Strategy and Nuclear Deterrence [Course Book ed.] 9781400856121

Citation preview

Strategy and Nuclear Deterrence

Strategy and Nuclear Deterrence EDITED BY

Steven E. Miller

PRINCETON UNIVERSITY PRESS P R I N C E T O N , N E W JERSEY

A N International Security READER

Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom Pnnceton University Press, Guildford, Surrey Selection and preface copyright €> 1984 by Princeton University Press All rights reserved First Princeton Paperback printing, 1984 First Hardcover printing, 1984 LCC 84-42549

ISBN 0-691-04712-x

ISBN 0-691-00597-4 (pbk )

Clothbound editions by Princeton University Press books are printed on acid-free paper, and binding materials are chosen for strength and durability Paperbacks, while satisfactory for personal collec­ tions, are not usually suitable for library rebinding Printed in the United States of America by Princeton University Press, Princeton, New Jersey

The contents of this book were first published in International Security, a publication of The MIT Press under the sponsorship of The Center for Science and International Affairs at Harvard Univer­ sity Copynght in each article is owned jointly by the Massachusetts Institute of Technology and the President and Fellows of Harvard College, and no article may be reproduced m whole or m part ex­ cept with the express written permission of The MIT Press Bernard Brodie, "The Development of Nuclear Strategy," !S 2, no 4 (Spring 1978), Colin Gray, "Nuclear Strategy A Case for a Theory of Victory," IS 4, no 1 (Summer 1979), Robert Jervis, "Deterrence and Perception," IS 7, no 3 (Winter 1982/83), Barry R Posen, "Inadvertent Nuclear War? Escalation and NATO's Northern Flank," IS 7, no 2 (Fall 1982), David Alan Rosenberg, "The Origins of Overkill Nuclear Weapons and American Strategy, 1945-1960," IS 7, no 4 (Spring 1983), Warner R Schilling, "U S Strategic Nuclear Concepts m the 1970s The Search for Suffi­ ciently Equivalent Countervailing Parity," IS 6, no 2 (Fall 1981), Desmond Ball, "U S Strategic Forces How Would They Be Used?" IS 7, no 3 (Winter 1982/83), Walter Slocombe, "The Counter­ vailing Strategy," IS 5, no 4 (Spring 1981), Benjamin Lambeth, "The Political Potential of Equiva­ lence," IS 4, no 2 (Fall 1979), Barry Blechman and Douglas Hart, "The Political Utility of Nuclear Weapons The 1973 Middle East Crisis," IS 7, no 1 (Summer 1982)

Contents

vii

The Contributors Steven E. Miller

ix

The Development of Nuclear Strategy

Bernard Brodie

3

Nuclear Strategy: A Case for a Theory of Victory

Colin Gray

23

Deterrence and Perception

Robert Jervis

57

Inadvertent Nuclear War? Escalation and NATO's Northern Flank

Barry R. Posen

85

Preface DETERRENCE AND DETERRENCE FAILURE

THE EVOLUTION OI AMERICAN STRATEGY

The Origins of Overkill: Nuclear Weapons and American Strategy, 1945-1960

David Alan Rosenberg

U.S. Strategic Nuclear Concepts in the 1970s·. The Search for Sufficiently Equivalent Countervailing Parity

Warner R. Schilling 183

U.S. Strategic Forces: How Would They Be Used?

Desmond Ball

215

The Countervailing Strategy

Walter Slocombe

245

113

THE POLITICAL IMPLICATIONS OF NUCLEAR WEAPONS

The Political Potential of Equivalence

Benjamin S. Lambeth 255

The Political Utility of Nuclear Weapons: The 1973 Middle East Crisis

Barry Blechman and Douglas Hart

273

The Contributors

STEVEN E. MILLER is Managing Editor of International Security and Assistant Director of the Center for Science and International Affairs at the Kennedy School of Government, Harvard University. BERNARD BRODIE was Professor Emeritus of Political Science at the University of

California, Los Angeles, at the time of his death in 1978. COLIN GRAY is President of the National Institute for Public Policy in McLean,

Virginia. ROBERT JERVIS is a Professor in the Political Science Department and the Institute

of War and Peace Studies, Columbia University. BARRY R. POSEN is a Rockefeller Foundation International Affairs Fellow in

residence at the Georgetown University Center for Strategic and International Studies. DAVID ALAN ROSENBERG wrote this article while Assistant Professor of History at

the University of Houston; since September 1983 he has been a Senior Fellow at the Strategic Concepts Development Center at the National Defense University. WARNER x. SCHILLING is the James T. Shotwell Professor of International

Relations and the Director of the Institute of War and Peace Studies at Columbia University. DESMOND BALL is a Fellow at the Strategic and Defense Studies Centre of the Australian National University. WALTER SLOCOMBE, formerly Deputy Undersecretary of Defense for Policy Planning, is now a member of the Washington, D.C., law firm of CapIin and Drysdale. BENJAMIN S. LAMBETH is a Senior Staff Member of the Rand Corporation. BARRY BLECHMAN is a Senior Fellow at the Georgetown University Center for Strategic and International Studies. DOUGLAS HART is a Defense Analyst with the Pacific-Sierra Research Corporation

in Arlington, Virginia.

Pr^jfj-JCG

Steven Ε. Miller

For nearly forty years now, scholars and policymakers have been struggling to cope with the political and military implications of nuclear weapons. The vast and incontestable destructiveness of these weapons has compelled the realization that their use on any large scale would bring untold horrors to both sides in a war. Acting upon this realization, statesmen and strategists have sought strategies that respond to the imperative to avoid the use of nuclear weapons. But there is no agreement on what configuration of nuclear forces best minimizes the likelihood of nuclear war. Nor can there be certainty that any strategy or doctrine can forever avert nuclear war. Moreover, the dangers and risks that attend nuclear threats neither deprive these weapons of political import nor prevent statesmen from seeking to manipulate nuclear risks for political gain. The essays collected in this volume address these several components of the nuclear dilemma that challenges modern statecraft. The effort to prevent the use of nuclear weapons has been dominated by the concept of deterrence. It is not an idea peculiar to the nuclear age nor is its appli­ cability limited to the nuclear balance. But it has assumed special centrality in the nuclear context. Indeed, although recent scholarship has called into question the adequacy of deterrence as a theory of state behavior in international politics (ques­ tioning its assumption of rationality in decision-making, its requirement for clear signalling of commitments between states, and its focus on perceptions of capability and resolve as prime determinants of conflict), issues relating to the requirements of deterrence remain at the heart of the nuclear debate. At base, deterrence in the nuclear context has a pristine simplicity: the threat of nuclear attack is negated by the possibility of nuclear reprisal, of retaliation in kind. This leads in rather straightforward fashion to the need for nuclear systems capable of surviving a nuclear surprise attack and performing the retaliatory mission. Vir­ tually all, of whatever political stripe, agree that this is the minimum condition for deterring an adversary in the nuclear age. But although this minimum condition has long since been met by both superpowers, there persists a highly contentious debate about the requirements of deterrence, for many important questions remain unanswerable in any definitive way: How large does the retaliatory threat need to be? Does the character of the retaliatory threat—counterforce, countervalue, limited, unlimited—affect the quality of deterrence? Is the certainty of retaliation necessary for deterrence, or will the mere possibility be sufficient? And if deterrence requires that one threaten assets that one's adversary values, what is it that the Soviet Union values and what does that mean for U.S. nuclear strategy? Disputants in the American strategic debate offer many and varied answers to these questions, a variety made possible by the dearth of direct evidence bearing

Strategy and Nuclear Deterrence | χ

on such matters and by the inevitably speculative nature of any effort to address such questions. As Bernard Brodie writes in the essay included in this volume, the strategic debate is to a large extent a contest of "conflicting intuitions." Despite the many variations on these arguments, however, there are two basic answers to the question "What deters?" In the first section of this volume, Bernard Brodie and Colin Gray provide examples of each: Gray insists that nuclear warfighting is necessary for deterrence; Brodie argues that it is not. Brodie's own acute intuitions on these matters made him not only the first, but also one of the foremost, of the civilian strategists, and made his an influential voice on issues of strategy and defense for more than three decades. In "The Development of Nuclear Strategy," published shortly before his death in 1978, Brodie was moved by what he saw as troubling developments in American thinking about nuclear war to comment again on issues he first tackled in his famous essay in The Absolute Weapon in 1946. In particular, he argued the unavoidability of a deterrent posture in a world of thousands of survivable nuclear warheads and dismissed as meaningless those strategies that go by the name warfighting or war-winning. Brodie viewed with extreme skepticism those who found ever more demanding requirements for deterrence. To the end of his life he believed that the assured destruction capability possessed by American nuclear forces was adequate for deterrence. In stark contrast, Colin Gray, in "Nuclear Strategy: A Case for a Theory of Victory," repudiates the notion of assured destruction (whose tenets he describes as "a sickly breed") and advocates instead that the notion of defeating the Soviet Union (rather than simply threatening it with damage as a result of some unspecified nuclear reprisal) be incorporated into U.S. nuclear strategy. In Gray's view, this requires that far more attention be paid to operational issues of actually fighting a nuclear war, that political and military targets be included in U.S. nuclear planning, and that the U. S. vigorously pursue active and passive defenses in order to reduce its vulnerability to nuclear attack. Such a warfighting posture, Gray argues, would be the most effective deterrent because it would confront the Soviet Union with the prospect of defeat. Those interested in deterrence have concerned themselves not only with how deterrence works but also with how it might fail. The essays by Robert Jervis and Barry Posen analyze two paths to the breakdown of deterrence. Jervis examines the role of perceptions in deterrence and explores the ways that misperception might lead to deterrence failure. He notes that not only are there many obstacles to accurate perceptions that derive from cognitive and emotional tricks that human beings play on themselves in order to avoid having to change their beliefs or confront unpleasant realities, but it can be very difficult to properly assess an opponent's values, to

Preface | xi

gauge the credibility of threats, and to fully appreciate an adversary's perception of his own choices. Since deterrence depends upon these sorts of perceptions, misjudgment can lead to deterrence failure. Posen examines a very different potential threat to deterrence, providing a detailed analysis of the dynamic that could cause conventional war to spill over into nuclear war as a consequence of the powerful—and possibly uncontrollable—escalatory pressures that would inhere in any large-scale East-West conventional conflict. This is an extremely important discussion, since most analysts agree that escalation from conventional war is one of the more likely paths to nuclear war. Especially pro­ vocative is Posen's argument that the normal operational requirements of conven­ tional forces in wartime could so jeopardize the survivability of strategic nuclear assets that political leaders might be unable to resist pressures to escalate to use of nuclear weapons, even if they wished to avoid escalation, because they could not long risk the degradation of their nuclear retaliatory force. If, as Posen suggests, inadvertent escalation is a real danger, then the notion of an extended phase of conventional conflict may be a fatal illusion. Turning from abstract analysis to the reality of American nuclear policy, the four essays in the second section provide a brief but fairly thorough account of the evolution of U.S. nuclear strategy from the dropping of the bombs on Japan in 1945 up through the most recent doctrinal controversy over the countervailing strategy embodied in PD-59. For the early years of the nuclear age, a large volume of archival materials is becoming available. These were exploited by David Alan Rosenberg in the writing of "The Origins of Overkill," which uses documentary evidence to reconstruct American nuclear weapons policy in the years between 1945 and the creation of the first integrated nuclear war plan (SIOP) in 1960. Rosenberg's analysis revises our understanding of this early period, for it reveals the interest of the Strategic Air Command and other nuclear war planners in striking first, in preventive as well as preemptive nuclear war, and in attacking Soviet nuclear and other military forces. Many had previously believed that American nuclear doctrine had not included these elements. Rosenberg also traces the bureaucratic and political struggles that accompanied and helped shape the effort to organize the nuclear war planning process. Many features of the current organizational arrangement for dealing with nuclear weapons policy have their origins in this earlier period. Rosenberg's account of the origins of U.S. nuclear strategy is followed by Warner Schilling's discussion of the subsequent evolution of American strategy through the end of the 1970s. "U.S. Strategic Nuclear Concepts in the 1970s" focuses on the effect that the emergence of parity in the late 1960s and early 1970s had on American thinking about nuclear war. Schilling analyzes the implications of the

Strategy and Nuclear Deterrence | xii

buildup of Soviet nuclear forces for American notions such as assured destruction and flexible options, and notes the appearance of such ideas as equivalence and sufficiency as Americans attempted to come to grips with the increase in Soviet capability. Finally, he examines the conditions of parity and explores whether there are any feasible alternatives to it. Desmond Ball's essay, "U.S. Strategic Forces: How Would They Be Used?," is a reaction to the uproar that followed the signing by President Carter in July 1980 of Presidential Directive (PD)-59, which was reported to be a major change in U.S. nuclear strategy. Ball argues that far too much attention is accorded to shifts in declaratory policy (that is, in what political leaders say about nuclear weapons policy) and that, as a result, the substantial continuity to be found in operational policy (that is, in how the U. S. actually plans to use nuclear weapons) is neglected. Ball's work addresses the latter set of concerns, and he examines questions relating to force employment and targeting, war termination, and limited nuclear options. On the basis of his study of operational issues, Ball is skeptical of claims that major strategic innovations have taken place, and he suggests that more careful attention to operational constraints would profitably temper the unreality he detects in aspects of declaratory policy. Accompanying the Ball essay in treating doctrinal controversies of the early 1980s is Walter Slocombe's authoritative account of the origins and meaning of the coun­ tervailing strategy contained in PD-59. Slocombe, himself a participant (as Deputy Undersecretary of Defense for Policy Planning) in the process that resulted in PD59, agrees with Ball that it did not represent a dramatic shift in U.S. nuclear strategy, but argues that it was nevertheless a sensible refinement of strategy in light of the evolving strategic situation. It was intended to enhance the deterrent posture by providing additional flexibility in nuclear war plans so that American decision­ makers have limited options that enable them to avoid the choice of suicide or surrender in response to Soviet attack. It also sought to diminish Soviet incentives to use nuclear weapons by explicitly threatening to attack targets in the Soviet Union thought to be highly valued by Soviet leadership, including the political and military leadership itself and the apparatus of control that supports it. These ad­ justments of strategy, Slocombe suggests, bolster the credibility of America's re­ taliatory capability and thereby serve the cause of deterrence. The third section turns to the role of nuclear weapons in foreign policy. The two essays included here seek to assess the impact of nuclear weapons on the perceptions and the behavior of states, to determine what sort of political shadow is cast by these frightful weapons. Benjamin Lambeth examines the Soviet achievement of parity, and speculates on the constraining and liberating effects that this could have

Preface | xiii

on American and Soviet diplomacy respectively. Lambeth also suggests that the USSR may not be content with mere parity, and may pursue strategic advantage because Soviet leaders believe it could provide coercive possibilities in crisis and would improve Soviet chances of prevailing should nuclear war occur. Blechman and Hart, on the other hand, focus on the American side of the equation and seek to illuminate the political role of nuclear weapons through the careful scrutiny of the American nuclear alert during the 1973 Middle East war. Prompted by a Soviet threat to intervene directly in the war, the alert explicitly introduced the nuclear specter into a situation in which a superpower confrontation seemed to be brewing. Blechman and Hart discuss the several purposes that were served by the alert, forempst of which was to signal America's high stakes in the situation, to emphasize the gravity of the crisis, and—of course—to avert Soviet intervention. In the event, the latter goal was achieved but not before the risk of nuclear confrontation had been raised uncomfortably high. This selection of ten articles, then, surveys a number of the central analytical, historical, and political issues associated with nuclear weapons and provides a rep­ resentative sampling of the perspectives that a diverse group of specialists bring to these very controversial questions. Since nuclear weapons remain at the forefront of public concern and the disputes about their use and control are, if anything, intensifying, the nuclear debate seems likely to remain a prominent feature of the American political scene. This volume will not resolve these questions, but we hope it will illuminate the terrain of debate.

Strategy and Nuclear Deterrence

The Development of Nuclear Strategy

Bernard Brodie

That concept was put forward almost at once at the beginning of the nuclear age that is still the dominant concept of nuclear strategy—deterrence. It fell to me—few other civilians at the time were interested in military strategy—to publish the first analytical paper on the military implications of nuclear weapons. Entitled "The Atomic Bomb and American Security," it appeared in the autumn of 1945 as No. 18 of the occa­ sional papers of what was then the Yale Institute for International Studies. In expanded form it was included as two chapters in a book published in the following year under the title The Absolute Weapon, which contained also essays on political implications by four of my Yale colleagues. 1 I should like to cite one brief paragraph from that 1946 book, partly because it has recently been quoted by a number of other writers, usually with approval but in one conspicuous instance with strong disapproval: Thus, the first and most vital step in any American security program for the age of atomic bombs is to take measures to guarantee to ourselves in case of attack the possibility of retaliation in kind. The writer in making that statement is not for the moment concerned about who will win the next war in which atomic bombs are used. Thus far the chief purpose of our military establishment has been to win wars. From now on its chief purpose must be to avert them. It can have almost no other useful purpose. 2 It was obvious then as now that this description of deterrence applied mostly to a war with the only other superpower, the Soviet Union, who did not yet have nuclear weapons but was confidently predicted in the same book to be able "to produce them in quantity within a period of five to ten years." 3 Let me mention a few more points in that 1946 essay in order to indicate what any reflective observer of the time would have found more or less self-evident. It stated that among the requirements for deterrence were extraordinary measures

1. Bernard Brodie, ed., The Absolute Weapon (New York- Harcourt, Brace, 1946). Co-authors were Frederick S. Dunn, Arnold Wolfers, Percy E Corbett, and William T. R. Fox. 2. Ibid., p. 76. The Brodie chapters comprise pp. 21-110 incl. 3. Ibid., pp. 63-69.

This article is based on an address given to the last Plenary Session of the National Conference, Inter-University Seminar on Armed Forces and Society, University of Chicago.

Strategy and Nuclear Deterrence \ 4

of protection for the retaliatory force so that it might survive a surprise attack, that margins of superiority in nuclear weapons or the means of delivering them might count for little or nothing in a crisis so long as each side had reason to fear the huge devastation of its peoples and territories by the other, that while it was possible that the world might see another major war in which the nuclear bomb is not used, the shadow of that bomb would nevertheless "so govern the strategic and tactical dispositions of either side as to create a wholly novel form of war,"4 and that this latter fact had particular implications for the uses of sea power, the classic functions of which depended on an intact home base and the passage of considerable time. It was also observed that while the idea of deterrence per se was certainly nothing new, being as old as the use of physical force, what was distinctively new was the degree to which it was intolerable that it should fail. On the other hand, one could add that "in no case is the fear of the consequences of atomic bomb attack likely to be low,"s which made it radically different from a past in which governments could, often correctly, anti­ cipate wars that would bring them considerable political benefits while exacting very little in the way of costs. Since 1946 there has been much useful rumination and writing on nuclear strategy and especially on the nature of deterrence, but the national debates on the subject have revolved mostly around three questions, all relating directly to the issue of expenditures. These three questions are: 1) What are the changing physical requirements for the continuing success of deterrence? 2) Just what kinds of wars does nuclear deterrence really deter? and 3) What is the role, if any, for tactical nuclear weapons? Far down the course in terms of the public attention accorded it is a fourth question: If deterrence fails, how do we fight a nuclear war and for what objectives? The latter question has been almost totally neglected by civilian scholars, though lately some old ideas have been revived having to do with what are called limited nuclear options. Otherwise most ques­ tions about the actual use of nuclear weapons in war, whether strategic or tac­ tical, have been largely left to the military, who had to shoulder responsibility for picking specific targets, especially in the strategic category, and who were expected to give guidance about the kinds and numbers of nuclear weapons re­ quired. In that connection, one must stress a point which certain young historians

4. 5.

I b i d , , p. I b i d . , p.

83. 80.

Development of Nuclear Strategy | 5

who are new to the field have found it difficult to grasp. Virtually all the basic ideas and philosophies about nuclear weapons and their use have been gen­ erated by civilians working quite independently of the military, even though some resided in institutions like Rand which were largely supported by one or another of the services. In these matters the military have been, with no signifi­ cant exceptions, strictly consumers, naturally showing preference for some ideas over others but hardly otherwise affecting the flow of those ideas. Whatever the reasons, they must include prominently the fact that to the military man deter­ rence comes as the by-product, not the central theme, of his strategic structure. Any philosophy which puts it at the heart of the matter must be uncongenial to him. One military writer significantly speaks of the deterrence-oriented "mod­ ernist" as dwelling "in the realm of achieving non-events in a condition where the flow of events is guided, not by his initiatives, but by other minds." And further: "The obvious difficulty with deterrent theory . . . is the yielding of the initiative to the adversary." In the preceding sentence initiative has already been called the sine qua non of success. 8

The Requirements for Deterrence How does one preserve against surprise attack enough of one's retaliatory force so that the opponent, in anticipation thereof, is deterred? Obviously there is a political dimension to this question, because the need for precautionary measures does vary according to whether or not we think the opponent is straining at the leash to destroy us. There used to be current a notion that if the opponent saw his way clear to destroy us without suffering too much damage in return, that fact alone must impel him to do it. But whether or not that view was ever correct, which I doubt, it is not likely ever to be a feasible option for him. Still, to pre­ pare against all possible crises in the future, it is desirable to minimize that proportion of our retaliatory forces which the opponent can have high confi­ dence of destroying by a surprise blow and to help keep alive in his mind full awareness of the penalties for miscalculation. At the beginning, the United States had a period of grace when it did not have to worry about enemy nuclear attack. However, conditions changed with the passing years, but the first sharp public reminder that we had an important

6. Col. Richard L Curl, "Strategic Doctrine in the Nuclear Age," Strategic R e v i e w (U.S. Stra­ tegic Institute, Washington, D.C.), Winter 1975, p. 48

Strategy and Nuclear Deterrence | 6

vulnerability problem came only with the publication early in 1959 of Albert Wohlstetter's well-known article, "The Delicate Balance of Terror." 7 What is not generally known is that this article was precipitated by Wohlstetter's frustra­ tion with the Air Force. He had been leading for over a year a large research project at Rand which had been trying to find the best means of protecting our bomber aircraft against surprise attack. After considering various alternatives, including the "airborne alert" favored by the Air Force (which was itself second choice for them to the Douhetian idea of striking at the enemy before he gets off the ground), the project group decided that the most cost-effective defense of bombers against surprise attack was a slightly-below-ground concrete shelter for each aircraft. This solution the Air Force vehemently rejected, invoking slo­ gans which identified concrete with the Maginot Line and with excessive defense-mindedness. Wohlstetter's article was thus an appeal to the public, which by its response showed itself both surprised and alarmed at the situation he de­ picted—the more so as his elegant use of the facts and figures at his fingertips lent persuasiveness to his message. However, that problem too passed without ever being resolved, because the article appeared on the eve of the coming of the ICBM, which lent itself to being put underground without controversy, and not far behind was the Polaris submarine. Wohlstetter had been concerned with defending bombers, mostly against bomber attack. We still have bombers as one of the legs of our so-called triad, and now these have to fear enemy missile attack. The Air Force still has no shelters for these bombers and does not contemplate any. Obviously, it still relies on their being in the air when the enemy attack arrives. In fact, on the often-mentioned grounds that they can be sent off early because they are recall­ able, our bombers are frequently projected as virtually a non-vulnerable retalia­ tory force. Well, perhaps they are, if one knows just how to read and respond to various types of ambiguous warning. The problem is not only not to send them off so late but also not to send them off too early. However, I do support fully the belief implicit in the Air Force position that some kind of political warning will always be available. Attack out of the blue, which is to say without a condition of crisis, is one of those worst-case fantasies that we have to cope with as a starting point for our security planning, but there are very good reasons why it has never happened historically, at least in mod-

7. Foreign Affairs, Vol. 37, No. 2, January 1959, pp. 211-256.

Development of Nuclear Strategy | 7

em times, and iox comparable reasons I regard it as so improbable for a nuclear age as to approach virtual certainty that it will not happen, which is to say it is not a possibility worth spending much money on. For similar reasons, I must add before leaving the Wohlstetter article that I could never accept the implications of his title—that the balance of terror be­ tween the Soviet Union and the United States ever has been or ever could be "delicate." My reasons have to do mostly with human inhibitions against tak­ ing monumental risks or doing things which are universally detested, except under motivations far more compelling than those suggested by Wohlstetter in his article. This point is more relevant today than ever before because of the numbers and variety of the American forces that an enemy would need to have a high certitude of destroying in one fell swoop. The numbers of those forces, incidentally, grew during the nineteen sixties like the British Empire was said to have grown—in a series of fits of absentmindedness. There are reasons why the number 1000 was chosen rather than a lesser number for our Minuteman missiles, in addition to our 54 Titans, and also why we chose to build 41 Polaris-Poseidon submarines capable of firing 16 missiles each, in addition to the 400 plus B-52s we had at the time, not to mention the quick reaction alert forces we have long had in Europe. But what­ ever those reasons were, they were not a response to Soviet figures. They in fact gave us, or rather continued, an overwhelming superiority, later increased by the application of the MIRV system to over half our Minuteman missiles, and three-quarters of our submarines. We are still today far superior in numbers of strategic warheads, and we also have a marked advantage in the important factor of accuracy. In the mid-sixties the United States defense community could look with satisfaction on our immense superiority in retaliatory forces which appeared also utterly secure by virtue of being for the most part either underground or under­ water. However, our restless research efforts were already tending to undermine that stable and comfortable situation by pushing to fruition the most incredible advances in ballistic missile accuracy, even without terminal guidance. Those advances resulted partly from developments in micro-electronics circuitry, which made it feasible to put complex computers aboard missiles and to integrate them with hypersensitive inertial-guidance gyros. Reliable unclassified estimates put the Circular Error Probabilities of today's American ICBMs at well under 300 yards, which is utterly fantastic for an object being hurled some 4000 miles. When that kind of accuracy is combined with MIRV, the silo-emplaced

Strategy and Nuclear Deterrence | 8

ICBM begins to be at risk. To be sure, the United States has continued to lead in these developments by a wide margin, and our accuracy does not imperil our silos, but the long lead times that some systems require make it appear provi­ dent, usually, to anticipate the opponent's canceling out some specific techno­ logical advantage of ours. Meanwhile, new discoveries on the effects of X-rays above the atmosphere had rekindled interest in the development of an ABM system. Progress was pro­ ceeding towards the system originally called "Sentinel," which President Nixon was to rename "Safeguard," when the developments in missile accuracy seemed suddenly to give it an important purpose. Up to then it had been a system in search of a mission. Inasmuch as most of its advocates admitted that the ABM could not reliably defend cities, interest focused on its use in defending hard-point targets, which is to say missile silos. Then there arose the great ABM controversy, an amazing chapter in our story because of the intense pas­ sions engendered among the adversaries. Moreover, unlike certain other con­ troversies, it was not a debate between the informed and the ignorant. There was plenty of technological sophistication on both sides. My objectivity in this article will not, I hope, be utterly compromised if I admit that I could never fully understand the pro-ABM position. It always seemed to me that in the expensive and fixed "Safeguard" mode readied for adoption— and later actually deployed at one site in North Dakota—the ABM could in prin­ ciple .be defeated in a number of ways that were already on the horizon, includ­ ing hardening the reentry vehicle against the X-rays of the Spartan missile warhead, cheaply multiplying the number of reentry vehicles as is done with MIRV, adopting terminal avoidance maneuvering in the RV, or even, if all else failed, abandoning the long-range ballistic missile in favor of the cruise mis­ sile, against which the radar of "Safeguard" would be ineffective. Now, in retrospect, we can add that it is most questionable whether American ICBMs needed any such protection at the time that "Safeguard" with its complex but fixed technology was being readied for deployment, or whether they will need it for a long time to come, if ever. Only about 23 percent of U.S. strategic nuclear warheads are carried in ICBMs as compared with about two-thirds for the Soviet Union. Our silos have already been super-hardened. An attack on our ICBMs would involve enormous timing problems for the Soviet Union, especially with most of their ICBMs still of the liquid-fuel variety. No one yet knows what kinds of fratricidal problems will arise among RVs detonating near each other within a short space of time. It would be an optimistic Soviet planner

Development of Nuclear Strategy | 9

who did not count on some two or three hundred of our ICBMs surviving even a well-concerted attack from much more accurate Soviet missiles than they have today, and that says nothing about the 7 to 8 thousand warheads in the other two legs of OUT triad. Surely no sensible opponent would try to eliminate our ICBMs in an initial attack unless he believed that he could at the same time with high confidence eliminate by far the major portions of our other retaliatory forces. One important meaning of the triad, in other words, is that each of the legs helps protect the other two. Besides, how would the Russians know that we would not launch ouf ICBMs on tactical warning, or at least during an attack which could hardly be simultaneous? In short, the utility of the ABM over the long term was at best dubious, but surely it was desirable to avoid deploy­ ing a technologically-fixed system well before it was really needed. Anyway, our Safeguard ABM was effectively cancelled by the agreements of SALT I, incidentally demonstrating that the greater utility of arms control agree­ ments lies not in enhancing our security, which is usually beyond their power, but in helping to save both sides from wasteful expenditures. That brings us to where we are now. The B-I bomber has effectively been dispatched, but looming over the horizon is the monumentally costly MX mis­ sile. Each would be well over twice the weight of the Minuteman III and would carry up to 14 large nuclear warheads. The cost per missile in a 200 to 250 missile program would be some $100 million. Although the Air Force is not advocating the MX because of the alleged vulnerability of Minuteman, it seems reasonable that so costly a missile in which so much power is concentrated be put into a mobile configuration, though that nearly doubles the cost of the system. Whether we need it in addition to Trident, which expands the operating area of our stra­ tegic submarines from about 2.5 million square miles to about 40 million square miles, and in addition also to the cruise missile, is not going to be deter­ mined on any realistic consideration of vulnerability. Protagonists and oppo­ nents will take their positions rather on political grounds, that is, depending on whether they harbor such views as incline them towards the Committee on the Present Danger or whether they live on lower levels of anxiety. HOW MUCH POWER IS ENOUGH XO DETER?

This question has always been with us in one form or another, but it is amaz­ ing to find it raised anew with intensified vigor just now when the number of strategic nuclear warheads in the American arsenal must be on the order of at least 9 to 10 thousand. Nevertheless, we have anguished statements like that

Strategy and Nuclear Deterrence | 10

of Richard Pipes. Mr. Pipes, Professor of Russian History at Harvard University, compels our attention because he was the chairman of that Team B selected last year at the prodding of Major General George Keegan to straighten out the relevant groups at the CIA, whose interpretation of Soviet intentions had apparently been too relaxed for some tastes. His recent Commentary article, which bears the title "Why the Soviet Union Thinks It Could Fight and Win a Nuclear War," 8 was circulated in reprint form by the Committee on the Present Danger, of which Professor Pipes is a charter member, and also by the like-minded Na­ tional Strategy Information Center. The "why" in the title of Mr. Pipes' article preempts the prior question whether some entity called the Soviet Union thinks as he says it does. The appropriate question is: who in the Soviet Union thinks they can fight and win a nuclear war? The article tells us that it is some Soviet generals who think so, not a single political leader being mentioned. One could at this point dismiss the issue by remarking that there are also plenty of U.S. generals who think that the United States could fight and win a nuclear war and are even willing to give a definition for the word "win," though few of us would be comfortable with that defini­ tion. Still, it is interesting to notice what kind of evidence Mr. Pipes adduces for his thesis. A prime piece of evidence for him is that Russian generals believe that "the basic function of warfare, as defined by Clausewitz, remains permanently valid," and he quotes the late Marshal Sokolovskii to that effect, with his own added emphasis: "It is well known that the essential nature of war as a continuation of politics does not change with changing technology and armament." 9 It is obvious that Professor Pipes reads into that famous dictum of Clausewitz he is alluding to a meaning quite different from that which Clausewitz intended. That might leave open the question what Marshal SokoIovskii intended, except that we know from other portions of his work that Sokolovskii did study Clause­ witz, perhaps even with some care. Clausewitz's meaning, which is in fact basic to everything that one thinks about nuclear deterrence, is developed at length in Books I and VIII of On War. In essence it amounts to the idea that war would be only senseless destruction if it were not in pursuit of some valid political

8. Commentary, Vol. 64, No. 1, July 1977, 21-34. See also the several "letters to the editor" provoked by this article in the second issue following. Commentary, Vol. 64, No. 3, Sept. 1977, pp. 4-26. 9. Commentary, July 1977, p. 30.

Development of Nuclear Strategy | 11

objective. It is precisely the fact that one finds it difficult if not impossible to find a valid political objective that would justify the destruction inevitable in a strategic nuclear exchange that makes the whole concept of nuclear deterrence credible. In short, far from finding Marshal Sokolovskii's quoted statement omi­ nously, we should all be able to agree with it with a fervent "amen." Mr. Pipes takes exception to my thirty-year-old statement on the importance of deterring nuclear wars rather than winning them. He does so on the grounds that it has had a generally bad effect on American strategic thinking. I wonder how he would change it? Pipes and his many like-minded associates imply, or some may even state, that the Soviet Union has interests conflicting with ours which its leaders are so unwilling to inhibit or moderate that they are capable of coldly contemplating waging aggressive war against us. However, they fall short of telling us what those inexorable interests might be, usually falling back on the old standby, "miscalculation." Miscalculation in a crisis is indeed some­ thing to be concerned about, but surely not for the sake of circumventing deter­ rence but for making it work. General Keegan appears to think that the Sovet leaders simply want to elimi­ nate the United States as a superpower rival, and he has some absurd notions about how their allegedly heroic efforts in civil defense have reduced to a re­ markably low level the human casualties they would suffer, though it is still reckoned in millions. However, even he has to admit they would sacrifice vir­ tually all their capital investment above ground, not to mention the amount of fallout they would have to cope with in the environment. Others, including Mr. Pipes, point to the 20 to 30 million Russian lives lost during World War II, as though that had been a matter of choice among the Soviet leaders at the time —a sacrifice willingly offered up as a maneuver to entice the Nazis to their de­ struction. Mr. Paul Nitze, another charter member of the Committee on the Present Dan­ ger, offers us a scenario in which the Soviet Union delivers a surprise attack which does not, to be sure, eliminate more than a portion of our retaliatory forces but which leaves us so inferior that the President, whoever he is at the time, elects to quit the fight before making any reply in kind. 10 Thus, the Soviet Union succeeds in making that otherwise elusive first-strike-with-impunity! An interesting thought, but it would take an exceedingly venturesome and also

10. Paul H. Nitze, "Deterring our Deterrent/' Foreign Policy, No. 25, Winter 1976-77, pp. 195-210.

Strategy and Nuclear Deterrence \ 12

foolish Soviet leader to bank on the President's not retaliating. Even Mr. Nitze is not really sure; he says only that he believes the president would not. But one must ask: even if all their evaluations were correct, what would these people have us do? Start a preventive war? Of course not. In the few instances where an attempt at an answer is offered, it seems to run as follows: abandon deterrence strategies in favor of war-winning strategies. But what does that mean? So long as we would not initiate or welcome the outbreak of a war, our basic peacetime strategy would still be that of deterrence. However, a difference is intended. The military have long been fond of saying that the best deter­ rence force is a war-winning force. Such a statement has certainly not been without meaning in the past, and even in the nuclear age it has some meaning in terms of tactical or theater forces. But when we speak of strategic nuclear ex­ changes, it is virtually impossible to find a reasonable meaning except in nega­ tive terms. That is, we can say what a war-winning force is not, or, to be more precise, what the people who use these phrases seem to mean by them. They mean a force which is definitely not inferior in any one of the several meaningful, or allegedly meaningful, attributes by which rival forces are usually compared. The United States has about two-thirds the number of ICBMs deployed by the Soviet Union and is substantially inferior in terms of throw-weight. Actually, since 1970 the United States has built and deployed more ICBMs than the Soviet Union (550 versus 330), the difference being that our newer ones replaced older types which were retired, while the Soviet Union has kept active nearly all the ICBMs it has ever built—which incidentally tells us something about the difference in quality between the two forces. And so far as concerns throw-weight, about which so much is heard in some quarters, let us remember that it was a deliberate choice on the part of the U.S. military some two decades ago to go for smaller ICBMs than the Soviets fancied, for at least three good reasons: first, our people favored the increased readiness that would go with a solid-fuel propellant; second, we knew that our accuracy was much superior to that of the Russians, and likely to remain so for some time; and third, our smaller warheads were really not so very small. We keep forgetting that it was only a 14 KT weapon that devastated Hiroshima. Whether against hard-point targets or any other kind, our retaliatory force with its far greater number of warheads and its much better accuracy and readi­ ness remains today a clearly superior force to that of the Soviet Union—a "warwinning force" if one insists upon it. And if it remains that today, how about the recent past? Where the Committee on the Present Danger in one of its bro-

Development of Nuclear Strategy | 13

chures speaks of "the brutal momentum of the massive Soviet strategic arms build-up—a build-up without precedent in history," 11 it is speaking of some­ thing which no student of the American strategic aims build-up in the sixties could possibly consider unprecedented. Since we are looking so hard for the reasons for the Soviet build-up, one possibility that ought to be considered is that it was simply triggered by ours, and that it continues to be stimulated by a desire to catch up. Against What Do Nuclear Weapons Deter? This question hardly arose at all in the early years of the nuclear age. The dogma of the time was that every modern war is total war, and the only war one seri­ ously thought about was that between the United States and the Soviet Union. This view survived into the first part of the Korean War, which was regarded by the Pentagon as a Soviet ruse to draw us into the western Pacific while they pre­ pared to launch an attack on Europe. Then came the thermonuclear weapon, which caused some people to think about the necessity of separating general war from limited or theater war, and following that came the Eisenhower Adminis­ tration with its somewhat atavistic commitment to massive retaliation. The fa­ mous Dulles speech of January 1954 caused a strong reaction, which finally came into full bloom at the advent of the Kennedy Administration in 1961 with its remarkable Secretary of Defense, Robert S. McNamara. The idea developed that strategic nuclear power deterred only the use of stra­ tegic nuclear power and hardly any other form of violence. On the contrary, some argued that the existence of strategic nuclear weapons made lesser wars without nuclear weapons more rather than less likely, as though the pressure for war was more or less constant and the blockage of it in one direction made it only more insistent to break out in another. Then we began to hear of nuclear weapons being "decoupled" from diplomacy, for obviously they were not going to be used anyway. The next step was to argue that nuclear weapons must not be used in theater warfare even in Europe, and that the way to avoid their use was so to build up our conventional forces—together with our allies—that the thresh­ old for use of tactical nuclear weapons would be raised too high to be breached. Such strong conventional forces, proponents held, made a far better theater de­

ll. Committee on the Present Danger: "Where We Stand on Salt," July 6,1977, p. 5.

