Nuclear Strategy in the Modern Era: Regional Powers and International Conflict [Course Book ed.] 9781400850402

Table of contents :
Contents
List of Figures and Tables
Acknowledgments
Chapter One. Introduction
Chapter Two. The Sources of Regional Power Nuclear Postures: Posture Optimization Theory
Chapter Three. Pakistan
Chapter Four. India
Chapter Five. China
Chapter Six. France
Chapter Seven. Israel
Chapter Eight. South Africa
Chapter Nine. Deterring Unequally I: A Large-n Analysis
Chapter Ten. Deterring Unequally II: Regional Power Nuclear Postures and Crisis Behavior
Chapter Eleven: Conclusion
Bibliography
Index

Citation preview

Nuclear Strategy in the Modern Era

PRINCETON STUDIES IN INTERNATIONAL HISTORY AND POLITICS Series Editors G. John Ikenberry and Marc Trachtenberg Recent Titles Nuclear Strategy in the Modern Era: Regional Powers and International Conflict by Vipin Narang The Cold War and After: History, Theory, and the Logic of International Politics by Marc Trachtenberg Liberal Leviathan: The Origins, Crisis, and Transformation of the American World Order by G. John Ikenberry Worse Than a Monolith: Alliance Politics and Problems of Coercive Diplomacy in Asia by Thomas J. Christensen Politics and Strategy: Partisan Ambition and American Statecraft by Peter Trubowitz The Clash of Ideas in World Politics: Transnational Networks, States, and Regime Change, 1510–­2010 by John M. Owen IV How Enemies Become Friends: The Sources of Stable Peace by Charles A. Kupchan 1989: The Struggle to Create Post–­Cold War Europe by Mary Elise Sarotte The Struggle for Power in Early Modern Europe: Religious Conflict, Dynastic Empires, and International Change by Daniel H. Nexon Strong Borders, Secure Nation: Cooperation and Conflict in China’s Territorial Disputes by M. Taylor Fravel The Sino-­Soviet Split: Cold War in the Communist World by Lorenz M. Lüthi Nuclear Logics: Contrasting Paths in East Asia and the Middle East by Etel Solingen Social States: China in International Institutions, 1980–­2000 by Alastair Iain Johnston Appeasing Bankers: Financial Caution on the Road to War by Jonathan Kirshner The Politics of Secularism in International Relations by Elizabeth Shakman Hurd Unanswered Threats: Political Constraints on the Balance of Power by Randall L. Schweller Producing Security: Multinational Corporations, Globalization, and the Changing Calculus of Conflict by Stephen G. Brooks Driving the Soviets up the Wall: Soviet-­East German Relations, 1953–­1961 by Hope M. Harrison

Nuclear Strategy in the Modern Era Regional Powers and International Conflict

Vipin Narang

Princeton University Press Princeton and Oxford

Copyright © 2014 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, 6 Oxford Street, Woodstock, Oxfordshire OX20 1TW press.princeton.edu Jacket photo: LGM-118A’s first stage solid rocket ignites as the missile clears the silo. National Museum of the US Air Force / USAF photo. All Rights Reserved ISBN 978–­0–­691–­15982–­9 ISBN (pbk.) 978–­0–­691–­15983–­6 Library of Congress Cataloging-in-Publication Data Narang, Vipin. Nuclear strategy in the modern era : regional powers and international conflict / Vipin Narang. pages cm. — (Princeton studies in international history and politics) Includes bibliographical references and index. ISBN 978-0-691-15983-6 (pbk.) — ISBN 978-0-691-15982-9 (hardcover)  1. Nuclear weapons— Government policy—Case studies.  2. Nuclear warfare—Government policy—Case studies.  3. Deterrence (Strategy)—Case studies.  4. Security, International—Case studies.  I. Title. U264.N36 2014 355.02'17—dc23 2013034255 British Library Cataloging-­in-­Publication Data is available This book has been composed in Sabon LT Std Printed on acid-­free paper. Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

Contents

List of Figures and Tables

vii

Acknowledgments ix Chapter 1

Introduction

Chapter 2

T  he Sources of Regional Power Nuclear Postures: Posture Optimization Theory

1 13

Chapter 3

Pakistan 55

Chapter 4

India 94

Chapter 5

China 121

Chapter 6

France 153

Chapter 7

Israel 179

Chapter 8

South Africa

207

Chapter 9

Deterring Unequally I: A Large-­n Analysis

222

Chapter 10

D  eterring Unequally II: Regional Power Nuclear Postures and Crisis Behavior

253

Chapter 11

Conclusion 299

Bibliography

313

Index

333

Figures and Tables

Figures Figure 2.1. Optimization theory of nuclear posture Figure 2.2. Optimization theory’s predictions for regional power nuclear posture Figure 9.1. Deterrent power of asymmetric escalation nuclear posture Figure 10.1. The Brasstacks Crisis Figure 10.2. Kargil War Figure 10.3. Operation Parakram 2002 planned offensives Figure 10.4. Yom Kippur War Israeli-­Syrian Front Figure 11.1. Optimization theory’s performance against empirical record

32 53 245 263 269 276 290 302

Tables Table 2.1. Summary of regional power nuclear postures 22 Table 2.2. Relationship between motivations for nuclearization and nuclear posture 25 Table 5.1. Chinese and Soviet land force deployments on the Sino-­ Soviet border in the Far East 141 Table 9.1. Description and empirical frequencies for large-­n 238 dependent variable (DV) Table 9.2. Summary statistics for nuclear posture and frequency of attack 239 Table 9.3. Multinomial logit results for 1816–­1992 (relative risk ratios) 241 242 Table 9.4. Multinomial logit results for 1816–­2001 (relative risk ratios) Table 10.1. Summary of crisis-­dyad predictions based on large-­n results 256 Table 10.2. Summary of crisis cases 297 Table 11.1. Explanatory success for optimization theory and alternative explanations 303

vii

Acknowledgments

This book would not have been possible without the support of countless mentors, colleagues, friends, and institutions. The genesis of my interest in security studies began when I was an undergraduate chemical engineer interested in international security. Two key individuals were responsible for shepherding my shift from engineer to political scientist. First, Scott Sagan took a chance on me as an undergraduate and has never wavered in his support, being both tough and caring when necessary. I am forever grateful to Scott and Stanford’s Center for International Security and Cooperation for starting my trajectory, and welcoming me back in 2012 as a Stanton Junior Faculty Fellow. I am fortunate to have Scott as a mentor, colleague, and friend. The second catalyst at Stanford was Taylor Fravel, who was the first teaching assistant I had in an international relations course in 1999. He pushed me to consider writing an undergraduate thesis in international security, and I have not looked back since. We have teamed up again at MIT, where he has provided constant guidance and support, and is a trusted friend and colleague. My career, let alone this book, would not have been possible if not for Scott and Taylor. This book began when I was a doctoral student in Harvard’s Government Department. I first thank Yuen Foong Khong, my Oxford MPhil advisor, for pushing me to pursue a PhD. At Harvard, my advisors, Stephen Rosen, Alastair Iain Johnston, and, remotely, Allan Stam, provided tremendous support, humor, and guidance. Steve, in particular, supported my decision to write a dissertation on nuclear strategy when almost everyone else was telling me the subject was dead. Iain was a role model in academic and personal integrity, and I always value his advice. Al had to work especially hard, since he advised the dissertation remotely, but his input was critical to ensuring that my research design and methods met his high standards. Other faculty at Harvard whose advice shaped this book include Bear Braumoeller, Mike Hiscox, Gary King, and Andy Kydd. My graduate student colleagues at Harvard who provided guidance and feedback through the years made the journey memorable: Tyson Belanger, Amy Catalinac,

x

Acknowledgments

Andrew Coe, Sheena Chestnut Greitens, Mike Horowitz, Andy Kennedy, Alex Liebman, Siddharth Mohandas, Becky Nelson, Jonathan Renshon, Erin Simpson, and Jane Vaynman. I thank the Paul and Daisy Soros Foundation, the Weatherhead Center, and the Smith Richardson Foundation for financial support during graduate school. A number of individuals helped me during my time conducting research in South Asia over the years, often in the extremely hot summers, and I thank them for their time and efforts: Uday Bhaskar, Stephen Cohen, Sumit Ganguly, Jack Gill, Kapil Kak, Gaurav Kampani, Gurmeet Kanwal, Paul Kapur, Bharat Karnad, Feroz Hassan Khan, Michael Krepon, Arvinder Lamba, Pratap Bhanu Mehta, Anit Mukherjee, Raja Menon, Raja Mohan, Rostum Nanavatty, Shuja Nawaz, Arun Prakash, Srinath Raghavan, Shyam Saran, Vijay Shankar, the late K. Subrahmanyam, and several others who must remain anonymous. Large portions of the book were critiqued and revised when I was a predoctoral fellow, and I thank the Olin Institute and the Belfer Center at Harvard for hosting me for two years. For tremendous help during these years, I want to thank Emma Belcher, Malfrid Braut-­Hegghammer, Matt Bunn, Alex Debs, Alex Downes, Matthew Fuhrmann, Vaidya Gundlupet, Matthew Kroenig, Jason Lyall, Sean Lynn-­Jones, Marty Malin, Steve Miller, Negeen Pegahi, Daryl Press, Josh Rovner, Karthika Sasikumar, Todd Sechser, Maya Tudor, Elizabeth Saunders, Ben Valentino, and Steve Walt. In these years, I met two colleagues who became great friends and who painstakingly read multiple iterations of this manuscript: Paul Staniland and Caitlin Talmadge. I finalized this book manuscript while working at my first job at MIT. I owe the MIT Department of Political Science and the Security Studies Program a huge debt of gratitude for their support, and for providing me with an intellectual home. This book would not have been possible if not for the hallway chats, eyes on the page, and moral support of Barry Posen, Steve Van Evera, Dick Samuels, Rick Locke, and of course Taylor. MIT hosted a book workshop that allowed Scott, Charlie Glaser, and Bob Jervis to spend a day helping me improve the book immeasurably, along with Barry, Steve, and Taylor. I thank them all for coming on a bitterly cold winter day, assembling more intellectual firepower on nuclear strategy in one room than I could have ever dreamed of. My graduate student research assistants, notably the extremely capable trio of Mark Bell, Chris Clary, and Nick Miller, provided excellent research and editorial assistance. Workshop participants at the University of Chicago, Dartmouth, George Washington University, Princeton University, Yale University, University of Virginia, and the University of Wisconsin–­Madison, where I was also a faculty member for a year and met great colleagues and friends, have provided critical constructive feedback. The final stop on my journey with this book, perhaps

Acknowledgments xi

appropriately, was back home where my career began, at Stanford’s CISAC, where the Stanton foundation generously allowed me to spend a year as junior faculty fellow. The year there with Scott, Lynn Eden, Malfrid, Thomas Hegghammer, David Holloway, Rifaat Hussain, and Feroz Hassan Khan was crucial as I worked through the final revisions on the manuscript. I thank them for welcoming me, and for a wonderfully enriching year back on The Farm. At Princeton University Press, I would like to thank Chuck Myers for patiently shepherding the book through the editorial process, and Eric Crahan for catching the pass as Chuck moved on. Special thanks to Sara Lerner and Karen Verde, whose hard work during the production process vastly improved the final product. Philip Schwartzberg of Meridian Mapping created the maps, and Sherry Smith generated the index. I also thank the two anonymous reviewers of the manuscript, whose suggestions were critical to the final arguments and evidence, though any and all errors of course remain my own. My thanks go to the editors at International Security and Journal of Conflict Resolution for granting me permission to use some of my previously published material in this book. My apologies to anyone I may have forgotten to thank. I would not have been able to complete this book without the support of family and friends. My parents never batted an eye as I skipped out on medical or law school to pursue a PhD in political science. Their support has been tremendous, and I have never been able to properly express my thanks to them. Sameer and Neha, and Ricky and Anisha, have been unwavering pillars of support. My closest friends, Rohan Murty and Sujay Jaswa, who wonder why anyone would want to study nuclear strategy, have always lifted me up with good (and bad) humor. For their support and friendship through the writing of this book, I also thank Matt Abel, Antara Datta, Prithviraj Datta, Deepa Dhume Datta, Justine Fisher, Sheena Chestnut Greitens, Zach Kaufman, Allon Klein, Roohia Sidhu Klein, Neil Kumar, Adam Lauridsen, Neil Malhotra, Siddharth Mohandas, Newman Nahas, Mark Paustenbach, Jonathan Renshon, Mike Sulmeyer, Amy Rogers, and Jason Wasfy. Through this journey, though, Sana Aiyar has been with me every step of the way since we met in the bowels of Perkins Hall, even as we had to commute between Boston and Madison for years. She has opened the world to me: Delhi, Kerala, London, Dar es Salaam, Cape Town, Baltimore, Madison, and back to Boston. She has given me an extended family in Delhi. And just as our first babies (our books) are being completed, we welcome our first son, Ishaan, into our journey forward. For making me a better person, I dedicate this book to them.

Nuclear Strategy in the Modern Era

Chapter One

Introduction

The world has entered a second nuclear age. This age is not defined by a bipolar global superpower competition involving massive nuclear arsenals with the capability to destroy each other multiple times over. In this new era, regional nuclear powers will define the proliferation and conflict landscape. These states have small nuclear arsenals, are often ensnared in long-­ standing rivalries, participate in multiple active conflicts, and often have weak domestic political institutions. However, because scholarly attention has focused largely on the superpowers and the Cold War nuclear competition we presently have a poor understanding of these unfolding nuclear dynamics. The superpower model of nuclear strategy and deterrence does not seem to be applicable, for example, to India and Pakistan, which have small arsenals and where an active and enduring territorial rivalry is punctuated by repeated crises that openly risk nuclear war. In the Middle East, if Iran were to successfully acquire nuclear weapons, a cascade of nuclear weapons states could emerge that would create complicated multipolar competitions very different from the bipolar Cold War arms race. In East Asia, almost nothing is known about North Korea’s nuclear arsenal or the doctrine by which those weapons might be employed. And there are potential proliferators waiting in the wings: Japan, essentially a standby nuclear state, could one day be compelled to weaponize its nuclear capabilities, with significant consequences for East Asian and global security. There is little evidence that these countries’ nuclear arsenals will mirror those of the superpowers, or that their nuclear behavior will follow the same patterns. What strategies and choices

1

2

Chapter 1

will these states make about their nuclear weapons? And how will those decisions about nuclear strategy affect international relations and conflict? Examining the decisions that regional nuclear powers—­such as China, India, Pakistan, Israel, France, and South Africa—­have made about their arsenals thus far, and their resulting behavior, helps address these questions. Regional nuclear powers, for systematic and predictable reasons, choose clearly identifiable nuclear postures and these postures matter to a regional power’s ability to deter conflict. These countries’ nuclear choices, therefore, provide valuable insight into the crucial challenges of contemporary nuclear proliferation and international stability. This book thus focuses on the overlooked experiences of the regional nuclear powers. It concentrates on two important questions. First, which nuclear strategies do regional powers adopt, and why? Second, what effect do these choices have on their ability to deter conflict?

Regional Power Nuclear Postures What do scholars and policy makers presently know about the nuclear strategies and postures of regional powers and why they choose them? Unfortunately, very little, for two reasons. First, the bulk of the scholarly literature on nuclear weapons suffers from a Cold War hangover, focusing heavily on the superpower experience of the United States and the Soviet Union.1 However, seven of the nine current nuclear states, and all those that might emerge, are, by definition, regional powers. Compared to the superpowers, these states face different constraints and opportunities, have arsenals that are orders of magnitude smaller, and must manage different 1. See, for example, Lawrence Freedman, The Evolution of Nuclear Strategy, 2nd ed. (London: Macmillan, 1989); Scott D. Sagan, Moving Targets: Nuclear Strategy and National Security (Princeton: Princeton University Press, 1989); Charles L. Glaser, Analyzing Strategic Nuclear Policy (Princeton: Princeton University Press, 1990); Lynn Eden and Steven Miller, eds., Nuclear Arguments: Understanding the Strategic Nuclear Arms and Arms Control Debates (Ithaca: Cornell University Press, 1989); Marc Trachtenberg, “The Influence of Nuclear Weapons in the Cuban Missile Crisis,” International Security, vol. 10, no. 1 (Summer 1985); Thomas Schelling, Arms and Influence (New Haven: Yale University Press, 1966); Thomas Schelling, The Strategy of Conflict (Cambridge: Harvard University Press, 1960); Bernard Brodie, Strategy in the Missile Age (Princeton: Princeton University Press, 1959); John Lewis Gaddis, The Long Peace: Inquiries into the History of the Cold War (New York: Oxford University Press, 1987); Robert Jervis, The Meaning of the Nuclear Revolution (Ithaca: Cornell University Press, 1989); Robert Jervis, “The Political Effects of Nuclear Weapons: A Comment,” International Security, vol. 13, no. 2 (Autumn 1988), 80–­90; for a classic on specific doctrinal/postural effects, see John J. Mearsheimer, “Nuclear Weapons and Deterrence in Europe,” International Security, vol. 9, no. 3 (Winter 1984/1985), 19–­46.

