Thinking About Nuclear Weapons: Principles, Problems, Prospects 0199563942, 9780199563944, 9780191569791

Citation preview

Thinking About Nuclear Weapons

The Changing Character of War Programme is an inter-disciplinary research group located at the University of Oxford, and funded by the Leverhulme Trust.

Thinking About Nuclear Weapons Principles, Problems, Prospects

Michael Quinlan

1

3 Great Clarendon Street, Oxford ox2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Michael Quinlan 2009 The moral rights of the author have been asserted Database right Oxford University Press (maker) First published 2009 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Quinlan, Michael. Thinking about nuclear weapons : principles, problems, prospects / Michael Quinlan. p. cm. Includes index. ISBN 978-0-19-956394-4 1. Nuclear weapons. 2. Nuclear arms control. I. Title. U264.Q54 2009 355.02’17–dc22 2008046119 Typeset by SPI Publisher Services, Pondicherry, India Printed in the UK by the MPG Books Group ISBN 978–0–19–956394–4 1 3 5 7 9 10 8 6 4 2

To the victims at Hiroshima and Nagasaki whose terrible experience sent to the whole world and to future generations a dark message of warning that may well have saved many millions of lives

This page intentionally left blank

Contents

Foreword Preface Why this Book? My Background Disclaimers, Apologies, Acknowledgements Abbreviations

Introduction Structure of the Book

x xii xii xiv xvii xx

1 1

Part I. The Significance of Nuclear Weapons 1. The Nuclear Revolution The Shock of 1945 Interpreting the Revolution

5 5 8

2. The Tools of Thinking Evidence Learning Language

13 13 14 15

3. Deterrence The Concept Criticisms and Fallacies After the Cold War

20 20 25 30

4. Nuclear Deterrence in NATO Doctrine and Planning Armouries

33 33 40

5. The Ethics of Nuclear Weapons The Moral Problem The Inescapable Choice

46 46 50

vii

Contents

Part II. Managing Nuclear Weapons 6. Risks Instability Escalation Accident and Miscalculation Terrorism

59 59 62 67 71

7. Proliferation Why Proliferation Matters The Non-proliferation Regime The Treaty Bargains The Future of the Regime

76 76 78 81 83

8. Arms Racing, Costs, and Arms Control Arms Races and Costs Arms Control

88 88 90

9. Easements and Escape Routes No First Use Alternative Defence Minimal Armouries Virtual Armouries Defence Against Ballistic Missiles

99 99 104 105 108 109

Part III. National Nuclear-Weapon Postures and Policies: Britain, India, Pakistan 10. United Kingdom Doctrine and Policy History The Concept of Independence The Scale and Use of Capability After the Cold War Future Policy

115 115 119 124 126 127

11. Nuclear Weapons in South Asia History Doctrines Capabilities Stability and Security Looking Forward Non-proliferation

133 133 136 138 139 141 147

viii

Contents

Part IV. The Path Ahead 12. The Abolition of Nuclear Armouries? The Abolitionist Goal Paths and Conditions Political Change Disarmament Mechanisms The Case for Study

153 153 156 158 161 164

13. The Practical Agenda The Non-proliferation Regime Arms Control and Disarmament by the Nuclear-Weapon States Doctrines and Postures Resolving Political Disputes

166 167

Appendix 1: Nuclear Weapons and Preventing War Appendix 2: The Strategic Use of Nuclear Weapons Index

181 184 190

171 178 180

ix

Foreword

In the preface to this book, Michael Quinlan does me the honour of quoting a remark I made in a recent lecture to the effect that ‘the nuclear dragon is not dead, but sleeping’. This if true—and it is hard to believe otherwise—would be sufficient justification for his writing this book, and good reason why we should all read it. For Quinlan is uniquely qualified to enlighten us about the dragon. Not only has he spent a lifetime studying the problem of nuclear weapons, but as a senior civil servant he has played a leading role in formulating British policy for their acquisition, possession, and use. He has, to put it succinctly, taught our masters how to think. Unfriendly critics may therefore be tempted to dismiss him as a mere apologist for government policy. They had better read this book. They will find themselves engaged with a protagonist who possesses not only a brilliant mind but also a sense of profound moral responsibility. Quinlan faces head on the inescapable moral dilemmas confronting anyone who, as he puts it, cannot content himself with ‘hand-wringing’ but has to take decisions and act on them. Such decisions cannot be taken without a clear moral compass, a sense of practical possibilities, and a capacity for hard analytic thought. All these qualities Quinlan possesses in abundance. He explains, sharply but courteously, why he is sceptical about such nostrums as no first use, massive arms reductions, and the development of anti-ballistic missiles. He is deeply concerned about the problem of nuclear proliferation; and while understandably sceptical about the total abolition of nuclear weapons before all major sources of international conflict have been resolved, he nevertheless believes that we should think seriously about the manner in which this highly desirable goal might be approached. Quinlan warns the reader not to expect any interesting revelations about the people he has encountered or the politics, domestic or

x

Foreword international, in which he was involved during his four decades as a policy-maker; but he tells us enough, about, for example, the problems that NATO faced in formulating its deterrence posture, to make his book much more than a discussion of abstract ideas. Historians will certainly find it valuable; but for everyone concerned with taming the nuclear dragon, the guidance it provides is essential. Michael Howard Sir Michael Howard, OM CH CBE MC, is co-founder and President Emeritus of the International Institute for Strategic Studies; formely Regius Professor of Modern History, Oxford University; and Professor of Military and Naval History, Yale University.

xi

Preface

Why this Book? The existence of nuclear weapons, and humanity’s ineradicable knowledge of them, poses demanding challenges. Working on these took up a large part of my time at several stages of my career as a United Kingdom civil servant in the second half of the twentieth century. In the course of this I built up progressively an informal set of concepts—ways of thinking—for tackling the issues they raised, and I believed (as I still do) that this set eventually amounted to a usefully robust and coherent structure of analysis. I used it in my contributions both to formulating policy and to explaining and defending it, whether at first hand or through briefing ministers and others. In addition to using this structure to meet the direct demands of official work, from the beginning of the 1980s onwards I drew upon it in other contexts—in writing or speaking in lectures or seminars, in interchange with others interested, for clarifying ideas to myself, and occasionally in published journals. Until my retirement from government service in 1992, however, publication was limited by the constraints which in the United Kingdom system rightly bear upon serving non-political officials. It did however become known—rather unusually, and not by my choice—that certain governmental statements, with no attributed authorship, had been written by me. (One of them is reproduced in Appendix 1.) My first attempt to organize my ideas into in a general presentation was made in 1997, when the Royal United Services Institute for Defence Studies published a monograph 1 which I had composed under the same primary title as the present book. I am indebted to the Institute for its support and continued cooperation.

1

xii

‘Thinking About Nuclear Weapons’, RUSI Whitehall Paper Series 1997, No. 21.

Preface I wrote then that nuclear-weapon issues had not the same urgent force as in earlier years. That view reflected not only the ending of the cold war between East and West but also the fact that the international regime of restraint centred upon the 1968 Nuclear NonProliferation Treaty had recently had several notable successes and no major setbacks. The subsequent decade, however, has proved in a number of respects more sombre. In the globalizing world of the twenty-first century the practical attention of most governments has undoubtedly moved to a different agenda, and few issues about nuclear weapons feature much in public discourse—to most people the subject probably seems yesterday’s concern. But though it has not regained the special salience it had during the cold war, it has not yet receded into safe irrelevance to the degree that optimists might once have hoped for—it may be that, as Sir Michael Howard has suggested, the dragon is not dead but sleeping. I have therefore thought it useful to update and expand my 1997 account. That account was rooted in experience of the cold war. We know now that nuclear war never happened in that setting. But the possibility of it always had to be taken most seriously, even if at any particular time it was thought unlikely. (That last was consistently my own sense. Perhaps through a failure of imagination, I never felt deeply oppressed by the looming shadow, though for several years I used to keep in close view beside my desk in the Ministry of Defence a picture of post-attack Hiroshima to remind me of the underlying reality.) The body of thinking built up across the world in those years was very extensive, and absorbed huge effort. Many of the issues no longer have the same form as they had in those days. Some of the elaborate analyses that were developed related to scenarios which now look far-fetched in a global security scene no longer dominated by a mistrustful relationship between two massively armed superpowers of opposed ideologies. The fundamental challenges, however, arise irreversibly out of the nuclear discovery, not just that particular confrontation. Nuclear possibilities can never vanish entirely, even if the physical abolition of the armouries now existing were to become a more realistic and imminent prospect than it currently seems. (I explore this in Chapter 12.) Questions about the basic understanding built up during the cold war by the Western defence community—much less is known about the Eastern side—and whether it was or was not sound therefore remain relevant to shaping policy in a still-anarchic world which may yet

xiii

Preface yield unpleasant surprises upon which nuclear weapons might bear. Tackling such questions ought moreover to start, as a good deal of post-cold-war commentary does not, from an accurate grasp of what that earlier understanding was. The matters addressed in this book have attracted a vast literature, much of it showing detailed and sophisticated scholarly investigation which I am not attempting to rival. Almost every particular topic has been examined at greater length elsewhere, and there are numerous issues related to nuclear weapons with which I never had occasion to deal, either in or out of government, and which I do not now seek to tackle at all. The basis for this book is experience, reflection, and debate, not academic research. I believe however that there is value in an attempt, on such a basis, to delineate a coherent general outlook upon these weapons, and to set at least a substantial set of the practical issues within a consistent perspective. I hope therefore that what I have written may be of interest both to historians—if only as a depiction of what was in the mind of a typical participant in policy formation during the second half of the cold war—and to students of international security. I should like also, if I can, to help ease the task of generations later than my own in positions of public responsibility for nuclear-weapon issues. They may have to face difficult decision-making about such issues from a basis of relevant experience and discussion less extensive and continuous than my generation had occasion to accumulate. They have had to tackle other pressing defence preoccupations, like the long-running conflicts in Afghanistan and Iraq, of a kind that my contemporaries and I were largely spared by the rigidities of the coldwar world. But there has clearly been, on both sides of the Atlantic, much less attention to issues of nuclear policy than there was in the cold war, and even a good deal of forgetting of the thinking that was developed and the lessons that were learned. There are risks in that.

My Background This book is neither a work of historical record nor a reminiscent memoir of my own engagement in the nuclear-weapon field. In accordance with customary expectation in Britain’s Home Civil Service, I did not at any stage keep a private diary about my official

xiv

Preface activities. Readers may however find it useful to know what basis of relevant experience underlies the analyses and opinions I have offered. My main career was spent almost entirely within the span of the cold war, and for most of it I worked in the defence field. I was first assigned in 1954 as an administrative generalist to the Air Ministry, then a separate department of state dealing with the business of the Royal Air Force. I was concerned from the outset with some practical aspects of building up and maintaining the United Kingdom’s V-bomber nuclear-delivery force, and then beginning to consider how it should be developed and eventually replaced. I was however not prompted to think hard about issues like those in this book until 1960–1, when I was secretary to a group of distinguished external scientists invited by the Air Ministry to bring independent minds to bear on various topics in security policy. One of these topics was what concepts ought to shape the armoury and doctrine of the North Atlantic Treaty Organisation below what was termed the ‘strategic’ level. From early 1962 to late 1965 I was Private Secretary to successive holders of the post of Chief of the Air Staff, the professional head of the Royal Air Force. That brought me into considerable contact with aspects of Britain’s nuclear role, then still almost entirely in the hands of the Royal Air Force. In 1968–70, as a member of the newly created Defence Policy Staff, I had some collegial involvement in that staff’s input to UK work on NATO nuclear policy while the Alliance’s Nuclear Planning Group was framing its ‘Provisional Political Guidelines for the Initial Defensive Tactical Use of Nuclear Weapons’. I had lead responsibility within the Defence Policy Staff on arms-control questions during the formative period of US/Soviet Strategic Arms Limitation Talks, negotiations between NATO and the Warsaw Pact on ‘mutual and balanced force reductions’, and other arms-control projects. I remained in touch, albeit less closely, with all these matters from 1970 to 1973 as Defence Counsellor in the UK Delegation to NATO in Brussels. After a spell away from defence business I returned to the Ministry of Defence in the spring of 1977 to take up the post nowadays called Policy Director. This was the main focal point within the Ministry for work on NATO and national nuclear-weapon policy. NATO was heavily engaged in continuing issues about the Alliance’s collective nuclear doctrine, nuclear arms control, the ‘neutron bomb’

xv

Preface episode in 1977–8, 2 and then the modernization of intermediaterange nuclear forces. I led the UK team in the NATO High-Level Group working on that last matter. The United Kingdom itself in this period was tackling the problems of updating and then replacing its nuclear-deterrent force equipped with the Polaris submarinelaunched ballistic missile. From mid-1981 until early 1988 I was again out of the defence field, but I remained aware of nuclear-weapon issues and occasionally involved in debate about them. This was partly from general interest but also for a more particular reason. From the later 1970s onwards there had been a resurgence in Britain, as elsewhere in Western Europe, of public campaigning against nuclear weapons. One strand in this was a contention that the possession of such weapons, or much of the apparent planning for their possible use, was incompatible with moral imperatives, especially those accepted by Christians. As a Roman Catholic known to have been involved at senior level with nuclear-weapons policy I was drawn by challenge into this debate, both in my own church and in other parts of the Christian community. That is why this book contains what may seem unexpectedly extensive treatment (in Chapter 5 and Appendix 2) of ethical aspects. I returned to the Ministry of Defence in 1988 as Permanent UnderSecretary of State, the senior non-political civilian appointment. I had then no close responsibilities for nuclear-weapon policy, my remit being broader, but until my retirement in 1992 I continued— occasionally, I suspect, straining the patience of my colleagues—to take a particular interest in it. Since retiring from government employment I have maintained (partly through the privilege of being Director of the Ditchley Foundation from 1992 to 1999, and thus engaged in its wide-ranging international conference programme) a fairly high albeit unsystematic level of involvement in the flow of academic and similar activity on nuclear-weapon issues, not only in the United Kingdom but in Continental Europe, the United States, and farther afield. Since 1998 I have taken a special interest in their bearing upon South Asia, latterly as a Consulting Senior Fellow at the London-based International Institute for Strategic Studies (IISS).

2

xvi

This is explained on pages 43–4.

Preface

Disclaimers, Apologies, Acknowledgements This book is not directly concerned with political aspects, whether domestic or international, outside the security calculus. As I have explained, it does not attempt to chronicle the episodes in which I was involved, to comment on every nuclear issue that crossed my official desk, or to illuminate the inner workings of government. I fear that anyone hoping for the inside story on this or that episode, or entertaining personal vignettes, or psychological insight into what it felt like to be involved in making nuclear-weapon policy, will be disappointed. For the most part, this book mentions particular events only where doing so seems to me to help in explaining ideas. Some of those who may be drawn to read this book will quickly perceive my disagreement with their known views. I have tried, however, to convey that in the currency of ideas, not of personalities or polemic. In this field—already difficult enough in itself— debate is sometimes disfigured by assault on the motives or attitudes of opponents. Where issues matter so much feelings may naturally run high, and as in any complex human enterprise involving many people there is no doubt to be found on almost every side of the diagnoses and debates a diverse and uneven range of motives. But motives, whether good or bad, and however confidently imputed, are not a sure index of an argument’s objective merit, and I have tried to concern myself with that. I assure those whose interest may be primarily historical that though drawing together the material has entailed a good deal of carpentry in drafting and updating in expression, I have not modified my basic concepts retrospectively. It will be for readers to judge whether my continuing contentment with them shows initial wisdom or subsequent obstinacy. By much the same token, I apologize to those who recognize in these pages a good many phrases or passages already familiar from my sporadic earlier writings. This book is scantily equipped with evidentiary footnotes or scholarly indications of where its ideas and factual information came from. I built up my structure of thinking pragmatically, as a working apparatus, and mostly by reflection and debate rather than by reading and organized study with careful note taken of sources. I suspect indeed that academics might have been scandalized by how limited my direct acquaintance was, during most of my official career, with the writings of eminent scholars. I do not deceive myself, however,

xvii

Preface that my synthesis is full of unique or original insights. I know— even where I cannot now pin down chapter and verse, or occasion— that I stand on the shoulders of intellectual giants from Bernard Brodie onwards. I readily acknowledge a wide-ranging debt to all who have contributed, whether in official or in academic circles, to the storehouse of thinking from which I have directly or indirectly drawn my stock. The choice of components is however my own, and so is the fault if it be poorly made. That said, I offer posthumous tribute to four exceptional people who influenced me diversely in this field. Neil Cameron 3 was perhaps not himself in the first flight of defence analysts, but his energetic and forward-looking enthusiasm for fresh ideas was always invigorating. Frank Cooper’s 4 legendary shrewdness and practicality repeatedly kept my feet on the ground. Sydney Bailey’s 5 resolve to understand fully and fairly viewpoints which, as a distinguished Quaker with a long record in non-military peace-seeking, he could not share was a powerful stimulus to my concern with the moral dimension. And to find someone of Leonard Cheshire’s 6 remarkable personal history, integrity, openness, and Christian charity on basically the same side as myself on difficult ethical questions served as both comfort and discipline. I offer warm thanks to Andrew Brookes, Mark Fitzpatrick, Rahul Roy-Chaudhury, Hilary Synnott, and Rachel Yemini at the IISS, to Jack Gill of the Near East–South Asia Center for Strategic Studies, and to my long-term Ministry of Defence friends and colleagues Peter Hudson, Michael Legge, Richard Mottram, and Moray Stewart, for their ready generosity in giving help and advice. Mistakes and misjudgments, however, remain entirely my own. I am grateful also for the cooperation of editors and publishers in my use of material similar to that in some of my earlier writings: For Chapter 10, the chapter commissioned from me by the Nonproliferation Policy Education Center in Getting MAD: Nuclear Mutual 3 Marshal of the Royal Air Force Lord Cameron of Balhousie, GCB, DSO, DFC: Chief of the Defence Staff 1977–9, Principal of King’s College, London, 1980–4. 4 The Rt. Hon. Sir Frank Cooper, GCB: Permanent Under-Secretary of State, Northern Ireland Office 1973–6, Ministry of Defence 1976–82. 5 See Explorations in Ethics and International Relations: Essays in Honour of Sydney D. Bailey, London, Croom Helm, 1981. 6 Group Captain Lord Cheshire of Woodhall Spa, VC, OM, DSO, DFC: World War II bomber pilot of unique distinction, observer at the nuclear strike on Nagasaki in August 1945, founder of the worldwide Cheshire Homes movement for the care of the disabled.

xviii

Preface Assured Destruction, Its Origins and Practice, ed. Henry D. Sokolski, US Army War College Strategic Studies Institute, Carlisle, Pennsylvania, 2004, pp. 261–74. For Chapter 12, my article ‘Abolishing Nuclear Weapons: Policy or Pipe-dream?’ in Survival, Vol. 49, No. 4, Winter 2007–8, pp. 7–15. For Chapter 13, my chapter in US–UK Nuclear Cooperation After 50 Years, ed. Jenifer Mackby and Paul Cornish, CSIS Press, Washington, DC, 2008, pp. 160–71. I owe very special thanks to my wife, Mary, for her long and patient tolerance of unsociable preoccupation with these matters. Michael Quinlan October 2008

xix

Abbreviations

ABM

anti-ballistic missile

BMD

ballistic-missile defence

BW

biological weapons

CBM

confidence-building measure

CTBT

Comprehensive Test Ban Treaty

CTRI

Cooperative Threat Reduction Initiative (‘Nunn-Lugar’ US–Russia collaborative programme)

CW

chemical weapons

ERW

enhanced-radiation warhead

EU

European Union

F/GA

fighter/ground attack aircraft

FMCT

Fissile Material Cut-off Treaty

FRG

Federal Republic of Germany

GLCM

ground-launched cruise missile

HEU

highly enriched uranium

IAEA

International Atomic Energy Agency

ICBM

inter-continental ballistic missile

ICJ

International Court of Justice

IISS

International Institute for Strategic Studies

INF

intermediate-range nuclear forces

IRBM

intermediate-range ballistic missile (maximum range typically 2,500–5,500km)

LEU

low-enriched uranium

MAD

mutually assured destruction

MBFR

mutual and balanced force reductions

MDA

1958 Mutual Defence Agreement (UK/US)

MIRV

multiple independently-targetable re-entry vehicle

xx

Abbreviations MLF

multilateral force (NATO project)

MRBM

medium-range ballistic missile (maximum range typically 500–2,500km)

MRV

multiple re-entry vehicle (not independently targetable)

MTCR

Missile Technology Control Regime

NATO

North Atlantic Treaty Organisation

NFU

no first use (usually of nuclear weapons)

NNWS

non-nuclear-weapon-possessing state

NPG

Nuclear Planning Group (NATO)

NPT

1968 Nuclear Non-Proliferation Treaty

NSA

negative security assurance

NSG

Nuclear Suppliers Group

NWFZ

nuclear-weapon-free zone

NWS

nuclear-weapon-possessing state

PAL

permissive action link

PTBT

1963 Partial Test Ban Treaty

RRW

Reliable Replacement Warhead

RUSI

Royal United Services Institute for Defence and Security Studies

SALT

Strategic Arms Limitation Talks (USA/USSR)

SLBM

submarine-launched ballistic missile

START

Strategic Arms Reduction Talks/Treaty/Treaties (USA/Russia)

UK

United Kingdom of Great Britain and Northern Ireland

UN

United Nations

UNSCR

United Nations Security Council Resolution

USA

United States of America

WMD

weapons of mass destruction

xxi

This page intentionally left blank

Introduction

Structure of the Book The book has four main sections. Part I—Chapters 1–5—is concerned with the basic significance of nuclear weapons. Part II—Chapters 6–9—addresses particular issues in managing their existence. Part III—Chapters 10–11—looks at some national examples to illustrate how policies can take shape in practice, and the questions they may raise. The two instances chosen are those I know best—the United Kingdom, where I mostly worked, and South Asia, in which during the past decade I have taken a particular interest and have enjoyed generous access to senior figures in and out of government. Part IV—Chapters 12–13—discusses the worldwide future of nuclear weapons. In Part I, Chapter 1 considers the fundamental revolution which nuclear weapons have brought about. Chapter 2 warns of some of the limitations and potential traps in ways of thinking and speaking about them. Chapter 3 discusses the concept of deterrence, which has been so salient in the nuclear age. Chapter 4 reviews, as a notable exemplar of practical policy development, how NATO’s thinking about nuclear weapons took shape. Chapter 5 wrestles with the moral problems they raise. In Part II, Chapter 6 surveys a range of risks such as instability, escalation, accident, and terrorism. Chapter 7 discusses proliferation, and Chapter 8 considers what nuclear arms control can usefully contribute. Chapter 9 examines various other ways in which there have been attempts or proposals to ease nuclear-weapon risks or to escape from them.

1

Introduction In Part III, Chapter 10 reviews Britain’s nuclear-weapon history, thinking, current posture, and future plans. Chapter 11 looks at India and Pakistan. In Part IV, Chapter 12 explores the idea of abolishing all nuclear armouries. Finally, Chapter 13 makes suggestions for the practical agenda of national and international action about them in the short and medium term.

2

Part I The Significance of Nuclear Weapons

This page intentionally left blank

1 The Nuclear Revolution

The Shock of 1945 From the earliest times recorded in history successive advances in innovation have been heightening the destructive power that could be brought to bear in armed conflict. The sling, the stirrup, the crossbow, gunpowder, the machine-gun, and a wide range of other devices illustrate that. The nineteenth and twentieth centuries saw increasing depth of scientific understanding and technological ingenuity applied to carry forward the pace of advance, most notably in the mastery of flight. Even before that dramatic development, thinkers had been speculating about how far the process might eventually reach. In 1862 the American savant Henry Adams wrote gloomily 1 : Some day science shall have the existence of mankind in its power, and the human race commit suicide by blowing up the planet.

Soon afterwards another American, the novelist Wilkie Collins, writing in 1870 in the aftermath of his country’s murderous civil conflict and at the time of the Franco-Prussian War, took matters a little further and, in a sense, more operationally 2 : I begin to believe in only one civilising influence—the discovery, one of these days, of a destructive agent so terrible that War shall mean annihilation, and men’s fears shall force them to keep the peace. 1 The Letters of Henry Adams, ed. J. C. Levenson, Ernest Samuels, Charles Vandersee, and Viola Hopkins Winner, Harvard University Press, Cambridge, MA, 1988, Vol. 1, p. 290. 2 The Letters of Wilkie Collins, ed. William Baker and William M. Clarke, Macmillan Press, London, 1999, Vol. 2, p. 344.

5

The Nuclear Revolution The reality that Wilkie Collins foreshadowed had to wait for seventyfive years. World War I saw the advent of poison gas as well as the military aircraft and the effective submarine, but their lethal power recognizably represented further steps of broadly the same order as previous ones. World War II, for all its industrially driven ferocity, had run for nearly six years before anything plainly transforming emerged. In mid-July 1945, however, following a secret development programme in which, though its origins lay in cooperation between the United Kingdom and the United States, the latter had increasingly borne the main burden, a nuclear weapon was successfully tested at Alamogordo in the New Mexico desert. President Harry Truman, with the concurrence of his British counterpart Prime Minister Winston Churchill already given before the test, authorized the operational use of two bombs in order to bring the bitter war against Japan to a swift end. Less than a month later, on 6 and 9 August, they were dropped in close succession upon Hiroshima and Nagasaki, with destructive effects on a scale far beyond anything previously achievable by single weapons. The Japanese leadership recognized that this amounted to a proof of irresistible force available, and surrendered unconditionally within a week of the attack on Nagasaki. The long-term effects of the bombs have sometimes been exaggerated. The sites of the two cities did not become radioactive deserts—flowers bloomed again in the ruins of Nagasaki the following spring—and Japanese clinical statistics do not bear out assertions that the cancer-producing results of the radiation passed on to generations beyond those directly affected. 3 But the lethal enormity is nevertheless clear. In each of the cities the full death toll, immediate and delayed, almost certainly reached well into six figures. Both the necessity and the morality of what was done have been vigorously debated ever since. Was it justifiable to use these terrible weapons when the eventual outcome of the war with Japan could not, by that stage, have been in doubt, even if its timescale was? Did their use really shorten the war appreciably and thus yield a net saving of lives on both sides? Might the war-terminating message— ‘The game is up’—have been brought home compellingly to the dominant leadership in Japan by use on targets other than cities? Ought 3 I recall this from a talk at a seminar in Britain in March 2004 by a distinguished Japanese oncologist, Professor Masao Tomonaga, himself in childhood a survivor of the Nagasaki attack.

6

The Nuclear Revolution the Nagasaki strike to have followed so swiftly upon the Hiroshima one? None of these challenges—and hostile commentary suggesting discreditable ulterior motivations on the part of the United States has from time to time posed additional ones—has an uncontroverted answer, even with the hindsight conferred by the extensive access which historians have subsequently had to governmental papers of the day on both the warring sides. It is not necessary for the purposes of this book to enter the lists on these specific questions, though some of the concepts, principles, and dilemmas underlying them echo through much of the discussion offered in later chapters. It is incontestable, however, that the stark suddenness and near-stupefying magnitude of what had happened sent an enormous shock-wave through the international system. Even though the death toll in each of the stricken cities may not have exceeded that of, for example, the heaviest US fire-bombing raids on Tokyo some months earlier, it was widely, if not universally, realized that something entirely special—a deeply radical discontinuity in the business of armed conflict—had happened. The distinguished US Secretary of War Henry Stimson had already spoken in May 1945 of the bomb as bringing about ‘a revolutionary change in the relations of man to the universe’. 4 The words of Krishna quoted from the Hindu sacred text the Baghavad Gita at the time of the mid-July test by Robert Oppenheimer, one of the bomb’s scientific begetters, have become generally familiar—‘Now I am become death, the destroyer of worlds’ 5 — and not long afterwards the seminal strategic thinker Bernard Brodie, in an aphorism almost equally well known, noted that ‘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’. 6 It is perhaps less widely known that Clement Attlee—the least charismatic but far from the least considerable of twentieth-century British prime ministers, who had ousted Churchill in the general election results made known at the end of July 1945, between the nuclear-weapon test in New Mexico and the operational use against Japan—wrote in a laconic paper to some of his Cabinet colleagues within weeks of that operational use and the end of the war that ‘the modern conception of war to which in my 4

See Nemesis, Max Hastings, HarperCollins, London, 2007, p. 495. Baghavad Gita, Chapter 11, Verse 32. 6 Bernard Brodie, The Absolute Weapon: Atomic Power and World Order, ed. Bernard Brodie, Harcourt, Brace & Co., New York for Yale Institute of International Studies, 1946, p. 76. 5

7

The Nuclear Revolution lifetime we have become accustomed is now completely out of date . . . this invention has made it essential to end wars . . . the whole conception of war is [to be] banished from people’s minds and from the calculations of governments’. 7 The sense that the need to avert war between the major powers had acquired a compelling new force was henceforth a crucial pillar of security policy for those powers. It could not however be the only consideration. After all, if nothing else was of comparable moment, the goal could most dependably be achieved by pacifism, the unconditional forswearing of all recourse to arms. Policy had to work out in practical terms how to achieve the goal in a complex and difficult world environment without abandoning other goals that were rightly seen as no less essential, such as (for the countries of the West) the maintenance of free and open democratic societies.

Interpreting the Revolution Especially in the earlier post-1945 years, much discourse about nuclear weapons, whether sympathetic or hostile to their continued existence, failed—despite insights such as that of Attlee—to comprehend deeply and clearly enough how they had transformed the whole idea of all-out warfare between major powers. Without that understanding, discussion could slip too easily into applying concepts derived from ‘classical’ warfare which had ceased to be adequate or apt. In the field of astronomy, the concepts of Ptolemy, long predominant, postulated a central Earth around which other bodies, including the Sun, revolved. Within this structure of analysis the task of accounting for observed facts required more and more convoluted explanatory hypotheses. Copernicus however espoused a different idea of great simplifying and clarifying power: that the Earth revolved round the Sun. Once this was grasped, much that had seemed perplexing or enormously complex now fell into place. A good deal of commentary about nuclear weapons—some of it from distinguished figures—long remained of Ptolemaic character. It is important to grasp the nuclear equivalent of the Copernican perception. 7 GEN 75/1: The Atomic Bomb, Memorandum by the Prime Minister, 28 August 1945, paragraphs 4 and 14.

8

The Nuclear Revolution The sudden and enormous leap in destructiveness brought by the advent of nuclear weapons was of a different order from that caused by, say, gunpowder or aircraft. It is not enough to view it as merely a ghastly intensification of the human horror of war. It did something fundamental at a colder level of analysis. It carried the potential of warfare past a boundary at which many previous concepts and categories of appraisal—both military and political—ceased to apply, or even to have meaning. In the past the normal professional aim of military operations, at least in major war, had been to so damage or disrupt the enemy’s forces that he either had none or too few remaining effective, as with Hitler in 1945, or at least was denied the means of bringing them to bear, as would-be invaders of Britain were by the naval victories of 1588 and 1805. In essence, military victory customarily meant depriving the adversary of either the strength or the reach to land his blow. But nuclear explosive potential—massively multiplied within a decade of 1945 by the development of fusion-based weapons, with destructive force many times that used against the Japanese cities—soon became combined with the worldwide delivery capability of modern missiles and the diversity and elusiveness of new missile platforms such as long-endurance nuclear-powered submarines. All this, exploited by the formidable resources of developed states, made achievable what is for practical purposes infinite destructive power, unstoppable and inexhaustible at any humanly relevant levels. This may well reflect a development even more fundamental than the effect of one especially dramatic scientific discovery. It is misleading to think of nuclear explosive power as just an appalling freak of nature, regrettably left around like a box of matches for the children to find. It is not merely conceivable but downright likely that if nature did not contain this particular potential or it had not been discovered the world would before long, in the advance of knowledge of the physical workings of the universe, have reached by some other route a capability for intolerable destructive force, especially against the vulnerabilities which the complexity of advanced modern societies entails. What, for example, might chemical or biological weapons by now be capable of had they and their delivery systems received the same investment of resources, across sixty years and more, as has been devoted to nuclear armouries? From this standpoint the advent of nuclear weapons may have been no more than

9

The Nuclear Revolution the particular form first taken—perhaps to macabre advantage, in that its savage abruptness made the reality unmistakable—by an evolution that was anyway inexorable. The issues for strategy and statesmanship, and for ethics and law, need therefore to be recognized as not only novel but also broad and basic. They concern how we are to live for the rest of human history (of which we may be simply at the start) with what our technological mastery confers: that is, the potential availability of virtually boundless capacity to inflict harm upon one another. In the light of this, our thinking has had to grapple afresh with what victory in war can now mean; how nuclear weapons relate to the rest of the spectrum of possible force; and what that relationship means for their handling in political and arms-control contexts. A contest of strength between infinitely strong adversaries is a logical incoherence. Nuclear weapons have produced the reductio ad absurdum 8 of warfare in the classical sense, and former categories and standards of professional appraisal no longer suffice. Many of the paradoxes and dilemmas which beset thinking about nuclear weapons are no more than the reflection of this reductio as we hold it up to the light from various angles. If victory is taken to mean, as it has ideally done in the past, rendering the adversary incapable of doing further harm, it now becomes unattainable when adversaries with large modern nuclear armouries confront one another. In such settings the concept of using nuclear weapons to achieve it is indeed, as some senior military figures used to comment acerbically, ‘military nonsense’. But that merely casts into sharper relief what has always been one of the possible aims of military action (as von Clausewitz classically recognized) and has had to become—because alternatives have fallen away—the central aim, whether at the conventional or at the nuclear level: that of operating upon the adversary’s decisionmaking and resolve, and of doing so before intolerable loss has been suffered—inducing him, even though he still has the physical power to continue, nevertheless to step back, by convincing him that any previous assessment of net advantage to be reaped by going to or remaining at war was mistaken. This limitation of rational objective is the key ‘Copernican’ reality. The measure of success then is how far the terms upon which the adversary is finally brought to cooperate in war-termination are compatible with our key political aims, such 8

10

I owe this key encapsulation to Andrew Edwards, lately of the UK Treasury.

The Nuclear Revolution as not surrendering sovereign territory. That is inevitably a narrower and less neatly assured concept of victory than the ‘disarming’ concept. It is however not a vacuous or pointless one; and no other is available. The underlying fact is that nuclear weapons colossally extend the spectrum of possible force. As a result they cannot realistically be viewed and managed (or prohibited) as though they were just aberrations within it like previous sorts of weapon seen as especially inhuman or disagreeable, such as soft-nosed bullets or poison gas. Unlike those, nuclear arms are irresistible and conclusive. They stretch the spectrum of force, as those did not, to near-infinity, and they thereby transform the potential significance of the whole spectrum. It has never been possible for war between states on issues where they regard their vital interests as being engaged to work like a football match, a contest constrained by rules overseen and sanctioned by some external authority. Escalation in war is far from certain, as pages [62–7] explain. But given the commitment nations bring to war, the passions a massive conventional conflict would have aroused, the hostility between opposing polities, and the power of nuclear weapons to trump lesser weapons, we could never take it as sure—whatever might have been said beforehand—that losers would accept non-nuclear defeat in obedience to treaties, promises, or international law. Even if all nuclear weapons had been scrapped, there could never be total assurance that a new Hitler would go down to defeat without building some and using them, or that a Churchill or Roosevelt would risk letting him thus prevail rather than make counter-preparation. Accordingly, though we can recognize subdivisions of the spectrum of force, and abstractions like thresholds and firebreaks can have a limited place in thinking about it, no conceptual boundary could be wholly dependable amid the stresses of major war. There is no way of segregating nuclear potential in a compartment impermeably sealed from other levels of conflict—no way of physically ensuring, as between states with the necessary resources, that vital-interest conflicts which start at the bottom of the spectrum do not end up at the top. The nuclear risk permeates all war between such states, whether or not particular weapons are used or even exist at the outset, and the risk has to be tackled above all by preventing major war, not by attempting to discipline it. We cannot dissolve the reductio ad absurdum—no sure path is available to rerationalizing warfare of the

11

The Nuclear Revolution kind and scale seen in the two World Wars of the twentieth century, even if we somehow thought that desirable. And this truth is not temporary or reversible, since the knowledge of how to make nuclear weapons can never be erased. There is no possibility of return to pre-nuclear innocence (if indeed ‘innocence’ is an apt term for the mindsets that could produce those two World Wars). In brief, nuclear weapons profoundly revolutionize the notion and the significance of all-out warfare. Our thinking and our language have had to adjust to that, as the next chapter notes.

12

2 The Tools of Thinking

Evidence At one level understanding nuclear weapons is all too easy. Their uniquely appalling destructive power is manifest. Both the development and the evaluation of ideas about them pose special problems, however, for a welcome reason: that there is little hard evidence about the political, strategic, and military consequences of their use. Their physical working is now thoroughly understood by scientific specialists. There may remain some uncertainties about their effects in extreme situations, as was illustrated by disputes in the 1980s about the possibility that a catastrophic global ‘nuclear winter’ would be caused if very large numbers were detonated. But the main effects are well known, most vividly and terribly from what happened at Hiroshima and Nagasaki in August 1945. The data on their use in war, however, are confined to those two events, when nuclear weapons were used conclusively by a nuclear possessor against a non-possessor. It follows that almost all the enormous body of writing and the sophisticated conceptual structures which have been built up about the use of nuclear weapons for preventing or conducting conflict are in a strict sense speculative. Complex theorizing about nuclear weapons and war is sometimes referred to, a shade dismissively, as a sort of ‘theology’, and it has been wryly said that the two disciplines have at least in common that we shall find out for sure what is true and sound only when it is too late to do anything further about it. We have no empirical data beyond 1945 about how events may run if nuclear weapons are used, and none at all about nuclearweapon powers coming seriously to war with one another whether

13

The Tools of Thinking with or without such use. Even propositions about the achievement of nuclear weapons in deterrence cannot look for rigorous evidentiary proof, since such propositions are essentially about alternative history—about what would or might have happened had matters been other than they actually were. This ought to instil in all who make diagnostic or predictive assertions about these matters a degree of intellectual humility that is in practice not always evident.

Learning In the absence of empirical data we have to rely upon concepts, hypotheses, and inferences not directly or fully tested. There is by now a vast and diverse body of reasoning and conjecture about what factors may prevent or may cause nuclear war, and about how it might run if it ever started. Certainty is not available, especially across the huge range of possible situations. The causes, circumstances, and course of conflict involving the use of nuclear weapons could vary widely, and it is a simplistic fallacy to talk as though it were a single undifferentiated phenomenon. But in matters so untested yet so important the temptation to hyperbole and over-assertion, whether for emphasis or to seize attention, is very strong. Contrary opinions, imposingly argued or proclaimed by distinguished individuals, are to be found over the years on almost every issue, and it is hard—especially for those lacking time or inclination to immerse themselves deeply in the debates—to judge immediately where wisdom lies. The concepts, though often complex ones which strain longer-established frameworks of analysis, are not in themselves or in their language deeply inaccessible to the non-specialist, as in some branches of mathematics or science; but for that very reason it is difficult even to assess readily what constitutes expertise. No neatly identifiable qualifications are available. Neither senior military rank nor scientific distinction, for example, can claim presumptive authority except on limited technical points. There is no substitute for looking at the merits of what is said rather than the eminence of who said it. All this is far from meaning that the writings are worthless or guesswork, that one opinion is as good as another, that we must despair of choosing between sound and unsound analysis and prescription, or that deep feeling and the revulsion with which any reasonable

14

The Tools of Thinking human being must contemplate the effects of using nuclear weapons will suffice instead of hard thinking. The utterances of nuclear theorists may often be thought convoluted or remote, their content or expression repugnant, or their conclusions over-elaborate. But now that nuclear knowledge irrevocably exists there have to be policies about nuclear weapons, and those policies have to rest on concepts of some kind. It matters a great deal whether the concepts are good or bad, since the consequence of getting policies seriously wrong could be unparalleled calamity. We need therefore to think as clearly and realistically as the powers of the human mind allow. The strategic-studies effort worldwide on nuclear weapons in the last sixty years and more represents in the round a massive intellectual achievement. We are perhaps fortunate that the early years of our nuclear learning took place when the horrors of world war even before the use of nuclear weapons were still vivid in widely shared memory, and when the weight and immediacy of the cold war’s ideological and geographical confrontation provided both a powerful simplifier of analysis and an acute sense of danger to discourage risk-taking or experiment. However that may be, the cumulative and shared understanding of nuclear issues among policy-makers and decision-takers, at least in the West, progressively became far deeper, more subtle, and more secure than it had been in the late 1940s—or even in the early 1960s, before collective effort in NATO’s Nuclear Planning Group (NPG) and US/Soviet dialogue in the framework of SALT (Strategic Arms Limitation Talks) and then START (Strategic Arms Reduction Talks/Treaties) began to take root.

Language There is another health warning to be suggested. It concerns the vocabulary of discourse about nuclear weapons. Again because of the lack of empirical data, we have to rely upon careful conceptual thought. For almost all of us, language is the necessary vehicle of that. Its accurate use and comprehension are accordingly of special importance in this field. Over the years there has grown up, as happens in most fields of professional study, a large array of customary metaphors and terms of art. Both the metaphors and the terms of art can carry risks to understanding.

15

The Tools of Thinking The metaphors are useful shorthand for referring to abstractions which conceptualize external reality in order to facilitate thought and dialogue. But discourse may slip into reifying the abstractions— treating them as though they denoted specific and certain phenomena with some sort of concrete existence outside the mind. The two examples of this trap which I encountered most often concerned the ‘nuclear threshold’ and ‘escalation’. The ‘nuclear threshold’ normally referred to the stage at which the leaders of a nuclear-weapon-owning state whose forces were being overpowered in a major non-nuclear conflict were envisaged as facing the hard choice between using their nuclear weapons and accepting defeat. (During the cold war planners in Western countries customarily assumed that it would be NATO, not the Warsaw Pact, that found itself in this situation.) This was a complex notion, yet the metaphorical threshold was occasionally alluded to as though one might almost stub one’s toe on it. Risks of miscomprehension were compounded by the adjective ‘nuclear’. Comment sometimes appeared to suppose that the location of the nuclear threshold was fixed mainly by the characteristics of nuclear armouries. But the prime determinant, as any proper grasp of the concept must show, would not be those characteristics but the relative combat performance of non-nuclear forces. It used to be sometimes suggested, against this, that the forward deployment of NATO nuclear delivery units, especially on West German territory, risked posing, in the face of a postulated Warsaw Pact thrust, a ‘use-or-lose’ dilemma which— whether as an inescapable fact or as a menacing stratagem deliberately adopted—could drive the timing of NATO nuclear action and so set the threshold. This was however in no way part of NATO’s doctrine or planning. Deployments were modified over the years to reduce any risk that the dilemma might arise swiftly; and forward commanders had neither the authority nor, at least in later years, the physical power to launch nuclear weapons without political clearance. There were widely in place electronic locks (‘permissive action links’—PALs) for which weapon custodians did not have the codes in advance of high-level release. ‘Escalation’ is a concept the frequent misunderstanding or misrepresentation of which has an importance meriting the fuller treatment offered in Chapter 6. The point here is simply that in the military context the word is shorthand for possible sequences of human choices, whereas it is often presented as though it referred

16

The Tools of Thinking to something approaching the inexorability of a chain reaction in chemistry. The risk with terms of art lies in the use of familiar words under conventions recognized by cognoscenti as referring to ideas narrower than, or otherwise different from, those which the words may convey to the lay reader. Four examples are as follows:

r In established nuclear discourse, at least in the West, ‘first use’ means something quite different from ‘first strike’. The former refers to the first occasion on which nuclear weapons are used in a conflict, on any scale, by either side. The latter, by convention, refers to a scenario in which one side mounts a pre-emptive operation—typically requiring attack with a large number of nuclear weapons—in the hope of destroying or much reducing the other’s capability for nuclear retaliation.

r In ordinary usage, and indeed in the context of international law about the use of force, the word ‘threat’ has overtones of deliberate and menacing intent. In security discourse, however, it may often—even usually—refer simply to a physical capability to do harm in the hands of a state regarded as conceivably an adversary. It does not then carry any necessary imputation about that state’s political attitude or planning.

r ‘Weapons of mass destruction’ (WMD) is generally accepted, following a definition set by a United Nations body, 1 as a convenient shorthand for referring collectively to nuclear, biological, chemical, and (less invariably) radiological weapons. This customary usage does not necessarily imply that inflicting mass destruction, in the sense of indiscriminate destruction of cities or populations on a huge scale, is the only way in which weapons in these categories could be used. Nor, conversely—as the disaster of 11 September 2001 demonstrated—is it the case that only such weapons could be used for mass destruction in that sense. The usage happens in any event to be an undesirable lumpingtogether of weapon categories which are in practice of widely different security significance. It has, however, become too deeply entrenched to be easily discarded.