Strategy and Nuclear Deterrence \ 14

terrence than could any reliance upon tactical nuclear weapons. These ideas were fervently adopted by President Kennedy and Secretary McNamara, and, after some hesitation, by the armed forces, especially the Army and Navy. After all, large conventional forces meant more of everything, and it also meant the kind of war that the military leaders felt they knew how to fight. Thus, what some sixteen years ago was a novel and even radical idea has long since become the conventional wisdom, especially in the United States. And, whether a good idea or not, it has cost this country a great deal of money. The massive retaliation idea was after all justified by its advocates mostly on grounds of economy, and the more we departed from it the higher the costs were likely to go. Because the alleged advantages of the conventional build-up policy have had a good press and are well known, I shall confine myself to expressing a few op­ posing arguments. THE ROLE FOR TACTICAL NUCLEAR WEAPONS

First, one of the strongest reasons consistently advanced by people like Alain Enthoven for doing away with tactical nuclear weapons is their alleged escalatory effect.12 According to this view, to use one nuclear weapon in the field, how­ ever small, is simply to pull the plug on the use of any and all nuclear weapons. One use thus leads quickly and directly to holocaust. It is interesting that our European partners favor the deployment of tactical nuclear weapons because they entertain the same belief, which means they favor it for the same reason that so many of our people oppose it. In my opinion, both are wrong. There was no problem in distinguishing between tactical and strategic bombing in World War II, and in avoiding the latter where it seemed politically desirable to do so. It is hardly self-evident that the distinction would be more difficult where nuclear weapons are involved, especially if we shift to much smaller tactical nuclear weapons than many that presently exist—which we should want to do anyway. Second, the strategic nuclear forces of each of the superpowers do inhibit the other from any kind of warlike action against it. This was proved abundantly during the Cuban missile crisis, where our side we know, and the other side we

12. Alain C. Enthoven has for some fifteen years been the outstanding spokesman for this school of thought, which has, however, included such prominent thinkers as AJhert Wohlstetter and Thomas C. Schelling. One of Enthoven's most recent articles on the subject was his "U.S. Forces in Europe: How Many? Doing What?/' Foreign Affairs (April 1975), pp. 512-532.

Development of Nuclear Strategy | 15

have good reason to guess, each thought it was looking down the barrel of a strategic nuclear war. It is also noteworthy that our people seemed to derive little comfort from the knowledge of having an overwhelming superiority in nu­ clear arms. 13 It is indeed the fear that escalation is possible, some think probable, that causes the ultimate sanction to have this general deterrent effect, at least between the two superpowers. Granted that we need some military power in the theater so that the opponent knows he could not make a truly aggressive move there without provoking a war, it is nevertheless an area where we can afford to be interested in economies. Actually, our allies have left us no choice in the matter, and it was predictable that they would not. They simply do not see the Soviet Union as threatening to attack. They have therefore firmly and consistently refused, despite long and continued prodding by our government, to build up to anything like the levels demanded by the conventional war thesis. Thus, we do not presently have and are not likely to get a real conventional capability in Europe, that is, one capable of dealing with a deliberate conventional Soviet attack against the West. For that matter, neither do we have a good tactical nuclear capability, but the latter is much more easily attainable. It would mean mostly changes in doctrine, in training, and in types of weapons, but not large net increases in men or equip­ ment. Third and most important, we should recognize that inasmuch as all our war plans are predicated upon an act of major Soviet aggression, the choice would not be really ours to make. The Soviets are hardly likely to enter a duel where they leave so critical a choice of weapons up to us. At any rate, we could not as­ sume in advance that they would leave the choice to us. We are presently com­ mitted to using tactical nuclear weapons if the Russians use them first or if we find ourselves losing without them. The Russians, if they were making a deliber­ ate attack, would refrain from using them, if they refrained at all, only if they were confident they could quickly overwhelm us without them. It is folly to think we could wait to see. Major adjustments in posture and tactics during a fast moving battle are not so easily made, especially if one is in the process of being overwhelmed. Finally, there is virtually nothing in the voluminous open Soviet

13. See especially Robert F. Kennedy, Thirteen Days: A Memoir of the Cuban Missile Crisis, (New York: W. W. Norton, 1969). It is interesting that at no time in his gripping narrative does Kennedy bring up or mention anyone else's bringing up the very clear American superiority in nuclear arms, from which they might have been expected to derive additional hope of de­ terring Soviet resort to war. That, of course, does not mean that nobody thought about it.

Strategy and Nuclear Deterrence | 16

tactical doctrine to support the notion that they make the sharp and important distinctions we do between use and non-use of tactical nuclear weapons. Far from embracing our favored doctrines, they have quite explicitly rejected them. The Russians do not appear to be straining at the leash to pounce upon us in Europe. Quite the contrary. It is therefore an area where our investment need be only relatively modest; yet if we have any substantial force at all it ought to be one that is truly effective as a fighting force and certainly as a deterrent. It is nonsense to hold that a force trained and equipped to fight conventionally— even though it has some essentially unusable nuclear weapons behind it— makes a better deterrent than one of comparable size trained and equipped to fight from the beginning with nuclear weapons designed exclusively for tactical use. By the latter I mean something like the enhanced radiation bomb (the socalled neutron bomb), which should have been presented to the public for what it is—a means of making bombs smaller while retaining optimum effectiveness— and which makes an ideal anti-tank weapon. The Wartime Use of Strategic Nuclear Weapons

I shall confine myself to commenting briefly on the special views identified with James R. Schlesinger as former Secretary of Defense, though they are more accurately described as a revival of some old ideas. 14 In the form offered by Schlesinger, they have been best described and advanced by an enthusiastic ad­ vocate, Benjamin Lambeth. 15 The general idea is to find some options outside what is regarded as the straight-jacketed posture of deterrence, which is viewed as waiting supinely to be struck and then attempting to expend our whole nu­ clear stockpile as rapidly as possible. The alternative suggested by the Schlesinger school is that we should be prepared during a crisis to initiate the use of strategic nuclear weapons, but only a few at a time, perhaps only one or two, the purpose usually mentioned being to show our "determination" or "resolve." In some-

14. An earlier exposition of similar ideas is found in Klaus Knorr and Thornton Read (eds.), Limited Strategic War (New York: Praeger, 1962). 15. Benjamin S. Lambeth, Selective Nuclear O p t i o n s i n American and S o v i e t Strategic Policy, RAND Report R-2034-DDRE, December 1976. Despite my criticisms in the text, I should point out here that Lambeth explicitly disavows belief that circumstances of the future are ever likely to favor American limited first use. He simply wants it adopted as a "standing contingency capability," though his argument endorsing that capability does wax exceedingly enthusiastic. His Report incidentally also includes a brilliant analysis of recent Soviet tactical and strategic thinking.

Development of Nuclear Strategy | 17

what more hushed tones, another purpose is sometimes whispered, one that immediately calls up references to the "surgical scalpel," that is, to get the enemy's top command and control apparatus before the few persons who are the key to it have a chance to go underground. Various people have long questioned the wisdom or even the purpose of plan­ ning to deliver all our nuclear weapons as rapidly as possible in replying to an attack, for many targets are not time-urgent and most of our own weapons are not directly threatened. The main war goal upon the beginning of a strategic nuclear exchange should surely be to terminate it as quickly as possible and with the least amount of damage possible—on both sides. The U.S. military have indeed told us that they take a different view. According to General George Brown, Chairman of the Joint Chiefs of Staff, "What we are doing now is target­ ing a war recovery capability." 18 Their object, in plainer words, is to see to it that in a strategic nuclear exchange the Soviet Union will suffer so much greater damage to its industrial plants and population than we do, that its recovery is much more prolonged. Whatever else may be said about this idea, one would have to go back almost to the fate of Carthage to find an historical precedent. In more recent times the United States has usually put itself to considerable effort and expense to help a defeated enemy recover, normally with the expectation that a vastly changed government in the defeated power will make it adopt an atti­ tude towards the world that is much more congenial with ours. It is inciden­ tally curious, and totally opposed to all the Clausewitzian canons, that so farreaching a military decision should escape any political input, a situation which no doubt reflects the very low priority which our top political leaders automati­ cally put on war fighting as opposed to deterrence concerns. Nevertheless, one has to acknowledge that targeting "recovery capability" may put some premium on rapid expenditure of one's stockpile. However, the proposal of the SchIesinger-Lambeth school is not concerned really with the issue of possibly excessive or excessively speedy retaliation to an attack upon us. Its central concern is with getting out of the retaliatory bind, with opening American considerations to the possibility of striking first in the hope of winning advantages hitherto disregarded or at least insufficiently considered. The idea of keeping to a low level the number of weapons used does sometimes seem like a mere palliative to the central notion that we should be

16. Quoted in The Defense Monitor (Center for Defense Information, Washington, D.C.), Vol. 6, No. 6, August 1977, p. 2.

Strategy and Nuclear Deterrence | 18

prepared to initiate the exchange, though of course the idea that weapons be used for mere warning does require that their numbers be kept low. Otherwise the opponent is sure to do what he is all too likely to do anyway, which is to mis­ apprehend our moderately benign intentions. The Schlesinger-Lambeth school makes much of the advantage of expanding the options of any president during a crisis, a theme which had an especially high rating with Mr. McNamara when he was Secretary of Defense. The notion that it is incontestably good to expand the chief executive's options is rather peculiar. For one thing, it runs directly counter to the basic tenets of constitution­ al government. The long conflict between the Congress and President Nixon during the Vietnam War concerning what seemed to be the limitations laid down by the Constitution upon presidential prerogatives and war-making pow­ ers is a case in point. From both the legal and the pragmatic view, it was not then and it certainly is not now clear that the national interest lay in expanding the President's options. More recently we saw President Ford denied by Congress the right to send a large quantity of munitions to an obviously collapsing South Viet­ namese government, and later denied the option of intervening in Angola with munitions if not with troops. In both instances President Ford and his Secretary of State complained most bitterly and offered up dire warnings of the evils that would befall American interests from such unseemly opposition to their wills. In both instances enough time has passed to suggest that the fears conjured up in those warnings were exaggerated, if not wholly imaginary. In short, the President does not always know best. The above examples refer to legal limitations upon the power of the executive. However they may affect the national interest in any specific instance, the idea that such limitations should exist is considered absolutely basic to democracy as opposed to dictatorship. Of another character, however, are limitations im­ posed by the lack of specific physical capabilities such as might have been pro­ vided in good time if a different outlook had prevailed. In recent times it has generally been taken for granted that the wider the President's range of options in physical means, the better. But why this should be so is also not altogether clear. Actually, one notices that the thesis has usually been defended in terms of extending the range of options downward, as in the arguments referred to above that to increase conventional forces is to reduce or remove the pressure upon the President in particular circumstances to use the nuclear option. Unfor­ tunately, when President Kennedy did expand the conventional ground forces

Development of Nuclear Strategy | 19

for that reason, his successor found himself with the means early in 1965 of sending combat forces to Vietnam without calling up the reserves! His options had been expanded in a way that particularly suited his own nature and out­ look, but it is not obvious that the national interest benefited from the result. Naturally, one should not hinge too much on single cases. President Johnson's conduct does not make a case against having a sufficiency of ground forces. It should, however, be enough to cast doubt on the notion that expanding the Presi­ dent's military options is always a good thing. It may be a good thing if we can always be sure of having wise presidents, but it also throws a heavier burden upon that wisdom, so that any lack of it becomes the more telling. It is an old story that one way of keeping people out of trouble is to deny them the means for getting into it. We have put in the President's hands a huge military power be­ cause we believe that the country's security demands that we do so, and we are obliged to trust that he will use it wisely. But to expand that power simply for the sake of expanding his options is to push hard that obligatory trust. Anyway, if the President did not already have the power to make the kinds of first strikes envisioned in the Schlesinger proposals, which inevitably he does, it would seem a very dubious proposition indeed that we should bestir ourselves to provide it. Certain kinds of weapons, like the cruise missile, may because of their extreme accuracy lend themselves particularly well to the kind of "surgi­ cal strikes" which in some people's fantasies bring victory without cost or even danger, but there had better be other reasons for calling for them than the desire to free the President for conducting such experiments. In fact, the overwhelming odds are that when and if the crisis comes, the man occupying the seat of power in the United States will exercise at least the cau­ tion of a John F. Kennedy during the Cuban missile crisis, who by his brother's intimate account was appalled by the possibility that any precipitous use of physical power by the United States might unleash the nuclear holocaust. 17 He had firmly determined upon a confrontation and had made his requirements plain to the adversary. There is no reason to suppose that President Kennedy was any­ thing but well above the mean level of courage we might expect in any president finding himself in a comparable situation. Yet we know from his conduct that the last thing he was interested in experimenting with during the crisis was some violent means of proving his resolve.

17. See above, Thirteen Days.

Strategy and Nuclear Deterrence | 20

There may indeed be worse crises than that of October 1962, but during such crises the man who must make the ultimate decisions for the United States will likely be searching desperately for the resolve he is supposed to show, and may well be the wiser man for having difficulties in finding it. In short, the notion that in an extremely tense crisis, which may include an ongoing theater war, any useful purpose is likely to be served by firing off strategic nuclear weapons, how­ ever limited in number, seems vastly to underestimate both the risks to the na­ tion in such a course and the burden upon the person who must make the deci­ sion. Divorced from consideration of how human beings actually behave in a crisis, it fits Raymond Aron's definition of "strategic fiction," analogous to "sci­ ence fiction." 18 Where Lambeth argues that the Schlesinger proposals introduce flexibility into an area of thinking hitherto marked by extreme rigidity, and that it introduces also strategy (in the form of choice) where no possibility of strategy existed before, he is simply playing with words. The rigidity lies in the situation, not in the thinking. The difference between war and no war is great enough, but that between strategic thermonuclear war and war as we have known it in the past is certain to be greater still. Any rigidity which keeps us from entering the new horrors or from nibbling at it in the hopes that a nibble will clearly be seen as such by the other side, is a salutary rigidity. And we need not worry whether the choices the President is obliged to make during extremely tense situations fill out anyone's definition of strategy. The important thing is that they be wise choices under the circumstances. It is especially curious that the notions discussed above should be advanced by many who continue to oppose the use of tactical nuclear weapons because of their alleged escalatory danger! That danger must be by orders of magnitude greater when the weapon or weapons detonated are by type and by choice of tar­ get clearly of the strategic variety. Though it would indeed be desirable to condition the people who control our own retaliatory forces not to regard the arrival of one or a few enemy warheads as necessarily the onset of an all-out attack, that kind of conditioning would hardly be dependable even for our own side, and we should certainly not seek to depend upon its existing with suitable firmness on the other. Finally, a word about what one puts on the line for such proposals. The de-

18. Raymond Aron, "The Evolution of Strategic Thought," included in Alastair Buchan (ed.) Problems of Modern Strategy (London: Chatto & Windus, 1970), p. 30.

Development of Nuclear Strategy | 21

fense community of the United States is inhabited by peoples of a wide range of skills and sometimes of considerable imagination. All sorts of notions and propo­ sitions are churned out, and often presented for consideration with the prefatory words: "It is conceivable that..." Such words establish their own truth, for the fact that someone has conceived of whatever proposition follows is enough to establish that it is conceivable. Whether it is worth a second thought, however, is another matter. It should undergo a good deal of thought before one begins to spend much money on it. In defense matters sums spent on particular proposals can easily become huge. Schlesinger's original proposal for "limited nuclear options" seems to have been loosely connected to the then new cruise missile, which fortunately has other factors to recommend it. Last year, however, Mr. Colin S. Gray, in a "Letter to the Editor" of The New York Times, 10 opined that inasmuch as it is now established that the United States will initiate the coming strategic nu­ clear exchange, concerning which he seems to have no doubt, we should proceed at once to buy the MX missile system. Fortunately, Mr. Gray is not a senior official in the Defense Department or the military establishment, but his con­ siderable writings have brought him attention within the defense community. Just why he considers the MX especially suited for initiating strikes is not clear— the cruise missile has at least the advantage of extreme accuracy—but his readi­ ness to spend billions of dollars to cover a contingency that most of his peers would regard as extremely dubious is a little breath-taking. Surely the chances of his prediction being realized, even if more than infinitesimal, are not so great that we could not make-do, if and when the time came, with vehicles deployed for retaliatory purposes. In these matters, to be sure, we are dealing fundamentally with conflicting in­ tuitions. There is no doubt that some people's intuitions are better than others, but the superiority of the former, though sometimes definable and explicable, may be difficult to prove. Still, the thinking up of ingenious new possibilities is deceptively cheap and easy, and the burden of proof must be on those who urge the payment of huge additional premiums for putting their particular notions in­ to practice.

19. The Nem York Times, October 19, 1977.

Nuclear Strategy: the Case for a Theory of Victory

Colin

s

Gray

JP or good or ill, or even perhaps for some of both, 1979 is almost certain to see the most intensive debate over strategic postural and doctrinal issues since the days of the misprojected missile gap back in 1959-60. SALT II is bringing it all together: the state of the balance, predictions of trends, the relevance (or otherwise) of strategic forces to superpower diplomacy, developments in high technology, Soviet intentions and Soviet performance, and the character of a desirable strategic doctrine. The great SALT II debate, when finally joined, will probably cast as much shadow as light because much of the argumentation will avoid reference to truly fundamental issues. Indeed, a similar problem besets the quality of debate over individual weapon and related program questions (i.e., does the United States need a follow-on [to Minuteman III] ICBM, and if so of what kind?—or, does the United States need a civil defense program?—and so forth). Much of the earnest and even occasionally rather vitriolic debate over SALT, the MX-ICBM, cruise missiles, and the like, is almost purely symp­ tomatic of disagreement over basic strategy—indeed, so much so that if attention were to be focused on the latter, then the generic, though not detailed, solutions to the former problems should follow fairly logically. As a somewhat inelegant axiom, this author will argue that a defense community which has not really decided what its strategic force posture is for, has no business either engaging in strategic arms control negotiations, or in passing judgment on the merits of individual weapon systems. A Need for Strategy

Notwithstanding the popular, and indeed official, nomenclature which clas­ sifies our centrally based nuclear launch systems as strategic, the fact remains that there is an acute deficiency of strategic thinking pertaining to those forces. To many people, apparently, it is not at all self-evident that there are any issues of operational strategy relevant to the so-called strategic nuclear forces. Strategic nuclear war, presumably, is deterred by the prospect of the employment of those forces; while, should such a war actually occur, again presumably, each side executes its largely pre-planned sequence of more and more punishing strike options in its Single Integrated Operational Plan

Strategy and Nuclear Deterrence | 24

(SIOP) and then dies with the best grace it can muster. This author has difficulty seeing merit (let alone moral justification) in executing the post­ humous punishment of an adversary's society, possibly to a genocidal level of catastrophic damage, and hence has some difficulty discerning the value of such an option brandished as an intended prewar deterrent.1 Of course, the U.S. Government has not been planning to execute even a rough facsimile of genocide for many years. But official, and even Presidential, language (and perhaps thinking),2 and war planning, have long been recognized to be somewhat different activities. This author is not confusing post-NSDM 242 nuclear weapon employment policy (NUWEP) guidance with assured de­ struction thinking,3 although he believes that both would prove fatal to the U.S. prospect of success in the event of war. In addition, this author does not accept the argument that U.S. war plans are in good order: the real deficiency lies in the strategic forces that have been acquired to attempt to implement them (though there is considerable merit in that argument). Absurd and murderous though mutual assured destruction (MAD) rea­ soning is to a strategic rationalist, one has to admit that the world, perhaps fortunately, is not ruled by strategic rationalists. Readers should be warned that this author does believe that there is a role for strategy—that is, for the sensible, politically directed application of military power in thermonuclear war. However, it is entirely possible that politicians of all creeds and cultures are, and will be, deterred solely by the undifferentiated prospect of nuclear war—which may be translated as meaning the fear of suffering societal punishment at an unacceptable level. Even if one suspects that the politician, 1 The actual execution of SIOP-IeveI attacks upon Soviet population and economic targets, on the canonical scale advertised in the late 1960s, would be either an act of revenge (and without political purpose), or—as initiative—would likely trigger a Soviet response in kind Assured destruction would leave an adversary's (presumably surviving) political leaders with nothing left to lose Prominent among the political weaknesses of assured destruction reasoning is the consideration that just as not all credible threats need deter (if the threat is insufficiently awesome), so not all awesome threats need deter (if they are insufficiently credible) 2 President Carter, in his State of the Union Message for 1979, advertised the "overwhelming" deterrent influence that reposed in only one Poseidon SSBN (nominally bearing 160 reentry vehicles of 40 kt 16 χ 10) The President neglected to mention that although 40 kt warheads could destroy a lot of buildings, it was not obvious that one Poseidon SSBN could accomplish anything very useful by way of forwarding the accomplishment of U S war aims "Transcript of President's State of Union Address to a Joint Session of Congress," The New York Times, January 24, 1979, ρ A 13 3 See U S Congress, House Committee on Armed Services, Full Committee Consideration of Overall National Security Programs and Related Budget Requirements, Hearings, 94th Congress, 1st session (Washington, DC US Government Printing Office, 1975), testimony of Edward Aldndge

Nuclear Strategy | 25

a rank amateur in strategic analysis, will be deterred where a professional strategic analyst would advise that he should not be, there remain good reasons for listening to the cautionary words of the professional. First, however unlikely the possibility from the perspective of American political culture, there could come to power in the Soviet Union a leader, or a group of collegial leaders, who would take an instrumental view of nuclear war. Whether or not such a group already is in place is very much a moot point. It could be profoundly imprudent simply to assume that strategic analysis has no bearing on the likelihood of occurrence of nuclear war. In a political context where a decision to act or not to act was finely balanced, military confidence and promises, or the lack thereof, could have a large influence on the political decision. One of the essential tasks of the American defense community is to help ensure that in moments of acute crisis the Soviet general staff cannot brief the Politburo with a plausible theory of military victory.

Second, it should not be forgotten that an important role for strategic analysis is the underpinning of a strategic doctrine which makes for the orderly management of, and choice between, defense programs.4 If sufficient deterrent effect is believed to repose simply in the undifferentiated threat of nuclear war (of doing a lot of damage in a short space of time), on what basis does one choose what to buy? The essentially arbitrary guidelines for the "required" levels of assured destruction make some sense on this reasoning. (i.e., nobody claims that some "magic fraction" of threatened damage is needed for deterrence—even if annual Posture Statements do lend them­ selves to being misread in that fashion—but, some doctrinal guidance, be­ yond simply doing a lot of damage, is required for the provision of rules of thumb and for the suggestion of appropriate measures of merit.)5 Unfortu­ nately, arbitrary doctrinal guidance for force sizing (and even quality) devised for the convenience of orderly administration tends to acquire an aura of strategic authority that was not originally intended and which it cannot bear.6 4. See Benjamin S. Lambeth, How to Think About Soviet Military Doctrine, P-5939 (Santa Monica, California: Rand, February, 1978), pp. 15-16. 5. On the rationales for the "magic fractions" of damage that permeated assured destruction reasoning in the 1960s, see Alain C. Enthoverv and K. Wayne Smith, How Much is Enough? Shaping the Defense Program, 1961-1969 (New York: Harper and Row, 1971), Chapters 5-6. 6. NATO's 23/30 guideline is a case in point. For planning convenience, a baseline "threat" had to be identified in order to ensure that NATO did not underestimate its possible operational problems. It was assumed, as a guideline only—not as a strategic prediction—that the Warsaw Pact would take thirty days to mobilize for war in Europe and that NATO would identify the character of the threat only Seven days into the Pact mobilization, thereby granting twenty-three days for countermobihzation. The thirty-day assumption was never intended to stand as a

Strategy and Nuclear Deterrence | 26

Third, and most important of all, it is sometimes easy to forget that a central nuclear war really could occur. Whatever the pre-war feelings, think­ ing, and even instincts, of a politician may have been, in the event of war it is safe to predict that he would demand a realistic war plan. The promise of imposing catastrophic levels of damage on Soviet society may, or may not, have merit as a pre-war declaratory stance, but the politician would find that his learning curve on nuclear strategy rose very rapidly indeed following a deterrence failure.7 Killing people and blowing down buildings, on any scale, cannot constitute a strategy—unless, that is, one has some well developed theory which specifies the relationship between societal damage, actual and threatened, and the achievement of (political) war aims. Unless one is willing to endorse the proposition that nuclear deterrence is all bluff, there can be no evading the requirement that the defense community has to design nu­ clear employment options that a reasonable political leader would not be self-deterred from ever executing, however reluctantly. Several years ago, Kenneth Hunt argued that among NATO's more im­ portant duties was the need to guarantee to the Soviet Union that it could not avoid having to initiate a "major attack" should it move westwards in Europe.8 NATO's function, on this theory, is not to defend Western Europe (at least, not directly); rather, it is to impose a high threshold for militarypolitical adventure—to compel Soviet leaders, of any degree of intelligence or rabid hostility, that there are no relatively cheap or risk-constrained mil­ itary options available. A similar logic underlies the policy positions of a major school of thought on strategic nuclear issues: all that we should, or need to, ask of the U.S. strategic nuclear posture is that it be capable of inflicting a lot of punishment on Soviet society. Precisely how much punish­ ment, and of what kinds, we need to promise, must be a matter for conjec­ ture, but fortunately for the robustness of a nuclear deterrent regime, pre­ judgment that the Soviet Union would attack only after such a lengthy period of mobilization, rather it was intended to generate a large, as opposed to a more modest, theater threat. Almost needless to say, 23/30 came to assume doctrinal significance. 7. As James Schlesinger once said: "[b]ut I might also emphasize, Mr. Chairman, that doctrines control the minds of men only in periods of non-emergency. They do not necessarily control the minds of men during periods of emergency. In the moment of truth, when the possibility of major devastation occurs, one is likely to discover sudden changes in doctrine." Testimony in U.S. Congress, Senate Committee on Foreign Relations, Nuclear Weapons and Foreign Policy, Hearings, 93rd Congress, 2nd session (Washington, D.C.: U.S. Government Printing Office, 1974), p. 160. 8. In The Alliance and Europe: Part II: Defence With Fewer Men, Adelphi Paper No. 98 (London: IISS, Summer 1973), p. 20 and passim.

Nuclear Strategy | 27

cision is not required. Essential to this thesis are the beliefs that nuclear war cannot be won, that there is no way in which damage in such a war can be held down to tolerable levels in the face of an adversary determined to impose major damage, and that notwithstanding the many differences be­ tween the superpowers in strategic culture, each side should have no diffi­ culty identifying, threatening, and, if necessary, effecting a level of damage that the other would find unacceptable. Strategic debate of recent years on SALT and strategic forces' issues has become so polarized and has involved such a high level of polemical "noise" that the transmission of signals between contending camps has been difficult. It is a considerable oversimplication to assert that there are two schools of thought on nuclear deterrence—there are not: there are many. However, while admitting the many nuances that separate the exact philosophies and policy prognoses of individuals, it is useful to recognize that the impending major debate over SALT II is being nurtured by what amounts to a funda­ mental dispute over the requirements, and even the place, of a deterrent policy. It is argued that the premises of the two loose coalitions of policy contenders drive the debate that surfaces all too often with reference to specific defense and arms control issues9—at a level of detail where the policy action (or inaction) advocated can have integrity only if it is related to basic assumptions and explicit desiderata (it is analogous to discussing strategy without reference to war aims). Assured Destruction and its Descendants: A Sickly Breed The first school of thought,10 which currently holds policy-authoritative sway in Washington (though it is not unchallenged within the government), may be thought of as the heir to the assured destruction ideas of the mid- to late 9. The "two camps" premise is not defended in detail in the text because (a) it is very close to being a self-evident truth, and (b) such an exercise in description would divert the discussion away from ideas and towards a summary of debate—with details required that are really of secondary importance, at most, to the theme of the article. Opinion, of course, exists on a spectrum. However, this author predicts that if one designed a simple questionnaire containing, say, ten "litmus paper-type" test questions of an either/or character, and submitted this ques­ tionnaire to 100 members of the U.S. national security community, inside and outside of government, there would be little cross-voting by individuals between "liberal" and "conserv­ ative" replies. Moreover, if one knew what an individual's final judgment was on SALT II, yea or nay, that fact would be extremely helpful in predicting his/her position on a wide range of other security issues. 10. In some important respects, it is more accurate and more satisfactory, at least for the limited purposes of this article, to talk of two schools, really loose coalitions of functional allies, of thought, than it would be to attempt to design a sophisticated multi-dimensional categorization

Strategy and Nuclear Deterrence | 28

1960s. In 1964-65, the U.S. defense community substantially abandoned the concept of damage limitation. It was believed that strategic stability ( the magic concept—far more often advanced and cited than defined),11 largely by virtue of a logic in technology (a truly American theme), could and should repose in what would amount to a strategic competitive stalemate. Each side could wreak unacceptable damage on the other's society, and neither could limit such prospective damage through counterforce operations or through active or passive defenses. Ballistic missile defense (BMD), in principle if not in contemporary technical realization, did of course pose a potentially fatal threat to this concept. A good part of the anti-ABM fervor of the late 1960s, which extended from Secretary of Defense Robert McNamara to local church and women's groups, can be traced to the strange belief that the goals of peace, security, arms control, stability and reduced resources devoted to defense preparation, could all flow from a context wherein societies were nearly totally vulnerable and strategic weapons were nearly totally invulner­ able. It is important to note both that U.S. operational planning never re­ flected any close approximation to the assured destruction concept,12 and that the legatees of MAD reasoning in the late 1970s have made some ad­ justments to the doctrine for its better fit with contemporary reality. The adjective "strange" was applied to MAD reasoning in the paragraph above from the perspectives of the historian of strategic ideas and of the sociologist of strategic culture. A detached observer might well ask and observe as follows: —Is MAD a matter of a logic in technology (wherein offense-dominance is a physical law), or is it more in the nature of a self-fulfilling prophecy? —If one side to the competition pursues the assured destruction path, how great a risk is it taking should the other side, for whatever blend of reasons, choose differently? of attitude and opinion. The latter implies a commitment to an accuracy in personal detail that verges upon the trivial and yet which could never really be complete. Probably the most satisfactory attempt at the categorization and analysis of strategic attitudes was Robert A. Levine, The Arms Debate (Cambridge, Mass.: Harvard University Press, 1963). However, even this excellent book suffered from the vices of its virtues. The very comprehensiveness of its coverage compelled the author to take at least semiserious note of opinions that are of no policy relevance. 11. See Edward Luttwak's contribution to "The Great SALT Debate," The Washington Quarterly, Vol. 2, No. 1 (Winter 1979), particularly pp. 84-85. 12. U.S. strategic nuclear planning was essentially unrevised from the Kennedy years until the early 1970s. See Desmond J. Ball, The Strategic Missile Program of the Kennedy Administration, 1961-1963. Unpublished manuscript (no date), Part 3, Chapter 2. For a definitive judgment we will have to await the eventual publication of the war plans (SIOPs), of the 1960s under the auspices of the Freedom of Information Act.

Nuclear Strategy | 29

—MAD and its variants assume a noticeable measure of functional conver­ gence of strategic ideas. But, strategic sociology tells us that each security community tends to design unique solutions to uniquely defined prob­ lems.13 —History may not tell us much with assurance, but certainly it suggests that technology cannot be frozen through arms control regimes: some qualita­ tive boundaries upon its inventory expansion may be accomplished, but the slender historical arms control record suggests that politicians are as likely to freeze the wrong, as the right (i.e., stabilizing!), developments, and that prohibitions in one area serve to encourage the energetic pursuit of capabilities in other areas.14 —Although the existence of nuclear weapons encourages nuclear-armed states to be extremely careful in their mutual dealings, the fact of nuclear weapons has yet to be transcribed into some absolute injunction against war. The late Bernard Brodie has offered the thought that "if it is not yet an established fact it is at minimum a strong possibility that, at least between the great powers who possess nuclear weapons, the whole char­ acter of war as a means of settling differences has been transformed beyond all recognition."15 Notwithstanding some recent claims by Raymond Garthoff,16 it is not at all certain that Brodie was correct with respect to the 13. The concept of strategic culture is a fascinating one and is as obvious as it has been neglected. For a brief and interesting introduction to the subject, see Jack L. Snyder, The Soviet Strategic Culture: Implications for Limited Nuclear Operations, R-2154-AF (Santa Monica, California; Rand, September 1977). The protracted SALT history has served to diminish enthusiasm for the strategic intellectual convergence thesis, but the U.S. Government is only at the beginning of attaining a due appreciation of the policy implications of the distinctive strategic culture thesis. This is one of those cases of rediscovery of the wheel. Most American strategic thinkers have always known that there was a uniquely "Soviet way" in military affairs, but somehow that realization was never translated from insight into constituting a serious and enduring factor influencing analysis, policy recommendation, and war planning. 14. Naval arms limitation by treaty in the1920s and 1930s (with its heavy focus upon battleships and, eventually, cruisers, should stand as a classic lesson for all time). Also, it is worth recalling Bernard Brodie's judgment on the complex naval competition of the last decades of the Nine­ teenth Century. "It is very likely that a more costly and politically more dangerous competition was avoided because the Powers permitted the building to go on steadily, subject only to selfimposed restraints, which in a period of such rapid obsolescence of new material were certain to be very real." Sea Power in the Machine Age (Princeton, N.J.: Princeton University Press, 1941), p. 254. A brilliant contemporary analysis of the unintended damage that can be wrought through the (mis)-control of technology is Edward N. Luttwak," SALT and the Meaning of Strategy," The Washington Review of Strategic and International Studies, Vol. 1, No. 2 (April 1978), pp. 16-28. 15. "The Continuing Relevance of On War," in Carl von Clausewitz, On War (Michael Howard and Peter Paret, eds.), (Princeton, N.J.: Princeton University Press, 1978), p. 49. 16. "Mutual Deterrence and Strategic Arms Limitation in Soviet Policy," International Security, Vol. 3, No. 1 (Summer 1978), pp. 112-147. For a preliminary reply to Garthoff, see the com­ mentary by Donald G. Brennan in International Security, Vol. 3, No. 3 (Winter 1978/1979).

Strategy and Nuclear Deterrence | 30

Soviet Union, while—even if he were correct—it remains the case that a nuclear war could occur. Mutual assured destruction, whatever its (highly dubious) merits as a pre-war deterrent declaratory stance, clearly has no appeal as operational guidance. Indeed, MAD is the antithesis of strategy— it relates military power to what?—to the punishment of a society for the sins or misjudgments of its rulers. —That the pre-war deterrent focus of MAD reasoning is appealing, but the historian in him/her is distressed by the realization that the advent of nuclear weapons has affected, but has not transformed the character of, international politics. Insane, drugged, or drunk American chief executives might seek to punish Soviet society, but a more responsible leadership has to be presumed to be likely to wish to adhere to an ethic of consequences (rather than revenge). Pre-War Deterrence: A Misleading Focus?

The second school of thought embraces a coalition of people who are con­ vinced: that Soviet strategic-nuclear behavior is difficult to equate even with a very rough facsimile of MAD reasoning; that the technical-postural basis for the American MAD thesis of the late 1960s has been eroded severely;17 and that the theory of mutual assured destruction, even as amended officially in the 1970s in favor of greater flexibility, appears to have little of merit to offer as an operational doctrine. To state the central concern of this article, U.S. official thinking and planning does not embrace the idea that it is necessary to try to effect the defeat of the Soviet Union. First and foremost, the Soviet leadership fears defeat, not the suffering of damage—and defeat, as is developed below, has to entail the forcible demise of the Soviet state. The second school of thought is edging somewhat tentatively towards the radical thesis that the theory of nuclear deterrence espoused, for example, by Ber­ nard Brodie from 1946 until 1978, a theory which stressed the "utility in 17. Assured destruction may have residual merit today in the strict context of deterring a Soviet counter-societal assault, but U.S. strategic forces have the same formal extended-deterrent duties that they have always had. As the Soviet Union has cancelled the more obvious U.S. strategic nuclear advantages, and as the U.S. continues to decline to seek to secure some measure of strategic superiority, so the attempt has been made to design "strategy offsets" for the adverse trend in the basic weapons balance. Very selective nuclear strike options, countereconomic recovery targeting, selective counter-military (and perhaps, in the 1980s, counterpolitical control) targeting, are all—to some degree—endeavors to effect an end run around the logical implications of an eroding military balance. This problem is well described with reference to the probable needs of NATO-Europe in Lawrence Martin's contribution to "The Great SALT Debate," pp. 29-37.