Introduction 3

conflict environments. As a result, the Cold War lexicon that was used to describe American and Soviet nuclear strategies cannot simply be superimposed on the regional nuclear states. These states have, in fact, chosen a diverse array of nuclear postures and strategies—­the capabilities, envisioned employment modes, and command-­ and-­ control procedures that go into operationalizing a nuclear weapons capability—­ that look very different from those of the superpowers. Yet, the choices of the regional nuclear powers—­reflecting many additional decades of thinking on nuclear strategy—­that are most relevant to the present and emerging international nuclear landscape have been largely ignored. Second, to the extent that scholars have paid attention to the non-­ superpower nuclear weapons states, they have focused primarily on their initial acquisition of nuclear weapons.2 Little systematic thought has been given to what, if anything, regional powers choose to do after they acquire nuclear weapons, largely because deterrence scholarship has convinced itself that the mere existence of nuclear weapons, rather than their strategy and posture, drives patterns of stability and conflict. This “acquisition” or “existential” bias is an untested and dangerous assumption. If regional powers’ thinking on nuclear strategy does not follow the superpower pattern, the effects of their choices are unlikely to be the same either. At present, however, scholars and policy makers lack a vocabulary and common analytical lens with which to categorize and understand regional power nuclear strategy. I fill this vacuum in the first part of the book by analyzing the experiences of the regional nuclear powers, or the non-­superpower states that have developed independent nuclear forces: China, India, Pakistan, Israel, South Africa, and France.3 I discuss these states’ choices about nuclear 2. See, for example, Scott D. Sagan, “Why Do States Build Nuclear Weapons? Three Models in Search of the Bomb,” International Security, vol. 21, no. 3 (Winter 1996–­97), pp. 54–­86; T. V. Paul, Power versus Prudence: Why Nations Forgo Nuclear Weapons (Montreal: McGill-­Queen’s University Press, 2000); Etel Solingen, Nuclear Logics: Contrasting Paths in East Asia and the Middle East (Princeton: Princeton University Press, 2007); and Jacques Hymans, The Psychology of Nuclear Proliferation: Identity, Emotions, and Foreign Policy (Cambridge: Cambridge University Press, 2006). 3. The United Kingdom is excluded because it does not have an independent nuclear arsenal. Because of its tight integration with U.S. nuclear forces since 1958, such as dual-­key control of tactical forces, leasing of forces, and strategic targeting coordination with the United States, I consider its nuclear weapons as essentially an adjunct force of America’s. Unlike France, which retained full independent control of its nuclear forces, Britain’s subordination to American nuclear planning makes studying its independent deterrent effect problematic. As a result, I do not classify the United Kingdom as having an independent nuclear posture. See Jeffrey Lewis, “Britain’s Independent Deterrent,” ArmsControlWonk. com, December 22, 2008. Available at: http://lewis.armscontrolwonk.com/archive/2139 /britains-independent-deterrent. In relation to the subsequent theory I develop, one might indeed classify the United Kingdom as having adopted the most extreme version of the

4

Chapter 1

strategy in terms of nuclear posture. Nuclear posture is the incorporation of some number and type of nuclear warheads and delivery vehicles into a state’s overall military structure, the rules and procedures governing how those weapons are deployed, when and under what conditions they might be used, against what targets, and who has the authority to make those decisions.4 Nuclear posture is best thought of as the operational, rather than the declaratory, nuclear doctrine of a country; while the two can overlap, it is the operational doctrine that generates deterrent power against an opponent. To put it bluntly, states care more about what an adversary can credibly do with its nuclear weapons than what it says about them. I thus use the term “nuclear posture” to refer to the capabilities (actual nuclear forces), employment doctrine (under what conditions they might be used), and command-­and-­control procedures (how they are managed, deployed, and potentially released) a state establishes to operationalize its nuclear weapons capability. This can also be thought of as “nuclear strategy,” and I use these terms interchangeably with both referring to the preceding definition. As Tara Kartha colorfully put it, without a nuclear posture or strategy, “a much vaunted [nuclear] test remains simply a loud bang in the ground.”5 Thus, I shift the critical variable of interest from simply the possession of nuclear weapons to how a regional power chooses to operationalize those capabilities as a nuclear posture. I identify three analytically distinct and identifiable nuclear postures: a catalytic strategy that attempts to catalyze superpower intervention on the state’s behalf; an assured retaliation strategy that threatens certain nuclear retaliation in the event a state suffers a nuclear attack; and an asymmetric escalation strategy that threatens the first use of nuclear weapons against conventional attacks. Understanding catalytic posture: complete subordination of its forces to its envisioned third-­party patron. However, since it does not meet the scope conditions of an independent regional nuclear power, I choose to exclude the United Kingdom from my analysis. Similarly, Ukraine, Kazakhstan, and Belarus are excluded because, although they inherited nuclear weapons as newly independent states, they never exerted independent control over them before transferring them back to Russia. North Korea is excluded for temporal reasons since we currently have too little knowledge about the ways in which it chooses to employ its nuclear weapons capability. I discuss the North Korean case in the conclusion. Because Russian nuclear posture is a legacy of the Soviet Union, I continue to classify Russia as a nuclear superpower. 4. As a definitional aside, the focus is intended to be on a state’s observable nuclear posture as defined above, as opposed to a state’s declared nuclear doctrine. A state’s nuclear posture is essentially its peacetime nuclear orientation and procedures for deployment and signaling during crises. Because of the challenges of studying doctrines in general, and nuclear doctrines in particular—­which are highly classified, often unarticulated, untested, and of questionable credibility—­I have chosen to focus on a critical component of doctrine, a state’s nuclear posture, in order to gain some leverage on the questions of interest. 5. Tara Kartha, “Ballistic Missiles and International Security,” in Jasjit Singh, ed., Nuclear India (Delhi: IDSA, 1998), p. 115.

Introduction 5

why states choose one of these postures over the others provides novel insights into the nuclear strategies and conflict dynamics across important areas of the world including the Middle East, East Asia, and South Asia.

Nuclear Strategy and Deterrence in the Modern Era The variation in regional power nuclear strategy is only important if it makes a difference to international security and conflict dynamics. So, which of these nuclear postures are required to deter conflict? Scholars and policy makers still do not know, due again to shortcomings related to the twin problems of the Cold War hangover and the acquisition bias. However, this lack of understanding is not because they have always believed that posture is irrelevant. During the Cold War, the United States and Soviet Union invested substantial effort and thought into how and whether the nuclear postures they adopted affected the strategic balance. The American vocabulary on nuclear strategy was littered with phrases such as “assured destruction,” “flexible response,” “massive retaliation,” “countervailing strategy,” and “damage limitation.”6 Scholars and practitioners believed that these different postures would affect the American ability to deter the Soviet Union, and invested billions of dollars on the basis of those beliefs. Practitioners of deterrence during the Cold War, and a nontrivial proportion of theorists, believed that it took quite a bit—­often maximalist nuclear postures—­to deter both conventional and nuclear conflict between the superpowers. As advocated by the likes of Albert Wohlstetter and Herman Kahn, the superpowers developed massive nuclear arsenals.7 They did so without ever answering how much of an arsenal, and of what type, it would take to deter conventional conflict. The development of overkill arsenals outpaced thinking on this critical issue. The superpowers were also so much more militarily powerful than the other states in the system that their ability to deter non-­superpower states was overdetermined, making it difficult to isolate the deterrent power of their nuclear arsenals. Because of this overdetermination, the superpower nuclear balance is a poor guide for analyzing the relationship between nuclear weapons and deterrence. Moreover, because of resource constraints and perhaps because the regional powers have learned from the superpower experience, the large arsenals and nuclear architectures of the superpowers have not been, and are unlikely to be, replicated by another state. The superpower balance 6. See note 1. 7. Glaser, Analyzing Strategic Nuclear Policy; Keir Lieber and Daryl Press, “How Much Is Enough? Nuclear Deterrence Then and Now,” Working Paper presented at Wisconsin International Relations Colloquium, University of Wisconsin–­Madison, February 3, 2009.

6

Chapter 1

is thus a methodologically and empirically unreliable guide to identifying the type of nuclear forces required to deter conflict. Despite this, virtually the entire corpus of existing scholarship on nuclear strategy and deterrence focuses on these two globally dominant states with massive nuclear architectures. None of this work addresses the question of whether the smaller arsenals and different strategies of the regional nuclear powers are sufficient to deter nuclear and conventional attack. Nearly seventy years after the advent of nuclear weapons, we still do not know exactly what size and type of nuclear arsenal is sufficient to deter. Toward the end of the Cold War, moreover, a revisionist line of thinking emerged in the writings of scholars such as Kenneth Waltz and Robert Jervis.8 This line of argument veered in the opposite direction, suggesting that deterrence could be achieved with far more minimal nuclear postures than the United States and Soviet Union developed. Indeed, the current thinking on nuclear proliferation evinces a deep “existential deterrence” bias, a concept first coined by McGeorge Bundy positing that the “mere existence of nuclear forces,” even ambiguous or non-­weaponized forces, should induce caution in adversaries and deter conflict, both nuclear and conventional.9 States are treated as equivalent once they acquire even a single nuclear weapon. This line of thinking has caused scholars and policy makers to worry about emerging nuclear powers such as North Korea. The implication is that with only a handful of nuclear weapons, a state can deter much more powerful actors, like the United States, because even a small risk of escalation to the nuclear level will deter nuclear and conventional attacks. By the end of the Cold War, and with the emergence of new regional nuclear powers, it became almost an article of faith that the mere possession of even a small number of nuclear weapons generated significant deterrent effects.10 As noted earlier, this belief in existential deterrence resulted in a focus on regional powers’ initial marches toward nuclear weapons capabilities, based on the belief that the decision to acquire nuclear weapons was critical and that the choices that came afterward were largely irrelevant. Little attention was paid to the strategies and forces regional powers developed once they first acquired nuclear weapons—­that is, the nuclear postures

8. See, for example, Kenneth Waltz, The Spread of Nuclear Weapons: More May Be Better, Adelphi Papers No. 171 (London: IISS, 1981); Jervis, The Meaning of the Nuclear Revolution; Jervis, “The Political Effects of Nuclear Weapons: A Comment”; and John J. Mearsheimer, The Tragedy of Great Power Politics (New York: Norton, 2003), p. 129. 9. See, for example, Marc Trachtenberg, “The Influence of Nuclear Weapons in the Cuban Missile Crisis,” International Security, vol. 10, no. 1 (Summer 1985), p. 139. 10. See, for example, Matthew H. Kroenig, “Time to Attack Iran: Why a Strike Is the Least Bad Option,” Foreign Affairs, January/February 2012, pp. 76–­86.

Introduction 7

regional powers adopted—­ because scholars assumed that these details were not relevant to deterrence or conflict dynamics. This consensus among scholars and policy makers is wrong. As I show in the second part of the book, regional nuclear powers have achieved widely different deterrence results with their nuclear arsenals. Even during the Cold War, the superpowers’ postures eventually evolved to establish a degree of mutual stability,11 with differential deterrent effects as their postures evolved over time.12 The effect for regional powers is even more pronounced. All of the regional nuclear powers developed nuclear forces along similar orders of magnitude (less than several hundred), and all faced the constraints and opportunities imposed by having to operate and maneuver below the superpowers. Yet some of them have been more successful in deterring their adversaries than others. Pakistan has successfully deterred Indian conventional attack on numerous occasions, but India has not been able to do likewise, as the 1999 Kargil War demonstrated. Even with nuclear weapons, Israel experienced serious deterrence failures against its Arab opponents in 1973 and 1991. Arguments about existential deterrence have no way to explain this variation in the rates of deterrence success. These two bodies of literature on nuclear weapons therefore leave a major theoretical gap in our knowledge on deterrence. On the one hand, there is a significant body of work focusing on superpower nuclear strategy during the Cold War and why it was important to deterrence dynamics. On the other hand, since the end of the Cold War the rise of an existential deterrence bias has suggested that everyone else’s nuclear strategy has no impact on their ability to deter conflict—­a belief that Cold War superpower deterrence theories and contemporary evidence suggest is fundamentally mistaken. Why have states with similarly sized arsenals had such differential success in deterring conventional conflict? I return to the intuition that guided early nuclear strategists and practitioners: nuclear posture and strategy, particularly for the regional powers, matters for deterrence and conflict stability. The conflict and deterrence experiences of the regional nuclear powers, which have thus far been overlooked, can provide insight into what kinds of nuclear forces are required to deter conflict. I find that differences in nuclear posture generate variation in a state’s ability to deter different types and intensities of conflict.13 There is little evidence that regional 11. This is a vast literature but see, for example, Freedman, Evolution of Nuclear Strategy. 12. Richard K. Betts, Nuclear Blackmail and the Nuclear Balance (Washington, DC: Brookings Institution Press, 1987); and Lieber and Press, “How Much Is Enough? 13. For an excellent discussion of the deterrence tradeoffs in theory facing emerging nuclear powers in terms of security environment versus civil-­military relations, see Peter Feaver, “Command and Control in Emerging Nuclear States,” International Security, vol. 17, no. 3 (Winter 1992–­3), pp. 160–­87.

8

Chapter 1

powers have reaped systematic deterrence benefits from the mere possession of nuclear weapons. Rather, their deterrence success is a function of their nuclear posture, and some postures fail to deliver on their promise to deter attacks. Nuclear posture, not just nuclear weapons, determines critical patterns of conflict and peace in international relations.

The Argument in Brief Largely forgotten in the scholarly and policy obsession with the Cold War and with nuclear acquisition is the fact that the regional powers have chosen different nuclear strategies. These differences matter greatly to their ability to deter conflict. I take these issues in turn. The first part of the book asks: which nuclear postures have regional powers adopted, and why? I begin in chapter 2 by identifying the diverse nuclear postures adopted by the regional powers, using original data collected from the field, and then develope a new theory for their selection. I classify and characterize three possible regional power nuclear postures, arrayed across a spectrum of capabilities and deployment options. • A catalytic posture, which consists of only a handful of nuclear weapons, threatens the explicit breakout of nuclear weapons in the event the state’s survival is threatened in order to compel—­or catalyze—­third-­party intervention on the state’s behalf. For example, Israel and South Africa adopted this posture for a significant portion of their nuclear histories. • An assured retaliation posture involves the development of secure second-­ strike nuclear capabilities that enable a state to threaten certain nuclear retaliation should it suffer primarily a nuclear attack. This posture has been adopted by India and China. • An asymmetric escalation posture develops capabilities and procedures that credibly enable the rapid and first use of nuclear weapons in the event of a conventional attack. France and Pakistan have each, at some point, adopted this posture.