1 United Nations Commission for Conventional Armaments, 1948—see S/C.3/32/ Rev.1.

17

The Tools of Thinking

r By semantically arbitrary convention NATO air planning for assault on ground targets assigned the term ‘strike’ to action with nuclear weapons and ‘attack’ to that with conventional ones. Attempts to distinguish in some fixed conceptual way between the terms ‘strategic’ and ‘tactical’, whether applied to weapons and delivery systems or to targets and particular operations, can be another source of confusion. Operational uses would not come neatly labelled, and different countries may well have widely divergent perceptions about what, in their particular circumstances, ranks as strategic. The nuclear revolution, as Chapter 1 has conveyed, has in any event changed the basic significance of the distinction away from that attaching to it in past warfare. In arms-control business the dividing line, if one is needed, has been a matter more for ad hoc negotiation than for the application of pre-existing and consistent objective criteria. In such business the distinction has mostly been found convenient for static classification of systems where measurable performance features such as maximum delivery range could be laid down as its basis. The term ‘strategic’ in these contexts has customarily been used as shorthand for nuclear delivery systems of intercontinental reach of a kind and destructive power supposed to be appropriate for ultimate punitive action. By most normal standards of understanding, of course, any use whatever of nuclear weapons would have to be regarded as of strategic significance. As the context of thinking about nuclear weapons has moved away from the almost-exclusive concept of encounter between the two huge adversaries of the cold war, efforts to distinguish in some neat way between ‘strategic’ and ‘tactical’ (or ‘theatre’, ‘sub-strategic’, or ‘pre-strategic’) have become less and less illuminating. The UK government in 2007 2 explicitly abandoned any such efforts. The word ‘deterrence’ also generates occasional misunderstanding, in two almost opposite ways. Chapter 3 brings out that it is inherently of far wider application than just to nuclear weapons, but it is sometimes used over-narrowly as though that were its sole reference. On the other hand, as its derivation from the Latin word meaning ‘frighten’ indicates, it strictly means preventing unwanted action by displaying the prospect of pain or penalty. It has however become 2 See for example Minister of State for Defence Lord Drayson, House of Lords Official Report, 24 January 2007, column 1107.

18

The Tools of Thinking increasingly customary in English-language strategic discourse to speak of ‘deterrence by denial’, meaning foiling hostile action rather than scaring it off, as one form of deterrence. (French does rather better in this regard, through using the word ‘dissuasion’.) Purists like me may disapprove of calling this ‘deterrence’; but in such matters usage is ultimately sovereign, and the purists must probably by now accept defeat on the point. That said, an interesting linguistic usage in English, observable latterly with increasing frequency, classes as ‘dissuasion’ such concepts as the idea voiced in the United States of maintaining military capability (whether offensive or defensive) so evidently weighty and sophisticated that potential adversaries will be led to conclude— before any question arises of direct operational deterrence—that efforts to compete with it cannot be expected to generate benefits worth the political and economic costs. That is indeed a legitimate use of the word, but its users should recognize that it is just one form of dissuasion. Dissuasion is strictly a wider term than deterrence, covering all paths to convincing another party that it will not profit by a given course of action. As some early US usage recognized, 3 deterrence, relating to one type of path, is properly a subset of dissuasion, not a separate idea. There may be risks of confusion in discourse if particular users seek arbitrarily to assign to dissuasion a narrower sense than normal and long-established practice understands. In brief, the message of this chapter is that thinking about nuclear weapons must be constantly on the alert—the more so in the merciful absence of historical experience to anchor and calibrate discussion— both to recognize uncertainty and to probe behind words and customary expressions so as not to lose sight of the underlying realities, or misconstrue them. War-preventing deterrence in the nuclear age, the main concern of the next chapter, is otherwise at risk of being built less efficiently and dependably than it could and should be.

3 See for example the chapter by Percy Corbett in The Absolute Weapon, ed. Bernard Brodie, Harcourt, Brace and Company for Yale Institute of International Studies, New York, 1946, p. 148.

19

3 Deterrence

The Concept Si vis pacem, para bellum. The idea of deterrence goes back a long way—it was not new even when a Roman formulated that Latin axiom: ‘If you want peace, make ready for war’. Deterrence arises from basic and permanent facts about behaviour which have always had a part to play in the management of human relationships. In deciding how to act, people customarily seek, whether consciously or not, to take into account the probable consequences of what they do. They refrain from actions whose bad consequences for them seem likely to outweigh the good ones. And we exploit these universal realities as one means of helping to influence others against taking action that would be unwelcome to us, by putting clearly before them the prospect that the action will prompt a response that will leave them worse off than if they had not taken it. When a small boy is told that if he bullies his little sister again he will be sent to bed without supper, he is being subjected to deterrence. Even when the warning is not voiced explicitly, if improvement in his behaviour is shaped by his sense of the risk of punishment he is being deterred. It is however only since the colossal shock of the nuclear revolution that deterrence has become so special and salient a concept in discourse about international security. Old-style defence, in the sense of physically warding off attack, could no longer play as large a part as it customarily had in the search for security. In face of the vast new destructive power only virtually impermeable protection would be good enough, and such a standard was simply not attainable. So the prevention of war became imperative; and given that there

20

Deterrence could be no realistic expectation, amid the world’s tensions during the decades following World War II, that political and diplomatic processes alone could be relied upon for that, deterrence had to step forward. The power of nuclear weapons conferred a uniquely enormous ability to display the prospect of disadvantage. The deep distrust between the two cold war alliances, with the Soviet Union seen by the West as a massively armed adversary of alien ideology, oppressive behaviour, and perhaps expansionist propensity, caused leaders to fear that intolerable things might sooner or later be done unless this prospect was exploited to rule them out. More generally, the sudden leap in military capability that nuclear weapons embodied increasingly convinced most people—vividly typified as early as 1945 by British Prime Minister Clement Attlee, already cited on pages 7–8—that the need to avert war, rather than have to wage it, had acquired a new and special cogency. War to the maximum of physical capability could never again be viewed as just an inferior and unpleasant way of managing international affairs. It had ceased to be a tolerable way of managing them at all. The world faced both the necessity and the possibility—these being in effect two sides of a single coin—of unmistakably convincing anyone, however unprincipled or sanguine, who might have been minded to initiate war with an advanced power on a crucial issue that doing so could not possibly, on any calculus, yield net benefit. Libraries-full of writing have accumulated about deterrence theory centred upon nuclear weapons. But however elaborate the analyses and the intellectual architecture become, they have to rest ultimately on the simple paradigm of human behaviour summarized earlier. The idea itself is not hard to understand. Teasing out how best to make it work amid the complexities of international affairs and in the terrifying presence of nuclear weapons, however, is by no means straightforward. Most discussion of deterrence has understandably tended to centre upon the immense reality of nuclear weapons. But it must not be supposed either that preventing their use is the sole aim of deterrence, or that they are its sole instruments. The imperative has been to prevent all war between major powers, not just nuclear war. There are two reasons for that. First, the fearsome events of August 1945 did not mean that war suddenly became nasty having previously been nice. It had not previously been nice. The carnage of the Western

21

Deterrence Front trenches in World War I is not to be forgotten; and World War II took, directly or indirectly, something like fifty million lives even before Hiroshima and Nagasaki were struck. The cosy-sounding adjective ‘conventional’ for classes of weapon like those of World War II and their still-more-potent modern successors should not mislead us. Second, non-nuclear war is not just appalling in itself. It is also the likeliest route to nuclear war. In practice, indeed, it is the only likely route, since (as Chapter 6 will argue) scenarios of the holocaust being launched by accident or through technical malfunction are farfetched to the point of fantasy. The risk would be at its highest when bitter conflict had already broken out at a lower level. War-prevention needs therefore to operate on all levels of military conflict between nuclear-capable states. Partly on that account, military deterrence cannot operate only by means of nuclear weapons. Their use can scarcely be credible, and their deterrent power therefore scarcely effective, against aggression at much lower levels of force or of national interest engaged, as US experience in the Korean peninsula and Vietnam illustrated. Military capability in the hands of states is moreover, as page 11 has noted, a continuum within which there can ultimately be no firebreaks guaranteed in advance or for all circumstances. Its effect whether in war or in deterrence works (or fails to work) as a package, not a stack of sealed boxes. A combatant failing to get his way at one level would always be able to consider his options at another. The various levels of military force are therefore complementary and interdependent; all can contribute to deterrence. (I came to dislike, and increasingly sought to avoid, the unqualified phrase ‘the deterrent’ used as referring just to nuclear capability, since that might seem to imply that only nuclear weapons contributed to the task of deterrence.) It is worth noting also that the interaction of various levels of capability within overall deterrence may helpfully ease the weight that would fall on some of the levels if they had to be taken in isolation. During the cold war, as NATO strategy and force provision exemplified, the overshadowing nuclear component was rightly seen as making it unnecessary to pursue the goal of assured defensive adequacy or parity with the Warsaw Pact at non-nuclear levels (whether conventional or other, such as that of chemical weapons [CW]). Without nuclear weapons that goal—which would have been more

22

Deterrence elusive, more unstable, and much more costly in both money and manpower—would have returned to centre stage. Effective deterrence has two main components, not just one. It undoubtedly requires that the adversary perceive the existence of capability, and of general will to use it if necessary, to exact in one form or another costs that he would find unacceptable. But it also requires that he have a sufficiently clear understanding of what is the action from which he must refrain. That was one of the key reasons why the East/West deterrent stand-off in Europe proved so secure, at least once the Berlin issue had been stabilized early in the 1960s. The Iron Curtain was deeply unpleasant, but its line and its importance were unmistakable—a harsh illustration of the axiom that good fences make good neighbours, or at worst non-warring ones. The breakdowns of deterrence most evident to Western countries since World War II include the crises over Korea in 1950, Cuba in 1962, the Falkland Islands in 1982, and Kuwait in 1990. In most if not all of those instances the existence of adequate military power to defeat and reverse the offence was manifest. The breakdowns were not caused by misperception of that on either side. They resulted from failure—whether blameworthy or not—to implant unmistakably and convincingly in the adversary’s mind, in advance, an accurate sense of what would absolutely not be acquiesced in. This is best done not just by one-off assertion but through understandings of interest, and perceptions of consequent resolve, built up by behaviour over time. A qualification to this may be necessary. It is sometimes suggested, for example in the context of South Asia, that adversaries in perceived confrontations could with advantage spell out fully what their ‘red lines’ are—that is, exactly what they would regard as intolerable. The drawback to this is that (quite aside from the risks of generating unsettling and abrasive political argument in the process of spelling out) it might be misinterpreted as implying that every unwelcome action that did not actually cross a pre-specified red line would be acquiesced in. Such misconstruction might open the way to overconfident calculation by an adversary of what he could get away with, and thus to risk-taking that would damage stability. In circumstances like these the maintenance of a degree of ambiguity— of uncertainty—may make for surer deterrence. Situations differ, and there is no escape from pragmatic judgement. But the need

23

Deterrence for consciously considering the issue of timely communication, by whatever method, ought always to be recognized. All this said, deterrence does not require a precise specification of what form the non-acquiescence will take. It requires only that it be made plain that the objectionable action will not be allowed to stand; that there is the power to prevent its doing so; and that there is also the resolve to use robustly, from within the range of capability available, whatever is found necessary for the purpose. Over-exact advance specification of intended means may actually weaken deterrence. For instance, it may help the adversary in one way or another to calculate how to evade or head off the response, or to reduce its impact. In open societies, it may prompt contentious national or international debate that may come to be seen as impairing the credibility of determination to act. That is why I have argued elsewhere, 1 for example, against the notion of declaratory policies that would purport to guarantee a nuclear response to the use of biological weapons (BW) or chemical ones. We usually ought, within the limits of any permanent international commitments or moral imperatives that may constrain our options, to avoid ruling any specific means either in or out in advance. The essence of the deterrent message, however communicated, should be simply that we will do whatever we find necessary to achieve our aim—preferably the minimum necessary, but not less. There is also a dimension about probabilities. In ideal circumstances deterrence is most securely provided by certainty both that grave misbehaviour will be promptly detected and that condign retribution or rectification will follow. Amid the world’s complexities, however, that partnership of assurances is not always achievable. Where that is so there is no option—for all parties—but to deal in probabilities and possibilities. That need not destroy deterrence. During the East/West confrontation leading figures as diverse as President Charles de Gaulle and Dr. Henry Kissinger at one time or another, in their own styles, reminded their hearers that it was at best unprovable that the United States would risk embarking, for European causes, on sequences of action that might imperil its own cities (‘sacrificing Chicago for Hamburg’). But the contrary was also unprovable for Soviet leaders. Even a modest chance of a huge penalty can have 1

24

See Survival, Vol. 37, No. 4, Winter 1995–6, pp. 189–91.

Deterrence great deterrent force. In different jargon, the ‘expected value’ of a disbenefit—the magnitude of the penalty factored with the perceived probability of its being levied—can outweigh even a high valuation of hoped-for benefit.

Criticisms and Fallacies Much misunderstanding has formed around the relationship between deterrent possession of nuclear weapons and their use. At one edge of the debate some have attempted—perhaps in the search for a way around uncomfortable political questions, or ethical ones such as those to be discussed in Chapter 5—to establish a complete disjuncture between possession and use, claiming that the former need carry no implication at all about the latter. At the other edge lie suggestions that deterrent possession must entail a fixed and implacable—almost an automatic—disposition to press the nuclear launch button given grounds for doing so. The truth is more complex than either of these extremes. There is clearly some force in the notion of ‘existential’ deterrence, 2 taken as meaning that if a state possesses a substantial nuclear armoury this is bound to make others cautious about gravely provoking it, whatever its leaders may or may not have said (or others may know or think they know) about intentions for the use of the weapons. We cannot however infer from this that our own capability will be durably effective in contributing to deterrence—especially in times of stress, when it is most needed—if there are no realistic concepts for its use, or if we have a settled resolve never to use it. The concepts do not have to extend to detailed targeting plans against specific adversaries. Especially in the post–cold war world, a deterrent stance addressed simply ‘To whom it may concern’ may be entirely appropriate. That is, after all, the basis on which most armed forces are maintained. But a categoric policy of non-use would be a different matter. Deterrence and use can in logic be distinguished, but not wholly disconnected. We cannot say that nuclear weapons are for deterrence 2 This term was used by McGeorge Bundy in a notable article ‘To cap the volcano’ in Foreign Affairs, Vol. 48, No. 1, October 1969, pp. 1–20.

25

Deterrence and never for use, however remote we judge the latter eventuality to be. Weapons deter by the possibility of their use, and by no other route. The concept of deterrence accordingly cannot exist solely in the present. It inevitably contains a reference forward to future action, however contingent. The reference need not entail automaticity, or even firm intention linked to defined hypotheses. It need entail no more than a refusal to rule out all possibility of use. But it cannot entail less than that. There are also practical reasons, of two kinds, why purely existential deterrence will not serve. Both spring from the fact that an operational nuclear armoury is not an inert pile of matériel but has to involve a large number of people. First, it is unrealistic to suppose that a firm intention never to use weapons, and the lack of any plans for their use, could be permanently and dependably concealed from an adversary; yet if he perceives them the deterrent effect of the armoury is hugely diminished if not indeed destroyed. Second, it cannot reasonably be supposed that the commitment of many thousands of individuals serving in or working for the forces concerned—often in career-long and very demanding tasks—could be durably sustained in the known absence of any policy for use or of contingency planning which they could regard as seriously intended. In brief, and whatever ultimate decision-takers may envisage in the privacy of their own thinking, a structure of deterrence cannot be built upon a state policy of absolute non-use, or without genuine concepts of possible use. These facts can be entirely compatible with a deep and sincere belief by a nuclear possessor that the need to use the weapons would be a disaster, that the option of simply not using them in the event must always be recognized, and that the crucial aim of possessing them must be to help, alongside other instruments of policy, prevent any circumstances from arising that might leave their use as apparently the least bad option available. But there lies at the heart of deterrence in the nuclear age an inescapable paradox: the more seriously the possessor is believed capable in extremis of using the armoury, the less likely it is that others will cause or allow circumstances to arise challenging its use. And the converse is also true. The point goes beyond the matter of basic refusal to exclude use. Especially though not only in the particular setting of the cold war confrontation, plans and capabilities had to provide options

26

Deterrence for use that could be credible. This meant, for example, developing weapons of greater accuracy and lower explosive yield, and plans for a more limited scale of use and more constrained selection of targets, than might feature in an unconstrained apocalyptic holocaust. The development of such weapons and plans was intermittently assailed by anti-nuclear campaigners in the West as implying that nuclear warfare was thought probable, or as betokening a dangerously increased inclination to regard it as a tolerable enterprise. The assault sometimes seemed to be seeking to imprison defenders of nuclear deterrence in a manufactured dilemma: if nuclear weapons were too powerful they were indiscriminate, and that was wicked; if measures were taken to make them less indiscriminate, this would make them more usable, and that too was wicked. But such criticisms failed to recognize the inevitability of the paradox. Ultimately, accepting them would lead towards less credible deterrence and thus more risk of war, not less. The evident possession of practical options was directed entirely to making war as remote an eventuality as possible. 3 Deterrence is a concept for operating upon the thinking of others. It therefore entails some basic presuppositions, whether or not made explicit, about that thinking. In the nuclear context critics sometimes maintained that these presuppositions must impute to adversaries very cool and sophisticated rationality, or deep malevolence, or both; and that the first imputation might well be precarious and the second unwarranted. But deterrence does not depend upon either. As to the first, though constructing an effective deterrence system embodying nuclear weapons is in some respects a complex affair, recognizing its message once it is credibly in place requires no more than simple rationality. Only a state ruler possessed by a reckless lunacy scarcely paralleled even in pre-nuclear history would contemplate with equanimity initiating a conflict that seemed likely to bring nuclear weapons down upon his country. As to the second, a policy of deterrence does not assume the presence of a profoundly evil adversary continuously resolved upon striking massively the moment guard is lowered. It is preferably directed on a long-run basis at closing off hostile options, obviating temptations before they can be seriously felt, and fostering a settled mindset which even in 3 Public explanation of this concept in the United Kingdom is illustrated in Appendix 1, an essay which I wrote and which the government approved for inclusion in the 1981 Statement on the Defence Estimates (Cmnd. 8212–1).

27

Deterrence sharply adversarial situations will not entertain military aggression as a possible course of action. To judge that nuclear weapons probably played a part in the remarkable maintenance of peace through over forty years of East/West confrontation, with many potentially combustible interfaces, does not have to postulate that the Soviet Union always harboured a constant formed intent otherwise to embark upon sweeping military conquest. 4 It rests on a sense that had armed conflict not been so manifestly intolerable the ebb and flow of frictions might have been managed with less caution, and a slide sooner or later into major war, on the pattern of 1914 or 1939, might have been less unlikely. Another occasional criticism argued that deterrence in the nuclear setting could not ultimately be credible either because no possessor could hope to be better off after a nuclear conflict than before it, or alternatively because in some situations the deterrer might be at risk of suffering proportionately more than the ‘deterree’ in any substantial exchange. The latter hypothesis might have arisen, for example, if a medium-sized nuclear-weapon state like Britain or France had found itself facing on its own a massive one like the Soviet Union. But such arguments mistake in two respects what is the crucial calculation. First, the harm likely to be suffered by wouldbe deterrer A is of course not wholly immaterial in the structure of deterrence, since it will bear upon the to-be-deterred B’s assessment of A’s resolve. The key calculation to be made, however—the key comparison which deterrence seeks to impose upon B—is between how B himself fares if he embarks on aggression and how he fares if he does not. Deterrence ‘du faible au fort’—of the strong by the weak—as French analysts used to express it, can be entirely coherent. Neither Britain nor France could have obliterated the Soviet Union, as it could them; but to pose fearful damage as the cost of reaping the advantages of their removal was nevertheless a valid approach. Second, the relevant comparison, for both A and B, is not between before and after, but between alternative futures—the future ‘if I do’ and the future ‘if I don’t’. Britain, unquestionably one of the victors in World War II, was plainly worse off in almost all material respects in 1945 than it had been in 1939; but that does not mean that the decision to resist Hitler proved foolish. 4 Recognition of this can be found very early in the nuclear age—see Arnold Wolfers in The Absolute Weapon, ed. Bernard Brodie, Harcourt, Brace and Company for Yale Institute of International Studies, New York, 1946, p. 134.

28

Deterrence The fallacy of supposing that the comparison to be influenced is between what A suffers and what B suffers has a further aspect. In the course of debate during the cold war on the ethics of possessing nuclear armouries anti-nuclear moralists sometimes argued that because the Soviet Union had to be supposed to have counterpopulation concepts of nuclear-strike, and given a view (discussed in Chapter 5) that a counter-population strike would always be immoral, no Western nuclear planning adequate to underpin deterrence could be morally legitimate. But that argument, too, rested on the misconception just described. Effective deterrence does not require symmetry of prospective suffering, either in scale or in kind. It did not require the West to envisage doing in retaliation to the Soviet Union whatever it had itself suffered. Another fallacy about deterrence, more persistent than most, is the claim (from which radical deductions for disarmament policy have often then been drawn) that nuclear weapons can deter nothing save other nuclear weapons. This claim seeks to impose upon the spectrum of military force a rigidly segmented structure which, as pages 11–12 have noted, ignores the passionate realities of major war. But the fallacy can be illustrated less abstractly. ‘Nuclear deters only nuclear’ has to hold that Western possession of nuclear weapons could have had no impact upon the Soviet Union’s assessment (for example at the time of the 1962 Cuban Missile Crisis) of its options in relation to West Berlin, which it could always have overrun swiftly by non-nuclear force. It has to hold that Pakistan’s possession of nuclear weapons makes no valid contribution to its sense of security in face of India’s superiority at other levels of military force. It has to hold that if governments implacably opposed to Israel’s existence came widely to power among its neighbours and successfully made common military cause in aggression, they could and would feel confident that Israel would accept political obliteration by conventional or CW force rather than exercise the nuclear-weapon option. More concretely still, it has to hold that at the time of the 1990–1 Gulf War, Israel’s known albeit undeclared nuclear capability not only did not play but also could not have played any part in Saddam Hussein’s decision not to use his CW against Israel as an assured means of fracturing the country’s patience, and so of triggering the Israeli intervention which he deeply desired to engineer in order to split the coalition against him. And further such scenarios can be imagined elsewhere, even

29

Deterrence within the limits of the world’s political configurations visible to us today. A distinguished Ministry of Defence colleague 5 once commented to me that a nuclear state is a state that no one can afford to make desperate. That is true and to the point. The political, moral, and psychological barriers to any use of nuclear weapons are unquestionably— and healthily—very high, especially after over sixty years of non-use. But while attack by nuclear weapons would certainly be one way of rendering a nuclear state desperate, it is absurdly narrow to suppose that nothing else could, and moreover that other states could safely base deeply hostile actions upon such a supposition.

After the Cold War As the years passed after the demise of the Soviet Union it became uncomfortably evident that the history of armed conflict had not come to any sort of end. There remained a significant number of actors ready to use force illegitimately to pursue objectives inimical to international order and peace. The ending of the cold war had indeed loosened constraint on some of these actors, and the international order could not immediately count on holding them within limiting structures as solid, well-understood, and accepted as the central East/West stand-off had become. A few commentators seemed even to despair of deterrence—how could we, they asked, deter people with different value-systems and less to lose? Any agonizing of this kind was however overblown. Only the truly insane have no sense of weighing consequences. To have a different value-system does not mean having none, no currency in which the risk of unacceptable disadvantage can be posed. One may narrow the point down further. It is hard to think of any instance in history of a state or similar power grouping that had no assets which it wished to retain, no collective concern for the lives of its members, and (in particular) no interest in the survival of its ruling regime. Where potential leverages of this kind exist—that is, everywhere—it must in principle be possible to construct penalty systems influencing the behaviour of states. 5 Sir Hermann Bondi, KCB FRS, Chief Scientific Adviser to the Ministry of Defence, 1971–7, Master of Churchill College, Cambridge, 1983–90.

30

Deterrence There may well be differences in mindset, or in readiness and ability, to make entirely cool expected-value assessments of penalties and the likelihood of their being exacted. For example, some leaders— especially autocrats not well supported by candid counsellors—may be apt to let over-optimism or even fatalism cloud their sense of the weight and likelihood of medium- or long-term penalties as compared with nearer-term ones like losing face domestically. Such possibilities have to be studied carefully and borne in mind by others seeking to build frameworks of deterrence to constrain states like these. But it does not follow that frameworks cannot be built at all. The values by which Saddam Hussein governed his subjects were repugnantly alien to the rest of the world, but he nevertheless knew in 1991 that he stood to lose a great deal that he prized if he over-provoked powerful external adversaries. His abstention then from using, in order to strengthen his resistance to being expelled from Kuwait, the BW or CW which he did at the time possess vividly illustrated his recognition of that. Even North Korea, in its extremity of self-isolated and paranoiac deprivation, values regime survival. (The prime problem facing the world community as it seeks to modify that regime’s behaviour has been not that deterrence can have no leverage but that the combination of South Korea’s nearness as quasi-hostage and the unwelcome economic and social repercussions for neighbours if North Korea collapses into chaos generates counter-deterrence against at least some of the possible coercive options.) Deterrence works by displaying the prospect of costs in terms of values prized by the ‘deterree’. For almost every society known to history, physical existence and physical assets are not the only such values. Instruments for executing physical destruction—that is, primarily, military instruments—are therefore not the only possible tools of deterrence. Political, economic, social, judicial, and even religious or cultural ones (influencing whether actors believe that they have the support of those whose approval they value or depend upon) can also sometimes make a contribution. It may well be increasingly desirable and indeed necessary in future to construct— and to display credibly in advance—deterrence frameworks (wherever possible alongside positive inducements to good behaviour) in different and more varied ways than was customary during the cold war. The frameworks may need to be more broadly based and perhaps more multi-faceted, with a diverse array of penalties made plain, in

31

Deterrence order to display prospective costs that will both carry credibility (in which evident international legitimacy may be of growing importance) and weigh relevantly upon the values of particular adversaries whose future conduct is mistrusted. In such frameworks of deterrence nuclear weapons will generally have a far less salient part to play than they were assumed to do between East and West during the cold war—a part perhaps simply of ultimate underpinning. And they can certainly play no useful direct role against terrorists, who themselves have no evident asset base that nuclear weapons could credibly threaten, though states giving substantial support to them could be susceptible to various instruments of deterrent pressure. The problem of terrorists is further discussed in Chapter 6. In brief, war-preventing deterrence in the nuclear age is of cardinal importance, and organizing it needs careful understanding. It can take diverse forms, and even in the nuclear-weapon context these forms may not need always or exclusively involve nuclear weapons themselves. But it is valuable to grasp how the part of such weapons in deterrence came to be viewed in the special setting of the cold war. The next chapter explores that.

32

4 Nuclear Deterrence in NATO

The cold war set the first substantial context for fitting nuclear weapons into the framework of international security. Within that context the practical problems of incorporating them into an effective structure of war-prevention found their most testing arena, at least from the standpoint of public perception, in the North Atlantic Alliance. An understanding of NATO thinking therefore makes a good starting-point for identifying and examining the conceptual and physical apparatus that has to come into play in shaping deterrence centred upon nuclear weapons.

Doctrine and Planning The key features of mature NATO nuclear doctrine took shape during an extended evolution stretching from the Alliance’s inception in 1949 until the 1980s. Both the long time and the laborious effort absorbed by this illustrated the learning processes which the nuclear revolution imposed on thinkers and planners in every country concerned with security issues. They also, however, reflected three special features of the NATO situation. First, doctrine had to command the open support or at least acquiescence of many nations of widely different sizes and capabilities, with different histories and therefore different attitudes to defence and the use of armed force, and with different patterns of public opinion. Second, it had to accommodate the natural divergence of security perspective between those living close to the Soviet Union’s massive military power and those separated from it by great oceans

33

Nuclear Deterrence in NATO and so threatened far less directly, if at all, by anything other than nuclear weapons. Third, it had to manage the awkward but inescapable fact that the Alliance’s own nuclear capability lay overwhelmingly in the hands of a member of the latter and not the former group. During the late 1950s and early 1960s the management of these complications, especially as they bore upon the Federal Republic of Germany (FRG) and its contentment within the Alliance, caused much anxious debate. Once the FRG had joined the Alliance in 1954 it was, by any reckoning, a crucial member, especially as being geographically at the centre of the Alliance’s front-line interface with the Soviet Union and the Warsaw Pact. But recent German history meant that on both sides of the Iron Curtain it was regarded as out of the question that the FRG should become a nuclear-weapon possessor like France or the United Kingdom. Indeed, the 1948 Brussels Treaty, revised at the time of FRG accession to it in 1954, explicitly ruled that out. The problem for NATO was how, within these constraints, to have the FRG involved enough in the Alliance’s nuclear-weapon capability to feel confident that that capability was unequivocally committed to its protection. From early on in its participation in NATO the FRG, like several other European members of the Alliance, owned and operated delivery systems—initially aircraft, and subsequently also land-based missiles—to carry nuclear warheads. These warheads remained, however, in US ownership and custody, and there were fears that such arrangements would not always or in times of stress be thought to constitute sufficient involvement. For some years, therefore, considerable political effort was invested in devising and developing ideas of a ‘multilateral force’ (MLF)—that is, a force in which nuclearweapon delivery platforms like surface ships would each be operated by multinational teams drawn from several countries including the United States as well as the FRG. The MLF idea, extensively explored in the early 1960s mainly at US initiative, was however increasingly found to pose more problems than it seemed worth (especially as there was never any question of US surrender of ultimate control of the warheads themselves). Attention turned accordingly towards solutions of a more deliberative character, culminating in the setting up in 1967 of the Nuclear Planning Group, to meet regularly at both ministerial and lower levels for discussion of the Alliance’s nuclearweapon doctrines, policies, and deployments.

34

Nuclear Deterrence in NATO Rather to the surprise of some observers, this organizational device proved notably successful. From the late 1960s onwards the development of NATO doctrine on nuclear weapons, including procedures for consultation and decision-taking about them, was taken forward by systematic and active NPG work. Its success owed a good deal to the fact that though the United States was inevitably always the pivotal participant, its representatives habitually and wisely sought to avoid an over-assertive role. The NPG’s first major endeavour, the framing of political concepts to guide any initial use of nuclear weapons, was led jointly by the FRG and the United Kingdom; and a decade later the high-profile plans for the modernization of Europebased intermediate-range nuclear forces (INF) were far from being a US imposition upon passive or reluctant allies. Attention to FRG concerns and sensitivities was always a key consideration. The wider characteristics of the Alliance noted earlier often caused the work to be complicated and slow, but at the same time it meant that doctrine was tested by study, debate, and challenge. This gave the outcome not only political acceptance but also an intellectual solidity and sensitivity which an easier and less collaborative genesis might not have ensured. The doctrine was built around the Alliance’s strategic concept of flexible response, itself also developed during the 1960s. 1 That concept was recurrently misunderstood and even caricatured, especially by those—like occasional voices in Paris, following France’s withdrawal from the integrated military structure of the Alliance and its abstention from participation in the NPG’s work—who were anxious, often for political reasons, to portray or advocate wide differences between NATO doctrine and their own views. For example, the concept did not, as was sometimes supposed, envisage a pre-determined sequence of moves—an ‘escalation ladder’—to be followed in the face of aggression. It did not rule out first use or early use of nuclear weapons, but it was far from prescribing, assuming, or preferring either. On the contrary, NATO authorities continually urged member countries to improve their contribution of conventional forces, so as to reduce the likelihood or rapidity of the Alliance’s having to confront such hugely difficult options. 1 The basic document enshrining this concept was known as MC (for Military Committee) 14/3, accepted by NATO ministers at their meeting on 14 December 1967.

35

Nuclear Deterrence in NATO The core of the flexible-response concept was always the timely use of the minimum force, whether conventional or nuclear, adequate to deny an aggressor success in his objective. Aggression against so broad and diverse an entity as the North Atlantic Alliance could have a wide variety of forms and scales. Providing a capability for apt and credible application of minimum effective force to fit any scenario therefore meant that there had to be plainly available a substantial range of military options from which the Alliance could choose both for initial resistance and for how best to proceed if the first option did not succeed in halting aggression. A narrow set of prescriptions for response might have suited a single country disposed, as France long saw itself, to relate its nuclear concepts tightly to its own national circumstances. But such a set, or a rigid doctrine for the order or number of options to be entertained, would have been the antithesis of flexibility, at odds with the realities of the Alliance’s political, geographical, and strategic diversity (and perhaps also with the deterrent merit of fostering uncertainty in the mind of the adversary). NATO thinking was always clear that a major conflict was not to be conducted in sealed compartments, whether of territory or of force category, and still less in sealed compartments imposed by an aggressor to suit his own strengths and preferences. The idea of possible escalation, in the sense of being ready to change the terms of the encounter in scope or intensity beyond what the aggressor had chosen, was essential. But NATO also recognized that the prospect of having abruptly to cross a massive gulf in these respects could scarcely be either acceptable to its own peoples or credible to a determined and ruthless opponent. Deterrence required making it as hard as possible for any adversary to form the view that NATO would shrink from decisions on raising the conflict’s breadth or severity, or to dare to act on such a view. The range of options available must therefore be an unmistakable continuum without huge gaps. That in turn meant that there had to be nuclear forces, backed by will and doctrine for their possible use, intermediate between conventional forces (NATO had no large offensive CW capability to counter that of the Warsaw Pact) and the ultimate strategic nuclear capability. This was all the more necessary since, especially in earlier days, that ultimate capability often entailed very high explosive yield and low accuracy of delivery, and therefore little ability to act in a controlled and discriminate way.

36

Nuclear Deterrence in NATO From NATO’s inception the judgement was widely accepted that its conventional strength might prove unable to repel, or even to arrest for long, a large-scale and determined attack by Warsaw Pact non-nuclear forces upon NATO territory. The complexities and uncertainties of military conflict have often confounded predictions about their course, but this judgement seemed a prudent inference from the political characteristics and military capabilities of the Soviet Union and its bloc, and from the strategic geography of the East– West confrontation. Moreover, NATO, as a grouping of free and democratic sovereign states, could not espouse plans envisaging the large-scale giving up of any member’s territory for reasons of operational manoeuvre, and therefore had to adopt a posture—‘forward defence’—that was by no means optimal for military effectiveness. Though occasional opinion was to be heard that the judgement of likely defeat at the conventional level was by no means overwhelmingly probable (or that reasonably feasible steps could be taken to make it much less so), it could scarcely be contested that the Alliance’s construction of military response options must at least reckon with the possibility that the judgement might in the event turn out to be right. In brief, therefore, the Alliance had to think hard about what it could do if aggression seemed set to succeed at the conventional or CW level. The only available options, surrender apart, plainly had to envisage the use of nuclear weapons in some way. Reflection, analysis, and discussion however brought home the difficulty of devising such options in any form likely to bring dependable benefit in classical military terms. Besides the overwhelming general fact of virtually inexhaustible destructive power in the hands of the postulated adversary, substantial collective NPG studies of possible scenarios brought out—in retrospect, unsurprisingly—a more particular conclusion. If, finding itself losing in a conventional conflict in Europe, NATO were to use nuclear weapons for military effect and the adversary then responded more or less symmetrically, it must be expected—with all due allowance for the uncertainty in such evaluations—that NATO would still find itself losing, save perhaps in one or two narrowly specialized scenarios. (One such scenario, for example, was a hypothetical amphibious assault by Warsaw Pact forces upon the Danish territory of Jutland. A well-directed yet limited NATO nuclear strike could in theory have left the aggressor without enough suitable shipping to continue in the same mode.) In other words, NATO could not count on its nuclear weapons to

37

Nuclear Deterrence in NATO substitute military victory for military defeat. Recognition of this conclusion infused the work of the NPG for most of the later years of the cold war. It did not follow, however, that NATO’s nuclear weapons could have no value in averting the consequences of prospective defeat at lower levels of force. The studies simply cast into much sharper relief the underlying truth of the nuclear revolution already discussed in Chapter 1. NATO nuclear doctrine had to concentrate upon the idea of using weapons—on a limited scale, and therefore leaving the adversary still with much to lose—to convey effectively to him the message that he had mistaken NATO’s political tolerance and underrated NATO’s will to resist, and that for his own survival he must therefore back off. As pages 10–11 have noted, imparting that message is, short of the ultimate strategic strike, the only use of nuclear weapons that is rationally available, whether in NATO’s particular situation or in any other confrontation between substantial nuclear powers. It is moreover basically as true (see page 102 below) of second as of first nuclear use. The NPG mostly did not focus upon issues about any ultimate strike, although NATO military planning did cater—sometimes rather artificially, given the plethora and weight of forces available from US homeland bases and maritime forces—for the use of NATO-declared nuclear forces based in Europe at that final stage. Much NPG work centred upon questions of targeting principle to fulfil this basic concept of use for war termination. (It did not attempt detailed target selection—that was left to military operational planners.) The purpose would inherently be to convey a political message and induce a political response. It was generally accepted, however, that achieving this would require action with some substantial material effect going beyond just the shock inherent in any nuclear action, severe though that shock might be, especially in first use. ‘Demonstrative use’—the detonation of a weapon over the open sea in the Baltic, say, rather than against any real target—was occasionally canvassed, and the option continued to be theoretically recognized, but it found little real support. It was judged, surely rightly, that this might well suggest to the adversary precisely a lack of the toughminded resolve that it would be the whole purpose of the action to emphasize. The characteristics of the action to be taken would have to be judged—inescapably amid uncertainty, and with acute risk in a

38

Nuclear Deterrence in NATO situation of great tension—along several dimensions. These would have to include the number of nuclear strikes, their explosive yield, the character and location of the particular targets chosen, the likely scale of collateral damage, and the nature and penetrativity—the assurance of getting through defences—of the delivery systems to be used. The heart of the judgement required would be to find the right balance between doing too little to drive home the message and doing so much as to provoke a ferocious reaction in rage, spasm, or misunderstanding of the purpose. No neat formulae were or could have been found, but there was progressively established a salutary checklist of factors to be weighed, and some particular considerations were underscored. It was mostly thought that targets should preferably, though not with absolute necessity, be military ones with some bearing upon the hostile operations in progress, so that the aggressor could not immediately sustain those operations unchecked but would be compelled at least to pause and address fresh and dangerous decisions. Misinterpretation of this idea may have contributed, at least indirectly, to the intermittent complaint that NATO concepts amounted to a ‘war-fighting’ doctrine foolishly directed to seeking military victory. But mature NATO targeting concepts were war-fighting ones only in the very limited sense, sharply distanced from classical war-winning, just outlined. Tensions—again inherently unresolvable in any tidy, definitive, or uniform manner amid the Alliance’s complexities—were regularly apparent between the grave objections to choosing targets within NATO’s own overrun territory (or, to a lesser extent, non-Soviet Warsaw Pact territory) and the perceived retaliation-provoking dangers of choosing targets within the Soviet Union itself. During the later years of NPG work the former consideration rightly came to bulk increasingly large. But even if targeting were to embrace Soviet territory there was no suggestion that it should, or could with any hope of success, be directed to eliminating Soviet ability to strike back. The reality of ultimate capability on both sides for ‘mutual assured destruction’ (MAD) was not challenged. That reality too was occasionally misinterpreted, as though it were somehow NATO’s preferred strategy. MAD capability was not a strategy. It was seen as an inescapable fact of life which simply had to be acknowledged and lived with, and around which, for good or ill, any strategy had to be built.

39

Nuclear Deterrence in NATO Some argument sprang up in the later years of the cold war, and for a time after it, about whether or not NATO doctrine did, or should, envisage the use of nuclear weapons only as a last resort. This debate seemed largely artificial; it was far from clear that differences of substance were bound to underlie it. NATO always viewed the use of nuclear weapons as a last resort, in the sense that so grave a step would be taken only if no other course was judged likely to meet the need—that is, to avert defeat. In a developing conflict that juncture might, if it came at all, come quickly or it might come slowly. The appropriate sense of ‘last’ was not primarily a temporal one, implying that every other conceivable option had first to be extensively tried out almost regardless of practical judgement about its chances of success. (A similar misconstruction of ‘last resort’ was to be found in debates during the run-up to the 1991 operations to expel Saddam Hussein’s forces from Kuwait.) The proper sense was ‘least to be preferred’ and ‘to be adopted only if nothing else can be expected to serve the necessary purpose’. German colleagues in the NPG framework were wont to cite an aphorism: ‘As late as possible, as early as necessary’. That captured the essence of the NATO concept.