Nuclear Strategy | 31

nonuse [of nuclear weapons]",18 has had extremely deleterious effects upon the quality of Western strategic thinking and hence upon Western security. Above all else, our attention has been directed towards the effecting of pre­ war deterrence, at the cost of the neglect of operational strategy.19 Incredible though it may seem, it has taken the United States' defense community nearly twenty-five years to ask the two most basic questions of all pertaining to nuclear deterrence issues: these are, first, what kinds of threats should have the most deterring effect upon the leadership of the Soviet state?—and, second, should pre-war deterrence fail, what nuclear employment strategy would it be in the United States' interest actually to implement? The debate has yet fully to be joined, but this revisionist school of strategic theorists sees little merit in contemporary official U.S. deterrent policy (though the trend, as is discussed below, is mildly encouraging). The argu­ ment launched in public in late 1973 by then Secretary of Defense James Schlesinger concerning selective nuclear targeting20 served, in retrospect at least, more to encourage persuasive fallacies than it did to focus attention upon the real problem. Our real problem, according to this view, is that the United States (and NATO-Europe) lacks a theory of victory in war (or sat­ isfactory war termination). If, basically, one has no war aims (one has no image of enforced and favorable war termination, or of how the balance of power may be structured in a post-war world), on what grounds does one select a strategic nuclear employment policy, and how does one know how to choose an appropriate strategic posture? The answer, in this perspective, is that one does not know.21 The answer provided by the first school of thought is that one chooses an employment policy (at least at the declaratory level), with roughly matching equipment, that has little if anything to do 18. The title of Chapter 9 of his magnum opus, War and Politics (New York: Macmillan, 1973). The historical boundaries of Bernard Brodie's career as a theonst of nuclear affairs were marked by The Absolute Weapon (New York: Harcourt, Brace, 1946), pp. 21-110; and "The Development of Nuclear Strategy," International Security, Vol. 2, No. 4 (Spring 1978), pp. 65-83. 19. Note the scorn which Brodie pours upon the idea of "war-winning strategies" in "The Development of Nuclear Strategy," op. cit., p. 74. Nonetheless, a little earlier Brodie did observe that civilian scholars have "almost totally neglected" the question of "how do we fight a nuclear war and for what objectives?"—if deterrence fails (ibid., p. 66). 20. Probably the most powerful single exposition of "the Schlesinger doctrine" was Schlesinger's testimony in U.S. Congress, Senate Committee on Foreign Relations, Subcommittee on Arms Control, International Law and Organization, U.S.-U.S.S.R. Strategic Policies, Hearing, 93rd Congress, 2nd session (Washington, D.C.: U.S. Government Printing Office, 4 March, 1974). 21. The strategic flexibility theme was much criticized by representatives of the first school of deterrence theory (see Herbert Scoville, "Flexible MADness", Foreign Policy, No. 14 [Spring 1974], pp. 164-177, and Barry Carter, "Nuclear Strategy and Nuclear Weapons," Scientific

Strategy and Nuclear Deterrence \ 32

with the intelligent conduct of war. By definition it is assumed that nuclear war cannot be waged intelligently for rational political ends: the overriding function of nuclear weapons is the deterrence, not the waging, of war. The second school of thought objects to the above reasoning on several grounds. First, the heavy focus upon nuclear threat, as opposed to nuclear execution, has encouraged a basic lack of seriousness about the actual con­ duct of a nuclear war—which feeds back into an impoverished deterrent posture and doctrine. Second, although peace may be its profession, one day—arising out of political circumstances that no one could foresee with any confidence—SAC might discover that war is its business, and it would be better for our future if, in that event, SAC were guided by some theory of how it should wage the war to a tolerable outcome. As noted earlier, the somewhat irresponsible ideas that pass for orthodox nuclear deterrent wis­ dom, with their bottom-line focus upon damaging Soviet economic assets, would (as a prediction) evaporate in their official appeal in the event of a deterrence failure. Fundamentally, they are not serious. SAC does, of course, have plans to wage a central war in a fairly serious way. But, and it is a very large but, (1) U.S. strategic weapon acquisition policy (under four Presidents) has failed to provide SAC with the means to prosecute the counter-military war very effectively, and (2) our counter-military planning (however well or poorly it could be executed) continues to be deprived of the overarching political guidance that it needs—a definition and a concept of victory. Superficially at least, Schlesinger's strategic flexibility, as reflected in NSDM 242 and eventually in actual nuclear employment plans,22 marked a noteworthy improvement in the quality of U.S. deterrent policy. A richer menu of attack options, small and large, would provide a president with less-than-cataclysmic nuclear initiatives, should disaster threaten, or occur, in Europe or elsewhere.23 Selectivity of scale and kind of attack, it was and is still argued, enhances deterrence because it promotes the vital quality of

American, Vol. 230, No. 5 [May 1974], pp. 20-31), but those representatives—reasonably enough, from their perspective—did not offer the most telling line of criticism: namely, that strategic flexibility, however desirable in and of itself (a view which Scoville and Carter did not share), does not constitute, or even approximate, a strategy. 22. See Lynn E. Davis, Limited Nuclear Options: Deterrence and the New American Doctrine, Adelphi Paper No. 121 (London: IISS, Winter 1975/6); and Desmond Ball, Deja Vu: The Return to Counterforce m the Nixon Administration (Los Angeles: California Seminar on Arms Control and Foreign Policy, December, 1974). The Davis characterization, with its focus upon limited nuclear options (LNOs) is very substantially misleading as to the basic thrust of NSDM 242. 23. See Martin, in The Washington Quarterly, op. cit., passim.

Nuclear Strategy | 33

credibility. As far as it goes, that line of thought has much to recommend it. Few would deny that a president should feel less inhibited over the pro­ spective dispatch of (say) thirty (or even 130) reentry vehicles, than he would over the dispatch of one to three thousand—particularly when the targets for those thirty to 130 reentry vehicles had been chosen very carefully with a view to inflicting the minimum possible population loss on the Soviet Union. This theme of restraint, selectivity, and usability—all in the interest of enhancing credibility for the improvement of deterrence—attracted pre­ dictable negative commentaries from quarters prone to argue that a more usable nuclear deterrent was a nuclear deterrent more likely to be used24 (similar arguments surfaced in connection with the protracted debate over enhanced radiation weapons). As Herbert Scoville explained: "[a] flexible strategic capability only makes it easier to pull the nuclear trigger."2S The second school of thought has no quarrel whatsoever with the ideas of flexibility, restraint, selectivity, minimal collateral damage and the rest. But, it does have some sizable quarrel with strategic selectivity ideas that are bereft of a superordinate framework for the conduct and favorable termina­ tion of the war. Against the background of a fairly steadily deteriorating strategic nuclear balance,26 the selectivity thesis simply adds what could amount to bigger and slightly more effective (i.e., the Soviets pay a higher military price) ways of losing the war. With a healthy strategic (im)balance in favor of the United States on the scale of, say, 1957 or 1962, one can see some logic to strategic flexibility reasoning. However, in the late 1970s and the 1980s, there are many reasons why a Soviet leadership might be less 24. For examples, see George W. Rathjens, "Flexible Response Options," Orbis, Vol. XVIII, No. 3 (Fall 1974), pp. 677-688, and Herbert Scoville, "'First Use' of Nuclear Weapons," Arms Control Today, Vol. 5, No. 7/8 Quly/August 1975), pp. 1-3. 25. Ibid, p. 2. 26. Assessment of trends in the strategic balance tend to be driven to a non-marginal degree by the doctrinal preferences of the assessor: somehow, people manage to find their beliefs sup­ ported by statistics. Nonetheless, it is difficult to analyze the contemporary strategic balance and find much comfort therein (almost regardless of one's doctrinal preferences). A group of analysts in ACDA has succeeded in achieving this quite remarkable goal. See U.S. and Soviet Strategic Capability Through the Mid-1980s: A Comparative Analysis (Washington, D.C.: U.S. Arms Control and Disarmament Agency, August, 1979). Needless to say, there are several important premises that one has to grant for the analysis to turn out as it does. Gloomier prognoses for the United States include: Santa Fe Corporation, Measures and Trends: U.S. and U.S.S.R. Strategic Force Effectiveness, DNA 4602Z (Washington, D.C.: Defense Nuclear Agency, March, 1978); John Collins, American and Soviet Military Trends since the Cuban Missile Crisis (Washington, D.C.: Center for Strategic and International Studies, Georgetown University, 1978); and John Collins and Anthony Cordesman, Imbalance of Power: An Analysis of Shifting U.S.-Soviet Military Strengths (San Rafael, California: Presidio, 1978).

Strategy and Nuclear Deterrence | 34

than fully impressed by constrained U.S. strategic execution, and might well respond with a constrained nuclear reply that would (and indeed should) most likely impose a noteworthy measure of escalation discipline upon the United States.27 Selective nuclear options, even if of a very heavily countermilitary character, make sense, and would have full deterrent value, only if the Soviet Union discerned behind them an American ability and will to prosecute a war to the point of Soviet political defeat. Targeting the Recovery Economy Of very recent times, much of the nuclear strategy debate has narrowed down to a dispute over the validity of the thesis that the real (and ultimate) deterrent to Soviet risk-taking/adventure is the threat that our strategic nuclear forces pose to the Soviet recovery economy. Orthodox assured destruction thinking has evolved since the late 1960s. Notwithstanding the worthy de­ terrent motives of their authors, it is a fact that the last several annual Posture Statements of Robert McNamara endorsed a mass murder theory of nuclear "war" (to stretch a term). In the event of a central nuclear war, our decla­ ratory policy was to kill tens of millions of Soviet citizens and destroy Soviet industry on a heroic scale.28 Fortunately, under Presidents Nixon and Ford, killmg people and blowing down buildings per se ceased to be strategic objectives (though, to repeat, this is not to impugn the motives of the MAD theorists of the 1960s—they wished to deter war: a highly ethical objective— it is only their judgment that is challenged here). Instead, it was noticed (belatedly—though welcome for all that) that recovery from war was an

27. This thesis is argued forcefully in Paul H. Nitze, "Deterring Our Deterrent," Foreign Policy, No. 25 (Winter 1976-77), pp. 195-210. On the subject of possible Soviet responses to American selective strike options see Benjamin S. Lambeth, Selective Nuclear Options in American and Soviet Strategic Policy, R-2043-DDRE (Santa Monica, California: Rand, December, 1976). Also note the very brief discussion in Harold Brown, Department of Defense Annual Report, Fiscal Year 1979 (Washington, D.C.: U.S. Government Printing Office, 2 February, 1978), pp. 55-56, 62. Rather more interesting is Harold Brown, Department of Defense Annual Report, Fiscal Year 1980 (Wash­ ington, D.C.: U.S. Government Printing Office, January 25, 1979), pp. 77-78. These two para­ graphs, weak though they are, constitute the strongest Posture Statement language in favor of (second-strike) hard-target counterforce that has been seen for well over a decade. 28. Statement of Secretary of Defense Robert S. McNamara on the Fiscal Year 1968-72 Defense Program and 1968 Defense Budget (Washington, D.C.: U.S. Government Printing Office, 23 January, 1967), Chapter II; and Statement by Secretary of Defense Robert S. McNamara on the Fiscal Year 1969-73 Defense Program and the 1969 Defense Budget (Washington, D.C.: U.S. Government Printing Office, 22 January, 1968), Chapter II. Also see Jerome H. Kahan, Security in the Nuclear Age: Developing U.S. Strategic Arms Policy (Washington, D.C.: Brookings, 1975), pp. 94-106.

Nuclear Strategy | 35

integral part of the Soviet concept of victory29—ergo, the United States should threaten the post-war recovery of the Soviet Union.30 The counter-recovery theory was not a bad one, but in practice several difficulties soon emerged. First, and most prosaically, American understand­ ing of the likely dynamics of the Soviet post-war economy was (and remains) far short of impressive. In the same way that arms controllers have been hindered in their endeavors to control the superpower strategic arms com­ petition by their lack of understanding of how the competition "worked," so our strategic employment planning community has found itself in the posi­ tion of being required to be able to do that which nobody apparently is competent to advise it how to do. To damage the Soviet recovery economy would be a fairly elementary task, but to damage it in a calculable (even a roughly calculable) way is a different matter. Furthermore, the discovery, year by year through the mid- to late 1970s, of more and more Soviet civil defense preparation, threw into increasing doubt the "damage" expectancy against a very wide range of Soviet economic targets. Second, it appears that the counter-economic recovery theme is yet another attempt to evade the most important strategic question. Should war occur, would the United States actually be interested in setting the Soviet economy back to 1959, or even 1929? Such an imposed retardation might make sense if it were married to a scheme for ensuring that damage to the American economy were severely limited. However, no such marriage has yet been mooted in policy-responsible circles.31 Third, it is possible that the posing

29. Counter-recovery targeting was not, of course, invented in the 1970s. In 1967, Robert McNamara said that "it seems reasonable to assume that in the case of the Soviet Union, the destruction of, say one-fifth to one-fourth of its population and one-half to two-thirds of its industrial capacity would mean its elimination as a major power for many years." (Emphasis added). Statement . . on the Fiscal Year 1968-72 Defense Program . . . , op. cit., p. 39. Counter-recovery targeting has come, in the 1970s, to imply attacks on a more discrete character than those suggested in McNamara's words. 30. General George Brown, then Chairman of the Joint Chiefs of Staff was very explicit on this subject. "We do not target population per se any longer. What we are doing now is targeting a war recovery capability." Quoted m The Defense Monitor, Vol. VI, No. 6 (August 1977), p. 2. 31. On the contrary, the current Secretary of Defense has written as follows: "I am not per­ suaded that the right way to deal with a major Soviet damage-limiting program would be by imitating it. Our efforts would almost certainly be self-defeating, as would theirs. We can make certain that we have enough warheads—including those held in reserve—targeted in such a way that the Soviets could have no expectation of escaping unacceptable damage." Department of Defense Annual Report, Fiscal Year 1979, op. at., p. 65. Of course the United States could impose unacceptable damage upon the Soviet Union, but there is no good reason to believe that the current administration (1) knows what unacceptable damage means in Soviet terms; (2) would be willing to fund a U.S. strategic posture capable of imposing truly unacceptable damage; or

Strategy and Nuclear Deterrence | 36

(even credibly posing) of major economic recovery problems to the Soviet Union might be insufficiently deterring a prospect if Soviet arms could ac­ quire Western Europe in a largely undamaged condition to serve as a recovery base; if the stakes in a war were deemed by Moscow to be high enough; and if the Soviet Union were able, in the course of the war, to drive the United States back to an agrarian economy.32 It is difficult to disagree with Henry Kissinger's comment on massive counterpopulation strikes. Every calculation with which I am familiar indicates that a general nuclear war in which civilian populations are the primary target will produce cas­ ualties exceeding 100m. Such a degree of devastation is not a strategic doc­ trine: it is an abdication of moral and political responsibility. No political structure could survive it.33 Targeting the Soviet State

Nonetheless, the counter-recovery theme of the 1970s has prompted an interesting line of speculation. Namely, perhaps the recovery that should be threatened is not economic in character, but rather political.34 Some revi­ sionists of the second school of deterrence theorists argue that any kind of counter-economic strategy is fundamentally flawed because it leads into Soviet strength. The Soviet Union, like Czarist Russia, knows that it can absorb an enormous amount of punishment (loss of life, industry, productive agricultural land, and even territory), recover, and endure until final vic­ tory—provided the essential assets of the state remain intact. The principal assets are the political control structure of the highly centralized CPSU and government bureaucracy; the transmission belts of communication from the center to the regions; the instruments of central official coercion (the KGB and the armed forces); and the reputation of the Soviet state in the eyes of its citizens. Counter-economic targeting should have a place in intelligent

(3) would be capable of understanding that our offensive strategy will avail us very little if our domestic assets are totally at risk. 32. This author sees some merit in Bernard Brodie's comment (on the targeting of war recovery capability) that "[w]hatever else may be said about this idea, one would have to go back almost to the fate of Carthage to find an historical precedent." The Development of Nuclear Strategy," op cit., p. 79. 33. "Kissinger's Critique [of SALT II]", The Economist, February 3,1979, p. 18. Kissinger watchers should note that their subject traditionally has been as poor a strategic theoretician as he has been a strong foreign policy analyst. 34. This idea has had some U.S. official status for at least five years, but its detailed meaning has never been probed rigorously.

Nuclear Strategy | 37

U.S. war planning, but only to the extent to which such targeting would impair the functioning of the Soviet state. The practical difficulties that would attend an endeavor to wage war against the Soviet state, as opposed to Soviet society, have to be judged to be formidable. However, one would at least have established an unambiguous and politically meaningful war aim (the dissolution of the Soviet political system) that could be related to a post-war world that would have some desirable features in Western (and Chinese) perspective. More to the point perhaps, identification of the demise of the Soviet state as the maximum ambition for our military activity, encourages us to attempt to seek out points of high leverage within that system. For examples, we begin to take serious policy note of the facts that: —The Soviet peoples as a whole have no self-evident affection for, as op­ posed to toleration of, their political system or their individual political leaders. —The Soviet Union, quite literally, is a colonial empire—loved by none of its non Great Russian minority peoples.35 —The Soviet state has to be enormously careful of its domestic respect and reputation, so fragile is the system deemed to be (evidence of Soviet official estimates of this fragility is located in the very character of the police state apparat that is maintained, and in the extreme sensitivity historically dis­ played in response to threats to Soviet authority in Eastern Europe). —The entire Soviet political and economic system is critically dependent upon central direction from Moscow. If the brain of the Soviet system were destroyed, degraded, or—at a minimum—isolated from those at lower levels of political command who traditionally have been discouraged from showing initiative, what happens to the cohesion, or pace of recovery, of the whole? —The peoples of Eastern Europe and the minority republics in the Soviet Union itself, respect the success and power of the Soviet state. What happens in terms of the acquiescence of these peoples in Soviet (and Great Russian) hegemony if Soviet arms either are defeated, or are compelled to wage a long and indecisive struggle?36 35. The Soviet imperial thesis has been advanced strongly by Richard Pipes. See: Russia Under the Old Regime (New York: Scribner's, 1974); and "Detente: Moscow's View," in Pipes, ed., Soviet Strategy in Europe (New York: Crane, Russak, 1976), pp. 3-44. 36. It is one thing if the Soviet state is able, as in The Great Patriotic War, to assume the mantle

Strategy and Nuclear Deterrence | 38

Improbable though it may seem to many, this discussion is beginning to point towards a not-implausible theory of victory for the West. The alter­ native theory of deterrence/war waging proposed by some people within the second school, by way of contrast to the mass murder, punishment theme of the first school, comprises essentially, the idea that the Soviet system be encouraged to dissolve itself. We resist the external military pressure of the Soviet Union, and effect carefully selected kinds of damage against the capacity of the Soviet state to function with authority at home. Soviet leaders can reason as well as Western defense analysts that large-scale countereconomic strikes would not serve Western interests (if only because of the retaliation that they would invite), whereas a war plan directed at the de­ struction of Soviet power would have inherent plausibility in Soviet estima­ tion. The Decline (but not Fall) of Assured Destruction The essential backcloth to this counter-political control strategy has to be the ability to deny the Soviet Union any outcome approximating military victory in a short war. No matter how intelligent our ultimate goals may be for World War III, if the Soviet Union can (or believes that it can) win a rapid campaign against NATO-Europe37 and, if need be, could escalate to do unmatchable damage to U.S. strategic forces, while holding virtually all American economic assets at nuclear risk, then the second school would have failed to think through the totality of the deterrence problem. Needless to say, scarcely less significant a weakness in orthodox deterrence thinking than the fact that it focuses upon the threat of effecting the kinds of damage to the Soviet Union that should not be of interest to American policy makers actually to execute, is the fact that it discounts totally the intra-war selfdeterrent implications of the vulnerability of American assets. Foreign policy, in good part, is about freedom of action. Mutual assured destruction think-

of defender of Mother Russia and if the general populace discerns no reasonable political alternative to Soviet power. It is quite another if the external enemy is being combatted militarily far from home, if "that enemy" seeks intelligently to exploit the latent fragilities of the Soviet system, and if the military damage suffered on Soviet soil is very substantially confined to Soviet-state type targets. 37. A very persuasive recent discussion of this area is C. N. Donnelly, "Tactical Problems Facing the Soviet Army: Recent Debates in the Soviet Military Press", International Defense Review, Vol. II, No. 9 (1978), pp. 1405-1412. Also see Peter Vigor, The Soviet View of War, Peace, and Neutrality (London: Routledge and Kegan Paul, 1975), pp. 14-15 and passim.

Nuclear Strategy | 39

ing, which still lurks in our declaratory policy and, presumably, in our war plans, virtually ensures self-deterrence and denies us the freedom of stra­ tegic-nuclear action that is a premise of NATO's strategy of flexible re­ sponse.38 It is no exaggeration to claim that still-orthodox punishment-oriented de­ terrence thinking stemmed to a notable degree from a group of theorists who tended to think of the superpowers as though they were two missile farms: the attainment of an assured destruction capability by both sides would encourage the establishment and endurance of a technologically imposed peace.39 The idea was fundamentally apolitical, astrategic, and was contrary to what the Soviet Union discerned, very sensibly, as its self-interest. Overall, as John Erickson has observed, American thinking on mutual deterrence, with its technological premises, reflects a "management" approach by way of contrast to "the Soviet 'military' inclination."40 This author has difficulty understanding how a country like the United States, which has accepted obligations to project power at great distances in support of forward-located allies, could have seen any noteworthy attractions in the mutual hostage theory of deterrence. Of all countries, the United States needs a credible strategic force posture married to a theory of feasible employment. The catastrophic retaliation thesis, whether or not preceded by very selective nuclear employment options, is an idea it would be hard to improve upon were one seeking to minimize the relevance of (American) strategic weapons to world politics. It is probably appropriate largely to dismiss the deterrencethrough-punishment ideas of the 1960s (or, at least, as formalized and cod­ ified in the 1960s) as the products of a defense community that was neither trained nor inclined to think strategically.41 After all, the codification of the 38. For an expansion upon this argument see Colin S. Gray, "The Strategic Forces Triad: End of the Road?", Foreign Affairs, Vol. 56, No. 4 (July 1978), particularly pp. 771-778. 39. A brief clear statement of this thesis permeated Robert McNamara's statement introducing the supposedly China-oriented Sentinel ABM system. "McNamara Explanation of 'Thin' Missile Defense System," The Washington Post, September 19, 1967, p. A.10. Also see Wolfgang Κ. H. Panofsky, "The Mutual-Hostage Relationship between America and Russia," Foreign Affairs, Vol. 52, No. 1 (October 1973), pp. 109-118. Interestingly enough, some of the more intense denials of McNamara-Panofsky reasoning seem to focus unduly upon the tactical merits of particular weapon systems. 40. "The Chimera of Mutal Deterrence," Strategic Review, Vol. IV, No. 2 (Spring 1978), p. 16. Also very useful is the discussion in Fritz Ermarth, "Contrasts in American and Soviet Strategic Thought," International Security, Vol. 3, No. 2 (Fall 1978), pp. 138-155. 41. Policy makers in Washington might profit from frequent reminders of CIausewitz' definition of strategy. Strategy teaches "the use of engagements for the object of the war" (emphasis in the original). On War, op. tit., p. 128,

Strategy and Nuclear Deterrence | 40

mutual punishment theory of deterrence as explicit policy—between 1964 and 1968—coincided exactly, and scarcely totally by chance, with gross stra­ tegic mismanagement in, and concerning, Vietnam. The same Department of Defense policy-making hierarchy that could not (or would not) design a theory of victory for Vietnam,42 similarly abandoned such an apparently extravagant notion in the realm of strategic nuclear policy. Until the mid-1960s, it is probably true to say that the quality of American strategic thinking concerning central war execution was a matter of relatively little importance: defeat for the Soviet Union was virtually implicit in the sheer scale of the strategic imbalance (i.e., even if the United States, in the event of war, had executed a foolish strategy, it would have done so on so massive a relative basis that the Soviet Union could not possibly have emerged from such a conflict with any net profit). But, as the capabilities of the two sides approached rough equivalence in the early to mid-1970s, the quality of strategic thinking, as reflected in actual plans, could easily make the difference between victory and defeat, or recovery and no recovery.43 The strategic debate referred to repeatedly in this article thus far is in a curious condition—with neither side quite sure of which positions are really

k

42 As Alexis de Tocqueville and many lesser commentators have observed, the conduct of foreign policy is not, and (given its political structure) cannot be, an American forte For a sense of perspective, it is worth noting that very few countries can wage long, losing (or perpetually inconclusive) wars and emerge with little, if any, domestic damage If Americans feel ashamed, in different ways, over their Vietnam record, they should consider what the war m Algeria did to France Admittedly, Algeria was a true colonial war, but still it was a case of a democracy attempting to cope with the consequences of military success and political defeat Any American president should know that the only kind of war his country can fight, and fight very well, is one where there is a clear concept of victory—analogically, the marines raising the flag on Mt Suribachi is the way m which a president should think of Amencan wars being terminated The more distant the Mt Suribachi analogue from the case at hand, the more doubtful a president should be over committing U S forces to action The U S public could have understood, and almost certainly would have approved, the U S Manne Corps seizing Hanoi (intact or rubble— no matter) in 1965 or 1966, and compelling Ho Chi Minh (or a successor—again, no matter) to sign a peace treaty That would be victory Amencan academic theonsts of "limited" (and "subhmited" war m the late 1950s and early 1960s) simply failed to understand their own country Most Americans believe that if wars are not worth winning (in fairly classical terms), they are not worth fighting 43 See T K Jones and W Scott Thompson, "Central War and Civil Defense," Orbis, Vol 22, No 3 (Fall 1978), pp 681-712, and Director of Central Intelligence, Soviet Civil Defense (Wash­ ington, DC CIA, July, 1978) This latter study claims that Soviet casualties (only half of which would be fatalities) could be held to "the low tens of millions", though only under the most favorable conditions for the Soviet Union (p 4) Some Boeing civil defense studies have sug­ gested, by way of contrast, that under the most favorable conditions Soviet population fatalities would be less than ten million Most commentators agree that a proper mix of offensive and defensive programs should make a dramatic difference to the prospects of early post-war recovery

Nuclear Strategy | 41

worth defending. Notwithstanding the high polemical noise level, there has been a very notable narrowing of real differences of opinion over the past three to four years. For prominent examples: —There is now widespread endorsement of the thesis that Soviet strategic thinking differs markedly from American. Indeed, recognition of what in the West we term a "war-fighting" focus (on the part of the Soviet Union) has helped greatly to promote insecurity in the minds of many over either the inherent wisdom, or the practical advisability (or both), of a punishment-oriented theory of nuclear deterrence. It is many years since com­ mentators in the United States have written about "raising the Russian learning curve."44 '—Today, there is virtually universal agreement that, notwithstanding the many and accelerating weaknesses in the Soviet system, most of the major military balances have been moving to the disadvantage of the West. There is no consensus over whether or not those trends will continue into and through the 1980s, nor as to whether or not those adverse trends constitute cause for alarm as opposed merely to concern.45 —To the knowledge of this author, in the United States' defense and arms control community today there are no strong adherents to anything ap­ proximating the pure theory of mutual assured destruction. But, those who have disengaged from the arguments of Robert McNamara's Posture Statements for 1968 and 1969 seem to be uncertain as to what other doctrinal haven there might be available, while many of those who have rejected MAD reasoning outright are less than confident that they have identified any superior alternative. A Catalogue of Confusion

The admittedly unsatisfactory designations, "first school" of thought and "second school" of thought, have been employed here because there is a considerable danger of unintended misrepresentation and undue simplifi44. John Newhouse, Cold Dawn: The History of SALT (New York: Holt, Rinehart and Winston, 1973), p. 4. Also of interest is Raymond L. Garthoff, "Salt 1: An Evaluation," Worid Politics, Vol. XXXI, No. 1 (October, 1978), particularly pp. 3, 24. 45. It is beginning to be fashionable to concede that the West will have to endure several years of unusual peril in the early 1980s, in terms of military balances considered narrowly, but that that condition will be transformed in the latter half of the decade as the U.S. strategic force posture accommodates cruise missiles, a follow-on ICBM and, eventually, the Trident 2 SLBM. A similar phenomenon is claimed for the trend in the theater balance in Europe: NATO's longterm defense program should have a very noticeable cumulative impact by the mid- to late

Strategy and Nuclear Deterrence | 42

cation should any less neutral titles have been chosen. At this stage in the article, it may be safe to introduce the claim that the "first school" corre­ sponds roughly to a focus upon "deterrence through punishment," while the "second school" tends to focus upon deterrence through the expectation of a militarily effective prosecution of war. Alas for neatness of description, neither group closely approaches its ideal type.46 The first school has recognized the immorality, inflexibility and plain in­ credibility of having a strategic force posture pre-programmed to deliver only massive strikes against Soviet economic assets per se. However, because it rejects any thoroughgoing "war-fighting" alternative as being certain to stim­ ulate the arms competition, perhaps to render war more likely (through the believed consequent increase in strategic instability), and to make the pros­ pects for negotiated measures of arms control far less encouraging, it has endeavored to design what might be termed "assured destruction with a human face." In place of the grisly (though superficially anodyne) prose of 1967-68 vintage McNamara, we are told about the deterrent virtues of stra­ tegic flexibility and the ultimately dissausive merits of impairing Soviet eco­ nomic recovery to a catastrophic degree. However, as observed above, the first school has yet to cope adequately with the rather obvious critical point that strategic flexibility and counter-recovery targeting are options that two can exercise. An intelligent strategy, if feasible, would be to design nuclear threats and employment options that the adversary either cannot or dare not match (or overmatch). Also, the first school has been increasingly overtaken by developments both in American weapon laboratories and, above all else, in the force posture that the Soviet Union is deploying. There is no logical reason why one should shift from a selective punishment thesis as a conse­ quence of observing the Soviet strategic developments of recent years (if one endorses the punishment thesis); but it does appear that many commentators have been uneasy and defensive in a context where the Soviet Union is apparently challenging every major tenet of the American theory of strategic stability.47 1980s. This author grants the possible validity of this theory, but is disturbed by the fragility of almost all of its premises. Henry Kissinger has commented persuasively on the early to mid1980s being "a period of maximum peril" in "Kissinger's Critique," op. ctt., p. 20. 46. The U.S. Department of Defense, in its declaratory policy, and even more in its actual operational planning (though not in its force acquisition), stands squarely between the two schools. DoD planning looks as though it is about the serious prosecution of war, but (a) the proper means are lacking, and (b) (to repeat a now familiar refrain) there is no theory of Soviet defeat to be discerned. 47. Soviet offensive-force development will, on current trends, pose an unacceptably high threat to the pre-launch survivability of the U.S. ICBM force by the early 1980s (for contrasting analyses

Nuclear Strategy | 43

First-school adherents are obliged by the contemporary climate of opinion in the United States to endorse the proposition that there should be an "essential equivalence" in strategic prowess between the superpowers, but what can this mean when there are very large asymmetries in strategic doctrine? Does it mean that the United States should invest in strategic capabilities that it deems to be destablizing (say, hard-target counterforce and civil defense), solely in order to provide a perceptual match with Soviet capabilities?48 In practice, first school theorists are finding it very difficult to resist venturing into program regions which really have no place in their philosophy. The result, as may be seen in the curious mix of half-heartedly promoted programs and ill-assorted ideas that constitute current strategic policy, is something for nearly everybody. Because the official theory of nuclear deterrence is so uncertain, one sees the following: —A new commitment to civil defense, qualified near-instantly by assurances that the new commitment will be neither very expensive nor so serious as to pose a threat to strategic stability. —A commitment to preserve a survivable ICBM leg to the strategic forces triad, but one that will pose as little (and as late) a threat to fixed site Soviet assets as the domestic SALT-related traffic will permit. —A commitment to the devising of a new strategic nuclear employment doctrine, but not one which challenges any of the basic premises of the deterrence through punishment thesis. —A commitment to second-strike hard-target counterforce prowess, on a scale which should fuel little first-strike anxiety in Moscow.49 —A commitment to a SALT process, and to a SALT II outcome, that has no reference to a stable strategic doctrine that has political integrity.50 As the period of intense debate over SALT II begins, it is fair to note that the United States Government sees merit in strategic flexibility, in some of this problem see John D. Steinbruner and Thomas M. Garwin, "Strategic Vulnerability: The Balance Between Prudence and Paranoia", International Security, Vol I, No. 1 (Summer 1976), pp. 138-181, and Colin S. Gray, The Future of Land-Based Missile Forces, Adelpht Paper No. 140 (Winter 1977)), while the Soviet Union continues to invest very heavily in active and passive defense of its homeland—thereby rejecting the thesis that it is desirable for Soviet assets to be at nuclear risk as hostages to the prudent behavior of Soviet (and American and Chinese) leaders. 48. On this subject see Paul C. Warnke "Apes On A Treadmill," Foreign Policy, No. 18 (Spring 1975), pp. 12-29. 49. Brown, Department of Defense Annual Report, Fiscal Year 1980, pp. 77-78. 50. It is difficult to tell a convincing story in support of SALT II, when the strategic doctrine that provides the political meaning in the strategic force posture is very uncertain. On what basis can one assess adequacy?

Strategy and Nuclear Deterrence ) 44

counterforce, in some degree of direct protection for the American public (though not much), and in the ability, in the last resort, to blow down large sections of the urban Soviet Union. This may be sufficient for deterrence, but a defense community should be capable of providing strategic direction that has more political meaning. Revisionist Claims: Myths and Reality

The second school of deterrence theory waxes eloquent on the absence of strategy in official policy, and indeed on the rarity of strategic thinking within the defense community,51 but remains slightly abashed at the boldness, and even apparent archaism, of the logic of its own position. Today's revisionists are challenging the mature judgment of the finest flowering of American strategic thought. In policy terms at least, "The Golden Age" of American strategic thought extended roughly from 1956 until (at the outside) 1965.52 The author of probably the single most highly regarded book to appear in this period53 has written as follows in the pages of this journal: The main war goal upon the beginning of a strategic nuclear exchange should surely be to terminate it as quickly as possible and with the least amount of damage possible—on both sides.54 Of course, the best prospect of all for minimizing (prompt) damage lies in surrendering preemptively. If Bernard Brodie's advice were accepted, the

51. For examples, see Daniel O. Graham, "The Decline of U.S. Strategic Thought," Air Force Magazine, Vol. 60, No. 8 (August, 1977), pp. 24-29; and Luttwak, "SALT and the Meaning of Strategy", op. at. The scope for strategic thinking may, of course, be reduced if one discerns no, or hardly any, political value in military action at the level of nuclear operations. In the words of Fritz Ermarth: "For many years the prevailing U.S. concept of nuclear war's consequences has been such as to preclude belief in any military or politically meaningful form of victory." "Contrasts in American and Soviet Strategic Thought," op. at., p. 144. One might reformulate Clausewitz' definition of strategy so as to read "the use of [the threat] of engagements for the object of the war" (On War, Op. Cit., ρ 128, my addition in brackets) in order to accommodate the strategy of deterrence and compellence, but there is grave danger in the judgment offered by Bernard Brodie in 1946: "Thus far the chief purpose of our military establishment has been to win wars. From now on its chief purpose must be to avert them. It can have almost no other useful purpose," The Absolute Weapon, op. cit., p. 76. This is a prime example of a good idea becoming a poor idea when it is taken too far: at worst, it is a doctrinal formula for losing wars. 52. Two as yet unpublished manuscripts discuss the rise of (civilian) nuclear-age strategic theorizing in great detail. These are James King, The New Strategy; and my own Strategic Studies and Public Polity: The American Experience. 53. Bernard Brodie, Strategy in the Missile Age (Princeton, N.J.: Princeton University Press, 1959). 54. Brodie, "The Development of Nuclear Strategy," op. cit., p. 79.

Nuclear Strategy \ 45

West would be totally at the mercy of a Soviet Union, which viewed war in a rather traditional perspective. The second school of nuclear deterrence is concerned lest its debating adversaries, neglecting the degree to which their ideas rest upon an unacknowledged measure of U.S. firepower (if not strat­ egy) superiority, which no longer exists, may mislead American policy mak­ ers into ignoring the possibility that nuclear age crises and wars can be lost, in a quite unambiguous fashion. The various arguments of the second school (really a loose coalition) of strategic theory do, it must be admitted, lend themselves fairly easily to grotesque misrepresentation. For example, responsible theorists of this per­ suasion do not claim: —That the Soviet Union believes that it will win a thermonuclear war (in­ stead, it is claimed that there is an impressive apparent consensus among Soviet authorities to the effect that victory [and defeat] is possible).55 —That the Soviet Union either wants or expects to have to wage a central war with the United States. Military power can be most useful, anH costeffective, when the mere promise of its exercise achieves desired deterrent and compellent outcomes. It is very likely indeed that the Soviet Union sees its strategic forces largely in a counterdeterrent role—functioning to seal off local conflicts from influence by U.S. strategic forces. However, any Soviet skepticism over the likelihood of central war does not (to the best of our knowledge) spill-over into defense programs and doctrine in the form of weapons and ideas that make little or no military sense. Because war is possible, one prepares sensibly for it. —That the Soviet Union anticipates achieving ultimate victory in war at little cost (much, though by no means all, of the argument of recent years in the United States concerning Soviet civil defense is really beside the point). Cautious committeemen in the Politburo could not afford to assume that Τ. K. Jones' optimistic studies (in Soviet perspective) were even close to the mark.56 Second-school theorists, by and large, anticipate Soviet ex­ pectation of the necessity of accepting human and industrial loss on a catastrophic scale. However, catastrophic loss need not be intolerable

55. See Richard Pipes, "Why the Soviet Union Thinks It Could Fight and Win a Nuclear War," Commentary, Vol. 64, No. 1 (July, 1977), pp. 21-34. 56. See U.S. Congress, House of Representatives, Committee on Armed Services, Civil Defense Review, Hearings, 94th Congress, 2nd session (Washington, D.C.: U.S. Government Printing Office, 1976), pp. 206-267.