Given these three alternative nuclear postures, why might states select one over the others? Current alternative theories in international relations that may explain a state’s nuclear posture—­security environment, technological determinism, and strategic culture—­are unsatisfactory and often indeterminate. Therefore, I propose a novel theory of the sources of nuclear posture: Posture Optimization Theory, or optimization theory for short. Taking inspiration from the neoclassical realist school of international relations theory, it explains how and why a regional power might select and optimize its nuclear posture in response to external security and internal

Introduction 9

domestic political and financial constraints. It offers a determinate prediction for a state’s nuclear posture based on several clearly identifiable and sequential variables. A state’s security en­vironment—­the availability of powerful allies and the severity of its immediate threats—­is critical and, indeed, the primary variable responsible for nuclear postures. But where that security environment is permissive, as is often the case, states have a range of choice in nuclear postures. Other considerations at the domestic level, such as civil-­military relations and resource constraints, regulate a state’s choice of nuclear posture. For example, the asymmetric escalation posture is not only financially and organizationally demanding, but it forces a state’s leadership to be prepared to devolve nuclear assets and authority to military end users in order to maintain the credibility of first-­use options. This increases the risk of unauthorized and accidental use of nuclear weapons and can impose tremendous strain on a state’s civil-­military organs. As such, states with permissive security environments that can opt out of an asymmetric escalation posture may find it rational to do so if adopting this posture is incompatible with other unit-­level preferences and constraints—­ especially those imposed by the configuration of the state’s civil-­military relations. This theory provides the most valid available framework for why states select the nuclear postures they do, providing testable and falsifiable predictions for the nuclear postures emerging and future regional nuclear powers might select in the future, and specifying when and under what conditions existing regional nuclear powers might shift postures. It is important to note that this theory attempts to predict and explain posture outcomes, not necessarily the process or efficiency by which a state adopts a particular posture (which is a separate question demanding a different theory and study). Chapters 3 through 8 focus in detail on the nuclear postures of each regional power: Pakistan, India, China, France, Israel, and South Africa. For each, I describe the specific nuclear strategy that the regional nuclear power has adopted over time. From an empirical perspective, this discussion on the actual nuclear postures of the regional powers over time provides the first comparative treatment of all six regional nuclear powers. I then test each state’s choice of nuclear posture against my theory and the three alternative explanations: structural realism, technological determinism, and strategic culture. I analyze whether the form and rationale for each state’s selection is consistent with each of these four theories. For example, I show that Pakistan has shifted from a catalytic posture to an asymmetric escalation posture and deploy my theory to explain these choices. Similarly, I offer an explanation for why India, China, and now Israel have all opted for small but secure assured retaliation postures, even though they have the capability to adopt more aggressive strategies.

10

Chapter 1

From a theoretical standpoint, these chapters provide the first broad explanation for why these states have adopted the nuclear postures that they have. They provide a framework that helps theorists and policy makers alike think about the variables that might drive regional powers to adopt a specific nuclear strategy, and understand the conditions under which states might shift strategies. The empirical tests establish the validity of the theory, thereby providing testable predictions for the postures that possible future nuclear states, such as Japan or Iran, might adopt. The variation in regional power nuclear strategies is only important, however, if there are consequences to the different choices these powers make. Thus, the second part of the book focuses on the ramifications of these choices about nuclear posture for international security. It asks the critical question posed earlier: what kind of nuclear strategy is required to deter conflict? Using both quantitative and qualitative analysis, I find that the mere possession of nuclear weapons fails to systematically deter conventional attacks. Chapters 9 and 10 demonstrate that there are very real differences in the deterrence consequences of these various nuclear strategies: some nuclear postures fail to deliver on their promise to deter conflict. In fact, only those states that adopt an asymmetric escalation posture enjoy significant deterrent success against conventional attacks. The catalytic and assured retaliation postures fail to do so because the risk of nuclear use even in intense conventional conflicts is so low that it does not deter opponents from attacking these nuclear powers—­sometimes resulting in conflicts of very high intensity. Chapter 9 conducts a large-­n analysis to isolate the average effects of these nuclear postures in reducing armed attacks at various levels of intensity. This analysis illuminates the general deterrence effects of each nuclear posture, providing an estimate of how many fewer attacks a state can expect to experience after adopting a particular nuclear posture. It shows that the asymmetric escalation posture is uniquely “deterrence optimal,” reducing conflict at each level of armed intensity against both nuclear and non-­nuclear powers. Contrary to conventional wisdom about the deterrent power of possessing any nuclear weapons, states adopting assured retaliation and catalytic postures have experienced serious deterrence failures even at high levels of conventional conflict intensity. The implication is that even secure second-­strike nuclear forces may not be sufficient to deter the initiation of full wars against a regional nuclear power. Certainly, nuclear weapons alone are insufficient to reap any significant deterrent effect. These findings fundamentally challenge the assumption that the mere possession of nuclear weapons provides substantial deterrent benefits. Chapter 10 focuses on the effects of nuclear postures in particular crises. These crises, selected in part based on the large-­n analysis in the previous chapter, tease out the mechanisms that connect nuclear posture to

Introduction 11

deterrence success and failure. I examine a series of crises in enduring rivalries between India and Pakistan, and between Israel and the various Arab states over time. Examining crises in enduring rivalries has the benefit of holding many variables constant in crises where many moving parts would otherwise make it difficult to isolate the effect of nuclear postures. In this chapter, I probe how different nuclear postures have affected states’ decisions to escalate or de-­escalate a crisis, measuring differential deterrence effects of postures on both the outbreak and the course of the crisis. I augment the empirical richness of the data on these crises with interviews with key national security decision makers, especially in the India-­Pakistan cases. This chapter confirms the findings from the large-­n analysis, showing that not only does the asymmetric escalation posture uniquely reduce armed attacks, but decision makers are deterred from attacking an asymmetric escalator because of the fear of nuclear first use. Catalytic postures have resulted in high-­intensity wars being waged against states, and even assured retaliation postures have been unable to deter high-­intensity wars, such as the 1999 Kargil War launched by Pakistan against India. On the other hand, when Pakistan shifted from a catalytic to an asymmetric escalation posture, it enjoyed a marked increase in deterrence success against India. Contrary to the conventional wisdom—­indeed, contrary to a bedrock article of faith in the canon of nuclear deterrence—­the acquisition of nuclear weapons does not produce a uniform deterrent effect against opponents. Despite the arguments of scholars including Kenneth Waltz, Robert Jervis, and John Mearsheimer, the possession of nuclear weapons by itself, and even the adoption of secure second-­strike forces, is insufficient to systematically deter conventional conflict. This finding overturns a central belief of modern deterrence theory, held for more than half a century, in the efficacy of nuclear weapons possession. The driver of deterrence success is not nuclear weapons, it is nuclear posture. Nuclear weapons may deter, but they deter unequally. States that wish to deter conventional attacks with nuclear weapons must explicitly orient their nuclear forces to do so. I conclude in chapter 11 with some implications for our understanding of nuclear deterrence and nuclear proliferation in a world where an increasing number of regional powers are pursuing nuclear weapons. The most important finding for theory is that nuclear weapons do not produce a uniform deterrent effect. Not only have limited nuclear arsenals had significant deterrence failures, but so have nuclear forces that have attained secure second-­strike capabilities. This significantly revises the conventional understanding of what it takes to deter conflict with nuclear weapons. The key variable in generating deterrent power against conventional conflict is not simply nuclear weapons, but nuclear posture. While the acquisition of nuclear weapons is an important step for regional powers, what comes

12

Chapter 1

afterward, and the pressures regional powers face in adopting nuclear strategies, are more important to the texture of international politics. Nuclear posture is the variable that produces differential deterrent effects, the factor that affects the frequency and the intensity of international conflict. The book also offers a vocabulary and a framework with which to think about the unique choices and pressures driving regional power nuclear strategy. It provides the first general framework for predicting how current and future regional nuclear powers might operationalize their nuclear weapons. Determining the type of posture a state might adopt provides insights about the frequency and intensity of conflict the region and the international system might expect, and thereby offers a blueprint for managing the existing proliferation landscape as well as the one that is unfolding. The coming decades are likely to bear witness to more volatile rivalries involving regional nuclear powers. Understanding a state’s likely nuclear strategy in that environment is important because these postures carry important tradeoffs that affect international security. For example, as Pakistan has demonstrated since 1998, states with an asymmetric escalation posture may reap significant deterrence benefits but they may simultaneously face substantial difficulties in securely managing that posture. Based on the projected risks to nuclear security, the international community might then be able to either ameliorate security predicaments driving the state toward such an aggressive nuclear posture, or provide command-­and-­control assistance so that the chosen nuclear posture can be implemented as safely as possible. In combination, understanding the sources and consequences of regional power nuclear postures provides a template for predicting—­and managing—­where and when nuclear proliferation could be particularly dangerous. The two analytical questions that animate this book—­the sources and deterrence consequences of regional power nuclear strategies—­advance our understanding of nuclear proliferation and deterrence dynamics by moving beyond the twin problems of the Cold War hangover and the acquisition bias that pervade the nuclear proliferation literature. A remarkable literature was penned on the evolution of the workings of the superpower nuclear balance, but these choices and experiences are unlikely to recur. The bulk of current nuclear weapons states, and every one that will emerge from this point forward, faces a different set of constraints and opportunities than the United States and the Soviet Union. But we have little systematic understanding of what comes after their initial decisions to “go nuclear.” I show that regional powers have developed innovative nuclear strategies and that these choices matter deeply to international relations and conflict.

Chapter Two

The Sources of Regional Power Nuclear Postures: Posture Optimization Theory

The regional nuclear powers have thought about and selected nuclear strategies and postures that are fundamentally distinct from those of the superpowers. In this chapter, I identify three main types of regional power nuclear postures, arrayed across a spectrum of capabilities and deployment procedures. I then present my theory for why regional powers might select a specific nuclear posture. That is, given the diversity of alternative nuclear postures available to regional powers, why do states select one posture over the others? I approach this question by exploring the variables that should be expected to regulate a regional state’s choice of posture. My theory, the Posture Optimization Theory, explains why the existing regional nuclear powers have adopted the nuclear postures and strategies they have, and generates testable predictions about what type of nuclear posture future regional nuclear powers might adopt based on a set of readily observable variables. Because of the dearth of regional states that have acquired nuclear weapons, this exercise has inherent limitations. However, by testing my framework against the existing empirical record, we can determine whether it provides a plausible framework with which to explain the choices of existing regional powers and to predict the choices that future nuclear powers might make. This first portion of the book has significant implications for how scholars and policy makers should think about nuclear proliferation as an increasing number of regional actors march toward nuclearization. By providing a testable and internally coherent framework for predicting the type 13

14

Chapter 2

of nuclear posture a regional power might adopt, and then the frequency and intensity of conflict that might result, we can gain analytical leverage regarding where and under what conditions nuclear proliferation might be most dangerous. I begin by characterizing the diversity of regional power nuclear postures, then review the few existing explanations for regional states’ choice of nuclear posture before developing my own theoretical framework. I operationalize that framework, as well as three alternative explanations, before proceeding to assess their performance against the empirical record in the subsequent chapters.

A Typology of Regional Power Nuclear Postures Unlike the superpowers, which both developed massive nuclear architectures capable of everything from tactical first use, to counterforce strikes, to assured destruction, the regional powers have had to make choices about how and where to allocate their deterrent power. For the superpowers, nuclear posture was often about the sequence, targets, and volumes in which tactical and strategic nuclear weapons might be used in a given contingency. Although this is an oversimplification, strategies popularly known as flexible response, massive retaliation, assured destruction, and damage limitation were largely variations on the same theme involving first-­use capabilities and sufficient retaliatory forces, and evolved as the superpowers’ nuclear architectures grew. On the other hand, regional nuclear powers—­which face different systemic and domestic constraints than the superpowers—­have adopted varied, but identifiable, nuclear postures across a spectrum of capabilities, management procedures, and transparency. These varied choices create a spectrum of distinct and identifiable postures available to the regional powers, which necessarily requires a different vocabulary because the size and character of regional power strategies is fundamentally distinct from those of the superpowers. That is, a typology of superpower nuclear strategies cannot be easily superimposed on the regional powers. To understand the choices that the regional nuclear powers have made, a novel classification scheme is required. In generating a typology of regional power nuclear postures, one must take care to categorize postures in an analytically distinct yet internally coherent way that does not devolve into identifying a separate posture for each regional power. I distinguish between three analytically distinct regional power nuclear postures: the catalytic posture, the assured retaliation posture, and the asymmetric escalation posture. Each of these postures gives different answers to the key questions motivating the selection of nuclear posture: what am I seeking to deter and how can I best do

Sources of Regional Power 15

so? These postures are clearly differentiated by their primary envisioned employment—­political catalysis, nuclear retaliation, and nuclear first use—­ their capabilities, their command-­and-­control architectures, and the levels of transparency regarding the latter characteristics. In practice, regional powers have to make choices under multiple constraints about how to allocate their scarce nuclear forces and toward what deterrent end. Unlike the superpowers, who could attempt to deter the full spectrum of conventional and nuclear conflict with their large nuclear forces, regional powers operate on the razor-­thin margins of nuclear force structure and therefore adopt conceptually and empirically distinct nuclear postures that are largely mutually exclusive—­only some regional powers have developed second-­strike capabilities, and only some have developed tactical nuclear weapons. These choices very clearly constrain the type of nuclear posture a regional power can adopt. Furthermore, because these postures are sticky and can take years to develop, a regional power’s peacetime orientation captures the posture available to it in a crisis or conflict as well—­a state cannot develop novel forces or command-­and-­control procedures during a crisis or conflict. Thus this typology of the distinct nuclear postures adopted by regional powers is both analytically useful and empirically accurate. Although the three postures may have minor variations within them, each regional nuclear power clearly falls into one of these three categories, suggesting that they are mutually exclusive and empirically exhaustive. Catalytic. A catalytic nuclear posture primarily envisions “catalyzing” third-­party—­often American—­military or diplomatic assistance when a state’s vital interests are threatened.1 It can do so by threatening to break out known nuclear weapons capabilities or previously ambiguous or non-­ operational nuclear capabilities and escalate a conflict if assistance is not forthcoming. It depends on the existence of a more powerful third-­party patron whose interests in regional stability—­or in forestalling the client state from overt nuclear breakout—­are sufficiently high, and the costs of assistance commensurately acceptable, that it might be compelled to intercede on the state’s behalf to effect de-­escalation. It is therefore a posture available only to regional powers, which can employ it to augment external balancing, and was by definition an option unavailable to the superpowers. 1. This term was used to describe South Africa’s nuclear posture, which envisioned using ambiguous capabilities as a strategy to compel American intervention on its behalf in Angola. In addition to South Africa, both Israel and Pakistan have employed this strategy at various points. For a description of the term with respect to South Africa, see Terence McNamee, “The Afrikaner Bomb: Nuclear Proliferation and Rollback in South Africa,” in Avner Cohen and Terence McNamee, Why Do States Want Nuclear Weapons: The Cases of Israel and South Africa (Oslo: Norwegian Institute for Defence Studies, 2005), p. 14.