Armouries In a wholly tidy, logical, and predictable world nuclear armouries, like any other military capability, would be developed, configured, and sized to fit previously determined doctrines of possible use. But for both good and bad reasons military provision is often not exactly like that. At least by the 1970s there seemed a wide mismatch between the concepts and doctrines outlined above and the nuclear inventories which had been built up during the 1950s and 1960s (especially in the categories customarily termed ‘non-strategic’—that is, of less than deep reach into the homelands of the two superpowers). Though Robert McNamara later became a passionate critic of nuclear armouries, during his tenure of office as US Secretary of Defense the segment of the US inventory deployed in Europe (aside from that held in the United States or aboard naval vessels) comprised at its peak over 7,000 warheads assigned to a wide range of delivery vehicles provided and operated by various Alliance members, including several European countries. It was always hard to see what practical and coherent strategic ideas could underlie this remarkable

40

Nuclear Deterrence in NATO technology-driven panoply. NATO subsequently found little difficulty (other than sometimes from the temporary tactical constraints imposed by arms-control negotiations—see page 95) in making a series of substantial reductions. The practicalities of military logistics and operations mean that planning for a dependable capability, in a diverse range of scenarios, to mount up to some possible total number of strikes will normally require an inventory of far more than that total. Even so, however, there can be no doubt that an ample range of options to fulfil the doctrines of possible use evolved by the NPG did not need anything like 7,000 warheads. The successive reductions in warhead numbers were matched—in some instances indeed preceded—by reductions in the number of categories of nuclear weapons and delivery systems. At one time or another the United States possessed, and often declared to NATO, nuclear systems for forward-infantry use, 2 for artillery use at various ranges, 3 for air-to-air use, 4 for anti-ballistic-missile (ABM) use, 5 as land-mines, 6 as depth-charges, 7 and in anti-submarine missiles and torpedoes, in addition to missiles 8 or aircraft-delivered weapons for striking targets well behind the assumed front line and back into Soviet territory itself. The progressive pruning of holdings in these categories, and the complete abandonment of some of them, reflected several influences. These influences included the political difficulty that certain of them—notably nuclear landmines (‘atomic demolition munitions’)—could scarcely be used other than on or over friendly territory; the increasingly evident fact that some were 2 The M-388 Davy Crockett nuclear-headed missile—fired from a tripod-mounted recoilless rifle—was operational with US infantry units in the 1960s. It was sometimes wryly described as ‘the only weapon in history with a lethal radius greater than its range’. 3 US nuclear-warhead shells were available for 280-mm, 8-in., and 155-mm howitzers. 4 Some US and Canadian interceptor aircraft carried Genie nuclear-headed (but unguided) anti-aircraft missiles from the late 1950s into the 1980s. 5 The US Nike-Zeus ground-based ABM system, from which the Spartan and Sentinel ABM systems were derived, carried nuclear warheads (as Soviet ABM systems did, and apparently still do in Russian service). 6 In the NATO–Europe environment, however, atomic demolition munitions posed severe problems over their prospective emplacement within NATO territory, especially in the FRG, and moreover in forward locations where for full operational utility they would probably have had to be used early in the expected circumstances of conflict (i.e. resisting a Warsaw Pact land invasion). 7 In addition to US deployments, UK warships carried variants of the UK’s WE 177 family of ‘non-strategic’ nuclear weapons for depth-charge use against submarines. 8 Ground-to-ground nuclear missile systems included at one time or another the Corporal, Honest John, Sergeant, and Lance short-range ballistic missiles, and later the Pershing I and II MRBMs and the Tomahawk GLCM.

41

Nuclear Deterrence in NATO relevant essentially to the military-war-winning aspiration which NATO had come to recognize as unrealistic; and the fact that advances in target-acquisition and similar technologies progressively undercut any case for using nuclear explosive power to compensate for poor accuracy of delivery. There remained nevertheless other considerations which suggested that the characteristics of the armoury as a whole could not be determined exclusively by the nature, number, and location of the targets to be put under threat. One such consideration held that non-strategic war-termination strike (or ‘pre-strategic’ or ‘tactical’ or ‘theatre’ strike—as page 18 has noted, the terminology never quite settled upon a uniformly accepted usage differentiating neatly among these) ought to be carried out by delivery systems evidently separate from the strategic ones that were seen as posing the ultimate threat of massive attack upon the Soviet heartland. An argument sometimes advanced for such a contention ran that unless this distinction was respected the Soviet leaders might mistakenly interpret the action as just the initial salvo in that ultimate assault, and might accordingly rush needlessly and calamitously, perhaps on use-or-lose grounds, to launch their own full-scale power. I was always sceptical of the weight of this argument. NATO would undoubtedly accompany any ‘wartermination’ nuclear action by a major effort in explicit communication to convey what its purpose was and was not. ‘Communication of Nuclear Intentions’ was indeed a specific theme of study in the NPG’s work. The Soviet leaders for their part would surely judge the action (as would NATO ones were the situation reversed) by the reality of how many and which targets were hit, not by the nature of the delivery system. They would know that all-out exchange meant their own country’s utter ruin, and would recognize the appalling dangers of precipitating it rashly or needlessly. They would know also that the scale and character of their own huge and diverse armoury meant that they need not fear being pre-emptively disarmed. There were however, in this non-technical category, further limiting considerations of greater weight. The collective nature of the Alliance meant that there was significant value in having other members share with the United States the political, material, and moral burdens of nuclear effort and so acquire a correspondingly enhanced understanding of policy and right to influence it. In practice the most substantial route available for achieving this, aside from participation

42

Nuclear Deterrence in NATO in the work of the NPG, was to involve as many of them as was politically feasible in the hosting, provision, and operation of Europebased delivery systems for which planning envisaged a genuine role. To have assigned nuclear-strike responsibilities exclusively to systems owned and operated entirely by the United States, however numerous, capable, and invulnerable, would have neglected this important dimension. It was also conjectured, rather more debatably but not trivially, that in the scenario of a massive Warsaw Pact assault upon Western Europe Soviet leaders might attach an extra degree of credibility to the likelihood that NATO nuclear systems based there rather than off-shore or on the far side of the Atlantic would actually be used. These were among the considerations shaping the plans for INF modernization which generated a notably sharp test of the Alliance’s collective political resolve, both internationally and domestically, at the end of the 1970s and through much of the 1980s. These plans, entailing US deployment in Europe of two new nuclear-weapon delivery systems—the Pershing II medium-range ballistic missile (MRBM), of longer reach than the Pershing I already in place, and a ground-launched version of the Tomahawk cruise missile (GLCM)—were carried through in the face of substantial domestic protest in the countries accepting stationing, though the deployment itself was short-lived as the easing of cold war tensions led in 1987 to the conclusion of the US/Soviet INF Treaty prohibiting all missile systems in this range category. As the reasoning just outlined illustrates, the usability of weapons, as judged likely to appear to the adversary, was constantly and properly in the minds of NPG participants. Almost inevitably, the idea sometimes gave rise to difficulties in public presentation. Critics of NATO policies were wont to assert that planned usability betokened a positive inclination to use, and indicated a mindset disposed to regard the waging of nuclear war as in some ordinary sense acceptable. As pages 26–7 have noted, argumentation of this sort misunderstands the working of deterrence. If taken to its logical conclusion, it would lead to a much less credible armoury, more precarious deterrence, and so greater risk of war, including ultimately nuclear war. The tensions in public understanding were nevertheless substantial, as was shown by their sharpest manifestation, when in the late 1970s a gross miscarriage in initial presentation—encapsulated by the media in the wholly misleading term ‘neutron bomb’—played a large

43

Nuclear Deterrence in NATO part in preventing NATO from deploying in Europe US enhancedradiation warheads (ERW). Nuclear explosions, all of which must involve the action of neutrons, generate energy in three main forms—blast, heat, and radiation. Advanced warhead design can modify the relative proportions of these within the total output. For action against heavily armoured targets like tanks the key destructive element is radiation, attacking the crews. In a standard warhead its radius of effect is the shortest of the three, so that heat and blast reach more widely than the direct military purpose would require in attack on armoured forces. An ERW would be one specially designed to provide, within a given total output of energy, more radiation effect and less unwanted (collaterally damaging, whether to friendly forces or to non-combatants) blast and heat. Against the background that a massive Soviet preponderance in offensive armoured forces was a constant concern of NATO military planners, the positive merits of such weapons in deterrence logic were scarcely disputable, even if their importance may have been overrated and the general political drawbacks of introducing a new class of weapons underrated. But the absurd caricature ‘Attacks people while leaving property untouched’ gained some campaigning leverage. That episode may prompt in retrospect a more general reflection. As I look back from the comparative comfort (in this regard) of the twenty-first century, I would be ready to recognize that in the pursuit of ideal logic and maximum assurance some aspects of NATO nuclear planning may have slipped into over-elaboration and over-insurance, insufficiently recognizing—in company with many and various opponents of NATO and its dispositions—the robustness which the huge reality of nuclear weapons conferred upon war-prevention even without extreme refinement. In Windsor, near London, there is a fine town hall with a pillared portico designed by the great architect Sir Christopher Wren. It is recounted that when he first submitted his plans the town corporation complained that there were not enough pillars holding up the roof of the portico. Under protest, he added extra ones; but he made his point, visible to this day, by arranging that the added pillars stop a few inches short of the roof. It may well be that some components of NATO deterrent planning and force provision were Corporation pillars rather than Wren ones. But it cannot be doubted that the portico

44

Nuclear Deterrence in NATO roof needed pillars; and that it was safer with too many than with too few. In brief, NATO came progressively, if not always smoothly, to a realistic and robust understanding of the part nuclear armouries should play in helping to sustain the confidence of its peoples in their security. But there remained among those peoples a significant strand of concern that the possible use of those armouries was so terrible that their possession even for deterrence could never be morally tolerable. The next chapter discusses that concern.

45

5 The Ethics of Nuclear Weapons

The Moral Problem Moral accountability is an integral part of what it means to be human, and no aspect of human affairs can claim exemption from it. The ethical assessment of war, however, has always been an especially difficult and challenging matter, since war entails the deliberate taking of life. But throughout history almost all cultures, whether or not moved by religious belief, have felt unable to condemn absolutely all engagement in it. They have needed accordingly to devise frameworks of evaluation to justify and discipline—and sometimes to forbid—the activity of armed conflict. The most systematically developed such framework is to be found in what is called the ‘Just War’ tradition, 1 originally shaped by Christian thinkers but based essentially not on scriptural or institutional authority but on concepts of natural law entirely accessible to those of any or indeed no religious faith (e.g. almost all the ideas in it can be also traced at one time or another in Islamic thinking 2 ). The tradition has a history stretching back to the fifth century of the Christian era, and there have been occasional attempts to discount it as an antiquated construct of no current utility. Such attempts fail to recognize its character as not a fixed doctrine but a living stream of practical reasoning which has continued to develop in line with the realities and challenges of new political and technological circumstances. Critics are, moreover, usually found readier to 1 For a short overview see Charles Guthrie and Michael Quinlan, Just War—The Just War Tradition: Ethics in Modern Warfare, Bloomsbury, London, 2007. 2 See for example John Kelsay, Arguing the Just War in Islam, Harvard University Press, 2007.

46

The Ethics of Nuclear Weapons assail it than to indicate what alternative structure of moral analysis they would propose to put in its place. Ideas from the tradition underlie much of the modern international law of war. It has always been the prime intellectual influence upon my own approach to the exceptionally severe ethical problems posed by nuclear weapons. The Just War framework offers two levels of evaluation: one addressing the rights and wrongs of having recourse to war (jus ad bellum), the other addressing those of how war is conducted (jus in bello). The moral acceptability of possessing nuclear weapons is contested at both these levels. At the jus ad bellum level, the criteria evolved in the tradition hold inter alia that war should be resorted to only if there is a reasonable prospect of success, and if the net balance of good likely to be achieved by undertaking it, as compared with the consequences of refraining from it, is reasonably judged to outweigh the inevitable costs. But, say the anti-nuclear critics, in a nuclear war everyone would lose; as US President Ronald Reagan notably said, there would be no winners. This is true in the limited sense that any substantial nuclear exchange is certain to leave both sides worse off than before it started. But that was true also of World War II, yet it was clear then that there were genuine victors in a meaningful sense. The proper test (as the discussion of deterrence in Chapter 3 has explained) is the comparison, so far as our inevitably imperfect predictive ability allows, not of before and after but of alternative prospects—the future if we go to war and the future if we do not. Moreover, as pages 14 and 38–9 have indicated, the use of nuclear weapons could take many forms other than that of wholesale exchange, and escalation to that is far from certain. The possibility does not have to be excluded entirely and for all scenarios that some use of nuclear weapons could serve to deny an aggressor the attainment of intolerable aims without entailing unlimited catastrophe, and so could lead to an outcome that might legitimately be termed successful and worth the costs. And the burden of possible fearsome consequences looms over both sides, not one alone. The risk of escalation cannot reasonably mean that there is a unilateral and unconditional duty upon one side to offer no adequate resistance, or no deterrent prospect of such resistance, to a nuclear-armed opponent. I regard the more difficult issues posed for nuclear-weapon possessors by Just War concepts as being those that arise at the other level of

47

The Ethics of Nuclear Weapons evaluation—jus in bello. The two criteria (which have their parallels in other ethical traditions) raising these problems most sharply are those of discrimination and proportionality. The concept of discrimination holds that ‘innocents’—noncombatants—must not be made the object of deliberate attack. In allout wars like the two World Wars of the twentieth century there was room for doubt and debate about exactly who should be classified as non-combatant—political leaders? civilian armaments workers? the operators of public utilities serving both civil and military purposes? The basic concept was however not in question (the aged, the sick, and the very young undoubtedly qualified as non-combatant, for example) even if applying it sometimes called for difficult linedrawing. And by ‘deliberate’ attack is meant attack in which the harm to non-combatants is positively desired and purposed, as being part of what is thought necessary to achieve the full objective of the attack, and not merely foreseen as an unavoidable and regrettable side-effect to be kept to the lowest possible scale compatible with not frustrating the legitimate military purpose. The concept of proportionality in bello, partnering that of discrimination, holds that actions in war should not be undertaken if their foreseen and unavoidable side-effects—the collateral harm that would be done to people not properly part of the military target— are reasonably judged to be an excessively high price to pay for the benefit that the action is expected to yield in the legitimate pursuit of the proper objectives of the war. Within this framework of analysis, the central problem posed by nuclear weapons has concerned the difficulty of using their monstrous destructive power in ways that would not cause huge loss of life to non-combatants. There are conceivable uses of nuclear weapons that need not inevitably cause massive collateral harm. Examples are use in depth-charges against submarines, or in exo-atmospheric interception of ballistic missiles or military satellites. But there is no possibility that an effective structure of war-preventing deterrence could rest on the threat solely of uses like these. In practice, the nuclear contribution to deterrence has to be built on the threat of use against targets on land; and virtually no such use could hope to avoid heavy loss of life among non-combatants. In face of this grave ethical conundrum there are, within the Just War framework of ethical principle and so long as nuclear armouries continue to exist, three potential stances available to be adopted by

48

The Ethics of Nuclear Weapons those concerned to act morally—and in ultimate essence only three, though detailed variants in their application to particular situations are possible:

r Position I maintains that it is morally imperative always and unconditionally to renounce the possession of nuclear weapons, regardless of circumstances and consequences.

r Position II maintains that the use of nuclear weapons must be always and unconditionally renounced, regardless of circumstances and consequences, but that the possession of them for deterrence—that is, with the aim of preventing war—can be morally tolerable if it is reasonably judged effective and necessary for that aim.

r Position III maintains that some use of nuclear weapons, in ways and on a scale the prospect of which could provide effective deterrence, might in extreme circumstances be morally tolerable; and that their possession for war-prevention can therefore be legitimate. Every one of these three positions faces difficulties so severe that, taken in isolation, they would normally have to be regarded as conclusive against it. But there is, in inexorable logic, no fourth position available. The internationally distinguished Quaker Sydney Bailey commented to me at the beginning of the 1980s that there was no policy about nuclear weapons that did not pose appalling moral and practical dilemmas. The ethical analyst concerned to identify a basis for responsible action (as distinct from hand-wringing about the tragic intolerability of the predicament) has no alternative but to choose one of these positions, deciding which of them poses least difficulty. It follows from this that in any honest debate about the choice it is not legitimate to focus exclusively upon the objections to the positions rejected. The objections to the position preferred must also be candidly acknowledged and tackled. All this is a further reflection of the perplexities and dilemmas set by the reductio ad absurdum of war which the advent of nuclear weapons generates. From another angle, it is an additional illustration of the problems which massive and accelerating advances in humanity’s scientific knowledge and technological skill can pose for long-established systems of moral evaluation, by creating practical

49

The Ethics of Nuclear Weapons possibilities previously beyond contemplation. (The modern medical field is increasingly rich in examples of that.) But this cannot be made a reason for abandoning the effort to bring moral concepts to bear.

The Inescapable Choice The absolutist Position I claims to prohibit the possession of nuclear weapons for all time and in all circumstances. It must therefore accept being tested against awkward scenarios. It cannot escape by claiming, for instance, that such scenarios will not materialize if courses of action have previously been chosen wisely. History demonstrates that human affairs can go gravely amiss, and decision-makers responsible for the security of their peoples cannot absolve themselves from facing up to hard practical questions on the ground that these ought not to have arisen. Position I has to be prepared to say that even if it had been known during World War II that Nazi Germany was succeeding in its efforts to develop nuclear weapons, it would have been the unconditional moral duty of Churchill and Roosevelt not to try to match that capability. It has to be prepared to hold that during the cold war it was the unconditional moral duty of the West to discard all its nuclear weapons even if the Soviet leaders refused to do so, kept all their own nuclear weapons, and were thereby in a position to impose their will by overwhelming force or the threat of it right across Western Europe or anywhere else they wanted. It has to be prepared to maintain that it is the unconditional duty of Israel to give up the nuclear weapons which everyone knows her to possess even if hostile countries around her, like Iran saying in the words of President Ahmedinejad that she ought to be wiped off the map, were to come to have massive military preponderance including nuclear weapons. These examples bring out the fundamental difficulty for Position I. It amounts in the end, in face of any adversary who possesses nuclear weapons and may have the will to use them, to full-blown pacifism—to forswearing recourse to arms and indeed to any other form of resistance, however appalling the character, past record, and present actions of the adversary may be. As even the deeply embattled Japanese leadership was forced to recognize in August 1945, no lesser type of force can hope to stand against nuclear weapons. There

50

The Ethics of Nuclear Weapons have been efforts (further discussed on pages 104–5) to portray passive dissent or non-cooperation of one kind or another as potential alternatives to complete submission, on the pattern for example of Mahatma Gandhi’s leadership of opposition to British rule in India between the two World Wars. But that rule, for all its faults, was not that of Hitler or Stalin either in inhumanity or in willingness to enforce obedience brutally. There can be no reality in hopes that non-violent resistance can offer a universally effective answer to the problem of how to protect peoples and keep them free in the face of nuclear weapons in hostile hands. Position II—‘use never allowable, but possession for deterrence potentially permissible’—has obvious having-it-both-ways appeal. It is however fundamentally incoherent, a claim to square a circle. As pages 25–6 have brought out, deterrence and use cannot be totally disconnected. Deterrence cannot exist if there is no possibility of use. Position II inescapably envisages the deliberate creation of a capability which it maintains must never be used; and at the same time it makes a renouncing statement about use which it has to hope will not be entirely believed. The whole concept must amount to the expression of a massive untruth: either the preparation is bogus or the renunciation is insincere. The nature and positive intent of each component is to convey a message directly contrary to that purportedly given by the other. The concept seeks deliberately to imply what it explicitly denies. Such a stance can scarcely be accepted as solving a profound moral dilemma. It tries to avoid the bitter choice between abandoning deterrence and allowing nuclear use to be potentially legitimate. The attempt is understandable, but it fails. Position II simply falls apart under scrutiny. (In a major Pastoral Letter of 1983 3 the Catholic Bishops of the United States, having clearly joined Pope John Paul II 4 in ruling out Position I, appeared on some interpretations of their words to incline towards Position II; but they plainly rejected it—in

3 ‘The Challenge of Peace: God’s Promise and Our Response’, Pastoral Letter of the US Catholic Bishops Conference, May 1983, reproduced in Origins, Vol. 13, No. 1. 4 Pope John Paul II, Message to the U.N. Second Special Session on Disarmament, 11 June 1982: ‘In current conditions “deterrence” based on balance, not as an end in itself but as a stage on the way toward a progressive disarmament, may still be judged morally acceptable.’ (He did not either then or later indicate whether his view rested on what I have termed Position II or Position III.)

51

The Ethics of Nuclear Weapons favour of Position III—in a clarification published five years later. 5 ) There is no way of evading the hard choice between Positions I and III. The central difficulty for Position III is to explain how it could ever be consistent with the imperatives of discrimination and proportionality to use the appalling destructive power of nuclear weapons in ways that would be certain to inflict heavy non-combatant deaths. Espousing this position is far from entailing approval of everything that has been done or contemplated over the years in the West, let alone elsewhere, to provide deterrence in the nuclear age. By about half-way through the cold war the armouries on both sides had become grotesquely excessive. There can be little doubt, in retrospect, that at least some of the contingency plans formulated for their potential use on a massive scale would have been morally intolerable to execute in any circumstances whatever. But the uses that effective deterrence must be ready to contemplate do not have to be directed, even at a stage of final strategic strike, to hitting populations and exterminating cities (whether or not we believe that an adversary may do that). There could and should be devised ways of dealing an agonising blow to an aggressor regime—with the purpose not of revenge but of ending or reducing the risk that it might feel able or disposed to continue or repeat its offence—other than simply by devastating its cities. When the UK government decided in 1980 to acquire the Trident missile system it said in its explanatory document 6 that it wanted to be able to pose a credible threat to ‘key aspects of Soviet state power’. Those words were deliberately intended not to mean crude counter-city strike. The United States has given similar indications, 7 and from its 1994 Defence White Paper onwards France (which avowedly had at one time a counter-city concept) has altered its doctrine to essentially the same effect. 8 Appendix 2 reproduces a published statement which I wrote in 1984 on the problem of whether or how any use at all of nuclear weapons could ever be morally tolerable. The environment in which 5 ‘A Pastoral Reflection on the Response to “The Challenge of Peace” ’, 21 July 1988, reproduced in Origins, Vol. 18, No. 9, p. 137. 6 Defence Open Government Document 80/23, July 1980, paragraph 12. 7 See for example letter of William P. Clark, National Security Adviser, to Cardinal Joseph Bernardin, 15 January 1983, cited at paragraph 179 of the US Catholic Bishops’ 1983 Pastoral Letter. 8 See Bruno Tertrais, Debating 21st Century Nuclear Issues, ed. Price and Maltby, CSIS, Washington, DC, 2007, p. 122.

52

The Ethics of Nuclear Weapons it was composed was that of the cold war, but I would still stand by the general structure of the argument it sets out. In brief, any substantial use of nuclear weapons, however carefully directed to legitimate targets, must be expected to result in fearful numbers of deaths, including many people who must indisputably be categorized as non-combatants. That would have to be weighed very gravely in the assessment of whether the terrible harm could be a proportionate price to pay for denying victory to the adversary. But assessing proportionality is a matter of evaluating not one magnitude only but the relationship between two. The other magnitude might have to be submission to a Hitler. The outcome of the assessment does not have to be prescribed beforehand solely as a prohibiting negative for every conceivable scenario. Moreover, as pages 25–6 have explained, deterrence does not require that we have an implacable determination to use our weapons near-automatically. It requires no more than that the adversary perceive a real possibility that we will use them if we are made desperate. That in turn requires no more than that in our own thinking and planning we should not have entirely ruled out the possibility of use, while recognizing also and always that should the awful moment for decision ever come, the right moral and prudential decision in the particular situation might be not to use them, whatever the provocation. All this is, without question, a hard argument to sustain—it is at full stretch. But that is true, at some point or another in the analysis, of that for any other stance. The argument defending Position III is in my view less difficult than the inescapable alternative. That alternative—Position I—has to hold that right-thinking countries are under an absolute obligation, for the rest of human history and irrespective of whatever lessons future experience may yield, not to counter-balance nuclear weapons even if they are in the hands of a Hitler, a Stalin, or a Pol Pot. The rejection of that alternative is in line with the underlying approach of the Just War tradition. That approach has always judged, as a matter almost of moral common sense, that it simply could not be right, especially for public authorities charged with the well-being and protection of their peoples, to interpret the powerful message of concern for all humanity conveyed in the Christian scriptures and their like as meaning that no effective resistance should be offered to Attila, Genghis Khan, or Hitler. Such an approach is, if

53

The Ethics of Nuclear Weapons anything, reinforced in the presence of the reductio ad absurdum of warfare between major developed states created by nuclear weapons. Key governments during the cold war increasingly judged that careful policies of deterrence involving the possession of nuclear weapons, and thus exploiting the reductio ad absurdum, rendered any war between East and West (and any actual use of the weapons) highly unlikely. Not everyone agreed with that judgement, or does so even now that we know that the cold war never became a hot one, but it was in any event a matter of practical appraisal, not moral analysis. Moral analysis must therefore be ready to reckon with the possibility that the judgement was right (as most strategic commentators have continued to believe) and also that it may remain sometimes relevant in current world circumstances. If that is so, it does not absolutely and automatically follow—or anyway not for ethicists whose approach is not purely consequentialist—that Position I must be wrong. It does mean, however, that a massive burden of demonstrating inevitability must rest on the shoulders of anyone who seeks to maintain that that position is inescapable and that Position III is untenable. All that said, it cannot be right to acquiesce uncritically, for the rest of human history, in a system that maintains peace between potential adversaries partly by the threat of colossal disaster. 9 Any serious moral stance must recognize a duty, alongside that of striving to reduce the costs and risks of nuclear armouries and maximizing their war-prevention benefits so long as they continue to exist, to work towards a world in which security can be maintained without incurring at all the burdens which they entail. (Chapter 12 examines this further.) To espouse Position III as the right basic ethical analysis for the choices confronting states in the circumstances of today is in no way incompatible with such a stance, or with believing—for example—that in specific circumstances it is no longer necessary or justifiable for a particular nation, or indeed any nation, to continue to accept the responsibilities, penalties, and dangers of having nuclear weapons. But that would be a judgement about feasibilities, consequences, and costs. Such considerations have indeed a moral dimension, but they cannot purport to settle difficult questions by 9 See the acceptance of this in the final paragraph of the UK government’s 1981 statement reproduced in Appendix 1.

54

The Ethics of Nuclear Weapons reference to a categoric moral imperative overriding all practical and prudential appraisal. In brief, amid the difficult realities of the world the possession of nuclear armouries cannot reasonably be dismissed as always intolerable in ethical terms. But there remains a grave moral duty both to take seriously as a long-term vision the possibility of eventual complete escape from their shadow and meanwhile to manage their continuance with as little risk and at as low a cost as possible. Chapter 12 in Part IV will discuss the former component of that duty. Part II now considers the latter and more immediate component.

55

This page intentionally left blank

Part II Managing Nuclear Weapons

This page intentionally left blank

6 Risks

Instability The concept of mutual deterrence requires the reciprocal cancellation of options for war at any level between advanced powers. These powers have to accept—counter-intuitively, by past standards of security appraisal—that they must acquiesce in their own vulnerability. Through most of the nuclear era it has been a concern increasingly shared on all sides that this concept, as the best available underpinning of peace amid political disagreements and in the presence of fearsome destructive potential, should be stably implemented by the key countries in the international system. The stability needed has been of two main kinds which are not always sufficiently distinguished. The first concerns deterrent efficacy in normal times—that is, the best practicable assurance that actions or other events are not likely to imperil the war-preventing balance by seeming to call into doubt the effective cancellation of options on both sides. The second concerns how to keep to a minimum the risks, if in time of crisis or conflict peace-time deterrence has faltered or failed, that (as is customarily believed to have happened with European mobilization and transport schedules in 1914) operational, logistical, or political features of the situation may work against taking decisions on adequate information and after proper consideration, with time available for communication and potential negotiation. Such risks could increase dangers of misjudgement precipitating an appalling conflict that might have been avoided. During the cold war there were occasional surface disturbances— whether actual or threatened—of the military balance, and of the

59

Risks confidence in it that underpinned deterrence. The Soviet missiledeployment adventure in Cuba in 1962 was a vivid example, and different viewpoints might suggest others, such as concern—albeit not, so it seemed, shared by General Secretary Gorbachev—that President Reagan’s 1983 aspiration for perfect anti-missile defence (see pages 109–10) might undermine Soviet deterrent sufficiency. But at no stage—at least after the Soviet Union’s acquisition, by the late 1950s, of a substantial inter-continental delivery capability— was the ultimate cancellation of options at real risk. Military staff on either side could never even remotely have hoped (to borrow Professor Colin Gray’s phrase 1 ) to brief their political leaders with a plausibly beneficial plan for launching attack even if, improbably, either leadership had been minded to wager national survival upon the accuracy and reliability of untested calculations and projections about enormous events reaching far beyond any available experience. It used occasionally to be suggested, during the cold war, that the ‘nuclear arms race’ threatened the basic stability of the deterrent structure. Pages 88–9 later comment on the supposition of such a race. Briefly, however, the facts bore it out in only a very limited sense. It was virtually impossible (and by no means universally or necessarily desirable) to freeze technological advance; but to suppose that this automatically undermined stability was to confuse ‘stable’ with ‘static’. Given the size, diversity, and security of armouries and the prodigious destructive force which they commanded, the margins of safety within the balance of the cold war stand-off were enormous in terms of operational calculus. To postulate that one bloc might outstrip the other decisively, having regard to what ‘decisively’ would have had to mean in order to have practical value for an aggressor, would have been like postulating that in an Olympic two-lap race one competitor might lap the rest of the field. Nuclear stability between the United States and the Soviet Union (and now Russia) could readily assume, given the massive size, wide diversity, and technical sophistication of the armouries, that two key underlying conditions were met. The first was that each side should be able to feel assured that the other could not by defensive measures, pre-emptive attack, or a combination of these reduce the weight of retaliatory strike power to a level that could tolerably be suffered. 1 See Professor Colin S. Gray, Strategy and History: Essays in Theory and Practice, Routledge, Abingdon, 2006, p. 92.

60

Risks The second condition, less often noted, was that the strike power likely to survive such an attack should still suffice to provide a choice of credible options for the form and scale of any nuclear response at subsequent stages. The point of this second condition was that, if it were not met, the state attacked might then find itself in the deterrence-weakening dilemma that may face any nuclear possessor lacking the flexibility to do anything short of precipitating all-out final holocaust. States with nuclear armouries markedly smaller than those of the two biggest possessors cannot take the satisfaction of these two conditions of peace-time stability in deterrent balance so readily for granted. Those of France and the United Kingdom, however—based, as they are, largely or wholly on advanced long-range submarinelaunched ballistic missiles (SLBMs)—have the necessary combination of weight and other characteristics such as near-invulnerability to pre-emption. That may well be true also of Israel, though such information as is available publicly—officially none—does not make it possible to judge with certainty. The evolving balance in South Asia between India and Pakistan, though not precarious, may not yet have fully reached this position on both sides. (The South Asian situation is further discussed in Chapter 11.) The position of China may be more open to doubt. It is not yet entirely clear that the United States accepts a relationship of mutual deterrence with China, and Chinese voices sometimes express unease about the aims or effects of US plans for ballistic-missile defence (BMD) and for long-range conventional strike capability. This USA/China issue exemplifies a problem that may arise more widely. The United States, in particular, is reluctant to acquiesce in a possible relationship of mutual deterrence with North Korea and Iran, which it regards as ‘rogue’ states, not trustworthy participants in the global order. It seeks accordingly, by the deployment of BMD systems, to remove or attenuate the ability of such states to strike home with nuclear weapons. But planners in other countries such as Russia and China, wary about the future capabilities of such systems following US withdrawal from the constraints of the 1972 ABM Treaty, can have no physical guarantee that the capabilities will not come to bear upon their own striking power, despite US declarations that that is not the intention. The need to recognize and manage this potential dilemma for deterrent stability and confidence may arise also in South Asia, as pages 142–4 will note.

61

Risks The needs of stability in time of crisis or conflict are of a different though not unrelated kind. What is then required of military forces and dispositions is that they should be so configured as to present to decision-makers on either side the minimum temptation or provocation to move swiftly to the use of nuclear force, whether to seize advantage or to forestall disadvantage, and the maximum incentive to hold back. The aim is that crucial decisions should to the greatest extent possible be taken—even amid the confusion, stresses, and damage of battle—in a measured way; by the top political authorities; and in the fullest possible understanding by each side of what the other is thinking and doing. All this is not peculiar to the nuclear age. The nuclear reality poses, however—again, as two sides of one coin—both a heightened imperative to achieve it and (because the incentive is both shared and compelling) heightened prospects of doing so. That is illustrated by much that was done during the cold war by way of precaution, both unilaterally and through East/West agreement. Though conjecture has occasionally suggested that there were from time to time severe perturbations or near-misses, 2 the practical record, in the round, is reassuring. (Fears about accident and miscalculation are reviewed later in this chapter.)

Escalation The term ‘escalation’ refers to the familiar general fact that in situations of competition or conflict actions by one side are apt to induce reactions by the other in order to recover advantage or redress disadvantage, and that in war this process may progressively raise the intensity of fighting. But the new particular fact that the nuclear discovery had stretched out the spectrum of possible force to the extreme of destructive apocalypse, especially between combatants with the colossal power and resources of the cold war adversaries, gave the idea of escalation a special importance. The possibility of escalation arises as soon as fighting starts, not just when nuclear use starts. The customary focus of concern, however, was primarily 2 For example, there have been suggestions that in 1983 some elements of the Soviet system worked themselves into a high state of alarm lest the NATO command-post exercise code-named ABLE ARCHER prove to be intended as cover for a massive surprise attack. But on the NATO side there was never the highest such idea or any actual movement of major nuclear-weapon takes in abnormal ways, and so no likelihood of concrete NATO actions that could credibly have triggered Soviet nuclear response.

62

Risks upon what might happen if nuclear weapons began to be used. Was nuclear retaliation certain? And could the process of response and counter-response ever be halted short of all-out world-destroying holocaust? Two points about these questions should be recognized at the outset. The first is that we cannot know the answers for certain. Anyone who asserts or implies that we can be sure or nearly sure cannot be on firm ground. Nor can we measure the probabilities neatly. No one knows how political leaders and armed forces will react in the unprecedented situations in question. Escalation is neither a physical process like a chemical chain-reaction nor a sequence of random events like outcomes on a gambling machine. It is a matter of interactive choices by people. It has to be considered therefore in human and political terms, not just as a matter of military or technical mechanics. The second point is that the emergency could arise in a wide variety of ways and settings. Assertions claiming uniform predictive authority throughout the range of possibility are very unlikely to be well-founded. So too, a fortiori, are deductions and evaluations purporting to rest on them. There are good reasons for fearing escalation. These include the confusion of war; its stresses, anger, hatred, and the desire for revenge; reluctance to accept the humiliation of backing down; the desire to get further blows in first. Given all this, the risks of escalation are grave in any conflict between advanced powers, and Western leaders during the cold war were rightly wont to emphasize them in the interests of deterrence. But this is not to say that they are virtually certain, or even necessarily odds-on; still less that they are so for all the assorted circumstances in which the situation might arise, in a nuclear world to which past experience is only a limited guide. It is entirely possible, for example, that the initial use of nuclear weapons, breaching a barrier that has held since 1945, might so horrify both sides in a conflict that they recognized an overwhelming common interest in composing their differences. The human pressures in that direction would be very great. Even if initial nuclear use did not quickly end the fighting, the supposition of inexorable momentum in a developing exchange, with each side rushing to overreaction amid confusion and uncertainty, is implausible. It fails to consider what the situation of the decisionmakers would really be. Neither side could want escalation. Both would be appalled at what was going on. Both would be desperately

63

Risks looking for signs that the other was ready to call a halt. Both, given the capacity for evasion or concealment which modern delivery platforms and vehicles can possess, could have in reserve significant forces invulnerable enough not to entail use-or-lose pressures. (It may be more open to question, as noted earlier, whether newer nuclearweapon possessors can be immediately in that position; but it is within reach of any substantial state with advanced technological capabilities, and attaining it is certain to be a high priority in the development of forces.) As a result, neither side can have any predisposition to suppose, in an ambiguous situation of fearful risk, that the right course when in doubt is to go on copiously launching weapons. And none of this analysis rests on any presumption of highly subtle or pre-concerted rationality. The rationality required is plain. The argument is reinforced if we consider the possible reasoning of an aggressor at a more dispassionate level. Any substantial nuclear armoury can inflict destruction outweighing any possible prize that aggression could hope to seize. A state attacking the possessor of such an armoury must therefore be doing so (once given that it cannot count upon destroying the armoury pre-emptively) on a judgement that the possessor would be found lacking in the will to use it. If the attacked possessor used nuclear weapons, whether first or in response to the aggressor’s own first use, this judgement would begin to look dangerously precarious. There must be at least a substantial possibility of the aggressor leaders’ concluding that their initial judgement had been mistaken—that the risks were after all greater than whatever prize they had been seeking, and that for their own country’s survival they must call off the aggression. Deterrence planning such as that of NATO was directed in the first place to preventing the initial misjudgement and in the second, if it were nevertheless made, to compelling such a reappraisal. The former aim had to have primacy, because it could not be taken for granted that the latter was certain to work. But there was no ground for assuming in advance, for all possible scenarios, that the chance of its working must be negligible. An aggressor state would itself be at huge risk if nuclear war developed, as its leaders would know. It may be argued that a policy which abandons hope of physically defeating the enemy and simply hopes to get him to desist is pure gamble, a matter of who blinks first; and that the political and moral nature of most likely aggressors, almost ex hypothesi, makes them the less likely to blink. One response to this is to ask what is

64

Risks the alternative—it can only be surrender. But a more positive and hopeful answer lies in the fact that the criticism is posed in a political vacuum. Real-life conflict would have a political context. The context which concerned NATO during the cold war, for example, was one of defending vital interests against a postulated aggressor whose own vital interests would not be engaged, or would be less engaged. Certainty is not possible, but a clear asymmetry of vital interest is a legitimate basis for expecting an asymmetry, credible to both sides, of resolve in conflict. That places upon statesmen, as page 23 has noted, the key task in deterrence of building up in advance a clear and shared grasp of where limits lie. That was plainly achieved in cold-war Europe. If vital interests have been defined in a way that is clear, and also clearly not overlapping or incompatible with those of the adversary, a credible basis has been laid for the likelihood of greater resolve in resistance. It was also sometimes suggested by critics that whatever might be indicated by theoretical discussion of political will and interests, the military environment of nuclear warfare—particularly difficulties of communication and control—would drive escalation with overwhelming probability to the limit. But it is obscure why matters should be regarded as inevitably so for every possible level and setting of action. Even if the history of war suggested (as it scarcely does) that military decision-makers are mostly apt to work on the principle ‘When in doubt, lash out’, the nuclear revolution creates an utterly new situation. The pervasive reality, always plain to both sides during the cold war, is ‘If this goes on to the end, we are all ruined’. Given that inexorable escalation would mean catastrophe for both, it would be perverse to suppose them permanently incapable of framing arrangements which avoid it. As page 16 has noted, NATO gave its military commanders no widespread delegated authority, in peace or war, to launch nuclear weapons without specific political direction. Many types of weapon moreover had physical safeguards such as PALs incorporated to reinforce organizational ones. There were multiple communication and control systems for passing information, orders, and prohibitions. Such systems could not be totally guaranteed against disruption if at a fairly intense level of strategic exchange—which was only one of many possible levels of conflict— an adversary judged it to be in his interest to weaken political control. It was far from clear why he necessarily should so judge. Even then, however, it remained possible to operate on a general

65

Risks fail-safe presumption: no authorization, no use. That was the basis on which NATO operated. If it is feared that the arrangements which a nuclear-weapon possessor has in place do not meet such standards in some respects, the logical course is to continue to improve them rather than to assume escalation to be certain and uncontrollable, with all the enormous inferences that would have to flow from such an assumption. The likelihood of escalation can never be 100 per cent, and never zero. Where between those two extremes it may lie can never be precisely calculable in advance; and even were it so calculable, it would not be uniquely fixed—it would stand to vary hugely with circumstances. That there should be any risk at all of escalation to widespread nuclear war must be deeply disturbing, and decision-makers would always have to weigh it most anxiously. But a pair of key truths about it need to be recognized. The first is that the risk of escalation to large-scale nuclear war is inescapably present in any significant armed conflict between nuclear-capable powers, whoever may have started the conflict and whoever may first have used any particular category of weapon. The initiator of the conflict will always have physically available to him options for applying more force if he meets effective resistance. If the risk of escalation, whatever its degree of probability, is to be regarded as absolutely unacceptable, the necessary inference is that a state attacked by a substantial nuclear power must forgo military resistance. It must surrender, even if it has a nuclear armoury of its own. But the companion truth is that, as page 47 has noted, the risk of escalation is an inescapable burden also upon the aggressor. The exploitation of that burden is the crucial route, if conflict does break out, for managing it to a tolerable outcome—the only route, indeed, intermediate between surrender and holocaust, and so the necessary basis for deterrence beforehand. The working out of plans to exploit escalation risk most effectively in deterring potential aggression entails further and complex issues. It is for example plainly desirable, wherever geography, politics, and available resources so permit without triggering arms races, to make provisions and dispositions that are likely to place the onus of making the bigger and more evidently dangerous steps in escalation upon the aggressor who wishes to maintain his attack, rather than upon the defender. (The customary shorthand for this desirable posture used to be ‘escalation dominance’.) These issues are not further discussed here. But addressing them needs to start from acknowledgement that

66

Risks there are in any event no certainties or absolutes available, no options guaranteed to be risk-free and cost-free. Deterrence is not possible without escalation risk; and its presence can point to no automatic policy conclusion save for those who espouse outright pacifism and accept its consequences.