Strategy and Nuclear Deterrence | 46

loss—and may indeed be loss of a kind that the Soviet Union is willing to absorb, if the political stakes in the conflict are high enough (and if the alternatives to extreme measures of military action are very unattractive in their likely returns). It is fundamental to the Soviet theory of victory that the essential (and as much else in the) homeland be preserved. It is a sobering thought that the loss of 30 or 40 million people might well be compatible with a context defined by a Soviet leadership as victory: it would depend very much on who was among that 30 or 40 million.57 —Any certain knowledge concerning the requirements of deterrence or the proper conduct of thermonuclear war for a politically acceptable outcome. What is claimed is that the ideas of the 1960s (the assured destruction of people and industry) and the 1970s (small- and large-scale attack options of a carefully constrained nature, counter-economic recovery targeting, and the currently increasing interest in even more counter-military options [than in the past—which was fairly extensive]) cannot withstand critical examination, given the adverse evolution of the major East-West military balances, and the more mature Western understanding of the Soviet ap­ proach to the waging of war. Counter-Military Targeting

Newspaper reports in late 1978 and early 1979 suggested that the Department of Defense was attracted to the idea of a substantially counter-military tar­ geting doctrine, in contrast to the counter-economic recovery theme.58 How­ ever, intra-governmental opposition to this idea is substantial, in part for reason of its budgetary implications, and in part because it offends some still fairly authoritative notions pertaining to the sacrosanct concept of stability. In 1978, a State Department publication claimed that "it is our policy not to deploy forces which so threaten the Soviet retaliatory capability that they would have an incentive to strike first to avoid losing their deterrent force."59 57. For an unsympathetic but useful review of revisionist arguments on Soviet avil defense, see William H. Kincade, "Repeating History: The Civil Defense Debate Reviewed", International Security, Vol. 2, No. 3 (Winter 1978), pp. 99-120. 58. Richard Burt, "U.S. Moving Toward Vast Revision of Its Strategy on Nuclear War," The New York Times, November 30, 1978, pp. Al, A7; Bernard Weinraub, "Pentagon Seeking Shift in Nuclear Deterrent Policy," The New York Times, January 5, 1979, p. A5; and Richard Burt, "Carter Shifts U.S. Strategy for Deterring Nuclear War," The New York Times, February 10, 1979, p. 5. As Mark Twain said of the story of his alleged death: these reports are highly exaggerated. 59. The next sentence is intriguing. "However, this policy is contingent on similar Soviet

Nuclear Strategy | 47

Counter-military targeting is not, of course, even close to being a novel idea in U.S. war planning. Indeed, one may speculate to the effect that countermilitary targeting already comprises the lion's share of strategic resource allocations in SIOP planning—a thought supported amply by the bevy of official commentaries offered in 1974-75 in support of the "Schlesinger shift" in targeting, largely following the guidance provided by NSDM 242 (of 1974). If there is a shift impending in favor of (still more) large countermilitary strike options in the SIOP, one can speculate that such a shift might imply the paying of heavier attention to Soviet projection forces or, at a more basic level, a commitment to purchasing the ability to neutralize a far higher fraction of really hard Soviet military (and political) targets than is the case at present. For the U.S. Government to endorse a full-fledged war-fighting doctrine in the strategic realm would constitute a doctrinal revolution. Such a doctrine would deny the validity of the stability theory that has informed U.S. defense and arms control policy since the mid-1960s.60 Strategic stability, in the standard formula, requires that societal assets (people and industry) be near totally vulnerable, while strategic weapons be invulnerable. The Soviet Union has always believed in the value of the assured destruction option vis-a-vis the United States, but not in mutual assured destruction. It is too early to be certain, but even if the United States under President Carter might be willing to shift its declaratory focus (and eventually its actual targeting plans)—and to invest in actual military capability—more toward military targets than is the case at present, it is unlikely that it will be able to overcome its funda­ mental skepticism over the wisdom of approaching a central nuclear war as one should approach (or did approach, in pre-nuclear days) non-nuclear war. Pending the occurrence (and resolution in favor of change) of a sophis­ ticated debate over the worth of still-fashionable ideas concerning crisis and arms race instability, American strategic policy will be shifted at the margin rather than rewritten. Also, a particular strategic posture, even one as large

restraint." The Strategic Arms Limitation Talks, Special Report No. 46 (Washington, D.C.: The Department of State, July, 1978), p. 3. This official logic fails, even on its own terms, if U.S. strategic forces are not vulnerable to preemptive attack. 60. But it had its intellectual genesis in the late 1950s. For examples, see Thomas C. Schelling: "Surprise Attack and Disarmament," m Klaus Knorr, ed., NATO and American Security (Prin­ ceton, N.J.: Princeton University Press, 1959), Chapter 8; "Reciprocal Measures for Arms Sta­ bilization," in Donald G. Brennan, ed., Arms Control, Disarmament, and National Security (New York: Braziller, 1961), Chapter 9.

Strategy and Nuclear Deterrence \ 48

as that to be maintained in the 1980s by the United States (with the blessing of SALT II), is not omni-competent. At the present time the United States does not have a strategic posture capable of seeking a military outcome to a war in which Western political authorities could place any confidence. Moreover, on the record extant, the interest of the Carter Administration in purchasing such a posture rapidly has to be judged to be distinctly lukewarm. Carter's record on the character and timing of the MX ICBM program, on the B-I bomber (aborted), and on civil defense, would have to fuel incredulity over the likely postural matching for any proclaimed new strategic doctrine with a war-waging, as opposed to a pre-war deterrent, orientation. If, as (almost certainly over-) reported, the U.S. Government should inch towards a very heavily counter-military stra­ tegic nuclear employment doctrine, it will need to understand the reouirements and limitations of such an approach. A non-defense professional might be somewhat puzzled by this discussion. As a general rule, he might observe, U.S. war planning surely has always been oriented most heavily towards Soviet military targets (strategic forces, projection forces, command and control targets, and war-supporting industry and transport networks)— so what is new? The answer lies in the scope of the military targeting, in its ability to cope with a much harder target set than before, and in its design for separation from civil society. Anybody who sought to argue that the United States suddenly had discovered counter-military targeting as an in­ teresting option, would of course be guilty of misleading his audience. For example, Richard Burt of The New York Times wrote recently that: The Carter Administration has revised the United States strategy for deter­ ring nuclear war by adopting a concept that would require strategic forces to be capable of large-scale precision attacks against Soviet military targets as well as all-out retaliatory blows against cities. The new strategy, which has emerged after months of debate in the Pen­ tagon, represents a significant departure from the long-held concept that the United States needs only to threaten all-out retaliation against Soviet cities to deter Moscow from launching a nuclear attack.61 Clausewitz wrote of war that "[i]ts grammar, indeed, may be its own [i.e., war should be waged in a way that makes military sense, given its unique dynamics], but not its logic."62 A U.S. SIOP oriented towards different kinds 61. "Carter Shifts U.S. Strategy for Deterring Nuclear War", 62. On War, Op. Cit ., p. 605.

op. cit.

Nuclear Strategy | 49

of military targets should be guided by a political logic—what are our war aims? A rewriting and recomputing of the SIOP in an even more heavily counter-military direction than is the case at present could place the United States in a somewhat worse position than that occupied by the (major) allied politicians of World War I—there could be a determination to do military damage to the enemy (which is very sensible), but a lack of commitment to the idea of prosecuting the war to the point where the enemy is defeated militarily (unlike World War I). The question of just how the military damage to be wrought is to be translated into political advantage could easily be evaded. It may be ungenerous to proffer such a negative (or, at least, skeptical) verdict, but it does seem that the official (at least in the Department of Defense) redirection of U.S. strategic nticlear targeting preferences continues to neglect factors that bear upon the issue of desired war outcomes. As noted earlier, a counter-political control strategy cannot succeed unless a Soviet military offensive, at the theater and/or intercontinental levels, is thwarted. However, it would be foolish to wage the military war without taking proper prior account of the Soviet perspective upon Soviet vulnerabilities. There is considerable danger that the United States, looking to the damage promise of an inventory of cruise missiles and (much later) MX ICBMs, will neglect the very important political criteria for strategic targeting. A theory of victory over the Soviet Union can be only partially military in character—the more important part is political. The United States and its allies probably should not aim at achieving the military defeat of the Soviet Union, considered as a unified whole; instead, it should seek to impose such military stalemate and defeat as is needed to persuade disaffected Warsaw Pact allies and ethnic minorities inside the Soviet Union that they can assert their own values in very active political ways. It is possible that a heavily counter-military focused SIOP might have the same insensitivity to Soviet domestic fragilities as may be found-in the counter-economic recovery orientation of the 1970s. In important respects, a heavily counter-military θΙΟΡ would be the kind of war plan that the Soviet Union is well equipped to counter. Notwithstand­ ing its apparent war-waging focus, the American authors and executors of such a doctrine would be unlikely to have considered the conduct of war as a whole; really they would still be seeking, very substantially, to be respon­ sive to pre-war deterrence needs. With a clear political war aim—to encour­ age the dissolution of the Soviet state—much of the military war might not need to be fought at all. The apparently resolute determination of the Amer-

Strategy and Nuclear Deterrence | 50

ican defense community not to think through its deterrence needs, which would involve addressing the question of war aims, promises to produce yet another marginal improvement in doctrine (after all, U.S. strategic forces have always been targeted against Soviet military power—whatever annual Posture Statements may have said).63 It may be worth reminding American policy makers in 1979 that the United States had a counterforce doctrine in Vietnam. A focus upon counter-military action, bereft of an overarching political intent, save of the vaguest kind, is unlikely to serve American interests well, except by unmerited luck. The war-fighting theme which now has limited, though important, official support in Washington, comprises no more than half of the change in think­ ing that is needed. It is essential for pre-war deterrent effect that Soviet leaders not believe they could wage a successful short war. But, for reasons that none could predict in advance, war might occur regardless of the pre­ war theories and the postures of the two sides. In that event, it will be essential that the United States has a theory of war responsive to its political interests.64 Because a counter-military focus in the SIOP is not informed by a clear goal of political victory against the Soviet state, the United States is unlikely to be able to wage an intercontinental nuclear war in a very intelli­ gent fashion. In World War II, American wartime leaders declined to attempt to look beyond the battlefield, so long as the war as still in progress, with results of impressive negative educational value for succeeding generations. How much more intelligent it would be to have explicit war aims that should, in and of themselves, have considerable pre- and intra-war deterrent value. One hesitates to criticize the reported current trend in official thinking, so healthy a change is it in its war-waging focus from the murderous and impolitic counter-economic themes of the 1960s and (most of the) 1970s. Nonetheless, the point has to be made that there continues to be an absence of political judgment overseeing U.S. strategic nuclear employment policy and, ergo, there is a neglect of strategy. A possible change in the 1980s in 63. In his valuable study of the counterforce debate of the early 1970s, Desmond Ball quotes an Air Force general as claiming (in February 1973) that the SIOP was "never reworked under (President) Johnson. It is still basically the same as 1962." Deja Vu: The Return to Counterforce in the Nixon Adminstration, op. cit., p. 17. 64. Such a war might not be tripped by a military accident that related to political intention on neither side. It might pertain to matters of vital interest to both sides. In short, the U.S. defense community might discover that it did have political goals that far transcended Brodie's prediction that the earliest possible war termination would likely be the superordinate objective (see Footnote 54).

Nuclear Strategy | 51

strategic employment orientation towards the counter-military, is fully com­ patible with a U.S. defense community which would not be able to bring itself to think of thermonuclear war in terms of victory or defeat. The U.S. defense community, substantially coerced in its thinking by the adverse trends in the major East-West military balances, has progressed from a counter-economic, to a counter-military focus in its nuclear employment reasoning (although the mechanical details of war planning may well have focused more upon Soviet military assets than the U.S. defense community generally understood to be the case), but it has yet to accept a strategic focus and advance to a counter-political control thesis. Unlike Soviet defense an­ alysts, Western commentators continue to be bemused by the reality-numb­ ing concept of "war termination." Wars are indeed terminated, but they are also won or lost. Moreover, if the U.S. defense community envisages (as it must, realistically) the sacrifice (presumably unwilling) of tens of millions of Americans in a thermonuclear war, that sacrifice should be undertaken only in a very worthwhile cause. If there is no theory of political victory in the U.S. SIOP, then there can be little justification for nuclear planning at all. Stability and the Need for Defense

The principal intellectual culprit in our pantheon of false strategic gods is the concept of stability. For more than fifteen years, influential members of the U.S. defense and arms control community have believed that it is useful, or even essential, that the Soviet Union have guaranteed unrestricted strategic access to American societal assets. Such unrestricted access was believed to have a number of stabilizing consequences. In and of itself it should limit arms competitive activity (such activity as remained would stem from "nor­ mal" modernization and from efforts to offset counterforce-relevant devel­ opments on the other side),65 while—more basically—it should promote some relaxation of tension, in that the Soviet Union would, belatedly, be assured of its ability to deter (through punishment) the United States.66 (This

65. Classic "period-piece" statements of the arms-race stability thesis were George W. Rathjens: The Future of the Strategic Arms Race: Options for the 1970s (New York: Carnegie Endowment for International Peace, 1969); and "The Dynamics of the Arms Race," Scientific American, Vol. 220, No. 4 (April, 1969), pp. 15-25. 66. Strange to note, the theory of arms race dynamics that featured as its centerpiece the proposition that each side acts and reacts in a fairly mechanistic fashion in pursuit of a secure assured destruction capability, has now been discredited pretty well definitively by the historical

Strategy and Nuclear Deterrence | 52

theory has some features in common with the view that the four-fold rise in oil prices in 1973-74 was "good for us"—compelling us to confront the implications of our own profligacy in the energy consumption field.) Analysts of all (or perhaps most) doctrinal persuasions have come at last to accept the view that the Soviet Union does not relax as a consequence of its achieving a very high quality assured destruction capability: the excellent reason for such continued effort is that the assured destruction of American societal assets plays no known role in Soviet deterrent or wartime planning— save as a threat to deter American counter-economic strikes. In addition, Soviet planners probably see considerable political coercive value for a post­ war world in a very large counter-societal threat. Backward though it has seemed to some, the Soviet Union has provided unmistakable evidence of believing that wars, even large nuclear wars, can be won or lost. The massmurder of Americans makes a great deal of sense in terms of the authority structure of a post-war world (since the Soviet Union cannot consummate a victory properly through the physical occupation of North America), but such a grisly exercise has little or nothing to do with the prosecution of a war (save as a counterdeterrent threat). American strategic (and arms control) policy, since the mid-1960s, has been misinformed by stability criteria which rested (and rest) upon a near-total misreading of Soviet phenomena. Soviet leaders are opportunists with a war-waging doctrine as their strategic leitmotiv. Supposedly sophisticated self-restraint in American arms competitive activity, designed so as not to stimulate "destabilizing" Soviet responses, has simply presented the Soviet Union with an upcoming period of strategic superiority of uncertain duration. The American stability theorem held only for so long as both sides endorsed it. There is legitimate dispute today over the quality of Soviet strategic programs, but no one, to the knowledge of this author, disputes the conten­ tion that the Soviet Union is seeking both to protect its societal assets (assured survival, not mutual assured destruction) and to pose the maximum threat to American strategic forces, compatible with the adequate manipulation of Western hopes and fears for the future for the purpose of discouraging a

facts, but the strategic policy premises that flow from that flawed theory have not been over­ hauled thoroughly. Since virtually all U.S. commentators agree that the Soviet Union is not attracted to MAD reasoning, the long-familiar "instability" case against urban-area BMD and non-marginal civil defense provision, is simply wrong. We are still in search of an adequate explanatory model for the strategic arms competition. See Colin S. Gray, The Soviet-American Arms Race (Lexington, Mass.: Lexington, 1976).

Nuclear Strategy | 53

strong American competitive response. Unlike the Soviet Union, the United States has declined to recognize (courtesy of its still-authoritative stability theory) that an adequate strategic deterrent posture requires the striking of a balance between offensive and defensive elements. There is a painful irony of several dimensions in this American intellectual failing. First, among the more pertinent asymmetries that separate the U.S. from the Soviet political systems, is the acute sensitivity of the former to the personal well-being of its human charges. It is little short of bizarre to discover that it is the Soviet Union, and not the United States, that has a serious civil defense program.67 Second, potentially the strongest element in the overall Western stance vis-a-vis the Soviet Union is its industrial mobilization ca­ pacity. Reasonably good American BMD carries healthily terminal implica­ tions for Soviet opportunism or adventure. A BMD system that works well enables the United States to wage a long war and to mass produce the military means for eventual victory. So great is American mobilization po­ tential, vis-a-vis the extant strategic posture, that U.S. defense policy, logi­ cally, should endorse a defensive emphasis. Such an emphasis is the guar­ antor of strategic forces in overwhelming numbers tomorrow. Third, if U.S. strategic nuclear forces are to be politically relevant in future crises, the American homeland has to be physically defended. It is unrea­ sonable to ask an American President to wage an acute crisis, or the early stages of a central war, while he is fearful of being responsible for the loss of more than 100 million Americans. If escalation discipline is to be imposed upon the Soviet Union, even in the direst of situations, potential damage to North America has to be limited. Damage-limitation has to involve both counterforce action and active and passive defenses. The claim that actually to protect (even very imperfectly) Americans and their industry would be destabilizing, is a doctrinal cliche whose shallowness merits uncompromising exposure. Since virtually all Western commentators recognize that the Soviet Union is not moved in its strategic policy by assured destruction criteria, and since no one can deny that an American President could not threaten, or implement, even highly intelligent strikes against the Soviet body politic if American society is totally open to Soviet retaliation, the stability concept is in need of fundamental redefinition. As long as American society is essen­ tially unprotected by BMD, air defense, and civil defense, the United States

67. To have a serious civil defense program does not mean that a country is preparing for war, any more than equipping a ship with lifeboats means that the shipping line is preparing to operate the ship in a dangerous manner.

Strategy and Nuclear Deterrence \ 54

will have to lose any process of competitive escalation against the Soviet force posture anticipated for the 1980s. Fourth, even if the arms controllers' argument were correct, that a defen­ sive emphasis would stimulate the Soviet Union into working harder so as to be able to overcome it through offensive force improvement,68 so what? Generically, the claim that this or that American initiative will catalyze Soviet reactions tends to be accorded far too respectful a hearing. Certainly it is sensible to consider adversary reactions and to take a full systemic look at possibilities, but a country as wealthy (and as responsible for international security) as the United States should not be deterred by the mere prospect of competition from undertaking necessary programs. (For example, an MX ICBM, deployed in a multiple protective structure [MPS] mode, will certainly have some noteworthy impact upon Soviet arms competitive activity, but such recognition does not constitute proof of the folly of deploying MX/MPS).69 Crude though it may sound, the United States would probably achieve more in the field of arms control if it decided to achieve and sustain a politically useful measure of strategic superiority,70 than if it continues its endorsement of the elusive quality known as essential equivalence. Superiority for Stability

If it is true, or at least probable, that a central war could be won or lost, then it has to follow that the concept of strategic superiority should be revived in 68. An argument central to the case against urban-area ABM defense was that its banning by treaty would break the action-reaction cycle of the arms race: the Soviet Union would not need to develop and deploy offensive forces to overcome such an American deployment (in order to preserve their assured destruction capability). It is a matter of history that the ABM treaty banned the ABM defense of U.S. cities, but Soviet offensive force improvements have marched steadily onward. The action-reaction thesis was logical and reasonable; it just happened to be wrong (it neglected the local color, the domestic engines of the arms competition). 69. If U.S. MX/MPS should induce the Soviet Union to proceed down a similar path, then stability (by anyone's definition) would be promoted. Rubles spent on MPSs are rubles not spent on missiles and warheads. It is true that an MPS system in place might attract the Soviet Union to producing large numbers of missiles, undetected, to be surge-deployed in a period of acute need. However, the Soviet Government can produce ICBMs secretly now—m the absence of an MPS system they could be fired from pre-surveyed "soft" sites. The verification argument against MX/MPS is not a telling one, but—as a hedge—deployment of a fairly thin (preferentially assigned) ballistic missile defense system around the MPS could purchase an extraordinary degree of leverage vis-a-vis any secretly (or suddenly) deployed Soviet missiles. See Colin S. Gray, The MX ICBM: Multiple Protective Structure (MPS) Basing and Arms Control, H1-2977-P (Croton-on-Hudson, New York: Hudson Institute, February, 1979). 70. On the political meaning of strategic power, see Edward N. Luttwak, Strategic Power: Military Capabilities and Political Utility, The Washington Papers, Vol. IV (Beverly Hills, California: SAGE, 1976).

Nuclear Strategy | 55

popularity in the West. Superiority has a variety of possible meanings, rang­ ing from the ability to dissuade a putative adversary from offering resistance (i.e., deterring a crisis), through the imposition of severe escalation discipline on opponents, to a context wherein one could prosecute actual armed conflict to a successful conclusion. There is certainly no consensus within the United States defense community today over the issue of whether or not any central war outcome is possible which would warrant description as victory. How­ ever, a consensus is emerging to the effect that the Soviet Union appears to believe in the possibility of victory, and that the time is long-overdue for a basic overhaul of our intellectual capital in the nuclear deterrence field.71 At the very least, most defense analysts would endorse the proposition that it is important for the United States to be able to deny the Soviet Union victory on its own terms. There is a need for Western strategic thinkers to address and overcome the emerging tension between the (probable) requirements of high-quality deterrence, and the still-authoritative and inhibiting ideas of crisis and arms race instability which have directed the U.S. defense community away from programs that speak to Soviet reality. A false choice has misinformed the structure of our thinking. The historical record of the arms competition since the mid-1960s shows that the choice has not been between, on the one hand, a measure of U.S. restraint which would facilitate Soviet acquisition of an assured destruction capability—an achievement that would promote the prospects of arms control negotiations intended to codify a more stable strategic environment—and, on the other hand, an absence of American re­ straint which would serve to stimulate Soviet countervailing programs and which would be doomed to failure anyway. Instead, the choice has been between a measure of American restraint which facilitated the Soviet drive to achieve a not implausible war-winning capability, and a relative absence of restraint which would greatly complicate the life of Soviet defense plan­ ners. Overall, the evidence suggests that the Soviet Union has not been

71. Noteworthy endeavors since the mid-1970s include: Robert Jervis, "Deterrence Theory Revisited," World Politics, Vol. XXXI, No. 2 (January, 1979^ pp. 289-324; Patrick Morgan, Deterrence: A Conceptual Analysis (Beverly Hills, California: SAGE, 1977); and Richard Rosecrance, Strategic Deterrence Reconsidered, Adelphi Paper No. 116 (London: IISS, Spring 1975). Even wellconsidered judgments published as recently as 1975 can look a little fragile in 1979. Consider these words of Professor Rosecrance: "Thus it is possible to say that although the deterrent requirements that were deemed necessary to protect Europe in the 1950s are probably not currently being met, they may not have to be met. An improvement in the Soviet-Western relationship makes them less necessary now than they were then." Ibid., p. 36.

Strategy and Nuclear Deterrence | 56

seeking a deterrent, as that concept and capability has been (mis?)understood in the West for nearly twenty years. The choice the United States confronts today is whether or not it will tolerate Soviet acquisition, unmatched, of an emerging war-fighting capability which might, with some good judgment and some luck, produce success in crisis diplomacy and in war. The instability arguments that are leveled against those who urge an American response (functionally) in kind-are somewhat fragile. For example, there is good reason to believe that the Soviet Union would be profoundly discouraged by the prospect of having to wage an arms competition against an American opponent no longer severely inhibited by its long-familiar sta­ bility theory. In addition, an American war-fighting-oriented strategic pos­ ture, if well-designed, should not contribute to crisis instability. The fact that the United States might pose a theoretical first-strike threat to much of the Soviet strategic posture, should not give aid and comfort to the "use them or lose them" argument. A central purpose informing U.S. strategic posture would be its denial of any plausible Soviet theory of victory. Why the Soviet Union would be interested in starting a war that it would stand little, if any, prospect of winning is, to say the least, obscure. The contemporary debate over strategic doctrine, whatever its eventual effect may be upon U.S. war planning and declaratory policy, has registered a qualitative advance over most of the strategic thinking of the past fifteen years. The debate has focused upon what might be needed to deter Soviet leaders, qua Soviet leaders, and some (still unduly limited) attempt has been made to consider operational, as opposed to pre-war declaratory, strategy. Theories of pre-war deterrence, however sophisticated, cannot guarantee that the United States will never slip into an acute crisis wherein a president has to initiate strategic nuclear employment or, de facto, surrender. In such a situation, a president would need realistic war plans that carried a vision of the war as a whole and embodied a theory of how military action should produce desired political ends. In short, he would be in need of strategy. Fortunately, still orthodox wisdom notwithstanding, there is no necessary tension between a realistic wartime strategy (and the posture to match) and the pre-war deterrence of undesired Soviet behavior.

Deterrence and Perception

Robert

Jervis

In the most elemental sense, deterrence depends on perceptions. But unless people are totally blind, we need not be concerned with the logical point that, if one actor's behavior is to influence another, it must be perceived. Rather what is im­ portant is that actors' perceptions often diverge both from "objective reality" (or later scholars' perceptions of it, which is as good a measure as we can have) and from the perceptions of other actors. These differences, further­ more, both randomly and systematically influence deterrence. Unless states­ men understand the ways in which their opposite numbers see the world, their deterrence policies are likely to misfire; unless scholars understand the patterns of perceptions involved, they will misinterpret the behavior. An example both shows that the problem extends to perceptions of third parties as well as main adversaries and underlines the way in which attempts at deterrence can not only fail but backfire if the assumptions about others' perceptions are incorrect. In order to mobilize British assistance in the Amer­ ican-Japanese political conflict of 1907-1908, President Theodore Roosevelt sought to portray the situation as quite tense. He expected that Britain would then aid him by restraining Japan. Unfortunately, and contrary to the Pres­ ident's assumption, the British perceptions of both him and the Japanese differed from those that Roosevelt held: "The British felt it was Washington, not Tokyo, which stood in need of a warning." As Hardinge, the Permanent Under Secretary of the Foreign Office, put it: "the President is playing a very dangerous game, and it is fortunate that he has such cool-headed people as the Japanese to deal with."1 Thus, rather than moving Britain closer to the An earlier version of this article appeared as Research Note No 11 of the Center for International and Strategic Affairs, the University of California, Los Angeles. It will be included in the forthcoming volume National Security and International Stability, ed. Bernard Brodie, Michael Intnligator, and Roman Kolkowicz (Cambridge, Mass.: Oelgeschlager, Gunn, & Hain, Publish­ ers, 1983), and appears here by permission of the author, editors, and publisher.

Robert Jervis is a Professor in the Department of Political Science and the Institute of War and Peace Studies, Columbia University, and the author of Perception and Misperception in International Politics and The Logic of Images m International Relations.

1. Charles Neu, An Uncertain Friendship (Cambridge: Harvard University Press, 1967), p. 199.

International Security, Winter 1982/1983 (Vol 7, No. 3) 0162-2889/83/030003-28 $02 50/0 © 1983 by the President and Fellows of Harvard College and of the Massachusetts Institute of Technology.

Strategy and Nuclear Deterrence | 58

United States, as Roosevelt expected, his actions made that country less willing to cooperate in opposing Japan. In light of the dangers inherent in misperceptions, one might expect that statesmen would pay careful attention to how others perceive them. In fact, this is usually not the case. While they are aware that determining others' intentions and predicting others' behavior is difficult, they generally believe that their own intentions—especially when they are not expansionist—are clear. As a result, they rarely try to see the world and their own actions through their adversary's eyes,2 although doing so would be to their advan­ tage. If a policy is to have the desired impact on its target, it must be perceived as it is intended;3 if the other's behavior is to be anticipated and the state's policy is a major influence on it, then the state must try to determine how its actions are being perceived. One would think, therefore, that every gov­ ernment would establish an office responsible for reconstructing the other's view of the world and that every policy paper would have a section that analyzed how the alternative policies would be seen by significant audiences. One theme of this essay is that the failure to undertake this task—and I do not mean to imply that it would be easy to accomplish—explains many cases of policy failure. It is hard to find cases of even mild international conflict in which both sides fully grasp the other's views. Yet all too often statesmen assume that their opposite numbers see the world as they see it, fail to devote sufficient resources to determining whether this is actually true, and have much more confidence in their beliefs about the other's perceptions than the evidence warrants. Misperception and the Failure of Deterrence One actor deters another by convincing him that the expected value of a certain action is outweighed by the expected punishment. The latter is com­ posed of two elements: the perceived cost of the punishments that the actor

2. The British tried to do this, with some success, during World War II. See Donald McLachlan, Room 39 (New York: Atheneum, 1968), pp. 252-258. ό. Of course accidents can lead to desired ends in ways decision-makers had not intended, but I do not think this is common. One example may be the U.S. Navy's unauthorized harassment of Soviet submarines in the Cuban Missile Crisis which probably helped convince the Soviet leaders that the confrontation was too dangerous to be permitted to continue. See Alexander George, David Hall, and William Simons, The Limits of Coercive Diplomacy (Boston: Little, Brown, 1971), pp. 112-114.

Deterrence and Perception \ 59

can inflict and the perceived probabilities that he will inflict them. Deterrence can misfire if the two sides have different beliefs about either factor. (MIS)PERCEPTIONS OF VALUE

Judging what constitutes harm is generally easier than estimating whether threats will be carried out, but even here there is room for differences which can undermine deterrence. On occasion, what one person thinks is a pun­ ishment another may consider a reward. The model is Br'er Rabbit. Only rarely do states in international politics want to be thrown into a brier patch; but Teddy Roosevelt's threat to intervene in the Cuban internal conflict of 1903 comes close. He declared that, if American property were raided in the course of the fighting, he would have to send in troops. Unfortunately, both factions believed that American intervention would work in their favor and busily set to work harassing Americans and their property.4 One could not have coerced Pol Pot by threatening to destroy his cities and a similar, if less extreme, point lies behind some of the current U.S. strategic policy. As former Secretary of Defense Harold Brown has argued, the United States must "take full account of the fact [s;c] that the things highly valued by the Soviet leadership appear to include not only the lives and prosperity of the peoples of the Soviet Union, but the military, industrial and political sources of power of the regime itself."5 This requires targeting the army, internal security forces, and the Communist Party. A related argument is that the Soviet leaders are ethnic Russians who care about maintaining the dominance of Great Russia and who would be deterred by

4. Allan Millet, The Politics of Intervention (Columbus, Ohio: Ohio State University Press, 1968). The point is nicely made in an anecdote about a British General made by B.H. Liddell Hart: Jack Dill was a delightful man for any enthusiast to meet or serve. But he was quite unable to understand that the average officer did not share his burning ardour for professional study and tactical exercises. An illuminating example of that incomprehension occurred in his way of dealing with the major commanding a battery attached to his brigade who had failed to show the keenness Dill expected. To emphasize his dissatisfaction Dill told this officer that he would not be allowed to take part in the remaining exercises—a punishment, drastic in Dill's view, which was a great relief to the delinquent, who had been counting the days until he could get away to join a grouse-shooting party in Scotland. —The Liddell Hart Memoirs, 1895-1938 (New York: Putnam's, 1965), p. 72. 5. Harold Brown, Department of Defense Annual Report Fiscal Year 1981 (Washington, D.C.. U.S Government Pnnting Office, 1980), p. 67. It should also be noted that if these arguments are correct, the threat to carry out these attacks would be no more credible than the threat to attack Soviet cities because there would be no reason for the Soviet response of retaliating against American cities to be different.

Strategy and Nuclear Deterrence \ 60

the threat to attack it but spare the other areas of the USSR, thereby enabling the other nationalities to rise up and either gain their independence or dominate the postwar state. Without endorsing the answers he provides, one can completely agree with Brown's argument that "our strategy has to be aimed at what the Soviets think is important to them, not just what we might think would be important to them."6 But this kind of analysis must be carried to its logical conclusion, not stopped at a point which is convenient to the analyst's political predilections. To argue that the Russians could be deterred by threatening to destroy the party and internal security forces implies not only that these instruments are needed to maintain Communist rule, but also that the Soviet leaders realize this. This may be correct, but if they believe what they say, they will think the regime enjoys the support of the population and so might conclude that the party would regenerate after the war. American leaders do not think that the destruction of the state apparatus in a war would permanently end democracy; would the Soviets have so little faith in their regime that they would lack comparable beliefs? As we have seen, threats of coercive war can misfire if the state does not understand what the opponent values. Threats to use brute force, on the other hand, do not involve this pitfall, but they do require the state to determine how its adversary evaluates the military balance—how it estimates who would win a war. This issue arose in the 1930s as the British leaders debated how to deter Hitler. Some felt that "economic stability"—which required that military spending be kept relatively low—contributed to this goal: "The maintenance of our economic stability . . . [could] be described as an essential element in our defense system . . . without which purely military effort would be of no avail. . . . Nothing operates more strongly to deter a potential aggressor from attacking this country than our stability. . . . This reputation stands us in good stead, and causes other countries to rate our power of resistance at something far more formidable than is implied merely by the number of men of war, aeroplanes and battalions which we should have at our disposal immediately on the outbreak of war. But were other countries to detect in us signs of strain, this deterrent would at once be lost."7 On the other hand, Churchill stressed the need for larger military

6. U.S., Congress, Senate, Committee on Foreign Relations, Hearings on Nuclear War Strategy (Washington, D.C.: U.S. Government Printing Office, 1981), p. 10. 7. Sir Thomas Inskip, Minister for Coordination of Defense, quoted in Martm Gilbert, Winston S. Churchill, Vol 5 1922-1939 (London: Heinemann, 1976), p. 891.

Deterrence and Perception | 61

forces: "an immense British army cast into the scales" was a great deterrent "and one of the surest bulwarks of peace."8 Neither side in the argument, however, tried as hard as it might have to learn exactly how Hitler saw the world and what sort of configuration of forces might have deterred him. Deterrence can also be undercut if the aggressor does not understand the kind of war which the status quo state is threatening to wage. The Japanese had no doubt that the United States would fight if they attacked Pearl Harbor. But many of Japan's leaders thought that the stakes for the U.S. were not sufficiently high to justify an all-out effort and that the Americans would instead fight a limited war, and, being unable to prevail at that level of violence, would agree to a settlement which would give Japan control of East Asia. Similarly, Hitler expected Britain and France to fight in September 1939 but doubted that they would continue to do so after Poland was defeated. Britain especially, he believed, had sufficient common interest with Germany to conclude a peace treaty after limited hostilities. In neither case did either side understand the other's beliefs or values. Indeed, the German and Jap­ anese perceptions of their opponents would have seemed to the latter so out of touch with reality as to hardly deserve consideration. British and American statesmen knew their own outlooks so well that they thought it obvious that others knew them also. To have recognized that alternative views were possible would have implied that their self-images were not unambiguously correct and that their past behavior might be interpreted as indicating a willingness to sacrifice friends and agree to less than honorable settlements. Because Britain, France, and the United States did not understand the other side's expectations, their deterrence strategies could not be effective. Their task was not only to convince their adversaries they would fight if pushed too far, but also that they would continue to fight even after initial reverses.9 Doing this would have been extremely difficult since it would have involved presenting evidence and making commitments about how they would behave a few years later under grave circumstances. But had the statesmen been aware of the German and Japanese perceptions, they might

8. Quoted in ibid, p. 945. 9. Churchill had a better understanding of the problem. In 1938 he stressed to a German diplomat that "a war, once started, would be fought out like the last to the bitter end, and one must consider not what might happen in the first few months, but where we should all be at the end of the fourth year." Quoted in ibid., p. 964. For a related argument, see Alan Alexandroff and Richard Rosecrance, "Deterrence in 1939," World Politics, Vol. 29, No. 3 (April 1977), pp. 404^424.

Strategy and Nuclear Deterrence | 62

have at least made some efforts. For example, President Franklin Roosevelt could have stressed the American tradition of vacillating between isolation and extreme involvement in international politics, of seeing the world in Manichean terms, of fighting only unlimited wars. Prime Minister Chamber­ lain might have done better explaining why he had abandoned appeasement, why Britain could not allow any power to dominate the continent, and why it would have no choice but to resist even if the military situation was bleak. Similarly, throughout the 1960s, the U.S. misjudged how much North Vietnam valued reunification and believed that an American threat to fight a prolonged war and inflict very heavy punishment on the North10 could dissuade the North from continuing its struggle. American decision-makers paid a great deal of attention to how to make their threats credible, but their misjudgment led them to ignore what was actually the crucial problem—that the North was willing to fight the sort of war the U.S. was threatening rather than concede. The Americans might not have been able to solve the problem even had they been aware of it, but as it was they never even came to grips with it. (MIS)PERCEPTIONS OF CREDIBILITY

Misperceptions of what the target state values and fears probably are less important causes of deterrence failure than misperceptions of credibility. Conclusions are difficult to draw in this area, however. Although many arguments about deterrence turn on questions involving credibility, scholars know remarkably little about how these judgments are formed and altered. For example, how context-bound are these estimates? Obviously the credi­ bility of a threat is strongly influenced by the specific situation in which it is issued. The threat to go to war in response to a major provocation could be credible when the threat to so respond to a minor insult would not. But there also is a component of credibility that inheres in the threatener, not the situation. In the same circumstance, one country's threat can be credible where another's would not be. Part of this difference of course comes from the country's strength, its ability to carry out the threat, and its ability to

10. As Walt Rostow put it, "Ho has an industrial complex to protect; he is no longer a guerilla fighter with nothing to lose." Quoted in Department of Defense, Pentagon Papers, Senator Mike Gravel, ed. (Boston: Beacon Press, 1971), Vol. Ill, p. 153. That North Vietnam absorbed almost unprecedented punishment is shown by John Mueller, "The Search for the Single 'Breaking Point' in Vietnam: The Statistics of a Deadly Quarrel," International Studies Quarterly, Vol. 24, No. 4 (December 1980), pp. 497-519.