16

Chapter 2

Because even a small risk of nuclear use may be sufficient to trigger third-­ party intercession, this posture can be executed with a limited arsenal that may not even be fully assembled or functional. Because it often relies on high levels of ambiguity surrounding capabilities and conditions of use, the catalytic posture tends to emphasize centralized control and does not integrate nuclear weapons into a state’s military doctrine. The key feature of the catalytic posture is that the state in question does not have survivable second-­strike forces or tactical nuclear weapons, and deterrent signals are not sent primarily to the envisioned opponent but rather to a third party in an attempt (successful or not) to induce or blackmail its intervention on its side. Certainly, as an extreme last resort, the state could potentially use nuclear weapons on an adversary, and so this posture ought to directly deter attempts to extinguish a state outright. However, direct use is not the primary envisioned employment and, indeed, the nuclear forces may be so recessed that assembly and use may not be a trivial matter. Indeed, employing small numbers of nuclear capabilities to catalyze third-­party intervention to augment conventional deterrence or to effect de-­escalation through external balancing could offer more deterrent power than attempting to directly deter adversaries with such limited nuclear forces. Therefore, theoretically and empirically, it would be optimal for regional powers with this form of nuclear capabilities and command and control to adopt primarily a catalytic posture, which is backstopped by the threat of use as a last resort. There is thus a core element of existential deterrence, but it is a secondary signal, should the catalytic posture fail. This posture might best be thought of as “existential-­plus” deterrence where there may be ancillary existential deterrent effects, but the primary aim is to compel a third party to intervene on the state’s behalf. The attempt to draw in third-­party intervention is thus the defining feature of a catalytic posture, regardless of whether that attempt succeeds (though empirically it often has). Because third-­party intervention is only probabilistic, the catalytic posture still may not have a strong direct deterrent effect on adversaries, since they may calculate that they can achieve limited conventional war aims before third-­party intervention occurs or without triggering intervention altogether. As a quintessential illustration, Israel adopted this posture and executed it during the 1973 Yom Kippur War. Several days into the war, as Israel feared that Syrian and Egyptian forces were threatening its survival—­ Israeli leaders had no way to know at the time whether Arab aims were limited or total—­Israel conducted operational checks on delivery vehicles in a manner that was easily detectible only to U.S. intelligence, signaling that it was considering unsheathing its opaque nuclear weapons capability. Israel’s explicit goal was to galvanize the U.S. government into both

Sources of Regional Power 17

rearming its military with conventional weapons and pressuring the Soviet Union to rein in its Syrian and Egyptian clients.2 The distinguishing feature of the catalytic posture is that Israel intentionally directed its nuclear signal at the United States, not at Egypt or Syria (indeed, Israel’s nuclear capabilities failed to deter their initial assaults and subsequent escalation). Two other states have also adopted catalytic postures, anticipating American intervention on their behalf and orienting their nuclear forces to ensure that Washington might do so: South Africa during the 1980s and Pakistan in the late 1980s.3 Because bipolar competition enabled regional powers to more easily exploit superpower behavior through the risk of nuclear escalation, this posture may have been easier to implement and execute during the Cold War. However, there is no theoretical reason why the catalytic posture should be limited only to the Cold War, since it simply requires the perceived reliability of a greater power third party. Assured Retaliation. Unlike the catalytic posture, which relies on indirect deterrence through uncertain third-­party intervention, the assured retaliation posture seeks to directly deter nuclear attack and coercion. It does so by threatening an opponent with certain nuclear retaliation even after sustaining damage. The assured retaliation posture is therefore distinguished by the presence of survivable second-­strike forces capable of targeting an opponent’s key strategic centers with definite, though not necessarily immediate, retaliation. Survivability can be achieved by a variety of stewardship procedures (e.g., dispersion, concealment, and deception) or technical means (e.g., sea-­based systems) that render it impossible for opponents to be confident of achieving a disarming first strike. In addition to possessing survivable nuclear forces, a state employing an assured retaliation posture must have forces that are capable of penetrating the adversary’s defenses and imposing certain retaliation. There must also be greater transparency a­bout capabilities than in the catalytic posture, such that the intended opponent has no doubt about the state’s ability to retaliate with nuclear forces 2. See Avner Cohen, “The Last Nuclear Moment,” New York Times, October 3, 2003; Sey­mour Hersh, The Samson Option: Israel’s Nuclear Arsenal and American Foreign Policy (New York: Random House, 1991), ch. 10; Hermann Eilts in Richard Parker, ed., The October War: A Retrospective (Gainesville: University of Florida Press, 2001), p. 121; and Avner Cohen, “Nuclear Arms in Crisis under Secrecy: Israel and the Lessons of the 1967 and 1973 Wars,” in Peter Lavoy, Scott D. Sagan, and James J. Wirtz, eds., Planning the Unthinkable (Ithaca: Cornell University Press, 2000), p. 118. 3. See Peter Liberman, “Israel and the South African Bomb,” Nonproliferation Review, Summer 2004, pp. 1–­35; Peter Liberman, “The Rise and Fall of the South African Bomb,” International Security, vol. 26, no. 2 (Autumn 2001), pp. 45–­86. In addition, though it is outside of the scope of this study because it does not yet have a credible nuclear weapons capability, it could be argued that Japan has a broad catalytic nuclear posture, using the threat of potential breakout to ensure continued American military support and alliance.

18

Chapter 2

following a first strike,4 but deployment patterns can be ambiguous to enhance survivability. To be sure, nothing precludes a state with an assured retaliation posture from using nuclear weapons first in a conflict, but the posture’s centralized deployment patterns and procedures would make it difficult for nuclear weapons to be released immediately for battlefield use.5 Thus, rather than planning for nuclear use in a deterrence by denial mission against conventional forces (indeed, deployment patterns for tactical nuclear use can undermine survivability), the assured retaliation posture is oriented primarily for deterrence by punishment against high-­value targets. The clearest indicator of an assured retaliation posture is the development of secure second-­strike forces, but the absence of tactical nuclear weapons within a state’s arsenal. The assured retaliation posture can be operationalized in a variety of ways and, within this posture, states have an array of choices regarding capabilities and management procedures. However, the key differentiating feature of the posture is a nuclear force structure that is arrayed to potentially and plausibly survive an attempted disarming conventional or nuclear first strike and be capable of retaliating with nuclear forces. A state has achieved a credible assured retaliation posture when adversaries can no longer be confident of executing a fully disarming first strike—­and this point is often reached relatively quickly at the regional level since uncertainties around numbers and deployment patterns that renders a single nuclear weapon and delivery system survivable can be achieved at low cost and very early in a state’s nuclear program. China and India have both adopted assured retaliation postures. Each relies on a small but secure and survivable nuclear force, arrayed for an assured retaliatory strike against their primary opponents’ strategic targets.6 Both have paired a declaratory 4. At the regional level, targeting has been primarily countervalue, because the lack of accuracy and deployment procedures often preclude rapid tactical or hard counterforce use. The aim is not assured destruction or massive retaliation, but assured retaliation. Preauthorization may occur to survive a decapitation attempt, but not for war-­fighting purposes. 5. Centralization does not preclude rapid release of strategic nuclear weapons (sea-­or land-­based, for example), but does make it more difficult to deploy nuclear weapons with front-­line forces for tactical nuclear use. 6. On India see, for example, Ashley Tellis, India’s Emerging Nuclear Posture (Santa Monica: RAND, 2001); George Perkovich, India’s Nuclear Bomb (Berkeley: University of California Press, 1999); Raju G. C. Thomas and Amit Gupta, eds., India’s Nuclear Security (Boulder: Lynne Rienner, 2000); Jasjit Singh, ed., Nuclear India (New Delhi: IDSA, 1998); D. R. SarDesai and Raju G. C. Thomas, eds., Nuclear India in the 21st Century (New York: Palgrave, 2002); Rajesh Basrur, Minimum Deterrence and India’s Security (Stanford: Stanford University Press, 2006). On China, see for example, Jeffrey Lewis, The Minimum Means of Reprisal: China’s Search for Security in the Nuclear Age (Cambridge: MIT Press, 2007); Alastair Iain Johnston, “China’s New ‘Old Thinking’: The Concept of Limited Deterrence,” International Security, 20:3 (Winter 1995–­6), pp. 5–­42; Evan Medeiros, “Evolving Nuclear Doctrine”; John Lewis and Xue Litai, China Builds the Bomb (Stanford: Stanford

Sources of Regional Power 19

no-­first-­use policy with operational procedures that make the first use of nuclear weapons unlikely. But both assure nuclear retaliation should they sustain a nuclear strike (or, adversaries must assume, if a level of unacceptable conventional damage were sustained). Asymmetric Escalation. The asymmetric escalation posture is explicitly designed to deter conventional attacks by enabling a state to respond with rapid, asymmetric escalation to first use of nuclear weapons against military and/or civilian targets. Peacetime deployments can be centralized but, to credibly deter conventional attack, nuclear weapons must be operationalized as war-­fighting instruments. An asymmetric escalator therefore must have the ability to disperse and deploy nuclear assets quickly, pre-­delegating authority for their release to military end users on the front edge of the battle, who would be charged with employing tactical or strategic nuclear weapons in a deterrence by denial mission against an adversary’s advancing conventional forces or war-­production capacity.7 This can include both deterrence by denial—­using nuclear weapons to deny an opponent its military objectives on the battlefield—­and deterrence by punishment missions. This posture is thus the most aggressive option available to nuclear states.8 It does not require numerical superiority of nuclear weapons—­posture and the number of weapons are often erroneously conflated—­but depends instead on how a state arrays its nuclear forces and how it could credibly use them. To achieve credibility, asymmetric escalators must be transparent about capabilities, deployment patterns, and broad conditions of use, requirements that can generate significant command-­and-­control pressures and increase the risk of inadvertent use of nuclear weapons. The key distinguishing feature of the asymmetric escalation posture is the capability and expressed intention to use nuclear weapons in a tactical setting against an

University Press, 1988); John Lewis and Xue Litai, China’s Strategic Seapower: The Politics of Force Modernization in the Nuclear Age (Stanford: Stanford University Press, 1994); John Lewis and Xue Litai, Imagined Enemies: China Prepares for Uncertain War (Stanford: Stanford University Press, 2006); Chong-­Pin Lin, China’s Nuclear Weapons Strategy: Tradition within Evolution (Lexington: Lexington Books, 1988); and Shu Guang Zhang, Deterrence and Strategic Culture (Ithaca: Cornell University Press, 1992). 7. It is theoretically possible for a state with an asymmetric escalation posture not to have an assured retaliatory capability, though empirically we observe that states that adopt this posture do attempt to ultimately develop sufficient tactical and potentially survivable second-­strike strategic weapons to absorb potential retaliation. 8. While this posture may have a significant deterrent effect against conventional attacks, it can also enable a revisionist state to aggress at low levels against an opponent and use the posture as a shield behind which to achieve those objectives. Pakistan has always had limited revisionist preferences with respect to India and, though its primary aim seems to be to deter India, it has been able to more aggressively seek its revisionist preferences with the asymmetric escalation posture.

20

Chapter 2

adversary’s conventional forces. This can place extensive costs and strain on the operational management of nuclear weapons. States that select asymmetric escalation postures are often those that face extremely binding security constraints and therefore have little choice but to adopt an aggressive posture. France and Pakistan have both adopted asymmetric escalation postures. During the Cold War, French forces faced a conventionally superior nuclear-­armed proximate threat: the Soviet Union. In order to deter the Warsaw Pact’s superior conventional forces, and seized with the fear that it would be left alone to face the Red Army, France threatened first use of nuclear weapons against Soviet forces and strategic targets should they breach Western Europe.9 After testing nuclear weapons in 1998, Pakistan shifted to an overt asymmetric escalation posture to deter India’s superior conventional power. The precise strategy used with an asymmetric escalation posture can vary (e.g., massive retaliation vs. flexible response), but the key feature is the enabling of a credible, asymmetric first use of nuclear weapons against conventional aggression, in an attempt to deter its outbreak. The catalytic posture attempts to signal primarily third parties, and assured retaliation attempts to directly deter adversaries by holding strategic targets at risk, but both envision nuclear employment as last-­resort measures; asymmetric escalation is oriented toward nuclear use as a potential option in the early stages of a conflict. Although a state with an asymmetric escalation posture may be able to “do more” with its nuclear force structure because of the costlier investment in capabilities and command-­and-­control structure, the primary envisioned employment of these three postures is mutually exclusive. Thus, for example, although the asymmetric escalation posture may have unintended catalytic effects because an outside party might be alarmed, the critical distinction is that an asymmetric escalator explicitly sends early first-­use deterrent threats directly to its adversary and not to a third party; and an asymmetric escalator may or may not have survivable second-­strike forces (France does, for example, but for a long time Pakistan did not).10 However, a state with a catalytic posture cannot simultaneously adopt either assured retaliation or asymmetric escalation postures; and assured retaliation is both conceptually and practically distinct from a catalytic or asymmetric 9. For some classic works on French nuclear doctrine, see Wilfred L. Kohl, French Nuclear Diplomacy (Princeton: Princeton University Press, 1971); and Wolf Mendl, Deterrence and Persuasion: French Nuclear Armament in the Context of National Policy, 1945–­1969 (New York: Praeger, 1970). 10. It is theoretically possible to distinguish between an asymmetric escalation posture with secure second-­strike forces and one without. In my classification and coding of postures, once a state has an asymmetric escalation capability, regardless of the state of its retaliatory forces, I code it as having an asymmetric escalation posture, even though these are distinguishable possibilities.

Sources of Regional Power 21

escalation posture. I code a state’s nuclear strategy by what its posture is maximally capable of doing.

Summary of Postures Table 2.1 summarizes the features of these three nuclear postures and the criteria used for coding them. To classify each state’s nuclear posture, I examine its nuclear capabilities (does it have survivable forces and/or tactical nuclear weapons?), its command-­and-­control arrangements (is control pre-­ delegated or assertive and centralized?), the level of transparency regarding capabilities and deployment patterns, and finally, the envisioned use and orientation of a state’s nuclear arsenal. I apply these coding rules to each of the regional nuclear powers over time to generate the codings given below, and explicate these codings in substantial detail in the coming chapters. Some states, such as India, have maintained the same posture throughout their nuclear histories, while others, like Israel and Pakistan, have shifted postures at some point. The stickiness of the capabilities and organizational inputs that define a state’s nuclear posture has two virtues. First, it reduces measurement uncertainty. We can be confident that peacetime postures are both accurate and effectively capture what a state has available to deploy and employ should a conflict occur. A state that physically lacks tactical nuclear weapons or survivable second-­strike forces, or the organizations to employ them as such, cannot develop them in the time frame of a crisis or conflict. As Barry Posen put it with respect to conventional posture, which is less sticky than nuclear posture: “Initial battles of a war are fought with the equipment at hand. Decisions made long before the war will determine some operational possibilities during the war.”11 A state may operationalize its posture in a conflict or crisis, but it cannot easily change its posture. This maintains the mutual exclusivity of these postures, both in peacetime and war. Second, in practice, it means that regional powers adopt one of these postures, and adopt the same posture toward all of their potential opponents.12 These three postures are the basic array of choices available to regional powers. The catalytic posture is a strategy designed to ensure the intervention of an external balancer backstopped by the extreme last resort measure 11. Barry R. Posen, Sources of Military Doctrine: France, Britain, and Germany between the World Wars (Ithaca: Cornell University Press, 1984), p. 31. 12. For example, China’s assured retaliation posture against the United States gives it the same capability against all of its relevant opponents; India’s assured retaliation posture against China ensures it against all its relevant opponents. Similarly, states with a catalytic or asymmetric escalation posture toward one state impart it with the capability to employ it against all politically relevant opponents as well.

22

Chapter 2 Catalytic

Assured Retaliation

Asymmetric Escalation

Primary Envisioned Employment

Breakout capabilities to accelerate third-­ party assistance

Nuclear retaliation following significant damage

Nuclear first use, primarily on conventional forces in denial mission

Capabilities

Ability to assemble a handful of nuclear weapons

Survivable second-­ strike forces

First-­use capabilities

Management

Recessed and opaque

Assertive political control

Delegative (assets and authority integrated into military forces and doctrine)

Level of Transparency

Ambiguous capability and deployment

Unambiguous capability; ambiguous deployment

Unambiguous capability and deployment

Empirical Codings

Israel (1967–­1990) South Africa (1979–­1991) Pakistan (1986–­1997)

China (1964–­present) India (1974–­present) Israel (1991–­present)

France (1960–­present) Pakistan (1998–­present)

Table 2.1. Summary of regional power nuclear postures

of threatening nuclear use if all else fails. The assured retaliation posture is oriented toward nuclear retaliation. The asymmetric escalation posture is geared toward nuclear first use. There are certainly variations within each of these postures, but the basic conceptual distinctions are mutually exclusive. For example, the amount of damage a state threatens to inflict is not a distinguishing feature of a posture—­if it is threatened in retaliation, it is classified as an assured retaliation posture; if threatened as a first use, it is classified as an asymmetric escalation posture. Furthermore, all of these postures can be coupled with damage limitation strategies (e.g., hard counterforce or missile defenses), but those measures do not alter the basic nuclear posture a state has chosen. However, a distinct posture that has yet to be chosen by regional powers because of prohibitive resource demands is a “splendid first-­strike” counterforce posture that seeks to fully disarm a nuclear adversary’s nuclear forces and thus enable a state to establish nuclear superiority or wage conventional war after disarming its adversary. Because such a posture, as in the Cold War, requires a sophisticated real-­time intelligence architecture to locate and destroy an adversary’s nuclear forces, and is a recipe for an expensive nuclear arms race, the regional powers have entirely abjured it thus far. But it is important to note that this could be an attractive theoretical possibility for an enterprising regional power. In this study, however, I limit my analysis to

Sources of Regional Power 23

the postures that have been chosen by regional powers to date. Therefore, although there may be theoretical possibilities outside of my typology, they have so far proven empirically implausible. Additionally, while my typology may have variations within each category listed, the basic postures are distinguished by the key characteristics described here. This categorization allows me to ask the primary questions in which this book is interested: (1) which nuclear posture does each country adopt and why? and (2) do these differences in posture generate variation in the state’s ability to successfully deter conflict? That is, is simply having nuclear weapons enough, or does a state—­particularly a regional power—­have to do something more in order to realize deterrence benefits against conventional attacks? I now focus on the first question.

