Page 64
APPENDIX A Revising the Practice of Deterrence
The concept of deterrence is a product of the Cold War. Though the underlying principles can be detected in the military writings of all historical periods, the word itself and the elaborate conceptualization that accompanies it have been developed over the past 50 years in the course of establishing a rationale for the deployment of nuclear weapons.1 The familiar central doctrine holds that nuclear weapons are maintained to prevent their use and, by extension, any large-scale form of warfare by threatening retaliation destructive enough to override any rational motive for aggression.
This concept rests on a theory of human behavior. Assuming that the primary danger is that of a war arising from deliberate calculation, the theory posits that a countervailing threat displayed with sufficient probability and sufficient destructive potential can dominate any aggressive calculation that might be made, no matter how perverse or myopic it might be. It is apparent from the historical record that this theory did not inspire the creation of nuclear weapons in the first place, nor did it very directly determine the size or composition of the deployments that occurred. Nonetheless it is arguably as consequential as any theory of human behavior has ever been. The single word "deterrence" has been widely accepted as a summary statement of the most fundamental national security objective and indeed as the central pillar of foreign policy. Within the United States, it is perhaps the most solidly established element of political consensus—the least disputed function that the increasingly beleaguered national government performs. Moreover, within the military establishments that deploy nuclear weapons, the conceptual elaboration of deterrence provides the main guidelines for practical decisions on the size and composition of forces and for the daily management of their operations.
The entrenched practice of deterrence has survived the declared ending of the Cold War essentially unaltered—a fact that is hardly surprising given the critical organizing functions that the concept has come to perform. The rhetoric of confrontation that originally accompanied the doctrine has been replaced with more polite forms of political discourse, and overall nuclear weapons deployments are being reduced to less than one-quarter of their peak levels. Nevertheless the main forces still continuously preserve the ability to initiate deterrent retaliation within 30 minutes—the nominal intercontinental flight time of a ballistic missile. And even with their scheduled reductions fully
1 The history of the concept is briefly reviewed in Alexander L. George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice, Columbia University Press, New York, 1974.
Page 65
accomplished, the residual capabilities of the United States and Russia will be virtually as lethal to each other as they were at the height of mutual antagonism. In terms of political consensus and institutionalized procedure, neither establishment knows how to do it in any other way.
Understandable as this situation may be, it cannot be continued indefinitely. The entire context of international security is being radically altered, and the emerging problems require different organizing principles. Moreover, all along there have been underlying dangers whose importance was obscured by the ideology of the Cold War. It is not responsible to tolerate those dangers in the new context. The prevailing practice of deterrence will have to be substantially revised. The sooner this is appreciated and the more systematically it is accomplished, the better off we all will be.
CHANGING CONTEXT
There is as yet no agreed formulation or summarizing imagery to characterize the period of history that is to follow the Cold War, but already it should be evident that it will involve a major transformation of international relationships.2 A globally extended economy is forming, driven by a revolution in information technology. The scale of this extended economy will have to undergo an unprecedented expansion as the world population surges over the next five decades. The effects associated with these two phenomena can be expected to generate extensive changes within most societies and will certainly alter their interactions.
The revolution in information technology is already a familiar event in terms of its effects on consumer products and thereby on daily life. Over the past two decades the inherent costs of performing the basic functions of storing, processing, and long-range transmission of information have undergone precipitous declines. Though agreed measures of these cost declines have not been fully established, they clearly amount to several orders of magnitude—factors of a thousand to a million or more. That appears to be the largest efficiency gain of any commodity in economic history. Highly facilitated information flows are enabling the production of goods and services to be conducted on a global scale and the market forces derived from that fact are spontaneously inducing an integrated international economy. This process is also diffusing technology and basic cultural information so extensively that the entire pattern of social organization seems likely to be affected.