Deterrence and Perception | 63

defend against the other's response. But there's more to it than this. Some states have reputations for being bolder, more resolute, and more reckless than others. That is, states are seen to differ in the price they are willing to pay to achieve a given goal. But it is not clear how these reputations are established and maintained or how important they are compared to the other influences on credibility. We cannot predict with great assurance how a given behavior (e.g., refusing to change one's position on an issue) will influence others' expectations of how the state will act in the future. To start with, does reputation attach to the decision-maker, the regime, or the country? If one president acts boldly, will other states' leaders draw inferences only about him or will they expect his successors to display similar resolve? After a revolution, do others think the slate has been wiped clean or does the reputation of the earlier regime retain some life? If one kind of regime (e.g., a capitalist democracy) displays willingness to run high risks, do others draw any inferences about the resolve of similar regimes? How fast do reputations decay? On these points we have neither theoretically grounded expectations nor solid evidence. In another area, we at least can be guided by a good theory. One of the basic findings of cognitive psychology is that images change only slowly and are maintained in the face of discrepant information. This implies that trying to change a reputation for low resolve will be especially costly because statements and symbolic actions are not likely to be taken seriously. Only the running of what is obviously a high risk or engaging in a costly conflict will suffice. On the other hand, a state with a reputation for standing firm not only will be able to win disputes by threatening to fight, but has the freedom to avoid confrontations without damaging its image. But these propositions, although plausible, still lack empirical evidence. The question of the relative importance of beliefs about the state's general resolve as compared to the role of other factors is also impossible to answer with any precision. How much do states make overall judgments about the prices others are willing to pay as opposed to looking primarily at the specific situation the other is in? In other words, how context-bound are estimates of how others will behave?11 The debate over the validity of the domino theory reminds us both of the importance of this topic and the difficulty of 11. This question can be linked to the "levels of analysis" problem in international politics— i.e., the question of whether the main causes of a state's behavior are to be found in its internal characteristics or its external environment—but a full discussion would take us far afield.

Strategy and Nuclear Deterrence | 64

coming to grips with it. If others were more impressed by America's eventual defeat in Vietnam than by the fact that it was willing to fight for years for a country of little intrinsic value, they would adjust downward their estimate of American resolve. But by how much? If there is another Berlin crisis, will the Vietnam-influenced reputation be as significant as others' judgment of the value of Berlin to the U.S.? When the new situation closely resembles a previous one in which the actor displayed low resolve, others are likely to expect similar retreat.12 But when the situation is very different, it is not clear whether a judgment of the state's overall resolve has much impact on others' predictions of its behavior. Even when these questions are not hypothetical, they are usually hard to answer, as is illustrated by the ambiguous nature of the events that followed the American defeat in Indochina. Has the Soviet Union drawn far-reaching inferences from the American retreat? Have the NATO allies lowered their estimates of the probability that the United States would respond to Soviet pressure or military moves in Europe? Have the Third World countries come to see the U.S. as less reliable? Since 1975, the Soviets have taken a number of actions inimical to American interests, the Europeans have voiced doubts about the credibility of the U.S. promise to protect them, and Third World states have been quite troublesome. But these problems do not present a sharp break from the pre-Vietnam era. It is easy to attribute any behavior contrary to American wishes to the lack of resolve which some observers think the U.S. displayed in Indochina. But it is much harder to establish that this is a better explanation than local conditions or general trends such as the increase of Soviet power. We can turn this example around and ask about the impact of the U.S. attempt to rescue the hostages in Iran. Others probably have raised their estimates of the likelihood that the United States would respond similarly in other cases in which American citizens were taken prisoner. But have per­ ceptions of American resolve to run risks in other kinds of situations been altered? One of the main arguments in favor of using force was that they would be, that U.S. promises and threats would be more credible. But scant evidence supports this view. The cost the U.S. foresaw in this case was not Soviet intervention, but adverse Third World reaction; would others expect the United States to act strongly in later situations when the costs to be

12. For a paradoxical exception to this generalization, see discussion of Mayaguez incident below.

Deterrence and Perception \ 65

incurred were of a very different kind? Others would draw such an inference if they employed the concept of "willingness to incur costs" or "propensity to act with boldness" as a homogeneous category. They might, of course, be correct to do so. The willingness to act in the face of Third World opinion might be linked to a willingness to defy the threat of a Soviet military response. But we know little about whether such global characteristics are possible or whether statesmen make them. One can also ask whether the inference would have been different had the rescue mission succeeded, or had it resulted in the death of the hostages. Ironically, this logic dictates that the impact on U.S. credibility would have been greater in the latter case than in the former. Had the hostages been killed, observers would probably have thought that the American leaders knew the operation was terribly risky. If they projected this pattern of risktaking onto later events, they would conclude that the U.S. would act even when it might not succeed. By contrast, if force had succeeded and others assumed that the Americans had been confident that this was going to be the result, they would not see the act as so bold. I admit this argument is strained, and indeed I doubt that observers would follow the train of rea­ soning I have presented. But this uncertainty underscores the difficulty of determining the inferences people do draw in these situations. The crucial question is the degree to which observers make general judg­ ments about others' credibility rather than basing their predictions largely on the nature of the specific situation and, if the situation is a continuing one, on the history of the other's behavior concerning it. To a significant extent, deterrence theory rests on the assumption that such general judgments are important. It is this which makes it both possible and necessary for a state to credibly threaten to react to an attack on an unimportant third country by a response which will involve greater costs than the intrinsic value of the third area. Such a threat can be credible because what the state will lose by not responding is not just the third country, but also its reputation for protecting its interests, a reputation that is more valuable than the costs of fighting. By the same logic, this response is necessary, because to fail to rise to the challenge is to lead others to doubt the state's willingness to pay costs to defend the rest of the status quo. Both prongs of this reasoning depend on actors' making relatively context-free judgments of credibility. Even if they do, the way in which these judgments are made can defeat significant aspects of the theory and practice of deterrence. When an actor either carries out or reneges on a threat, observers can make either or both

Strategy and Nuclear Deterrence \ 66

of two kinds of inferences that will influence his future credibility. First, they may alter their estimate of what I have elsewhere called his "signaling rep­ utation"—i.e., his reputation for doing what he says he will do. 13 The bar­ gaining tactic of commitment, so well known in deterrence literature, is supposed to be effective because the state increases its cost of retreating by staking its reputation on standing firm. But this tactic will work (and this explanation of actors' behavior will be appropriate) only if actors try to determine how likely it is that others will live up to their promises and threats rather than predicting their behavior solely on the basis of estimates of what they value and the prices they are willing to pay to reach various objectives. This is the second kind of inference actors draw from others' past behavior. It ignores statements and other signals that can be easily manip­ ulated and looks only at whether the other stood firm, compromised, or retreated in the past, irrespective of what he said he would do. If this kind of inference is dominant, then signals of commitment have little impact. To use Schelling's terms, actors would be able to issue warnings, but not threats. 14 This would mean that an actor could not deter others by symbolic­ ally committing himself to a course of action and staking his reputation on living up to his pledges. Finally, an ironic possibility should be noted. A concern for reputation can lead states to act and draw inferences in a pattern opposite from the one that we—and most other analysts—imply. This is not to dispute the common starting point; states often refuse to back down not because of the immediate and direct costs of doing so, but because of the belief that a retreat will be seen as an indication of general weakness and so lead others to expect retreats in the future. But the desire to counteract such undesired conse­ quences may lead a state that has retreated on one issue to pay especially high costs to avoid defeat on the next one. Thus the United States was not only willing but anxious to use force to free the Mayagueζ because it wanted to show others that its evacuation of Indochina did not mean that it would not defend its other interests—the very consequence which it had predicted would follow from a defeat in Vietnam and which justified its participation in the war. If others understand this logic and expect states to behave in this

13. Robert Jervis, The Logic of Images in International Relations (Princeton: Princeton University Press, 1970), pp. 20-26, 66-112. 14. Thomas Schelling, The Strategy of Conflict (New York: Oxford University Press, 1963), pp. 123-124.

Deterrence and Perception | 67

way—to follow retreats with displays of firmness—then reputations for carry­ ing out threats do not influence estimates of credibility because—to com­ pound the paradox—reputations are so important that states must rebuild them when they are damaged. If you have been caught bluffing in poker, are others likely to call you in the next round in the belief that you bluff a lot or are they unlikely to do so because they think that you know it is no longer safe to bluff? To the extent that the latter is the case, perceptions of credibility are influenced by the state's recent behavior, but in a way which produces equilibrating negative feedback rather than the positive feedback of the dom­ ino dynamics. 15 JUDGING THE ADVERSARY'S ALTERNATIVES

Deterrence works when the expected costs of challenging the status quo are greater than those of accepting it; deterrence may fail and defenders be taken by surprise not only if their threats are insufficiently credible or directed at the wrong values, but also if the defenders fail to grasp the expansionist's dismal evaluation of the alternatives to fighting. Although the deterring state realizes that its adversary has strong incentives to take action—or else deter­ rence would not be necessary—it usually thinks that the latter has a wide range of choice. Furthermore, the deterring state almost always believes that the adversary is tempted to act because of the positive attraction of the gains he hopes to make. In fact, however, the other state often feels that it has little choice but to act because, if it does not, it will not merely forgo gains, but will suffer grave losses. 16 Status quo powers often underestimate the pressure that is pushing the other to act and therefore underestimate the magnitude of threat and/or the degree of credibility that will be required to make the other refrain from moving. The pressures felt by Japan in the fall of 1941 and by China in the fall of 1950 illustrate why the target state can feel it must act even though it knows some sort of war will result. China and Japan perceived the alternative to fighting not as maintaining the status quo—which was tolerable—but as permitting a drastic erosion of the posi­ tions they had established. Because the status quo states did not understand 15. It is possible, of course, that under some circumstances a retreat leads statesmen to expect other retreats and that under other conditions they draw the opposite inference, but we do not know enough to specify the conditions. 16. Ole Holsti, "The 1914 Case," American Political Science Review, Vol. 59, No 2 (June 1965), pp 365-378; Richard Ned Lebow, Between Peace and War The Nature of International Conflict (Baltimore: Johns Hopkins University Press, 1981).

Strategy and Nuclear Deterrence | 68

this, they did not grasp the difficulty of the job of deterrence that they were undertaking. This is one reason why they thought that their superior power was clearly sufficient to keep the adversary at bay. The case of the Chinese entry into the Korean War is especially striking since the United States did not even grasp the Chinese fear that, if the U.S. conquered North Korea, it would threaten China. American leaders had no such intention and thought this clear to everyone, just as they felt that their unwillingness to fight a limited war in 1941 was clear to all. Again, not only was there a major difference in perceptions, but one of which both sides were unaware. Deterrence failed; but more than this, the deterrence strategy could not be adequately crafted since it was not based on a correct assess­ ment of what the other side valued and feared. Similarly, the basic question of whether deterrence was possible was not adequately faced. In neither instance did the United States consider that even a well-developed deterrence policy might fail and therefore that it should balance the costs of war against the costs of making concessions; since deterrence seemed likely to succeed, the painful alternative of sacrificing some values and abandoning some for­ eign policy goals was not to be taken seriously. Self-Deterrence The previous sections provided some reasons why inaccurate or conflicting perceptions can lead to failures of deterrence. Most treatments of this subject deal with cases like surprise attack in which statesmen incorrectly believe that they have deterred others. While this problem is fascinating and important, we should not neglect the less dramatic other side of this coin: states can successfully deter others unintentionally or unknowingly. Because actors can perceive things that are not there, they can be deterred by figments of their own imagination—"self-deterrence," if you will. An example is the British fear throughout the 1930s that Germany would wipe out London at the start of a world war.17 Although the Germans fed this fear by exaggerating their air strength, the enormity of the gap between the British beliefs and the German activities indicates that most of the explanation must lie with the former's perceptual predispositions. Ingenious deception schemes rarely work unless they fit with what the target already believes. 17. The most thorough treatment is Uri Bialer, The Shadow of the Bomber: The Fear of Air Attack and British Politics, 1932-1939 (London: Royal Historical Society, 1980).

Deterrence and Perception \ 69

The British made two notable errors. First, they greatly overestimated the damage that would be caused by each ton of bombs dropped. Perhaps even more startling than the fact that their estimate was off—by a factor of 25—is the low level of effort that they put into developing the estimate. 18 Since British policy rested in significant measure on the belief that war would entail what would later be called "unacceptable damage," one would think that great care would have been devoted to estimating how much damage aerial bombardment would cause. In fact, almost all British analyses rested on a simple and badly biased extrapolation from the few raids on London during World War I. No competing studies were generated; no alternative sources of data or methods were used. This error was compounded by a fundamental misreading of German air policy and air strength. The British belief that Hitler had the intent and the capability to make British cities his prime target was incorrect on both counts. The German air force was predominantly designed to support ground troops. Doctrine, plans, and aircraft for strategic bombardment did not exist.19 The effort Germany mounted in the summer of 1940 in circumstances that neither side anticipated was an improvised one. Part of the explanation for these errors is that the German bombing raids of the First World War left a strong imprint on the decision-makers. The public had demanded greater protection and panic had been a significant problem. But I do not think that purely cognitive or unmotivated factors were of primary importance. That is to say, the misperceptions and miscal­ culations cannot be accounted for by innocent intellectual and informationprocessing errors—such as mislearning from history—that would have been corrected had they been pointed out to the decision-makers. Rather, the errors served important functions and purposes for those who were making them. To a significant extent, the errors were motivated ones, in the sense of being useful to the actors, of facilitating valued actions, positions, or atti­ tudes. We usually adduce perceptions and calculations as proximate expla-

18. For a good discussion, see Paul Bracken, "The Unintended Consequences of Strategic Gaming," Simulation and Games, Vol. 8 (September 1977), pp. 300-315. 19. Even during the first years of the war, Hitler did not pay careful attention to the bombing campaign against Britain. See R.J. Overy, "Hitler and Air Strategy," Journal of Contemporary History, Vol. 15 (July 1980), pp. 410-412. Later Hitler placed great faith in the new terror weapons, the V-I and the V-2, but he never analyzed the probable effect of these weapons with any care. For an argument that takes partial exception to the view expressed here, see Williamson Murray, "The Luftwaffe Before the Second World War: A Mission, A Strategy?" Journal of Strategic Studies, Vol. 4 (September 1981), pp. 261-270.

Strategy and Nuclear Deterrence \ 70

nations of decisions. But in this case the main causation runs the other way: the pessimistic assessments of German bombing were as much the product of policies as they were a cause of them. This seems a particularly odd argument in this context because the deci­ sion-makers were conjuring up mythical threats that restricted their country's freedom of action and eventually undermined its security. Nevertheless, different sectors of the British elite had different reasons for finding the fear of bombardment congenial. The Royal Air Force (RAF), which produced and analyzed much of the intelligence on which the estimates were based, was predisposed to believe in a potent German bombing threat because its iden­ tity as a separate service rested on the efficacy of strategic bombardment. To have recognized that the German air force's main mission was ground sup­ port would have introduced the question of whether Britain's air force should not be similarly employed. (For the same reason, the RAF resisted the idea that defense against bombers might be possible and insisted that counterbombardment was the only effective peacetime deterrent and wartime strat­ egy. It was the civilian leaders, especially the Minister for Co-ordination of Defence, Sir Thomas Inskip, who saw that changing technology allowed fighters to destroy a sufficient proportion of bombers to make defense against prolonged bombing feasible.) Proponents of both major foreign policy positions—appeasement and antiappeasement—also had reasons to accept the pessimistic air estimates. For the appeasers, the estimates were useful by showing that the costs of war would be terribly high, thus indicating the need for international conciliation. The British could contemplate opposing Germany only if they were sure that their vital interests were at stake. If the issue were only the British abhorrence of the German domestic regime and its uncouth behavior or the mere pos­ sibility that German aims were unreasonable, confrontations were too costly to be justified. Furthermore, if the threat were from the air, the British re­ sponse had to be in the same realm. Little money could be spared for the other services, especially the army. This fit the appeasement policy nicely because a defense posture based on air power would limit spending and facilitate a foreign policy that would remain within British control rather than requiring close cooperation with allies. Before 1914 the cabinet had become partly committed to France through joint naval planning and, when it de­ cided for war, it found that the only war plan available subordinated the British Army to the French. In the 1930s, such cooperation would imply prewar ties which could interfere with appeasement and drag England into

Deterrence and Perception | 71

a dangerous anti-German stance. This danger could be avoided by a military policy that shunned a large army. Ironically, the anti-appeasers also had reasons to overestimate German air strength. They thought Hitler was highly aggressive and therefore expected him to build what they thought would be a maximally effective air force. Failing to see the German weaknesses and inefficiencies, they expected the air fleet would be larger than it was. Being preoccupied with their own fears—they vastly underestimated the staying power of the working class— they were sure that Germany was planning to rely on weapons of terror. A month after Hitler came to power, Robert Vansittart, the Permanent Under Secretary of the Foreign Office, argued that the Germans were "likely to rely for their military power . . . on the mechanical weapons of the future . . . and above all [on] military aircraft. . . . Aviation in particular offers Germany the quickest and easiest way of making their power effective."20 "It must. . . be remembered," Churchill said in 1936, "that Germany has specialized in long-distance bombing aeroplanes."21 This misreading also fit nicely with the attempts to mobilize the British public. The greater the German air force, the greater the British air force should be. Furthermore, the high estimates im­ plied that Germany was aggressive, since it was building more than its defense required. The British, then, did much of Hitler's work for him. While he did seek to deter Britain, the British perceptions cannot be completely explained by the German behavior. British fantasies, developed by different groups for differ­ ent reasons, inhibited accurate analysis of the German air threat and led decision-makers to accept pessimistic views. As a result, the fact of deterrence far outran the German policy of deterrence. Current American fears about Minuteman vulnerability and Soviet nuclear "superiority" may be similar examples of self-deterrence. Some argue that if the Soviets could destroy many of America's strategic forces by using a relatively small proportion of their missiles (an outcome made possible by MIRVs, a technology ironically pioneered by the U.S.), they might start a war either in the hope of gaining world dominance or, more likely, in a preemptive blow during a crisis in which they feared a grave political setback

20. Quoted in D.C. Watt, "British Intelligence and the Coming of the Second World War in Europe: The Assessment of the Enemy," in Ernest May, ed., Knowing One's Enemies: Intelligence Assessment Before the Two World Wars, forthcoming. 21. Quoted in Gilbert, Winston S. Churchill, p. 797.

Strategy and Nuclear Deterrence | 72

or even an American first strike. A related argument is that the Soviets are gaining a "war-fighting" ability that significantly exceeds America's. The Soviet Union, some fear, could do much better than the United States in all levels of warfare; the American knowledge of this Soviet advantage places it in a situation not unlike that of England in the 1930s. Rather than debating the validity of this position,22 I want to raise the question of whether the analogy of the 1930s applies in the less obvious aspect I discussed earlier. While it is unlikely that statesmen are now re­ peating the error of overestimating the casualties from bombing, U.S. com­ mentators are creating self-deterrence because the scenarios they are contem­ plating probably are mythical. The best example is a Soviet attack on Minuteman silos and other U.S. strategic forces. Although abstract American models may indicate that these forces are vulnerable, these calculations in­ volve several simplifying assumptions—e.g., that the Soviets could fire a carefully coordinated salvo of hundreds of missiles, that the figures for accuracies derived from firings over test ranges would hold true when the missiles were fired over different parts of the earth with different gravitational anomalies, that all systems will work as expected in the wartime environ­ ment. Since we lack experience with nuclear war, the models obviously are necessary, but it is not clear how seriously the results should be taken. At the very least, decision-makers should know the assumptions they are ac­ cepting when they rely on them. They should also note the political questions which are begged. No decision-maker has ever taken an action which ac­ cepted uncertainties as portentous as those which would be involved in a first strike. Would the side that was behind in the counterforce exchange continue to spare the other's cities? Even if both sides wanted to fight a limited counterforce war—and this would not be consistent with the Soviet approach to war—would the leaders be able to retain the necessary control over their emotions and their forces? If the alarmist models are far removed from reality, the United States may deter itself by paying so much attention to these calculations. It may act more hesitantly, become less confident, refuse new commitments or retract old ones, and even—although I doubt that this would occur—encourage the Soviets to believe that it is safe to undertake actions they previously shunned. A narrowed and distorted focus on implausible contingencies has led to an 22. I have done so in "Why Nuclear Superiority Doesn't Matter," Political Science Quarterly, Vol. 95, No. 4 (Winter 1979-80), pp. 617-633.

Deterrence and Perception | 73

exaggeration of Soviet strength which could restrict U.S. freedom of action to a greater degree than Soviet deterrence policy does. All other things being equal,23 those who believe that the Soviets are militarily superior to the U.S. and that this margin will give them an advantage in conflicts with the West will be more likely than those who reject these views to avoid confrontations with the Soviets. Taken to its extreme, the result would be a form of selffulfilling prophecy in which the U.S. would act as though it were weak, thus permitting the USSR to make gains which would confirm the belief that Soviet "superiority" was a potent weapon. It is sometimes argued that while nuclear "superiority" is militarily meaningless, the Soviets believe that it matters and so will be more likely to stand firm if they believe they have this "advantage." If the Soviet leaders did think that the state of the current nuclear balance permitted them safely to embark on adventures, they would be more likely to provoke confrontations. In fact there is little evidence that this is the Soviet view,24 but if American leaders think it is, they will give the Soviets an unnecessary bargaining advantage. Those who argue that the USSR has strategic superiority which can be used for political gains see themselves as Churchills, but they may be helping to produce the timidity that they decry. Their aim is to spur increases in U.S. arms sufficient to produce favorable results in the war-fighting calculations and therefore, they believe, to favorably influence U.S. and Soviet behavior. But if they succeed only partly and convince people that the calculations have real referents but do not convince them to build more missiles, they will have magnified, if not created, the danger that so worries them. Limits to Rationality

Most arguments about deterrence, including those made above, assume that both sides are fairly rational. Some of the general problems raised by this 23. Often they are not: those who think that the Soviets are militarily strong also tend to believe that they are very aggressive and so see retreating as extremely costly. 24. As George Quester notes, it is "remarkable . . . that the overwhelming bulk of the discussions . . . of growing Soviet relative power and of possible threats come from the west." "Defining Strategic Issues: How to Avoid Isometric Exercises," in Robert Harkavy and Edward Kalodziej, eds., American Security Policy and Pohcy-Makmg (Lexington, Mass.: D.C. Heath, 1980), p. 204. For an evaluation of the Soviet view, see Karl Spielman, The Political Utility of Strategic Superiority: A Preliminary Investigation Into the Soviet View (Arlington, Va.: Institute for Defense Analysis, May 1979). For the general Soviet view of the current situation, see Seweryn Bialer, "The Harsh Decade: Soviet Policies in the 1980s," Foreign Affairs, Vol. 59, No. 5 (Summer 1981), pp. 9991020.

Strategy and Nuclear Deterrence | 74

claim have been treated elsewhere.25 Here I want to focus on four barriers to accurate perception which reduce actors' sensitivity to new information and limit their ability to respond to unexpected situations. The first three barriers are cognitive; the fourth springs from emotions. OVERCONFIDEN CE

First, there is solid evidence from laboratory experiments and much weaker, but still suggestive, evidence from case studies that people overestimate their cognitive abilities. For example, people's estimates of facts usually are less accurate than they think. When asked to give a spread of figures such that they are 90 percent certain that the correct answer lies somewhere between them, most people bracket the true figure only 75 percent of the time.26 Similarly, people generally overestimate the complexity of the way they use evidence. They think they are tapping more sources of information than they are, overestimate the degree to which they combine evidence in complex ways, and flatter themselves by thinking that they search for subtle and elusive clues to others' behavior. Acting on this misleading self-portrait, people are quick to overreach themselves by trying mental operations they cannot successfully perform. Thus, when people are given a little clinical training in judging others' psychological states, they make more errors than they did previously because they incorrectly think they can now detect all sorts of peculiar conditions.27 Overconfidence is also exhibited in the common rejection of the well-established finding that simple computer programs are superior to experts in tasks like graduate student admission and medical diagnosis which involve the combination of kinds of information amenable to fairly objective scoring.28 People believe that, unlike a simple computer program, they can accurately detect intricate, interactive configurations of explanatory or predictive value. In fact, their abilities to do so are very limited.

25. See, for example, Philip Green, Deadly Logic (Columbus, Ohio: Ohio State University Press, 1966); Patrick Morgan, Deterrence (Beverly Hills: Sage, 1977); Robert Jervis, "Deterrence Theory Revisited," World Politics, Vol. 31, No. 2 (January 1979), pp. 299-301, 310-312. 26. Baruch Fischhoff, Paul Slovic, and Sara Lichtenstein, "Knowing with Certainty: The Appro­ priateness of Extreme Confidence," journal of Experimental Psychology: Human Perception and Performance, Vol. 3 (1977), pp. 552-564. 27. Stuart Oskamp, "Overconfidence in Case-Study Judgments," journal of Consulting Psychology, Vol. 29 (1965), pp. 261-265. 28. For a review of this literature, see Lewis Goldberg, "Simple Models or Simple Processes? Some Research on Qinical Judgments," American Psychologist, Vol. 23 (July 1968), pp. 483-496.

Deterrence and Perception | 75

Although a full explanation of this phenomenon is beyond the scope of this paper, overconfidence is probably fed by three factors. First, many of our cognitive processes are inaccessible to us. People do not know what information they use or how they use it. They think some information is crucial when it is not and report that they are not influenced at all by some data on which in fact they rely.29 This makes it easier for them to overestimate the sophistication of their thought processes. Second, a specific aspect of this lack of awareness is that people often rely more than they realize on analogies with past events, especially recent events that they or their country have experienced first-hand. Since these events seem clear in retrospect, much of this certainty is transferred to the current situation. A third cause of overconfidence, also linked to lack of self-awareness, is that people not only assimilate incoming information to their pre-existing beliefs, a point to which we will return, but do not know they are doing so. Instead, they incorrectly attribute their interpretations of events to the events themselves; they do not realize that their beliefs and expectations play a dominant role. They therefore become too confident because they see many events as providing indepen­ dent confirmation of their beliefs when, in fact, the events would be seen differently by someone who started with different ideas. Thus people see evidence as less ambiguous than it is, think that their views are steadily being confirmed, and so feel justified in holding to them ever more firmly. Some of the consequences of overconfidence for deterrence strategies are best seen in light of the two other cognitively rooted perceptual handicaps and so the discussion of them should be postponed. But some effects can be noted here. First, statesmen are likely to treat opposing views quite cavalierly since they are often quite sure that their own beliefs are correct. Cognitive dissonance theory asserts that this intolerance arises only after the person has made a firm decision and has become committed to a policy, but our argument is that it occurs earlier, when even a tentative conclusion has been reached. Second, decision-makers tend to overestimate their ability to detect subtle clues to the other's intentions. They think it is fairly easy to determine whether the other is hostile and what sorts of threats will be effective. They are not sufficiently sensitive either to the possibility that their conclusions are based on a cruder reading of the evidence or to the likelihood that highly complex explanations are beyond their diagnostic abilities. Third, because 29. Richard Nisbett and Timothy Wilson, "Telling More Than We Can Know: Verbal Reports on Mental Processes," Psychological Review, Vol. 84 (1977), pp. 231-257.

Strategy and Nuclear Deterrence | 76

decision-makers fail to realize the degree to which factors other than the specific events they are facing influence their interpretations, their consid­ eration of the evidence will be less rational than they think it is and less rational than some deterrence strategies require. For example, while people realize that it makes no sense to believe that another country is likely to be an aggressor just because a state they recently faced was one, in fact the previous experience will greatly increase the chance that the state currently under consideration will be seen as very dangerous. Similarly, beliefs about the kinds of deterrence strategies which will be effective are also excessively affected by recent successes and failures. Extraneous considerations then influence both conclusions as to whether deterrence is necessary and deci­ sions as to how they will be sought. Decision-makers, furthermore, do not recognize this fact (if they did, presumably they would act to reduce its impact) and so overestimate the extent to which their policies are grounded in valid analysis. NOT SEEING VALUE TRADE-OFFS

The second important cognitive process that influences deterrence is the propensity for people to avoid seeing value trade-offs.30 That is, people often believe that the policy they favor is better than the alternatives on several logically independent value dimensions. For example, those who favored a ban on nuclear testing believed that the health hazards from testing were high, that continued testing would yield few military benefits, and that a treaty would open the door to further arms control agreements. Opponents disagreed on all three counts. This kind of cognitive consistency is irrational because there is no reason to expect the world to be arranged so neatly and helpfully that a policy will be superior on all value dimensions. I am not arguing that people never realize that a policy which gains some important values does so at the price of others, but only that these trade-offs are not perceived as frequently and as severely as they actually occur. This cognitive impediment has several implications for deterrence. First, it complicates the task of balancing the dangers entailed by issuing threats with the costs of making concessions. Rather than looking carefully at this trade­ off, statesmen are likely to be swayed by one set of risks and then evaluate the other costs in a way that reinforces their initial inclinations. For example, 30. For a further discussion of this, see Robert Jervis, Perception and Misperception m International Politics (Princeton: Princeton University Press, 1976), pp. 128-142.

Deterrence and Perception \ 77

a decision-maker who is preoccupied with what he and his state will sacrifice if he compromises on an issue is likely to convince himself that the danger of war if he stands firm is slight; the statesman who concludes that this danger is intolerably high is likely to come to see the costs of retreating as low. As long as the risk on which he focuses is in fact the greater one, and as long as the situation remains unchanging, this minimization of the trade­ off is not likely to lead the decision-maker to choose a policy that differs from the one he would have adopted had he been more rational. But if either of these two conditions is not met, then the quality of the policy will suffer. Thus, if the decision-maker focuses first on the risks of war and finds that it looms large, he may incorrectly judge the costs of retreating as less. He could then abandon a policy of deterrence when rationality would dictate main­ taining it. In other cases, a decision-maker who has decided to stand firm may minimize the value trade-off by failing to take full account of the costs of his position. For example, he may come to believe that, while conciliatory mea­ sures would lower the short-run risk, they would increase the danger over a longer period by leading the adversary to think that it was safe to trifle with the state's interests. In this arrangement of perceptions and evaluations, standing firm appears preferable to being conciliatory on both the dimension of prevailing on the issue in dispute and the dimension of avoiding war.31 The failure to face trade-offs also helps explain the tendency for states to become overextended, to refuse to keep ends and means in balance, and to create more enemies than they can afford. For example, in the years preced­ ing World War I, Germany added Russia and Britain to its list of enemies. On top of the conflict with France, this burden was too great even for a state as strong as Germany. Although both international and domestic factors were also at work, the psychological difficulty of making trade-offs must not be overlooked. When the German leaders decided to drop the Reinsurance Treaty with Russia in 1890, they perceived minimum costs because they expected that ideological conflict would prevent Russia from joining forces with Germany's prime enemy, France. Similarly, the decision to build a large navy and pursue a belligerent policy toward England was based on the

31. Jack Snyder, "Rationality at the Brink," World Politics, Vol. 30, No. 3 (April 1978), pp. 345365. But for the phenomenon to fit the analysis here, the value dimensions must be logically independent. This will not be true if both the perceptions of the need to stand firm and evaluations of the costs of not doing so are produced by a coherent image of the adversary.

Strategy and Nuclear Deterrence | 78

assumption that England's conflicts with France and Russia were so deep that eventually British leaders would have to seek an understanding with the Triple Alliance. German statesmen did not see that their policy involved a greater risk of turning Russia and Britain into active enemies than was entailed by the rejected alternative policy of conciliation and compromise. This failing was not peculiar to Germany. French policy between 1882 and 1898 sought both to rebuild a position of strength against Germany and to contest English dominance of Egypt. To pursue either objective meant risking war with one of these countries. This might have been within the bounds of French resources; war with both was not. So an effective policy required France to set its priorities and decide whether it cared more about its position in Europe or about colonial issues. For over ten years, however, French leaders refused to choose, instead thinking that the same policy could max­ imize the chances of gaining both goals. It took the shock of England's willingness to go to war in the Fashoda crisis for French statesmen to realize that they could not afford too many enemies and had to make a hard choice. President Jimmy Carter's foreign policy provides a final example. To most of the goals of the preceding Ford Administration, the President added an increased concern with preventing proliferation and protecting human rights. He and his advisers did not seem to appreciate that pushing states on one front might diminish their ability to push them on others. Only when crises arose to clarify the mind did they decide to relax the more recent pressures in order to increase the chance of enlisting support for what were taken to be the more important national security goals. But by this time, a large price had been paid in terms of antagonizing others and appearing hypocritical; the overly ambitious initial policy jeopardized America's ability to achieve more limited goals. ASSIMILATION OF NEW INFORMATION TO PREEXISTING BELIEFS

The third cognitive process I want to discuss is probably the most pervasive and significant. It is the tendency for people to assimilate new information to their preexisting beliefs, to see what they expect to be present. Ambiguous or even discrepant information is ignored, misperceived, or reinterpreted so that it does minimum damage to what the person already believes. As I have discussed at length elsewhere,32 this tendency is not always irrational and 32. Jervis, Perception and Misperception, pp. 143-172.

Deterrence and Perception \ 79

does not always decrease the accuracy of perception. Our environment pre­ sents us with so many conflicting and ambiguous stimuli that we could not maintain a coherent view if we did not use our concepts and beliefs to impose some order on it. Up to a point—which cannot be specified with precision— rejecting or providing a strained interpretation of discrepant evidence is the best way to account for all the available information. It is the way scientists behave in treating their data because science would be impossible if they altered their theories to take account of each bit of discrepant information. As Michael Polanyi puts it: The process of explaining away deviations is in fact quite indispensable to the daily routine of research. In my laboratory I find the laws of nature formally contradicted at every hour, but I explain this away by the assump­ tion of experimental error. I know that this may cause me one day to explain away a fundamentally new phenomenon and to miss a great discovery. Such things have often happened in the history of science. Yet I shall continue to explain away my odd results, for if every anomaly observed in my laboratory were taken at its face value, research would instantly degenerate into a wildgoose chase after imaginary fundamental novelties.33 Similarly, statesmen who miss, misperceive, or disregard evidence are not necessarily protecting their egos, being blind to reality, or acting in a way which will lead to an ineffective policy. The evidence is almost always am­ biguous and no view can do justice to all the facts. In retrospect, one can always find numerous instances in which decision-makers who were wrong overlooked or misunderstood evidence that now stands out as clear and important. But one can also note, first, that many facts supported the con­ clusion that turned out to be wrong and, second, those who were right treated the evidence in the same general way—i.e., they also ignored or misinterpreted information which conflicted with their views. Even if the assimilation of incoming information to preexisting beliefs is not as pernicious as is often believed, it creates a variety of problems for deterrence strategies, especially since this cognitive process operates in con­ junction with the other two just discussed. First, images of other states are difficult to alter. Perceptions are not responsive to new information about the other side; small changes are not likely to be detected. Once a statesman 33. Michael Polanyi, "The Unaccountable Element in Science," in Knowing and Being, Essays by Michael Polanyi, ed. Marjorie Grene (London: Routledge and Kegan Paul, 1969), p. 114.

Strategy and Nuclear Deterrence | 80

thinks he knows whether the other needs to be deterred and what kind of strategy is appropriate, only the most dramatic events will shake him.34 The problem is compounded by the common belief to the contrary that, if the initial hunches about what the other side is up to are incorrect, the other's behavior will soon set the statesman straight. For example, those who see the other side as an aggressor usually argue that if this image is incorrect the other can easily demonstrate that its bad reputation is not warranted. In fact, the ambiguity of most evidence coupled with the absorptive power of most beliefs means that an inaccurate image may not be corrected at a point when the situation can still be controlled. A second and more general consequence of the cognitive limitations we have discussed is that political perceptions are rarely completely accurate and policies rarely work as designed. Statesmen cannot then afford to de­ velop policies which are so fragile that they will fail very badly if others do not act exactly as expected. A large margin of error must be built in. The statesman who is sure that his beliefs and calculations are correct in all their details is likely to encounter serious trouble, just as defense strategies that are based on the need to receive tactical warning of when and where the other side is planning to move are likely to fail. For example, it was not reasonable to have expected the military commanders to have anticipated an attack on Pearl Harbor or to have kept the base on constant alert. The latter procedure would have greatly disrupted the urgent training program. Instead the decision-makers should have sought a way to gain some measure of insurance against an attack with the lowest possible interference with train­ ing. The same principle applies to the construction of deterrence strategies. If they are based on an unrealistic assessment of our abilities to perceive our environment and choose among alternatives, they are likely to attempt too much and to fail badly. Third, cognitive impediments place sharp limits on the degree to which deterrence strategies can be fine-tuned, limits that are more severe than statesmen generally realize. For example, states commonly try to develop policies that exert just the right amount of pressure on the other—that is, enough to show the other that the state is very serious, but not enough to provoke desperate behavior. Or, they try to indicate a willingness to ease tensions with an adversary without cooperating to such an extent that third 34. Glenn Snyder and Paul Diesmg, Conflict Among Nations (Princeton: Princeton University Press, 1977), pp. 389^04.

Deterrence and Perception \ 81

parties would feel menaced. At the tactical level, intricate bargaining maneu­ vers are planned and subtle messages are dispatched. For example, in the discussions within the U.S. government in early 1965 about what sort of troops to send to Vietnam, Assistant Secretary of Defense John McNaughton dissented from the view that the initial deployment should be Marines. The problem, he argued, was that the Marines would bring with them "high profile" materiel such as tanks which would indicate to the North that the U.S. was in Vietnam to stay. It would be better to send the 173rd Airborne Brigade which lacked heavy equipment; this would signal Hanoi that the U.S. would withdraw if a political settlement could be reached.35 But even if the actions are carried out as the decision-maker wants them to,36 precision is often defeated by the screen of the other side's perceptual predispositions. As a result, while subtlety and sophistication in a policy are qualities which observers usually praise and statesmen seek, these attributes may lead the policy to fail because they increase the chance that it will not be perceived as it is intended. It is hard enough to communicate straightfor­ ward and gross threats; it will often be impossible to successfully apply complex bargaining tactics which involve detailed and abstruse messages. Decision-makers often underestimate these difficulties and so try to develop plans that are too intricate to get across. Furthermore, because it is very hard to tell what others have perceived, statesmen often fail to see that they have failed to communicate. Finally, since discrepant information is likely to be misinterpreted, deter­ rence strategies must be tailored to the other's preexisting beliefs and images, thus limiting the range of strategies that can succeed. Because the inferences which the other draws are largely determined by its initial beliefs, acts which will deter one decision-maker will be ignored or interpreted differently by another. If the perceiver thinks that the state is deeply concerned about the issue and has high resolve, deterrence will be relatively easy. If he has the opposite view, it will take great efforts to make a credible threat. But unless the state's leaders know what the other side thinks, they will neither know what they have to do to deter it nor be able to judge the chances of success. A frequent cause of deterrence failure is the state's misdesign of its actions

35. Pentagon Papers, p. 421. 36. Most studies of policy implementation reveal that this rarely happens. For a nice analysis that combines bureaucratic and perceptual factors that complicate attempts at coercion, see Wallace Theis, When Governments Collide (Berkeley. University of California Press, 1980).