The Sources of Posture: The Prevailing Proliferation Literature Why might states select one of these specific nuclear postures over the others? There is much scholarship on why states seek nuclear weapons and why they test them when they do.13 For example, Scott Sagan’s landmark article identifies various pressures pushing states toward nuclear weapons: security concerns, domestic politics, and the pursuit of international prestige. Etel Solingen has identified the political economic reasons why some states might develop nuclear weapons and why others might not. Finally, Jacques Hymans develops a psychological model focusing on oppositional nationalist leaders who view testing nuclear weapons as a means to exert superiority over some “other” adversary. The question of why states seek nuclear weapons and why they test them when they do represents the overwhelming portion of the proliferation literature. However, I argue that the decision to acquire or test nuclear weapons is only one element of nuclear posture—­the capabilities that constitute an arsenal and how they are arrayed. We simply do not have a good understanding for why regional powers choose different nuclear postures and strategies. All of the nuclear powers have had at least latent security pressures that drove them to nuclearize in the first place, yet they have chosen different postures. Either other variables are necessary to explain the choice of posture or the nature of the security threat against these regional nuclear powers varies and requires further explication. In addition, states that supposedly have sought or tested nuclear weapons for the same 13. See, for example, Scott D. Sagan, “Why Do States Build Nuclear Weapons? Three Models in Search of the Bomb,” International Security, vol. 21, no. 3 (Winter 1996–­97), pp. 54–­86; Paul, Power versus Prudence; Solingen, Nuclear Logics; and Hymans, Psychology of Nuclear Proliferation.

24

Chapter 2

reasons (e.g., security) have sometimes chosen different postures, such as China and Pakistan. Likewise, states that have tested for ostensibly different purposes—­security versus international prestige—­have selected similar postures, such as China and India. Table 2.2 depicts the existing literature’s primary explanation(s) for why a particular regional state acquired nuclear weapons and then its resultant posture, illustrating that the correlation between the two is imperfect at best. Thus, while the pressures that drive a state toward nuclearization are clearly not unrelated to the resultant choice of posture, considerations unique to each step demand that decisions to nuclearize and decisions to adopt a particular posture require different theories. To date, there has been no systematic exploration of the choices nuclear states make afterward: how and why do states operationalize nuclear forces in the varied ways that they do? We presently have very poor answers to this question. The current literature on nuclear postures is an array of some very good studies of individual state postures, such as those of the United States,14 France,15 Israel,16 India,17 China,18 Pakistan.19 Most of these, however, tend toward historical and descriptive treatments of these states’ nu­­ 14. See for example Freedman, Evolution of Nuclear Strategy; Sagan, Moving Targets. 15. For some classic works on French nuclear doctrine, see Kohl, French Nuclear Diplomacy; and Mendl, Deterrence and Persuasion. 16. See for example Avner Cohen, Israel and the Bomb (New York: Columbia University Press, 1998); Yair Evron, Israel’s Nuclear Dilemma (Ithaca: Cornell University Press, 1994); Michael Karpin, The Bomb in the Basement (New York: Simon & Schuster, 2006); and Zeev Maoz, “The Mixed Blessing of Israel’s Nuclear Policy,” International Security, 28:2 (Fall 2003), pp. 44–­77; Hersh, The Samson Option. 17. See Tellis, India’s Emerging Nuclear Posture; Thomas and Gupta, India’s Nuclear Security; Singh, Nuclear India; SarDesai and Thomas, Nuclear India in the 21st Century; Basrur, Minimum Deterrence; S. Paul Kapur, Dangerous Deterrent: Nuclear Weapons Proliferation and Conflict in South Asia (Stanford: Stanford University Press, 2007); Bharat Karnad, India’s Nuclear Policy (Westport: Praeger Security International, 2008); Verghese Koithara, Managing India’s Nuclear Forces (Washington, DC: Brookings Press, 2012). 18. See for example M. Taylor Fravel and Evan S. Medeiros, “China’s Search for Assured Retaliation: The Evolution of Chinese Nuclear Strategy and Force Structure,” International Security, vol. 35, no. 2 (Fall 2010), pp. 48–­87; Johnston, “China’s New ‘Old Thinking’ ”; Evan Medeiros, “Evolving Nuclear Doctrine”; Lewis and Xue, China Builds the Bomb; Lewis and Xue, China’s Strategic Seapower; Lewis and Xue, Imagined Enemies; Lin, China’s Nuclear Weapons Strategy; Zhang, Deterrence and Strategic Culture; and Lewis, Minimum Means of Reprisal. 19. See Feroz Hassan Khan, Eating Grass: The Making of the Pakistani Bomb (Stanford: Stanford University Press, 2012); Samina Ahmed and David Cortwright, eds., Pakistan and the Bomb (South Bend: University of Notre Dame Press, 1999); Samina Ahmed, “Pakistan’s Nuclear Weapons Program: Turning Points and Nuclear Choices,” International Security, vol. 23, no. 4 (Spring 1999), pp. 178–­204; and Timothy Hoyt, “Pakistani Nuclear Doctrine and the Dangers of Strategic Myopia,” Asian Survey, vol. 41, no. 6 (2001), pp. 956–­77.

Sources of Regional Power 25

Regional Power Israel

Primary Ascribed Motivations for Nuclearization

Posture

Security

Catalytic & Assured Retaliation

Security & Domestic Politics

Catalytic

India

Status & Domestic Politics

Assured Retaliation

China

Security

Assured Retaliation

Pakistan

Security

Catalytic & Asymmetric Escalation

Status & Security

Asymmetric Escalation

South Africa

France

Table 2.2. Relationship between motivations for nuclearization and nuclear posture

clear postures, with few exploring the sources of nuclear posture in these particular states. On the one hand, single-­country studies provide an entry point into comparatively studying the nuclear postures of regional powers. On the other hand, they themselves do not provide a systematic comparative lens with which to view the development of nuclear postures across regional powers. There is very little existing work on the choices and trade­ offs confronting regional nuclear powers as a class of states. There are two exceptions to this. The first is Avery Goldstein’s work on Chinese, British, and French nuclear postures.20 Goldstein’s major contribution is to identify and systematically evaluate the unique security situation of the “lesser” nuclear powers during the Cold War, unified by their operation below the superpower plane and thus analytically distinct from the superpowers themselves. He argues that structural pressures led China, Britain, and France to pursue nuclear weapons, conditioned somewhat by their alliance options (or lack thereof). Goldstein, however, addresses a different question than the one of interest here. First, he focuses on these states’ decisions to rely on nuclear deterrence for their security and not on the different types of postures they chose, with Britain virtually forsaking an independent deterrent by placing its forces under de facto NATO control in 1958, France opting for a more robust asymmetric escalation capability, and China choosing limited “city-­busting” capabilities in its adoption of an assured retaliation posture. Second, by overlooking this variation in regional nuclear postures, Goldstein adopts a strict neo-­realist logic that privileges international security considerations and constraints above all else. While latent security pressures are clearly critical in a state’s decision to acquire both nuclear weapons and the resulting posture, other variables 20. Avery Goldstein, Deterrence and Security in the 21st Century (Stanford: Stanford University Press, 2000).

26

Chapter 2

may intervene to produce variation in the latter. Finally, in focusing on a subset of regional nuclear powers limited to the Cold War—­France, China, and Britain (which I exclude due to the lack of independence of British nuclear forces, as mentioned earlier)—­Goldstein’s study does not consider the full universe of cases in either space or time, particularly the critical experiences and choices of India, Pakistan, Israel, and South Africa. This may lead to erroneous inferences about the variables that might drive other regional powers’ choice of posture, both empirically and in the future. The second effort at a comparative analysis of emerging nuclear powers is Peter Feaver’s work on nuclear command-­and-­control arrangements in emerging nuclear states.21 This is the best theoretical treatment of how nascent nuclear states balance the so-­called always/never problem—­ensuring that nuclear weapons can “always” be launched when desired but “never” launched when not (this is also known as “positive” and “negative” controls, respectively)—­based on a state’s time urgency (security environment) and the stability of its civil-­military relations.22 His framework is a useful starting point that makes a valuable contribution by introducing civil-­ military relations as a key independent variable, but it is theoretically underdeveloped as a theory for nuclear posture. First, it does not attempt to explain why regional powers will choose the form of capabilities that they do. Command-­and-­control procedures are a critical component of nuclear posture, but they are only one of several components. Second, the framework is often indeterminate: when a state’s security environment predicts one form of command and control, but its civil-­military structure predicts another, which variable is determinate and under what circumstances? Finally, Feaver does not attempt to test his framework against any empirical cases. Does the pattern of command-­and-­control arrangements match the theoretical predictions? Are there other variables that might intervene to generate the observed empirical pattern of posture choice? Therefore, we presently have no systematic theory for why regional nuclear powers select the nuclear postures that they do. Answering this question would be an important advance since, as the second part of the book demonstrates, the acquisition of nuclear weapons alone does not have a consistent or uniform effect on conflict outcomes. It is nuclear postures that seem to matter, so having a theory of the sources of nuclear postures is of critical importance to understanding the effects of nuclear proliferation. In this chapter, I therefore attempt to identify the variables that might regulate the choice of posture and develop a deductive framework for the choices and tradeoffs a regional nuclear state may face. 21. Feaver, “Command and Control in Emerging Nuclear States,” pp. 160–­87. 22. Ibid., p. 182.

Sources of Regional Power 27

Sources of Nuclear Posture: Posture Optimization Theory At the most general level, I argue that states select nuclear postures in a way that optimizes their force structure for their external security environment and their internal threats and constraints. That is, states carefully calculate what strategy they require to deter their likely foes and what they are organizationally and financially capable of doing, and optimize their choice of posture in response. I call this “Posture Optimization Theory,” or just “optimization theory” for short. Specifically, I argue that the choice of posture is regulated by a series of sequential variables facing a state that are prioritized—­from the structural to the unit level—­to produce a specific posture. A state’s relative power position and the international structure determine what exactly a state must deter with its nuclear forces. Sometimes structural variables alone can determine a state’s nuclear posture. In other cases, however, the international threat environment permits a range of possible postures and it is domestic-­level variables that ultimately determine the final choice, based on what the state can organizationally manage and what it can afford. The framework I develop is thus a rationalist decision tree that privileges a state’s structural opportunities and constraints. It is important to note that, as a structurally based framework, this theory is about optimal outcomes and choices, not a theory about optimal process. There is no doubt that states may select what they believe to be an optimal posture but for a variety of factors, the process may lag or be slow and inefficient. But process and implementation demands a different theory or may be unique to each state. This is therefore strictly a theory of optimal choice of posture. When a regional nuclear power has a powerful patron state that might be compelled to intervene on its behalf and act as a decisive external balancer, it should select a catalytic posture since it can benefit from the third-­ party patron’s assistance without having to fully face the complications of managing an overt nuclear arsenal: the management costs and, in some cases, potentially crippling international sanctions or reactive nuclear proliferation by adversaries. If a state’s available options to augment deterrence through external balancing disappear and it faces extremely binding security constraints, a regional nuclear power has no option but to adopt an asymmetric escalation posture. However, when such binding security constraints do not obtain and a state has greater latitude, a series of unit-­ level intervening variables can lead a state to select an assured retaliation posture. The theory I develop, importantly, is determinate and testable—­ determinate in that structural and unit-­level variables should intervene at specific nodes to produce, in combination, a prediction of a single posture for every given state. As a result the theory can be tested against the

28

Chapter 2

empirical record, and used to predict the behavior of potential future nuclear-­armed states. My theory draws insights from the neoclassical realist school. Slim structural realist theories23 which argue that the dynamism in international politics is strictly a function of the distribution of aggregate material capabilities are often difficult to deploy to explain variation in state behavior.24 In the realm of nuclear posture and strategy, structural realism is deafeningly quiet about any variation that regional nuclear powers—­which all occupy roughly the same relative power position as each other—­might have.25 These are, after all, states that possess one of structural realism’s crucial power equalizers: nuclear weapons. Beyond that threshold, however, structural realism has little to say about the way in which nuclear capabilities will be chosen and managed. The paradox of structural realism is that it purports to explain the behavior of the system’s major powers, but it is precisely these states that have the power to thwart structural constraints. The indeterminacy in the structural realist school in general has led to the development of richer theories from within the realist camp that attempt to better capture the complexity of state behavior. The neoclassical realist school, in particular, has generated useful frameworks that can be applied systematically to state foreign and security policies.26 As Gideon Rose writes, neoclassical realists “argue that the scope and ambition of a country’s foreign policy is driven first and foremost by its place in the international system and specifically by its relative material power capabilities. This is why they are realist. They argue further, however, that the impact of such power capabilities on foreign policy is indirect and complex, because systemic pressures must be translated through intervening variables at the unit level.”27 There are thus two key variables in neoclassical realism, in analytic sequence. Pride of place goes to a state’s 23. See Kenneth Waltz, Theory of International Politics (New York: McGraw Hill, 1979); and Mearsheimer, Tragedy of Great Power Politics. 24. See, for example, Colin Elman, “Horses for Courses: Why Not Neorealist Theories of Foreign Policy,” Security Studies, vol. 6, no. 1 (1997), pp. 7–­53; Jack Snyder and Thomas Christensen, “Chain Gangs and Passed Bucks: Predicting Alliance Patterns in Multipolarity,” International Organization, vol. 44, no. 2 (Spring 1990), pp. 137–­69. 25. For example, Kenneth Waltz vaguely refers to states achieving secure second strike forces. But not all have, and there are various ways to achieve that state of affairs. See Waltz, “The Spread of Nuclear Weapons: More May Be Better,” Adelphi Papers, no. 171 (London: IISS, 1981). 26. See Gideon Rose, “Neoclassical Realism and Theories of Foreign Policy,” World Politics, vol. 51, no. 1 (1998), pp. 144–­72; Randall Schweller, Deadly Imbalances: Tripolarity and Hitler’s Strategy of World Conquest, (New York: Columbia University Press, 1998); William Wohlforth, The Elusive Balance: Power and Perceptions During the Cold War (Ithaca: Cornell University Press, 1993). 27. Rose, “Neoclassical Realism,” pp. 146.

Sources of Regional Power 29

security environment. The freedom that a state has to maneuver within the system is a direct function of how binding that security environment is; this is a dynamic condition dependent on the structure and severity of a state’s security environment as well as the state’s relative power in that environment. In the extreme, states facing a grave proximate security threat are forced by structural pressures toward security-­seeking behavior to mitigate the gravity of that threat. For these states, the predictions of neoclassical realism collapse into the neorealist prediction of strictly security-­seeking actors, where the constraints of the international system are so binding that an analysis of unit-­level variables is largely unnecessary. In this way, relative security environment “establishes the basic parameters of a country’s foreign policy.”28 States in less constraining security environments—­either because they have a relative power advantage or because the security environment is less threatening—­have more options from which to choose when acting in the international arena, beyond the desperate pursuit of security. For states facing less severe and immediate security threats, there is greater room to maneuver and unit-­level intervening variables, such as domestic political considerations, may ultimately determine a state’s security policy, since the permissive security environment may not punish the state for behavior that is not purely security-­seeking. With respect to nuclear postures, states finding themselves in more relaxed security environments may have the latitude to opt out of an asymmetric escalation posture that would place intense burdens on civil-­military arrangements and resource outlays. In particular, because of all the concerns of nuclear weapons management highlighted by scholars such as Scott Sagan and Peter Feaver, states that are able to trade off some “deterrent power” for arsenal security may be rational to do so.29 The strand of neoclassical realism from which I derive inspiration adopts an implicitly rationalist approach and treats states as relatively unitary actors, though not as strictly as pure structural realism.30 In the realm of nuclear posture, these are not heroic assumptions. Decisions regarding nuclear capabilities are made at the highest levels of leadership, though clearly organizational pathologies can intervene, as I explore below. In addition, choices about posture evolve over long periods because they involve the deliberate development of capabilities and organizational procedures. They are thus not produced by time-­urgent constraints, which is when 28. Ibid. 29. See Scott D. Sagan, The Limits of Safety: Organizations, Accidents, and Nuclear Weapons (Princeton: Princeton University Press, 1993); Peter D. Feaver, Guarding the Guardians: Civilian Control of Nuclear Weapons in the United States (Ithaca: Cornell University Press, 1992). 30. This is in contrast to the strand of neoclassical realism that privileges state intentions, which is not as strictly rationalist, and very difficult to ex ante test.