At the same time we are encountering an unprecedented surge of the world population—the rapid rise associated with an exponential growth sequence before it reaches some natural or induced limit. Barring a cataclysm, the world population will increase by roughly 1 billion people per decade over the next three decades and will exceed 8 billion by 2025. The trend thereafter is not yet
2 Steinbruner, John. 1994. "The Problems of Strategic Realignment," paper prepared for the 1994 meeting of the Atlantic Conference of the Chicago Council on Foreign Relations.
Page 66
determined, but a trajectory that reaches 10 billion by 2050 is a plausible possibility. More than 95 percent of whatever increase occurs will come in the poorest communities. The absolute magnitude and the distribution of this surge is a combination without precedent in human history and will clearly give tremendous impulse to the internationalizing economy.
As an obvious consequence of this impulse, economic performance is likely to become the principal determinant of national viability and therefore the central objective of policy for all governments. Moreover, performance will necessarily be defined not only in terms of overall growth but also in terms of distribution. Unless the globalizing economy successfully extends its reach to those people in the lower economic strata, where the population surge is occurring, the coherence of many if not all political systems is likely to be in question and some would almost certainly be torn apart. It is difficult to imagine a successfully operating international economy of 10 billion people, 6 billion of whom live under conditions of endemic austerity and another 2 billion who experience continuously declining standards of living. The amount of violence generated in an integrated economy of that sort would presumably be massive, more than the prosperous 2 billion could reasonably expect to contain by coercive means.
The expansion of economic participation required to assure a favorable trend in economic equity—that is, an absolute and relative improvement in the standards of living of the poorest population segments—implies that the global economic product will have to increase by a factor of five or more, including a probable tripling of energy and agricultural production. Were that to be attempted on the basis of current technologies even taking their natural evolution into account, the environmental consequences would probably be severe enough to preclude the economic growth objectives, at least in some of the more burdened regions of the world. That implies that massive investment programs will have to be undertaken to alter the core production and consumption patterns on a schedule commensurate with the population surge, a process that would entail large structural and technical shifts within virtually all national economies. It also implies an increasing sensitivity to the balances of material flows and to their environmental effects, a development likely to be of decisive importance in the more burdened regions and potentially so on a global scale as well.
As these implications emerge, there will also undoubtedly be a diffusion of political power. National governments struggling to assure economic performance will not have autonomous means to do so. Information technology is enabling, probably in fact compelling, the decentralization of many decision processes, thereby eroding the degree of control that national governments are able to exercise within their societies. It is simultaneously driving the global extension of basic economic activities, thereby dispersing control into the international economy as a whole. The predictable longer-term effect of this pattern is to drive national governments into more consequential collaboration.
Page 67
That in general will be the only realistic means of extending their effective authority.
STRATEGIC IMPLICATIONS
It is possible to project conceptions of political intention that appear to justify the continuation of Cold War deterrent practices in this emerging situation. The conceivable reversion to authoritarian government and to expansionist policies in Russia is being advanced in official documents as a major reason for the preservation of U.S. deterrent forces.3 Informally there are more venturesome variants. One can observe, for example, that China and India together encompass one-half of the world population and that both are poised for rapid economic growth, as is Indonesia. If one assumes that national identity can somehow override the globalizing trend and that these powers will join Japan on the frontiers of technology, then one can posit the emergence of an Asian power center or alternatively a major confrontation centered in Asia.
It is worth noticing, however, that none of the countries in question is yet making the extensive investments required to develop classic military power projection capabilities on a global or even a regional scale. In fact the United States is currently the only country sustaining investments of that magnitude, and there are strong reasons why others would not attempt to match what we have done. Those reasons have to do with the deeper implications of the transformation that is in progress.