Strategy and Nuclear Deterrence | 82

growing out of incorrect beliefs about its adversary's perspective. For ex­ ample, American leaders were taken by surprise in October 1962 because they thought it was clear to the Soviet Union that placing missiles in Cuba would not be tolerated. Since the Americans believed—correctly—that the Soviets were not likely to run high risks, they found it hard to imagine that the USSR would try to establish a missile base abroad. U.S. leaders did not think that great efforts at deterrence were necessary because they did not realize that the move would not look risky to the Soviets.37 Just as the best way to understand the conclusions a person draws from a "fact-finding" mission is to know his initial beliefs rather than to know what evidence he was exposed to, so one can often make better predictions about how a state will interpret others' behavior by knowing the former's predis­ positions than by knowing what the latter actually did. Unfortunately, states­ men rarely appreciate this and, to compound the problem, usually have a much better idea of what they think they are doing and what messages they want to convey than they do of what the others' perceptual predispositions are. The difficulty is two-fold and two-sided. The fact that perceptions are strongly influenced by predispositions means that it is very difficult to convey messages that are inconsistent with what the other already believes. And the fact that statesmen do not understand this influence reduces their ability to predict how others will react. Even if decision-makers understood the prob­ lem, prediction would be difficult because it is so hard for them to grasp the way in which others see the world. But in this case they would at least realize that many of their messages would not be received as they were sent. Since this understanding is often lacking, decision-makers' messages not only con­ vey different meanings to each side, but each is usually unaware of the discrepancy. Statesmen are then likely to err both in their estimates of what the other side intends by its behavior and in their beliefs about how the other is reading their behavior. Severe limits are thus placed on the statesman's ability to determine whether and what kind of deterrence strategy is called for and to influence the other's perceptions in a way which will allow this strategy to succeed. A failure to understand these limitations imposed by the way people think will make it more difficult for scholars to explain state

37. Klaus Knon, "Failures m National Intelligence Estimates: The Case of the Cuban Missiles," World Politics, Vol. 16, No. 3 (April 1964), pp. 455—467. For an alternative argument, see Richard Ned Lebow, "Soviet Risk Taking: What Are the Lessons from Cuba?" Political Science Quarterly,

forthcoming.

Deterrence and Perception | 83

behavior and will lead a statesman to attempt overly ambitious policies that are likely to bring his country to grief. Defensive Avoidance

A final impediment to accurate perception that can complicate or defeat deterrence is affective rather than purely cognitive. In a process known as defensive avoidance, people may refuse to perceive and understand ex­ tremely threatening stimuli.38 For example, the failure to see the value trade­ offs discussed above can be motivated by the need to avoid painful choices. At this point we do not know enough about the phenomenon to determine when these errors occur and how influential they are in comparison with other factors. But it seems clear that on at least some occasions powerful needs, often arising from domestic politics, can produce badly distorted perceptions of other countries. Thus Paul Schroeder has argued that the British images of Russia in the period leading up to the Crimean War cannot be explained either by Russian behavior or by long-standing and deeply imbedded cognitive predisposition but rather were caused by shifting British needs to see Russia as threatening or accommodating.39 Whether England tried to deter Russia or conciliate it then depended on internal factors that were neither rationally related to foreign policy goals nor susceptible to Russian influence. Similarly, states may come to think that it is relatively safe to challenge the adversary's deterrent commitments when a modicum of rational analysis would indicate that the risks far outweigh the slight chances of success if domestic or foreign needs for a challenge are very strong. This is not only to argue that the costs of forgoing gains and accepting the adversary's deterrence can be so high as to rationally justify a challenge that the statesman knows is likely to fail; this may be unfortunate but it is not troublesome in terms of perceptions. Rather the knowledge of the high costs of accepting the status quo can lead statesmen to ignore or distort information about the costs of challenging it. Thus Lebow shows that the reason why India in 1962, the United States in the fall of 1950, and the Soviet Union

38. The fullest discussion is in Irving Janis and Leon Mann, Decision Making (New York: Free Press, 1977). 39. Paul Schroeder, Austria, Great Britain, and the Crimean War (Ithaca, N.Y.: Cornell University Press, 1972).

Strategy and Nuclear Deterrence | 84

before the Cuban Missile Crisis were not able to see that their adversaries would inflict painful rebukes if they persisted was that they were preoccupied with the costs they would pay if they did not.40 To return to a case mentioned earlier, the American attempts to deter Japan failed because Japan thought that the war would be limited. This error may have been at least in part a motivated one. The feeling that acquiescing in the American demands was intolerable led the Japanese to adopt an unrealistically favorable view of the alternative—the only way they could avoid facing the need to sacrifice very deeply held values was to believe that the U.S. would fight a limited war. That their conclusion was driven by this need rather than by objective anal­ ysis is indicated by the quality of their deliberations: "Instead of examining carefully the likelihood that the war would in fact be a short, decisive one, fought under optimum conditions for Japan, contingency plans increasingly took on a strangely irrational, desperate quality, in which the central issue, 'Can we win?' was shunted aside. Rather, it was as if Japan had painted itself into a corner."41 The result is that deterrence can be difficult if not impossible. Threats that should be credible and effective, even when the cognitive impediments discussed above are not operating, may be missed or misread. It usually will be hard for the deterrer to realize that it is facing this danger and even an understanding of the situation will not easily yield an effective policy since the other's perceptual screens are often opaque. 40. Lebow, Between Peace and War and "Soviet Risk Taking." Also see Richard Cottman, Foreign Policy Motivation (Pittsburgh: University of Pittsburgh Press, 1977); Alexander George and Rich­ ard Smoke, Deterrence in American Foreign Policy (New York: Columbia University Press, 1974); and Jack Snyder, Defending the Offensive: Biases in French, German, and Russian War Planning, 1870-1914 (Ph.D. dissertation, Columbia University, 1981), which does a particularly fine job of separating motivated from unmotivated errors. Sharp-eyed readers will note a shift from some of my earlier views on this point. For further discussion, see Robert Jervis, "Political Decision Making: Recent Contributions," Political Psychology, Vol. 2 (Summer 1980), pp. 89-96. 41. Robert Scalapino, "Introduction," in James Morley, ed., The Fateful Choice: Japan's Advance Into Southeast Asia, 1939-1941 (New York: Columbia University Press, 1980), p. 119. Also see Gordon Prange, At Dawn We Slept (New York: McGraw-Hill, 1981), pp. 16, 21.

Inadvertent Nuclear War?

ν κ·Posen

Βαη

Escalation and NATO's Northern Flank Could a major EastWest conventional war be kept conventional? American policymakers in­ creasingly seem to think so. Recent discussions of such a clash reflect the belief that protracted conventional conflict is possible, if only the West fields sufficient conventional forces and acquires an adequate industrial mobiliza­ tion base. Indeed, the Reagan Administration has embraced the idea of preparing for a long conventional war, as evidenced by its concern with the mobilization potential of the American defense industry.1 Underlying this policy is the belief that the United States should be prepared to fight a war that, in duration and character, resembles World War II. American decision­ makers seem confident of their ability to avoid nuclear escalation in such a war if they so desire. That confidence is dangerous and unwarranted. It fails to take into account that intense conventional operations may cause nuclear escalation by threat­ ening or destroying strategic nuclear forces. The operational requirements (or preferences) for conducting a conventional war may thus unleash enor­ mous, and possibly uncontrollable, escalatory pressures despite the desires of American or Soviet policymakers. Moreover, the potential sources of such escalation are deeply rooted in the nature of the force structures and military strategies of the superpowers, as well as in the technological and geograph­ ical circumstances of large-scale East-West conflict. If the escalatory pressures While the author bears sole responsibility for all views expressed here, he is grateful to Bruce G. Blair, Joshua Epstein, William W. Kaufmann, John Mearsheimer, Stephen Van Evera, and Kenneth Waltz for their suggestions.

Barry Posen is a Council on Foreign Relations Fellow. This article was written while he was a Postdoctoral Fellow at the Center for International Affairs, Harvard University.

1. See, for example, the accounts of Secretary Weinberger's views in George Wilson, "Wein­ berger Order: Plan for Wider War," The Boston Globe, July 17, 1981; and Richard Halloran, "Weinberger Tells of New Conventional-Force Strategy," The New York Times, May 7, 1981. For further indications of the Administration's views on this subject, see also Richard Halloran, "Needed: A Leader for the Joint Chiefs," The New York Times, February 1, 1982; Richard Halloran, "Reagan Selling Navy Budget as Heart of Military Mission," The New York Times, April 11, 1982; and Caspar W. Weinberger, Secretary of Defense, Annual Report to Congress, FY 1983, pp. 1-13, 16-17, 28-29.

International Security, Fall 1982 (Vol. 7, No. 2) 0162-2889/82/010028-27 $02 50/0 © 1982 by the President and Fellows of Harvard College and of the Massachusetts Inshtute of Technology.

Strategy and Nuclear Deterrence \ 86

that could attend a major conventional war are to be prevented from driving decision-makers toward decisions they neither intend nor wish to make, those pressures must be recognized and guarded against by the leaders of both superpowers.2 Moreover, underestimating the escalatory risks that would accompany conventional war has several significant negative consequences, even in peacetime. First, American decision-makers pay insufficient attention to the details of conventional posture and operations essential to the limitation of war. Too many agree with the observation that "both sides understand conventional warfare, they know that it can be controlled in the present age."3 Second, leaders who fail to appreciate fully the dangers of nuclear escalation may not be cautious enough about both the initiation and the conduct of direct confrontations between Soviet and American military power. Third, nuclear deterrence may be undermined by excessive public confidence in the limitability of superpower conventional war. The "threat to lose control" is an important element of NATO's flexible response strategy, and must be preserved. It would be unfortunate if the public pronounce­ ments of Western strategists encouraged the Soviets to believe that they could easily avoid nuclear punishment for "conventional" aggression. Fourth, emphasis on protracted conventional conflict weakens Western Eu­ rope's confidence in America's nuclear guarantee. Emphasizing instead the difficulty of keeping conventional war conventional might ameliorate Alliance fears that the U.S. nuclear umbrella no longer shields them. Unfortunately, surprisingly little attention has been devoted to analyzing and understanding this path to nuclear war. Escalation has generally been conceived of as either a rational policy choice, in which a leadership decides to preempt or to escalate in the face of a conventional defeat, or as an accident, the result of mechanical failure, unauthorized use, or insanity. But escalation arising out of the normal conduct of intense conventional conflict falls between these two categories: it is neither a purposeful act of policy nor an accident. What might be called "inadvertent escalation" is rather the unintended consequence of a decision to fight a conventional war. Defense 2. Because the American defense posture and war plans are more easily studied and influenced than those of the Soviet Union, this essay will focus mainly on American choices of plans and deployments that affect escalation to the use of nuclear weapons. It should be emphasized that unless Soviet leaders are also sensitive to this problem, their decisions about the conduct of conventional war could as easily cause escalation as those of their American counterparts. 3. Morton Halperin, Limited War in the Nuclear Age (New York: John Wiley, 1963), p. 64.

Inadvertent Nuclear War? | 87

analysts have long believed that nuclear war is most likely to emerge out of a conventional conflict, but the idea of inadvertent escalation has somehow escaped their attention. It nevertheless is as plausible a route to nuclear conflict as many of the scenarios that have received extensive scrutiny, and probably more likely than many. Because nuclear escalation could destroy everything that we value, the West must be careful that the way it intends to defend itself conventionally does not bring about the very destruction it hopes to avoid. Since inadvertent escalation is a consequence of conventional military prac­ tices, its likelihood is affected by choices regarding force postures and op­ erational plans. There is little evidence, however, that the problem of esca­ lation control has influenced such choices. As a result, the huge investments that the United States has made in conventional forces for the purpose of buffering against precipitate nuclear escalation may be negated by imprudent operational planning and reckless execution of those plans. Under the prevailing conceptions of escalation, most NATO strategists have focused their attention on improving the conventional balance in Cen­ tral Europe as the most important barrier to the early use of nuclear weapons (and, indeed, the collapse of NATO's conventional defenses could lead to nuclear escalation under the strategy of flexible response). Concern about inadvertent escalation, however, causes NATO's Northern Flank to loom much larger in importance, for it illustrates vividly the escalatory pressures inherent in many conventional operations. It is from a thorough examination of this area that I shall principally draw in this essay. The danger of inadvertent escalation arises with particular acuteness in the Northern Flank, especially in the region comprised of the Norwegian and Barents Seas, Northern Norway, and the Kola Peninsula in the Soviet Union. In Northern Norway, NATO territory and NATO conventional capabilities are situated near a critical element of Soviet strategic nuclear power—the ballistic missile submarine force based at Murmansk in Kola. Conventional military operations in the region could put this force at risk. The U.S. Navy frequently argues in favor of attacks with conventionally armed, carrier-based aircraft against Soviet naval bases and naval forces in this area, a strategy endorsed by the Reagan Administration.4 If such attacks should threaten the

4. Congressional Budget Office, Building a 600-Ship Navy: Costs, Timing, and Alternative Approaches (Congress of the United States, March 1982), pp. xi-xii, xxiii, 17; Halloran, "Reagan Selling Navy Budget"; John Lehman, Secretary of the Navy, "America's Growing Need for Seaborne

Strategy and Nuclear Deterrence | 88

survivability of Soviet ballistic missile submarines, as they might, how would the Soviet Union react? It cy to General Nathan F. Twining, 6 February 1951, Twining Folder, Box B-60, LeMay Papers, LC. 52. "Briefing on Exercise 'Dualism' Conducted by Colonel John A. Armstrong," 9 December 1948, A/AE (1948) 354.2 Exercise Dualism, CSAF; Rosenberg, "U.S. Nuclear Stockpile, 1945 to 1950," pp. 29-30. On SAC problems, see, m particular, Letter, Charles A. Lindbergh to Secretary of the Air Force Thomas K. Finletter, 8 June 1950, in Office of the Administrative Assistant,

Strategy and Nuclear Deterrence \ 130

October 1951, SAC was composed of 28 bomber and 7 fighter wings, with 15 consid­ ered combat ready.53 The emergence of SAC was rooted in Air Force experience. The air planners were all veterans of the bombing campaigns against Germany and Japan. They were convinced of the efficacy of strategic bombing, and believed that the Air Force must serve as the nation's new first line of defense. Despite doubts about individual weapons systems such as the B-36 bomber, Air Force leaders were determined to project an image of strength and confidence.54 They were also motivated by the fear, largely unfounded, that the Navy and its aviators were attempting to take over the strategic bombing mission. One Air Staff officer warned in 1948: "if the Greyhound Bus Company can demonstrate a capability of delivering [atomic] bombs better than any other agency, that company will get the job. . . ."55 LeMay instilled a spirit in his command which made SAC for a time the nation's most elite military unit. SAC air crews trained under a competitive reward system that fostered high morale and military proficiency by awarding spot promotions to those personnel who achieved Select and Lead crew status. LeMay sponsored the development of innovative technology to increase SAC readiness as well. He was particularly proud of the specially etched lucite plate radar screen overlays which simulated radar images of Russian target areas. These projections were used to train air crews to recognize fixed geographic features as guides in locating assigned aim points. Special ultrasonic maps and radar intensity charts were also developed, special radar reconnaissance procedures were adopted, improvements were made in bomb­ ing and navigational equipment, and mock Soviet warning radars were built.56 By

Correspondence Control Division, Special Interest File, 1950, Record Group 340, Papers of the Secretary of the Air Force, Modern Military Branch, National Archives; Letter, LeMay to Vandenberg, 21 September 1950, Vandenberg Folder, Box B-61, discussing B-29 and B-50 engine problems, and paper "Major Deficiencies in the B-47 Aircraft," 26 October 1950, Command Papers, Box B-98, both in LeMay Papers, LC. 53. "Strategic Air Command Progress Analysis, 1 November 1948 to 31 October 1953," Box B100, LeMay Papers, LC, pp. 11-12. 54. For contrasting views of the B-36 and U.S. strategic capabilities, see the rather pessimistic remarks in "Aircraft and Weapons Board Proceedings, 29 December 1948-6 January 1949," Vol. 1, pp. 54-141, in Executive Office, House Investigation, August-October 1949, CSAF; and the confident predictions of General Vandenberg in JCS 1952/1, 21 December 1948, CCS 373 (10-2348) Sec. 1, JCS. For a useful overview of Air Force thinking, see John T. Greenwood, "The Emergence of the Postwar Strategic Air Force, 1945-1953," in Alfred F. Hurley and Robert C. Ehrhardt, eds., Air Power and Warfare, The Proceedings of the 8th Military History Symposium, U.S. Air Force Academy, October 18-20, 1978 (Washington, D.C.: U.S. Government Printing Office, 1979), pp. 215-244. 55. Little, Air Force Atomic Energy, Vol. 2, p. 152. The officer was Colonel J. Fickel. 56. "Strategic Air Command Progress Analysis, 1948-1953," pp. 37-38, 44-50; "Strategic Air Command Progress Analysis, 1 November 1948-31 December 1956," pp. 28-30, 65-83, Box B100, LeMay papers, LC; "Memorandum, Op-36C to Op-36, Subject: Briefing given to the rep­ resentatives of all services at SAC Headquarters, Offutt Air Force Base, Omaha, on 15 March, 1954," Document One in David Alan Rosenberg, "Ά Smoking Radiating Ruin at the End of Two Hours': Documents of American Plans for Nuclear War with the Soviet Union, 1954-1955," International Security, Vol. 6, No. 3 (Winter 1981/1982), pp. 3-38.

Nuclear Strategy, 1945-1960 | 131

1952, SAC was training in earnest for the blunting mission, using its radar prediction and offset bombing techniques to simulate strikes against airfields whose precise coordinates were unknown.57 Improvements in intelligence gathering complemented SAC's training techniques. Beginning in the late 1940s, Special Electronic Airborne Search Operations were flown by Air Force bombers and transports and Navy patrol planes to ferret out Soviet radar capabilities for the purpose of developing electronic countermeasures and locating Soviet air defenses. By 1950, SAC RB-36 and RB-50 aircraft were flying electronic and photo reconnaissance missions, although few if any of these spy flights actually crossed Soviet borders.58 On rare occasions, Central Intelligence Agency operatives were sent to check out specific targets. A Siberian airfield was reconnoitered in 1952, for example, to see if it could be used by Soviet bombers for nuclear missions.59 The most bizarre intelligence operation of this period was the CIA-sponsored "Moby Dick" program, which sent large "Skyhook" balloons equipped with Air Force ca­ meras drifting across the Soviet Union from Western Europe to Japan. Unfortunately, since it was not possible to track the balloons in flight, it was difficult to identify what they had taken pictures of.60 The bulk of Air Force basic intelligence on the USSR between 1949 and 1953—well over 50 percent in 1951—was provided by Project WRINGER. WRINGER employed 1300 military and civilian personnel in Germany, Austria, and Japan to interrogate thousands of repatriated prisoners of war from the Soviet Union and correlate the reports for use in target and other planning.61 TRUMAN'S EXPANSION OF NUCLEAR CAPABILITY

The buildup of U.S. nuclear capability was greatly facilitated by a shift in presiden­ tial thinking. On July 14, 1949, Truman told his top policy advisers: "I am of the 57. "Headquarters, Strategic Air Command, Report: Springfield Radar Bombing Evaluation," 7 July 1952, Box B-106; "Strategic Air Command Progress Analysis, 1948-1953," pp. 27-29, Box B100; Letter, Curtis LeMay to Major General Frank A. Armstrong, 9 July 1953, Armstrong Folder, Box B-49; and Info for Vice Commander, D/Ops, on Location of Airfields from Radar Photos, Commanders Diary #5 Folder, 1953, Box B-64, all LeMay Papers, LC. 58. Special Electronic Airborne Search Operations are described in Memoranda, General Omar Bradley to Secretary of Defense Louis Johnson, 5 May and 22 July 1950, and Johnson to the President, 24 May 1950, all in Omar N. Bradley Folder, General File, PSF, HSTL. President Truman's approval of flights is noted on the 5 May letter. Strategic Air Command Progress Analysis, 1948-1953, pp. 39-50. Letter, Curtis E. LeMay to Professor Harland Moulton, ca. December 1980, copy courtesy Professor Moulton. Prados, The Soviet Estimate, pp. 24-30. 59. Peer DeSilva, Sub Rosa, The CIA and the Uses of Intelligence (New York: Times Books, 1978), pp. 58-61; Harry Rositzke, The CIA's Secret Operations (New York: Readers' Digest Books, 1977), pp. 18-38. 60. Prados, The Soviet Estimate, pp. 29-30. The "balloon operation" is also mentioned in A.J. Goodpaster, MCP, February 10, 1956, in Joint Chiefs of Staff, January-April 1956, Folder 2, Subject Series, Defense Department Sub Series, Box 4, WHO-SS, DDEL. 61. Alfred Goldberg, ed., History of Headquarters, USAF, 1 July 1950 to 30 June 1951 (Washington, D.C.: Department of the Air Force, 1955), p. 7; James M. Erdmann, "The WRINGER in Postwar Germany: Its Impacit on United States-German Relations and Defense Policies," in Clifford L. Egan and Alexander W. Knott, eds., Essays in Twentieth Century Diplomatic History Dedicated to Professor Daniel M. Smith (Washington, D.C.: University Press of America, 1982), pp. 159-191.

Strategy and Nuclear Deterrence | 232

opinion we'll never obtain international control. Since we can't obtain international control we must be strongest in atomic weapons."62 The budgeting and programming decisions of the next several years reflected, in part, Truman's reluctant acceptance of nuclear weapons as the centerpiece of American defense policy. In the fall of 1949, Truman approved the substantial increase in nuclear production requested by the JCS in response to the Harmon Report. In October 1950, following the outbreak of the Korean War, he approved a second increase.63 In December 1950, NSC-68/4 established objectives for a phased and balanced buildup of American armed forces through June 1952.64 During budget deliberations the following summer and fall, however, Air Force Secretary Thomas K. Finletter and Chief of Staff Vandenberg forcefully argued that only the Air Force could deliver the nation's expanding nuclear arsenal, and that failure to provide adequate delivery capability would nullify the effect of the nuclear production increases. As Vandenberg explained: In the event of war, there will be concurrent requirements for the destruction of Soviet atomic delivery capability, direct atomic attack on Soviet ground and tactical air forces, and destruction of the critical components of the enemy's war sustaining resources. It must be pointed out that if we do not provide an air force tactically strong enough to deliver atomic weapons on target with a high degree of reliability (and we thereby run out of delivery capability while appropriate targets and unex­ pended bombs remain) we will have committed a military blunder which will defy logical explanation to the American people. We will have failed to make provision to exploit our one major military advantage over the USSR.65 The JCS subsequently recommended that the Administration assign primacy to the buildup of the Air Force and SAC. The Fiscal Year 1953 defense budget was based on an Air Force objective for June 1954 of 143 wings, 48 more than had been proposed in NSC-68/4, with no corresponding increases in Army and Navy force objectives. President Truman cut this goal to 133 wings by the end of Fiscal 1954, but the Air Force retained its priority position, receiving over 40 percent of the military budget.66 The growth of the nuclear stockpile was linked to escalating target estimates, just as expansion of SAC was linked to the stockpile. In January 1952, President Truman

62. Department of State, FRUS, 1949 Vol. I, National Security Affairs, Foreign Economic Policy (Washington, D.C.: U.S. Government Printing Office, 1976), pp. 481-482. 63. Rosenberg, "American Atomic Strategy and the Hydrogen Bomb Decision," pp. 75-79; James S. Lay, Jr., Memorandum for the President, October 2, 1950, summarizing report of Special Committee of the National Security Council, with President's Approval, October 9, 1950, noted, Expansion of the Fissionable Material Folder, NSC-Atomic File, PSF, HSTL. 64. Poole, The JCS and National Policy, Vol. 4, pp. 65-77. 65. JCS 1800/164, 6 September 1951, CCS 370 (8-19-45) Sec. 34, JCS. See also Memorandum for the Chief of Staff, by TKF (Thomas K. Finletter), no date (ca. mid-September 1951), discussing relationship between Air Force program and nuclear weapons production; and Finletter, Mem­ orandum for the Secretary of Defense, 16 July 1951, all in Folder 2A, Memos Re 138 Wings, Box 83, Hoyt S. Vandenberg Papers, LC. 66. Poole, The JCS and National Policy, Vol. 4, pp. 79-131.

Nuclear Strategy, 1945-1960 | 133

approved a third increase in fissionable material production, amounting to a 50 percent increase in plutonium production and 150 percent in uranium 235. General Vandenberg, who had originated the request within the JCS, justified it in terms of prospective Soviet targets. Even allowing for incomplete intelligence, he told the President, there appeared to be "perhaps five or six thousand Soviet targets which would have to be destroyed in the event of war."67 This would require a major expansion in weapons production. Critics in the other services labeled this line of argument "bootstrapping." Air Force-generated target lists were used to justify weapons production, which in turn justified increased appropriations to provide matching delivery capability. The new blunting mission greatly facilitated the bootstrapping dynamic. Soviet atomic capa­ bility in 1952 was estimated at only about 50 bombs and 800 TU-4 bombers, but related targets included nuclear production and storage facilities, fuel storage sites, and, most numerous of all, airfields from which a nuclear strike might be launched. As intelli­ gence improved, Air Force target lists steadily outpaced accelerating stockpile growth.68 The three approved increases in nuclear production may well have been the most substantive actions taken by the Truman Administration in the area of strategic nuclear policy during its last three years. By January 1953, a construction program was underway which would add eight plutonium production reactors and ten gaseous diffusion U-235 production plants to the five reactors and two gaseous diffusion plants operating in mid-1950. These plants and reactors were capable of supporting an enormous expansion of the nuclear weapons stockpile. According to recent au­ thoritative unclassified estimates, the stockpile grew from approximately 1000 weap­ ons in the summer of 1953 to nearly 18,000 by the end of the decade.69 No subsequent administration found it necessary to authorize any further expansion of nuclear production facilities to meet weapons requirements. 67. Memorandum for the President, January 17, 1952, summarizing meeting of NSC Special Committee on Atomic Energy with the President, January 16, 1952, and President's approval on Memo, James S. Lay, Jr. to the President, January 16, 1952. The amount of increase is discussed in Memorandum, Robert A. Lovett to Executive Secretary, National Security Council, 11 December 1951, Subject: Secretary of Defense Comment on Expansion Program, all in Ex­ pansion of the Fissionable Material Folder, NSC-Atomic File, PSF, HSTL. 68. Soviet atomic capability is described in JCS 2101/75, 25 October 1952, Appendix C to Enclo­ sure A, CCS 381, US (1-31-50) Sec. 20, JCS, and N.I.E. 64, Part I, Soviet Bloc Capabilities through Mid-1953,12 November 1952, National Intelligence Estimates, 64-66 Folder, Box 254, Intelligence File, PSF, HSTL. The bootstrapping dynamic was described in author's interview with Mr. John P. Coyle, veteran Navy operations analyst, Washington, D.C., 29 March 1978. 69. Nuclear production facilities are listed in Richard G. Hewlett and Francis Duncan, History of the United States Atomic Energy Commission, Vol. Π, Atomic Shield, 1947-1952 (University Park, Pa.: The Pennsylvania State University Press, 1969), p. 669. Stockpile size is estimated in Rosenberg,"Ά Smoking Radiating Ruin,"' pp. 7-8; William M. Arkin, Thomas B. Cochran, and Milton M. Hoenig, "The U.S. Nuclear Stockpile," Arms Control Today, Volume 12 (April 1982), pp. 1-2; and is estimated for 1952-1954 in Memorandum, 2 November 1951, with page with figures labelled "Atomic Conf. Finletter's Office 11/2/51" in Diaiy Notes Folder, Dan A. Kimball Papers, Box 2, HSTL.

Strategy and Nuclear Deterrence \ 134

The weapons in the American stockpile had grown increasingly sophisticated and powerful in the last years of the Truman Administration. The "nominal" 20 kiloton yield of the Mark 3 bomb was multiplied more than 25 times as a result of innovations perfected between 1948 and 1952. These included advances in the design, composi­ tion, stability, and power of the high explosives used to detonate a fission core, and improvements in the mechanics, structure, and composition of the fissile "pit" itself. The pit improvements included development of "composite" U-235 and plutonium cores, and exploitation of the levitation design concept which utilized an air space to allow the detonation shock wave generated by the high explosive assembly to gather momentum before imploding the core, resulting in higher yields and' increased effi­ ciency in the use of fissionable material. These developments led to the November 1952 test of the Mark 18 Super Oralloy (U-235) bomb which yielded 500 kilotons.70 By this time, atomic weapons were being mass produced, and the various fission cores were interchangeable among the weapons assemblies being manufactured. The technology of "boosted" weapons, which employed a small amount of fusion fuel within a "hollow implosion" core to further increase efficiency and yield was first tested in April 1951, and appears to have been capable of producing yields approach­ ing one megaton. In prospect were "true" thermonuclear weapons. The first of these was tested in October 1952. A 21-ton, cryogenically cooled, liquid-fueled monster, it yielded 10.4 megatons, but could not be delivered against Soviet bloc targets. Devel­ opment of "dry" lithium deuteride-fueled fusion bombs weighing between 4,000 and 40,000 pounds was already underway, and "emergency" delivery capability in SAC B-36s was ready by 1953.71 Air Force planners perceived an increasing need for such extremely powerful weapons. In the 1940s, military requirements for high yield weapons had been minimal and easily calculated. These rose rapidly, however, as a result of construction of Soviet industry with improved blast resistance behind the Urals; General LeMay's emphasis on bonus damage; and the 1951 "Buster" tests, which revealed that mea­ surements of earlier nuclear explosions had exaggerated blast effects by as much as

70. Information on nuclear weapons design is drawn from the following sources· Rosenberg, "U.S. Nuclear Stockpile, 1945 to 1950," pp. 27-28; Office of Public Affairs, U.S. Department of Energy, Nevada Operations Office, Announced United States Nuclear Tests, July 1945 through December 1981 NVO-209 (Rev. 2) January, 1982; JCS 1823/8, 6 November 1948, JCS 1823/11, 20 December 1948, and JCS 1823/35, 30 November 1950, CCS 471.6 (8-15-45) Sees. 13 and 20, JCS; Lee Bowen, History of the Air Force Atomic Energy Program, Vol. IV, The Development of Weapons (Washington, D.C.: U.S. AirForce Historical Division History, 1955, declassified with deletions June 1981), pp. 1-143; table appended in letter, Robert H. Duff, Office of Classification, U.S. Department of Energy, to David A. Rosenberg, December 4, 1980; Herbert York, The Advisors: Oppenheimer, Teller, and the Superbomb (San Francisco: W.H. Freeman, 1976); Affidavit, Dimitri A. Rotow, May 3, 1979, in "United States of America vs. the Progressive, Inc." (copy courtesy of Mr. Aaron Meyers, Esq.); and Charles Hansen, "U.S. Nuclear Bombs," Replica in Scale, Vol. 3 (January 1976), pp. 154-159. 71. Bowen, Air Force Atomic Energy, Vol. IV, pp. 174-233; JCS 1823/35; York, The Advisors, pp. 82-87. Yield possibilities in boosted weapons are discussed in David Holloway, "Research Note: Soviet Thermonuclear Development," International Security, Vol. 4, No. 3 (Winter 1979-1980), pp. 192-197; and in JCS 1823/34, 1 November 1950, CCS 471.6 (8-15-45) Sec. 20, JCS.

Nuclear Strategy, 1945-1960 | 135

one-half to two-thirds, indicating that weapons requirements had been underesti­ mated in all previous war plans.72 SUMMARY: NUCLEAR WAR PLANNING IN THE TRUMAN YEARS

By the time Truman left office in January 1953, the framework for nuclear strategy had been established. The United States had been launched into the era of nuclear plenty, a series of targeting categories had been approved which emphasized preemp­ tion of Soviet nuclear capability, and the Strategic Air Command had gained a major voice in how targets were selected and damage criteria established. In addition, American war planning was being conducted on a routine basis with a regularity and intensity unprecedented in peacetime. Through the Korean War, the JCS stressed emergency war planning, although efforts were made to institutionalize mid- and long-range planning as well. In the summer of 1952, the Joint Chiefs authorized creation of a new "family" of war plans: the Joint Strategic Capabilities Plan (JSCP) designed to govern the wartime operations of U.S. forces-in-being during the current fiscal year; the Joint Strategic Objectives Plan (JSOP) which established force and mobilizatioij requirements for the next three to five years; and the Joint Long Range Strategic Estimate, which looked ahead five years or longer as a guide to research and development. These plans were supposed to be completed annually, but the latter two were often delayed by interservice debates and appeared somewhat more irregularly. In the 1950s, the JSOP became increasingly a "wish list" rather than a realistic estimate of requirements. Nevertheless these three efforts provided a per­ manent structure for organizing strategic planning into the 1970s.73 The JSCP and the operational plans it guided, including the SAC Emergency War Plan, were generally prepared on an annual basis. They fostered a process of intensive interservice debate and analysis which, in the absence of real global conflict, served as a kind of "surrogate war" for generating and testing forces and concepts. Each new planning effort built on the tradeoffs and compromises endorsed by the preced­ ing "war," thereby creating a dynamic which tended to discourage radical changes. The NSC made little effort to guide nuclear war planning, beyond the summary guidance provided in NSC-30. NSC-68 of April 1950 explicitly rejected preventive war ("a military attack not provoked by a military attack on us or our allies"), but left the door open for preemption with the parenthetical notation that: The military advantages of landing the first blow . . . require us to be on the alert in order to strike with our full weight as soon as we are attacked, and, if possible, before the Soviet blow is actually delivered.74

72. Rosenberg, "American Atomic Strategy and the Hydrogen Bomb Decision," p. 81; JCS 1823/ 35; Memorandum, David Driggs to Hoyt S. Vandenberg, Subject: The Drastic Implications of BUSTER Air Blast Measurements, 17 December 1951, Folder #8, Buster Data, Box 88, Vanden­ berg Papers, LC; JCS 1953/11, 28 May 1952, CCS 373 (10-23-48) Sec. 7, JCS; Poole, The fCS and National Policy, Vol. IV, p. 167. 73. Ibid.; JCS Policy Memo 84, 14 July 1952. 74. NSC-68, "United States Objectives and Programs for National Security," April 14, 1950, Department of State, FRUS, 1950, Vol. I: National Security Affairs; Foreign Economic Policy (Wash­ ington, D.C.: U.S. Government Printing Office, 1977), pp. 281-282.

Strategy and Nuclear Deterrence | 136

NSC-114/3 of· June 1953 described the highest priority Air Force mission in equally ambiguous terms: "To defend by both offensive and defensive air operations critical areas in the Western Hemisphere, with particular emphasis on defense against atomic attack."75 The NSC did little to clarify how the President would determine where and when atomic bombs might be used. An eighteen month staff study, begun in the spring of 1951 but never formally approved, outlined the procedures by which he could obtain advice if necessary. It recognized that the JCS, as the President's chief military advisers, would play a key role: Unless there is an initial determination by the Joint Chiefs of Staff that the use of atomic weapons in a given situation is militarily desirable, it is difficult to see how the question of such use can arise in any realistic way.76 If the JCS recommended use of nuclear weapons, then the President would seek, at a minimum, the views of the Secretaries of Defense and State and the Chairman of the Atomic Energy Commission. If time permitted, consultation would also be carried out with congressional leaders, and thought given to notifying the American people and other governments. This study once again stressed presidential prerogative, and, although hinting at the problem, did little to close the gap between policy and planning which had been codified by NSC-30. It is not clear whether Truman ever fully grasped the fundamental dilemma posed by Soviet possession of nuclear weapons. How could the United States plan to defend itself against a potential Soviet strike while ruling out the option of striking first? Truman was quite clear in his own mind that preventive war was contrary to American values and could not be considered a policy option. "Starting an atomic war is totally unthinkable for rational men," he told the nation in his farewell address.77 But whether he had resolved the issue of preemption to his own satisfaction is not known. When AEC commissioner Thomas Murray asked whether the President had really meant to rule out the option of preempting Soviet nuclear capability under the threat of imminent Soviet attack, Truman sidestepped the question. "I rather think you have put a wrong construction on my approach to the use of the Atomic bomb," he wrote. "It is far worse than gas and biological warfare because it affects the civilian

75. NSC 114/3, Report to the National Security Council on United States Programs for National Security, June 5, 1952, NSC Meeting 105 Folder, Box 215, Subject File, PSF, HSTL. 76. Study, June 11, 1952, appended to Memorandum, Everett Gleason to the President, October 23, 1952, NSC-Atomic Weapons—Procedures for Use Folder, NSC Atomic File, PSF, HSTL. See also the other documents in this folder. The process began with a draft memo by R. Gordon Arneson for Secretary of State Acheson, April 24, 1951, in FRUS, 1950, Vol. I; National Security Affairs; Foreign Economic Policy (Washington, D.C.: U.S. Government Printing Office, 1979), pp. 820-826. See also Hewlett and Duncan, Atomic Shield, p. 539. 77. "The President's Farewell Address to the American People, January 15, 1953," in Public Papers of the Presidents: Harry S Truman, 1952-1953 (Washington, D.C.: U.S. Government Printing Office, 1966), pp. 1197-1202.