Accident and Miscalculation Ensuring the safety and security of nuclear weapons plainly needs to be taken most seriously. Detailed information is understandably not published, but such direct evidence as there is suggests that it always has been so taken in every possessor state, with the inevitable occasional failures to follow strict procedures dealt with rigorously. Critics have nevertheless from time to time argued that the possibility of accident involving nuclear weapons is so substantial that it must weigh heavily in the entire evaluation of whether war-prevention structures entailing their existence should be tolerated at all. Two sorts of scenario are usually in question. The first is that of a single grave event involving an unintended nuclear explosion—a technical disaster at a storage site, for example, or the accidental or unauthorized launch of a delivery system with a live nuclear warhead. The second is that of some event—perhaps such an explosion or launch, or some other mishap such as malfunction or misinterpretation of radar signals or computer systems—initiating a sequence of response and counter-response that culminated in a nuclear exchange which no one had truly intended. No event that is physically possible can be said to be of absolutely zero probability (just as at an opposite extreme it is absurd to claim, as has been heard from distinguished figures, that nuclear-weapon use can be guaranteed to happen within some finite future span despite not having happened for over sixty years). But human affairs cannot be managed to the standard of either zero or total probability. We have to assess levels between those theoretical limits and weigh their reality and implications against other factors, in security planning as in everyday life. There have certainly been, across the decades since 1945, many known accidents involving nuclear weapons, from transporters skidding off roads to bomber aircraft crashing with or accidentally dropping the weapons they carried (in past days when such carriage was

67

Risks a frequent feature of readiness arrangements—it no longer is). A few of these accidents may have released into the nearby environment highly toxic material. None however has entailed a nuclear detonation. Some commentators suggest that this reflects bizarrely good fortune amid such massive activity and deployment over so many years. A more rational deduction from the facts of this long experience would however be that the probability of any accident triggering a nuclear explosion is extremely low. It might be further noted that the mechanisms needed to set off such an explosion are technically demanding, and that in a large number of ways the past sixty years have seen extensive improvements in safety arrangements for both the design and the handling of weapons. It is undoubtedly possible to see respects in which, after the cold war, some of the factors bearing upon risk may be new or more adverse; but some are now plainly less so. The years which the world has come through entirely without accidental or unauthorized detonation have included early decades in which knowledge was sketchier, precautions were less developed, and weapon designs were less ultra-safe than they later became, as well as substantial periods in which weapon numbers were larger, deployments more widespread and diverse, movements more frequent, and several aspects of doctrine and readiness arrangements more tense. Similar considerations apply to the hypothesis of nuclear war being mistakenly triggered by false alarm. Critics again point to the fact, as it is understood, of numerous occasions when initial steps in alert sequences for US nuclear forces were embarked upon, or at least called for, by indicators mistaken or misconstrued. In none of these instances, it is accepted, did matters get at all near to nuclear launch—extraordinary good fortune again, critics have suggested. But the rival and more logical inference from hundreds of events stretching over sixty years of experience presents itself once more: that the probability of initial misinterpretation leading far towards mistaken launch is remote. Precisely because any nuclear-weapon possessor recognizes the vast gravity of any launch, release sequences have many steps, and human decision is repeatedly interposed as well as capping the sequences. To convey that because a first step was prompted the world somehow came close to accidental nuclear war is wild hyperbole, rather like asserting, when a tennis champion has lost his opening service game, that he was nearly beaten in straight sets. History anyway scarcely offers any ready example of major war

68

Risks started by accident even before the nuclear revolution imposed an order-of-magnitude increase in caution. It was occasionally conjectured that nuclear war might be triggered by the real but accidental or unauthorized launch of a strategic nuclear-weapon delivery system in the direction of a potential adversary. No such launch is known to have occurred in over sixty years. The probability of it is therefore very low. But even if it did happen, the further hypothesis of its initiating a general nuclear exchange is far-fetched. It fails to consider the real situation of decision-makers, as pages 63–4 have brought out. The notion that cosmic holocaust might be mistakenly precipitated in this way belongs to science fiction. One special form of miscalculation appeared sporadically in the speculations of academic commentators, though it was scarcely ever to be encountered—at least so far as my own observation went—in the utterances of practical planners within government. This is the idea that nuclear war might be erroneously triggered, or erroneously widened, through a state under attack misreading either what sort of attack it was being subjected to, or where the attack came from. The postulated misreading of the nature of the attack referred in particular to the hypothesis that if a delivery system—normally a missile—that was known to be capable of carrying either a nuclear or a conventional warhead was launched in a conventional role, the target country might, on detecting the launch through its earlywarning systems, misconstrue the mission as an imminent nuclear strike and immediately unleash a nuclear counter-strike of its own. This conjecture was voiced, for example, as a criticism of the proposals for giving the US Trident SLBM, long associated with nuclear missions, a capability to deliver conventional warheads. Whatever the merit of those proposals (it is not explored here), it is hard to regard this particular apprehension as having any real-life credibility. The flight time of a ballistic missile would not exceed about thirty minutes, and that of a cruise missile a few hours, before arrival on target made its character—conventional or nuclear—unmistakable. No government will need, and no nonlunatic government could wish, to take within so short a span of time a step as enormous and irrevocable as the execution of a nuclear strike on the basis of early-warning information alone without knowing the true nature of the incoming attack. The speculation tends moreover to be expressed without reference either to any realistic political

69

Risks or conflict-related context thought to render the episode plausible, or to the manifest interest of the launching country, should there be any risk of doubt, in ensuring—by explicit communication if necessary—that there was no misinterpretation of its conventionally armed launch. It may be objected to this analysis that in the cold war the two opposing superpowers had concepts of launch-on-warning. That seems to be true, at least in the sense that successive US administrations declined to rule out such an option and indeed included in their contingency plans both this and the possibility of launchunder-attack (that is launch after some strikes had been suffered and while the sequence of them was evidently continuing). The Soviet Union was not likely to have had more relaxed practices. But the colossal gravity of activating any such arrangements must always have been recognized. It could have been contemplated only in circumstances where the entire political context made a pre-emptive attack by the adversary plainly a serious and imminent possibility, and where moreover the available information unmistakably indicated that a massive assault with hundreds or thousands of missiles was on the way. That was a scenario wholly unlike that implicit in the supposition that a conventional missile attack might be briefly mistaken for a nuclear one. The other sort of misunderstanding conjectured—that of misreading the source of attack—envisaged, typically, that SLBMs launched by France or the United Kingdom might erroneously be supposed to be coming from US submarines, and so might initiate a superpower exchange which the United States did not in fact intend. (An occasional variant on this was the notion that ‘triggering’ in this way might actually be an element in deliberate French or UK deterrent concepts. There was never any truth in this guess in relation to the United Kingdom, and French thinking is unlikely to have been different.) The unreality in this category of conjecture lay in the implication that such a scenario could develop without the US government making the most determined efforts to ensure that Soviet (or now Russian) leaders knew that the United States was not responsible for the attack, and with those leaders for their part resorting, on unproven suspicion, to action that was virtually certain to provoke nuclear counter-action from the United States. There used occasionally to be another speculation, that if the Soviet Union suffered heavy

70

Risks nuclear strikes known to come from France or the United Kingdom, it might judge its interests to be best served by ensuring that the United States did not remain an unscathed bystander. But even if that were somehow thought marginally less implausible, it would have been a different matter from misinterpretation of the initial strike. As was noted earlier in this chapter, the arrangements under which nuclear-weapon inventories are now managed are in several important respects already much less open to concern than they were during much of the cold war. Worries voiced more recently sometimes relate to ‘cyber-attack’—hostile interference, whether by states or by other actors such as terrorists, with information systems used in the control of armouries. It is highly unlikely, though details are (again understandably) not made public, that regular reviews of control arrangements are oblivious to any such risks. Perceptions of them do however reinforce the already-strong case that whatever arrangements still remain in place for continuous high readiness to launch nuclear action at short notice should be abandoned. Chapter 13 returns to this.

Terrorism The contingency of nuclear weapons being brought to bear by covert means, as an extreme form of sabotage, was discussed almost from the outset of the nuclear age. 3 Since the cold war ended, however, and especially since the murderous outrages of 11 September 2001 against the US homeland, there has been much conjecture about the possibility of their use not by states but by terrorists. This possibility needs to be clearly distinguished from that of terrorist use of a ‘radiological’ weapon that would use high explosive to scatter toxic radioactive material. Such an act—much easier to execute, though it is noteworthy that no group seems yet to have done so—could be highly disruptive in an urban environment, and the International Convention on the Suppression of Acts of Nuclear Terrorism, which came into force in July 2007, covers this as well as true nuclear explosions. It could not, however, approach the lethality of the latter. 3 See for example Bernard Brodie in The Absolute Weapon, ed. Bernard Brodie, Harcourt, Brace and Co. for the Yale Institute of International Studies, New York, 1946, pp. 49–52.

71

Risks In almost any conceivable setting a terrorist-caused nuclear explosion would be an immense calamity (and a transforming event politically in the global scene). Opinion among commentators differs widely about its degree of probability. Even those, however—with whom my own judgement inclines to side—who regard the likelihood as slight would share the view that every reasonable step ought to be taken to reduce it to the lowest possible level. To succeed in a nuclear-explosion attempt a terrorist organization would have three main needs:

r to construct or otherwise acquire a warhead; r to find means of delivering it to a desired target area; r to have the will to use it. Risk-reduction efforts have to work on increasing the difficulty and lowering the probability of one or more of these three needs being met. To take them in reverse order: in respect of the third need, it is by no means clear and certain that every terrorist organization would in fact have the will to execute so extreme an act and would think its purposes to be advanced by that, given the certainty of massive and worldwide condemnation and proscription. There is however little that risk-reduction efforts can do, at least directly, to strengthen the likelihood of judgement being made in that abstentionist direction. Optimism about limits to the hardihood and reckless inhumanity of terrorists cannot carry sufficient weight, especially given what is known or credibly believed about the attitude and aspirations of alQaeda and its like, to render unnecessary the best possible efforts to reinforce the blocking of the other two needs, the matériel-related ones. As regards the second need, one can imagine several ways in which a terrorist organization, having somehow obtained a warhead, might seek to deliver it. This could be attempted by launching it on a missile, for instance, or by carrying it in a private aircraft, or by sailing a ship carrying it into a major port, or by smuggling it by one means or another into the target country and then moving it by road to the chosen objective. Though such options are unlikely always to be as easy as thriller-writers may make them seem, they cannot all be conclusively closed off. It is nevertheless worthwhile to enhance the delivery difficulties which a terrorist would face, for example, by strengthening the ability at ports to detect radioactive material

72

Risks in incoming cargoes, or by improving constraints—as through the international Missile Technology Control Regime (MTCR)—upon the ability to acquire missiles capable of carrying substantial payloads over long distances. The prime focus of prevention must however be upon frustrating the first need, the acquisition of a warhead. There are two conceivable paths to acquisition. The first is constructing a warhead, and the second is stealing or buying one from the stockpile of a nuclearweapon state. The basic principles of constructing a nuclear warhead are now widely known. It is a long step from understanding those to possessing—and keeping hidden and secure—the advanced tools and engineering skills required actually to build one, especially one that would not be awkwardly large for handling, concealment, and transport. It would doubtless be imprudent, however, to assume that an organization of al-Qaeda’s apparent scale and resources would always find that wholly out of reach. The biggest problem for the clandestine builder would be the acquisition of the essential fissile material. That can be formidably difficult, as was shown by the experience of Saddam Hussein’s Iraq, where many years of effort, by a substantial state, had still not led to the acquisition of such material by the time of the 1990–1 Gulf War. Nevertheless, the shock given by the revelations in 2004 (see pages 135–6 below) of the scale and spread of illicit trafficking led by the senior Pakistani scientist Abdul Qadeer Khan in weapon-related know-how and matériel—albeit not, so far as is known, fissile material itself—showed that there were real dangers in this field. It also, however, stimulated collective action by the United Nations. Security Council Resolution (UNSCR) 1540 of 2004, passed under Chapter VII of the UN Charter and so binding on every member state, required that effective measures be put in place by all to guard against any repetition or emulation. The other path to acquisition—procuring somehow a weapon from the stockpile of a possessor state—has attracted more frequent public speculation. This has usually related to the massive size and wide dispersion of the ex-Soviet inventory and to misgivings, in the wake of the break-up of the Soviet Union in 1991, about whether that inventory was everywhere dependably safeguarded. Extensive efforts—many of them internationally funded, notably by the United States in the ‘Nunn-Lugar’ Cooperative Threat Reduction Initiative (CTRI) programme, but also from European states and the European

73

Risks Union (EU)—have been made to achieve more assured security of Russia’s inherited armoury and its key components. Though anxieties have not been entirely laid to rest, they seem legitimately less acute now than they did initially. There is a further potential instrument available for reducing the risk of nuclear-weapon use by terrorists: deterrence. As page 32 has noted, terrorists typically have no evident asset base the loss of which they would regard as disastrous. As a result military deterrence can rarely, if ever, bite directly upon them. But despite the remarkable multi-faceted enabling power of modern global communications such as the Internet, major terrorists do not often function without support or at least tolerance from states, directly or indirectly, as al-Qaeda’s relationship with the Taleban regime in Afghanistan illustrated. Since states or their governing regimes always have something considerable to lose, deterrence (essentially by means other than nuclear weapons themselves, since the use of these could not easily be credible or proportionate in such settings) can be brought to bear upon them. The forcible overthrow of the Taleban regime in 2001–2—approved by the UN Security Council—undoubtedly weakened al-Qaeda, and it sent a deterrent message to others. In the nuclear context, deterrence of terrorist-supporting states might be able to gain added leverage through the fact that weaponusable fissile material typically has a distinctive ‘signature’ peculiar to its source. Given collective political will and cooperation on the part of states possessing such material, it would be technically possible to develop international arrangements to ensure that the source could almost always be identified. This would make clear, with benefits to deterrence, that allowing the acquisition of such material by terrorists—whether deliberately or by culpably inadequate security— could be brought home to the state of origin. Work on realizing and exploiting this possibility is being conducted both in the United States and in Europe. 4 This chapter does not attempt to review all that is being or might be done in non-nuclear ways to reduce the risks or costs of a terroristcaused nuclear-weapon attack, or to contain and alleviate its effects if it does happen. There is however one other nuclear aspect that may merit attention. It is conceivable that some of what would 4 For a convenient summary, see the section on nuclear forensics in Royal Society policy document 07/08, ‘Detecting nuclear and radiological materials’, London, March 2008.

74

Risks need to be done by a country suffering such an attack, especially by way of alleviation, would need a degree of understanding of nuclear weapons and their working that would not normally exist within non-nuclear-weapon states (NNWS) and could not easily be provided to them in advance without some risk of spreading proliferationsensitive information. If study confirms that that is so, contingency planning by the nuclear-weapon states (NWS) ought to cater for the possible need to render very rapid expert help to any NNWS victim. In brief, the continued existence of nuclear armouries inescapably carries a range of potential risks. These have been successfully survived for over sixty years, and none of them seems extreme when weighed against the drawbacks, dangers, or impossibilities of getting rid of weapons swiftly and entirely in the present world. Their management nevertheless needs constant and careful attention and effort. The next chapter considers the particular and high-profile matter of the potential spread of weapon possession by states.

75

7 Proliferation

Why Proliferation Matters ‘Proliferation’ has long been the accepted shorthand for the spread of nuclear-weapon ownership beyond existing possessors (though its customary use has more recently widened to include that of BW and CW, and sometimes also of ballistic missiles). From almost the start of the nuclear age there was concern about this, on both particular and general grounds. The particular ground was that this or that new possessor, whether or not explicitly identified, might import special dangers. They might be, or be unsettlingly thought to be,

r of uncertain motivation or risk-taking propensity; r disposed towards securing disruptive change in the international order;

r of an internal character not wholly trusted to maintain secure long-term control of an armoury and its materials;

r involved in combustible regional confrontations; or r likely, if they acquired weapons, to prompt rivals to take similar steps. The wider ground of concern was simply a generalization of all this: a fear that the more countries had nuclear weapons the higher the risk that sooner or later something would go wrong somewhere and weapons would be used. A minority strand of counter-argument occasionally questioned these worries. 1 It typically suggested that the possession of nuclear 1 See for example Kenneth N. Waltz, The Spread of Nuclear Weapons: More May Be Better, Adelphi Paper No. 171, London, IISS, 1981.

76

Proliferation weapons, bringing with it more vivid awareness of what they could do, was a formidably sobering influence on possessors, so that wild behaviour was highly improbable. The East/West nuclear stand-off had proved to be a stable conflict-preventing condition over a long period, and there was no reason to expect such markedly different effects in other confrontations that this great benefit should be withheld from them. Alternatively or additionally, the spread of weapons to new possessors was in the long run so inexorable that the global community would do better to concentrate on managing and optimizing its consequences rather than fruitlessly bemoaning its prospect. There is some substance in elements of this commentary. For example, though in the wake of the 1998 nuclear explosive tests by India and Pakistan the governments of most other states understandably preferred not to say so (because of feared repercussion or emulation, or general damage to the non-proliferation regime), it was in objective security terms by no means evident that a nuclear stand-off between these two countries amid their historic antipathies and alongside the Kashmir flashpoint was wholly to be deplored. (Chapter 11 discusses the South Asian scene further.) In the round, however, worldwide opinion, both within and beyond governments, has overwhelmingly and rightly regarded the laissez-faire case as outweighed by the contrary arguments. It has not always been easy for the international community, over the years, to sustain a sensibly balanced judgement in evaluating the risks of proliferation and in agreeing the priority, and the degree of urgency, of counter-action against them alongside competing policy considerations. Accusations of hysteria at one end of the judgemental spectrum and of complacency or inertia at the other used to be freely traded. There are one or two states whose character or behaviour is viewed internationally with such mistrust that others reasonably regard any acquisition of nuclear weapons by them as a development to be swiftly and robustly opposed. Apart from these special instances, however, the near-term consequences of particular events that might heighten the possibility of proliferation, taken individually, can rarely be judged calamitous. The prime case for a strong worldwide anti-proliferation strategy is of a broad character, resting on cumulative long-term uncertainties and chain-reaction dangers.

77

Proliferation

The Non-proliferation Regime During the decades since the early 1960s, when China joined the group of open nuclear-weapon possessors, there has been progressively assembled an array of measures to prevent or discourage further proliferation. This array has included the following:

r treaties and legal agreements, global or regional—most importantly the 1968 Nuclear Non-Proliferation Treaty (NPT), but including also the various treaties establishing nuclear-weaponfree zones (NWFZs) 2 ;

r multilateral arrangements such as the MTCR, the Nuclear Suppliers Group (NSG), the US-led Proliferation Security Initiative, the Hague (or International) Code of Conduct for transparency and restraint in the acquisition of ballistic missiles, and the provisions of UNSCR 1540 of 2004, to limit the availability of know-how or matériel for weapons or delivery vehicles;

r security assurances through alliances or otherwise, notably in NATO and under the US/Japan Cooperation and Security Treaty, and in US relationships with South Korea and Taiwan, to allay motivations towards the acquisition of nuclear weapons;

r national and international efforts to reduce risks of leakage from existing capabilities, such as the US CTRI programme and other similar funding efforts, including those by the EU and its member countries such as the United Kingdom, to help secure the former Soviet arsenal and its supporting infrastructure; and

r political and economic pressure. None of these instruments on its own is a total bar to proliferation, and some could with advantage be further developed, more widely and consistently applied, or more dependably policed in order to strengthen the efficacy of the array as a whole. In the aggregate, however, for all the imperfections, the effort has worked notably well, or at least much better than many sober-seeming warnings once 2 Treaties of Tlatelolco 1967 (Latin America and Caribbean), Rarotonga 1985 (South Pacific), Bangkok 1995 (South-East Asia), Pelindaba 1996 (Africa), Semipalatinsk 2006 (Central Asia). This last, however, maintains certain continuing deployment rights for Russia which arguably impair its purity as a declaration of guaranteed freedom from nuclear weapons. At this writing the Pelindaba and Semipalatinsk treaties lacked the number of ratifications needed to bring them into force.

78

Proliferation suggested. In the 1960s it was commonplace to hear predictions of twenty, thirty, or forty nuclear-weapon possessors by the end of the twentieth century. By that yardstick it is a striking success that there are still, in 2008, only the original NPT-recognized five from the early 1960s together with India, Pakistan, Israel (an unacknowledged but undoubted possessor), and also—marginally and precariously—the beleaguered and isolated Stalinist regime of North Korea. Another once-unacknowledged but real possessor, South Africa, has certainly discarded its capability. Major countries like Argentina and Brazil, at one time accounted among the likeliest additions to the list, have credibly renounced the possibility (as also, more recently, has Libya). The break-up of the Soviet Union has left Russia alone as the one nuclear-weapon inheritor state rather than the four (adding Belarus, Kazakhstan, and Ukraine) that once seemed possible, given how widely the huge Soviet arsenal had been deployed. The keystone of the preventive barrier, though by no means its only component, has been the NPT. This enshrined formal tolerance of the continued possession of armouries by the five NWS which had conducted nuclear explosive tests before the Treaty was concluded— the United States, the Soviet Union, the United Kingdom, France, and China—alongside renunciation by all other states acceding to the Treaty. From the outset this feature, though accepted by all the parties through their ratifications of the Treaty, was criticized as unfair, and as incompatible with any expectation that the Treaty régime could remain stable and effective. The criticism continues in full voice, but the Treaty has endured through forty years and a succession of fiveyearly review conferences. One of these, in 1995, agreed unanimously to extend its duration indefinitely. At the debating level the criticism may seem formidable, and it undoubtedly fuels a good deal of discontent. Nevertheless, there are legitimate rejoinders. There is first an argument from realism. In the circumstances of the cold war it would have been fantasy to expect the United States and the Soviet Union (in particular, though not only them) to abandon their armouries. That remains essentially if less manifestly true now. Given this reality, what NNWS have always had to ask themselves is whether they would prefer a ‘discriminatory’ NPT or none. They have answered overwhelmingly in favour of the former alternative. More states have acceded to this treaty than to any other in history, with eventually only three abstentions (India, Israel, and Pakistan) and one declared withdrawal (North Korea). They have

79

Proliferation done so in their own interest—not as a favour to the nuclear powers, but as a benefit primarily conferred on one another. The NPT offers assurance that even amid possible stresses and strains they need not worry about or provide for local or regional nuclear-arms competition among themselves. The NPT’s recognition that all states are not necessarily equal for all purposes is moreover not without parallel in treaty or charter. The world is already accustomed, again as a matter of realism, to a structure that is formally discriminatory in the security field—a Security Council with five states having both permanent membership and special veto-wielding powers. (This is not to argue that the present coincidence between the Security Council five and the NPT five is inherently necessary or desirable. It is in fact neither.) Criticism of the two-tier NPT typically asserts that the inequality is a standing incitement to proliferation. It is certainly a useful debating-point, and so of political and presentational comfort, for states which for their own reasons of perceived national interest wish to acquire nuclear weapons; but it scarcely suffices to constitute on its own any large substantive element of such reasons. There is moreover an opposite side to the coin, at least as important. In some instances—including very crucial ones, such as Germany and Japan—the possession of nuclear weapons by NPT-recognized allies committed to their security has been a key reassurance obviating the need to seek nuclear capability of their own. More generally, it is no way irrational for the great majority of NNWS to make the judgement that in the world as it is, untidy and risky, the abolition of nuclear weapons—further discussed in Chapter 12—is as yet neither feasible nor automatically desirable; and that in those circumstances it is safest, in the interest of all, to draw and maintain a limit of possession around a small and wellestablished group of countries (which happen, moreover, to represent a very substantial proportion of the world’s population 3 ). If it is likely, as I believe along with most commentators, that nuclear weapons have played some part in preventing the recurrence of world-warlike conflict since 1945, that is a global boon from which NNWS themselves have reaped great benefit. 3 In 2007 the aggregate populations of the known nuclear-weapon states (both NPTrecognized and other) came to over three billion, around half of the global total. And states happy to accept nuclear-weapon protection from others account for a significant further preparation.

80

Proliferation

The Treaty Bargains One other NPT feature has served as a focus for recurrent criticism of the legitimated or tolerated NWS. Article VI of the Treaty is regularly cited as enshrining commitments which they have failed to discharge adequately. The criticism typically claims that the Treaty centres on a ‘grand bargain’ whereby the great majority of the parties undertook not to acquire nuclear weapons in return for a commitment by the NWS, embodied in Article VI, to proceed to the dissolution of their armouries. This is, at best, a notably one-dimensional and incomplete representation of the facts, as both the terms of the Treaty and its negotiating history show. 4 The Treaty embodies three major bargains, of which that indicated in its Article VI is the least specific. The other two are the bargain noted earlier between the NNWS, to avoid the costs and dangers of arms races among themselves, and the bargain expressed in Article IV whereby the NNWS are not to be denied—are indeed to be helped in—the exploitation of nuclear power to provide energy for non-military purposes. It is a distortion, and a disservice to nonproliferation, to portray the Treaty as though it were primarily about long-term complete disarmament by the NWS. To do so risks providing political cover to states that may be minded to evade the Treaty’s central purpose, which is that of preventing the spread of nuclearweapon ownership. Moreover, the criticism does not always focus accurately on what the relevant terms of the Treaty are. The preamble clearly aspires to an eventual nuclear weapon-free world, but the terms of the operative Article VI commitment are these: Each of the Parties to the Treaty undertakes to pursue negotiations in good faith relating to the cessation of the nuclear arms race at an early date and to nuclear disarmament, and on a treaty on general and complete disarmament under strict and effective international control.

During the 1970s and into the 1980s the performance of the NWS in cessation of the ‘arms race’ (even though the full aptness of that rhetorical term in the latter part of the cold war was increasingly 4 See for example articles by Joachim Krause, Michael Ruhle, and David Yost in the Chatham House journal International Affairs, Vol. 83, No. 3, May 2007.

81

Proliferation questionable, as Chapter 8 will note) was not uniformly impressive; but on any objective view there has long ceased to be any such race. In the next respect—nuclear disarmament (the text does not apply the adjective ‘complete’ to that part of the Article, as critics of the five seem often to imply, and measures well short of total renunciation can still properly be described as nuclear disarmament)— the record during the cold war was mostly poor; again, however, the position is markedly different now for all of the NWS except China (which, though it publishes no details, undoubtedly has only a modest nuclear arsenal). Finally, the NWS can point to the evident juxtaposition in Article VI of the nuclear-related commitments with general and complete disarmament. Indeed, in the Treaty’s preamble the desired elimination of nuclear arsenals is described as ‘pursuant to’ a treaty on general and complete disarmament under strict and effective international control. No country—and the obligation rests on all parties, not just the NWS—can yet claim much achievement, or policy effort, in advancing the prospect of any such treaty. In 1996 the International Court of Justice (ICJ), in the course of an inconclusive pronouncement 5 about the legality of the use or threat of use of nuclear weapons which notably divided its members and was by no means universally admired by lawyers on either side of the nuclear debate, opined that there was a duty upon the NWS to undertake and bring to a conclusion negotiations for the abolition of their armouries. This view was unanimous, but it was dictum 6 — a comment—and not a finding or ruling on an issue that had been submitted to the Court for decision and argued out before it. Such a comment was not capable of creating a fresh legal obligation that was not indisputably present in the Treaty. The Court would have had no special standing or expertise to set any timescale, or to lay down any new specificity of obligation. The UN General Assembly passed by majority vote a resolution 7 supporting the ICJ’s observation and indeed calling for negotiations to begin in 1997, but that too had no additional legal force.

5 ‘Legality of the Threat or Use of Nuclear Weapons’, Advisory Opinion, ICJ Reports 1996, pp. 226–67. 6 Ibid. pp. 263–4: Dissenting Opinion of Vice-President Schwebel. 7 UN General Assembly Resolution 51/45 M of 10 December 1996.

82

Proliferation

The Future of the Regime All that said, concern about the possible spread of nuclear weapons properly continues to engage especial international attention, discussion, and diplomatic effort. It is misconceived, against the background of the striking on-balance success of the non-proliferation regime over a long lifespan, to present this concern as reflecting a sense that outstanding problems, whether general or particular, mean that the regime stands at some sort of critical cliff edge, or that an inescapable choice must urgently be confronted between the abolition of nuclear armouries and unrestricted proliferation. The Treaty-centred regime has proved powerful and beneficial, and the prime forces making it so have not ceased to operate. There is no large pent-up weight of would-be proliferators pressing immediately upon the dam which the regime provides. Nevertheless, the political thrust of Article VI is clearly towards sustained reduction in nuclear armouries. Both that thrust and the long-term goal of complete divestment have been reaffirmed by the five recognized NWS themselves in political declarations at review conferences of parties to the Treaty, notably at the 1995 conference which agreed to extend the Treaty’s life without limit, and then again at the 2000 conference. Much has been done by way of disarmament since the end of the cold war—far more than anti-nuclear campaigners customarily recognize—but there remains considerable scope, even short of complete abolition, to do more, as Chapter 13 will suggest. Pursuing that, in addition to its intrinsic merits of reducing costs and risks and fostering confidence between states, will help to sustain acceptance of the Treaty regime as reasonable and legitimate (including the limitations, burdens, and duties it places on all parties, for example in regard to avoiding trade in certain items and to cooperating if necessary in enforcement measures). A sense thus strengthened among the NNWS of ownership of the Treaty— not merely acquiescence or grudging tolerance—could reinforce the efficacy of the anti-proliferation panoply as a whole. It is potentially helpful that on both sides of the Atlantic serious and realistic study, receiving governmental support from the United Kingdom 8 and also 8 UK Foreign and Commonwealth Secretary Margaret Beckett, in a speech at the Carnegie Endowment in New York on 25 June 2007, announced governmental financial support for a policy-neutral study project on the abolition of nuclear armouries being

83

Proliferation from Norway, has been initiated into the political conditions that would need to exist, and the technical arrangements that would need to be put in place, in order to make the eventual abolition of all nuclear armouries practically achievable without damage to stability and security. This theme is further explored in Chapter 12. Disarmament aside, experience has shown up weaknesses in the working of the regime which need to be rectified if it is to continue in the long term to achieve its purposes dependably and everywhere. One aspect of weakness is external to the regime itself. Key states are often found unwilling to attach high priority, as against potentially competing national interests of other kinds such as trade or other economic advantage, to enforcing the obligations of the regime by effective pressure upon states that breach them, or move towards doing so. Such pressure may need to include the credible prospect and ultimately the actuality—though preventive deterrence in advance is plainly better than punishment afterwards—of imposing serious penalties. The history of the international handling of the problem cases of North Korea and Iran illustrate the difficulties of achieving and sustaining international consensus on that. (In the case of North Korea, this may be partly because of the paradoxical deterrent effect, already noted on page 31, of the country’s socioeconomic vulnerability to a collapse that would impose heavy burdens upon its neighbours.) So too, indeed, does the saga of dealing with Iraq in regard to WMD in the mid-1990s, when a more unitedly determined effort by all five permanent members of the Security Council to enforce manifestly and strictly the Council’s 1991 resolutions prohibiting Iraqi possession of WMD might have weakened, if not destroyed, the case that was later made for the more drastic action taken in 2003 with such damaging wider consequences. There are however also three substantial weaknesses within the nonproliferation regime. The first such weakness relates to international verification that the purposes of the regime—especially the provisions of the NPT undertaken by the independent London-based IISS, eventually published in September 2008 as Adelphi Paper 396, ‘Abolishing Nuclear Weapons’, by George Perkovich and James Acton. Considerable study work of a broadly similar thrust has been under way also at several leading non-governmental institutions in the United States, some of it with support from the Norwegian government. In addition, UK Defence Secretary Des Browne, in a speech to the UN Conference on Disarmament on 5 February 2008, proposed exploratory discussions about verification among the weapon laboratories of the five NWS.

84

Proliferation itself—are being fully respected. There is a clear need both for comprehensive prohibitions and constraints and for effective guarantees that these are being complied with. ‘Full-scope’ safeguards called for by the Treaty apply the verification tools of the International Atomic Energy Agency (IAEA) to all fissionable material within a state, whether domestically produced or imported. Experience with Iraq, which after its defeat in 1991 was unmasked as having in earlier years successfully concealed extensive contraventions, highlighted the need to tighten the system. In the wake of that uncomfortable episode the IAEA formulated an Additional Protocol to the safeguards rules, entailing considerably stronger verification, in order to reduce the risk that such breaches might in future go undetected. At this writing the Additional Protocol had gained widespread support, but specific national acceptance and implementation were by no means everywhere in place. Universalization needs to be made an important aim—and rendering it effective might well require enhanced resources for the IAEA. The second weakness relates to the fact that Article X of the NPT allows parties to withdraw from the Treaty on giving six months’ notice. Such a provision has many parallels in other treaties, as United States withdrawal in 2002 from its bilateral 1972 ABM Treaty with the Soviet Union and its successor the Russian Federation exemplified. It is questionable, however, whether a let-out of this kind, exercisable without penalty, remains appropriate or acceptable in a field like non-proliferation where the purpose of a globally accepted multilateral treaty has become, at least politically, a major norm of the international system, and where withdrawal by one party may have a crucial effect on whether the treaty’s benefits continue to accrue to numerous other parties. The potential damage in this last regard is vividly illustrated by concerns about the regional repercussions if Iran were to become the possessor of a nuclear-weapon capability, or were reasonably believed at the least to be deliberately constructing an ability to become so at short notice if it so chose. It is almost certainly unrealistic to envisage amending the Treaty so as formally to remove the right to withdraw; but it would be desirable to devise, for future deterrence, an evident and internationally agreed package of disagreeable consequences which any state withdrawing without manifestly imperative reason must expect to suffer. The third weakness—and perhaps in the long run the one of greatest importance—relates to the fact that nothing in the Treaty

85

Proliferation directly debars non-nuclear parties from carrying the development of their capability in the field of nuclear-energy generation and use to a level and in ways (whether or not deliberately so intended) from which they could with comparative ease and swiftness move to the construction of nuclear weapons. This ‘threshold’ capability would come about through the ability to acquire fissile material—the critical enabling component of weapon construction—by enriching uranium to a high level or reprocessing weapon-usable plutonium from spent fuel. The significance of this weakness in the regime may be much widened by the likelihood that as the twenty-first century progresses the global importance of nuclear energy will be enhanced, and its spread accelerated. That would result from the combination of mounting pressure on limited oil and gas resources, of related concerns about security of energy supply, and of demands that the use of fossil-fuel resources should in any event be cut back on climate-change account. A surge of countries coming by this route to a threshold weapons capability would be seriously harmful to confidence in the non-proliferation regime. There is a strong case that the NWS—most of whom have generally taken a somewhat passive view of the offer implicit in the Treaty’s Article IV—will need to build a readier and more attractive system of access to nuclear energy (as well as maintaining impetus in disarmament) if they are to secure, in parallel, acceptance of new limitations to deal with the threshold problem. The review conference of Treaty parties in 2005 was a near-total fiasco, for a mix of reasons which included conflicting regional concerns among some of the parties, and also a damaging general scepticism within the US administration around that time about the merit of global arms-control pronouncements and agreements (including those accepted for the United States by its predecessor). The weaknesses just noted went largely unaddressed, and certainly unremedied. If they remain entirely so at future review conferences— the next one due in 2010—the long-term success of the Treatycentred regime may be endangered even if specific country-related problems such as those posed by Iran and North Korea have meanwhile been resolved or eased. Specific international arrangements to redress shortcomings are usually matters to be developed by other groupings or institutions such as the IAEA; the review conferences are not themselves executive instruments. But the conferences are a powerful—even an essential—forum for mobilizing the

86

Proliferation collective political impetus needed to achieve practical results in remedying deficiencies and enforcing compliance. A clear success in 2010 would underline the authority and legitimacy of the nonproliferation regime. That would enhance, with benefit to the deterrence of breaches, the political penalties both for any breakout and for any failure to share enforcement responsibilities. In brief, the proliferation of nuclear weapons has been and remains a serious concern. It has been tolerably well controlled for forty years, with the help of a diverse array of instruments, and there is no reason to despair of their continued success. To secure that, however, will require new or heightened efforts to deal with changing problems. For full credibility and acceptance, such efforts will need to be partnered by the exploitation of further possibilities for international agreement on the wider limitation and reduction of nuclear armouries. That path of action is examined in the next chapter.

87

8 Arms Racing, Costs, and Arms Control

Arms Races and Costs Over the past sixty years nuclear armouries have absorbed huge costs. As we look back upon the cold war most people would probably now agree that the costs need not have been so high in that long confrontation. There was over-insurance on both sides in numbers, diversity, and performance standards, even given acceptance of the basic case for maintaining a highly assured war-prevention stand-off in which over-insurance would be much better than under-insurance. Suspicion and mistrust sometimes led decision-makers to base force provision on extreme worst-case conjectures about possible scenarios, adversary intentions, and future capabilities. In addition, enthusiasm for exploiting interesting new technological possibilities, and the influence of what President Dwight Eisenhower called ‘the militaryindustrial complex’, almost certainly played a significant part from time to time in driving the insurance beyond what cool strategic appraisal might have suggested to be needed. All that said, however, a sound perspective needs to acknowledge points which criticism of the costs sometimes misses. The vivid term ‘arms race’ was often attached to the history of East/West nuclear force provision. It embodied a good deal of rhetorical hyperbole, at least as applied to the latter half of the cold war. There was undoubtedly a constant and vigorous competition in the technological sophistication of armouries, except where arms control agreements applied or facilitated constraint. (An example of such constraint was the 1972 US/Soviet ABM Treaty, which sharply

88

Arms Racing, Costs, and Arms Control limited defences against ballistic missiles and thereby also eased the pressure for greater increase in offensive-weapon inventories.) But though the realities for most of the confrontation’s forty-year duration were in some ways complex they did not, in the round, bear out the implication which the ‘arms race’ label carries—that is, of an unrelenting contest to outstrip the adversary in the size or weight of nuclear armouries, or in resource allocation. Serious thinkers recognized almost from the start of the nuclear age that once whatever level was judged adequate for deterrent sufficiency had been reached, there was little or no point in numerical superiority. 1 On the Western side average annual expenditure was much lower, in constant-price terms, in the 1970s and 1980s than it had been in the 1950s and 1960s. Numbers in most categories of delivery system were more often pruned than increased. Some categories, as pages 41–2 have noted, were abandoned altogether, and explosive yields were generally reduced. The facts on the Soviet side were generally much more opaque, but both superpowers accepted, from the 1972 SALT I agreement onwards, specific, substantial, and verified constraints on their strategic offensive armouries. And the trend of reduction has continued and even accelerated since the cold war ended. The total number of operational nuclear weapons worldwide cannot be assessed precisely, but figures around 25,000 are commonly cited. Though such a total remains excessive and appalling, it is probably only about one-third of the level at the cold-war peak, and a substantial proportion of it is in non-operational reserve or awaiting dismantlement. Criticism has frequently attacked the costs of nuclear armouries as a deplorable diversion from worthier purposes like the relief of domestic social disadvantage or Third World poverty. Money spent on nuclear armouries, as on other aspects of defence provision, is indeed money not spent on such purposes. It certainly ought to be kept as low as legitimate security concerns allow. But freedom from major war is not a luxury, proper to be foregone in a spirit of generous self-denial. It is both a massive good in itself and a condition of most other goods; and the immense deterrent force of nuclear 1 See for example Bernard Brodie, The Absolute Weapon, ed. Bernard Brodie, Harcourt, Brace and Company for Yale University Institute of International Studies, New York, 1946, pp. 46–9. Dr. Henry Kissinger later made essentially the same point in a muchnoticed press conference on 3 July 1974.

89

Arms Racing, Costs, and Arms Control weapons is widely and reasonably believed to have contributed to ensuring it. Moreover, enormous though the nuclear cost figures seemed in aggregate over the years, they were not, at least in the West, a large fraction of available wealth. In Britain, for example, the defence budget during the 1970s and 1980s typically absorbed around forty or fifty pounds sterling out of every thousand in gross domestic product. Nuclear forces on average took no more than about onetwelfth (often less) of that budget—at most three or four pounds out of the thousand. By 2008 the UK defence budget had fallen, in proportionate terms, to between twenty and twenty-five pounds out of a thousand. The share of this budget to be taken in future by nuclear-weapon provision stood to vary with the uneven incidence of capital expenditure, but it was not forecast to go beyond about two pounds in the thousand. 2 It would be strange to assign to such amounts, rather than to any other public or private expenditure in the entire remainder of the thousand, special blame for preventing, say, governmental overseas aid from reaching the level recommended by the United Nations. There is a wry irony about the costs of nuclear weapons. For what they can uniquely provide in contribution to preventing major war they are cheap, not expensive. NATO during the cold war consistently saw them as making it unnecessary to strive for full parity with the Warsaw Pact in conventional or CW military capability. Such parity would have cost much more in both money and manpower. The United Kingdom, for example, could not have abandoned conscription as readily as it did at the end of the 1950s. It was indeed essentially on such grounds that early in its existence NATO rejected proposals—the ‘Lisbon’ force goals of 1952—aimed at getting closer to equality in non-nuclear force levels.

Arms Control The concern to reduce or contain the costs and risks of nuclear weapons placed increasing emphasis on what could be achieved through explicit international agreement. By the end of the cold war 2 See UK Government White Paper Cm. 6994, The Future of the United Kingdom’s Nuclear Deterrent, December 2006, paragraphs 5.12–14.

90

Arms Racing, Costs, and Arms Control there was in place a substantial structure of negotiated deals to limit (or, more rarely, to reduce) the scale of armaments, or else to constrain types or deployments—in shorthand, arms control agreements. This structure undoubtedly contributed to the desired aims. The long processes of building it, and the record of what was and what was not achieved, brought into play a wide range of considerations. Some of the considerations were essentially political, whether internally or internationally. These included the impact on attitudes and expectations, on trust and goodwill (or the alleviation of ill-will) between potential adversaries, and on the willingness of domestic electorates and alliance partners to accept financial or other burdens. Political consequences of these kinds are customarily hard to measure, but not therefore unreal. Their value is often very substantial. But the direct effects of arms control agreements relate primarily to security—that is, to reducing dangers, including the instability that may flow from unpredictability—and to economy—that is, to reducing costs, including those that may result from over-insurance amid uncertainty. It is in those dimensions that possible agreements need in the first instance to be evaluated. The evaluation needs to look concretely and coolly at specific implications. Arms control is not an end in itself but one of the instruments available for achieving security, alongside others such as political accord, the maintenance of armed forces, and the acquisition of intelligence. In the security calculus there can be no automatic starting presumption that arms control must be of a standing superior to that of other instruments, whether morally or otherwise. It is entirely possible for a bad arms control agreement—a poorly judged bargain, or one not fairly and dependably delivered by all the parties—to endanger peace or impair security, or even to increase costs. There is no substitute for examining specific merits in context, on a cautious and long-run basis. Military-security considerations and political ones can sometimes seem to run counter to one another. In democratic states public opinion about the provision of armed forces may pull in more directions than one—it may occasionally demand projects, or want to preserve institutions, whose real contemporary value for money is questionable. But it may also generate impulses to strike arms control deals more swiftly, or with fewer mistrustful-looking safeguards, than objective appraisal would advocate. Striking a wise balance between such diverse considerations can be a complex task with any arms

91

Arms Racing, Costs, and Arms Control control proposal. But at least two factors can make the balancestriking particularly testing in the nuclear field. The first factor is that the enormity of nuclear weapons inevitably makes public debate about their treatment in arms control highly charged. Natural abhorrence gives special appeal to ideas which offer hope, by political agreement, of avoiding or lightening the burdens they impose. Expectation of action on such lines may put heavy pressure upon political leaders even where more sober evaluation might doubt the real utility of an arms control initiative. For example, from the late 1960s into the 1980s concern was widely voiced in some European countries of NATO, especially the FRG, that the Soviet Union’s numerous MRBMs and intermediate-range ballistic missiles (IRBMs) ought somehow be captured in arms-control constraint. The necessity for this rated poorly in coolly assessed security terms. Given the vast size, diversity, and reach of the Soviet nuclear armoury as a whole, it made no real difference to NATO how many missiles of these particular categories there were within it. The Soviet Union had a full panoply of ample destructive options available anyway against NATO’s European members. Limitation or reduction in this component was therefore not worth the payment of any large price from within NATO’s very different security rationale and structure of defence provision. But the political ‘do-something’ demand in the non-nuclear FRG and elsewhere plainly became formidable. An arms control agreement—the 1987 INF Treaty—was duly concluded between the United States and the Soviet Union, banning all missile systems of maximum ranges between 500 and 5,500km. Though by that late stage of the cold war this deal was undoubtedly good politics, both domestically and internationally, its merit in strict terms of established NATO strategy was open to question. Nevertheless, it is fair to recognize that public contentment with security policies has ultimately, in democracies, a security value in its own right. The question of Soviet M/IRBMs in arms control illustrates also the second complicating factor. For reasons inherent in the fundamentally novel security calculus which the nuclear revolution imposes, the interactions between the armouries of adversaries are less straightforward than in the past. With limited exceptions, 3 3 One exception—a narrow one—during the cold war was that the Soviet ABM system defending Moscow and the area around it relied upon exo-atmospheric detonation of nuclear warheads carried on interceptor missiles to destroy or disable incoming offensive

92

Arms Racing, Costs, and Arms Control nuclear weapons do not exist to fight other nuclear weapons—that notion mostly belongs to pre-nuclear concepts of warfare. The need for a particular category or a particular number is therefore rarely a function of the adversary’s corresponding inventory. The unfamiliar and sometimes complex concepts of the nuclear age are however not easy to put across neatly to non-specialist publics, and it is tempting for political communicators to fall back on simpler-seeming concepts like apparent equivalence. This happened, for example, in most NATO countries at the end of the 1970s, when leaders often chose to present the case for modernizing NATO’s INF armoury with Pershing II MRBMs and Tomahawk GLCMs as made necessary by Soviet deployment of the SS20 IRBM equipped with multiple independently targetable re-entry vehicles (MIRVs). This was not in fact the prime reason why, within the NATO strategic concept for deterrence, modernization was desirable. That prime reason was more complicated and less easy to present to non-expert audiences: it was the diminishing ability of existing NATO delivery systems—that is, at the time, primarily aircraft—to offer assurance of penetrating Warsaw Pact defences in deep strikes on a limited scale from European bases. This diminution generated doubt about whether evident capability for such strikes, which was perceived (as page 43 has explained) to be a significant element in NATO’s inventory of flexible-response options, could be dependably sustained. But the presentational use of ‘SS-20-countering’ as the main public justification for modernization would undoubtedly have made it hard thereafter to resist public expectation, even had the political will existed to do so, that the new NATO missiles be coupled with the SS-20 in arms-control bargaining. There are other examples of arms-control difficulty arising from the temptations of over-facile public presentation. Though the idea of an exceptionless prohibition on nuclear explosive tests, entrenched by a comprehensive test ban treaty (CTBT), has accumulated an international significance that is now rightly accepted as an important re-entry vehicles. (It appears that Russia continues to deploy systems of this kind.) A wider exception, at least in the perceptions of cautious planners, related to the US and Soviet land-based inter-continental ballistic missile (ICBM) forces. In assessments of stability, force requirements, and arms-control balance much concern was often focused upon whether one of these forces could pre-emptively eliminate the other. The concern sometimes seemed disproportionate to the practical reality of the scenario; but however that may be, direct operational relationships of such a kind are not normally to be found in the nuclear-weapon field, especially in the post–cold-war environment.