Perhaps the central fact is that deliberately calculated, large-scale aggression—the central focus of deterrent policy—is simply not a major temptation for a major government. The classic exercise of seizing territory by force is not worth the risk and expense involved, save for a few marginal situations. With the advanced capabilities of the United States and its major allies, those exercises can in principle be detected and preemptively defeated, and even an initial success could not be sustained. Basically the assertion of political jurisdiction by illegitimate force is ruinously inefficient in the globalizing economy. The variant of attempting political intimidation by threatening long-range destruction is such a blunt instrument and is so exposed to countervailing threat that it is not a credible policy for a major government. As long as the source of aggressive intent can be located and identified, the basic deterrent effect is not worth contesting and is therefore relatively easy to achieve.
The more serious security danger, moreover, is that emerging from spontaneous social violence and from small-scale but highly destructive threats whose originating source cannot be readily located or identified. The globalizing economy is making access to destructive technology inherently available, as dramatized but only indirectly illustrated by terrorist episodes in
3 Department of Defense, Office of the Secretary of Defense, Nuclear Posture Review briefing, Washington, D.C., September 22, 1994.
Page 68
Tokyo and in Oklahoma City. Small states and substate organizations can acquire weapons of mass destruction and sophisticated means of delivery. The proliferation of highly destructive clandestine threats of this sort could reach unmanageable proportions. So also could the instances of radical internal disintegration such as have occurred in Bosnia, Somalia, Rwanda, Tajikistan, and many other places as well. At the moment, the leading military establishments are poorly prepared to handle this array of problems, and traditional deterrent practices interfere with the intricate collaboration among them that would be required to develop relevant capabilities.
RUSSIAN CASE
The problem of adjusting traditional security commitments to fit the new circumstances clearly weighs most heavily on the Russian military establishment.4Russia, it is important to note, has been the principal victim of the Cold War. For 50 years as the core element of the Soviet Union, it sustained an effort to develop a competitive military establishment in confrontation against all of the major industrial economies. The cost in terms of economic opportunity was tremendous. In the aftermath of that period, Russia has inherited an oversized and unbalanced remnant of the Soviet military establishment poorly suited for its new political and territorial configuration. It faces the problem of relocating and redesigning this inherited establishment while simultaneously undergoing a massive regeneration of its economy, its political system, and indeed its entire society.
In responding to the problem, the Russian military planning system is attempting to preserve a military establishment of more than 1.5 million people. This is the minimum deemed necessary to preserve core nuclear deterrence, to protect against an imaginable conventional ground attack in the Far East and tactical air assaults from the West, and also to cope with flaring episodes of civil violence along their southern border. Though these images of potential threat may appear unlikely to the rest of the world, in the traditional logic of military planning they are at least as plausible as the ones the United States currently uses to set standards for its military deployments. So are the force structure conclusions derived from them.
Those conclusions, however, are wildly unrealistic in economic terms. Russia would have to spend nearly $100 billion per year to sustain a 1.5million-person establishment even if it could produce comparable equipment at half the cost the United States experiences. As prices in the Russian economy adjust to world standards, the full financial requirements of the planned military establishment would exceed $200 billion. The officially enacted defense budget of 40 trillion rubles, nominally comparable to $20 billion at the time it was approved, undoubtedly understates the resources that the Russian defense effort
4 Steinbruner, John. 1995. "Reluctant Strategic Realignment: The Need for a New View of National Security," The Brookings Review (Winter 1995).
Page 69
actually extracts from the economy. But whatever the true amount is, it almost certainly falls well below the minimum sustaining requirement.
There is very little practical prospect that the defense budget will be expanded to meet this requirement. The attempt to do so would threaten the more vital process of economic regeneration and would require a degree of coercive political recentralization that probably has become infeasible but at any rate would be self-defeating. The international reaction to that development would drive the burdens of military preparation to yet more unrealistic levels.
If indefinitely continued, underfinancing of the Russian military establishment will assuredly cause its internal deterioration, and a series of very grave consequences could readily result—the loss of control over large weapons inventories and a destructive interaction with the process of political reform foremost among them. The disintegration of Yugoslavia has provided a chilling hint of what could happen. If that large set of risks is to be avoided, the Russian establishment will probably have to be cut to a level substantially below the current planning aspiration in order to preserve its internal coherence. That in turn requires some very systematic international arrangements to provide reassurance. A coherently planned reduction will not be undertaken if those who are doing it believe the consequence is indefinite exposure to unmanageable external threat.
