensive measures against deployment and delivery of weapons of mass destruction by military units or by clandestine means such as disguised civilian ships and aircraft entering commercial ports and civilian airfields. A variety of passive defense measures at potential target sites is also included. Active defense against ballistic missile attack, including attacks against both the U.S. homeland and allies' homelands and deployed coalition forces helping to defend those homelands, must be viewed in this broader context. All experience in warfare tells us that none of the stages of this defense in depth can be expected to work perfectly. But each stage has an important contribution to make toward a strong cumulative defense. The potential for success of each stage of defense and the cost imposed by each on potential opponents will determine how resources should be distributed across the multistage defense.
Active Defense Against Ballistic Missile Attack
Active defense against ballistic missile attack has taken on an aura and importance beyond those of most weapons systems because during the Cold War the prospect of such defense affected relations between the United States and the Soviet Union and the size and nature of the offensive missile holdings of each of the countries. Deploying antimissile defenses threatens U.S.-Russian treaty obligations that have lent stability to the offensive strategic deterrence
equation. Many argue the need, now, for protection of the United States against small attacks by states other than Russia, and protection of allies against analogous attacks.9
The problem of intercepting a missile with high probability, in the terminal phase of its flight, is well on the way to solution, although full development and implementation of the capability have yet to be achieved. Other aspects of defense against ballistic missile attack are more difficult to solve, however; if the warhead is nuclear, it must be intercepted before the warhead can continue on a ballistic trajectory to the target vicinity and detonate with wide effect, and before detonation on contact with the interceptor can cause damaging effects from a greater distance. Intercept must occur during an even earlier stage of a missile's trajectory to ensure that it does not disperse submunitions carrying chemical or biological agents-the militarily more effective means of attacking with those weapons-thereby rendering terminal defenses only partly useful. Intercept early in the trajectory also avoids the difficult problem of differentiating decoys from true warheads, a problem that otherwise presses toward terminal defense after both warheads and decoys have reentered the atmosphere. These technical pressures lead to concepts for boost- or ascentphase intercept, and to "preboost" system concepts to find launchers and missile command, control, and targeting complexes and destroy them before missiles can be launched.
The growing number of potential attackers who may threaten our allies and affect our interests, or who may ultimately threaten our homeland, suggests a major expenditure on proliferation of antimissile defenses. Such defenses must also be deployable on short notice to protect allies in a crisis, if they are not already in place. Combined with the technical needs for boost-phase intercept and prelaunch attacks, and separating consideration of defending the United States for the moment, a desirable "theater-strategic" goal emerges that would place a "cap" over a hostile or rogue nation that might threaten to launch ballistic missiles against U.S. bases or allies' vital facilities and their populations, to keep such missiles from emerging beyond that nation's borders. It has been suggested that the United States might join with Russia to create such a capability. However, if the Russians had it, they might not agree with us about what a "rogue nation" is, and they would be able to threaten our allies in NATO Europe who have strategic deterrents of their own or those where the dual-capable aircraft of NATO's theater nuclear capability are based. Thus,
9 Wolfgang Panofsky points out that "attacks against the United States by states other than Russia, if they are to occur at all, and protection of allies against analogous attacks, are likely to be delivered by means other than ballistic missiles. The United States is vulnerable to such delivery across unguarded land boundaries, by air, by ships in harbor, and the like. Thus expending of large sums for defense against limited ballistic missile attacks seems inconsistent unless comparable efforts were instituted against other means of delivery. For example, the unsuccessful effort in drug interdiction has shown that an effective comprehensive defense seems unattainable without enormous expenditures and compromise to fundamental American values."
even in this context, strategic and tactical or theater missile defense objectives cannot be cleanly separated.
Part of the argument about defenses focuses on the desirability of such expenditures for U.S. continental defense, in relation to deterrence by offensive missiles and systems, in view of the treaty problems that the deployment of defenses would raise. In that context it has been noted that even under START II the United States will have enough nuclear weapons to devastate any attacker, in a situation that is unsymmetrical against any but the Russians. Therefore, it is argued by one side of the debate that the threat of totally devastating retaliation should be sufficient to deter an attack of the kind that has been of growing concern.
The other side of the dispute points to the attempt that is being made to differentiate between strategic defenses and those that we may either put in place or deploy on short notice to defend overseas base areas and allies as part of broader theater defense activity. There are differences in defenses designed to protect the U.S. homeland and those designed to be used at relatively short range in theater missile defense. Defense against theater ballistic missiles is acceptable within the constraints of the ABM treaty. However, as defenses are designed for boost-phase or ascent-phase intercept from greater distances, and as theater-level missile ranges increase, so that the speed and reach of terminal defenses must increase, the technical boundaries between theater missile defense and strategic defense of the U.S. homeland will blur. Russia has indicated that it feels strongly about what might appear as unilateral changes in the ABM treaty, and could hold the arms reductions of the START II treaty hostage to these concerns.