30

Chapter 2

assumptions of means-­ends rationality are more tenuous. Thus, assumptions of rationality and a relatively unitary actor are useful starting points. However, by appealing to insights from the neoclassical realist school, my theoretical framework allows for greater complexity so that unit-­level intervening variables that relax these assumptions slightly—­including civil-­ military relations for example—­can be introduced where appropriate. The value of neoclassical realism is that it trades “neorealism’s parsimony and rigor for greater richness and more determinate hypotheses about system dynamics and specific state behavior.”31 In privileging security constraints, it maintains the core of realist thought, but it accepts that powerful states have a more complicated set of options from which to choose in their international political behavior. This added layer of theoretical complexity offers the potential for added explanatory utility and determinacy, providing a more satisfying and useful approach for phenomena that do not necessarily imperil the survival of great powers. Clearly one objection to neoclassical realism is—­and has been—­that it simply smuggles in intentions or unit-­level variables in an ad hoc fashion in order to save a degenerative realist project.32 However, in order to avoid such an ad hoc or post hoc error, it is critical to develop theories in the neoclassical tradition that clearly, deductively, and systematically specify where and under what conditions unit-­level variables should intervene. In doing so, theories constructed in the neoclassical tradition can be—­and must be—­both testable and falsifiable.33 31. Schweller, Deadly Imbalances, p. 16. 32. See, for example, John Vasquez, “The Realist Paradigm and Degenerative versus Progressive Research Programs: An Appraisal of the Neotraditional Research on Waltz’s Balancing Proposition,” American Political Science Review, vol. 91, no. 4 (December 1997), pp. 899–­912; and Randall Schweller, “New Realist Research on Alliances: Refining, not Refuting, Waltz’s Balancing Proposition,” American Political Science Review, vol. 91, no. 4 (December 1997), pp. 927–­30. 33. I should add a word about the role of state intentions, that is, whether a state is status quo or revisionist vis-­à-­vis its primary adversary. It is certainly the case that this variable may be an analytically prior regulator for the selection of nuclear posture. States with revisionist preferences may select postures that diverge from those of states with status quo preferences, all else being equal. A state with revisionist preferences that aims to use conventional military power to achieve its revisionist aims against an adversary that has the ability to retaliate with land power—­against which nuclear weapons might be suitable for a deterrence by denial missions—­should theoretically adopt an asymmetric escalation nuclear posture that would provide a shield behind which to aggress. In other words, this posture, by threatening early first use of nuclear weapons in a conventional conflict, would enable a state with revisionist preferences to pursue aggression with virtual impunity as the nuclear posture would serve to curtail the opponent’s ground retaliatory options. This choice might be independent of the state’s security environment and domestic political configuration. However, I exclude this variable for two reasons. First, theoretically, including a variable on state intentions is often the source of unfalsifiability in neoclassical realist theories, since

Sources of Regional Power 31

A preview of my theoretical framework is outlined schematically in figure 2.1, with the sequence of the four relevant variables, each taking binary values, laid out to produce a specific prediction of regional nuclear posture. I now describe the variables of interest in sequence. I will return to how I operationalize them in the following section.

Security Environment Consistent with the neoclassical realist school, I prioritize and privilege a state’s security environment, as defined by the material external threats and opportunities confronting a state. States that pursue nuclear weapons do so because they face some baseline level of security threat, but there is certainly scope for variation in the immediacy and severity of those threats. Some regional nuclear powers may face existential threats, whereas others may face more latent security pressures that motivated their nuclear weapons programs. In addition, unlike the superpowers, which operated at the highest planes of the international system, the regional powers operate at a lower plane and therefore may have opportunities for superpower protection that may affect their choice of nuclear posture. I identify two structural conditions that, in sequence, should affect a state’s calculation about the optimal nuclear posture. These conditions determine the range of activity that a state must deter with its nuclear forces. Third-­Party Patron. The first structural condition a regional nuclear power ought to consider is whether it has the opportunity to avail itself of a reliable greater power third-­party patron for potential protection in a crisis or conflict. Irrespective of the nature of the state’s adversaries, the augmentation of a more powerful third-­party patron (often a superpower) in a conflict or crisis can be a decisive security determinant and deterrent to conventional aggression. States that believe that such a patron can be it is very difficult to measure ex ante, and thus one is often circularly left with behavior as an indicator of intentions. Second, in the case of regional nuclear powers, empirically, revisionist intentions—­especially systemic revisionism—­has been extremely rare. Of the six regional nuclear powers, five clearly had status quo preferences; only Pakistan had somewhat revisionist intentions, but even these were limited and its primary aim in adopting its nuclear posture is as a deterrent to Indian conventional invasion. Thus, although in theory a state with revisionist preferences against a proximate adversary may lean toward an asymmetric escalation posture, I cannot fully test this hypothesis given the empirical universe of cases. I therefore exclude it for the purposes of testability and falsifiability. I simply mention this variable here for completeness of theory. However, my theory begins with the empirical observation that all regional powers that have acquired nuclear weapons have had largely status quo preferences or, at worst, limited revisionist preferences (not systemic revisionist preferences). Thus the following theory does assume a logic for status quo states, who adopt nuclear postures to preserve something akin to the status quo.

32

Chapter 2 Regional Nuclear Power

Availability of Reliable Third-Party Patron? Yes

Catalytic

No

Facing Conventionally-Superior Proximate Offensive Threat? Yes

Asymmetric Escalation

No

Civil-Military Arrangements Assertive

Delegative

Assured Retaliation

Resource Constraints? Yes

Assured Retaliation

No

Asymmetric Escalation

Figure 2.1. The optimization theory of nuclear posture

compelled to reliably intervene on their behalf (as opposed to against them), perhaps through the threat of brandishing nuclear weapons, would be rational to select a catalytic nuclear posture for several related reasons. First, if it is possible for a regional power to do so, it can avail itself of a powerful external balancer to augment its security, either with conventional assistance or supplies, or through basic alliance dynamics. Second, for states with unoperationalized or ambiguous nuclear forces, the appeal to an external balancer may be preferable to nuclear breakout, which is the internal balancing alternative. Nuclear breakout can generate significant costs for the regional state including sanctions or a regional nuclear cascade, which might negatively impact the state’s security, particularly since most of the regional nuclear powers historically operated under a robust post-­1968 NPT nonproliferation regime. This third-­party option was by definition unavailable to the superpowers, so it is a unique option for regional powers that operate at the plane below. The availability of a reliable third-­party patron is generally a system-­ level variable that depends on the perceived ability of the regional power to compel or blackmail a greater power to intervene on its side. That is, it is dependent upon the availability of a great(er) power which the state believes is willing to and capable of decisively intervening in a crisis or

Sources of Regional Power 33

conflict on its behalf. Specifically, the regional power has to calculate (1) that the third party would have interests in intervening on its behalf in particular scenarios, and (2) that the costs to the third party of intervention, of whatever nature required to be successful, would be acceptable to the patron such that the likelihood of its acting is believed to be sufficiently high and reliable. For example, diplomatic intervention or even arms shipment may be an acceptable cost to a third-­party patron to achieve regional security goals or forestalling client state nuclear breakout in some cases, while risking a nuclear strike from the Soviet Union may not be. If these conditions exist, then the regional nuclear power has the option to select a catalytic nuclear posture. This posture offers a state the ability to compel and ensure that the third-­party patron intervenes before a mortal threat to the state arises. In these cases, the regional power exploits the third party’s incentive to prevent conflict or nuclear proliferation in that particular region and uses nuclear weapons as a bargaining chip to effectively blackmail the third party to intervene before the conflict escalates uncontrollably. This is essentially a specific form of chain-­ganging, as highlighted by Snyder and Christensen: “As a result, any nation that marches to war inexorably drags its alliance partners with it. No state can restrain a reckless ally by threatening to sit out the conflict, since the demise of its reckless ally would decisively cripple its own security.”34 Importantly though, this condition may not mean that the superpower is a full-­fledged ally of the regional power. Indeed, if the regional power were an ally it might not need to manipulate nuclear risks to compel the superpower to intervene on its behalf. Instead, the relationship between the superpower and the regional nuclear power may simply be one of overlapping interests in which the superpower has an overriding interest to prevent the regional power from fully unsheathing its nuclear capabilities or becoming engaged in a crisis that may escalate to nuclear use. The third party’s interest in intervening could be predicated on anything from avoiding nuclear use in a particular region of perceived vital interest, avoiding further horizontal proliferation should the regional power in question openly declare its nuclear capabilities, to the maintenance of stability in that region, to preserving the independence of a particular state (ally or not). If a reliable third party is available and nuclear weapons could be used to manipulate its intervention, a catalytic posture is optimal for the regional power because it provides an insurance policy for the state while reducing the risk of triggering a cascade of regional nuclear proliferation or bringing acute and costly nonproliferation pressure upon the state for becoming

34. Thomas Christensen and Jack Snyder, “Chain Gangs and Passed Bucks: Predicting Alliance Patterns in Multipolarity,” International Organization, vol. 40 (1990), p. 140.

34

Chapter 2

an overt nuclear power.35 It is also a dominant option over existential deterrence because of the augmenting of power provided by third-­party diplomatic or military intercession (this is one of the reasons why “existential deterrence” is not a theoretically attractive or empirically adopted nuclear posture at the regional level; states for whom a catalytic posture is not possible or attractive should move to one of the other two postures rather than attempting to rely on small or unoperationalized forces for deterrence). As I later show, the catalytic posture was successfully employed by Israel and Pakistan at points during the Cold War, and was explicitly adopted by South Africa during the entirety of its nuclear program. In all three cases, the external patron was envisaged to be the United States, whose interests during the Cold War, it was believed, would compel Washington to intervene on these states’ behalf in the event of a conflict that risked escalation to the nuclear level. The nature of American intervention envisioned by these regional powers ranged from diplomatic intercession in the cases of Pakistan and South Africa, to conventional military supplies in the case of Israel. France was similarly faced with determining whether the United States could be a reliable third-­party patron. However, since the nature of intervention France required to be decisive for its security predicament involved active American military involvement against the Soviet Union and risked nuclear attack on the American homeland, France would ultimately calculate that reliance upon the United States could not meet its security needs. It should be noted that in some cases, the belief that the United States would reliably intervene may have been a delusional belief held only in the capital of the regional power (e.g., Pretoria) and nowhere else. While the belief in the availability of a third-­party patron may have been greater during the Cold War, when regional states could exploit superpower competition to augment their own security, it is not simply a relic of the Cold War. For example, states continue to believe that the United States extends its nuclear umbrella to Japan and NATO allies, as well as a perceived security guarantee, and commitment to maintaining a qualitative military edge, to Israel. If there is no reliable third-­party patron available, or the credibility of the third-­party patron’s commitment is questioned by the regional nuclear power, then a regional state’s choice of posture can be determined by another structural variable. 35. As identified by Solingen in Nuclear Logics, political economic considerations play a role here. But since all of these potential catalytic states are already nuclear powers, it is not clear how far this theory goes in explaining the variation in vulnerability states already possessing nuclear weapons might have to sanctions if they conduct nuclear tests. This is a rich avenue for future research.

Sources of Regional Power 35

Immediate Security Environment. A state that lacks a third-­party patron must obtain security on its own. It must therefore operationalize its nuclear posture according to the severity of its immediate security environment. In particular, a state facing a conventionally superior proximate offensive threat should have no option but to adopt an asymmetric escalation posture. The condition just specified is one of the most binding—­if not the most binding—­security environment a state can face. It captures the state of the world where a country faces an existential threat from a state or combination of states that possesses conventional superiority. Each of the components of the definition above must obtain in order for the state to be forced by its security environment to adopt an asymmetric escalation posture. For example, a state facing a conventionally superior proximate threat that does not pose an offensive or existential threat—­for example one separated by territory, advantageous terrain, or water—­has the option to improve its conventional forces as a deterrent. It is only when that conventional superiority is usable across geographies and generates a potential existential threat that a state has no option but to adopt an asymmetric escalation nuclear posture. That is, in order to mitigate the deployable conventional superiority of its proximate opponent, the only option available is a posture that threatens the first use of nuclear weapons in the event of a conventional breach and that offsets the state’s conventional inferiority. Such a posture is the only credible position a state in such a security environment can adopt to deter a conventionally superior proximate offensive threat. However, this is the only security environment that is determinate with respect to nuclear posture. And this environment, though it does occur, is quite rare. Most regional nuclear powers are faced with conventional threats against which they have relative conventional parity or superiority (either quantitatively or qualitatively due to geography or net assessments). Others have substantial peripheral territory that enables a “defense in depth” strategy that can mitigate the conventional power of potential neighboring adversaries. This variable is readily observable since it is based strictly on deployable material capabilities and geography. For example, while Pakistan attempts to compensate for its numerical inferiority against India by operating on interior lines of communication, it cannot escape the fact that a large-­scale conventional war with its larger neighbor would occur across easily traversable plains and desert sectors. This condition renders Pakistan extremely vulnerable to India’s numerically superior conventional power. India, on the other hand, is buffered against China’s larger conventional land capability by the inhospitable terrain over which any land war between the two neighbors would have to be fought. As such, Pakistan’s primary security threat is presently a conventionally superior proximate offensive threat, whereas India’s primary opponents are adversaries against which it either has conventional superiority or natural

36

Chapter 2

buffers. This security environment conditions these states’ relative freedom in their choices of nuclear posture. So while regional nuclear powers that face a conventionally superior proximate offensive threat have little option but to choose the asymmetric escalation posture, regional nuclear powers in more permissive security environments have significant room to maneuver. In these cases, security environment is an insufficient predictor for nuclear posture, since these states can select either an asymmetric escalation or an assured retaliation posture without jeopardizing their survival. I identify two unit-­level intervening variables that might push a state toward a particular posture in those circumstances.

Unit-­Level Variables Midrange theories in the neoclassical realist tradition recognize that, at times, structural variables can be determinate. Often, however, they are not, because states often have significant power to overcome structural  con­ straints, or the constraints themselves are insufficiently binding. In these cases, unit-­level variables may be required to explain state decisions. Here, I identify the critical unit-­level variables that might affect a state’s calculation about nuclear posture in a systematic, testable and falsifiable, and ad hoc way. When relative power and the international structure non–­ allow for a range of possible nuclear postures, the variables that ought to determine the ultimate choice of strategy are: the state’s preferences and ability to organizationally manage its nuclear forces, and what the state can afford to build. Civil-­Military Arrangements. The first unit-­level regulator of posture choice is the nature of civil-­military arrangements. It is precisely the relationship between these two groups—­and associated organizations with their own pathologies and preferences—­that defines a state’s ability to manage its nuclear assets. The dilemma generated by civil-­military relations is nicely captured by Peter Feaver’s description: “The civil-­military problematique is thus a simple paradox: the very institution created to protect the polity is given sufficient power to become a threat to the polity.”36 The greater the capability and authority ceded to the military to protect the state from external threats, the greater the potential the military can be a threat to the state. How a state manages that paradox is critical to its ability to address external and internal threats. Civil-­military relations is a complicated variable, one that can be produced by a whole host of 36. See Peter D. Feaver, Armed Servants: Agency, Oversight, and Civil-­Military Relations (Cambridge: Harvard University Press, 2003), p. 4.