THE SUBORDINATION AND REVISION OF DETERRENCE
In fact it seems likely that reassurance will eventually emerge as the central security objective of the new era. It is the natural guideline for international security relationships not only between the United States and Russia but also more generally. If no major government is straining to commit aggression or to practice intimidation, then all share a common interest in protecting against these traditional threats as efficiently as possible. To the extent that they can reassure each other in that regard, force deployments, alert levels, and defense budgets can be reduced. Much of the expense and the inherent danger of Cold War forces has been driven by the perceived need to prepare for war on short notice. Beyond that the exercise of reassurance would establish the foundation for close cooperation in responding to instances of spontaneous violence and the potential proliferation of clandestine threats.
At the outset, at any rate, an international security arrangement based on systematic reassurance would necessarily subordinate but presumably not eliminate the practice of deterrence. The legitimacy of preserving a residual deterrent capability would be accepted, but the primary commitment would be to reassuring measures designed to provide convincing indication that nuclear force operations were being restricted to that single legitimate purpose. Abstract as that principle may seem at first glance, its full implementation would involve extensive, indeed revolutionary, changes in prevailing operational practice. Specifically, it would terminate continuous alert operations; it would impose
Page 70
international standards of accounting and physical security on weapons and fissionable materials inventories; and it would formulate agreed restrictions on operational doctrine. The main purpose of all three measures is to set higher standards of operational safety designed to be reassuring both to those being deterred and to those being protected.
Alert Practices
The termination of alert operations would be the most consequential of these measures. As noted, the preparations for rapid retaliation that were developed during the Cold War continue in the aftermath. The number of nuclear weapons currently being maintained on alert status is sufficient to execute a coordinated plan designed to do decisive damage to opposing military forces and their supporting industrial structure. With any serious indication of increased tension, current strategic forces are in the habit of adding additional weapons to their alert forces so as to increase the scale of attack they are immediately prepared to undertake. This operational pattern was developed in order to be sure beyond any practical doubt that a deviously calculated, skillfully concealed attack would not meaningfully degrade the capacity for retaliation. Implicit in the practice is the commitment to respond so quickly that retaliation would be effectively initiated before the initial attack had been completed.5
These alert practices have been accompanied by elaborate physical and procedural measures designed to prevent accidents and unauthorized actions, and they have been fundamentally successful in that regard.6 There has been no hostile or unintended explosion of a nuclear weapon since 1945. The record is replete with incidents that warn of the inherent danger, however, and there are particularly strong reasons for worrying about crisis conditions. The ostensibly good safety record applies primarily to routine peacetime circumstances. The few occasions when nuclear weapons were maneuvered in response to crisis circumstances have produced some unsettling episodes.7 The infrequency of crisis experience is merciful, but it means that there has been little opportunity for the discovery and correction of managerial errors under the conditions when they are most likely to occur. It also means that the empirical base provided by historical experience is not adequate to derive a comprehensive measure of safety. In particular there is reason to believe that the Cold War procedures
5Blair, Bruce G. 1993. The Logic of Inadvertent Nuclear War, Brookings Institution, Washington, D.C.
6Carter, Ashton B., John D. Steinbruner, and Charles A. Zraket, eds. 1987. Managing Nuclear Operations, Brookings Institution, Washington D.C.
7 Sagan, Scott. 1993. The Limits of Safety: Organizations, Accidents, and Nuclear Weapons, Princeton University Press, Princeton N.J.