93

Arms Racing, Costs, and Arms Control political fact, its real merits in security terms have usually been overrated. The 1963 Partial Test Ban Treaty (PTBT) long ago set a barrier to the global health risks associated with tests above ground. The explosive testing of warheads has not been a necessary route for nuclear proliferation beyond the NPT-recognized possessor countries, as the examples of Israel and South Africa almost certainly illustrate, nor does forgoing it make increases in the lethality of weapons technically impossible. The later decades of testing, at least by Western possessors, were mostly directed to improving the accuracy, reliability, or safety of weapons and reducing the need for very large explosive yield—purposes that were not obviously undesirable. The political momentum built up behind a CTBT, however, owed something to the widespread inclination among Western governments to use difficulties about its verification (genuine enough in the earlier years of the concept) as an all-purpose explanation of why progress was not being made, instead of presenting to publics an account—which would inevitably have been more complex—of why the advantages of a treaty were in any event less clear-cut and less weighty than enthusiastic proponents claimed. Another example, of a more hypothetical kind, may further illustrate the point. In the later cold-war years some political opinion within the United Kingdom, even when it accepted a case for the maintenance of UK nuclear-weapon capability in the near term, repeatedly suggested that at some stage that capability should be made the subject of arms control negotiation. The political attraction of hints of this sort was understandable; but no coherent or concrete indication ever seemed to be given of what might be the content of a sensible arms-control bargain—what the United Kingdom might offer to give up (short perhaps of the entire capability) and what security-enhancing change by a negotiating counter-party might credibly be looked for in return. The absence of any such indication should not be thought surprising. Whatever might be judged about the merit or otherwise of UK capability or of its particular form and scale, the considerations which led to its configuration—at what successive governments considered to be, in the circumstances of the time, the lowest level viable for its deterrent purpose—in no way rested upon the precise size or characteristics of the Soviet offensive armoury. At any levels remotely likely to emerge from arms control bargaining that armoury always manifestly possessed (and continues to possess in Russian hands) destructive capability far beyond any

94

Arms Racing, Costs, and Arms Control UK ability either to neutralize pre-emptively or to ride out. There was therefore no prospect that any negotiation of the kind postulated could achieve an outcome that made any objective difference to UK security or to what UK governments needed to maintain that. Bilateral arms control bargaining is by no means the only, nor necessarily the best, instrument or context for constraining and reducing nuclear armouries. It may even in some circumstances impair flexibility and inhibit or delay reduction. During the protracted East/West negotiations in the latter half of the cold war on ‘mutual and balanced force reductions’ (MBFR) NATO offered, as part of a possible bargain, to withdraw a substantial number of Europebased US nuclear warheads that were no longer seen as operationally needed. The Soviet Union’s initial reaction did not take up the offer; but NATO then judged it necessary to retain these surplus warheads in Europe for several years so as not to surrender a possible bargaining card prematurely or without return. Even in situations where what is necessary for security is not closely dependent upon what an adversary may do or may be ready to give up, the pressures of public opinion and of cost in any event impose on governments in democratic systems, especially in the post-cold-war world, powerful incentives not to sustain larger nuclear armouries than are necessary. Extensive decisions for reduction in recent years by all three Western NWS, outside any arms control framework, illustrate that. Interesting questions arise about the future of bilateral nuclear arms-control agreements of the US–Soviet cold war kind. The United States and Russia are no longer in a relationship of adversarial political competition worldwide between near-equals. In cold strategic terms the scale and nature of any nuclear armoury that Russia could afford to sustain would make little difference to what the United States—or NATO—needs. It is unlikely, for example, that a Russian withdrawal from the 1987 INF Treaty, as has sometimes been threatened, need be of material concern to NATO in strict security terms—it would be primarily a comment on the state of political relations. (It would also be a setback to the collective record of the NPT-recognized NWS in the pursuit of disarmament—a record whose maintenance and enhancement Russia ought surely to regard as an interest shared with the United States). And Russian talk 4 of targeting 4

See for example The Times, 16 August 2008, ‘Russia in Nuclear Threat to Poland’.

95

Arms Racing, Costs, and Arms Control nuclear weapons against European neighbours if actions which it dislikes are pursued is no more than bluster, of no credibility or security significance. If seriously meant, it would be merely a reflection of poverty and shallowness in Russian strategic thinking. The decline in the significance of bilateral deals between the former cold-war adversaries was partly exemplified by the 2002 Moscow Treaty setting limits to the total size of the US and Russian strategic offensive warhead inventories. By the rigorous standards of its SALT and START predecessors, that Treaty was more political theatre than serious arms control. It limited numbers only at a single distant point in time, the end of 2012; curiously, the limit was expressed as a wide bracket (1,700–2,200); and it related only to ‘operational’ inventories—reserve holdings were left entirely unconstrained. No matching system of verification was provided. The verification arrangements already in place, those supporting the START agreements of 1991 and 1993, remained due to expire in December 2009, by which time no limits on numbers under the Moscow Treaty would have come into force. A more robust successor to the Moscow Treaty, envisaging lower numbers, a wider ambit, and a more rigorous framework, might nevertheless have value in indirect ways. It could underscore, in the context of Article VI of the NPT, the good faith of the two parties in continuing to pursue nuclear disarmament. It could also reinforce confidence about the security in which the Russian armoury is held. Unease about that, and about consequent risks that weapons or key materials might fall into the hands of terrorists or other undesirable acquirers, has not been entirely dispelled despite the passage of years without known serious mishap and with extensive Western-supported effort in the management of inventories and infrastructure. A new and tauter treaty might, through the influence of verification procedures, strengthen supervision and reassurance. Perhaps more importantly albeit doubtless also with greater negotiating difficulty, it could yield an opportunity for seeking to bring under constraint and oversight the large holding of ‘non-strategic’ nuclear weapons which Russia is generally understood still to maintain. Russia makes no information public about the size or composition of this holding, but speculation that it is in the order of two or three thousand operational weapons (and some commentators put it higher still) is not countered. If that is right, the holding is several

96

Arms Racing, Costs, and Arms Control times larger than that maintained by the United States in comparable categories, and it is not easy to see why Russia should regard its retention as necessary on so large a scale. The absence of information and explanation makes it hard, however, to judge whether a clear and up-to-date rationale exists (beyond perhaps a general cautionary sense of ‘the more, the better’) and accordingly whether there is any coherent basis on which Russia might be brought to consider negotiations about its future as part of a wider bargain. Chapter 13 (pages 173–6) discusses this further. If bilateral nuclear bargains have become less significant (informal ones in South Asia perhaps apart, as Chapter 11 will discuss) than they were during the cold war, multilateral ones remain valuable. The most important are those in the proliferation field, already reviewed in Chapter 7. There is however another group, the salutary effect of which is to provide timely assurance against the migration of competition in nuclear weapons into areas where it has not and need not become established—that is, where no state has yet seen any need or scope to integrate their availability into its key security concepts and policies. One set within this group is made up of treaties defining NWFZs. At this writing there are five such treaties. 5 The possibility of further ones does not yet seem exhausted, though hopes that the concept could usefully extend into areas where nuclear weapons are (whether for good or for ill) part of the existing security structure, for instance in Europe, are probably misplaced unless there is radical political change. Another set within the group of multilateral bargains comprises treaties imposing global bans on placing nuclear weapons, or undertaking weapon-related activities, in certain sorts of environment. The main examples are the 1959 Antarctic Treaty (which extended also beyond the nuclear-weapon field), the 1963 PTBT already noted, the 1967 treaty banning the placement of nuclear weapons in outer space, and the 1971 treaty banning their placement on the sea-bed. Critics occasionally scorned some of these as vacuously forbidding activities which in any event no one wanted to undertake, but they have had and retain long-term precautionary value. More concretely, the desire recently declared by China and Russia for negotiations towards further global limits upon the military use of space may, in 5

For a list of these, see note 2 to Chapter 7.

97

Arms Racing, Costs, and Arms Control addition to its direct significance, have a bearing upon the willingness of those two key countries to accept negotiated constraint in other fields. In brief, the path of concluding formal international agreements for nuclear arms control, realistically followed, has yielded valuable dividends, and its possibilities are by no means exhausted. The next chapter reviews a range of suggestions for reaping similar dividends in additional and sometimes more radical ways.

98

9 Easements and Escape Routes

Virtually from the start of the nuclear era there was a continual and proper concern, widely if unevenly shared, to find ways of removing or at least reducing costs and risks. Much political effort and intellectual ingenuity was devoted to this search—sometimes with useful success, sometimes without it. Chapter 7 has considered how proliferation has been and can be tackled, and Chapter 8 has looked at what other arms-control agreements can contribute. The present chapter briefly reviews some further manifestations of the search.

No First Use The concept of ‘no first use’ (NFU) declarations customarily envisages that nuclear-weapon possessors would give an absolute and permanent promise that never, under any circumstances whatever, would they be the first side to use such weapons in a conflict. The idea had a long history in the Soviet Union’s proclaimed set of arms-control preferences (though there is retrospective reason, as archives have been opened, to doubt its sincerity there, and in any event Russia has withdrawn support for it). It was given special impetus in the West in 1982, when a distinguished group of former US public servants seemed 1 to espouse it. It attracted substantial further support thereafter from time to time. China has made much of endorsing it. India has also done so, albeit with a significant later qualification to the 1 The group comprised McGeorge Bundy, George F. Kennan, Robert S. McNamara, and Gerard C. Smith, writing in Foreign Affairs, Vol. 60, No. 4, pp. 753–68—Nuclear weapons and the Atlantic Alliance. I say ‘seemed’ only because in a direct written exchange with me in the margin of a conference at Ditchley Park in 1983, McGeorge Bundy disclaimed that the NFU undertaking recommended was intended to be wholly absolute.

99

Easements and Escape Routes effect that it reserves the option of nuclear response if it is attacked with BW or CW. NATO and its NWS members have always declined, as also has Pakistan, to adopt the idea. It is important, however, to note that this in no way means that NATO or these countries have ‘a policy of first use’, as careless or hostile commentators have alleged—not least in China, which seeks to place great emphasis upon the purported wisdom, virtue, and unqualified purity of its NFU stance. (Still less, of course, is there any policy of ‘first strike’ in the sense noted on page 17.) To refuse to exclude an option is very different from regarding it as the preferred choice, and NATO has never regarded first use as the latter. Moreover, it is in no way the case—as is also sometimes alleged—either that refusal to rule out first use must entail maintaining ‘hair-trigger’ arrangements for nuclearweapon release, or that it must imply a ‘war-fighting’ concept of the purpose of any use. The presentational attractions of the NFU idea are evident, and the desire is understandable to establish some sort of additional barrier against the escalation of armed conflict to the nuclear level, even beyond the powerful one already created by worldwide abhorrence and over sixty years without use—the nuclear ‘taboo’, in common parlance. From as long ago as the 1950–3 Korean War this taboo has increasingly meant that any thought of initiating the use of nuclear weapons would be burdened with huge global political costs, quite aside from any other consideration (and that has almost certainly been one of the reasons why, as the Korean and Vietnam Wars illustrated, any superpower use of nuclear weapons against small states or for non-vital interests can have virtually no credibility or purchase). But the idea of NFU promises rests ultimately on sand, as an attempt to pre-empt and alter by peacetime declaration the harsh realities of what would be immensely stressful and demanding situations, with huge interests at stake. No one for decades past, if ever, has seriously thought of using nuclear weapons save in defence of deeply vital interests where no other course would serve. It is without doubt most desirable, as NATO regularly urged its members during the cold war, that nuclear powers or alliances should do whatever they reasonably can, both by their political stances and actions and by military provision and deployment, to reduce the possibility of ever having to face such decisions. It cannot however be wholly abolished. An NFU undertaking by

100

Easements and Escape Routes NATO during the cold war would have purported to guarantee that the Warsaw Pact would be left militarily in possession of any gains— even the complete overrunning of a NATO member country—which aggression, however sudden, brutal, and cynical, could wrest and hold by the use of its conventional and CW preponderance. It may now, since the end of the cold war, be hard for most nuclear powers, especially those of the West, to see how their deeply vital interests are at all likely to be threatened by non-nuclear means that could not be countered in non-nuclear ways. They may have a confident expectation to that effect, and they will certainly wish to conduct their policies and shape their military capabilities, plans, and dispositions in order to sustain and strengthen as fully as possible the basis for such an expectation. It might well be helpful, in a world that is continually worried about nuclear weapons and eager to constrain their spread and reduce their salience, that nuclear powers should say all this more openly and clearly than most of them have so far done. But if in the end the expectation were confounded, and truly vital interests were indeed about to be overborne by non-nuclear means— heavy CW or BW attack, say, or trade strangulation—the idea that a nuclear power would let itself be overwhelmed simply because of an NFU promise is plainly absurd. The situation of Israel vividly illustrates that. If a nuclear-possessor country is desperate, whether it be nuclear attack or something else that has made it so, it will not let its options be narrowed by a past promise made in peacetime tranquillity and without reference to the calamities and iniquities which would have created the dire emergency. It might indeed decide that in all the circumstances (which could take many forms) it preferred defeat to the awful risks of embarking upon nuclear action. But the promise itself could not be conclusive, or even likely to weigh heavily, in that calculation. A nuclear-weapon possessor should certainly have a very strong preference against first use. That is not, however, what an NFU statement of the kind customarily proposed would say. A promise is more than a policy. It is an undertaking to others to act (or not act) in a specified way whether or not, when the time comes, the giver of the promise finds it convenient or advantageous to do so. An NFU declaration would therefore be claiming to express something more than preference and expectation. To the extent that it seeks to guarantee that, it cannot be dependable. It is moreover scarcely beneficial

101

Easements and Escape Routes to weaken general deterrence of major war by purporting to remove entirely the preventively helpful shadow of nuclear weapons. More broadly, we should have a deep distrust of taking up, on grounds of advantage in political presentation, positions that rest on false strategic premises. NFU as a universal, unconditional, permanent promise therefore ought not to find favour. Some NFU proponents have argued that since first use is anyway plainly irrational it cannot credibly contribute to deterrence, and that forswearing the possibility therefore involves no real sacrifice. The premise in this argument—the supposed irrationality—rests on obsolete concepts of warfare, as pages 10–11 have explained. Suppose however that we concede it: the argument then turns out to prove too much. If first use cannot be rational, what can be the specific nature and purpose of weapon use that is capable of making second use— that is, response to an adversary’s first use—rational? But if second use cannot be rational, then (on the argument proffered) it cannot deter; and we end with the strange proposition that the prospect of nuclear use deters nothing at all, save perhaps in a calculus of pure ultimate-holocaust revenge. If however there is some rational seconduse concept short of that—a concept related, as it would have to be, to war-termination—it is impossible to see why it should be available to second use yet never in any circumstances to first use. There is hereabouts (though it is not integral to the refutation set out above) another of the nuclear era’s wry paradoxes. No certainties are to be had, and national mindsets may differ substantially, so that mirror-imaging of the adversary is unwise. It is entirely conceivable, however, that in some of the possible range of scenarios first use would actually be less escalatory—that is, less apt to provoke savage response, and more conducive to war-termination—than second use. In the latter case the adversary would already have proved his willingness to engage in nuclear warfare. In the former he might have been hoping to avoid it, and so be more likely to be shocked into reappraisal of his aims when the defender demonstrated the hardihood to breach the nuclear taboo. A secondary point mostly particular to the cold war context in Europe may be historically worth noting. If an aggressor did believe a NATO NFU undertaking—or at least thought that it betokened greater NATO reluctance to take timely nuclear action—he might have felt more able to optimize his force dispositions for non-nuclear attack. Such attack can be helped by close massing of forces; but

102

Easements and Escape Routes the risk of nuclear strike upon such tempting targets, if thought credible, has to be in the mind of an attacking commander, so that his freedom to concentrate is inhibited. He might otherwise be more likely to prevail quickly at the non-nuclear level. The NFU promise, if believed, would then have had the perverse effect of lowering the nuclear threshold—the point where a defender being overrun at nonnuclear levels of combat must choose between nuclear action and defeat. It may be objected to all this that there is already in place, and internationally established, a particular category of NFU assurances. In 1978 the United States and the United Kingdom, in terms later substantially matched by the other NPT-recognized NWS, undertook that they would never use nuclear weapons against any state that was neither a nuclear power nor in alliance with one. Similar undertakings were given at various other times in the context of the establishment by treaty of NWFZs in the South Pacific and elsewhere. In pure theory these assurances—‘negative security assurances’ (NSAs)—are indeed vulnerable to the critical logic outlined above. The limiting clause (‘that was neither . . . ’) and the political make-up of the NWFZs were, however, seen as reducing to vanishing point the probability that any scenario could arise in which the assurances would close off an option otherwise seriously in contemplation. Even then, the NSAs could not ultimately be bankable. Iraq at the beginning of 1991 was formally a state protected by them, its extensive breaches of the NPT not being known then. A senior Iraqi military officer 2 later confirmed, however, that the fact of nuclear-weapon power in the hands of the opposing coalition (quite aside from those of Israel) was not absent from Iraqi minds during the Gulf War in their appraisal of the option of using their CW armoury. The United States will certainly have wanted that to be so when it uttered gravely menacing even if carefully unspecific warnings against such use. An NFU promise, if believed at all, could only lighten a potential aggressor’s perception of risk and so stand to weaken deterrence. Yet it would have done nothing dependable to diminish real risk. NATO and others always recognized this truth in relation to the NFU promise about CW conveyed by the terms of widespread qualified accessions, including that by the Soviet Union, to the 1925 Geneva 2 Lieutenant-General Kamal Hassan al-Majid, in interrogation during his temporary defection to the West in 1995.

103

Easements and Escape Routes Protocol prohibiting the use of such weapons. The qualifications to accession reserved no more than a right of retaliation—that is, of second use. But NATO could never prudently have counted, and in its defensive preparations it did not count, on Soviet abstention from first use of its massively superior CW armoury if major conflict were to break out. (Iraq was also a signatory to the 1925 Protocol and so committed to no first CW use, but that did not prevent its initiating the use of CW in its war of aggression against Iran in the 1980s.) There is no physical way of rendering nuclear weapons incapable of first use. The underlying reality remains that there is no possibility of arranging for major war on crucial issues to be conducted without the risk of nuclear use between advanced powers which have nuclear weapons or the power to acquire them. Policies which attempt to secure that possibility by declaration are doomed to fail. If they have any security effect at all, it may lie simply in the direction of lessening the fear of war, which in the absence of global political transformation is the surest available preventive.

Alternative Defence There have been suggestions that if only we think radically and imaginatively enough we can surely find methods of countering— and so, in prospect, of deterring—nuclear-capable aggressors entirely without involving nuclear weapons or perhaps even other types of massive force. Precedents cited include Indian civil disobedience, led by Mahatma Gandhi, to British imperial rule in the first half of the twentieth century; the French Resistance to German occupation in World War II; and Yugoslav, North Vietnamese, and Afghan achievements in face of the ostensibly greater power of Nazi Germany, the United States, and the Soviet Union, respectively. The difficulty is, however, that even if—very debatably—one regards all these examples (which do not include the Jews of Warsaw during World War II) as encouraging, none of them reflects the truly hard test case. Could such models secure the protection of an advanced state, with all the vulnerabilities of a complex modern urban-industrial society, against a ruthless adversary armed onesidedly with nuclear weapons? The Resistance was in some degree a

104

Easements and Escape Routes telling symbol, but it did not liberate France. Suppose moreover that a nuclear Hitler had been able to make a Hiroshima of, say, Avignon, and then warn the Resistance that Rouen would be next unless the nuisance stopped? It would scarcely have continued. The underlying reality remains that nuclear weapons provide overwhelming force. Unless we feel able to assume that a nuclear adversary will lack the will to use them or that we present no target against which they might be used effectively, arrangements which would confront them with far less force cannot be expected to provide dependable long-term protection. Governments responsible for the safety and freedom of their peoples cannot rest their security policies on hopes of such a kind.

Minimal Armouries For many reasons, including cost, nuclear armouries ought to be kept as small as their proper security purposes allow. At least those of the two biggest powers, and perhaps those of smaller ones too, have in the past mostly been considerably larger than was necessary. It is almost certain that the present levels of holdings, much reduced though they already are, could safely be lowered further. It has sometimes been suggested that the levels could with advantage eventually reach as low as a hundred or even fifty weapons for the United States and Russia (and, a fortiori, for other possessors). It is fallacious, however, to suppose that there is a straight-line graph in which the advantages—political, financial, safety, or other—of reducing armouries rise steadily and evenly as numbers approach zero. Fifty or a hundred weapons are not far more acceptable politically, more virtuous morally, safer, more stable in security terms, or even necessarily many times less expensive than, say, a thousand. The ‘advantage’ line in some of these respects may actually turn downward as numbers fall. If there is a serious case at all for the retention of a nuclear armoury, it is imprudent to take risks about its adequacy and stability by extreme pruning. There is no sufficient benefit to be had from taking risks at the margin. The drawbacks of a level of a hundred (or any other number of the same general order) will vary both with the circumstances and responsibilities of the possessor state and with the particular

105

Easements and Escape Routes characteristics of the armoury in such respects as warhead types and yields, delivery systems and platforms, penetrative and other performance, deployment, reliability, protection, and logistic support. But a basic list of potential disadvantages of a hundred-weapon armoury to be weighed might include these:

r It might be an easier and more tempting target for pre-emption or disruption (whether by nuclear weapons or in some other way, not necessarily of orthodox military character) than, say, a thousand. Moreover, a low limit on numbers, especially for the United States and Russia, would almost certainly be found acceptable only in an arms-control agreement. Such an agreement would have to have stringent verification provisions; yet verification would deprive the armouries of the contribution which secrecy can make towards protection against pre-emption.

r It must be more sensitive to erosion by accident, like fire at a storage depot.

r It must be more vulnerable to technical mishap, such as an unexpected design fault emerging in weapons, delivery vehicles, or platforms. The United Kingdom, which has always had an armoury of modest size, has vivid memories of two disconcerting events. In the 1960s the entire holding of a strategic bomber type—the Valiant—had to be retired virtually overnight when a grave airframe fatigue problem was found. More recently, continuous at-sea patrol came close to being lost when a serious fault came to light in the propulsion system of the Polaris-carrying submarines.

r Because the level is drawn with such tautness, verifying exact compliance with it becomes both more important and probably more difficult. At a thousand weapons, it does not matter much in bottom-line security terms if someone cheats by a margin of, say, fifty weapons. At a hundred it might matter considerably.

r Partly as a result of that, both the potential advantage and the potential ease of effective breakout from the limit are greater, and the potential temptation mounts accordingly. A quick clandestine spurt by a violator (possibly partnered by some well-timed pre-emption by one method or another) has a far better chance against a hundred than against a thousand of seizing a crucial

106

Easements and Escape Routes advantage—a key change in power balance—before others can react effectively.

r At such a level it becomes at best much more difficult to devise credible and acceptable doctrine for the use of weapons, and policy for their targeting. This last point may merit expansion. Earlier chapters have explained that if weapons are to deter they must be capable of realistic use— they cannot serve even a solid political function if they are nothing but symbols or negotiating counters—and also that there are good reasons for having options evidently available for different types and levels of nuclear strike. Those reasons, though first elaborated in the cold war context, are not exclusively related to that context; the underlying concepts are of permanent relevance. It is hard to imagine that an armoury of a hundred warheads, let alone of fifty, could provide an adequate range of options for a state with worldwide involvement and responsibilities such as those of the United States. Having a given number of warheads held somewhere in inventory is nowhere near equivalent to having a dependable capability, assured against the possibility of adverse circumstances, to execute that number of successful strikes at required times and places. The United Kingdom seeks to maintain no more than forty-eight warheads, in one submarine, at sea at any one time; yet its experience-based assessment of the operationally available stock needed for assured support of that posture was reduced in December 2006 merely from 200 to 160. As to targeting, it is another uncomfortable paradox of the nuclearweapon business that below a certain level the fewer weapons there are available the cruder their targeting may need to be. The planner may be driven to envisage population-killing counter-city strikes in order to construct an ultimate penalty grave enough to underpin deterrence of an adversary regime that is, virtually ex hypothesi, determined and perhaps brutal. But to reverse the trend there had been in the West during the later years of the cold war away from counterpopulation targeting would be of low credibility; of little political acceptability, either internationally or domestically; and ethically intolerable. A level of a hundred or fifty for a major power has many of the disadvantages of tokenism. Potential adversaries may be minded to

107

Easements and Escape Routes judge that having come down so far the possessor state does not really mean business at all, that it lacks ultimate resolve. And it would probably become harder rather than easier to hold off politically those segments of opinion that believe the proper level to be zero.

Virtual Armouries Another radical suggestion for reducing risks is that possessor states should negotiate an agreement whereby their nuclear capabilities would be verifiably held in a ‘virtual’ condition. This concept is usually intended to mean much more than just taking systems off high-alert postures (a change that is certainly feasible, and very desirable). The concept envisages that the capabilities would be rendered unable to be used operationally except following overt reconstitution measures that would take a considerable time—typically at least months. There are formidable difficulties about achieving such a situation globally even if it were regarded by all possessors as acceptable in principle. It is highly questionable whether it would be so regarded. The South Asian environment illustrates this. Pakistan would seem unlikely, in face of India’s wide margin of preponderance at non-nuclear levels of military force (and the steps being taken also, at this writing, to reduce the reaction time of those levels) to regard its security as enhanced by making its counter-balancing nuclear capability for deterrence less readily and evidently available. NATO during the cold war would certainly have felt similar reluctance. Even given political will, defining equitable rules would be very difficult, and verification arrangements would need to be complex and intrusive. The structure, composition, and deployment of present armouries differ widely between one possessor and another. It would be—at best—extremely hard to devise verifiable arrangements that would ensure, amid the diversity of circumstances (including perhaps problems about the willingness and training of personnel for this unusual role), that the feasible timetable for reconstitution was the same for all. To the extent that it was not so perceived, there could be dangerous temptations in time of tension or crisis (and also unsettling perceptions of such temptations) to seize advantage, whether for protection or for offence, by winning the reconstitution race. And

108

Easements and Escape Routes the problems stretch even further than that. How could realistic and equitably symmetrical allowance be made for the possibility, in time of harsh confrontation, that one side might seek, by military action or otherwise, to disrupt the other’s reconstitution timetable, profiting from the knowledge of the other side’s dispositions which verification arrangements would inescapably have made available and perhaps also from the nature of the dispositions likely to be required by the agreement, such as monitored central storage? Provision for national security is inescapably in the insurance business, and concepts have to be tested against awkward scenarios, not just benign ones or even most probable ones. In the round, the ‘virtual’ concept might well be seen as a minus rather than a plus for stability and confidence in war-prevention. Proponents of the concept have recognized that it is especially hard to fit missile-carrying submarines into a balanced and guaranteed scheme, for several reasons such as the likelihood that the demanding skills needed for submarine operation cannot easily be maintained in cold storage. But such submarines are already the most stable and least worrying class of platforms—the least pre-emptible, and so the least in need of high readiness to fire or of launch-on-warning plans. That surely suggests that there is something amiss with the concept. In reality, the concept’s elaborateness and difficulty are disproportionately severe by comparison with the problems it seeks to solve. The history of the nuclear age does not bear out suppositions that risks of accident, mistake, or overreaction leading to weapon launch are acute. Such risks of these kinds as do exist could and should be alleviated by less drastic and complex but more practicable measures to ease alert states. ‘Virtual’ armouries (or minimal ones of the kind discussed earlier) might perhaps be appropriate or even necessary as brief way-stations on a pre-defined and agreed road to zero holdings. As postures in their own right, however, they lack merit.

Defence Against Ballistic Missiles In a dramatic speech on 23 March 1983 (which soon acquired the populist label ‘Star Wars’) US President Ronald Reagan placed before his country the goal of a comprehensive defence shield against ballistic missiles—an impermeable ‘Astrodome’ dependably

109

Easements and Escape Routes protecting the entire homeland—to render nuclear weapons, as he envisaged, ‘impotent and obsolete’. A massive weight of sceptical criticism swiftly accumulated around this concept, and in its pure form it scarcely seems to retain any substantial body of convinced advocates. Vigorous and complex debate continues about the merits of systems for intercepting ballistic missiles, but with far more modest goals. This book does not attempt substantive appraisal of that debate, though I remain sceptical both of whether the colossal militaryindustrial investment in BMD undertaken by the United States—with expenditure already soaring well beyond a hundred billion dollars— has yielded or is likely soon to yield security dividends worth the financial and political costs, and also of whether in the nuclear field deterrence of aggression is generally in much need of an attempted supplement of this kind. I am still more sceptical of whether BMD is relevant and valuable enough to the security concerns of countries like the members of the EU to warrant heavy expenditure by them from within defence budgets already hard-pressed to meet demands of much more evident importance, such as enhanced strategic and tactical airlift. But whether or not these scepticisms are justified, it is incontestable that BMD offers no credible likelihood of providing the complete escape from harsh nuclear realities and dilemmas to which the Reagan speech aspired. No nuclear-weapon possessor—certainly not the United States—now supposes that BMD offers the prospect of making its offensive armoury unnecessary. And indeed in some settings, as Russian and Chinese reactions to potential US deployments of BMD suggest, it may be more likely to increase than to reduce the levels or salience of nuclear forces, since it risks prompting countries concerned about their maintenance of effective deterrence to invest more heavily in offensive capability. The utopian aspiration which President Reagan voiced (an aspiration that did not, however, address the various ways there are of delivering nuclear weapons other than by ballistic missile) illustrates the longing to find physically guaranteed escape from the nuclear revolution—to recover somehow a lost invulnerability. Reality cannot however be evaded or reversed in such ways. The pervasive and permanent truth with which humanity now lives is that science has put in our hands virtually infinite destructive power, from the shadow of which there is unlikely to be technologically assured exemption. Practical hope has to lie primarily with political

110

Easements and Escape Routes understanding and skills, based upon awareness on all sides that we have reached and cannot dispel the reductio ad absurdum of all-out war. That awareness needs to be progressively supplemented, and if possible one day made redundant, by structures to provide worldwide and dependable non-military routes for constraining and resolving disputes between states, alongside politically acceptable methods of verifying further limitation and perhaps eventual abolition of armouries. Chapter 12 will explore this further. In brief, this chapter has argued that none of the radical approaches surveyed offers a dependable prospect of escaping from the harsh burdens of nuclear weapons by stepping outside the familiar mainstream repertoire for managing risk and cost. Part III now sets out illustrative examples of how countries have sought in practice, against the background of the general realities sketched in Parts I and II and of their own particular circumstances, to establish and conduct their possession of nuclear weapons.

111

This page intentionally left blank

Part III National Nuclear-Weapon Postures and Policies: Britain, India, Pakistan

This page intentionally left blank

10 United Kingdom Doctrine and Policy

History The United Kingdom brought to the nuclear revolution a security mindset differing in significant respects from that of the United States. The United States, behind its huge two-ocean moat, enjoyed (despite the outlying Pearl Harbour shock in 1941) a perception of homeland sanctuary that in some degree endured, psychologically if not intellectually, until 11 September 2001. The combination of island geography and dominant maritime power had long given Britain a similar sense. With the advent of aircraft, however, the experience of the two twentieth-century World Wars had irreversibly erased that sense. In World War I German air bombardment of Britain began as early as January 1915, initially by Zeppelin airships and later by fixed-wing aircraft. The direct damage they inflicted was modest by comparison with the carnage of the Western Front, but the memory of attack and vulnerability lingered in public consciousness. In 1932 the leading British political figure of the time, Stanley Baldwin, notoriously warned that ‘the bomber will always get through’. 1 In World War II air bombardment of the homeland—especially though not only in the 1940–1 ‘Blitz’ on London—became part of common experience. New dimensions were added in 1944, when attacks began first from the V.1 cruise missile and then from the V.2 ballistic missile. The scale of all this did not approach that suffered by Germany or Japan, but over 50,000 civilians were killed. 1

House of Commons Official Report, 10 November 1932, cols. 613–18.

115

United Kingdom Doctrine and Policy This thirty-year history meant that the British people and their leaders entered the nuclear age with an awareness and indeed acceptance of inescapable vulnerability more vivid, yet at the same time less shocking because less unfamiliar, than did the United States. In addition, from 1940 until almost the end of the conflict in Europe in 1945, Britain’s own strategic bombing offensive was a massive component of the nation’s war effort. (There were heavier losses in action among Bomber Command aircrew in World War II than among British junior officers on the Western Front throughout World War I. Around half of those who flew operationally died.) The impact, the value-for-resources-used, and even the morality of the offensive were much debated after the war; but its weight and salience meant that awareness of the practical aspects of long-range homeland attack—the realities and limitations of targeting, for example—probably became more widespread, not only among professional servicemen but also with political leaders and in public discourse, than in most other countries. This awareness included an inclination to believe that such attack must tend to be directed—rather like maritime blockade, a historic form of Britain’s military leverage—to sapping an adversary’s political, economic, and social strength and will rather than just assailing his armed forces directly. In the immediate aftermath of August 1945 there were mixed views in Britain, as elsewhere, about the long-term import of what had happened to Hiroshima and Nagasaki. Some voices, including military ones, questioned its revolutionary significance, but air force leaders from the start took a different view, matching that of Prime Minister Clement Attlee. In a notable letter in September 1945, only two months after succeeding Winston Churchill in office, he argued to President Harry Truman—in line with his internal paper already cited on pages 7–8—that the new weapons entailed a qualitative, not just a quantitative, change in the nature of warfare. Existing conceptions, he said, were now ‘completely out of date . . . the only deterrent is the possibility of the victim of such a [nuclear] attack being able to retort on the victor’. The idea of deterrence by the prospect of retaliation, as the only protection against nuclear weapons, dominated government thinking from then on. It was reinforced, before long, as studies increasingly brought out that a relatively small number of thermonuclear weapons could virtually destroy Britain as a functioning modern society. The concern for deterrence centred upon nuclear

116

United Kingdom Doctrine and Policy weapons was given an even sharper edge by perceptions that Soviet non-nuclear military preponderance in Europe was so massive that without prompt and all-out US participation (not to be assumed until the creation of NATO in 1949, and even thereafter not in rapid prospect on a matching scale) a Soviet assault could reach the English Channel within weeks; and that the diversion of economic effort that would have been needed to offer any likelihood of altering this assessment was utterly out of the question for the war-devastated economies of Western Europe. The United Kingdom had begun work towards developing nuclear weapons as early as 1940, before the United States did, but its effort was absorbed into a joint US/UK project soon after the entry of the United States into the war in late 1941. At the end of the conflict, however, the United States abruptly terminated cooperation on nuclear-weapon development, and indeed legislation prohibiting it (the McMahon Act of 1946) was enacted. In January 1947 the UK government decided to develop a capability of its own. Though this decision was at first kept very secret and had indeed been formally opposed on affordability grounds by the Cabinet ministers with economic portfolios, it is hard to imagine that Britain—conscious of itself as one of the main load-bearing victors of World War II, and still with worldwide imperial responsibilities—might have decided otherwise. The development of jet-propelled bombers of strategic range had been initiated in 1946, and in the mid-1950s three types were brought into service—Valiant, Vulcan, and Victor, collectively known as the V-force, with the latter two types being at least the equal of contemporary US aircraft in most aspects of performance other than radius of action (which Britain needed less). At its peak the operational V-force numbered about 140 aircraft. The first UK nuclearweapon test had been held in 1952, and by 1958 free-standing UK competence in weapon design had been sufficiently demonstrated for the United States to repeal the McMahon Act barrier and resume close cooperation through a wide-ranging Mutual Defence Agreement (MDA) which remains in force. This cooperation was never subsequently interrupted, and it became very close, while taking care from 1968 onwards not to breach the constraints which the NPT came to impose. The MDA carried much advantage to the efficacy and economy of UK effort on warhead design and production, though the benefits were not exclusively one-way.

117

United Kingdom Doctrine and Policy The United Kingdom introduced delivery systems and warheads of its own making also at non-strategic levels, based initially on the Canberra light bomber (from which the US B-57 was derived) and subsequently on various fighter/ground-attack (F/GA) or strike aircraft, both Royal Air Force and Royal Navy. UK nuclear warheads came to be deployed also in anti-submarine depth-charges in Royal Navy surface ships. All these arrangements were entirely distinct from those under which from the mid-1950s onwards the United Kingdom operated various delivery systems—themselves UK-provided, save for the US Thor IRBM stationed at bases in the east of England from 1959 to 1963—to carry US-owned warheads on a ‘dual-key’ basis. Under this system ultimate control over warhead custody was to remain always with US personnel until the moment of operational release authorized by both governments. A pattern was thus set that came to be adopted also for several other members of NATO, though the Thor precedent of US provision of the delivery vehicle itself was not generally repeated. The initial creation of UK strategic nuclear capability was a fully independent enterprise, aside from a period when Valiant bombers had US weapons under dual-key arrangements pending the buildup of the full planned inventory of UK ones. Most early thinking and planning for the use of the capability, however, envisaged operation alongside US strike forces. The particular value of UK contribution—political influence apart—was seen initially as being added weight, assurance that targets of particular concern to Britain would be struck early on, and possible path-clearing disruption of Soviet defences. There was also the possibility of posing an awkward practical dilemma for any Soviet concept of pre-emptive missile strike on the West—‘simultaneous launch or simultaneous arrival on target?’—given that whichever option was chosen would give one or other of the two Western forces enhanced warning time. But the US armoury soon grew to a size and diversity that made any UK supplement, in narrow operational terms, no more than a modest optional extra. These realities partnered awareness of the inescapable future obsolescence of the V-force and of the cost of any replacement for it within a hard-pressed defence budget that still had to support extensive post-imperial commitments as well as NATO ones. The UK government was therefore compelled to think afresh and hard about the nature and strength of the case for continuing nuclear capability, and about the make-up of the capability.

118

United Kingdom Doctrine and Policy The need for such thinking, and for effective communication of its outcome, was heightened by the fact that—to an extent not matched, at least proportionately, in any other nuclear-weapon state—the case for continued capability was under vigorous domestic challenge not only on value-for-money and opportunity-cost account but also on deeper grounds of political and moral unease. The opportunity-cost aspect of the challenge often came from Army circles. It has been rare for senior Army officers to express enthusiasm for UK retention of an independent capability, perhaps partly because the British Army has never had nuclear weapons of its own as distinct from US dualkey ones. More widely, however, abhorrence of nuclear weapons has had a special appeal to the strand of internationalist moralism that has featured prominently (and not discreditably) in British politics, mostly on the left, since before World War I. Vocal albeit minority challenge to the possession of such weapons has remained a feature of the British scene almost continuously since the late 1950s, though with highs and lows largely according to whether new decisions were on the agenda. For most of the 1980s, indeed, one of the two major political parties—Labour—was directly and unconditionally opposed to the retention of nuclear capability. It is now generally believed that this stance worked to Labour’s electoral disadvantage, and it was abandoned by the party leadership by the end of that decade. It has, however, remained powerful on the left of the party. In the March 2007 debate in the House of Commons 2 on whether work should be undertaken for the long-term continuance of UK capability, about a quarter of Labour Members of Parliament voted against the proposal of their own party’s government.

The Concept of Independence The challenge to policy thinking became inescapable in the early 1960s. The United Kingdom had the inherent capacity to maintain a fully independent procurement base. It was no less wealthy or advanced technologically than France, which chose to do that. The opportunity cost, however, to other aspects of defence provision within unavoidable financial constraints would have been severe.

2

House of Commons Official Report, 14 March 2007, cols. 298–398.