Sources of Regional Power 37

conditions—­including strategic culture, societal structural features, and security environment itself.37 Indeed, one concern might be that civil-­military relations are produced by a state’s security environment writ large—­the balance of external versus internal threat—­and therefore epiphenomenal. However, this variable can still have an independent regulatory effect for several reasons. First, states in similar security environments with similar regime types can have different civil-­military structures; this produces independent cross-­national effects of civil-­military relations. Second, even if civil-­military structures initially form due to the balance of external/internal security threats, once they become institutionalized and ossify, they can produce an independent effect on a state’s management of military and security affairs. With respect to selecting nuclear posture, a particular feature of civil-­ military relations, as identified by Feaver and Huntington with respect to conventional forces, is important: whether the civilian government exerts assertive control over the military or whether there are so-­called normal, or stable, relations in which substantial authority and operational control are delegated to the military. States with an assertive structure feature indicators such as: centralization of command authority at strategic and tactical levels to civilian authorities, coup-­proofing techniques, frequent rotation of officer corps, privileging sectors of the officer corps based on regime loyalty rather than merit, and so on. States with a delegative arrangement have largely “professional” militaries in which a greater degree of autonomy is ceded to the military in doctrinal development and operations, there are increased levels of decision-­making autonomy at lower levels of the officer corps, and less centralized physical control over military assets (e.g., deployed with end users). This intervening variable is critical in determining whether a state opts for an assured retaliation posture, even when the state has the resources to develop an asymmetric escalation posture. States with highly centralized and assertive civil-­military structures should be unwilling to delegate both the nuclear assets and the launch authority to the conventional military required for an asymmetric escalation posture. That is, when this variable takes the value of assertive, states simply do not have the domestic structures or willingness necessary to implement an asymmetric escalation posture. Absent a third-­party patron and a proximate offensive threat, an assured retaliation posture is the only option for states with assertive 37. For classic works on civil-­military relations, see Samuel P. Huntington, Soldier and the State (Cambridge: Harvard University Press, 1957); Morris Janowitz, The Professional Soldier (Glencoe: Free Press, 1960); Michael Desch, Civilian Control of the Military: The Changing Security Environment (Baltimore: Johns Hopkins University Press, 1999); Feaver, Armed Servants; and Samuel Finer, The Man on Horseback: The Role of the Military in Politics (New York: Praeger, 1962).

38

Chapter 2

civil-­military relations. The assertive/delegative arrangements over the military in general translate into very specific controls over a state’s nuclear arsenal. States with largely assertive civil-­military structures will favor negative controls over a state’s nuclear assets to prevent their unauthorized or accidental use by the military or others. These negative controls can take a variety of forms, from the technological (e.g., physical locks, permissive action link-­type technology, etc.) to the organizational and procedural (e.g., competing units stewarding components that comprise a usable asset or elite centrally controlled units responsible for managing strategic assets, a stovepipe structure with a dedicated strategic arm that does not interface with regular conventional units, making the adoption of an asymmetric escalation posture difficult to operationalize). States with largely delegative authority will favor positive controls that assure nuclear use when necessary, such as devolving launch authority to the field, developing hardened communications and early-­warning systems, and avoiding physical impediments to nuclear use under chaotic wartime environments. Favoring positive controls ensures that the nuclear weapons can be used, first if necessary, under all possible contingencies; it also requires a great deal of certainty that the officers charged with nuclear assets will not release them without appropriate authorization or against unauthorized targets. A delegative civil-­military structure enables a state to potentially select an asymmetric escalation posture. As Barry Posen has argued with respect to conventional doctrine, militaries should be expected to have a bureaucratic preference for the proactive nature of the asymmetric escalation posture over the more reactive assured retaliation posture since it allows them to retain the initiative, incorporate nuclear weapons into their array of war-­fighting doctrines, and develop organizational missions that ensure the prestige and growth of the armed forces.38 Certainly an assertive civil-­military relationship may be a proxy for some underlying structural feature that generates civilian fear of its military (e.g., ethnic fractionalization, regime type, etc.). Feaver notes that the greater the fear of coups, the greater the likelihood that civil-­military relations will exhibit patterns of assertive control.39 However, because the same civil-­military arrangements can be produced by a variety of conditions—­ not just fears of coups—­and vary across or within regime types, it is this indicator that is most relevant to predicting the types of stewardship procedures a state might employ in managing its nuclear arsenal. That is, states may develop assertive or delegative control over their military for a variety of reasons, some of which may dissipate even though the particular arrangement persists. Therefore, the key indicator of interest with respect to 38. Posen, Sources of Military Doctrine. 39. Feaver, “Command and Control in Emerging Nuclear Nations,” p. 177.

Sources of Regional Power 39

choices about nuclear posture is whether the state evinces assertive or delegative control over the military, and not necessarily the features that might produce that arrangement. In particular, regime type does not provide the appropriate variable because not all democracies possess delegative structures, and not all autocracies have assertive civil-­military regime types. For example, India is a democratic state that exhibits assertive civil-­military relations, while Pakistan has been a frequently authoritarian state with delegative civil-­military arrangements. States that lack a third-­party patron and that do not face a conventionally superior proximate offensive threat, and with assertive civil-­military structures should select an assured retaliation posture. Because these states’ security environment does not demand immediate release of nuclear weapons under most likely contingencies, the assertive nature of the civil-­military relationship suggests that strategic assets should also be under central control both during peacetime and in crises. An assured retaliation posture avoids ceding control of nuclear assets and authority to end users until a decision to release nuclear weapons has already been made by the central political leadership. In this case, states privilege checking potential internal threats to security rather than external threats because of the latitude granted by a relatively permissive security environment. However, states lacking a third-­party patron, not facing a conventionally superior proximate offensive threat, but with delegative civil-­military structures, again have additional latitude in the posture they select. Lacking a third-­party patron, these states that have stable or “normal” civil-­military relations will not choose a catalytic posture, but they can adopt either assured retaliation postures or asymmetric escalation postures. What regulates that selection? I argue that there is one final variable that determines a state’s selection of posture. Resource Constraints. The final variable of interest is concerned with the resources available to a regional power against its primary envisaged threat(s). This variable basically assesses whether a state has the financial and organizational resources required to sustain an asymmetric escalation posture, which is more costly to develop and maintain than the assured retaliation posture for reasons specified earlier (particularly at the regional level, where opponents often have only limited counterforce capabilities, enabling a state to achieve a secure second-­strike threshold through a variety of relatively cheap measures such as dispersion, dummy sites, and concealment).40 In essence, the state is faced with the following problem: 40. At the regional level, states face limited counterforce threats and can take cheap but effective measures to achieve survivability. That is, a regional power can establish an assured retaliation posture without having to develop, for example, costly submarine-­based forces, though it may nevertheless choose to do so. Assurance of retaliation can be achieved

40

Chapter 2

does it have enough relative technological and economic capacity to develop all of the systems required for the adoption of a full asymmetric escalation posture against its primary threat? Obviously, developing a full array of nuclear capabilities—­from the tactical to the megaton-­range strategic warheads—­across a wide range of delivery azimuths with the necessary command-­and-­control architecture is a powerful deterrent to both conventional aggression and nuclear coercion. But depending on the capabilities and resource endowment of the adversary, it can be tremendously technically and operationally difficult as well as prohibitively expensive to erect and maintain. Regional states with significant capital and technological constraints relative to their primary adversary find themselves on the short end of an arms race capability, and any effort to adopt an asymmetric escalation posture therefore risks sparking a spiral of offensive and defensive acquisitions that could bankrupt the state or render it vulnerable. If the state faces significant resource constraints, even if it has delegative civil-­military structures, the optimal strategy is to adopt an assured retaliation posture that may serve as a deterrent at the nuclear level but accept that the nuclear arsenal may be unable to significantly deter lower-­level conventional conflicts. States such as China, whose primary opponents were more technically advanced and had larger military-­industrial complexes (the USSR and United States in China’s case), simply could not expend the necessary resources to compete in a warhead, missile, or architecture race. But what it did possess was a relative manpower and territorial advantage against its primary adversaries, which allowed it to absorb some level of conventional attack without threatening the state’s existence. It is rational for such states to adopt assured retaliation postures that rely upon the threat of assured nuclear retaliation—­even if delayed—­in the event of a nuclear attack. However, states that do not have relative capital or technological resource constraints vis-­à-­vis their primary adversaries and that have stable civil-­military relations that allow for a delegative military structure are in the best possible position to adopt an asymmetric escalation posture. In this case, they do not confront an immediate and severe external security threat, they face few internal threats to stability, and they are not relatively constrained in their technological and military expenditures. As such, they with inexpensive measures such as concealment, deception, and dispersion. On the other hand, an asymmetric escalation posture requires a broader array of forces, must be kept at higher levels of alert, and is more organizationally intensive. As such the marginal and operating costs of an asymmetric escalation posture are higher than an assured retaliation posture over time. For example, in contemporary debates in the United Kingdom and France on cost-­cutting measures to their nuclear postures, the cost of maintaining just an assured retaliation sea-­based capability is calculated to be cheaper than maintaining full asymmetric escalation postures, though France continues to adopt the latter.

Sources of Regional Power 41

can reap all the deterrence benefits of an asymmetric escalation posture without its major drawbacks: financial cost and threats of unauthorized or internal use. That is, relatively rich and secure regional nuclear powers with normal civil-­military relations may have the luxury of being able to adopt an asymmetric escalation posture. As figure 2.1 depicted, not all variables matter for each regional power—­ these variables are prioritized and generate determinate predictions based on the values each variable takes as one moves down the decision tree. While there are several pathways to the asymmetric escalation and assured retaliation postures, there is only pathway to a catalytic posture. The catalytic posture should only be adopted by states that have a credible third-­ party patron available who is likely to intervene on its behalf in conflicts. This is the most cost-­effective option for states in this particular situation since it does not require the development of significant capabilities or support assets and does not necessarily provoke an arms race or harsh nonproliferation sanctions. A catalytic posture allows a state to augment its security through the availability of a powerful third-­party patron and avoids the costs of explicit proliferation: the potential triggering of a cascade effect in its region which could be net-­security detrimental, the potential price it would have to pay for possibly violating the international nonproliferation regime, and the potential costs of transforming a recessed deterrent into an operational one. However, the tradeoff is that the catalytic posture gambles the state’s security on probabilistic third-­party intervention. The assured retaliation posture can be adopted via two distinct pathways. In both cases, the state operates without a reliable third-­party patron but also does not face a conventionally superior proximate offensive threat. The first pathway involves a state in the above condition with assertive civil-­military relations that demand centralized control of nuclear assets. In this case, the threat to the center—­either from the military itself or from domestic groups that may be able to seize nuclear assets—­is deemed sufficiently high to adopt extensive negative controls, which excludes the possibility of an asymmetric escalation posture. The second pathway involves a state in the above condition and with delegative civil-­military structures but with significant resource constraints that preclude its ability to bear the costs and internal risks of an asymmetric escalation posture. That is, the state is in a lax security environment, has normal civil-­military relations, but is resource-­constrained from developing a posture beyond a basic assured retaliatory capability. In both cases, there are rational reasons why a state might opt out of an asymmetric escalation posture in favor of an assured retaliation posture, sacrificing some deterrent power in order to optimize other internal constraints. The asymmetric escalation posture can also be adopted through two different pathways. The first is the strictly binding security environment

42

Chapter 2

in which a state lacks a third-­party protector and, regardless of its civil-­ military structures, has no option but to select an asymmetric escalation posture in order to deter a conventionally superior proximate offensive threat. This is perhaps the most extreme security situation in which a state can find itself. The resulting choice of posture is entirely driven by the severity of its security environment. The second pathway for an asymmetric escalation posture is for a state in a less binding security environment that does not have—­or does not believe in the reliability of—­a third-­party patron, has delegative civil-­military structures, and has a relative financial/technological resource advantage over its envisaged threat(s). States in this position are free to maximize the deterrent power of the posture without incurring its major drawbacks. The asymmetric escalation posture can thus be both the curse of the severely threatened and the luxury of the rich and stable. The schematic depicted in figure 2.1, a theory of rational posture optimization, not only illustrates the pathways for why states might select a particular nuclear posture, but it also provides the conditions under which nuclear postures might change. If a state believes that it no longer has its third-­party patron, it may find itself quickly jumping from a catalytic posture to an asymmetric escalation posture if its primary adversary is a conventionally superior proximate offensive threat. Similarly, if a conventionally superior proximate offensive threat dissipates, then the regional power has several options and would follow the nodes sequentially to determine its optimal nuclear posture. In other cases, an exogenous shock (e.g., a change in regime type) that results in a reoriented civil-­military relationship may alter a state’s decision calculus about nuclear postures. In this way, the theory above not only provides an account of static choices of nuclear postures but also the reasons and mechanisms for shifts in nuclear postures.

Testing Sources of Posture I now turn to how I intend to operationalize, measure, and test my theoretical framework in the subsequent chapters. I test my framework against the record of all six of the regional nuclear powers—­China, France, India Israel, Pakistan, and South Africa—­so as to not induce any selection bias.41 I first describe each state’s nuclear posture in detail and how I have classified it over time using the measures for nuclear posture presented earlier. Second, in each of the country case studies, I measure values for each regulatory variable and trace the pathway through each node to see if my predictions 41. This is not to deny that there are selection effects in the initial decision for these states to pursue nuclear weapons. I seek to explain posture conditional on possessing nuclear weapons, so I do not test this framework against non-­nuclear states.

Sources of Regional Power 43

are accurate. Finally, I test my framework against three general primary alternative explanations: structural realism, technological determinism, and strategic culture. Clearly, with six cases (and nine different choices of posture over time) and four variables, it is difficult to conclusively prove my theory. However, the goal of my empirical tests in the subsequent chapters is to establish the validity of my framework by determining whether these regional powers adopted nuclear postures consistent with my predictions for the reasons hypothesized. In this section I primarily discuss my strategy for operationalizing the variables for sources of posture and then I proceed to lay out the three main alternative explanations.

Operationalizing the Variables The first variable, availability of a third-­party patron who might be compelled to intervene on the state’s behalf, is relatively easy to measure since it is based on empirical alliance patterns and behavior at the system level. This variable measures whether the regional power believes that a third party, often a superpower and more often the United States, has reliable and consistent incentives to intervene on its behalf in a conflict, which it could do for a variety of reasons: to prevent the state from being harmed in a conflict, or to prevent overt nuclearization by the regional power for nonproliferation reasons, or simply because regional stability is in the third party’s interest. This variable assesses whether the regional power believes that the stronger power has both the incentive and the military or diplomatic capacity to intervene in the event of a severe threat to it. In the case of Israel during the Cold War and Pakistan in the 1980s, there was sufficient reason in Tel Aviv and Islamabad to believe that the United States could not and would not allow a significant threat to materialize against them before intervening on their behalf given American general nonproliferation interests and specific interests in the Middle East, and in Afghanistan during the Soviet invasion in the 1980s. In this operationalization of the variable, it is a system-­level feature that is relatively straightforward to measure. In other cases, however, the belief may be strictly a unit-­level attribute where the regional power believes that a third-­party patron may exist, even if the third party has no actual intention of serving as a patron. That is, the belief may exist entirely in the minds of the political leadership of the regional power, even if the state is not a client state or even an ally of a greater power (e.g., South Africa in the 1980s and Pakistan in the 1990s after the American interest in Afghanistan disappeared). This unit-­level conception of the variable is slightly more difficult to measure, since it involves tracing beliefs among decision makers in the regional states about the likely third-­ party responses to conflicts. Critically, though, at both the structural and

44

Chapter 2

unit levels, the belief about third-­party availability has to be held ex ante. I therefore attempt to measure this variable independent of whether the patron actually intervenes in a conflict and irrespective of what posture the regional state ultimately chooses. The second variable to consider is whether a regional nuclear power faces a conventionally superior proximate offensive threat. This variable has several subcomponents, some easier to measure than others. The measurement of conventional superiority is coupled with the proximity of the threat and the terrains over which conflict would occur. These measures are based strictly on material capabilities and geography. An opponent or coalition of opponents is conventionally superior to a state if, in likely conflict contingencies, the net assessments based on quantitative, qualitative, and organizational characteristics suggest that the opponent (or coalition) is decisively conventionally stronger to the regional power of interest. This is a binary condition that does not require precise measurement of the degree to which superiority may obtain. This can be measured empirically, if the two states in question have fought wars in the past, or analytically based on forces, force ratios, and operational considerations. The extent of the threat is related to the final subcomponent of this variable: proximity. Geography matters, and states that share an unimpeded land border have more to fear than those buffered by impassable terrain.42 John Mearsheimer argues that “land power is the dominant form of military power in the modern world” and “large bodies of water [as well as the Himalayas] sharply limit an army’s power projection capability.”43 As such, states that share a land border across a desert or plains have more to fear from each other than states separated by large bodies of water or inhospitable terrain. The final condition, offensive threat, is a measure of the ability of an adversary’s conventional capabilities to be employed over the relevant geographies in a manner that could pose an existential threat to the state. This is not an easy measure, and is complicated by the difficulties of distinguishing offensive and defensive military technologies, but can nevertheless be proxied by an adversary’s operational doctrines and whether, for example, a regional power can erect a “defense in depth” strategy to mitigate a neighbor’s conventional power. In order for this variable—­conventionally superior proximate offensive threat—­to obtain, all three conditions must be met. That is, a regional nuclear power’s primary threat must be a conventionally superior opponent that shares an easily passable land border with it and has the capability to employ that superiority in a fashion that poses an existential threat to the state. 42. See Mearsheimer, Tragedy of Great Power Politics. Also see Stephen M. Walt, The Origins of Alliances (Ithaca: Cornell University Press, 1987). 43. Ibid., pp .83, 114; parenthetical comment added by author.