Page 71
involve a meaningful risk of an inadvertent war arising out of the interaction of opposing alert procedures.8
In the new strategic situation, there is no reason to accept this risk. In principle, the legitimate deterrent effect can be preserved, essentially undiminished, with much higher standards of safety if physical and procedural measures are introduced to assure that no weapon is immediately available for use and that the process of preparing a weapon for use would reliably provide international warning that it is occurring. The most direct way of accomplishing that is to separate warheads from delivery vehicles or to otherwise configure the operating systems short of full readiness with international monitoring procedures to verify that condition. To do this in a manner that did not open up any good possibility for an effectively concealed initial attack would involve some substantial problems of technical design and would require systematic collaboration among all of the countries deploying nuclear weapons. At least in technical terms, however, there is no reason to believe that standard would be more difficult to achieve than the current rapid reaction posture. Deactivating nuclear weapons is a major part of the agenda for reassurance.
The Accounting for and Physical Security of Fissionable Materials
A supplementary part of the agenda concerns the accounting for and physical security of fissionable materials.9 In aggregate, the five states that explicitly developed nuclear weapons during the course of the Cold War produced hundreds of metric tons of plutonium and highly enriched uranium and fabricated nearly 100,000 nuclear weapons out of this material. Very exacting standards of accounting and physical protection were developed for the weapons themselves, but the same standards were not extended to the byproducts of the effort. Moreover, no provisions were made for disposition of the fissionable material other than incorporating it in weapons or holding it in reserve for that purpose. The critical isotopes will be suitable for weapons application for spans of time ranging from tens of thousands of years to hundreds of millions of years, and their radioactive decay products will be a severe health hazard for those durations. Ultimately some acceptable method of disposition will have to be devised. This material cannot be deployed as weapons or stored in its current sites for as long as it will remain dangerous.
The issue of ultimate disposition is being immediately posed by the release of weapons-grade material from active inventories in the course of implementing the arms control agreements that have recently been concluded by
8 Blair, Bruce G. 1995. "Global Zero Alert," Brookings Occasional Papers, Brookings Institution, Washington, D.C., May.
9 Carnegie Commission on Preventing Deadly Conflict. 1995. "Comprehensive Disclosure of Fissionable Materials: A Suggested Initiative," Discussion Paper, Carnegie Corporation of New York, April. Also, Committee on International Security and Arms Control of the National Academy of Sciences. 1994. Management and Disposition of Excess Weapons Plutonium, National Academy Press, Washington, D.C.
Page 72
the United States and Russia. The two countries will release approximately 100 metric tons of plutonium along with more than 600 metric tons of highly enriched uranium as they carry out the provisions of the START I and START II agreements. The uranium can be readily diluted to lower enrichment levels and stored indefinitely without radiological hazard. If it is eventually burned in commercial reactors, it will nonetheless contribute to the further accumulation of plutonium. For plutonium itself, at least in the United States, there is no agreed method of disposition other than holding it in guarded storage. It currently appears likely that the effort to begin the process of disposition will consume a decade or more, and completion of the process is likely to require several decades after it has been initiated.
In the course of attempting to solve that immediate problem, it can be expected that a broader issue will be recognized. More than 1,000 metric tons of plutonium have been produced throughout the world as a by-product of nuclear power generation, with approximately 100 tons of it held in separated form in several different locations. Though this material is of different isotopic composition than the plutonium produced for weapons application, it can be fabricated into nuclear explosives and is virtually as dangerous from that point of view. Under the provisions of the Non-Proliferation Treaty, most of this material is subject to International Atomic Energy Agency safeguards, which are in effect a set of international auditing arrangements overlaid on national accounting systems. Even under the most generous estimate of their effectiveness, it is apparent that plutonium generated in commercial reactors is not subjected to the same managerial standards as weapons-grade material, particularly not the same standards of physical security.
It is prudent to assume that this situation cannot be continued indefinitely without eventually producing a serious breach of control or an actual explosive catastrophe. It is a by-product of the Cold War practice of deterrence that will surely have to be refined under the imperatives of the new era. In order to establish more robust protection against the unmanageable profusion of clandestine threats, the nuclear weapons establishments will have to set more exacting and more comprehensive standards for the accounting and physical protection of fissionable materials. The key to that is making the status of these materials continuously transparent to the international community as a whole— a major revision of traditional practice.