119

United Kingdom Doctrine and Policy The UK government abandoned in 1960, as much on grounds of vulnerability to pre-emption as of cost, the development of a silobased MRBM (Blue Streak) to succeed the V-bombers. For a brief period thereafter it was envisaged, in agreement with the United States, that the United Kingdom would buy the planned US airlaunched ballistic missile Skybolt (WS 138A) to extend the effective life of the V-bombers, but the United States then decided not to go ahead with producing it. There can be no doubt now, for all that the Royal Air Force found it painful to relinquish the role and the Royal Navy was (and perhaps remains) by no means unanimously keen to possess it, that the shift thus forced to a submarine-launched system was right for long-term UK interests. The government chose, while continuing to design and manufacture its own warheads and platforms (in the form of nuclear-propelled submarines, with only limited initial help from the United States in their design), to rely on purchase from the United States for its strategic delivery vehicles— that is, the long-range missiles the development of which would otherwise have been the most demanding and costly element of force provision. In 1962, when the deal for this outcome was struck in a summit meeting at Nassau in the Bahamas, some opinion within the US administration—once thought to include that of Secretary of Defense Robert McNamara—was minded to see UK capability as potentially an unwelcome complication in NATO, in international arms control, and in the encouragement of a collective political Europe. This strand of opinion had accordingly been far from keen on having the United Kingdom helped in this substitute way after the demise of the Skybolt plan. President John Kennedy overruled it, however, and reluctance of such a kind became less and less evident in later years. There was, moreover, never any sign of US attempts to use the nuclear-force relationship as a lever to secure UK compliance with US wishes in other fields, for example over military participation in the Vietnam War. The basic choice to buy submarine-launched delivery systems from the United States—initially the Polaris A.3 missile with multiple reentry vehicles (MRVs), thereafter the Trident D.5 missile with re-entry vehicles that were (unlike those of Polaris) independently targetable (MIRVs)—has been maintained as subsequent major replacement decisions have arisen in 1980–2 and 2006–7. A limited exception

120

United Kingdom Doctrine and Policy was undertaken in the later years of the cold war, when the United Kingdom completed for itself the complex and costly Chevaline modification of Polaris A.3’s MRV-carrying front end. The UK force was too small to rely simply on saturating BMD, as the United States could. The aim of the Chevaline project was to ensure by technical means that, as deterrent credibility was perceived to require, the force remained evidently capable—whether or not the government would actually choose so to target its missiles—of defeating the exo-atmospheric BMD system which protected Moscow and a wide area around it. The size and shape of this area depended on the particular trajectory of an incoming missile, but its typical coverage was measured in tens of thousands of square miles. (Chevaline also offered some underlying insurance against the possibility of Soviet withdrawal from the 1972 ABM Treaty and its constraints.) No similar project to meet what was customarily referred to as ‘the Moscow criterion’ has been judged necessary in the era of Trident, with its more versatile capability—a capability which indeed the United Kingdom has never exploited to its full warhead-carrying capacity. It has always been clear that there would be no point, alongside the huge US capability, in shouldering the burdens of a UK strategic nuclear capability if it were wholly dependent upon the United States. The United Kingdom has therefore had to shape and communicate a careful concept of what independence needs to mean in this field, and why it is worth having. The key features of the thinking, as it matured, were two linked ideas (although political, institutional, and other motivations, as distinct from security rationales, were always more diverse across the wide span of time and of people involved). One idea was that of a second centre of nuclear decision-making within the Western political grouping. The other was that of operational independence. From early in the nuclear age the US armoury was more than adequate in material terms—numbers, diversity, reach, and technical and operational quality—for the needs of any alliance or coalition to which the United States was dependably committed. The security case for any of its partners to spend scarce resources on providing an independent supplement could rest only on hypotheses that in some scenario or other the US armoury might be thought not available, or not reliably available. For example, it might be feared that in the cold-war situation of effective nuclear parity between East and

121

United Kingdom Doctrine and Policy West, with the United States itself inescapably under mortal threat, the Soviet Union might calculate (or, as UK spokesmen were usually careful to say, miscalculate) that when real operational decisions had to be faced US nuclear power would not be used, or not fully and promptly used, in the defence of Western Europe. The existence of independent nuclear capability in Western Europe, far more directly threatened by possible Soviet aggression, was seen as a useful added insurance against any such assessment. Given such a premise, what independence needed to mean in practice (at least from the standpoint of security rationale—cloudier considerations of political posture or national image are not addressed here) depended on what were the scenarios of perceived US nonavailability to be insured against. These scenarios might be of broadly two kinds. The first would postulate that the United States, while still politically committed to its allies, might hold back when faced with the nuclear decision amid the heat and fear of war. The second would postulate a deeper and longer-term estrangement from Europe— a radically changed environment in which the United States had largely disengaged from European security concerns, and in particular had withdrawn its cooperation and abrogated any obligations to European allies in nuclear-force procurement and support. If it were desired to cater just for the first sort of scenario, what was needed was simply operational independence (it might be called Mark I): the evident capability to press nuclear-launch buttons whether or not the United States chose, or wished its allies, to do so. 3 But to insure also against the second sort—long-term US estrangement—required procurement independence (Mark II). It is unilluminating to argue about which Mark is ‘real’ independence. The practical point is that they are alternative insurance policies. As in most insurance situations, the wider the cover required, the higher the premium. The United Kingdom chose to take out the Mark I level of cover, and this rarely 3 Some anti-nuclear campaigners in Britain have wished to believe that there is not even operational independence—that should the use of UK capability ever come seriously into contemplation the United States would have, and would certainly activate, assured means, if it were opposed to such use, of physically preventing it. (For an egregious example, see New Statesman, 27 March 2006, pp. 16–17.) There is neither evidence nor likelihood in support of any such conjecture, and UK ministers have categorically dismissed it. If it were true, successive UK governments of both main parties for over forty years would have been massively deceiving Parliament and public, and knowingly wasting huge resources on a strategically pointless project.

122

United Kingdom Doctrine and Policy cost more than about one-twelfth, and indeed usually much less, of the defence budget. France’s impressive and sustained technological achievement in building, largely unaided, independence Mark II— within total defence expenditure of much the same magnitude as the United Kingdom’s—appears to suggest two or three times as much for Mark II. Inevitably, that carried opportunity costs in the field of conventional forces, as the disparity between the military contributions of the two countries in the 1991 liberation of Kuwait perhaps illustrated. A wide variety of considerations, including both domestic and international political ones, bore upon whether the United Kingdom ought to maintain a capability to meet this rationale of secondcentre operational independence Mark I. The design of the capability raised further issues, for example, about its composition, weight, level of assurance, and targeting. The central security question which successive UK governments needed to ask themselves, however, was whether a capability of this kind yielded the best ‘added value’ for the national interest as compared with other possible uses of the resources it absorbed, such as the provision of stronger conventional forces or indeed purposes outside the defence field. Not everyone who supported the case for a UK capability would have accepted this formulation of the central issue as a judgement weighing added value against cost. Some appeared, at least in their choice of justifying language, to attach to the capability an absolute importance—to believe that it lay so crucially at the heart of national security that it must be sustained whatever the cost, rather as French official doctrine has seemed to hold in relation to French nuclear capability. My own view was and remains less unqualified. For hypothetical example, if domestic political or economic exigencies had so strained the defence budget during the cold war that a choice had to be made between a substantial nuclear capability and a substantial NATO-committed land/air contribution on the European mainland, I should have advised in favour of the latter. During the cold war the pattern—both geographical and political— of full British commitment to NATO gave the capability’s added value, whether judged large or small, a particular Alliance dimension. The United Kingdom expressed no formal nuclear assurance to the FRG or other European NATO allies (nor was any sought, though FRG belief that UK nuclear capability was helpful to NATO

123

United Kingdom Doctrine and Policy deterrence was made plain) beyond that already implicit in the terms of the 1949 North Atlantic Treaty and the revised Brussels Treaty of 1954. There were, however, clear inferences to be drawn from a UK stance (unlike that of France) of unqualified and active membership in the Alliance’s integrated military structure and its nuclear planning effort, partnered by the forward deployment in Germany of the only field army the United Kingdom possessed together with substantial air elements some of which had UK-owned nuclear weapons. The US nuclear armoury itself constituted a massive insurance policy. A supplementary capability based on a second-centre rationale, as an insurance policy against the failure of the first insurance policy, was inevitably directed therefore to a scenario of low probability, albeit relating to eventualities that might otherwise be hugely disastrous. That low overall probability was part of what the balancing judgement—value against cost—properly had to weigh. Critics sometimes, however, misunderstood the structure of the judgement. In particular, it was occasionally argued that the United Kingdom must be at least as likely as the United States—perhaps much more so—to baulk at the nuclear decision, and that the notion of adding deterrent value by the second-centre concept was therefore empty. But the concept did not depend on any comparison between the two national probabilities. The point and effect of operational independence was that the UK probability, whether larger or smaller (and standing in any event to vary widely with circumstances), was a separate and additional probability, a further and different complicating uncertainty which an adversary would have to weigh, not a lesser included case.

The Scale and Use of Capability Successive UK governments have mostly been and remain very reluctant to go into detail publicly about how they arrive at judgements on the scale of capability, and on how it might be used if it ever had to be. The most explicit conceptual account of what it was thought, during the cold war, that the UK strategic nuclear force should be able to do is to be found in the 1980 document 4 explaining why the 4

124

Defence Open Government Document 80/23 of July 1980.

United Kingdom Doctrine and Policy Trident SLBM (at that stage envisaged in its C.4 version—the shift to D.5 came in 1982) had been chosen to replace Polaris:

The ‘Second-Centre’ Role 9. If Britain is to meet effectively the deterrent purpose of providing a second centre of decision-making within the Alliance, our force has to be visibly capable of posing a massive threat on its own. A force which could strike tellingly only if the United States also did so—which plainly relied, for example, on US assent to its use, or on attenuation or distraction of Soviet defences by United States forces—would not achieve the purpose. We need to convince Soviet leaders that even if they thought that at some critical point as a conflict developed the US would hold back, the British force could still inflict a blow so destructive that the penalty for aggression would have proved too high. 10. There is no way of calculating exactly how much destruction in prospect would suffice to deter. Clearly Britain need not have as much power as the United States. Overwhelming Britain would be a much smaller prize than overwhelming the United States, and a smaller prospective penalty could therefore suffice to tilt his assessment against starting aggression that would risk incurring the penalty. Indeed, one practical approach to judging how much deterrent power Britain needs is to consider what type and scale of damage Soviet leaders might think likely to leave them critically handicapped afterwards in continuing confrontation with a relatively unscathed United States. 11. The Soviet Union is a very large and powerful state, which has in the past demonstrated great national resilience and resolve. Its history, outlook, political doctrines, and planning all suggest that its view of how much destruction would constitute intolerable disaster might differ widely from that of most NATO countries. Appalling though any nuclear strike would be, the Government does not believe that our deterrent aim would be adequately met by a capability which offered only a low likelihood of striking home to key targets; or which posed the prospect of only a very small number of strikes; or which Soviet leaders could expect to ward off successfully from large areas of key importance to them. They might even be tempted to judge that if an opponent equipped himself with a force which had only a modest chance of inflicting intolerable damage there might be only a modest chance that he would have the resolve to use it at all. 12. Successive United Kingdom Governments have always declined to make public their nuclear targeting policy and plans, or to define precisely what minimum level of destructive capability they judged necessary for deterrence. The Government however thinks it right now to make clear that its concept of deterrence is concerned essentially with posing a potential threat to key

125

United Kingdom Doctrine and Policy aspects of Soviet state power. There might with changing conditions be more than one way of doing this, and some flexibility in contingency planning is appropriate. It would not be helpful to deterrence to define particular options further. The Government however regards the considerations noted in paragraphs 10 and 11 above as important factors in deciding the scale of capability we need.

The wording in paragraph 12 of this extract—‘. . . threat to key aspects of Soviet state power’—was of particular significance, though public commentary mostly did not pick this up and the government did not attempt to underscore it. The language was deliberately chosen—partly with ethical concerns in mind—to convey that, while cities could not be guaranteed immunity, the UK approach to deterrent threat and operational planning in the Trident era would not rest on crude counter-city or counter-population concepts. At the same time, the language sought to avoid the false exclusivity, which used sometimes to be found in strategic commentary, of supposing that a targeting strategy had to choose simply between the destruction of cities and the attempted neutralization of the adversary’s power to retaliate. It has never been appropriate, whether in regard to UK concepts or more widely, to interpret the familiar and sometimesuseful distinction between ‘counter-value’ and ‘counter-force’ in that narrow way.

After the Cold War Following the end of the cold war successive governments of both main political parties made extensive reductions—motivated in significant degree by political awareness of the NPT Article VI concerns noted on page 83—in UK nuclear forces. The force of four nuclear-powered submarines each capable of carrying sixteen Trident D.5 missiles became the sole system component, partly in recognition of the fact that the missile’s range, accuracy, and ability to target warheads independently, together with the scope for varying its weapon loading, conferred considerable flexibility in possible use. Weapons for delivery by aircraft and from surface ships were phased out; the United Kingdom no longer participated in NATO dual-key arrangements involving US-owned weapons; and eventually

126

United Kingdom Doctrine and Policy US weapons for US delivery aircraft ceased to be stored in Britain. 5 A series of cuts in operational holdings of nuclear warheads for the Trident missiles was also announced and implemented, the most recent bringing the declared number to ‘fewer than one hundred and sixty’ 6 —the lowest declared total among the five NPT-recognized NWS. It had long been generally assumed, and was briefly acknowledged by the government in its December 2006 White Paper 7 on the future of UK nuclear capability, that some of the Trident warheads were of an explosive yield deliberately reduced to considerably below that of the standard warhead (itself not formally disclosed, but commonly supposed to be in the order of 80–100 kilotons). The government has said 8 that in the post–cold-war setting it no longer regards terms such as ‘sub-strategic’ as appropriate descriptions of this alternative capability, but it is clear that this is a change of verbal practice, not of substantive policy. There seems no reason to discard concepts of using lower-yield options either in ‘war-termination’ use short of a final strike, or perhaps against targets where lower yield could suffice for the desired destructive outcome and avoid disproportionate collateral effects.

Future Policy Towards the end of 2006 technical estimates—conservatively assessed, given both the small size and the single-system character of UK capability—of limits to the life that could be counted upon for the current submarines, and of lead-times for any replacement, together with the need to keep alive industrial capability to build it, compelled the UK government to face the question of whether, and if so in what form and on what scale, it should envisage the continuance of the nuclear role beyond the early 2020s. The judgements were readily reached that a force based on submarines remained much the best form if capability was to be maintained, and that Trident D.5—with the UK participating at relatively modest cost in the US programme to extend its service life to about 2042—should continue to provide the 5 6 8

See the Guardian, 26 June 2008: ‘US removes its nuclear arms from Britain’. 7 Cm. 6994 of December 2006, paragraph 2.3. Ibid., paragraph 4.9. See footnote 2 to Chapter 2.

127

United Kingdom Doctrine and Policy delivery vehicle. (Notably, the United States has undertaken 9 that the UK will be offered cooperation in any US successor system to Trident, and has said that such a system should be compatible with new UK submarines in the light of the expectation that the life of these would be capable of stretching into the 2050s.) Given these judgements, the immediate question for decision was whether major design work for new boats should be undertaken, as the minimum step needed if the option of timely continuance was to be dependably preserved. Production orders for the boats would not be required for at least a further five years. The government took the view, explained in a substantially argued White Paper, 10 that at the currently projected costs and in a global environment still with so much uncertainty and potential danger, now was not the time to decide to abandon entirely a capability which the United Kingdom had possessed for half a century. Public debate was encouraged, but it never attained the vigour and salience of the 1980s. The government’s view was endorsed by Parliament 11 with majority support from both main political parties. No attempt was made in official presentations, and none credibly could have been, to justify the decision in terms of specific scenarios or adversaries. Nor was any suggestion put forward, as a few voices in outside commentaries had earlier conjectured, that there was an argument related to the ambitions of Europe collectively as a more effective global influence in terms of ‘hard’ as well as ‘soft’ power and to the undesirability, in such a context, of leaving France as the sole European nuclear-weapon possessor (a situation that France itself does not want). The idea that UK capability might have European relevance had a history in Britain reaching as far back as a speech by Edward Heath in 1967 12 —‘held in trust for Europe’. In my view such considerations, though they could not be conclusive, are by no means without weight. It was evident in 2006–7, however, that a European dimension, however lightly indicated and even without any question arising of a collective institutional aspect, would not enhance the decision’s domestic appeal in the climate of British public opinion at the time about the European Union (EU). In addition, 9 Exchange of letters on 7 December 2006 between Prime Minister Tony Blair and President George W. Bush. 10 11 Cm. 6994. House of Commons Official Report, 14 March 2007, col. 398. 12 Godkin Lectures, Harvard University, 1967: ‘Old World, New Horizons—Britain, Europe and the Atlantic Alliance’.

128

United Kingdom Doctrine and Policy the EU’s own agenda was already more than fraught enough without the addition of hints about a subject as sensitive as that of nuclear weapons would be in most member countries. The presentation of the decision reflected the continuing concern—customarily felt more strongly in the United Kingdom than in any other NWS, in part for the reasons indicated on page 119—to be seen to respect the requirements and spirit of NPT Article VI. It noted, in addition to the Article’s lack of specificity, the UK’s good record of reduction and openness, the improbability that UK abjuring of nuclear capability would alter the actions of any other possessor, and the unreasonableness of any supposition that the Treaty imposed a duty of renunciation upon the United Kingdom regardless of what others did. As noted earlier, the strategic case was made in very broad terms. It related in essence to the unsettled and still-anarchic character of the international environment; to the continued intention of the United Kingdom to be a major load-bearing actor in it, and to have the confidence to accept responsibilities and risks accordingly; to the impossibility of predicting dangers far enough ahead for it to be acceptable to defer provision against them until they had become evident; and to the effective finality of any decision to withdraw from the possession of a nuclear armoury. The government’s language presented the decision on continuance as a settled matter. In the March 2007 Parliamentary exchanges, 13 however, it committed itself to being prepared to have Parliament look at the issue again, in the light both of international circumstances and of the latest cost estimates, when the next major decision points were reached—that is, probably not later than about 2013. The central decision of principle might at that stage be significantly influenced by whether the cost estimates remained of the same order as those assessed in 2006–7 14 ; by whether there remained room to accommodate these estimates within a hardpressed defence budget without more damaging displacements in other fields; and by whether the global security environment, the non-proliferation regime, and the international agenda for further

13

House of Commons Official Report, 14 March 2007, col. 284. Cm. 6994 (paragraph 5.11) gave the procurement cost estimates for a force of four new submarines and associated support as £15–20 billion, mostly spread over fifteen years. In this field (unlike that of many other advanced defence capabilities) the UK record of bringing projects to fruition in line with original estimates is good. 14

129

United Kingdom Doctrine and Policy reducing the numbers or risks of nuclear weapons had meanwhile fared well or ill, for example in respect of Iran and its neighbourhood. My own view is that the revisiting of the issue in the light of such factors should be approached seriously. The case for going ahead is not so plainly and unconditionally compelling that it should be taken as entirely beyond reconsideration. But as a matter of political reality—and perhaps national psychology—it does not at present seem probable that any likely UK government would at that point take (or that general public opinion would necessarily welcome) the irreversible step of discontinuance unless one or more of the factors just noted had changed radically and unmistakably. In essence, the stance taken by both main political parties holds that maintaining nuclear capability is an ultimate safeguard against the world security scene’s going badly wrong—as it sometimes has in the past— in ways not currently foreseeable. It further holds, by implication, that the fact that most other countries are debarred by their NPT commitments from taking out such insurance is not a reason why the United Kingdom—which happens for largely historical reasons not to be thus constrained, and which is moreover ready to accept wide-ranging international burdens—should unilaterally renounce an insurance the discarding of which would do nothing to enhance the security of others, and would be most unlikely to prompt emulation by any other possessor. Beyond the basic question, important issues of a more specific kind will arise at the next major decision point. One is whether it remains necessary to envisage a force of four submarines (though a fourth would probably not have to be ordered until several years further on), or whether three would suffice to provide a reasonable expectation of maintaining the continuous operational availability— one submarine always at sea—that has been regarded as essential and sustained without even momentary interruption since the late 1960s. A further submarine-related question—though this was not raised in the 2006 White Paper—might be whether it remains appropriate to have sixteen missile tubes in each boat rather than some smaller number. Given the carrying capacity of the Trident D.5 missile and the declared intention that each submarine should carry no more than forty-eight live warheads, twelve tubes (or even fewer) would seem enough. Having some margin might be an insurance against the long-term possibility that a successor missile to D.5 might be more demanding of hull space to provide a given strike potential,

130

United Kingdom Doctrine and Policy but this would need to be weighed against financial and any other advantages from building smaller boats. A further possible aspect of submarine design concerns whether the submarine propulsion system should be based upon the use of lowenriched (LEU) rather than, as in the current boats, highly enriched uranium (HEU). Exploration of arms control options, whether in the form of a fissile-material cut-off treaty (FMCT) or in a wider context, has sometimes suggested that global verification arrangements sharply constraining the possession of fissile material could find that leaving a loophole for naval propulsion entailed awkward complication. The use of LEU might carry drawbacks of higher machinery noise (increasing submarine detectability and potential vulnerability) and of a need—avoidable with HEU—for mid-life refuelling that could affect levels of operational availability. All this is at present highly conjectural, but it might have to become a factor by the time that the choice has to be made between three and four submarines. The characteristics of LEU propulsion just noted might, if it were adopted, weigh in the scales in favour of four. A separate set of issues is whether—for reasons that might be broadly analogous to those underlying the US administration’s case for a ‘Reliable Replacement Warhead’ (RRW)—new nuclear warheads should be designed (as distinct from mere refurbishment or remanufacture of the existing design), and if so what their characteristics should be, especially in respect of explosive yield. In that last regard there must surely be a case for considering a standard yield a good deal lower than that attributed to the current inventory, against the background of a targeting philosophy which will presumably continue, as paragraph 12 of the 1980 document cited earlier 15 conveyed, to reject a counter-population approach. A reduction in yield might tend actually to enhance deterrent credibility. The United Kingdom has never been directly drawn into armscontrol negotiations about the magnitude or make-up of its nuclear armoury. This has been both because of the armoury’s modest scale and for the reason already indicated on page 94: that since the perceived UK requirement—especially after the move to submarine-based systems which were substantially invulnerable to pre-emption—was in no way a function of the size or character of a potential adversary’s holdings in comparable categories of weapon, 15

Defence Open Government Document 80/23 of July 1980.

131

United Kingdom Doctrine and Policy there was no credible or logical prospect of a useful arms-control deal that would significantly alter the requirement. Successive governments have however regularly reaffirmed—more often and more clearly than most of the other NPT-recognized NWS—the UK’s fundamental acceptance, in the light of the terms of the Treaty and of related undertakings at successive Treaty review conferences, that the eventual goal should be to abolish all nuclear armouries. 16 As footnote 8 on page 83 has recalled, in June 2007 the then Foreign and Commonwealth Secretary, Margaret Beckett, announced that there would be a governmental financial contribution to work in support of a systematic and independent study by the IISS—being undertaken entirely without preconceptions about the outcome for policy—on what would be required, politically and technically, to achieve a nuclear-weapon-free world not less secure for the international community as a whole than that which would otherwise be in prospect. This is further discussed in Chapter 12. In brief, this chapter has recalled the nuclear-weapon history of one of the earliest possessor states, and has outlined its adjustment— perhaps not yet complete—to a post–cold-war world. To partner this portrayal of considerations facing the possessor of a long-mature capability, the next chapter looks at a pair of more recent acquirers. 16 See for example ‘The National Security of the United Kingdom’, Cm. 7291, March 2008, paragraph 4.19.

132

11 Nuclear Weapons in South Asia

My career in government service brought me into scarcely any contact with South Asian matters. After the nuclear explosive tests by India and Pakistan in May 1998, however, I was prompted in various ways to take a particular interest in nuclearweapon issues there, and I paid several visits to the region.

History India had a long record of condemning nuclear weapons and their retention by the five NWS recognized (and, in the Indian view, inappropriately privileged) by the NPT. Indian concern with nuclear technology from the 1950s onwards, however, undoubtedly had in mind possible military application, and India refused to accede to the NPT as an NNWS. In 1974 it conducted an underground nuclear explosive test, claimed at the time and for many years afterwards as having been related to possible peaceful use. This claim—never widely believed outside India—was abandoned in the 1990s, and the test’s relationship to weapon development was acknowledged. Pakistan inevitably matched India’s refusal to accede to the NPT. Already deeply concerned about the bitter and seemingly endless dispute over Kashmir, and scarred by the 1971 break-away of East Pakistan to become Bangladesh—an episode in which India played a part—Pakistan viewed the 1974 Indian test with alarm. The test deepened Pakistan’s determination to press ahead with its own clandestine nuclear-weapons programme, already initiated in 1972. There is little doubt, despite persistent denials by both countries, that China

133

Nuclear Weapons in South Asia helped considerably in this programme, though it is not entirely clear whether or how far this help continued in breach of treaty obligations after China’s accession to the NPT in 1992. It seems certain also that there has been help from North Korea (and perhaps again, though less certainly, from China) with ballistic-missile delivery vehicles or related technology. Though India and Pakistan long maintained the practice of concealment and denial, it became increasingly widely assumed— including by one another—that both of them either had, or were close to having, some operational nuclear capability. This belief and its deterrent effect may have played some part in the fact that confrontations and suspicions in the 1980s and early 1990s did not lead to renewed warfare. Though there were continual unrest and frequent exchange of fire across the de facto Line of Control dividing Kashmir, the two countries did not come openly to blows between 1971 and 1999. In May 1998 India broke nuclear cover, announcing the conduct of five explosive tests. Within weeks Pakistan—which must have had preparations in an advanced state of readiness—announced the conduct of six tests (balancing the numerical account with India’s tally, taken as including the 1974 test). Some external assessments, based mostly on seismological data, questioned whether all the tests, in either set, had achieved the full results claimed for them—India fortythree and Pakistan twenty-five kiloton yield for the largest of their tests—and voices within the Indian scientific community later argued that further testing was necessary to ensure that India had a full range of proven capability, including thermo-nuclear yield. There was no doubt, however, that both countries had proved themselves capable of making nuclear weapons of at least the Hiroshima/Nagasaki destructive power. By no means all opinion, notably in respect of India, regarded what had happened as enhancing the national security of either country. 1 But both were henceforth nuclear-weapon powers, whether the fact were liked or not. The international community—with a nearunanimity that came as an uncomfortable surprise to at least some in India—did not like it. The UN Security Council passed a resolution (UNSCR 1172) condemning what had been done, demanding its 1 For an example of a sceptical view, see Professor Amartya K. Sen, The Argumentative Indian, Penguin, 2006, pp. 251–69.

134

Nuclear Weapons in South Asia effective reversal, and calling for the imposition of punitive sanctions. There was never any likelihood, however, that either country would step back from its course. Significant sanctions were imposed upon both, but in the ensuing decade these were progressively lightened and eventually abandoned as other policy concerns—Pakistan’s value as an ally of the West against international terrorism, and India’s growing global economic and political importance—came to outweigh the anti-proliferation case for continuing to demonstrate disapproval. Two dramatic episodes in the years immediately after the tests were widely seen as trials of how reciprocal nuclear deterrence, now overtly in place, might bear upon political behaviour. In 1999 Pakistan ventured upon a substantial military incursion across the Line of Control, near Kargil. In 2001–2, following a terrorist outrage against the Indian Parliament in the heart of Delhi and then another big terrorist attack in India, with much Indian opinion believing Pakistan to have been at least indirectly complicit in both, there were for several months massive military mobilizations in place on both sides of the frontier between them. There was considerable international concern, especially in the early aftermath of each episode, about the risk of a nuclear conflagration. Commentators on the first one differed about whether nuclear deterrence should be blamed for having failed to prevent it from happening at all (or even for having stimulated it), or credited with having helped—alongside US pressure on Pakistan to withdraw—to avoid its developing into a major conflict. There was rather broader agreement, after the event, that nuclear deterrence had probably helped significantly to prevent the second episode from leading to a clash of arms. In any event, however, it may reasonably be conjectured that the two episodes, in their different ways, amounted to salutary caution-inducing experiences for the learning of lessons by both sides in the early years of open nuclear confrontation, much as the Berlin crisis of 1961 and the confrontation over Soviet missiles in Cuba in 1962 probably did during the cold war. The episodes may also have brought home usefully to the two countries, as well as to others farther afield, that their nuclear relationship was of more than just bilateral concern. One other episode forms an important part of the background to evaluating current nuclear-weapon policies, attitudes, and prospects in the region. In February 2004 it was revealed publicly that

135

Nuclear Weapons in South Asia Western intelligence had uncovered an extensive profit-motivated network, led from within Pakistan, which over several years had sold clandestine and illegal aid in matériel and know-how to potential nuclear proliferators, notably Iran, Libya, and North Korea. Though commercial involvement was by no means confined to Pakistan, the central animating figure—by his own admission (albeit, so he later claimed, under duress)—was Dr. A. Q. Khan, widely revered in Pakistan as the key scientific leader of Pakistan’s own nuclearweapon programme. He was removed from office, disgraced (at least formally), and placed under house arrest, but not otherwise punished. Firm denials of wider governmental awareness or complicity were issued. These were not universally believed. Khan was not made available for questioning by other countries, and much irreversible harm had been done. There is little doubt, however, that the Pakistan government found the affair damaging and humiliating, and that it moved effectively to terminate the activity and guard against any recurrence.

Doctrines Indian and Pakistani pronouncements about the doctrines governing their possession of nuclear armouries have been patchy (as were comparable utterances by NATO and its members in the formative years of the cold war). There have clearly been learning and development processes under way, but published information does not make clear how far they have gone, save in a few limited respects. In India an eminent government-appointed advisory panel delivered in the summer of 1999 a published report 2 with notably wideranging and ambitious proposals for the nature and scope of doctrine and force-planning concepts. Ministers did not endorse its content, however, and an official statement at the beginning of 2003 3 was of considerably narrower scope. Perhaps because the Indian programme had been driven mostly by political and scientific actors, with military participation apparently in a more subsidiary role, such statements as have been made have been primarily about political 2 3

Draft Report of National Security Advisory Board, 17 August 1999. Press release, Indian Cabinet Committee on Security, 4 January 2003.

136

Nuclear Weapons in South Asia and strategic doctrine. Virtually nothing has been said on operational doctrine—that is, on matters in respect of which military planners need guidance, such as deployment, readiness, the use of warning, options and specific objectives for use, and targeting concepts. Official statements have declared that India seeks only a minimum nuclear-deterrent capability. That is not a precise concept, and most other nuclear-weapon possessors would probably claim the same. The statements have emphasized, however, that India does not seek to compete with anyone in the size of the armoury—that the required size is not a function of any other holding. In the immediate aftermath of the 1998 tests some ministers suggested 4 that India’s need for nuclear-weapon capability related primarily not to Pakistan but to China; but such suggestions have not remained current, and potential adversaries have ceased to be specified. This apart, India has from the outset of its declarations affirmed an NFU promise, though (as pages 99–100 have noted) that was later qualified so as not to apply if an adversary used biological or chemical weapons, the possession of which India has renounced. Pakistan has said still less about nuclear doctrine, at least in formal official statements, except that the idea of purporting to guarantee NFU has been explicitly rejected. That rejection cannot be thought surprising. As NATO policy in the cold war showed, it is virtually inevitable that in an environment of political confrontation a country perceiving itself as weaker at the conventional-force level will decline to make such promises (though, as again page 100 brought out, that is not to be construed as implying a positive policy of first use). In other substantive respects the Pakistan government has been silent about doctrine, but some broad principles have become widely understood. 5 In addition, considerable information has been published about governmental organization for dealing with nuclearweapon issues, in respect of both matériel and operations. Perhaps as a reflection of sensitivities in the wake of the A. Q. Khan affair, this information has laid especial stress on thorough arrangements to maintain both physical and personnel security. 4 See for example Defence Minister George Fernandes, reported in the Financial Times, 5 May 1998. 5 See for example Major-General M. A. Durrani, Pakistan’s Strategic Thinking and the Role of Nuclear Weapons, Sandia National Laboratories Cooperative Monitoring Center, July 2004, pp. 23–4.

137

Nuclear Weapons in South Asia

Capabilities Neither country declares officially what operational weapons it has or plans to have, or what types and numbers of delivery vehicles are or will be assigned for them. External speculations about warhead holdings, based mostly upon estimates (also speculative) of stocks of plutonium or HEU, vary widely. At this writing typical guesses have imputed about fifty warheads to Pakistan and slightly more to India. There is however no solid information in the public domain, and conjectures do not uniformly accept even that India’s arsenal is larger than Pakistan’s. In strategic logic, the eventual requirement would need to reflect the doctrine chosen in such respects as the nature of targeting and whether there should be potential to ride out a hostile pre-emptive attack, to mount more than one level of strike, and (for India) to deal with one adversary while maintaining deterrence of potential others. It is probable that for delivering their nuclear weapons both countries would at present still have to rely to a substantial extent upon F/GA aircraft. These need special adaptation for the nuclear role, and neither country has said how many or which types in their holdings are so adapted. It seems likely that among India’s total of nearly 500 F/GA aircraft the preferred choice would be the Mirage 2000, with a radius of action of 600km for a typical unrefuelled operational flight profile. This could cover from Indian bases all major cities in Pakistan, but would give no strategic capability against China. Pakistan has about 110 F/GA aircraft, the farthest-reaching being the F.16. Its typical radius of action, at 1,200km, would cover Delhi and Mumbai but not southern or eastern India. Pakistan has test-fired ground-launched land-attack cruise missiles, and envisages air-launched applications in due course. This could extend target coverage more widely, though it is not declared whether the missiles would be designed to be able to carry nuclear warheads. Both countries have sought increasingly to move towards concentrating capability on land-based ballistic missiles, and have tested systems of varying range. Pakistan currently has some of the 1,500kmrange Ghauri system in service, and has conducted several tests of the Shaheen-2 missile with a design range of 2,000–2,500km. India has the more demanding operational requirement, given that it seeks a deterrent posture in relation also to China. It has tested a 3,500kmrange Agni-3 system to supplement the 2,000km-range Agni-2, and

138

Nuclear Weapons in South Asia has stated 6 that development is under way of a missile with a reach of 5,000–6,000km. Ranges like those of the above-mentioned systems seem plainly to imply nuclear-weapon delivery, and so do those of most other ballistic missile systems held in or tested for the two armouries, but there has been ambiguity about India’s intentions for the 150km-range Prithvi I deployed near Pakistan’s eastern frontier. Neither country possesses sea-based nuclear delivery systems or has them in early prospect, though India is believed to have tested underwater launch of a ballistic missile (albeit of limited range). India has made several successful tests of a high-performance cruise missile—Brahmos, developed in collaboration with Russia—for use from surface ships (as well as from land and, eventually, from aircraft and submarines) but this seems clearly intended for delivering conventional warheads. The 1999 Indian advisory-group report mentioned earlier envisaged an eventual nuclear triad (aircraft, land-based missiles, sea-based missiles) such as the United States and Russia possess, and some Indian commentators have argued that in the long term resources should be concentrated on long-range SLBMs. It will however take many years and heavy expenditure to make such a capability fully operational even if it is made the goal of policy. There is little evidence that Pakistan is at present minded to pursue seabased options with any vigour, though the possibility has not been formally ruled out.

Stability and Security In the immediate aftermath of the 1998 tests concerns voiced by outside observers related not only to damage done to the solidity of the global non-proliferation regime but also to two further fears. The first of these was that the deterrent balance between India and Pakistan might prove unstable, given their difficult relationship and their close proximity—in particular, the fact that crucial Pakistani assets such as the major city of Lahore and key north–south communication links lay close to the border between them and thus to threat from Indian superiority in conventional forces. The second was that weapons 6 Announcement by Defence Research and Development Organisation, 12 December 2007.

139

Nuclear Weapons in South Asia might be held in conditions not fully proof against unauthorized use, or acquisition by terrorists. To the extent that the first fear rested on perceptions of active political hostility between the two countries, it was eased by improvements in the relationship from 2003–4 onwards. Extensive dialogue built up, a ceasefire along the Line of Control held, and the Kashmir problem, though (as disturbances in August 2008 reminded) far from solved, began to seem a less dominant problem than before in the political relationship, partly perhaps because other preoccupations began to bulk relatively larger for each country. ‘Hot-line’ communication links were improved, and both countries openly avowed at foreign minister level 7 that they regarded nuclear capabilities on each side as a force for stability (an avowal that might indirectly have reflected a degree of reappraisal after the crises of 1999 and 2001–2). In strict operational terms the two nuclear armouries have not evidently attained the exceptional—and hugely expensive— standard of insurance for stability that was attained between the main cold war adversaries, with even far-fetched scenarios of hostile pre-emptive strike meticulously guarded against. In practice, however, the developing nuclear-force configurations of the two South Asian countries do not suggest serious risks of instability in crisis. Especially as debate and habituation have deepened understanding, neither can currently suppose either that it has credible options for disarming the other, or that the other might reasonably think that possible. The second fear has related to the in-country security of weapons and key materials. Both countries maintain that they give very high priority, and great effort, to this. External observers rarely express worries about India, but Pakistan is another matter, partly in the aftermath of the A. Q. Khan affair and partly because of perceptions of uncertain political stability and, more particularly, of conjectures about the strength of militant Islamist sentiment even within government and the armed forces. It is doubtless to assuage such worries that, as noted earlier, Pakistan has provided a good deal of information—more than India has—about its organization, resources, and procedures for ensuring the security of the armoury. 7

140

Joint statement following New Delhi meeting, 29 June 2004.

Nuclear Weapons in South Asia In the wake of UNSCR 1540 in 2004 mandating all UN members to ensure that they have effective safeguards against leakages of WMD or related materials to non-state actors, both countries have moved to systematize their legislation and supporting structures to meet the imperative. It will be important, as both surely recognize, that they continue to provide substantive reports on the practical working of their arrangements to the UN Commission tasked, on behalf of the international community, with global oversight of the responses to UNSCR 1540.

Looking Forward In both countries strategic-affairs communities, both in and out of government, have since the 1998 tests notably deepened their grasp of the issues raised by the possession of nuclear weapons. It is not to be expected that they will view with automatic reverence the suggestions of outside commentators formed in a different environment. But the long saga of the cold war—the world’s first experience, and ultimately a successful one, of managing the existence of nuclear weapons—may still have points of relevance, even if these are not everywhere acknowledged. For Pakistan, concern about deterrent stability inevitably relates above all to the security interface with India. The Indian outlook is not the converse of that—India understandably sees itself as having to weigh wider aspects in its security policy, even though these are no longer illustrated in the nuclear field by explicit governmental references to other neighbours. The interface with China is, however, quiescent, despite the continued disagreement about sovereignty over the large Aksai Chin territory where India suffered a sharp military defeat in 1962. In current circumstances operational deterrence has to be evaluated primarily in the context of the relationship with Pakistan. Concerns about peaceful stability in that context have become less acute and issues about nuclear weapons less salient in recent years, both because of the 2003–4 rapprochement and because— albeit for different reasons—in each country the focus of national attention has moved away from matters contentious between them. But amicable relations are not irreversible. They might be at risk, for

141

Nuclear Weapons in South Asia example, if frictions over Kashmir were to sharpen again, whether related to Pakistani perceptions of Indian foot-dragging over political progress or to Indian ones of Pakistani complicity in terrorist activity. Other unsettling possibilities might arise if India decided to deploy operationally a BMD system which some in Pakistan then interpreted as designed to undermine Pakistan’s deterrent capability; or conceivably if Indian leaders came to suppose (or Pakistani ones suspected that they intended) that the improvements recently undertaken in the readiness of their conventional forces, in the programme known as COLD START, would provide usable new options of a limited-war kind for punitive response to unwelcome behaviour by Pakistan. There is therefore a case—especially in the continued absence of stable long-term agreement on the status and governance of the whole of Kashmir, which would be by far the biggest possible contribution to stability—that it would be prudent to use the period of comparative calm deliberately to build a basis for enhanced confidence that would be robust against mishaps or adverse stresses. It may be unlikely that formal arms-control agreements about force levels or weapon-system types, of the cold-war US/Soviet kind, can have much part to play. The strategic asymmetries between the two countries are wide, and there is no reason to disbelieve statements by each of them that they do not regard the scale or character of their requirements as being a function of the other’s armoury. Nevertheless, long-term confidence would be strengthened—and the prospect of holding the nuclear armouries at modest levels valuably improved—if it were found possible to establish a practical and dependable common understanding of how best to manage the relationship of mutual nuclear deterrence calmly and economically. A considerable amount of dialogue at multiple levels has taken place. There has been agreement to observe a moratorium on any further nuclear explosive tests, and several confidence-building measures (CBMs) have been put in place. These however are peripheral to the central issue of deterrent balance, and broader proposals advanced by each country at one time or another have not made shared headway. Each country must surely accept, whatever might be thought about past history and comparative political or social merits, that the other is entitled to exist as an independent and prosperous state within recognized borders (without prejudice, in that last regard, to their

142

Nuclear Weapons in South Asia deeply divergent positions over Kashmir). There must also be a shared recognition that nuclear war would be a vast and intolerable calamity for both of them—and for the whole region—in environmental terms as well as in direct and massive loss of life in crowded areas. This would far outweigh any benefit that nuclear conflict could conceivably yield even on the most sanguine assumptions about the likely course of events or about national resilience. Some candidate themes for building on such foundations of common understanding are sketched in the following paragraphs. Each country has a right to secure deterrence—that is, to be confident that the other has no easy options for inflicting grave harm on its vital interests by force. In other words, each is entitled to feel assured that it could, if ultimately necessary, impose penalties that would outweigh any benefits which forcible change to the status quo damaging its vital interests might have hoped to achieve. This might be unpalatable to some Indian opinion, which rejects a facile equation between the two countries and legitimately notes that India is a much larger state with broader concerns and a more complex international environment, notably though not only in its interface with China. But the search for bilateral stability to which both countries profess commitment can scarcely prosper without a pragmatic acceptance of at least this minimum of balance and equivalence, if necessary alongside a recognition of India’s wider context. Each country should be enabled to feel confident that the other has neither the intention nor the capability to deprive it of deterrent sufficiency, whether by pre-emptive strike, by warding off retaliation, or by a combination of these. This follows from the preceding principle. To fulfil it each country needs to have an adequate understanding of the other’s capabilities and plans. Facilitating this for the long term may come to require more deliberate openness than either has so far shown. (The openness may indeed sometimes need to reach beyond merely verbal assertion—in situations of potential confrontation countries may legitimately be reluctant to rest their security on assurances which they cannot in some way verify.) Some Pakistani commentators have voiced concern about the long-term objectives or effects of Indian decisions, especially on BMD, though worries on that score have not been uniformly shared in Islamabad. India has undertaken

143

Nuclear Weapons in South Asia substantial development work and testing towards a BMD system, 8 but no governmental decision seems yet to have been taken on whether, how, on what scale, and at what cost it might be deployed operationally. Nor has there been any clear official indication of the strategic concepts that might, in India’s political and geographical circumstances, underlie any such deployment—to protect what assets, against which possible adversary countries, with what likelihood of operational success (given, for example, that ballistic missiles are not the only possible method of attack), and with what expectation of or allowance for likely reactions and counter-moves by such countries. These seem important dimensions for decision-makers to weigh. Each country should be enabled to maintain nuclear sufficiency as economically as possible. Both countries face massive economic, social, and environmental demands upon national resources. Expenditure on nuclear forces, as on other defence purposes, therefore carries heavy opportunity costs. It is moreover in the interest of neither that pressures should be intensified (especially upon Pakistan, as having the smaller and less robust economy) that might damage domestic stability, or lead to risk-taking or under-provision on nuclear security and safety. Even if the historical thesis were to be accepted, as only a minority of commentators have maintained, that the economic pressures of keeping up militarily with the United States helped to undermine the Soviet Union and its Communist system, it would be deeply dangerous to regard that as a relevant or useful model for gaining advantage in the circumstances of South Asia. Whether or not either country chooses to make politically binding statements about when or how it would or would not use nuclear weapons, each should seek, so far as is consistent with maintaining security at reasonable cost, to reduce to the minimum the possibility that either might feel impelled to have early recourse to nuclear weapons. As has already been noted, it cannot be expected that Pakistan will emulate Indian NFU declarations, since that would imply willingness to acquiesce in any defeats, however far-reaching the consequences, which India’s conventional superiority might be able to impose in Pakistan’s vulnerable 8 Statement by V. K. Saraswat, Chief Controller, Defence Research and Development Organisation, reported by Agence France Presse, 13 January 2008.