Nuclear Strategy, 1945-1960 \ 157

population and murders them by wholesale."78 Whether Truman deliberately avoided Murray's question, or simply had not understood it, will probably never be known. Eisenhower's New Policies and Programs

Dwight D. Eisenhower entered the presidency in January 1953 with a more thorough knowledge of nuclear weapons than any president before or since. As Army Chief of Staff in 1946, he had jurisdiction over the Manhattan Engineer District and was the sole channel to high civilian policymakers for nuclear stockpile information. He was the chief advocate for the development of the first JCS nuclear war plan in 1947, and in 1949, while temporary chairman of the JCS, championed increased nuclear production. As the first NATO Supreme Allied Commander from 1950 to 1952, he became familiar with JCS targeting categories and priorities, and encouraged planning for the tactical nuclear defense of Europe.79 Eisenhower was not awed by nuclear weapons, but neither was he sanguine about them. In March 1953, he admonished his Cabinet against thinking of the bomb as "a cheap way to solve things." "It is cold comfort," he noted, "for any citizen of Western Europe to be assured that—after his country is overrun and he is pushing up daisies— someone still alive will drop a bomb on the Kremlin."80 Nevertheless, the President knew that nuclear weapons could not be "disinvented," and that they had perma­ nently altered the character of warfare.81 Shortly after taking office, he reversed Truman's long-standing policy of near complete civilian control of the atomic stock­ pile, and took steps to make the bombs immediately available to the military. On June 20, 1953, responding to a Defense Department request for custody of the entire national war stockpile, he transferred a sizable number of complete atomic weapons to the military for deployment to specified bases afloat and ashore.82 Such transfers would continue throughout his administration, both to decrease the vulnerability of the stockpile through dispersal and to increase operational readiness, until by 1961

78. Letter, Truman to Murray, January 19, 1953; and letter Murray to Truman, January 16, 1953, both in Atomic Bomb Folder, General File, PSF, HSTL. 79. On Eisenhower's nuclear background, see in particular Rosenberg, "U.S. Nuclear Stockpile, 1945 to 1950," pp. 27-28; JCS 1725/2, 12 March 1947, CCS 004.04 (11-4-46) Sec. 3, JCS; Letter, Major General K.D. Nichols, USA (Ret.) to David A. Rosenberg, October 27, 1980; Oral History, General Lauris Norstad, Interview by Dr. Thomas Soapes, 11 November 1976, DDEL, pp. 6-29, 39-42; Senior Officers Debriefing Program, Conversations between General Andrew J. Goodpaster and Colonel William D. Johnson and Lt. Colonel James C. Ferguson, 29 January 1976, pp. 56-57; 25 February 1976, pp. 6-14; U.S. Army Military History Research Collection, U.S. Army Military History Institute, Carlisle Barracks, Pa. (hereafter USAMHI). 80. Emmet Iohn Hughes, The Ordeal of Power (New York: Dell Books, 1964), p. 88. 81. Oral History, General Andrew J. Goodpaster, Interview by Dr. Thomas Soapes, Ianuary 16, 1978, DDEL, pp. 104-106. 82. Memo, C.E. Wilson, to Chairman, ICS, Subject: Transfer and Deployment of Atomic Weap­ ons, 24 June 1953, in JCS 2019/62, 25 June 1953, CCS 471.6 (8-15-45) Sec. 39, JCS. The original JCS request for custody is in Memo, William M. Fechteler for the Joint Chiefs of Staff to the Secretary of Defense, Subject: Custody of Atomic Weapons, 11 March 1953, in Atomic Energy-

Strategy and Nuclear Deterrence \ 138

less than 10 percent of the stockpile remained in civilian control.83 Where Harry Truman viewed the atomic bomb as an instrument of terror and a weapon of last resort, Dwight Eisenhower viewed it as an integral part of American defense, and, in effect, a weapon of first resort. The new President was not happy with the crisis-oriented defense program and record peacetime defense budgets he had inherited from the Truman Administration. He was convinced that budgeting to meet a projected "year of maximum danger" would result in runaway expenditures and undermine the American economy. Be­ lieving that the United States must approach military preparedness "only on a defen­ sive, which means a long-term basis," he sought to develop a balanced, integrated defense posture for the indefinite future.84 EISENHOWER'S NUCLEAR POLICIES

It took the Eisenhower Administration almost a year to thrash out its national security policy and defense program. This multi-level process included NSC studies on se­ curity policy, economic defense, and the defense of the continental United States; a high level review of the policy options of containment, spheres of influence, and rollback codenamed Operation SOLARIUM; and an analysis of defense strategies and priorities prepared by the newly appointed Joint Chiefs of Staff.85 On October 30, 1953, Eisenhower approved NSC-162/2 as his administration's Basic National Security Policy. NSC-162/2 explored the nature of the Soviet threat, the implications of U.S. alliances and foreign commitments, and the economic ramifica­ tions of defense, and concluded that national security required: (1) A strong military posture, with emphasis on the capability of inflicting massive retaliatory damage by offensive striking power;

Miscellaneous 1953-1954 Folder #2, NSC Series, Subject Subseries, Box 1, White House Office, Special Assistant, National Security Affairs (hereafter WHO-SANSA), DDEL. The record of JCS discussions and presidential decisions through 1953 is in the JCS 2019 series, CCS 471.6 (8-1545) from 1949 on, and the Folders: "NSC-Atomic Weapons, Agreed Concepts," "NSC-Atomic Weapons, Non-Nuclear Components," and "NSC-Atomic Weapons, Department of Defense," all in NSC-Atomic File, PSF, HSTL. 83. Andrew J. Goodpaster, MCP, January 13, 1961, dated January 17, 1961, in AEC, December 1960-January 1961, Vol. Ill, Folder #6, Subject Series, Alphabetical Subseries, WHO-SS, DDEL. Eisenhower's philosophy on dispersal is explored in A.J. Goodpaster, Notes on Meeting with the President, 1 December 1954, with Letter, Secretary of State to the President, December 1, 1954 and Letter, Eisenhower to Lewis L. Strauss, 1 December 1954, Atomic Energy Commission, 1953-1954, Folder 2, Administration Series, ACWF, EPP, DDEL. 84. Diary entry, January 22, 1952, in Robert H. Ferrell, The Eisenhower Diaries (New York: W.W. Norton, 1981), p. 209. 85. The process is best documented in John Lewis Gaddis, Strategies of Containment: A Critical Appraisal of Postwar American National Security Policy (New York: Oxford University Press, 1982), pp. 127-163; Kinnard, President Eisenhower and Strategy Management, pp. 1-36; Glenn H. Snyder, "The New Look of 1953," in Warner R. Schilling et al., Strategy, Politics, and Defense Budgets (New York: Columbia University Press, 1962), pp. 379-524; and Samuel F. Wells, Jr., "The Origins of Massive Retaliation," Political Science Quarterly, Vol. 96 (Spring 1981), pp. 31-52.

Nuclear Strategy, 1945-1960 \ 139

(2) U.S. and allied forces in readiness to move rapidly initially to counter aggression by Soviet bloc forces and to hold vital areas and lines of communication; and (3) A mobilization base, and its protection against crippling damage, adequate to insure victory in the event of general war.86 In December 1953, a three-year defense program was approved that reflected these priorities. 1. OFFENSIVE STRIKING POWER. The high priority assigned to massive retaliatory striking power meant that the Air Force would continue to play a central role in national defense. Although the Navy's 16 attack aircraft carriers were considered an integral part of the nation's "offensive striking power," their air groups were not oriented exclusively to nuclear attacks on the Soviet homeland, nor were the "targets of naval interest" they would strike in a" war the same as the BRAVO and DELTA targets assigned to SAC by the JCS. By the end of 1953, SAC contained 10 heavy and 25 medium bomb and reconnaissance wings, nearly 23 of which were considered combat ready, along with 28 refueling squadrons, totalling, in all, over 1,500 aircraft, including 1,000 nuclear-capable bombers. The propeller driven medium bomb and reconnaissance force was in the process of converting to the all-jet 600 mile per hour B-47, greatly enhancing SAC striking power.87 Although the Air Force had borne the brunt of Eisenhower's $5 billion in defense budget cuts in the spring of 1953, the object, the President explained, was not to downgrade SAC, but to get "as much effectual strength as we can immediately" by assuring that all operational wings would be manned and equipped, and "then concentrating on a steady program."88 The Fiscal Year 1954 to 1957 defense program approved in December 1953 had an ultimate objective of 137 wings. During these three years, the Air Force received an average share of 47 percent of total defense appropriations, compared to 29 percent for the Navy and 22 percent for the Army.89 2. TACTICAL NUCLEAR WEAPONS. The second element of Eisenhower's nuclear strat­ egy was defense against local aggression, especially in Western Europe. This was to be achieved through a manifest willingness to use not only "massive retaliatory striking power," but also tactical nuclear weapons. By the fall of 1953, efforts had been underway for nearly two years to prepare the tactical nuclear defense of Europe. In January 1952, the JCS authorized General Eisenhower to begin planning for the

86. NSC 162/2, Review of Basic National Security Policy, October 30, 1953, NSC papers, Modern Military Branch, National Archives (hereafter NSC-MMB). These statements are also contained in NSC 162, September 30, 1982 and NSC 162/1, October 19, 1953, ibid. 87. Strategic Air Command Progress Analysis, 1948-1953, pp. 1-25. The nature of "Offensive Striking Power" is discussed in JCS Info Memo 922, 10 February 1954, CCS 381 U.S. (1-31-50) Sec. 35, JCS. 88. L.A. Minnich, Jr., Legislative Leadership Meeting, May 12, 1953, Supplementary Notes, Staff Notes, January-December 1953, DDE Diary, Box 4, ACWF-EPP, DDEL. The controversy over Air Force budget cuts is recounted in E. Bruce Geelhoed, Charles E. Wilson and Controversy at the Pentagon, 1953 to 1957 (Detroit: Wayne State University Press, 1979), pp. 59-80. 89. Alfred Goldberg, ed., A History of the United States Air Force, 1907-1957 (Princeton, N.J.: D. Van Nostrand, 1957), p. 117.

Strategy and Nuclear Deterrence \ 140

use of a specific number of atomic bombs by already deployed Navy and soon to be deployed Air Force tactical air units.90 One month later, Project VISTA, a scientific and technical study completed under the auspices of the California Institute of Tech­ nology, concluded that "the tactical employment of our atomic weapons resources holds outstanding promise" for defending Western Europe.91 A draft chapter had even proposed that nuclear weapons be employed initially only in a tactical role, and that the strategic offensive be withheld, for humanitarian as well as political and military reasons. General Lauris Norstad, Commander, U.S. Air Forces, Europe, subsequently convinced most of the scientists responsible for the draft (although not the most prominent one, J. Robert Oppenheimer) that tactical nuclear air forces would not be fully constituted for several years, and that SAC would be critical in halting any attack in the interim. The final VISTA report made no mention of SACs offensive, but emphasized small weapons with yields of 1 to 50 kilotons, and urged the creation of a Tactical Atomic Air Force that would rival SAC in both professionalism and appropriations.92 By 1953, technological advances were rapidly expanding possibilities for tactical and other uses of nuclear weapons. The 10,000-pound implosion fission bombs of 1945-1949 were being replaced by the 3,000-pound Mark 5 and the 2,700-pound Mark 7. The 1,000-pound Mark 12 would be available in 1954. Each of these weapons could detonate cores with yields as high as their larger predecessors, and could be fitted to most existing and planned Air Force fighter bombers and Navy attack planes. In addition, they could be utilized as warheads for the Regulus and Honest John surfaceto-surface guided missiles, shells for long-range artillery, depth charges for anti­ submarine warfare, atomic demolition land mines, and, eventually, warheads for surface-to-air and air-to-air missiles.93 Despite the increasing feasibility of tactical nuclear weapons, such weapons had not been fully integrated into operational plans at the time Eisenhower took office. The reluctance of civilian policymakers to use nuclear weapons in Korea had created doubts in the minds of many military planners as to when, if ever, such use would

90. JCS 2220/4, 31 January 1952, with SM-271-52, 28 January 1952 appended, CCS 471.6 (4-1849) Sec. 7, JCS. Preparations for tactical nuclear planning are well documented in a series of "Red Line" Messages between Lauris Norstad and Hoyt Vandenberg from March through December 1951, Box 86, Vandenberg Papers, LC; Norstad Oral History, DDEL, pp. 6-42; and Goodpaster Conversations, February 1976, USAMHI, pp. 6-14. 91. Final Report, Project VISTA, A Study of Ground and Air Tactical Warfare with Especial Reference to the Defense of Western Europe, Pasadena, California, California Institute of Tech­ nology, 4 February 1952, Modern Military Branch, National Archives, pp. xxi-xxvi. 92. Red Line Message, Personal to Vandenberg from Norstad, 7 December 1951, Red Line Messages, 1 May-31 December 1951 Folder, Box 86, Vandenberg Papers, LC. See also Charles J.V. Murphy, "The Hidden Struggle for the Η-Bomb," Fortune, May 1953, pp. 109-110, 230; VISTA Final Report, pp. xxi-xxviii, 138-142, 165-192. 93. Information on the size and capabilities of smaller nuclear weapons came from the sources cited in notes 70 and 71, ante; undated list, Los Alamos Stockpiled Nuclear Weapon Designs, Los Alamos Scientific Laboratory (copy courtesy Thomas B. Cochran); and documents in the JCS 2012 series through 1954, CCS 471.6 (5-31-44), JCS.

Nuclear Strategy, 1945-1960 | 141

be authorized.94 NSC-162/2 was the first statement of national policy to directly address this issue. The first draft, dated September 30, contained only an ambiguous reference to the use of "special weapons" in the interests of "national security."95 During the NSC meeting on October 7, that sentence was strengthened to read: "In the event of hostilities, the United States will consider nuclear weapons to be as available for use as other munitions." This statement, approved as national policy on October 30, was seen by the JCS as an assurance that they could confidently plan for the use of nuclear weapons in both limited and general conflicts. The December 1953 defense program identified, for the first time, "the provision of tactical atomic support for U.S. or allied military forces in general war or in a local aggression" as a major component of U.S. military posture and strategy.96 Over the next few years tactical nuclear weapons were increasingly integrated into calculations of force requirements and effectiveness. By the late 1950s, requirements for tactical nuclear weapons were routinely submitted by all unified and specified commanders for inclusion in the overall JCS request which established atomic production levels.97 3. DEFENSE AGAINST NUCLEAR ATTACK. The third element in Eisenhower's nuclear strategy, as spelled out in NSC-162/2, was the defense of "our striking force, our mobilization base, and our people." From the beginning of the nuclear age, the problem of direct defense against nuclear attack had appeared all but insurmountable. It was estimated in 1950 that just sixteen atomic weapons, if properly targeted, could "most seriously disrupt" the U.S. government.98 Initial air defense efforts were started by the Air Force in 1947, and a number of scientific studies of the problems of air and civil defense were undertaken. It was not until December 1952, however, that President Truman authorized construction of an early warning radar system which would provide three to six hours notice of approaching bombers. One month later, he created a special subcommittee of the NSC to "evaluate the net capabilities of the Soviet Union to inflict direct injury on the United States, up to July 1, 1955."99 94. Memorandum, Kirkpatrick to Admiral Radford, 7 November 1955, Subject: Review of Events Leading Up to JCS Action of 10 December 1953 concerning "Military Strategy and Posture," CJCS 381, Military Strategy and Posture; Memorandum for Admiral Radford by "W," 11 Sep­ tember 1954, CJCS 381 (Continental Defense) January-December 1954; and Memorandum, E.H.J. Cams to Admiral Radford et al., Subject: Estimate of Military Risk, SM-1721-53, 16 October 1953, CCS 381 U.S. (1-31-50) Sec. 30, all JCS. 95. NSC 162, 30 September 1953, NSC-MMB, p. 26. 96. JCS 2101/113, 9 December 1953, with decision, 10 December 1953, CCS 381 U.S. (1-31-50) Sec. 31, JCS. Copy also in Charles E. Wilson, Secretary of Defense 1953, Folder #2, Box 44, Administration Series, ACWF-EPP, DDEL. 97. For example, see Message CINCLANT to JCS, 14 January 1959, Subject: Fiscal Year 1960 Weapons Requirements, CCS 4614 (Allocation) 6 January 1959; JCS 1844/299, 19 November 1959, and SM-1207-59, H.L. Hillyard to Chief, Defense Atomic Support Agency, 27 November 1959, Subject: Allocations of Atomic Weapons, CCS 4614 (Allocation) 20 April 1959, Group 2, all JCS. 98. Presentation to the National Security Council, 1 February 1950, "Need of Defense Measures Against Increasing Threat of Atomic Attack Against the Continental United States," appended to Memorandum, Alfred M. Gruenther to the Deputy Secretary of Defense, 31 January 1950, CD-22-2-2, Record Group 330, Papers of the Office of the Secretary of Defense, Modern Military Branch, National Archives. 99. NSC-139, An Early Warning System, December 31, 1952, NSC Memoranda Approvals, Box

Strategy and Nuclear Deterrence \ 142

The Net Evaluation Subcommittee reported to the NSC on May 18, 1953. Its still classified findings are reflected in a June 1953 study which stated that the continental defense programs then underway were not now adequate either to prevent, neutralize, or seriously deter the military or covert attacks which the USSR is capable of launching, nor are they adequate to ensure the continuity of government, the continuity of production, or the protection of the industrial mobilization base and millions of citizens in our great and exposed metropolitan centers.100 This, the NSC concluded, "constitutes an unacceptable risk to our nation's survival." Although complete invulnerability was impossible to achieve, "a reasonably effective defense system can and must be obtained." The NSC recommended expanding U.S. nuclear deliveiy capability to match the growing Soviet Air Force, and developing a defense and early warning system adequate to "prevent disaster, and to make secure the mobilization base necessary to achieve U.S. victory in the event of general war."101 The pressure to develop a "reasonably effective" system of continental defense increased two months later, when the Soviet Union tested a 300 to 400 kiloton boosted fission weapon and publicly claimed that it had achieved thermonuclear capability.102 NSC-162/2 projected that the Soviet Union might soon have the ability to strike a "crippling blow" against the U.S. industrial base in a surprise attack. From this time on, the Net Evaluation Subcommittee was reconstituted annually to assess the balance between Soviet and American nuclear capabilities.103 During the next seven years, the Eisenhower Administration gave high priority to completing radar early warning networks extending above the Arctic Circle; developing and deploying conventional, and beginning in 1957, nuclear-armed jet interceptor squadrons and surface-to-air missile batteries throughout the United States; completing a low frequency analysis and recording sound surveillance system (SOSUS), for the distant detection of sub­ marines; constructing alternate joint and national command posts for the relocation of governmental and military leadership in the event of attack; and initiating an annual series of Operation ALERT exercises, in which even the President -and the Cabinet took part, to test the ability of the government to relocate and continue functioning in an emergency.104 193, Subject File, PSF, HSTL. NSC-140, January 19, 1953, and NSC-140/1, May 18, 1953, both listed with titles in Annotated List of Serially Numbered National Security Council Documents, appended to Memo by James S. Lay, Jr. for the NSC, Februaiy 17, 1959, in Subject Filing of NSC Papers Folder, Subject Series, Alphabetical Subseries, Box 20, WHO-SS, DDEL. 100. NSC-159, Report of the Continental Defense Committee, July 22, 1953, quoted in NSC5408, Continental Defense, February 11, 1954, NSC-MMB. 101. NSC-153/1, June 10, 1953, quoted in ibid. 102. Holloway, "Soviet Thermonuclear Development," pp. 192-197; National Intelligence Esti­ mate 11-3-55, Soviet Capabilities and Probable Courses of Action through 1960, 17 May 1955, in Folder, same title, NSC File, Subject File, Box 10, WHO-SANSA, DDEL, pp. 24-25. 103. NSC-5423, June 23, 1954; NSC-5511, Febmary 14, 1955; NSC-5605, May 24, 1956; NSC5728, December 24, 1957; NSC-5816, July 1, 1958, all contain directives to the Net Evaluation Subcommittee. Annotated List of Serially Numbered NSC Documents, February 17, 1959. 104. NSC-5408; NSC-5606, Continental Defense, June 5, 1956; NSC-5707/4, Review of Basic

Nuclear Strategy, 1945-1960 | 143

AMERICA'S FIRST-STRIKE OPTIONS: PREVENTIVE WAR AND PREEMPTION

Anxiety about the threat which the United States would soon be facing generated serious consideration of whether the nation might have to initiate war in its own defense. Although seldom explicitly discussed in writing, preventive war was implicit in some of the major policy deliberations of the time. In the spring of 1953, Eisen­ hower's top foreign policy advisers rejected a proposal by the SOLARIUM Steering Committee, headed by retired Air Force Reserve General James Doolittle, that a fourth policy alternative be studied: giving the Soviet Union two years to agree to terms, with the understanding that failure to cooperate might lead to general war.105 In August 1953, an Air Force study of "The Coming National Crisis" raised a similar argument, noting that the time was approaching when the U.S. would find itself in a "militarily unmanageable" position. Before that time arrived, the nation would have to choose whether to trust its future to "the whims of a small group of proven barbarians" in the USSR, or "be militarily prepared to support such decisions as might involve general war."106 President Eisenhower and Secretary of State John Foster Dulles seriously discussed this dilemma in August 1953. When the Soviet Union acquired thermonuclear weap­ ons, Eisenhower wrote in a summary memorandum, the United States might find itself vulnerable to a crippling strike. Its only security would then lie in the deterrent effect of its ability "to inflict greater loss against the enemy than he could reasonably hope to inflict upon us." But if the contest to maintain this relative position should have to continue indefi­ nitely, the cost would either drive us to war—or into some form of dictatorial gov­ ernment. In such circumstances, we would be forced to consider whether or not our duty to future generations did not require us to initiate war at the most propitious moment we could designate.107 In March 1954, a new statement of U.S. objectives for general war was approved, to replace that in NSC-20/4. The first objective enunciated in the new document, NSC-5410/1, was to "achieve a victory which will insure the survival of the United

National Security Policy: Military and Non-Military Aspects of Continental Defense, March 26, 1957; NSC-5802, Continental Defense, February 3, 1958; and NSC-5802/1, February 19, 1958, in Boxes 6, 9, 17, 20, and 23 of NSC Series, Policy Papers Subseries; and Department of Defense Progress Report on Status of Military Continental Defense Programs, 1 June 1954, Study of Continental Defense by R. Sprague 2953-4, FoJder #6, NSC Series, Subject Subsenes, all WHOSANSA, DDEL. See also papers in the JCS 1899 and 1851 series, CCS 381 U.S. (5-23-46) through 1958, and 3210 Air Defense Operations, by date, thereafter; and the JCS 2084 Series, CCS 373.24 U.S. (9-8-49) through 1958, all JCS; and discussions of Operation ALERT in Cabinet Meeting Minutes in the DDE Diary, ACWF-EPP, DDEL. 105. Snyder, "The New Look of 1953," pp. 407-409. 106. Major General Robert M. Lee, USAF, Memorandum for the Chief of Staff, U.S. Air Force, 21 August 1953, Subject: The Coming National Crisis, Top Secret File (1), 1952-1957, Subject File, Box 121, Nathan F. Twining Papers, LC. 107. Memorandum, Eisenhower to Dulles, September 8, 1953, DDE Diary, August-September 1953, Folder 2, DDE Diary, Box 3, ACWF-EPP, DDEL.

Strategy and Nuclear Deterrence | 144

States."108 Formal recognition that the survival of the U.S. might be in doubt generated renewed pressures for consideration of preventive war. In May 1954, Eisenhower was briefed on a paper by the JCS's Advance Study Group which proposed that the U.S. consider "deliberately precipitating war with the USSR in the near future," before Soviet thermonuclear capability became a "real menace." Army Chief of Staff Matthew Ridgway denounced this course as "contrary to every principle upon which our Nation had been founded," and "abhorrent to the great mass of American people."109 The President did not commit himself at this briefing, but subsequently approved an updated Basic National Security Policy paper in the fall of 1954 which stated unequivocally that "the United States and its allies must reject the concept of preventive war or acts intended to provoke war."110 If preventive war was ruled out as a strategic option, preemption was not. In this era before ballistic missiles, preemption appeared to be both militarily and constitu­ tionally feasible. Warning of a Soviet attack—even a surprise attack—was still being calculated by the CIA in terms of days or even weeks, because of the time needed to prepare Soviet forces and bases for strikes on the United States. It was estimated that even after the first wave struck there would be time to blunt an attack, since it would fake up to thirty days for the Soviets to deliver all of the nuclear weapons.111 Response to an impending or actual strike, Eisenhower told a group of congressmen in January 1954, "will be a very quick thing as fast as Congress can meet." In fact, he added, "if you were away and I waited on you (before taking retaliatory action) you'd start impeachment proceedings against me."112 If war broke out in Western Europe, he told the JCS later that year, the United States would employ tactical nuclear weapons to hold the line, while "first priority" was given to launching SAC "to blunt the enemy's initial threat." After "averting disaster" in this initial phase, the U.S. would go on to mobilize to win the war over a possibly extended period.113 No change in JCS policy was required to implement this scenario. The blunting mission was maintained as the highest targeting priority, followed by retardation of Soviet troop advances and the destruction of the urban-industrial base and govern­ ment control centers. General LeMay, the man who would determine "the exact

108. NSC 5410, United States War Objectives, 19 February 1954, and NSC 5410/1, 29 March 1954, in NSC 5410/1 U.S. War Objectives Folder, Box 9, NSC Series, Policy Papers Subseries, WHO-SANSA, DDEL. See also Rosenberg, "Ά Smoking Radiating Ruin,'" p. 29n. 109. General Matthew B. Ridgway, Memorandum for the Record, 17 May 1954, Historical Record, 15 January to 30 June 1954 Folder, Box 30, Matthew B. Ridgway Papers, USAMHI. 110. NSC 5440/1, December 28, 1954; approved as Basic National Security Policy in NSC 5501, January 6, 1955, NSC-MMB, Paragraph 35. See also M.B. Ridgway, Memo for the Special Assistant to the President for National Security Affairs, Subject: Review of Basic National Security Policy, 22 November 1954, Historical Record Folder, Box 30, Ridgway Papers, USAMHI. 111. Draft Special N.I.E., 11-8-54: Probable Warning of Soviet Attack on the U.S. Through Mid1957, 10 September 1954, SNIE 11-8-54 Folder, Box 47, WHO-SANSA, DDEL. 112. L.A. Minnich, Minutes, Bipartisan Legislative Meeting, January 5, 1954, Staff Notes, Jan­ uary-December 1954 Folder, Box 4, DDE Diary, ACWF-EPP, DDEL. 113. A.J. Goodpaster, MCP, 22 December 1954, ACW Diary Dec. 1954 (2) Folder, ACW Diary Series, Box 3, ACWF-EPP, DDEL.

Nuclear Strategy, 1945-1960 | 145

manner in which SAC will fight a war," was firmly committed to preparing his command to preempt Soviet nuclear delivery capability, believing that "if the U.S. is pushed in a corner far enough we would not hesitate to strike first." By 1953, SAC had identified 409 Soviet airfields which could be used to launch nuclear strikes, and the next year, LeMay proposed development of a "comprehensive plan for the defeat of Communist air power."114 SAC WAR PLANS AND CAPABILITIES

As of March 1954, SACs Basic War Plan called for up to 735 bombers to hit Soviet early warning screens from all directions simultaneously, using chaff and other elec­ tronic countermeasures, to overwhelm the extensive Soviet defenses. All designated targets would be struck in this single massive blow, in order to minimize the time U.S. bombers would have to remain in hostile air space, maximize destruction, and reduce the need for costly follow-up strikes. If adequate warning time were available, the basic "optimum" strike plan would be executed in full, although there were a "multitude of timing variations and options" to allow for such variables as aborts, malfunctions, and losses of bases and aircraft to a surprise Soviet attack.115 The combined SAC offensive against military and urban-industrial targets was not necessarily what the President and his top policy advisers had envisaged. In October 1953, JCS Chairman Admiral Arthur Radford unsuccessfully urged the NSC to estab­ lish a clear policy for the graduated use of nuclear weapons that gave top priority to striking "military forces operating against us or our allies," and targets that would support those forces, before launching "unrestricted atomic operations in retaliation for such attacks on the west."116 In June 1954, Eisenhower expressed his own pref­ erence for concentrating on military targets. "If we batter Soviet cities to pieces by bombing," he asked his JCS in June 1954, "what solution do we have to take control of the situation and handle it so as to achieve the objectives for which we went to war?" Eisenhower urged the war planners to use more imagination in envisioning "the best way to fight the next war in order that we could attain our national objectives with the minimum cost and the least dislocation to the world."117 Even within the Air Force community, there were critics of the SAC approach to strategic bombing. In 1952, Bernard Brodie had helped prepare a RAND Corporation study which warned that American cities were much more vulnerable to thermonu114. Document One in Rosenberg, "Ά Smoking Radiating Ruin,'" pp. 18, 27; Letter, Nathan Twining to LeMay, 17 June 1954, Twining Folder, Box B-60, LeMay Papers, LC. 115. Document One in Rosenberg,'"A Smoking Radiating Ruin/" pp. 24-26; Historical Division, Headquarters, Strategic Air Command, "Historical Study 73A, SAC Targeting Concepts" (April 1959, extracts only declassified, 1980), pp. 2, 3, 5-7; Lemmer, The Air Force and Strategic Deterrence, pp. 21-23. 116. A. W. Radford, Memorandum for Mr. John Foster Dulles, 13 October 1953, CJCS 040 Atomic Energy Commission, JCS. 117. L.L.L. (Lyman L. Lemnitzer) to General Ridgway, 21 June 1954, Historical Record, January to 30 June 1954 Folder, Box 30, Ridgway Papers, USAMHI. For a different report of the Presi­ dent's statement by Press Secretary James Hagerty, see Gaddis, Strategies of Containment, p. 135n.

Strategy and Nuclear Deterrence \ 146

clear attack than those in the Soviet Union, and recommended that the soon to be tested H-bomb be used strictly in a tactical role to reduce the dangers of escalation. The study was widely briefed in the Air Staff and the Defense Department, but was not well received. The Air Force, Brodie noted, proved "decidedly hostile to any notions that it might be necessary to find ways of limiting war by eliminating strategic bombing."118 By 1954, other RAND strategists led by James Digby, a number of Air Staff planners including the future head of ballistic missile development, Brigadier General Bernard A. Schriever, and the Evaluation Staff at the Air War College were quietly promoting a "no-cities" strategy which would focus U.S. nuclear strikes on Soviet nuclear ca­ pability and military forces. This strategy was publicly advocated by former Air Force Secretary Thomas K. Finletter as well.119 The Air Force high command, however, continued to favor the proposed combined offensive. In the fall of 1953, the Weapons Systems Evaluation Group completed a year-long assessment of what impact the planned SAC offensive would have. Count­ ering the arguments put forward by the VISTA report and RAND, WSEG-IO con­ cluded that unless strikes were carried out against industrial and government control centers, the Soviet Union "could support immense armed forces for at least two years of intensive warfare." Even with the successful completion of these strikes, it might take four to six months for fuel and lubricant shortages to affect Soviet forces in the field. That period could be reduced, however, by application of greater force. In fact, it now appeared that "some level of atomic attack would be decisive against the civicpolitical structure of the interior of Russia." The possibility that SAC might be able to deliver a single war-winning blow was an irresistible temptation to Air Force planners.120 Within two years, that capability appeared to be almost within reach. In the spring of 1954, the United States began testing thermonuclear weapons with yields as high as fifteen megatons, a level which was thought by both the AEC and the military to be "as large as need is foreseeably seen."121 With one of these weapons, which would begin entering the stockpile within the year, SAC could destroy a city, a hardened

118. Brodie, War and Politics, p. 394n. This episode is also discussed in Norman Moss, Men Who Play God: The Story of the Hydrogen Bomb (New York: Harper and Row, 1969, Penguin Books Reprint with revisions, 1972), pp. 58-60. 119. Alfred Goldberg, A Brief Survey of the Evolution of Ideas About Counterforce, RM-5431-PR (Santa Monica, Calif.: The Rand Corporation, October 1967, Revised March 1981), pp. 8-12; Lemmer, The Air Force and Strategic Deterrence, pp. 55-57. 120. Paper, no author, "Comments on WSEG #10 for General Twining," n.d. (ca. early 1954), Ready File, General Twining (1954), Box 120, Twining Papers. The tasking of WSEG #10, Evaluation of the Effects of the Mid-1954 First Phase Atomic Offensive Against Fixed Industrial Targets in the Soviet Bloc, is in CCS 373 (10-23-48) Sec. 6, 7, and 8, JCS. 121. Announced U.S. Nuclear Tests, p. 4; Entry, April 5, 1954, Phone Calls, January-May 1954— 1, DDE Diary, Box 5, ACWF-EPP, DDEL. Even before this test, the JCS had increased their request for thermonuclear weapons production. See JCS 2096/19,18 January 1954, with Decision, 19 January 1954, CCS 471.6 (12-14-49) Sec. 4, JCS.

Nuclear Strategy, 1945-1960 | 147

command center, or an airfield. Possibilities for bonus damage and destruction of multiple targets were also greatly enhanced, factors which were still "thought to be quite important in view of the many targets requiring destruction and the limited size of the Strategic Air Command."122 In June 1955, SAC received the first of its eight-jet B-52 heavy bombers. The B-52, which could carry up to four megaton yield bombs over an unrefueled combat radius of 3,000 miles at more than 550 miles per hour, would entirely replace the lumbering B-36 by 1959.123 SACs expanding capability was enhanced by its growing independence. From 1951 on, General LeMay did not submit his annually updated Basic War Plans as required for JCS review, believing that the details of operational planning should be closely guarded. In June 1955, Air Force Chief of Staff Twining sent LeMay a formal request for information on SAC missions, priorities, and concept of operations to facilitate inter-service coordination. That October, LeMay provided a briefing to the JCS, and a limited number of summaries of the current plan.124 In the fall of 1955, SAC achieved virtual control over target selection. In October 1954, Army Chief of Staff Ridgway, noting that SAC's target list was not fully consistent with the JSCP list approved by the JCS, urged tighter control over SAC planning. "I consider it imperative," he wrote, that the Joint Chiefs of Staff insure that the great striking power of the Strategic Air Command be employed in accordance with the sound military principle of economy of force and in accordance with a national policy which seeks to attain national objectives without indiscriminate mass destruction of human life.125 In April 1955, Ridgway proposed establishing a Joint Target Selection and Evaluation Group to achieve this objective.126 The Joint Staff committees assigned to study his proposal, however, reported back after Ridgway's retirement that joint, Pentagonbased planning had become infeasible because of the expanded applications of nuclear weapons to targets which could not be pre-identified. They recommended that two joint committees prepare the damage criteria to guide target selection, but that the actual selection process be delegated to the unified and specified commands, includ­ ing SAC. This recommendation was approved by the JCS on November 18. Although SAC's target list was still subject to JCS approval, the Joint Staff had neither the time nor the resources to thoroughly evaluate the finished product, particularly since SAC had just acquired its own IBM 704 computer to analyze and prioritize the increasingly

122. Futrell, Ideas, Concepts, Doctrine, pp. 551-552. 123. Office of the Historian, Headquarters, Strategic Air Command, Development of Strategic Air Command (Omaha, Neb.: Office of the Historian, Headquarters, Strategic Air Command, 1976), pp. 43-53, 77. B-52 capabilities are described in Walter Boyne, Boeing B-52, A Documentary History (New York: Jane's Publishing Company, 1981), pp. 135-137, 147. 124. Letter, Nathan Twining to Curtis LeMay, 9 June 1955; LeMay to Twining, 18 June 1955; and Twining to LeMay, 22 July 1955, Twining Folder, Box B-60, LeMay Papers, LC; CSAFM 284-55, 7 October 1955, appended to JCS 2057/32, 7 October 1955, CCS 381 (12-1-47) Sec. 5, JCS. 125. JCS 2057/47, 18 October 1954, CCS 373.11 (12-14-48) Sec. 20, JCS. 126. JCS 2056/71, 12 April 1955, forwarding Ridgway Memo of 9 April 1955, ibid., Sec. 23.

Strategy and Nuclear Deterrence \ 148

complex target data base. Since none of the other services could duplicate this ca­ pability, they found it even more difficult to intervene in target planning and thus interrupt the ongoing bootstrapping process.127 THE GROWING SOVIET NUCLEAR THREAT

While U.S. capability was expanding, Soviet capability was kept under constant review. In March 1954, President Eisenhower urged the Science Advisory Committee of the Office of Defense Mobilization (ODM) to undertake a study of the problem of surprise attack in the thermonuclear age. Modern weapons, he noted, "had made it easier for a hostile nation with a closed society to plan an attack in secrecy and thus gain an advantage denied to the nation with an open society." The Technological Capabilities Panel (TCP), chaired by James R. Killian, Jr., president of the Massachu­ setts Institute of Technology, was convened to prepare a future-oriented study of this problem.128 The February 1955 report of the TCP, "Meeting the Threat of Surprise Attack," provides an important benchmark in the evolution of Eisenhower's thinking about nuclear strategy. The report focused on the technical aspects of defense, and at­ tempted to project "a sense of urgency without despair." The Soviet Union already had enough mid-range bombers and bombs of up to one megaton yield, it concluded, to seriously damage the United States, and this threat would expand geometrically as the Soviets acquired true thermonuclear weapons and long range delivery aircraft. The U.S. could expect to maintain its strategic superiority for three to five years, but would be vulnerable to a devastating and possibly decisive surprise attack during that time. The report gave highest priority to improving intelligence and tactical warning capabilities, and preparing for instantaneous response, including the use of nuclear armed air defense missiles. It urged the President to proceed with the dis­ persion of nuclear weapons to offensive and defensive forces, and to grant "advance authority for the instant use of the atomic warheads wherever needed over the land areas of the United States and Canada."129 The TCP report also reaffirmed the JCS BRAVO, ROMEO, and DELTA targeting priorities, while stressing the interrelationship between offensive and defensive forces: "Our striking forces must blunt the attack at its source: defense must protect our retaliatory power as well as our people and our cities. Together they provide overall strength and a substantial deterrent to war."130 Although dispersing the

127. JCS 2056/75, 10 November 1955, with Decision, 18 November 1955, ibid., Sec. 24; "Strategic Air Command Progress Analysis, 1948-1956," pp. 65-69. 128. James R. Killian, Jr., Sputniks, Scientists, and Eisenhower (Cambridge, Mass.: The MIT Press, 1977), pp. 68-71. 129. "Meeting the Threat of Surprise Attack," The Report to the President by the Technological Capabilities Panel of the Science Advisory Committee, Office of Defense Mobilization, February 14, 1955, Technological Capabilities Panel of the S.A.C., Report to the President, February 14, 1955 Folder, Subject Series, Alphabetical Subseries, Box 11, WHO-SS, DDEL, Vol. I, pp. 10-22, 31-46, Vol. II, pp. 50-71, 73-111. 130. Ibid., p. 17. See also Killian, Sputniks, Scientists, and Eisenhower, pp. 71-93.