Restraints on Operational Doctrine
The deactivation of deployed weapons and the control of fissionable materials are measures that are directed, as it were, to the hardware of deterrence policy. Under Cold War practice, arms control measures designed to stabilize the interaction of deterrent forces relied primarily on hardware constraints-the number of weapons to be deployed, the number of warheads they could carry, the number of tests allowed, the physical parameters of testing, and such things. These were matters that admitted to exact definition and
Page 73
independent verification. In the adversarial bargaining process of the era, they emerged naturally as the focus of control.
If the practice of deterrence is subordinated to the broader objective of reassurance, then additional measures having to do with the ''software" of deterrence practice become both feasible and desirable. Among the more important of these would be restrictions on operational doctrine.
The Cold War practice of deterrence was bedeviled by some serious tensions involving the doctrine of force operations. The entire concept of deterrence required, in principle, a strict policy of retaliation; but, as a practical matter, if a nuclear war even appeared imminent, there were powerful incentives to initiate it. Largely because it was so difficult to protect command systems from a dedicated assault, a force able to initiate a coordinated attack might fare substantially better than the one slavishly adhering to the rule of retaliation. That realization drove the two principal establishments into their rapid reaction postures and created the possibility that one side in the heat of intense crisis might misjudge the actual imminence of war and might initiate what it considered to be a protective preemptive attack.
In terms of the underlying theory of calculated behavior, this hair-trigger situation did not seem to be unacceptably dangerous. Indeed it could be used to solve another doctrinal dilemma—the fact that the massive threats used in advance of war to present an overwhelmingly stark deterrent threat would be irrational if war were actually to occur. Lest a truly cold-blooded aggressor count on that anomaly to prevent retaliation, the hair-trigger preparations introduced an element of chance that would preserve the deterrent effect. If one considers the fact, however, that no large organization can be absolutely controlled by central calculation, particularly not one whose central authority is inherently vulnerable to preemptive destruction, then the element of chance becomes a danger rather than a comfort.
In order to remove the underlying risk of mistaken or inadvertent preemption, it would be necessary to remove the doctrinal commitment to rapid reaction, and that would be desirable, perhaps also necessary, even if deactivation had been achieved. It would clearly not be wise to set up a situation in which the major forces were programmed to move from a normal state of deactivation directly to one of rapid reaction. To prevent that, some important doctrinal limitations would have to be set—most notably, no targeting of command systems and no completed authorization for retaliation in response to tactical warning. Those measures of reassurance would segregate the deterrent effect from its more dangerous by-products.
During the course of the Cold War, doctrinal limitations of this sort were summarily rejected on grounds that they could not be verified with absolute certainty by independent means. That standard of verification with its presumption of an intent to cheat promoted the confrontational atmosphere of the era and precluded a number of prudent safety measures that would have been highly desirable even if they could not be verified with absolute confidence. If the possibility of cooperative verification is admitted on behalf of
Page 74
the general objective of reassurance, then some very useful mutual constraints on force operations can be introduced with reasonable assurance. Tactical warning can and should be made a collaborative venture. If there were a desire to do so, arrangements could be devised to give controlled access to targeting plans without revealing their full details. If the normal pattern of force operations is made transparent, then nefarious alternatives that might be secreted away are forced to carry the considerable burden of detachment. If military forces are precluded from training for an operation, there is reasonable assurance that they will not attempt to do it.
CONCLUDING PERSPECTIVE
There are two simple conclusions that emerge from this assessment. If the practice of deterrence is liberation from the belligerent political attitudes that originally inspired it, then it can be made a good deal safer than it has historically been. The process of making this adjustment, moreover, is one of the things that must be done in responding to the security imperatives of the emerging era. The determining fact is that any identifiable actor can be readily deterred. It is the impersonal processes and the actors that cannot be identified that we most need to worry about.