144

Nuclear Weapons in South Asia geographical circumstances. It is nevertheless, as NATO increasingly recognized, damaging to stable deterrence if the nuclear threshold— determined primarily by conventional-force relativities, not nuclear ones—appears to be markedly low, or to be set by use-or-lose dilemmas resulting from risks that nuclear systems might be either disabled or (because of deployment near the front line of military encounter) overrun at an early or low-level stage of conflict. Each country should so frame both its nuclear-force dispositions and its concepts for possible use that top-level and considered political control of all nuclear forces remains assured on both sides at every stage of a crisis or conflict. This implies recognition that it would be against the interests of both countries if either were at risk of having to take nuclear-release decisions hastily or on imperfect information, or to delegate crucial decision-taking to lower levels of command. It also implies that both should be very wary of adopting in targeting doctrine and planning, whether at the conventional or the nuclear level, decapitation concepts designed to eliminate or disable key elements of the other’s central control system. Swift and dependable communications both between political leaders and between military commanders on both sides should remain always immediately available before and throughout any armed conflict. This partners the preceding theme. Communications need to be securely available in two dimensions: vertically, to maintain assured control within each country’s system even in crisis conditions of high-readiness force dispersal, and laterally, so that even in the heat of conflict each could still talk to the other at key levels to maximize the chance of averting escalation to ultimate catastrophe. Each country should, so far as practicable, so design its arrangements that maintaining the secure availability of its nuclear capability does not require that in time of crisis steps have to be taken that could appear provocative or open to misinterpretation. The factors for keeping nuclear forces in a peace-time posture that facilitates ready political control and minimizes costs and accident risks (e.g. by keeping warheads and delivery vehicles widely separated), and those for having them operationally ready and invulnerable in time of crisis, pull against one another in some ways. There is no easy formula either for striking the balance, or for guaranteeing that in transition from one posture to another steps genuinely intended

145

Nuclear Weapons in South Asia as prudent defensive measures of protection are not seen by the other side as possible preparations for attack. At the least, however, the potential tensions and sensitivities need to be recognized, and options for managing them—perhaps by direct communication— carefully weighed. Each country should so design its arrangements as to keep to a minimum any risk of accident involving nuclear warheads, or of accidental or unauthorized launch of nuclear-capable delivery systems; and should notify the other immediately if any such event nevertheless occurs. Both countries should take account of any lessons available from the experience of others, notably during the cold war, when there was never an accidental nuclear explosion or an accidental or unauthorized strategic missile launch in a hostile direction. The NWS recognized by the NPT should be as forthcoming as possible, within the constraints set by the Treaty and without prejudice to their wider views on nuclear developments in South Asia, in sharing their security and safety experience and expertise. (There have been circumstantial press reports 9 of extensive US help to Pakistan in this field, and it is to be hoped that Pakistan will use any such help to the full.) It is welcome that India and Pakistan have concluded a significant agreement 10 on reducing the risks of accidents and communicating swiftly if they do happen. Much if not indeed all of what has been outlined above has by now become well understood among key figures in both India and Pakistan—it is offered here just as a contribution to a checklist. A general issue for consideration is whether, and if so how, a set of such understandings might be more securely entrenched on a shared footing, given that nuclear stability is a fundamental and enduring common interest irrespective of the general political climate at any particular time. Dialogue cannot always be entirely divorced from that climate. During the cold war, for example, the initiation of the first SALT negotiations between the United States and the Soviet Union was set back several months by the Soviet action in Czechoslovakia in 1968, and the Soviet invasion of Afghanistan in late 1979 halted US ratification of the second SALT deal. But so far as possible, businesslike action in the pursuit of the stability that is a shared 9

New York Times, 17 November 2007. ‘Agreement on Reducing the Risk from Accidents Relating to Nuclear Weapons’, signed in New Delhi, 21 February 2007. 10

146

Nuclear Weapons in South Asia interest should not be made hostage to the ebb and flow of general political relations, or used as a vehicle for unrelated political demonstration. (Once the SALT I bargains had been established in 1972, the work of the joint commission supervising their implementation and observance continued almost independently of other US/Soviet dealings.)

Non-proliferation A separate key question is how the two countries should see their relationship to the global non-proliferation regime centred on the NPT. That question was given added and awkward prominence by the widely criticized—albeit as vigorously defended—deal for nuclearenergy cooperation between the United States and India agreed in principle at the meeting of President George W. Bush and Prime Minister Manmohan Singh in July 2005 and signed at a further meeting between them in March 2006. On any view, whatever its overall political and strategic merits, this deal sat uncomfortably with the terms of UNSCR 1172, passed at the time of the 1998 tests, and more broadly with the strict logic of a uniform worldwide nonproliferation regime. The deal needed acceptance by legislatures in both countries (as well as a degree of acquiescence by the IAEA and the NSG, as the prime international bodies concerned with proliferation). It was long stalled by internal political difficulties in India. These were eventually overcome, however. IAEA and NSG acceptance was secured, and the US Congress approved the agreement. It is effectively impossible for India and Pakistan formally to accede to the NPT. They will plainly not consider joining as NNWS, and amending the Treaty to admit them as NWS is certain to be nonnegotiable with almost all other parties. They might however consider committing themselves explicitly to act fully, in matters covered by relevant articles about not transferring matériel or know-how, as though they were parties to the Treaty. (France took such a position for over twenty years before acceding to the Treaty in 1992.) The matter could probably not be formally expressed in quite this manner, since both countries have maintained a position of basic disapproval of the Treaty as being discriminatory. Given the will, however, ways could surely be found of expressing clear commitment to the substance of the articles.

147

Nuclear Weapons in South Asia A commitment of this kind might valuably be extended also to the content of other proliferation-related arrangements such as the MTCR and the NSG—indeed, the terms formulated for the US/India deal on nuclear energy contained such an undertaking, in line with domestic legislation which India had already enacted. Both India and Pakistan might declare an intention to respect the various NWFZs, though the terms of the relevant treaties mostly do not allow for formal accession or association by states not party to the NPT. They might also consider the possibility of some declaration about limiting their production of fissile material, though it is unlikely that the build-up of their armouries has yet reached a point where they would be ready to commit themselves forthwith to joining a fissile-material cut-off treaty (FMCT). There is however little prospect that such a treaty will be in place for some years to come, and the two countries might be better able to contemplate accession by the time that it materializes. It is probable also that neither country would wish to remain the final obstacle to the entry into force of the CTBT—for which their ratifications are required—once all other necessary ratifications had been given. And both countries have already adopted export control laws matching the best practice of NPT signatories. In the round, action on the lines sketched in the two preceding paragraphs might both enhance the good name of the two countries themselves and contribute positively to confidence in the global nonproliferation regime. The successful continuance of that regime is in their interest no less than that of others, as they have indeed clearly recognized. 11 That interest, moreover, may increasingly require that the two countries find ways—and that others should help them to do so— of engaging positively as stake-holders in the future improvement of the regime in respects such as those reviewed on pages 84–6. Devising such ways may call for diplomatic flexibility and ingenuity, given that their stance towards the NPT means that they could not easily take a direct part in the work of the 2010 review conference. But it is plainly desirable, given their weight within the international system, that their involvement and responsibility-sharing should go beyond mere conformity with the present constraints of the regime. In particular, both countries will surely profit if the reshaping of the 11 See for example working papers submitted by India and Pakistan under Agenda Item 4 of the Disarmament Commission session in New York, 9–27 April 2007.

148

Nuclear Weapons in South Asia regime in respect of access to cooperation in nuclear energy can create a new context in which such access can be available to them without encountering the regime-related awkwardnesses that beset the 2005 US/India project. In brief, the avowed entry of India and Pakistan into the ranks of nuclear-armed countries has, after early disquiets, developed less alarmingly than many observers had feared. Their proximity entails for them both, despite wide asymmetries, crucial common concerns that need deliberate and perhaps collaborative attention. They also share a deep security interest in the long-term future of nuclearweapon issues worldwide. Part IV now considers that future.

149

This page intentionally left blank

Part IV The Path Ahead

This page intentionally left blank

12 The Abolition of Nuclear Armouries?

The Abolitionist Goal As Chapter 7 has explained, there is a widespread global commitment in terms of political declarations—the formal legal position is less clear-cut, save in a remote context of general and complete disarmament—to the eventual abolition of all nuclear armouries. The idea has a history reaching back to 24 January 1946, when at its inaugural meeting in London the UN General Assembly set up an Atomic Energy Commission and tasked it with making proposals for ‘the elimination from national armaments of atomic weapons and of all other major weapons adaptable to mass destruction’. Regular reaffirmations of this goal can be heard from many states. There have also been, from time to time, high-level suggestions of giving real political impetus to its pursuit. Examples are to be found in the Reagan–Gorbachev dialogue at Reykjavik in 1986 and in the speech by Indian Prime Minister Rajiv Gandhi to the UN General Assembly on 9 June 1988. With a few notable exceptions, however, and despite the recurrent rhetoric, the subject for a long time attracted curiously little examination, at least by states, at a level that could be regarded as of truly serious quality, depth, and objectivity. There was a wide divergence observable—for the most part it scarcely deserved to be called a debate—between two polarized extremes. One pole, which might be called that of the righteous abolitionists, pointed to the commitment and insisted that countries possessing these weapons ought to get on, more or less forthwith, with disposing of them; there were regular demands that abolition negotiations be initiated without delay. The other pole, that of

153

The Abolition of Nuclear Armouries? the dismissive realists, asserted that complete abolition was fanciful dreaming. They recalled the truism that nuclear weapons can never be disinvented, and maintained that the world must expect to have to concentrate on managing their existence for the rest of human history—or perhaps, to put the matter slightly less starkly, that successful abolition must imply an international environment so vastly different from that of the present or the foreseeable future that it was idle to spend time now on talking about it. Both of these viewpoints were wrong. The righteous abolitionists tended to talk of giving up nuclear weapons as though it were a sort of international equivalent of giving up smoking—the kind of thing that any sensible and strong-minded state ought to be able to do without long-drawn-out shilly-shallying. But this ignored the fact that states do not acquire or retain a nuclear armoury, with all its costs and other drawbacks, just as a matter of idle whim. They do so for reasons centred upon, even if often by no means confined to, their national security (and in at least three instances—Israel, Pakistan, and Russia—it is far from self-evident that their perception of the case for possessing nuclear weapons would necessarily be reversed if others gave them up). One may think such reasons in this or that or indeed every case to be mistaken or overrated, but they cannot be simply brushed aside. Demands for early negotiation on abolition are little more than posturing if the political conditions do not yet credibly exist and the groundwork has not been laid for it to have a chance of success. (The Berlin Wall was always plainly deplorable, but to have demanded in the 1970s that negotiations for its demolition should be embarked upon would have been near-frivolous.) In January 2006 Pope Benedict XVI assured the world that the idea that nuclear weapons could contribute to security was ‘completely fallacious’. 1 The Pope’s words customarily command widespread attention, even in matters of practical judgement where the Vatican has no inherent or distinctive expertise. But no analytical exegesis was offered, and mere assertion cannot suffice. The dismissive realists were wrong because whether or not it may now be believed that the implication of a long-term goal of eventual abolition conveyed in the NPT and subsequent declarations at its five-yearly review conferences was unwise or unreal, it was a goal 1 Message of Pope Benedict XVI for the Celebration of the World Day of Peace, 1 January 2006.

154

The Abolition of Nuclear Armouries? clearly accepted, notably though not only at the 1995 conference which agreed on the unlimited extension of the Treaty’s duration. It cannot be shrugged off as simply pious dreaming. It has often been reaffirmed and invoked, and it continues to be relied upon as a load-bearing component in the set of bargains that constitutes the global non-proliferation deal, the deal that is the best and indeed only generally accepted international regime that exists for controlling the spread of nuclear weapons. The longer possessor countries continue to act, or are thought to be acting, as though ‘eventual’ meant something like ‘contemporaneously with the abolition of all evil in the world’, the greater the danger that this element of the multi-part deal centred upon the NPT will cease to bear its load, with peril to the entire regime. And the peril may go wider. As Chapter 6 has discussed, the risk of nuclear-weapon use is at any one time extremely low. The fact that the world has come through over sixty years without its happening is not just a benignly miraculous fluke, and there can be no meritorious ground for occasional attention-seeking claims that use is near-certain within some specified period. But the probability of use, whether by states or by terrorists, cannot be zero. However low it may be thought to be at any particular time it cannot, in the multiplication of the passing years and decades, be regarded as merely trivial, especially if the much wider spread of nuclear energy— which must for several reasons be highly likely—puts at least some of the potential ingredients of weapon capability in increasingly numerous hands. That some of these hands may be unstable adds to the concern. Given all this, there is a need for cool and careful examination—if possible neutrally approached rather than driven by pre-determined campaigning preference in the direction of either pole—of what an acceptable non-nuclear world that was at least as satisfactory in other respects as today’s (or, to express more accurately the comparison which policy-makers have to address, as the world that is judged likely to exist in future if nuclear armouries are not abolished) would look like, and what would be needed to create and sustain it. The words in the United Kingdom government’s 1981 Defence White Paper 2 still seem apposite: 2 ‘Statement on the Defence Estimates’, April 1981, Cmnd. 8212–1, p. 14. The full text of the section in which this appears is given in Appendix 1 to this book.

155

The Abolition of Nuclear Armouries? Any readiness by one nation to use nuclear weapons against another, even in self-defence, is terrible. No-one . . . can acquiesce in it comfortably as the basis for international peace for the rest of time. We have to seek unremittingly, through arms control and otherwise, for better ways of ordering the world. But the search may be a very long one . . . and impatience would be a catastrophic guide . . . .

Article VI of the NPT does not explicitly call for complete nuclear disarmament, but understanding to that effect was affirmed unanimously at the 1995 review conference, and again—very clearly— at the 2000 conference as part of the programme to which the five NWS pledged themselves politically (though at the 2005 conference the US administration of President George W. Bush and the government of France declined to reiterate it). UK governments have repeatedly, for over twenty years and most recently in the White Paper of December 2006 on the future of UK nuclear-weapon capability, 3 declared their adherence to the ultimate abolitionist goal.

Paths and Conditions Examination of what would have to be done to achieve a reasonably secure non-nuclear world might fall initially into two parts. One part concerns what might be termed the disarmament mechanisms, the other the political conditions. The latter constitute if anything the more important and difficult segment of the task. In January 2007 a highly interesting and weighty statement 4 on nuclear abolition made by George Shultz, Henry Kissinger, William Perry, and Sam Nunn— not a group of natural nuclear peaceniks—rightly attracted considerable attention. Their initiative (which was strongly reaffirmed a year later 5 ) was intended as a major push towards taking abolition seriously, but the text was primarily about the disarmament-mechanisms 3 ‘The Future of the United Kingdom’s Nuclear Deterrent’, Defence White Paper, December 2006, HMSO Cm. 6994, paragraph 2.2. 4 ‘A World Free of Nuclear Weapons’, Wall Street Journal, 4 January 2007. In Britain a comparable group later issued a joint statement in similar if slightly more guarded terms: ‘Start worrying and learn to ditch the bomb’, Douglas Hurd, Malcolm Rifkind, David Owen, and George Robertson, The Times, 30 June 2008. 5 Public statement of 15 January 2008.

156

The Abolition of Nuclear Armouries? element, with only one point in a set of eight addressing the politicalcontext element that is ultimately cardinal. US President Harry Truman once said 6 : Let us not become so preoccupied with weapons that we lose sight of the fact that war itself is the real villain.

The memoirs of Javier Solana’s great-uncle, Salvador de Madariaga, a distinguished figure in League of Nations disarmament striving between the wars, comment 7 : The trouble with disarmament was (it still is) that the problem of war is tackled upside down and at the wrong end . . . Nations don’t distrust each other because they are armed; they are armed because they distrust each other. And therefore to want disarmament before a minimum of common agreement on fundamentals is as absurd as to want people to go undressed in winter . . . Disarmers would avoid wars by reducing armaments. They run to the wrong end of the line. The only way . . . consists in dealing day by day with the business of the world . . . the true issue is the organisation of the world on a cooperative basis.

The most dependable path to abolishing unwelcome security apparatus, whether in the form of weapons or of walls like that which once divided Berlin, is to bring about the political change that would make it irrelevant. But as the Berlin Wall’s persistence for over a quarter of a century illustrated, such change is often not achievable easily or in predictable timescales. Against that background, the existence of nuclear armouries is widely judged to have made a crucial contribution to the prevention since 1945 of conflict on the scale of the two World Wars. It is nowadays legitimately noted that nuclear deterrence is essentially about preventing armed collision between major states, and that on current indications conflict in the twentyfirst century seems generally unlikely to take that form (though nuclear deterrence may well itself have something to do with that unlikelihood). It is undoubtedly true, and welcome, that the shadow of nuclear weapons has lost much of the dark immediacy it seemed to have 6 President Harry S Truman, Mr. Citizen, Bernard Geiss Associates, New York, 1960, p. 267. 7 Professor Salvador de Madariaga, Morning Without Noon, Westmead, Saxon House, 1973, pp. 48–9.

157

The Abolition of Nuclear Armouries? during the cold war, and that deterrence based on them mostly plays a much more limited and less forward part among the instruments for managing international affairs. It would however be shallow, unhistorical, and potentially dangerous to assume now that global interdependence and sixty years of freedom from that level of conflict have so irreversibly altered mindsets and perceptions of interest that the continuance of such freedom can henceforth be taken for granted without the nuclear contribution yet at the same time without need to consider, before it is discarded, what ought to be done to reinforce non-warlike instruments within a dependable toolbox for ensuring the freedom. The need for such a toolbox is, on a long view, made all the more compelling by the likelihood that, as pages 9–10 have suggested, the march of science will place in the hands of advanced states other weapons of a destructive effect similarly capable of rendering war between such states utterly intolerable. To insist that nothing should be done in military and technical terms towards the eventual abolition of nuclear armouries until a long and ambitious list of political ameliorations has been securely achieved would be neither necessary nor reasonable. But the two streams of action cannot be wholly separated. Some degree of parallel advance is essential.

Political Change The political-context element of the striving toward nuclear-weapon abolition might be subdivided into two parts. The first sub-division concerns particular disputes of a grave and long-lasting character, the type of issue to which the Shultz group referred by implication in general terms—the Israeli/Palestinian problem, India/Pakistan issues especially over Kashmir, perhaps also the relationship of Taiwan to China. It seems absurd to imagine that key actors, especially Israel and Pakistan, will be found willing permanently to scrap their nuclear-weapon insurance unless the relevant dispute has either been resolved or else somehow reduced to a condition in which, rather as with Greco-Turkish disagreements over Cyprus and still more between Spain and the United Kingdom over Gibraltar, all parties can be confident that major war is dependably absent from the options that either side might even consider. This chapter makes no attempt either to propose or to predict the means by which, or the timescale

158

The Abolition of Nuclear Armouries? in which, such situations might be attained. The point for present purposes is simply that all this has to be part of any serious agenda for global nuclear abolition. It might be added, even though the matter does not turn on any particular territorial dispute, that it would be necessary also to think hard about what would have to change in respect of Russia—in her political condition and attitudes, and in her security environment— to induce her to give up the nuclear armoury that now constitutes probably the most significant remaining feature enabling her to feel in some sense still a special international power. This armoury has moreover received if anything greater emphasis, not less, in the adaptation of Russian defence concepts to the changed postSoviet setting, and (unlike in the United States) the introduction of new weapon systems is in hand. That changed setting includes unrest at several points close to her enormous borders as well as pressure from Chinese immigration into eastern Siberia, alongside grave weaknesses in the bulk of Russian conventional forces. Russia moreover is perhaps the most vivid exemplar of the problem that the massive and versatile conventional military strength which leads some US voices to suggest that the United States might be well suited by a non-nuclear world—Defense Secretary Les Aspin, for example, speculated along these lines in the early 1990s 8 —may lead others to feel that that is precisely why they would be disadvantaged by it. They may not easily forgo the ultimate equalizing power of nuclear weapons. The second main sub-division of the political-context element is less specific, but perhaps in the long run not less important. As the foregoing discussion has noted, it is reasonable to judge (even if impossible to prove) that nuclear weapons have played a large part in the remarkable absence of war between major advanced states since 1945. That absence has been a colossal blessing to the whole world, not just to those states themselves. We should not lightly tinker with the structures which seem to have helped to achieve it. Nuclear weapons, bringing the unmistakable reductio ad absurdum of all-out war between advanced states, have meant that such states have been compelled, like it or not, to accept that such warfare is permanently off the table—that it has to be absolutely excluded 8 I recall this from conversation at an informal meeting in the margin of a NATO conference in 1993.

159

The Abolition of Nuclear Armouries? from the menu of options which they can entertain for resolving or managing disagreements among them. And the nuclear-abolition agenda has to consider what would have to be changed in other respects, from the world we live in today, to achieve that hugely valuable exclusion as reliably by less disagreeable and less costly means, whether or not with a continuing armed-force component. That may be both intellectually and practically the hardest part of the whole enterprise. What would we have to envisage, short of the utopian notion of a world government? We may be optimistic that old-fashioned territorial disputes of the Palestine/Kashmir/Taiwan kind are largely now a matter of historical legacy. Those three have been with us since the 1940s, and scarcely any comparable new ones involving major advanced states have arisen since then save perhaps in the wake of the break-up of the Soviet Union. But it would be a very sanguine analyst who predicted that there would never again in future be disputes of similar severity over, say, water, energy supplies, or other natural resources such as key scarce minerals, migration (perhaps driven by climate change or demographic pressure), or humanitarian outrages. So we might well need more dependable and universal agreement on rules of international behaviour, procedures for handling disputes, and instruments (whether global or regional) for enforcing conformity with those procedures. Detailed suggestions are not attempted here, but by way of example one might envisage that the UN community would need to make more headway than its 2005 summit meeting managed to do both with Security Council reform and with the effective entrenchment of internationally agreed concepts to discipline recourse to military force, on lines such as those put forward in the powerful 2004 report by the eminent international study group established by Secretary General Kofi Annan. 9 In a world still made up of sovereign states there can never be absolute assurance of conformity with rules and their enforcement. It has been said that one defining feature of sovereignty is the ultimate ability to default on one’s commitments; and uncertainty about readiness to abide by done deals does not relate only to ‘rogue’ states, or to small ones. But we can strive for better probabilities of respected order than exist today, and the striving should be seen as a more important 9 Report of the Secretary-General’s High-level Panel on Threats, Challenges and Change, December 2004: ‘A more secure world: Our shared responsibility’.

160

The Abolition of Nuclear Armouries? task than some of the key actors at the 2005 summit appeared to accept. The need for stronger political arrangements, and better probabilities of obedience to them, would not be merely in order to prevent the re-emergence of old-style major conventional war. If, in a non-nuclear world, technologically advanced countries came seriously to blows over what they regarded as vital interests, the temptation might be powerful—if only as insurance against breakout by the adversary—to acquire or reacquire nuclear weapons, with all the dangers that could then flow from a competitive rush to rearm amid the pressures of immediate crisis or conflict. It is sometimes suggested that the very fact of this reconstitution risk could serve as a deterrent to war—weaponless deterrence, it has been called, a sort of deterrence at one remove. But that implies a worldwide and longsighted wisdom on which it would surely be imprudent to count unless there had been real, systematic, and dependable advance in structures, methods, and attitudes for the handling of disputes.

Disarmament Mechanisms The disarmament element of the necessary agenda was given fuller treatment in the Shultz group’s declaration, though their specific suggestions were mostly the familiar arms-control agenda such as the CTBT and an FMCT. These are important projects both individually and, still more, in the aggregate, and they are amply worth pursuing in their own right. But while they would helpfully clear some of the ground for abolition, and much improve the political climate for it, they do not actually lead to it at all closely. The key fact is that having no nuclear weapons is politically, strategically, and technically a very different thing from having a modest number, or de-emphasizing them in policy and doctrine. Carrying through the current set of arms control ideas, or other complementary concepts such as improving still further the technical ability to achieve by nonnuclear means military tasks previously thought to require nuclear explosive power, all have merit. But complete abolition would have to have, quite aside from the wider political aspects touched upon earlier, a more radical agenda on the disarmament-mechanisms side. It would need, in outline example, to have plans for at least three aspects:

161

The Abolition of Nuclear Armouries?

r First, identifying accurately the starting baselines of existing capability, and defining what long-term denuclearization was required to mean—what matériel, apparatus, and facilities must no longer exist.

r Second, devising worldwide verification arrangements to provide all countries (not merely a few technologically advanced ones) both with adequate and lasting assurance, technical and political, of that non-existence, and with evidence that could not be ignored if break-out was attempted.

r Third, devising a path and a timetable by which current weapon possessors were to move to abolition without at any stage in the process creating new instabilities perceived as damaging to their security. As the discussion of minimal armouries on pages 105–8 has suggested, the risks to stability and confidence might be at their highest as numbers of weapons fell very low and the proportionate effect of imbalance or evasion became more significant. Establishing in concrete terms how all this might be done, and still more agreeing the result internationally, would be a formidable task. Comparatively little of it has as yet been undertaken, especially by governments. Even if the eventual political environment for its comprehensive application is judged remote, there is a strong case for seeking to carry exploration forward so that implications, problems, and options are understood thoroughly and in good time. This book does not seek to consider the disarmament-mechanisms segment of the agenda in any detail. 10 Two preliminary comments may however be pertinent. The first is that it would surely be essential, both for the global legitimacy of the abolition regime at its inception and in order to enhance the prospect of united enforcement action against subsequent defaulters, that verification arrangements should be universal and non-discriminatory—no exemptions because, for example, our country is ‘the special country’ or ‘the good guy’, or has lucrative commercial secrets to protect, or a legislature apt to make difficulties. The second comment is that perfect and assured certainty in every aspect of definition and verification is bound to be unattainable, however massive the international 10 For a substantial discussion see IISS Adelphi Paper 396, Abolishing Nuclear Weapons, by George Perkovich and James Acton, September 2008.

162

The Abolition of Nuclear Armouries? effort—already foreshadowed by UK proposals for exploratory discussion among NWS weapon laboratories 11 —that comes to be devoted to developing systems and putting them in place. (That development and its implementation would on any view be enormously demanding, both in resources and in the time needed to design and agree them and then carry them to completion.) The technical apparatus, structures, and procedures for an abolition regime would therefore have to be complemented by a substantial and worldwide measure of trust that could flow only from extensive political changes in the directions sketched earlier in this chapter. Even the disarmament-mechanisms as distinct from the political segment of the task would itself inevitably raise many issues with a political dimension. It might turn out that the process of moving together towards the goal valuably reinforced confidence, but awkward questions would remain unavoidable. Would key countries, especially current nuclear-weapon possessors, be prepared to live with the inescapable margin of technical uncertainty in verification? Would they be prepared to shoulder a fair share of the costs? Would they be prepared to accept possible restrictions upon economically useful activities which, even when genuinely not intended for weapon-creating purposes, had to be prohibited because they were capable of being so used? Would they be prepared to accept a depth of intrusive inspection almost certainly exceeding anything that previous arms-control agreements have entailed? Would they be prepared to guarantee participation in tough enforcement action against breaches even in instances that were politically or economically uncomfortable, and moreover to trust such guarantees given by others? Still further issues not examined here would have to come into consideration. For example, the nuclear-weapon prospect might well stand to be influenced by what happens over BW and CW. It would not be easy to achieve denuclearization unless the effective prohibition of those was generally regarded as secure. That goal seems at present within less difficult reach; but its entrenchment would have to be made dependable for an environment where the lastresort underpinning of nuclear deterrence, which has in the past facilitated the political shaping of arms control agreements with 11 See speech by UK Secretary of State for Defence Des Browne to the UN Conference on Disarmament, Geneva, 5 February 2008.

163

The Abolition of Nuclear Armouries? imperfect verification or sometimes even with none, would no longer be available. And though it would be important to avoid overloading with ancillary issues an agenda already very formidable, questions might also arise about weapons in space or anti-missile defence, or, in the longer term, about new technologies of destruction not yet generally visible. There might well be difficult questions also—their weight perhaps depending upon how successfully and dependably there had been put in place more benign arrangements for the management of disputes—about whether, in order to maintain deterrence of adversaries or wrong-doers, greater conventional military power (costing more in both money and manpower) would have to be provided. In situations of political mistrust—as perhaps in South Asia, for example—wide divergences in conventional strength might become more troubling, as they would have done in Europe during the cold war.

The Case for Study The discussion in this chapter does not rest on any naive supposition that the abolitionist quest could be the sole or even the prime motor of all the political changes that seem desirable in order to make the goal securely achievable. In many respects it could make no more than a useful motivating contribution to those. The central point, however, is simply that the theme of abolishing nuclear weapons is one on which there is broad and serious analytical work to be done, and work moreover upon which, or at least upon the need for which, widely different viewpoints could initially converge—whether it be the righteous abolitionists hoping to prove that abolition is less intractable and less distant than sceptics suppose, or the dismissive realists expecting to demonstrate that there is no worthwhile ground for thinking that it will ever happen, or those in the middle inclined to believe (as I do) that the goal has to be taken seriously but will entail, at best, a long, difficult, and as-yet-uncertain road. The aim of study would be in the first instance not to establish or advocate a programme of action or to inaugurate a negotiation, but simply to lay a better foundation of understanding upon which debate about prospects, options, and possible path-clearing work might be advanced.

164

The Abolition of Nuclear Armouries? In brief, there is complex and demanding work to be done, by believers and non-believers alike, on the idea of abolishing nuclear armouries altogether. At best, however, that outcome could not be achieved swiftly. The next and final chapter considers what more might be done in the meantime—whether that meantime is judged to be of indefinite or merely long duration—to limit costs and risks.

165

13 The Practical Agenda

Serious exploration of the idea of completely abolishing nuclear armouries, on the lines discussed in Chapter 12, is merited and indeed necessary in its own right (and the more nuclear-armed states that engage in it, the better). But it can also serve a salutary indirect purpose. It is certain to bring out that on any objective assessment, however sanguinely approached, the world must expect to have to live for decades to come with the reality of nuclear armouries. Neat prediction is plainly impossible, but few informed commentators would be likely to rate at better than fifty-fifty the chances of their being entirely dissolved before, say, the centenary in 2045 of the Hiroshima and Nagasaki catastrophes. Their continued existence will therefore, on any view, have to be carefully and durably managed for a long time yet. The near-term prospect is not, in the round, gravely worrying, and headline-seeking rhetoric to the effect that we stand on the edge of an abyss risks discrediting and damaging the pursuit of what is reasonably and soberly achievable. Nevertheless, significant risks remain. Alleviating these with minimum cost and friction entails a wide, complex, and in some ways difficult agenda. Elements of the agenda can certainly help in an advance towards the abolitionist goal, and some would indeed be essential for preparing the way there. But the agenda needs to be tackled whether or not one believes in the realism of that goal—optimists and sceptics can find common ground. And it cannot responsibly be evaded or downplayed on a claim that it is of merely temporary relevance, or that the international community ought to devote all its political attention and energy to higher and more ambitious aims about nuclear armouries.

166

The Practical Agenda This chapter surveys in broad outline potential components of that agenda, some though not all of them already noted in earlier chapters.

The Non-proliferation Regime Chapter 7 has brought out that the nuclear non-proliferation regime centred upon the NPT, even if not globally precarious, is under threat of being weakened in three direct ways. The first is the risk that North Korea or Iran, or both, might escape entirely from the regime— and, still more damagingly, might do so without having to pay a manifest and lastingly heavy price. At this writing the prospects of confining or rolling back North Korea offer some encouragement, even though the regime’s remarkable negotiating history of promise– postpone–evade is plainly not yet ended and its activities in external proliferation have continued to arouse unease. Its environment is, however, less beset by conflict and tension than the Middle East, and though its behaviour is far from unimportant it is accordingly, in the breadth and weight of potential repercussions should there be an unsatisfactory outcome, the less worrying of the two countries. There are deeper grounds for concern about Iran. It is a much more important state than North Korea in size, resources, and regional influence. Under its current regime it is not a dependably statusquo-accepting participant in international affairs—both the extremist declarations of its president against the existence of Israel and its substantial support for violent non-state actors like Hezbollah and Hamas, and probably also in Iraq and Afghanistan, testify to that. The ‘high-confidence’ assessment by the US intelligence community published in December 2007, to the effect that Iran had no current nuclear-weapons programme, was, even if correct, 1 far from a certificate of innocence. It conveyed that Iran had indeed earlier had such a programme, in flagrant contravention of its NPT commitments, and that there was no assurance against its resumption in some guise. There had been other concealments from the IAEA, also seriously 1 Unusually, the estimate was later publicly (albeit informally) challenged as oversanguine by intelligence partners of the United States. See Financial Times, 6 March 2008, ‘US under fire for intelligence report on nuclear plans’.

167

The Practical Agenda in breach of Treaty obligations, and there had been clandestine dealings with the A. Q. Khan network. Iran continued selectively to evade key questions 2 posed by the IAEA (which formally voiced ‘serious concern’ 3 ) and refused to conform with explicit obligations about its activities repeatedly imposed by the UN Security Council. It was insisting, in the teeth of those requirements, on maintaining an indigenous uranium-enrichment programme when credible other paths of assured supply for peaceful energy purposes had been offered. It was not easy to judge, still less to prove unarguably, what were the long-term intentions of a government structure in which there were probably divergences of view and unsettled opinions at least as regards emphasis or priority, within an intellectual environment that had little apparent tradition of cool, expert, or open-minded strategic analysis. Iran’s stance was presented by its leaders as no more than the exercise of a right to achieve full national independence in the generation of nuclear energy. Such a right is not itself prohibited by the NPT; the absence of precautionary qualifications to it is, as pages 85–6 have noted, one of the current vulnerabilities of the non-proliferation regime. But whether or not achieving that was genuinely the limit of the purposes of the Iranian leaders, its effect would be at least to carry Iran to a threshold weapons capability—that is, a condition in which the timescale for developing an operational nuclear-weapon armoury, if at some later stage that were decided upon, would be relatively short. (Iran seemed already to have in service or in early prospect, and had indeed brandished in testing, ballistic-missile delivery vehicles with ranges in the order of 2,000km.) Such an effect, if unchecked, would both damage the global non-proliferation regime and risk provoking similarly damaging emulation by other countries in a volatile region where Iran and its possible aspirations are viewed with a good deal of misgiving. The judgement given by the US intelligence community helpfully weakened arguments for direct and early pre-emptive military action, whether by the United States or by Israel. Such action would almost certainly have conferred at best no more than temporary benefit, 2 See for example New York Times, 3 March 2008, ‘Vienna Meeting on Arms Data Reignites Iran Nuclear Debate’. 3 See New York Times, 27 May 2008, ‘Nuclear Agency Accuses Iran of Wilful Lack of Cooperation’.

168

The Practical Agenda at the cost of gravely harmful wider and more lasting repercussions. But the case remained strong both for finding internationally agreed ways—preferably inducements, or else the prospect of penalties, or a combination of the two—to persuade Iran to change course, and for considering soberly and realistically (even if privately) how best to manage and confine the damage if Iran did go ahead. It is not easy to see how the prospects of a satisfactory outcome can be maximized unless a broad dialogue, not over-hedged about with prior conditions, is undertaken between Iran and the United States as well as leading members of the EU, to lighten the regime’s sense of wronged beleaguerment, to assuage its more extreme apprehensions, and to challenge its unwillingness to give weight to the legitimate concerns of others and to its international obligations. The second direct danger to the long-term strength of the nonproliferation regime is that general weaknesses within its present working will continue unremedied. These weaknesses relate, as Chapter 7 has outlined, to shortcomings in the coverage of its safeguards and in the rigour of their verification; to the comparative ease, and freedom from penalty, with which parties can formally withdraw from the Treaty; and to the absence of satisfactory and globally accepted arrangements reconciling Treaty-recognized entitlement to the benefits of nuclear energy for civil purposes with avoiding the risks of facilitating clandestine or threshold weapons capability. This last may well be, in the long run, the most important of the weaknesses to tackle. A thorough, generous, and non-discriminatory package offer of help with nuclear energy, by the widest possible grouping of technologically capable states, might maximize the chances of persuading other states both to accept the inevitable constraints and to cooperate in condemning and penalizing any intransigent holdouts such as Iran appears to be. The third danger, linked closely to the second one, is that there might be in 2010 a harmful repetition of the virtual failure of the 2005 Review Conference of parties to the Treaty. The avoidance of such an outcome must be a high priority for the key parties, especially the United States and the other Treaty-recognized NWS. A successful conference, besides easing particular problems, would reinforce the authority of the regime as a whole and the deterrent likelihood of its enforcement against breaches; and another unsuccessful one would have the opposite effect.

169

The Practical Agenda These issues aside, the NPT-centred regime remains strictly incomplete, in that three important states—India, Pakistan, and Israel— have never been participants in it. There is no credible way of bringing them formally into the regime. They will not accede to the Treaty as NNWS, and the existing parties to it will not admit them as additions to the five acknowledged NWS. There would however be advantage in having India and Pakistan (Israel’s total silence on nuclear-weapon matters seems unlikely to be broken, and Israel is in any event now a very unlikely proliferator) declare explicitly that in all matters relating to the transfer of know-how or matériel they would act fully as though they were bound by all the relevant features of the non-proliferation regime, on the lines sketched earlier on pages 147–8. Such declarations, especially if partnered by action of the kinds outlined on pages 142–6 to enhance and entrench nuclear security and stability in their own relationship, could valuably relieve the disquiet which the international community has felt about their open emergence as nuclear-weapon powers. Action on these lines might also make a contribution, even if a modest one, to easing the widespread misgiving which the US/India agreement on nuclear energy generated in relation to the consistent application of the nonproliferation regime. There remains the question of nuclear proliferation to terrorists. Page 74 has noted the scope for putting in place, with useful if limited deterrent effect, systems to help ensure that the source of any transfer or leakage of fissile material from states to terrorist use would not go unidentified. In addition, there is still much to be done to fulfil the mandatory provisions of UNSCR 1540, which the Security Council passed in 2004 under Chapter VII of the UN Charter. All member states were required to put in place, operate effectively, and report upon legal and other arrangements throughout their jurisdictions to prevent the acquisition or movement of WMD-related matériel by non-state actors. At this writing the reporting (and, it must be suspected, the underlying action) is a long way from being globally complete. The life of the special commission established to receive implementation reports on behalf of the Security Council has had to be extended beyond what was originally planned. It seems evident that collective political pressure for the thorough fulfilment of the resolution’s requirements needs to be maintained and perhaps even intensified. If, as is sometimes suggested, some of the smaller or

170

The Practical Agenda less well-resourced states have difficulty with the task, better-placed others should surely offer help.

Arms Control and Disarmament by the Nuclear-Weapon States The effective working of the non-proliferation regime—including the tackling of the Iran problem—and still more its strengthening on the lines indicated above will continue to impose substantial demands and constraints upon the non-nuclear parties to the NPT. As pages 83–4 have discussed, their willingness to accept all this cannot be divorced from what the five Treaty-recognized NWS do to reduce the size and salience of their armouries and to lessen any risks which the retention of those may continue to entail. (There was an occasional inclination within the US administration of President George W. Bush—and also within the French government—to seek to establish a disjuncture between these two streams of action, but this cannot be credibly sustained either historically or politically in the light of the Treaty’s Article VI and of repeated subsequent declarations to which all the NWS have subscribed. Both streams are necessary for the continuing legitimacy of the Treaty-centred regime.) Four of the NWS have already done a great deal by way of disarmament, and there might well be advantage in their seeking more positively to make the facts of what they have done widely understood. 4 China, the fifth, is an exception, but that is understandable given its modest inventory (which indeed at one time it publicly claimed 5 to be the smallest among the five) and its lack of allies, as well as its uncertainty—against the background of concerns such as the possible future scope, purpose, and effect of US BMD deployment— about whether the United States truly accepts a stable relationship of mutual nuclear deterrence. These factors mean that China is likely 4 For a convenient summary of what the United States has done, see statement by Dr. Christopher A. Ford, US Special Representative for Nuclear Non-Proliferation, 17 March 2007. On 18 December 2007 the White House stated that the US nuclearwarhead stockpile at year-end would be less than one-quarter of its level at the end of the cold war. The United States has now no programmes to provide new capabilities in either warheads or delivery systems. 5 Dr. Jeffrey Lewis, ‘The Ambiguous Arsenal’, Bulletin of the Atomic Scientists, May/June 2007, pp. 52–5, cites a statement by the Chinese Foreign Ministry in April 2004 that ‘Among the nuclear weapon states China possesses the smallest nuclear arsenal’.

171

The Practical Agenda to remain reluctant either to provide concrete information about its capabilities (though it makes much of claimed openness about doctrine for their use) or to engage in negotiated arms-control constraints upon them. There remains however considerable room to do more, even if one believes—as I do, for reasons summarized on pages 105–9—that ideas of minimal or ‘virtual’ armouries are misconceived save possibly as temporary devices on an agreed path towards complete divestment, and that in themselves they would be more likely to erode than to enhance confidence in stable security even if it proved possible to surmount the formidable difficulties which their complexity would pose for multilateral negotiation. The political need for manifest and substantial advance in disarmament will be all the more necessary to the extent that the nuclear powers decide—however legitimately in their own terms—to undertake ‘continuance’ projects such as Russia’s extensive modernization programme, the United Kingdom’s construction of new missile-carrying submarines, and US pursuit of its RRW plan for a new warhead. (It is salutary that the US Congress has made clear, and RRW proponents are understood fully to accept, that if that plan goes ahead it should be on the basis that no new military mission or increase in explosive yield is contemplated, that explosive testing would not be sought, and that substantial reductions in reserve holdings of weapons would be facilitated.) The CTBT, still lacking the ratifications needed to bring it into force, and the projected FMCT are high-profile projects which the great majority of NPT parties desire and expect to see tackled and brought to fruition. Even those observers who (again like me) are not wholly persuaded that in terms of cold strategic logic they carry all the weight that enthusiasts habitually ascribe to them must recognize that they have come over the years to carry a global political importance which by now has force in its own right. Their future inescapably bears upon perceptions of whether NWS are serious about reducing the salience and risks of their armouries, and therefore upon the willingness of NNWS to continue meeting their own responsibilities within the overall regime. And even if diverse difficulties continue to impede bringing the two treaty projects to full fruition, there is valuable scope meanwhile for nuclear-weapon possessors to exploit further the possibilities which undoubtedly exist to begin realizing the objectives of the projects through strengthened moratoria and similar political undertakings, whether unilateral or

172

The Practical Agenda multilateral, about the activities which the projects are intended to terminate or constrain. Most of the reductions in nuclear armouries since the end of the cold war have been made by unilateral national choice. That remains the speediest and least complicated path for action, and it can certainly be used further—perhaps even within the United Kingdom’s small force, in the limited respects touched upon on pages 130–1. There is still, however, a role for international agreements. They may have particular value in persuading Russia— conscious of uncertainty and unrest around its enormous borders, and also of nuclear weapons as being still a special factor in its postSoviet international standing—to accept additional limitation and cut-back. The 2002 Moscow Treaty between the United States and Russia had notable imperfections as an arms control measure, as page 96 has explained. It is for several reasons to be hoped that a more lasting and rigorous successor treaty can be negotiated. Some exploratory discussion to that end between the two countries is understood to have taken place already. Two general questions arise about the prospects for a new bargain. The first concerns the chill which the events of August 2008 in Georgia cast over US/Russian political relations. This was certain to impede dialogue for some time ahead, and it soon led the US administration to withdraw from congressional consideration 6 the deal on nuclear energy that had been agreed between the two governments in April 2008. Nevertheless, if they both come to perceive worthwhile shared interests in seeking a deal, the impediment need not be total or long-lasting. The 1963 PTBT was signed less than a year after the Cuban Missile Crisis, and though the launch of SALT I discussions was set back by the Soviet invasion of Czechoslovakia in August 1968 preparatory work continued and serious negotiations, eventually fruitful, began fifteen months later. The second general question, however, concerns whether there will be, especially on the Russian side, any perception of worthwhile shared interest in a substantial new bargain. Russia is now, in its international dealings, more confident—more combative, even—than it was in 2002 when it agreed the Moscow Treaty (and made little fuss about US withdrawal from the 1972 ABM Treaty). It may be aware also 6 See Financial Times, 9 September 2008, ‘US puts Moscow Nuclear Deal on Hold Amid Chilly Relations’.