As this debate proceeds, the results of theater missile defense developments, and the results of allowable ballistic missile defense research and development, will influence it and will ultimately change the dynamics of the argument.
ANALYSIS, MODELING, AND PLANNING
Although a variety of theoretical works relate to deterrence, the practice is based mainly on the collective wisdom of those who have had to devise ways of preventing bad things from happening in specific cases. Collecting assessments of real experiences results in some general notions about whether deterrence worked or failed in the past and why. Often these notions are based on cases in which deterrence was not an explicit part of the strategy. Such efforts to apply the concept of deterrence retroactively are not useful because they often represent the use of selective examples to support a preconceived idea.
Deterrence has not benefited from a great deal of practical study because there are relatively few cases available in which an explicit strategy involving deterrence was developed. The major case, the strategy vis-à-vis the Soviet Union, offers only partial instruction applicable to present and future conditions. Furthermore, understanding of those cases in which deterrence played a major role tends to be one-sided because the necessary level of detailed information
from opponents is not available, although increased access to former senior Soviet officials and archives will now make more balanced studies possible. Deterrence theory has also been subject to implicit constraints imposed by the prevailing cultural values, and the "tyranny of the best estimate," in which alternatives posited by different members of the intelligence community are argued out and compromises are reached to give the "best estimate" of a situationa process that tends to focus views according to majority judgments devised exclusively within our own value system.
Current research on deterrence expands to draw on notions derived from behavioral theory, systems analysis, decision science, and other related fields. At the microlevel, behavioral models include a simple linear set of relationships involving a stimulus affecting an organism resulting in a response followed by consequences. In thinking about such relationships in a more complex way for the purpose of understanding the dynamics of deterrence, the key is to focus on the orientation of the organism both as an individual and as an organization within a culture and an environment that can affect its or their calculations of potential risk, loss, and gain. In particular, this analysis should seek to identify the relevant needs, vulnerabilities, ideas, feelings, and experiences that are most likely to elicit certain behavioral responses.
Of particular significance is the focus on feedback in developing models of relevant systems. Models of decision making can usefully compare the options of a party to be deterred in terms of most likely, best-case, and worst-case outcomes, within the value system and environment of the party. Such models may incorporate different images of the opponentsuch as one that assumes a pragmatic incrementalist, and another that assumes an exceedingly ambitious, frustrated leaderand different values, such as attitudes toward loss of human life. Moreover, given the difficulty of understanding and predicting how different personalities in different cultures and circumstances may behave, it is important to consider alternative images of specific adversaries and allies, when using these tools for assessment. The ability to vary context and the decision-making personalities of the individuals and the organizations in such analysis and modeling allows formal exploration of different possible responses, guards against premature or one-sided assessments, and encourages escape from the "tyranny of the best estimate."
Target analysisi.e., analysis of the "deterree"for the purpose of devising a strategy incorporating deterrence should clearly identify assumptions and include a comprehensive examination of the needs and vulnerabilities of the target. These analyses should be incorporated into net assessments of the United States compared with potential adversaries and allies. Net assessments should define the problem, identify U.S. and potential opponents' and coalition partners' objectives, and place the situation in the proper context of political, economic, and security trends and developments, and consequences of success and failure for each participant. Conclusions should include explicit judgments about the prospects for success and the level of confidence in the information and analytical results. Gaps in information should be identified along with key
indicators that suggest that assumptions may no longer be valid or developments are taking a different course, calling for new assessments.
Neither modeling nor gaming can predict outcomes with any confidence. However, modeling and gaming can provide useful insights to strengthen deterrence programs. They are particularly useful in understanding the dynamics of deterrence that often are not apparent in static analysis. Modeling can integrate a wide set of variables and may be particularly helpful in understanding key relationships and linkages that may not otherwise be apparent. Modeling also can help in understanding the likely consequences of alternative strategies. Similarly, gaming with expert surrogates provides opportunities to observe the interplay between two or more sides and to understand the rationale behind key responses in different value systems.
Both of these analytic tools can be useful in examining important strengths and weaknesses of all sides-essential information for effective deterrence. Both can and should be applied a priori to anticipate potential crisis situations and the field of possible responses. They can be especially helpful to the group of ''strategic worriers" called for above and for familiarizing strategic decision makers with situations and possible responses they may actually be called upon to face during their tenures.