Nuclear Strategy, 1945-1960 | 149

bomber force would provide interim protection against surprise attack, the best solution, the TCP concluded, would be to ensure that the bombers were enroute to their targets before any attack arrived. It also warned that the United States must prepare for a period of transition and instability after about 1958 or 1960, during which the advantage would rest with whichever side acquired and deployed ballistic missiles in threatening numbers, and urged a continued commitment to ballistic missile development programs. Eisenhower was very much impressed by the TCP report, and from this time on relied increasingly on the advice of scientists, whom he viewed as honest brokers with regard to the complex and often politicized issues of nuclear strategy. A second report issued in February 1955 provided a more detailed analysis of U.S. thermonuclear capability, and raised a disturbing question not addressed by the TCP. What if the U.S. no longer had the capability to blunt an impending Soviet attack "at its source"? WSEG-12, a study by the Weapons System Evaluation Group, projected that the atomic offensive outlined in the JSCP for Fiscal Year 1956 might have achieved the ultimate goal of ending a general war in a single blow. It would destroy virtually all Soviet atomic production capability, obliterate 118 out of 134 major cities, cause 60 million deaths, and "virtually eliminate the Soviet bloc industrial capabilities, and preclude any significant recuperation for at least one year." But WSEG-12 also pointed out that even if SAC destroyed the 645 targeted airfields, there would still be at least 240 remaining to which Soviet bombers could be dispersed and survive attack. "To achieve a high degree of assurance of destroying all known Soviet operational and staging bases," the report concluded, "would require an allocation of approximately twice that evaluated." Even these additional bombs and bombers, it added, "cannot prevent the Soviets launching a strike unless we hit first."131 The projection that the United States might no longer be able to substantially disarm the Soviet Union was based on potential, rather than actual, dispersion of the Soviet bomber force. The problem would become increasingly real as the size and range of the Soviet force expanded. In May 1955, a National Intelligence Estimate (NIE) set current Soviet medium bomber strength at more than 1,300. Some 1,100 of these were obsolete TU-4 "Bull" propeller driven imitations of the U.S. B-29, and the rest were "Badger" jet bombers equivalent to the B-47, but despite their deficiencies, all were considered capable of striking the U.S. on one-way missions. The NIE also estimated that the Soviet Union had some 20 "Bear" turbo-prop and 20 "Bison" jet heavy longrange bombers. Much more disturbing was the forecast that there would be approx­ imately 700 Soviet heavy bombers by mid-1959, as well as 700 "Badgers."132 Expanding Soviet delivery capability was complemented by advances in weapons technology. In November 1955, a 1.6 megaton test confirmed Soviet achievement of 131. Briefing of WSEG Report No. 12, "Evaluation of An Atomic Offensive in Support of the Joint Strategic Capabilities Plan," April 8, 1955, Document Two in Rosenberg, '"A Smoking Radiating Ruin/" pp. 29-38. WSEG-12's summary report was completed February 28, 1955, according to John Ponturo, Analytical Support for the Joint Chiefs of Staff: The WSEG Experience, 1948-1976 (Arlington, Va.: Institute for Defense Analyses Study S-507, July 1979), p. 101. 132. N.I.E. 11-3-55, 17 May 1955, pp. 30-31; Prados, The Soviet Estimate, pp. 38-47.

Strategy and Nuclear Deterrence \ 150

a "true" two-stage thermonuclear capability.133 With such weapons, the USSR could seriously damage the United States even if only a few bombers managed to escape destruction and penetrate U.S. defenses. The "bomber gap" was widely perceived as a technological or numerical challenge. But, Eisenhower never accepted the intelli­ gence estimates of such a gap, and his JCS were split over both the gap's existence and significance. To them, the 1956 controversy suggested something more complex and ominous: the declining U.S. ability to preempt and disarm an impending Soviet strike. In mid-January 1956, Eisenhower sought to impress on the NSC the appalling "chaos and destruction" the United States would suffer in a thermonuclear war. He urged the NSC to continue to maintain a five-year strategic materials stockpile, believing it was "crazy" to think that any war could be "won or lost in a period of thirty days." Nevertheless he asked his advisers to seriously consider what would happen when "we reach a point where we will have passed the limits of what human beings can endure."134 The President's concerns were echoed by the Net Evaluation Subcommittee report of January 23, 1956, which projected that by mid-1958, even with advance warning, the nation would be virtually defenseless against a devastating Soviet attack. The U.S. could probably inflict three times as much damage in retaliation, Eisenhower wrote in his diary, but there was little we could do during the month of warning in the way of dispersal of population, of industries, or of perfecting defenses that would cut down losses. The only possible way of reducing losses would be for us to take the initiative sometime during the assumed month in which we had the warning of an attack and launch a surprise attack against the Soviets.135 Since this would not only violate national tradition, but would require rapid, totally secret congressional action and immediate implementation, he added, "it would appear impossible that any such thing would occur."136 REAFFIRMING MASSIVE RETALIATION

The prospect that the United States might be confronting an enemy which it could virtually destroy, but could not effectively disarm, raised serious questions about massive retaliation as a strategy of national defense, and sparked renewed interest in exploring the possibilities and difficulties of nuclear deterrence. President Eisen­ hower, however, despite his increasing pessimism regarding the outcome of a general 133. Holloway, "Soviet Thermonuclear Development," pp. 192-197; Rear Admiral Edwin T. Layton, Memorandum for Chairman, Joint Chiefs of Staff, Subject: Implications of Soviet Ar­ maments Programs and Increasing Military Capabilities, 9 December 1955, CJCS 091 Russia, 1953-1955, JCS. 134. Memorandum, by S. Everett Gleason, Subject: Discussion at the 272nd Meeting of the National Security Council, Thursday, January 12, 1956, January 13, 1956, 272nd Meeting of the NSC Folder, NSC Series, Box 7, ACWF-EPP, DDEL. 135. Diary entry, January 23, 1956, Ferrell, ed., The Eisenhower Diaries, pp. 311-312. 136. Ibid.

Nuclear Strategy, 1945-1960 | 151

war, responded to the challenge by strengthening rather than revising his scenario for nuclear exchange. During the winter and spring of 1956 he took steps to bring U.S. force structures more clearly into line with a policy of instant nuclear response, and sought more explicit NSC endorsement of such a strategy. In April 1956, he proposed a $500 million supplemental defense appropriation to expand B-52 produc­ tion from 17 to 20 a month. Although there might be no winners in a thermonuclear war, the President told his JCS, "we don't want to lose any worse than we have to."137 From the beginning, the JCS had been divided over the concept of massive retal­ iation and in particular what force structure was needed to effectively implement it. The four-day marathon in August 1953 aboard the yacht Sequoia during which the Chiefs personally hammered out the New Look succeeded more as a result of ex­ haustion than real consensus. The recurrent theme in the JCS debates was to what degree the United States should rely on strategic air power to achieve its national security objectives. From 1953 until his retirement in 1955, Army Chief of Staff Ridgway repeatedly argued that greater emphasis should be placed on the problems of limited conflicts and the need for conventional forces. This argument was taken up even more forcefully by his successor, General Maxwell Taylor.138 By January 1956, primarily as a result of Army dissents, the JCS were deeply divided over the strategic concepts for the JSCP and JSOP. In late January, determined to unify the Chiefs behind the President's policy, Secretary of Defense Wilson requested a new three-year statement of force requirements and strategy, to replace the 1953 New Look. This statement was to be based on the assumptions, approved by Eisen­ hower, that "if general war should be forced upon us, the U.S. will employ atomic weapons from the outset of hostilities," and would employ them in lesser conflicts, in the absence of overriding political considerations, "whenever it is of military advantage to do so."139 Taylor was convinced that these assumptions were inconsis­ tent with both national interest and policy and took every opportunity to challenge

137. A.J. Goodpaster, MCP, February 10, 1956; Eisenhower's reactions to the Bomber Gap are recounted in Prados, The Soviet Estimate, pp. 43-50; Department of Defense, Semi Annual Report of the Secretary of Defense, January 1 to June 30, 1957 (Washington, D.C.: U.S. Government Printing Office; 1957), p. 8. 138. The "New Look" debates are chronicled by Stephen Jurika, Jr., ed., From Pearl Harbor to Vietnam: The Memoirs of Admiral Arthur W. Radford (Stanford, Calif.: Hoover Institution Press, 1980), pp. 327-330; and papers in the JCS 2101 series, CCS 381 U.S. (1-31-50) Sees. 26 through 32, JCS. Later debates are covered in the CJCS 381 (Military Strategy and Posture) files of JCS Chairman Radford's Office Files, JCS; and exemplified by JSCP debates in JSPC 877/299, 23 November 1955, CCS 381 (11-29-49) Sec. 27. See also Matthew W. Ridgway with Harold H. Martin, Soldier (New York: Harper and Row, 1956), pp. 266-322; and General Maxwell D. Taylor, The Uncertain Trumpet (New York: Harper and Row, 1959), pp. 23-46. 139. Memorandum, C.E. Wilson to Chairman, JCS, 27 January 1956, Subject: Military Strategy and Posture, and Memorandum, Lt. Colonel C.J. George to Wilson, February 6, 1956, CJCS 381 (Military Strategy and Posture) 1956, JCS. The Wilson memo is also in Military Planning 19567, Folder 2, Subject Series, Defense Department Subseries, Box 6, WHO-SS, DDEL. Significantly, this copy, declassified in 1980 vice 1978 for the copy in the JCS files, has the section on nuclear weapons use quoted here deleted.

Strategy and Nuclear Deterrence \ 152

them, despite Eisenhower's admonishments that the JCS must achieve consensus on such critical issues. On March 15, 1956, the NSC adopted a revised statement of basic national security policy which called for flexibility and emphasized the importance of containing limited conflicts, but ultimately deferred to presidential prerogative in establishing nuclear strategy. NSC-5602/1 declared that nuclear weapons "will be used in general war and in military operations short of general war as authorized by the President." It again rejected preventive war, while affirming national "determination to prevail if general war eventuates." Most significant, it anticipated the need for pre-authorization for the use of nuclear weapons. "Such authorization as may be given in advance," it stated simply, "will be determined by the President."140 On March 30, Eisenhower met with the JCS to explain his interpretation of this policy statement. "Any war in which Russian troops were involved directly against United States forces or the United States" would be general war, the President ruled, and any such Soviet attack would be met by launching SAC "as soon as he found out that Russian troops were on the move," while simultaneously requesting Con­ gress to declare war. "The attack once in the air could be recalled," he remarked, but "if it stayed on the ground it might not ever get off."141 On May 24, Taylor met with Eisenhower a final time to argue for modification of this strategy. By 1960, he argued, both the U.S. and the USSR would have such large thermonuclear stockpiles that "a situation of mutual deterrence must be envisaged." The greatest danger then would be posed by the possible escalation of an initially small conflict into general war. By focusing on the "worst case" scenario of a fullscale nuclear exchange, the U.S. was preparing for the least likely eventuality, and freezing out conventional forces needed to contain limited conflicts. Taylor also urged the President to reconsider the use of nuclear weapons in local conflicts, given the dangers of escalation. Eisenhower remained unconvinced. The use of tactical nuclear weapons would not provoke escalation any more than "twenty ton blockbusters," and he was determined not to let U.S. forces become bogged down in small conventional wars. Taylor's arguments, Eisenhower stated, depended "on an assumption that we are opposed by people who would think as we do with regard to the value of human life." If the Soviets decided to risk a "life and death struggle," they would certainly use atomic weapons at once, "and in full force." The only prudent course would be "to get our striking force into the air immediately upon notice of hostile action by the Soviets." Massive retaliation would be the "key to survival."142

140. NSC 5602/1, Basic National Security Policy, March 15, 1956, NSC 5602/1, Basic National Security Policy Folder, NSC Series, Policy Papers Subseries, Box 17, WHO-SANSA, DDEL, pp. 1-11.

141. A.J. Goodpaster, MCP, March 30, 1956, dated April 2, 1956, April 1956-Goodpaster Folder, DDE Diaries, Box 15, ACWF-EPP, DDEL; Arleigh Burke, Memorandum for the Record, Subject: Conversation with the President, 30 March 1956, Originator's File, AAB, NHC. 142. A.J. Goodpaster, MCP, May 24, 1956, May 1956-Goodpaster Folder, DDE Diaries, Box 15, ACWF-EPP, DDEL.

Nuclear Strategy, 1945-1960 \ 153

Recognizing defeat, Taylor abandoned his protest. The President's strategic con­ cepts requiring nuclear response in a general war were incorporated into both the JSCP and JSOP. The Army resigned itself for the time being to consequent force reductions, and by September had begun reorganizing its combat forces into "Pentomic" divisions to maximize their effectiveness on a nuclear battlefield.143 To further strengthen the U.S. position, Eisenhower also approved a JCS request for increased production of both very high yield thermonuclear weapons and small warheads suitable for air defense. In February 1956, when informed that air defense requirements might absorb a considerable portion of U.S. nuclear production in the future, the President responded that "nuclear production is now so vast that it may be possible to have everything that is needed."144 In early 1957, the first MB-I "Genie" two kiloton yield atomic air-to-air missiles were deployed, followed in 1958 by nuclear armed "Nike Hercules" surface-to-air missiles.145 In April 1956, Eisenhower apparently authorized the Air Defense Command to use these missiles immediately in case of surprise attack when it would be infeasible to wait for orders. This type of advance authorization had been proposed by the TCP report in 1955, and endorsed by NSC5602/1. The 1952 instructions and procedures for the "Interception and Engagement" of enemy aircraft were utilized to implement the President's directive.146 In endorsing the JCS production goals for high yield weapons, Eisenhower overrode strong objections within the JCS to the Air Force statement of weapons requirements. Increased production was justified through 1959, the President ruled on February 29, although he thought there should be a "levelling off " after that time. He saw merit, however, in the objections raised to Air Force plans for the use of such weapons, and urged the JCS to review their targeting plans to reduce application of thermo143. The record of JSCP/JSOP debates and their resolution may be found in CJCS (Military Strategy and Posture) and in CCS 381 (11-29-49) Sec. 30, JCS. The approved strategic concepts are in SM-423-56, 23 May 1956, m ibid., and SM-763-56, 19 Sept. 1956, CCS 381 U.S. (1-31-50) Sec. 66, JCS. "Pentomic" reorganization is recounted in A.J. Goodpaster, MCP, October 11, 1956, Diary-Staff Memos, October 1956, DDE Diary, Box 19, ACWF-EPP, DDEL. 144. A.J. Goodpaster, MCP, February 10, 1956. 145. MB-I Genie deployment is noted in Naval Message, Basegram, from Admiral H.D. Felt, 192309Z February 1957, CCS 471.6 (8-15-45) Sec. 90, JCS. Genie yield is based on information in Tom Compere, ed., The Air Force Blue Book, 1961, Vol. Il (New York: Military Publishing Institute, 1960), p. 341, and Announced U.S. Nuclear Tests, p. 6. Deployment of Nike Hercules with nuclear warheads is noted in letter, Neil McElroy to the President, January 12, 1958, Neil McElroy, Secretary of Defense 1959, Folder 4, Administration Series, Box 28, ACWF-EPP, DDEL; list, Los Alamos Stockpiled Nuclear Weapon Designs; and U.S. Army Antiaircraft Artillery and Guided Missile School, Fort Bliss, Texas, Joint Doctrine, Army Surface to Air Atomic Fires in Air Defense, January 8, 1957, File No. AAGMS, SAAF, JD, USAMHI. 146. The 1956 Authorization based on the 1952 instructions is noted in Arthur W. Radford, Memorandum for the Secretary of Defense, Subject: Department of Defense Report on Status of U.S. Military Programs as of 31 December 1956, March 20, 1957, CCS 381 U.S. (1-31-50) Sec. 68, JCS. "Interception and Engagement Instructions," n.d., but noted "approved Nov. 52," in Nuclear Weapons, Instructions for Use, June 1958 Folder, Administration Series, Box 31, ACWFEPP, DDEL. The April 1956 date is based on the titles of documents 16-22 on the Withdrawal Sheet in "AEC—Policy on Use of Atomic Weapons," Briefing Notes Subseries, NSC Series, Box 1, WHO-SANSA, DDEL.

Strategy and Nuclear Deterrence | 154

nuclear weapons "inappropriately against industrial targets of too little overall im­ portance," and to avoid overlapping "of large numbers of large yield weapons on specific targets." Bombs with yields "greater than two megatons should be used only for very special purposes," he warned, and targeting should be tailored not to cause "unnecessarily high population losses. "u? This directive, which reaffirmed Eisenhow­ er's preference for targeting military capability rather than the urban-industrial base, marked the beginning of a serious effort to contain escalating force and weapons requirements during the difficult years which lay ahead. Entering the Missile Age: ICBMs, Bomber Vulnerability, and Interservke Rivalry, 19571961

It was the dawning of the ballistic missile era, and in particular the projected deploy­ ment of Soviet intercontinental ballistic missiles (ICBMs) which finally forced Presi­ dent Eisenhower to reevaluate his scenario for general war. The Soviet missile threat posed a twofold strategic challenge. Because tactical warning of a surprise attack would be measured in minutes, it would vastly complicate the problem of getting SAC's bombers in the air before the attack arrived. Because ballistic missiles would be very difficult to locate and destroy, the problem of neutralizing Soviet nuclear capability would be greatly compounded. In coming to grips with these challenges, the Eisenhower Administration faced a series of crucial decisions between 1957 and 1961. These included how to deal with bomber force vulnerability, what types and numbers of ballistic missile systems should be authorized, how forces should be based and targeted, and whether they were intended primarily for preemptive or retaliatory attacks. During these four years, real alternatives to the planned SAC offensive emerged, but failed to take hold. Although Eisenhower repeatedly ex­ pressed a desire to place limits on the proposed offensive, he ultimately failed to do so. DEVELOPING BALLISTIC MISSILES

All three services had worked on missile development since World War II, although progress through 1950 was slowed by the Truman Administration's budget ceilings. Work on an ICBM was further delayed by doubts as to whether it would be possible to develop missiles capable of delivering the large and heavy weapons then in production, and, if so, whether they would have the accuracy or yield necessary to destroy specific targets. The Korean War rearmament program, with its two to four year time frame, facilitated development of Army and Navy air defense missiles and atomic armed tactical ballistic and cruise missiles, but gave little impetus to the ICBM. The successful test of a true thermonuclear weapon in October 1952, portending yields high enough to compensate for poor accuracy, and the warhead weight re-

147. A.J. Goodpaster, Memorandum for Admiral Arthur W. Radford, February 29, 1956, Joint Chiefs of Staff, January-April 1956, Folder 2, Subject Series, Defense Department, Box 4, WHOSS, DDEL.

Nuclear Strategy, 1945-1960 \ 155

ductions predicted by the AEC greatly improved prospects for the ICBM. In December 1952, an Air Force ad hoc committee recommended loosening accuracy requirements from 1,500 feet to one mile, and lowering warhead weight from 10,000 to 3,000 pounds in order to facilitate development. Because of the change in administrations, and Eisenhower's cuts in the Air Force budget, however, there was no immediate followup on these recommendations.148 In November 1953, during the New Look, Trevor Gardner, the Secretary of the Air Force's special assistant for research and development, set up the Strategic Missiles Evaluation Committee, which produced its report in February 1954. This report, and a concurrent RAND Corporation study, projected that an ICBM could be deployed by 1960 or soon after, if yield, payload, and accuracy requirements were broadened, additional funds were allocated, and the entire program was given high priority. Steps were subsequently taken to reorient the ongoing Atlas ICBM program, and to provide financing from Air Force funds adequate to ensure that "the extent to which it is accelerated [would] be limited only by technical progress."149 The report of the Technological Capabilities Panel (TCP), which had been formed as a result of a 1954 briefing by Trevor Gardner, gave crucial support to ballistic missile programs. The TCP emphasized the importance of maintaining the U.S. technological lead over the Soviet Union, and recommended accelerated development of both land-based ICBMs and land- and sea-based Intermediate Range Ballistic Missiles (IRBMs), all armed with megaton warheads. By the end of 1955, the Army, Navy, and Air Force had each organized an IRBM development program, and these, along with the Air Force ICBM effort, were accorded highest national priority.150 Despite Eisenhower's support for these programs, he retained serious reservations. In December 1956, the President told the JCS that he did "not think too much of the ballistic missiles as military weapons," and had only approved the acceleration rec­ ommended by the TCP because of the "psychological importance" of the new tech­ nology. He did not foresee the need for a large missile force once the nation proved

148. The best sources on ballistic missiles are Edmund Beard, Developing the ICBM: A Study in Bureaucratic Politics (New York: Columbia University Press, 1976); Ernest G. Schwiebert, A History of the U.S. Air Force Ballistic Missiles (New York: Frederick A. Praeger, 1965); Nathan F. Twining, Neither Liberty nor Safety (New York: Holt, Rinehart and Winston, 1966), pp. 299-309; Michael A. Armacost, The Politics of Weapons Innovation: The Thor-Jupiter Controversy (New York: Columbia University Press, 1969); Lemmer, The Air Force and Strategic Deterrence, pp. 11-19, 40-46; Max Rosenberg, The Air Force and the National Guided Missile Program, 1944-1950 (Washington, D.C.: USAF Historical Division Liaison Office, June 1964, declassified 1981); and Robert L. Perry, The Ballistic Missile Decisions, P3686 (Santa Monica, Calif.: Rand Corporation, October 1967). 149. Decision on AFC 12/8, Subject: USAF Strategic Missile Program, 6 May 1954, Air Force Council Decisions, Vol. I, Box 103, Twining Papers, LC; Beard, Developing the ICBM, pp. 142187; Schwiebert, USAF Ballistic Missiles, pp. 67-85, 217-219. 150. Technological Capabilities Panel Report, pp. 38-39, 63-65; James S. Lay, Jr., Memorandum for the National Security Council, Subject: Recommendations of the Report to the President by the Technological Capabilities Panel of the Science Advisory Committee, ODM, Continental Defense, Study of, by R. Sprague, 1955, Folder 8, NSC Series, Subject Subseries, Box 1, WHOSANSA, DDEL.

Strategy and Nuclear Deterrence \ 156

it could master the technology. Bombers would be the mainstay of SAC for ten years or more, Eisenhower predicted, especially since the U.S. already had large bomber forces, and "we must consider how to use them." Secretary of Defense Wilson proposed, and Eisenhower apparently concurred, that "150 well-targeted missiles might be enough."151 COPING WITH BOMBER VULNERABILITY

By 1956, the Air Force was beginning to address the problem of the vulnerability of bomber forces in the missile age, and in particular the need to adjust to drastically reduced tactical warning time. As early as 1951, the RAND Corporation had begun a series of studies for the Air Staff on "the effectiveness of our strike force under surprise attack." The first reports focused on the problem of protecting overseas bases, and resulted in the Air Force's decision to use such bases only for pre- and post-strike refueling.152 In 1956, echoing the findings of the TCP report, RAND ana­ lysts warned that bombers based in the United States might soon be in danger of air or even missile attack, and recommended improvements in early warning systems and SAC response time, dispersal of the bomber force, and improved airfield defen­ ses. In response to such concerns, SAC moved to construct or acquire additional bases in the U.S. and Canada, and conducted a series of tests from November 1956 through September 1957 to evaluate the human and material requirements of holding one-third of the bomber force on fifteen minute "ground alert."153 In the spring of 1957, responding to growing concern about national vulnerability, the Federal Civil Defense Administration proposed an ambitious $32 to $50 billion civil defense program. At the request of the NSC, the ODM Science Advisory Com­ mittee set up the Security Resources Panel to analyze the merits of the proposal. The chairman of the panel, H. Rowan Gaither, a long-time RAND trustee and adviser, subsequently requested and received permission to expand the scope of the study to include the entire question of national defense in the ballistic missile era.154 The final report of the Gaither Committee, "Deterrence and Survival in the Nuclear Age," was briefed to the President on November 4, and to the NSC three days later. 151. A.J. Goodpaster, MCP, December 19, 1956, dated December 20, 1956, December 1956 Diary—Staff Memos Folder, DDE Diary, Box 20, ACWF-EPP, DDEL. A similar comment may be found in A.J. Goodpaster, MCP, March 15,1956, dated March 16, 1956, Joint Chiefs of Staff, January-April 1956 Folder 2, Subject Series, Defense Department Subsenes, Box 4, WHO-SS, DDEL. 152. Decision on AFC 22/4a, Subject: Vulnerability of the USAF Strategic Striking Complex, 10 March 1953, Air Force Council Decisions, Vol. I, Box 103, Twining Papers, LC. See also A.J. Wohlstetter et al., Selection and Use of Strategic Air Bases, R-266 (Santa Monica,- Calif.: Rand Corporation, April 1954). 153. Staff Report, Protecting U.S. Power to Strike Back in the 1950s and 1960s, R-290 (Santa Monica, Calif.: Rand Corporation, Contract No. AF 18 [600]-1600, September 1, 1956); Development of Strategic Air Command, pp. 58-60; Strategic Air Command Progress Analysis, 1948-1956, Fore­ word, pp. 117-135; Decision on AFC 20/7, 6 March 1956, Subject: Strategic Plan of Operations, Air Force Council Decisions, Vol. I, Box 103, Twining Papers, LC. 154. Robert Cutler, No Time For Rest (Boston: Atlantic-Little Brown and Co., 1956), pp. 347-361; Prados, The Soviet Estimate, pp. 51-75.

Nuclear Strategy, 1945-1960 \ 157

It expressed only moderate support for fallout shelters, and instead assigned highest priority to improving SACs survivability. The only way to protect the American people, the report concluded, would be to deter war through the maintenance of a secure and effective retaliatory striking force. It warned that "by 1959 the U.S.S.R. may be able to launch an attack with ICBMs carrying megaton warheads against which SAC will be almost completely vulnerable under present programs," and urged that steps be taken to reduce SAC reaction time, improve tactical warning, increase bomber dispersion, and improve active and passive airfield defenses. It also recom­ mended further acceleration of U.S. ballistic missile programs, with particular em­ phasis on the Polaris submarine-launched IRBM because of its "advantages of mobility and greatly reduced vulnerability."155 Three of the committee members, the President was told privately, had advocated reconsideration of the option of preventive war as a means of avoiding a situation of extreme jeopardy.156 Eisenhower responded to the Gaither Committee briefing cautiously. It was his belief, he stated, that the threat was being considerably overestimated, and that Soviet ballistic missile technology was not progressing as fast as projected. For at least the next five years, "aircraft would still be the primary means of carrying out destruction, and during this period we have the capability of delivering the greater blow." He agreed that ballistic missile development programs should be further accelerated, and that SAC response time should be reduced, since "there is in reality no defense except to retaliate."157 SAC must realize, he stated after the briefing, that "we must not allow the enemy to strike the first blow."158 This aspect of the problem was underlined on November 7 when Eisenhower was briefed by the chairman of the Gaither steering committee, R.C. Sprague, who had conducted his own highly secret investigation of SAC readiness. Sprague reported that SAC had not yet implemented a ground alert, and on a particular randomly selected day "not a single plane could have left the ground within six hours except for a few that were by chance in the air on a test at the time." In an actual surprise attack, the U.S. could probably count on getting only 50 to 150 large weapons off the ground, and these would be totally inadequate to achieve a "substantial retaliatory attack."159

155. A.J. Goodpaster, MCP, November 4, 1957, dated November 6, 195, Science Advisory Committee, November 1957-April 1958, Folder 3, Subject Series, Alphabetical Subseries, Box 23, WHO-SS, DDEL; NSC 5724, "Deterrence and Survival in the Nuclear Age," Report to the President by the Security Resources Panel of the Science Advisory Committee, 7 November 1957, in Gaither Report, November 1957-January 1958, Folder 2, ibid., Box 13. 156. Appointments, November 9, 1957, November 1957 Folder 2, ACW Diary, Box 9, ACWFEPP, DDEL. 157. A.J. Goodpaster, MCP, November 4, 1957. 158. Appointments, November 9, 1957. 159. A.J. Goodpaster, MCP, November 7, 1957, following NSC Meeting, Military Planning, 1958-1961, Folder 3, Subject Series, Defense Department Subseries, Box 6, WHO-SS, DDEL; Robert Cutler, Memorandum for the President, Subject: SAC Concentration in U.S. and Reaction Time, October 25, 1957, Robert Cutler 1956-7, Folder 1, Administration Series, Box 11, ACWFEPP, DDEL.

Strategy and Nuclear Deterrence | 158

Despite its political volatility, the Gaither report was more a symptom than a cause of change. Eisenhower believed that the U.S. was facing a gradual deterioration of strategic advantage, as forecast by the TCP report, but not an immediate crisis. "Until an enemy [has] enough operational capability to destroy most of our bases simulta­ neously and thus prevent retaliation by us," he stated in February 1958, "our deterrent remains effective."160 Recalling the illusory "bomber gap" dissipated by U-2 recon­ naissance, the President was skeptical of the 1958 National Intelligence Estimates which projected that there might be 100 Soviet ICBMs by mid-1959 or 1960, and 500 by 1963, and subsequent NIEs suggested that his doubts had been justified. Unfor­ tunately, the "missile gap" would become a potent political issue exploited by Dem­ ocratic presidential hopefuls, despite Eisenhower's efforts to dispel it, and would not be shown to be illusion until the first !Discoverer and SAMOS satellites documented the miniscule size of Soviet missile capability in 1960-1961.161 Eisenhower's approach to protecting U.S. bombers in the missile age was not to increase their survivability on the ground, but to ensure that they would be in the air by the time any surprise attack arrived. The program of bomber blast shelters recommended by the Gaither Committee was not implemented, and from 1957 on, airfield defense programs such as air defense missiles, interceptors, and survivable command and control systems were cut back. On the other hand, a concerted effort was made to increase tactical warning time and rapid response. In January 1958, Defense Secretary NeU McElroy approved development of the Ballistic Missile Early Warning System (BMEWS) to complement the Distant Early Warning Line activated in August 1957. Construction was soon begun on BMEWS stations in England, Greenland, and Canada. In 1958, development work was begun as well on the "Midas" early warning satellite.162 To further reduce reaction time, on May 22, 1957, Eisenhower issued an "authori­ zation for the expenditure of nuclear weapons," still classified, which apparently went beyond the air defense authorization of the previous year, and gave permission for nuclear response under a variety of emergency conditions. The very complex instructions needed to implement this authorization required extensive consideration by the Departments of Defense and State, and were not approved until February 17, 1959. Other Emergency Action Documents, including possibly a secret executive order covering the transfer of nuclear weapons from the AEC to the military and their 160. A.J. Goodpaster, MCP, February 4, 1958, dated February 6, 1958, Missiles and Satellites, Vol. II, January-February 1958, Folder II, Subject Series, Defense Department Subseries, WHOSS, DDEL. 161. Anticipated Soviet ICBM deployments are laid out in JCS 1899/523, 20 October 1959, CCS 3340, Strategic Air and ICBM Operations, 10 September 1959, JCS; Prados, The Soviet Estimate, pp. 75-126. 162. Lemmer, The Air Force and Strategic Deterrence, pp. 45-51; Office of the Director, Defense Research and Engineering, Report No. 10, Military Space Projects, March-April, May 1960, Folder, same title, Subject Series, Defense Department Subseries, Box 9, WHO-SS, DDEL. JCS discussions of the Gaither reporf s implications for continental defense are in the JCS 2101 series, CCS 381 U.S. (1-31-50) Sees. 74-79, while those pertaining to SAC vulnerability are in the JCS 1899 series, CCS 381 U.S. (5-23-46) Sees. 94-98, both JCS.

Nuclear Strategy, 1945-1960 \ 159

subsequent utilization, were also promulgated during this time. Finally, in 1960, a series of still classified directives regarding the use of nuclear weapons was issued to each of the unified and specified commanders.163 Significant progress was also made in reducing SAC response time. In 1959, the goal of holding one-third of the bomber force on ground alert was achieved, by adding more bomber crews, restructuring maintenance schedules, and redesigning runway approach ramps.164 On March 1, 1958, the Chief of Staff of the Air Force authorized SAC to implement the "fail safe" or "positive control" system, which would permit rapid response without the risk of initiating war through miscalculation. Bombers would take off at the first sign of a Soviet strike, and proceed along Emer­ gency War Plan routes toward their targets, but their missions would be aborted at a pre-specified point unless they received the "go" code to continue. SAC also pushed ahead with its plans to establish an airborne alert consisting of one-fourth of the B52 force. In 1959, the JCS approved the airborne alert in principle, but because of the considerable cost of continuous operations, recommended that it be put into operation only when the President considered such special precautions necessary.165 MORE CAPABILITY, MORE TARGETS

The development of SAC alert plans was greatly facilitated by advances in weapons design. Through the mid-1950s, the fission cores used to fuel atomic weapons and to initiate the fusion reaction in "two stage" thermonuclear bombs could be inserted only immediately before takeoff, or in flight. This problem was solved by the perfec­ tion of a new "sealed pit" advanced bomb design which, among other things, con­ tained smaller amounts of high explosives and fissionable material. The sealed pit 163. The documentary evidence on advance authorization is fragmentary. This statement is based on B.L. Austin, Memorandum for Chairman, JCS, DM-93-58, 27 March 1958, Subject: Status of the "Instructions for the Expenditure of Nuclear Weapons under Special Circum­ stances," CCS 471.6 (8-15-45) Sec. Ill, JCS; Memoranda, James S. Lay, Jr. to the Secretaries of State and Defense, Subject: Policy Regarding Use of Atomic Weapons, March 19, April 2, April 9, and April 23, 1958; Memorandum, Gordon Gray to General Goodpaster, December 15, 1958, discussing proposed Executive Order; undated page containing chronology of developments regarding Instructions for the Expenditure of Nuclear Weapons from February 17, 1959 through May 18, 1960; and letter, Thomas S. Gates to the President, June 5, 1959, all in AEC-Policy on Use of Atomic Weapons Folder, Box 9, WHO, Project Cleanup, DDEL; withdrawal sheet of documents in ibid.; and withdrawal sheets of documents in Atomic Weapons, Presidential Approval and Instructions for Use Of, Folders 1-5; and Atomic Weapons, Correspondence and Background for Presidential Approval, Policy re Use, Folders 2-6, both in NSC Series, Subject Subseries, Box 1, WHO-SANSA, DDEL; and Instructions to Commanders, Folders 1-5, Subject Series, Alphabetical Subseries, Box 14, WHO-SS, DDEL. 164. Development of Strategic Air Command, pp. 67-77; Special Report on SAC in Aviation Week and Space Technology, June 20, 1960, pp. 101-144; WSEG Staff Study No. 77, The Feasibility, Cost and Effectiveness of Dual Runways at SAC Bomber Bases, 20 July 1959, CCS 4960, Air Bases, 27 July 1959, JCS. 165. Memo, Chief of Staff, USAF to the JCS, on Launching of the Strategic Air Command Alert Force, CSAFM-72-58, 10 March 1958, enclosed in JCS 1899/398, 10 March 1958, in CCS 381 U.S. (5-23-46) Sec. 95, JCS. JCS deliberations on Airborne Alert are in CCS 3340, Strategic Air and ICBM Operations, 10 September 1959, JCS, and Futrell, Ideas, Concepts, Doctrine, pp. 578-582.

Strategy and Nuclear Deterrence \ 160

weapons were efficient, lightweight, high yield thermonuclear bombs, equipped with safety devices which reduced the risk of inadvertent detonation to near zero. They could be stored fully assembled in alert bombers, or atop ballistic missiles, for long periods without substantial maintenance, permitting almost instantaneous response. By 1959, the majority of weapons assigned to SAC were of sealed pit design.166 SAC delivery capability also continued to expand. At its peak in 1959, SAC had nearly 500 B-52s, more than 2,500 B-47s, and over 1,000 propeller and jet driven tanker aircraft. Development was underway on the supersonic B-58 which would enter service in August 1960, and on the proposed follow-on to the B-52, the Mach 3 B-70 "Valkyrie." As recommended by the TCP report, the Air Force was also pressing for development of a nuclear powered bomber, which would have unlimited range and would greatly facilitate maintenance of an airborne alert.167 Growth of the bomber force was still closely linked to expanding target lists. The SAC target list totalled 2,997 in 1956, and 3,261 in early 1957. By the end of the decade, SAC had analyzed over 20,000 potential Soviet bloc targets, grouping the most important of these into Designated Ground Zeros consistent with the blast radii of large yield thermonuclear weapons.168 Expansion of the list was largely attributable to identification of additional "counterforce" targets, in particular airfields and sus­ pected missile sites, and to the poor quality of target intelligence through the 1950s, which encouraged creative guesswork. The most sophisticated intelligence gathering system, the U-2 program, initiated in 1956 in response to the TCP recommendations, apparently involved only twenty or thirty deep penetration overflights of the Soviet Union, mostly between 1956 and 1958. Because the program was designated as a national reconnaissance and warning system under CIA control, moreover, the Air Force was unable to dictate the timing or routing of these flights to maximize their usefulness for target identification or verification, although target planners made the most of the information it provided.169 THE ARMY AND NAVY FIGHT SAC DOMINANCE OF NUCLEAR POLICY

Continued expansion of SAC's proposed strategic offensive, and increased concern about the vulnerability of land-based bombers, generated a concerted effort within the JCS to break the bootstrapping dynamic. In the spring of 1957, possibly following up on the President's 1956 directive, Army Chief of Staff Taylor and Chief