173

The Practical Agenda that there is little sign of new US nuclear-weapon programmes which it might want to see capped. Nevertheless, there is or should be a common interest both in the reinforcing contribution which further disarmament could make to the non-proliferation regime, and in the long-term predictability which arms-control constraints provide. At the least, therefore, a new US administration should be ready to put Russian willingness to the test. The Moscow Treaty envisaged a level of 1,700–2,200 for each country’s operational stock of strategic weapons. It said nothing about either non-deployed reserve stocks or systems classed as nonstrategic—that is, of less than inter-continental range. Constraint upon reserve stocks may not easily be achieved, given verification difficulties and—perhaps still more—uncertainties, at least on the US side, about the long-term reliability of operational holdings unless and until warheads of new design, on the lines of the RRW project, can be expected to be generally available. Further reduction in operational strategic holdings, possibly to a specific limit somewhere in the order of 1,000–1,200 on each side, might be a proper objective. It is hard to envisage circumstances in which such a level would not suffice to give both parties stable and credible deterrence in all likely directions, especially if—as seems likely—the scope is by no means yet exhausted for further developing conventional capabilities to execute military tasks once thought to require the explosive power of nuclear weapons. It would be important also, for confidence and as evidence of resolve to sustain disarmament momentum in line with Article VI of the NPT—and perhaps also for greater assurance of weapon security— that non-strategic systems be brought into account and under verification. As pages 96–7 have noted, though Russia publishes no information about these it is generally believed to have several times the US holding. It would be very desirable to reach agreement on full transparency by both sides about such systems, and on equal (and low) levels of them. It might well, however, be difficult to conclude any agreement on those lines—where movement would have to be mostly on the Russian side—without some US movement to meet Russian concerns on one or both of two counts which Russia would almost certainly see as pertinent: the continued forward holding of US nuclear weapons in Europe for NATO, and the planned US deployment in Central Europe of defences against ballistic missiles.

174

The Practical Agenda On the first of these issues, the presence of US warheads in Europe breaches no law or agreement, and NATO has no need to apologize for it. But it is difficult now to see any compelling security or political case for their retention. Quite aside from the fading of potential threat to NATO since the cold war ended and the Soviet Union and its client alliance were dissolved, the United States plainly has in its inventory delivery systems that could swiftly bring nuclear weapons to bear if necessary at any point on the globe without the use of forces and equipment permanently based in Europe. The original political aspects of the case for the deployment—a combination of the desire to underscore US commitment to the defence of its European allies, and the merit of involving several of those allies in the realities and responsibilities of nuclear deterrence—no longer have great force. The warhead holding is currently believed to have been reduced to around 300, with perhaps half of these earmarked for non-US delivery systems under dual-key arrangements. Withdrawal of this stationed holding might moreover relieve European allies participating at present in those arrangements of awkward and costly decisions ahead about the replacement of the aircraft which they assign to this role, within over-stretched budgets and amid pressure for better contribution to collective European defence effort in other ways. It need do no harm to Alliance cohesion and confidence provided that its presentation evidently and explicitly maintained three key principles. The first would be that the full military capability, both conventional and nuclear, of the United States remained unequivocally committed to the defence of all NATO allies, and to the agreed NATO strategic concept (whether the present one or a new one if, as is understood to be under consideration, a revised version is commissioned in 2009 or thereabouts when the Alliance marks the sixtieth anniversary of its foundation treaty). The second would be that allies for their part maintained their full acceptance of the strategic concept, including its nuclear aspects, and took steps to keep alive their understanding of those aspects. The third would be that there was no question of declaring any part of Alliance territory a nuclear-weapon-free zone into which weapons could not be reintroduced if necessary for security and deterrent confidence should circumstances change radically. There seems no reason why affirmation of any of these principles should be found difficult.

175

The Practical Agenda The second issue on which movement by the United States might help in inducing Russia to accept cut-back and constraint in nonstrategic nuclear forces is that of US plans to station elements of a BMD system in Central Europe. Russian objection to plans for such stationing may reflect a general sense of resentment at new Western military deployments in areas that were once part of Russia’s own protective glacis. More concretely, it has rested upon suspicion that, while the plans have been explained by the United States as motivated by concerns about potential Iranian capability to threaten the West with nuclear weapons, their effect if not indeed their hidden intent might be to lay a foundation for later build-up that could erode the deterrent power of Russia’s own armoury. That suspicion, even if it be believed unreal or exaggerated in the light of the disparity between the scale of the Russian inventory and that of the proposed BMD deployment, must be likely to impair Russian willingness to reduce or constrain its offensive holdings. There is room to question the cogency and urgency of the security arguments for the BMD plans. Even if Iran does acquire significant operational nuclear-weapon capability—an eventuality that is, at worst, not in early prospect—massive Western deterrent power would still bear down upon its wielding, especially in the directions to which the BMD plans could be relevant. Given that reality, shelving the plans— or accepting a cooperative US/Russia BMD arrangement of some kind that was less than ideal in purely US operational terms, or shaping the plans in phased ways that would make advance to full operational capability dependent on the emergence of clearer threats—would not seem a disproportionate concession for a new US administration (perhaps less committed doctrinally and politically than its predecessor has been to the BMD plans) to make in return for Russian flexibility on non-strategic weapons, and perhaps also for more wholehearted Russian commitment to achieving a satisfactory outcome over the Iranian problem. There had indeed (albeit before the falling-out over Georgia) been preliminary signs 7 of possible compromise over the BMD issue. And neither of the policy shifts just postulated seems likely to be strongly opposed by the European members of NATO to whose security they primarily relate. European support for the BMD plans expressed at

7

See Financial Times, 7 April 2008, ‘US–Russia Optimism on Missile Shield Deal’.

176

The Practical Agenda NATO meetings 8 is likely to have been a reflection less of positive enthusiasm for them than of a desire not to disoblige the United States or appear to give in to Russian hectoring. Another aspect of behaviour by the nuclear powers that sometimes attracts censure is that—albeit in degrees that vary widely from one to another—public information about the size and composition of their armouries is more limited than critics would like. Transparency is not an automatic good in every situation. Uncertainty or ambiguity may sometimes be positively helpful to deterrent stability, especially for small or not-yet-matured armouries, where the absence of precise knowledge may usefully dispose potential adversaries to overestimate capability, or where full disclosure of the nature, number, or deployment of delivery systems and warheads might be thought to heighten vulnerability to pre-emption. To put the same point another way round, the loss of secrecy might drive states with armouries of small size or limited quality to adopt higher readiness states, or to expand or diversify their capabilities. This last might be notably so in respect of China, especially if it remained apparently the case that the United States is not entirely ready to acquiesce in mutual deterrence. In the South Asian situation precise public information might moreover prompt domestic pressures for emulation, for example, if India were to discover that (as is sometimes conjectured) Pakistan actually had the larger holding of weapons. Nevertheless, in most situations transparency is a positive factor for stability between potential adversaries or competitors, and for confidence across the non-proliferation scene as a whole. Britain and France have made full statements of their force levels. 9 It might be a helpful further contribution to confidence if both of them were voluntarily to commit themselves to not exceeding the declared levels, and also to accepting international verification of such a commitment in line with whatever scrutiny arrangements may come to be agreed in a new arms-control bargain between the United States and Russia.

8 North Atlantic Council summit declaration, Bucharest, 3 April 2008, paragraphs 37–8. 9 The United Kingdom has regularly made considerable information available, most recently when announcing further stockpile reductions in Cm. 6994 of December 2006. In a speech at Cherbourg on 21 March 2008 President Nicolas Sarkozy also announced reductions, and for the first time declared the size of France’s warhead holding.

177

The Practical Agenda

Doctrines and Postures There are several ways in which modifying or clarifying nuclearweapon doctrines and postures—even if done without reciprocal agreement or parallel action by others—might contribute usefully to reducing perceived salience and risk. One example is the matter of first use, discussed in Chapter 9. As page 101 has explained, while new purported promises of NFU would have little or no merit, it would be advantageous if as many as possible of the Treatyrecognized NWS were to make it clear that they now regard as highly unlikely the eventuality of their needing to contemplate first use, and that they would seek so to shape their policies, postures, and operational doctrines as to keep that eventuality as securely excluded as possible. Some collective NATO declaration on these lines (and perhaps also explicit repudiation of hostile caricatures about ‘policies of first use’ such as those noted on page 100) might usefully feature in the next public updating of the Alliance’s strategic concept. In much the same spirit, possessors could explicitly abandon any remaining special arrangements for keeping nuclear weapon systems on short-notice alert, and still more for any concept of launch on warning or launch under attack. Even given a judgement (as pages 67–9 have suggested) that the real risk of mistaken missile launch under such arrangements is remote, the circumstances in which such arrangements would actually need to come into play are now so improbable as no longer to warrant the political and presentational costs of maintaining them, for example in the face of speculative worries (even if these too are thought to be misplaced) about ‘cyberinterference’. In other words, the relative priority as between availability for prompt and certain nuclear use if necessary and evident assurance against precipitate or mistaken such use can surely be shifted in favour of the latter. There may perhaps remain a case for maintaining capability for short-notice action with conventional weapons, but that scarcely extends to the nuclear field. Chapter 4 has noted that, over the span of the nuclear age, there has been a progressive narrowing of the range of military roles in which the destructive power of nuclear weapons might be thought necessary in order to compensate for shortcomings in the accuracy or lethality of conventional weapons. The United States, in particular, has by now a formidably weighty, varied, and far-reaching

178

The Practical Agenda array of conventional capabilities for identifying major targets precisely and then striking and disabling them. It is increasingly hard to believe that deliberately maintaining nuclear weapons for any purely military function remains necessary on any significant scale. The repeated refusal of the US Congress to approve expenditure for research on new ‘bunker-buster’ nuclear capabilities points away from such ideas. The politico-strategic judgement implied by that refusal—a judgement which I would share—is that the range of circumstances in which, and of purposes for which, the availability of specialized operational capability of this kind would be crucial for either use or deterrence is too narrow and improbable to warrant incurring the political costs of developing new types of weapon at a time when the international community is looking, especially in the non-proliferation context, for further reduction in the salience of nuclear weapons. Perfect foreknowledge is not feasible about what might conceivably be found desirable in some extreme scenario, and formally absolute renunciation of particular options might be neither credible nor salutary. As Chapter 3 has explained, so long as nuclear armouries continue to exist there have to be some serious concepts for their ultimate use if their contribution to preventing war is to be sustained at all. There is a case, however, that in all NWS nuclear deterrence concepts, operational doctrine, and force provision should in future be geared essentially to the purposes of sending a final political warning to discourage an offending regime from continuing aggression, and thereafter—in the very last resort—of rendering it unable or very unlikely to reoffend. That ultimate stage might well include military assets among those to be targeted, but the aim would not be classical military victory or pre-emptive disarming. Such a reorientation or narrowing would move still further away than Western NWS have already done from interpretations of mutual assured destruction as envisaging the virtual demolition of societies, and would regard keeping the collateral loss of innocent life to the minimum possible as a crucial consideration. This might have implications not only for shrinking the size of future armouries but also for modifying their characteristics in such respects as reduced explosive yield.

179

The Practical Agenda

Resolving Political Disputes It is beyond the scope of this book to explore how the disputes and fears which lead states to believe that they may need nuclear weapons for security might be resolved or allayed. Amicable and durable settlement of contentious problems like those over Taiwan, Kashmir, or the Israeli/Palestinian confrontation would be likely to do more than any of the agenda items surveyed above to ease humanity’s anxieties about nuclear weapons. So too would advances in the rules, authority, instruments, and efficacy of the United Nations or other international bodies for managing without armed conflict any present or future inter-state disputes that could not be wholly resolved. For the future of humanity it is almost certainly more important, and probably more urgent, to work at dissolving the causes of conflict than at abolishing its weapons, though the two paths need sustained and parallel effort. Issues directly concerned with nuclear armouries, like those sketched in this chapter, present an extensive and highly important set of challenges, which need to engage the energetic attention of the international community. But it would be a disservice to peace if those concerned with such issues—whether as upholders or as critics of the existing nuclear order—allowed it to be supposed that they constitute the entire or even the prime agenda for reducing the risks and burdens of the irreversible nuclear discovery.

180

APPENDIX 1

The following essay appeared in the UK government’s Statement on the Defence Estimates 1981 (Cmnd. 8212-I, pp. 13–14)

Nuclear Weapons and Preventing War Nuclear weapons have transformed our view of war. Though they have been used only twice, half a lifetime ago, the terrible experience of Hiroshima and Nagasaki must be always in our minds. But the scale of that horror makes it all the more necessary that revulsion be partnered by clear thinking. If it is not, we may find ourselves having to learn again, in the appalling school of practical experience, that abhorrence of war is no substitute for realistic plans to prevent it. There can be opposing views about whether the world would be safer and more peaceful if nuclear weapons had never been invented. But that is academic; they cannot be disinvented. Our task now is to devise a system for living in peace and freedom while ensuring that nuclear weapons are never used, either to destroy or to blackmail. Nuclear weapons are the dominant aspect of modern war potential. But they are not the only aspect we should fear. Save at the very end, World War II was fought entirely with what are comfortably called ‘conventional’ weapons, yet during its six years something like fifty million people were killed. Since 1945 ‘conventional’ war has killed up to ten million more. The ‘conventional’ weapons with which any East–West war would be fought today are much more powerful than those of 1939–45; and chemical weapons are far more lethal than when they were last used widely, over sixty years ago. Action about nuclear weapons which left, or seemed to leave, the field free for non-nuclear war could be calamitous. Moreover, whatever promises might have been given in peace, no alliance possessing nuclear weapons could be counted on to accept major nonnuclear defeat and conquest without using its nuclear power. Non-nuclear war between East and West is by far the likeliest road to nuclear war. We must therefore seek to prevent any war, not just nuclear war, between East and West. And the part nuclear weapons have to play in this is made

181

APPENDIX 1 all the greater by the facts of military power. The combination of geography and totalitarian direction of resources gives the Soviet Union a massive preponderance in Europe. The Western democracies have enough economic strength to match the East, if their peoples so chose. But the cost to social and other aims would be huge, and the resulting forces would still not make our nuclear weapons unnecessary. No Western non-nuclear effort could keep us safe against one-sided Eastern nuclear power. An enormous literature has sprung up around the concepts of deterrence in the nuclear age. Much of it seems remote and abstruse, and its apparent detachment often sounds repugnant. But though the idea of deterrence is old and looks simple, making it work effectively in today’s world needs clear thought on complex issues. The central aim is to influence the calculations of anyone who might consider aggression; to influence them decisively; and, crucially, to influence them before aggression is ever launched. It is not certain that any East–West conflict would rise to all-out nuclear war; escalation is a matter of human decision, not an inexorable scientific process. It is perfectly sensible—indeed essential—to make plans which could increase and exploit whatever chance there might be of ending war short of global catastrophe. But that chance will always be precarious, whether at the conventional or the nuclear level; amid the confusion, passions, and irrationalities of war, escalation must always be a grave danger. The only safe course is outright prevention. Planning deterrence means thinking through the possible reasoning of an adversary and the way in which alternative courses of action might appear to him in advance. It also means doing this in his terms, not in ours; and allowing for how he might think in future circumstances, not just in today’s. In essence we seek to ensure that, whatever military aggression or political bullying a future Soviet leader might contemplate, he could not foresee any likely situation in which the West would be left with no realistic alternative to surrender. Failure to recognize this complicated but crucial fact about deterrence— that it rests, like a chess master’s strategy, on blocking off in advance a variety of possible moves in an opponent’s mind—underlies many of the criticisms made of Western security policy. To make provision for having practical courses of action available in nuclear war (or for reducing its devastation in some degree by modest civil defence precautions) is not in the least to have a ‘war-fighting strategy’, or to plan for nuclear war as something expected or probable. It is, on the contrary, a necessary path to deterrence, to rendering nuclear war as improbable as we humanly can. The further evolution in 1980 of United States nuclear planning illustrates the point. The reason for having available a wider range of ‘non-city’ target options was not in order to fight a limited nuclear war—the United States repeatedly stressed that it did not believe in any such notion—but to help ensure that even if an adversary

182

APPENDIX 1 believed in limited nuclear war (as Soviet writings sometimes suggest) he could not expect actually to win one. The United Kingdom helped to develop NATO’s deterrent strategy, and we are involved in its nuclear aspects at three main levels. First, we endorse it fully as helping to guarantee our security, and we share in the protection it gives all Alliance members. Second, we cooperate directly, like several other members, in the United States’ power which is the main component of the nuclear armoury, by making bases available and providing certain delivery systems to carry United States warheads. Third, we commit to the Alliance nuclear forces of various kinds—strategic and theatre—under our independent control. The details of all this are matters of debate, which the government welcomes. But the debate should recognize that positions which seek to wash British hands of nuclear affairs, while continuing (as NATO membership implies) to welcome United States nuclear protection through the Alliance, offer neither moral merit nor greater safety. Whether we like the fact or not, and whether nuclear weapons are based here or not, our country’s size and location make it militarily crucial to NATO and so an inevitable target in war. A ‘nuclear-free’ Britain would mean a weaker NATO, weaker deterrence, and more risk of war; and if war started we would if anything be more likely, not less, to come under nuclear attack. The East–West peace has held so far for thirty-five years. This is a striking achievement, with political systems so sharply opposed and points of friction potentially so many. No one can ever prove that deterrence centred on nuclear weapons has played a key part; but common sense suggests that it must have done. Deterrence can continue to hold, with growing stability as the two sides deepen their understanding of how the system must work and how dangers must be avoided. Not since the Soviet gamble over Cuba in 1962 have we come anywhere near the brink. It is entirely possible, if we plan wisely, to go on enjoying both peace and freedom—that is, to avoid the bogus choice of ‘Red or dead’. To recognize the success of deterrence is not to accept it as the last word in ensuring freedom from war. Any readiness by one nation to use nuclear weapons against another, even in self-defence, is terrible. No one—especially from within the ethical traditions of the free world, with their special respect for individual life—can acquiesce comfortably in it as the basis of international peace for the rest of time. We have to seek unremittingly, through arms control and otherwise, for better ways of ordering the world. But the search may be a very long one. No safer system than deterrence is yet in view, and impatience would be a catastrophic guide in the search. To tear down the present structure, imperfect but effective, before a better one is firmly within our grasp would be an immensely dangerous and irresponsible act.

183

APPENDIX 2

This text was originally written in 1984, and was published in 1987 as Annex A to an article in Theological Studies, No. 48 (1987), ‘The Ethics of Nuclear Deterrence: A Critical Comment on the Pastoral Letter of the U.S. Catholic Bishops’.

The Strategic Use of Nuclear Weapons 1. The judgement is often accepted, readily or reluctantly, that strategic use of nuclear weapons—taken here to mean use extensive enough for the prospect to provide adequate last-resort underpinning for deterrence of a potential aggressor superpower—could never be morally legitimate. This Note considers that judgement on the basis of three beliefs, not themselves here argued out: a. that mutual deterrence involving nuclear weapons is a highly stable safeguard of peace and thus, for the West, of freedom—that provided sensible plans and provision are maintained, the likelihood of ever having to choose between nuclear action and submission to totalitarian conquest is remote; b. that an alternative security system involving Western abandonment of nuclear weapons would carry far higher risk to peace and freedom; c. that Western possession of a substantial nuclear armoury nevertheless would not be justified if there were no circumstances in which its use could ever be right. On this basis, the judgement has consequences so grave that it should not be accepted unless established as inevitable for all reasonably conceivable circumstances. 2. This Note takes for granted that nuclear attack whose specific purpose was essentially the destruction of population and property, with little regard for combatant/non-combatant distinctions, could never be justified, even in reprisal. 3. The judgement usually rests on some or all of these arguments:

184

APPENDIX 2 i. that strategic nuclear action would certainly kill non-combatants, even if that were not its purpose; ii. that the scale of killing would inescapably be disproportionate to any good result that could reasonably be expected; iii. that even if such action could be brought within tolerable limits in respect of discrimination and proportionality, it would certainly or with overwhelming probability trigger further events exceeding those limits; iv. that radiation from fall-out causes deaths (and maybe genetic damage) in a way so certain, far-flung, and long-term that it must be regarded as lying beyond any reasonable view of unintended collateral effect as envisaged in just-war theory. 4. Argument (i) is incontrovertible. There are specialized uses of nuclear weapons—as in the anti-ballistic-missile or anti-submarine roles—which (aside from the limited fall-out inevitable with any non-buried explosion, because of the material of the device itself) might cause no non-combatant casualties; but no such use or combination of uses could inflict a penalty heavy enough for sure deterrence. But the certainty of harm to noncombatants does not, in itself and irrespective of degree and of proportion to other effects, mean that action entailing it cannot be legitimate. Major war—at least in the round, if not always in relation to clearly identifiable particular actions—has long entailed virtual certainty of such harm, and most ethical analysis has tolerated this within ‘double effect’, ‘lesser of two evils’, or similar concepts. 5. Argument (ii) is both more central and more disputable. It involves considering, first, what damage to objectives that might legitimately be attacked might suffice, in prospect, to deter a potential aggressor; and second, whether the harm to non-combatants likely to accompany such damage could ever be a lesser evil than letting the aggressor prevail. 6. The first of these questions requires us to consider whether a damage plan sufficient to deter could be devised without envisaging massive attack on cities as such. This is not precisely calculable, because it depends on the particular circumstances and vulnerabilities of a potential aggressor state and the value judgements of its leaders. But we might ask ourselves whether we could imagine types of nuclear attack upon the West which, without wholesale assault upon cities, would still leave our complex societies too preoccupied with reconstruction and social cohesion to have either will or capability for avoidable military enterprises abroad on a large scale. Plainly we could; and even though a totalitarian state might have a different tolerance level, it must be possible that attacks not essentially counter-population in character could still—for example, by dislocating or distracting economic effort, by shaking internal political acquiescence, or by reducing military capacity for external

185

APPENDIX 2 conquest and occupation—render such a state highly unlikely to be able or willing to continue major external aggression. (It is not, however, supposed here that successful ‘counter-force’ attack, in the particular sense of attack which seeks to neutralize an adversary’s nuclear capability before it can be used, is a feasible option, or that pursuit of capability for it would be other than unhelpful to confidence in stability.) 7. Nevertheless, even with strategic targeting policies which kept total scale as low as possible, excluded direct attack on populations as such, and regarded the minimization of civilian casualties as a major consideration, very large numbers of non-combatants would probably be killed. (This might well be likely also of major war with modern non-nuclear weapons.) The second of the questions in paragraph 5 above—proportionality—then arises. 8. Deaths on a huge scale would be an appalling calamity. But proportionality involves not one factor only, but the relationship between two. World conquest or domination—even for a short time, as of the Nazi conquests— by a tyranny like those of Hitler or Stalin would also be an appalling calamity. Each was responsible, within a limited time and geographical sway, for tens of million of deaths and immeasurable other suffering. Whatever we think of present regimes, the structures and traditions of totalitarian states rarely guarantee their neighbours that they could never assume the malignity of, say, the Stalinist system. The assessment, moreover, is not just a matter of counting lives lost on alternative hypotheses; the defence of truth and human rights has a separate and major weight, as the [Pastoral Letter] recognizes. The price of defeating Hitler was enormous—many millions of lives, including a substantial proportion of non-combatants—yet most people would agree that it was better than letting him prevail. There seems no reason why we must conclude, for every possible future case, that we have to reach a contrary judgement in respect of very large losses—even perhaps running again to many millions—as the undesired price of resisting an extreme tyranny engaged in global aggression. Hundreds of millions would be another matter, but the discussion in paragraph 6 above suggests that halting the aggressor need not automatically be assumed to involve such a penalty. 9. To be both effective in deterrent prospect and justifiable in execution, the level of nuclear strike would need to be such that the expected damage to the aggressor should suffice to rob him of will or capability to go on, while at the same time the unpurposed harm expected to non-combatants was less than the evil expected if the aggressor prevailed. It is not obvious that there could never be a level of strike satisfying these two conditions together. (Indeed, it might be argued that the higher we set the aggressor’s assumed tolerance of loss, the greater we imply to be his inhumanity and therefore the harm from letting him prevail.)

186

APPENDIX 2 10. The question can be viewed from another angle. Suppose, artificially, that major aggression by a tyrannical regime could be frustrated by counteraction which would inflict a hundred unpurposed non-combatant deaths. Almost everyone would agree that this was morally tolerable. At the other extreme, almost everyone would agree that counter-action involving a hundred million non-combatant deaths would in any circumstances be disproportionate. Where exactly, along the vast spectrum between a hundred and a hundred million, the cross-over point between proportionate and disproportionate would fall would be a difficult assessment, requiring judgement which would vary widely with circumstances, such as the nature and record of the aggressor. But in principle there must be a cross-over point. It would be possible to frame the provision of Western deterrent capability at a level judged adequate, in the worst reasonably conceivable military circumstances, to inflict damage at or just below the cross-over point related to the worst reasonably conceivable aggression to be prevented. Force provision so calibrated might still be very substantial and thus highly effective in deterrence; yet virtually by definition there is no compelling reason why its potential use should be condemned out of hand, if planning did not assume (as it need not and should not) the automatic use of the full capability in circumstances falling short of the worst cases to which force provision had been geared. 11. Some further points are worth noting:

a. A deterrent policy on the basis of paragraph 6 does not depend on the adversary’s having a similar one. It could still be adequate even if his envisaged massive counter-population strike. Deterrence requires not that the penalty for the aggressor should be greater than for the defender but that it be greater than the prize the aggressor could expect to gain. b. Even if non-combatant deaths are truly unpurposed—that is, if the targeting seeks to achieve its results without them, and to keep their numbers to a minimum—the prospect of them may still contribute to deterrence; and this does not morally invalidate the concept. c. An aggressor state—particularly a totalitarian one judging others by its own standards—might never feel able to rule out the risk that, whatever the defender’s declared policy, a different one might be followed under the stress of war (especially if it itself attacked cities). This too might reinforce deterrence; and provided that the declared policy was truly intended and reflected in planning, the deterrent bonus of feared breach would not make immoral the possession of weapons properly justified on other grounds.

187

APPENDIX 2 d. If deterrence failed, moral judgement would still remain to be made about whether the particular circumstances warranted the execution of nuclear-strike plans at all, and if so on what scale. e. The concept of strategic strike implied in paragraph 6 might well never warrant anything like the all-out launch of warheads in the many thousands of which present armouries are capable. Evaluation of how large armouries ought still to be in conformity with that concept would need, however, to be approached warily. For example, even if it were judged that the delivery on appropriate targets of X warheads was the most that could ever be legitimately planned, the calculation of inventory would be entitled to allow conservatively for the capability of the opposing armoury in a well-timed pre-emptive strike; unserviceabilities; defences, and their possible development; and possible change in target sets. The outcome might legitimately be several times X. 12. Argument (iii) of paragraph 3 concerned adversary response. In ethical theory it is not clear that conjecture, however confident, of further immoral reactions by an aggressor must determine the morality of the victim’s reaction at or near the situation of last resort; it seems unlikely, for example, that moral theologians would regard it as a woman’s absolute duty—as distinct from her right if she so chooses—to submit to rape if she believes that resistance would lead to worse. It would plainly be natural for the victim of aggression to consider whether the outcome of resistance would be worse than that of submission. But we do not have to prescribe the answer in advance. We cannot tell surely now, for all possible circumstances, what course nuclear war would take; in particular, we have no ground for assuming that Western resistance along the lines of paragraph 6 would inevitably precipitate an unlimited counter-city response (the point in paragraph 11(c) is relevant). It seems unreasonable therefore to claim that opinions, inevitably uncertain, about how an unjust aggressor would react must absolutely rule out the option of resistance. 13. Argument (iv) of paragraph 3 concerned radiation effects. These— especially genetic damage, if a large-scale likelihood of this were substantiated—are particularly grave and repugnant, and our lack of sure knowledge about them imposes an extra duty of prudence. Yet there seems no reason why moral evaluation must regard them as fundamentally different from other kinds of collateral harm. They would certainly cause non-combatant suffering and death, and long-term harm; but so would any modern war even if radiation effects did not exist. Suppose that we knew that the undesired non-combatant deaths from radiation caused by the nuclear defeat of a neo-Hitler would number ten, or a hundred? We would surely not view this as making that defeat immoral. The implication must be that death

188

APPENDIX 2 or other harm from radiation should enter the assessment of proportionality rather than have some overriding significance. 14. In brief, if moral evaluation of deterrent possession of nuclear weapons is approached on the basis outlined at the start of this Note, the case for condemning all strategic use needs to meet a very high standard of proof. The discussion above indicates that it is not established to such a standard. We are therefore not bound to conclude that the strategic use of nuclear weapons, on a basis adequate for the prospect to deter, must be immoral. 15. This Note does not evaluate current Western planning and force provision against the criteria it implies. Much past planning and provision has clearly not had all such criteria closely in mind. In recent years, however, the need has been more and more fully recognized for targeting policies not entailing deliberate attack on populations; increases in weapon accuracy, and reductions in explosive yield have helped the shaping of more discriminate options; and as repeated Western proposals for deep cuts in strategic forces show, Western leaders plainly believe that armouries much smaller than present ones could still give stable deterrence.

189

Index

Able Archer exercise 62 abolition of nuclear armouries 54, 153–65 accidents 67–71, 146 Adams, Henry B. 5 Additional Protocol 85 Ahmedinejad, President Mahmoud 50, 167 Alamogordo 6 alert states 178 al-Qaeda 72–4 alternative defence 104–5 Annan, Secretary-General Kofi 160 Antarctic Treaty 97 anti-ballistic-missile systems see ballistic missile defence Anti-Ballistic-Missile Treaty 1972 61, 85, 88–9, 121, 173 anti-nuclear campaigning 119 Argentina 79 arms control 90–8, 171–7 arms racing 60, 81–2, 88–90 Aspin, Hon. Les 159 atomic demolition munitions see nuclear land-mines attack, NATO air planners’ definition of 18 Attlee, Rt. Hon. Earl (Clement) 7–8, 21, 116 Bailey, Sydney D. xviii, 49 Baldwin, Rt. Hon. Stanley 115 ballistic missile defence 41, 61, 92–3, 109–10, 121, 143–4, 175–6 Bangkok, Treaty of 78 Beckett, Rt. Hon. Margaret 83, 132 Belarus 79 Berlin 23, 29, 135, 154, 157 biological weapons 9, 24, 163 Blair, Rt. Hon. Tony 128 Blue Streak 120 BMD in Central Europe 175–6

190

Bomber Command 116 Bondi, Professor Sir Hermann 30 Brazil 79 Brodie, Bernard xviii, 7, 71, 89 Browne, Rt. Hon. Des 84, 163 Brussels Treaty 34, 124 Bundy, Professor McGeorge 25, 99 bunker-buster weapons 179 Bush, President George W. 128, 147, 156, 171

Cameron, Marshal of the Royal Air Force Lord xviii Canberra bomber aircraft 118 chemical weapons 9, 22, 24, 36, 101, 103–4, 163 Cheshire, Group Captain Lord xviii Chevaline system 121 China, People’s Republic of 61, 82, 99–100, 133–4, 137, 141, 143, 158, 171–2, 177 Churchill, Rt. Hon. Sir Winston 6–7 Clark, William P. 52 Collins, Wilkie 5–6 communication of nuclear intentions 42 Comprehensive Test Ban Treaty 93–4, 148, 161, 172 confidence-building measures 142 Cooper, Rt. Hon. Sir Frank xviii Cooperative Threat Reduction Initiative 73, 78 Copernicus 8 Corbett, Professor Percy E. 19 Cornish, Dr. Paul xix costs 88–90 counter-force strike 17 counter-population strike 29, 52, 107, 126 counter-value strike 126 Cuba 23, 29, 60, 135

Index Davy Crockett missile 41 de Gaulle, President Charles 24 delegated authority 16 Delhi terrorist attack 135 de Madariaga y Rojo, Professor Salvador 157 Democratic Republic of Korea, see North Korea demonstrative use of nuclear weapons 38 deterrence 18–32 ‘deterrence by denial’ 19 disarmament 80–8, 171–7 discrimination 48, 52 dissuasion 19 Ditchley Foundation xvi Drayson, Lord 18 dual-key arrangements 34, 118, 127, 175 Durrani, Major-General M. A. 137

Gandhi, Mahatma 51, 104 Gandhi, Prime Minister Rajiv 153 general and complete disarmament 82 Geneva Protocol on chemical weapons 103–4 Georgia 173, 176 Germany Brussels Treaty limitation 34 relationship with NPT 80 role in NATO nuclear planning 34–5, 40 view of I/MRBM arms control 92–3 view of UK nuclear role 123–4 Gorbachev, Mikhail 60, 153 Gray, Professor Colin S. 60 ground-launched cruise missiles 41, 43, 93, 138 Gulf War 1990–1 40 Guthrie, General Lord 46

Edwards, Andrew 10 Eisenhower, President Dwight D. 88 enhanced radiation weapons, see neutron bomb escalation 16–7, 35, 47, 62–7 escalation dominance 66 European Union 73–4, 78, 110, 129 existential deterrence 25 explosive yield 36, 131, 179

Hague Code of Conduct on ballistic missiles 78 Hastings, Sir Max 7 Heath, Rt. Hon. Sir Edward 128 highly-enriched uranium 86, 131 Hiroshima xiii, 6–7 Howard, Sir Michael xiii Hurd, Rt. Hon. Lord 156

Falkland Islands 23 false alarms 68 Federal Republic of Germany (FRG), see Germany Fernandes, Minister George 137 fighter/ground-attack aircraft 118, 138 first strike 17, 100 first use 17, 100 fissile material cut-off treaty 131, 148, 161, 172 flexible-response strategy 35–6, 93 Ford, Hon. Christopher A. 171 forward defence 37 France cost of nuclear capability 119, 123 deterrence concept 28, 35–6 deterrent stability 61 European dimension 128 force levels and transparency 177 independence 123–4 nuclear abolition 156, 171 relationship with NPT 147 Resistance 104–5 targeting policy 52

independence in nuclear capability 121–3 India ballistic missile defence 142–4 capabilities 138–9 deterrent relationship with China 137, 141, 143 deterrent relationship with Pakistan 141–6 deterrent stability 139–41 no first use 99–100, 137 non-proliferation 79, 147–9, 170 nuclear energy deal with USA 147–8, 170 nuclear doctrine 136–7 nuclear history 133–5 nuclear tests 77, 133–4 security of armoury 140–1 INF Treaty 43, 92, 95 inter-continental ballistic missiles 93 intermediate-range ballistic missiles 92–3 intermediate-range nuclear forces 35, 93 International Atomic Energy Agency 85–6, 147, 167–8 International Court of Justice 82

191

Index International Code of Conduct on ballistic missiles 78 International Convention on the Suppression of Acts of Nuclear Terrorism 71 International Institute for Strategic Studies xvi, 83–4, 162 Iran 50, 61, 84, 167–9, 176 Iraq 29, 31, 73, 84–5, 103–4 Iron Curtain 23 Israel 29, 50, 61, 79, 101, 154, 158, 170 Japan 6, 50, 78, 80 jus ad bellum 47 jus in bello 47–8 Just War tradition 46–8, 53 Kamal, General Hassan al-Majid 103 Kargil 135 Kashmir 77, 133–5, 140, 142–3, 158 Kazakhstan 79 Kelsay, Professor John 46 Kennan, Hon. George F. 99 Kennedy, President John F. 120 Khan, Abdul Qadeer 73, 136–7, 140, 168 Kissinger, Hon. Henry A. 24, 89, 156 Korean peninsula 22–3 Krause, Professor Joachim 81 Kuwait 23, 31, 40 last resort 40 launch on warning 70, 178 launch under attack 70, 178 Lewis, Dr. Jeffrey 171 Libya 79, 136 Lisbon force goals 90 low-enriched uranium 131 McMahon Act 117 McNamara, Hon. Robert S. 40, 99 Mackby, Dr. Jennifer xix Manmohan Singh, Prime Minister 147 medium-range ballistic missiles 92–3 Military Committee (MC) 14/3 35 minimal armouries 105–8, 172 minimum deterrence 137 MIRVs 93, 120 miscalculation 67–71 Missile Technology Control Regime 78, 148 Moscow BMD system 121 ‘Moscow criterion’ 121 Moscow Treaty 2002 96, 173–4

192

MRVs 120–1 Multi-Lateral Force 34 mutual and balanced force reductions xv, 95 mutual assured destruction 39, 179 Mutual Defence Agreement US–UK 117 Nagasaki 6–7 NATO arms control 41 communication of nuclear intentions 42 Exercise Able Archer 62 flexible response 35–6, 93 forward defence 37 High-Level Group xvi multilateral force project 34 no first use 100, 178 nuclear authorization 16 nuclear doctrine 16, 33–40 Nuclear Planning Group xv, 15, 34–41 strategic concept 35–6, 175, 178 negative security assurances 103 neutron bomb 43–4 no first use 99–104, 137, 178 non-proliferation 76–87 arguments for and against 76–7 regime components 78 regime successes 79 Non-Proliferation Treaty (NPT) accessions 79 bargains 81 disarmament (Article VI) 81–3, 129, 156, 171 nuclear energy (Article IV) 81, 169 review conferences 79, 83, 86–7, 148, 155, 169 verification 84–5, 169 withdrawal (Article X) 85, 169 non-strategic weapons 18, 40–4, 96–7, 174 North Atlantic Treaty 124 North Atlantic Treaty Organisation see NATO North Korea 31, 61, 79, 84, 134, 167 Norway 84 nuclear energy 81, 85–6, 155 nuclear land-mines 41 Nuclear Suppliers Group 78, 147–8 nuclear ‘taboo’ 100 nuclear threshold 16 nuclear-weapon-free zones 78, 97, 103, 148, 175

Index ‘nuclear winter’ 13 Nunn, Hon. Sam A. 156 Nunn-Lugar programme see Cooperative Threat Reduction Initiative Oppenheimer, Dr. J. Robert 7 Outer Space Treaty 97 Owen, Rt. Hon. Lord 156 Pakistan ballistic missile defence 143 capabilities 138–9 dealings with USA 135, 146 deterrent relationship with India 29, 108, 142–3 deterrent stability 139–42 Khan, Abdul Qadeer 135–7 no first use 100, 137, 144 non-proliferation 79, 147–9, 170 nuclear doctrine and organization 137 nuclear history 133–6 nuclear tests 77, 134 security of armoury 137, 140–1 Partial Test Ban Treaty 94, 97, 173 passive resistance 104 Pelindaba, Treaty of 78 Perkovich, Dr. George 84, 162 permissive action links 16, 65 Perry, Hon. William J. 156 Pershing MRBM 41, 43, 93 Polaris SLBM 120–1 Pope Benedict XVI 154 Pope John Paul II 51 proliferation see non-proliferation Proliferation Security Initiative 78 Proportionality 48, 52–3 Ptolemy 8 radiological weapons 71 Rarotonga, Treaty of 78 rationality in deterrence 27, 30 Reagan, President Ronald W. 47, 60, 109, 153 ‘red lines’ 23 reductio ad absurdum of warfare 10–11, 49, 54 Reliable Replacement Warhead 131, 172 Reykjavik meeting 1986 153 Rifkind, Rt. Hon. Sir Malcolm 156 Robertson, Rt. Hon. Lord 156 Royal Society 74 Royal United Services Institute for Defence Studies xii

Ruhle, Dr. Michael 81 Russia continued importance of nuclear weapons 154, 159 deterrent stability 60–1 future arms control 95–7, 173–6 INF Treaty 95 military use of space 97 no first use 99 non-strategic armoury 96–7, 174 sole nuclear-weapon successor of USSR 79 Treaty of Semipalatinsk 78 US BMD plans 175–6 US nuclear weapons in Europe 174–5 Saddam Hussein 29, 31, 73 safeguards (IAEA) 85 Saraswat, Dr. V. K. 144 Sarkozy, President Nicolas 177 Schwebel, Judge Stephen M. 82 Sea-Bed Treaty 97 ‘second-centre’ role 121–3 Semipalatinsk, Treaty of 78 Sen, Professor Amartya K. 134 Shultz, Hon. George P. 156 Skybolt missile 120 Smith, Hon. Gerard C. 99 South Africa 79 South Asia 23, 61, 133–49 South Korea 31, 78 Soviet Union BMD system 41, 121 break-up 73–4, 79, 160 Cuba 1962 23, 29 deterrability 28 Exercise Able Archer 62 no first use 99 strike planning 29 Western perceptions 21 SS20 missile 93 stability in deterrence 59–62 ‘Star Wars’ speech 109 Stimson, Hon. Henry L. 7 strategic, meaning of 18 strategic armouries 174 Strategic Arms Limitation Talks xv, 15, 89, 96, 146–7, 173 Strategic Arms Reduction Talks/Treaties 15, 96 strike, NATO air planners’ definition of 18

193

Index submarine-launched ballistic missiles xvi, 61, 70, 109, 139 sub-strategic, use of term 18, 42, 127 tactical, meaning of 18, 42 Taiwan 78, 158 Taleban 74 terminology 15–19 terrorists 32, 71–5, 170–1 Tertrais, Dr. Bruno 52 threat, concept of 17 Thor IRBM 118 Tlatelolco, Treaty of 78 Tokyo 7 Tomonaga, Professor Masao 6 Transparency 177 Trident SLBM 69, 120, 125–30 Truman, President Harry S 6, 116, 157 United Kingdom abolition of nuclear armouries 132, 155–6 anti-nuclear attitudes 119 arms control 94–5, 131–2 cooperation with United States 117, 120, 122, 128 costs of nuclear capability 90, 119, 122–3, 128–9 disarmament 129 dual-key US weapons 118 Europe 128–9 future plans 119, 127–131 independence 122–3 negative security assurances 103 nuclear history 106, 115–9 participation in NATO doctrine 35, 124 pre-nuclear experience 115 reductions in armoury 126–7 second-centre concept 121–2 sub-strategic capability 18, 118, 127 targeting policy 52, 126–7 technical risk experience 106 transparency 177 warhead inventory 107, 126–7 United Nations General Assembly 82, 153 United Nations High-level Panel 160 United Nations Security Council 74, 160, 168 UNSCR 1172 134–5, 147

194

UNSCR 1540 73, 78, 141, 170–1 United States alert arrangements and doctrine 178–9 Anti-Ballistic Missile Treaty 1972 88–9 attitudes to arms control 86 ballistic missile defence 61, 109–11, 171, 175–6 bunker-buster weapons 179 Cold War armoury in Europe 40–4 cooperation with UK 117–8, 120, 128 Cooperative Threat Reduction Initiative 78 dealings with Pakistan 135, 146 deterrent relationship with China 61, 171–2, 177 dual-key weapons in NATO 34, 118, 174–5 future nuclear arms control 95–7, 173–6 Japan security treaty 78 minimal armouries 105–7 NATO doctrine 35 negative security assurances 103 no new capabilities planned 171, 179 non-strategic weapons 40–4 nuclear energy deal with India 147–8, 170 nuclear energy deal with Russia 173 Proliferation Security Initiative 78 reductions made 89, 171 relations with Iran 167–9 Reliable Replacement Warhead 131, 172 targeting policy 39, 52 warheads in Europe 40–1, 174–5 United States Catholic Bishops 51–2 V-bomber force 106, 117–8, 120 virtual armouries 108–9, 172 Waltz, Professor Kenneth N. 76 war-fighting concepts 39 war termination 10–11, 38, 42, 102 Warsaw 104 Warsaw Pact 16, 34, 43 weapons of mass destruction 17 Wolfers, Arnold 28 Wren, Sir Christopher 44 Yost, Professor David S. 81