Sources of Regional Power 45

Third, measuring civil-­military relations relies upon the structure of conventional military relations with the political center to avoid the tautology of using nuclear command-­and-­control structures to predict nuclear posture. Some of the indicators used to measure general assertive versus delegative structures were discussed earlier: centralization of command authority at strategic and tactical levels to civilian authorities, coup-­proofing techniques, frequent rotation of officer corps, privileging sectors of the officer corps based on regime loyalty rather than merit. Using these indicators avoids the circularity of measuring civilian-­nuclear command structures to predict the same thing. The key feature I attempt to measure is the level of autonomy ceded to the military by a state’s political leadership. That in turn is usually an answer to part of the civil-­military paradox: how fearful is the political leadership of the military turning its coercive power back at the state itself? On one extreme are states with a tradition of military obedience to its civilian masters wherein the military professionalizes and executes and implements the directives of civilian leadership with high levels of trust, competence, and reliability. This is the Huntingtonian ideal,44 and the American example is often cited as the quintessential example of “normal” civil-­military relations evincing high degrees of autonomy and delegative authority, since fears of coups are minimal. On the other extreme is a military, or praetorian, regime where the political leadership and its coercive instrument, the military, are one and the same. In such cases, the military has a monopoly over the state, though there could certainly be internal challenges to the particular military head-­of-­state from within the military. In both these cases, however, the military often has a large degree of delegative authority—­the political leadership may provide the mission and set broad parameters, but strategy, operations, and tactics are devised and executed by the military without interference from civilians; in praetorian states, this is simply definitional.45 In the middle of this spectrum lie a set of states that are neither military states nor states with normal civil-­military relations. These states may vary in their degree of assertive control over the military, employing a variety of increasingly intrusive measures such as: selection and screening mechanisms, generating competing organizations to monitor and check the military, direct civilian monitoring of the military (e.g., “police patrol” monitoring), and finally, direct civilian intervention into military affairs at the strategic, operational, or tactical level.46 Again, the level of intrusion could be due to a whole host of structural issues, many of which might be proxied by the center’s fear of a coup, but I focus on the nature of the 44. See Huntington, Soldier and the State. 45. Feaver, Armed Servants, pp. 75–­77. 46. Ibid., p. 86.

46

Chapter 2

resulting civil-­military relationship since this feature is most theoretically relevant to nuclear command-­and-­control structures. I classify states with normal civil-­military relations or pure military regimes as having delegative structures. I measure the degree of political intrusion into military affairs to assess whether states have assertive structures that might predict nuclear command-­and-­control structures that are similarly tightly centralized. Finally, resource constraints is also relatively straightforward to measure since it depends strictly on observable resource endowments. This is a dynamic measure that depends on the material resource endowments of both the regional nuclear power and its primary adversary or adversaries. Not only does the production of the first nuclear weapon require significant financial and industrial resources, but some postures—­particularly the asymmetric escalation posture—­can be expensive to develop and manage, both in terms of capital and industrial outlays but also organizational and maintenance costs.47 What is critical for my theory, however, is whether a state can sustain the cost of developing that posture vis-­à-­vis the primary envisioned opponent(s). For example, in the China-­U.S. dyad, China has a relative advantage in manpower but a severe relative disadvantage in financial and advanced technical capacity (for now). Thus, if the choice of posture reached the “relative resource constraint” node, I would predict that China would be at a severe relative disadvantage in selecting an asymmetric escalation posture since the United States could outspend and out-­develop China’s capabilities in a spiraling arms race. Thus, if China were to follow the theoretical pathway down to this node, with delegative civil-­military structures, the prediction would be that resource constraints against its primary envisioned adversary would push it toward an assured retaliation posture. This variable is thus measured by examining the relative material resource endowments in a given dyad to determine if the regional power might be constrained in choosing a particular posture.

Alternative Explanations While there are country-­specific alternative explanations for why a particular state adopted a given nuclear posture, from a comparative perspective, there are no existing works that provide obvious alternative explanations to the general question of why regional powers adopt particular nuclear postures. Thus, I derive several alternative explanations from the broader proliferation literature that do provide alternative accounts for why states, in general, might select different nuclear postures over time. These are 47. See Michael Horowitz, The Diffusion of Military Power: Causes and Consequences for International Politics (Princeton: Princeton University Press, 2010).

Sources of Regional Power 47

necessarily broad alternative explanations to capture the comparative nature of the empirical record; they are not country-­specific explanations. If there are particularly strong alternative explanations for a specific case, I include it in my tests in addition to the broad alternative explanations. But for the purposes of testing my own generalizable theory against other generalizable theories, I use the following three alternative explanations: structural realism, technological determinism, and strategic culture.

Structural Realism Structural realism offers a macro-­ material alternative explanation arguing that regional power nuclear posture is a function of the material constraints of the international system, and a state’s immediate security environment in particular. As noted earlier, slim structural theories are difficult to deploy where one observes variation across states or time that are situated in relatively similar security environments. Hence my framework appropriates several additional variables in an effort to capture the variation in nuclear postures across regional powers. However, it could very well be that the addition of these variables adds little to no explanatory power to a strict and slim structural explanation. I therefore test the structural realist explanation as an alternative to my neoclassical realist framework. If a parsimonious structural realist explanation is sufficient to capture the variation and texture of regional power nuclear postures, then my richer framework would be unnecessary. The core structural realist hypothesis is that regional nuclear powers simply adjust their choice of posture to their security environment. This hypothesis has been argued, at times, by Kenneth Waltz and more recently by Avery Goldstein.48 That is, regional nuclear powers that find themselves to be in a more Waltzian defensive realist environment49 in which security is abundant and their security not persistently threatened may opt for an assured retaliation posture, whereas states in a more Mearsheimerian offensive realist environment in which security is scarcer may be forced to select asymmetric escalation postures. If this were the case, one would expect regional powers experiencing high levels of conflict pre-­nuclearization to trend toward the asymmetric escalation posture, and those facing less severe threats to adopt assured retaliation or catalytic postures. And states that adopt, for example, an assured retaliation posture but that continue to 48. See Waltz, “The Spread of Nuclear Weapons: More May Be Better”; Goldstein, Deterrence and Security in the 21st Century. 49. See Waltz, “The Spread of Nuclear Weapons: More May Be Better”; Mearsheimer, Tragedy of Great Power Politics.

48

Chapter 2

experience significant conventional conflict should shift toward an asymmetric escalation posture. In this way, there should be some “security-­ optimization” of regional power nuclear postures, where states shift pos­ tures as their security environments dictate and so that all experience roughly the same level of deterrent power against conventional conflict. In this view, a state’s posture would be conditioned by a “second-­image reversed” process where the particular security environment in which a state finds itself wholly determines its choice of posture, and states select postures that are optimal for only their security environment.50 The implication of the structural realist explanation is that posture itself is epiphenomenal with respect to its effect on the frequency and intensity of attacks a state experiences after it adopts a particular posture. In this view, structural pressures both compel states to acquire nuclear weapons and also determine the resultant posture based on the severity of those security pressures. In this view, posture itself has no independent effect. That is, if states optimize their choice of nuclear posture based on their security environment, they should all deter conflict equally well. As I will show later, this is not the case. States that adopt catalytic or assured retaliation postures have experienced significant and systematic deterrence failures, suggesting that regional states are not selecting their nuclear postures entirely based on security considerations. Nevertheless, it is critical to test this macro-­ material explanation for posture choice, since it cannot easily be dismissed out of hand. My framework generates the same prediction as the structural realist explanation in the event that a regional nuclear power lacks a third-­party patron and faces a conventionally superior proximate threat: asymmetric escalation. Certainly an asymmetric escalation posture produced by a conventionally superior proximate threat for a state that has no external options is entirely determined by security environment. However, if a state has a credible third-­party patron in that security environment, my theory departs from structural realism by predicting that the state should adopt a catalytic posture. Structural realists would have a tough time arguing that a state should rely on external balancing when an internal balancing option was available. The logic of self-­help, particularly for a state that is nuclear-­capable, should render a catalytic posture extremely unappealing to structural realists of all stripes. Indeed, structural realism should be allergic to the catalytic posture, where a nuclear state does not take full advantage of its nuclear weapons but instead outsources its defense to a third-­party patron. Because structural realists are generally skeptical of any state’s ability to know the intentions of other states with much certainty, 50. On second-­image reversed, see Peter A. Gourevitch, “The Second-­Image Reversed,” International Organization, vol. 32, no. 4 (Autumn 1978), pp. 881–­912.

Sources of Regional Power 49

structural realists are likely to view as unacceptably dangerous any strategy that relies too heavily on another state to guarantee one’s security, as the catalytic posture does. If structural realism is correct, we should never observe a catalytic posture. Furthermore, my framework argues that the international structure and a state’s security environment are generally indeterminate. In cases where a state does not face an acutely dangerous security environment, structural realism has difficulty generating a determinate prediction for whether that state should adopt an asymmetric escalation or assured retaliation posture. Thus, in order to discriminate between an assured retaliation and asymmetric escalation posture in states that do not face nuclear-­armed conventionally superior proximate threats, my theory argues that the consideration of several unit-­level variables is necessary. But, in order to test whether these variables are indeed necessary, I search for evidence that states at those nodes choose the postures they do for systematic external security considerations as the structural realist explanation would predict.

Technological Determinism Structural realism offers a macro-­ materialist explanation for a state’s choice of nuclear posture. By contrast, technological determinism offers a more micro-­material alternative explanation that focuses on particular nuclear capabilities at the unit level. Technological determinism, at its core, argues that capabilities drive a state’s posture and, in the extreme, that there is little to no political choice to make about the particular posture a state ultimately adopts.51 Rather than political considerations driving the development of technologies for a particular purpose, this view argues the reverse: that technology and capabilities shape political considerations. With respect to nuclear posture, this theory would argue that a regional power should adopt the posture that its technical capacity allows—­nothing more and certainly nothing less. The key determinants for nuclear posture would therefore be, for example, the amount and quality of available fissile material, weapons design capabilities, availability and type of delivery vehicles, command-­and-­control technology, and early warning capabilities. 51. For example, see Itty Abraham, The Making of the Indian Atomic Bomb: Science, Secrecy, and the Postcolonial State (New York: St. Martin’s Press, 1998); also more generally, Barry Buzan, An Introduction to Strategic Studies: Military Technology and International Relations (London: MacMillan Press, 1987); Merritt Roe Smith and Leo Marx, eds., Does Technology Drive History?: The Dilemma of Technological Determinism (Cambridge: MIT Press, 1994); Bernard Brodie, “Technological Change, Strategic Doctrine, and Political Outcomes,” in Klaus Knorr, ed., Historical Dimensions of National Security Problems (Lawrence: University Press of Kansas, 1976), pp. 263–­306.

50

Chapter 2

As a state’s technical capabilities evolve and its stockpile of available fissile material grows, there should be an attendant shift in nuclear posture. This explanation treats technology as a driver of posture, as if posture were simply a function of technological inertia or momentum. If the technological determinism view is correct, each regional nuclear power should follow a similar trajectory in the evolution of its posture: from catalytic, to assured retaliation, to asymmetric escalation as its capabilities evolve over time. That is, in this view, all states should follow the same postural evolution as their capabilities evolve, and all states should eventually adopt the most technologically demanding posture—­an asymmetric escalation posture. Further, the timing of any shift should be dictated by the development and incorporation of technologies that enable and characterize each evolving posture. Certainly, technological constraints do limit a state’s choice of posture in the early phases of nuclearization. States with limited fissile material and delivery capabilities simply do not have the immediate capability to adopt an asymmetric escalation posture. However, because organizational and management procedures are key to distinguishing, for example, catalytic from assured retaliation postures, cases in which states adopt a catalytic posture when they have the technical capability to adopt an assured retaliation or asymmetric escalation postures would be disconfirming to the technological determinism argument. Similarly, states that have the technological capability to adopt an asymmetric escalation posture but opt for assured retaliation postures would also challenge this theory. The key to distinguishing this alternative explanation from my framework is untangling the causal direction between technological development and postural choice: does a political choice for a posture drive technological developments required to implement that posture or vice versa? Hence, any evidence that choices about posture trail technological developments would support the technological determinism hypothesis. However, evidence that technological developments were driven by political imperatives in a top-­down, rather than bottom-­up, fashion would support my framework.

Strategic Culture In addition to the two materialist alternative explanations for nuclear posture, I also test a strategic culture alternative explanation, which may be the strongest challenge to my framework. Strategic culture has been used in various forms in security scholarship, most extensively by Iain Johnston, as well as in earlier works by Colin Gray who describes a “national style” in American and Soviet nuclear doctrines; by Jack Snyder, who writes about

Sources of Regional Power 51

Soviet strategic culture and nuclear doctrine; and Elizabeth Kier, who argues that conventional military doctrines in World War I/II France and Britain were produced by a specific organizational culture which varied as a function of domestic political configurations.52 I deploy Johnston’s testable definition of strategic culture, which is an “ideational milieu which limits behavioral choices” derived from deep and shared historical sources. It comprises “basic assumptions about the orderliness of the strategic environment, that is, about the role of war in human affairs (whether it is inevitable or an aberration), about the nature of the adversary and the threat it poses (zero-­sum or variable sum), and about the efficacy of the use of force (about the ability to control outcomes and eliminate threats, and the conditions under which applied force is useful)” which produce a “limited, ranked set of grand-­strategic preferences that is consistent across the objects of analysis . . . and persistent across time,” providing options for effectively dealing with a state’s threat environment.53 This analytic lens views culture as a shaping force that conditions the range of options that states—­or groups within states—­might consider in their security policy. The implication of this theoretical approach is that states in similar strategic environments, if they have decision makers and strategists imbued with different strategic cultures, should evince different preference orderings and thus different behaviors or nuclear postures. By the same token, states in different security environments that have groups with similar strategic culture (e.g., the warfighters, or alternatively the realpolitikers) should select the same postures. In addition to cross-­national influences of strategic culture, there can be intra-­state temporal effects if exogenous shocks shift the weight of preferences or groups within a given state, since multiple strategic cultures may be held within a single state.54 In this view, shifts in a state’s posture should therefore be due to an external shock that alters the ordering or weighting of strategic culturally generated preferences to produce a different posture. In addition, since strategic culture tends to be a deeply rooted, relatively slow-­moving variable—­though certainly mutable over time—­its effects could also manifest themselves cross-­nationally in distinct styles of nuclear posture that moderate the influence of material considerations. Evidence for this alternative explanation would exist if

52. Colin Gray, Nuclear Strategy and National Style (Lanham: Hamilton Press, 1986); Jack Snyder, Soviet Strategic Culture: Implications for Limited Nuclear Operations (Santa Monica: RAND Reports, September 1977); and Elizabeth Kier, “Culture and Military Doctrine: France Between the Wars,” International Security, vol. 19, no. 4 (Spring 1995), pp. 65–­93. 53. Alastair Iain Johnston, “Thinking about Strategic Culture,” International Security, vol. 19, no. 4 (Spring 1995), pp. 46–­48. 54. Ibid., pp. 53–­54.

52

Chapter 2

internal debates about nuclear posture among the military, strategists, and decision makers reflect a weighing of the three questions posed by Johnston earlier to produce a specific posture consistent with a country’s strategic or organizational cultures. It should be noted that my framework is not inconsistent with a strategic culture explanation, just as it is not wholly inconsistent with a structural realist explanation in some scenarios. I argue, however, that strategic culture as a variable should intervene in a very specific way: through a state’s civil-­military relations. It is not inconsistent with my argument that a country’s civil-­military relations might be conditioned by—­or be an expression of—­strategic culture; as noted earlier, a whole host of variables might produce a state’s particular civil-­military arrangements, including strategic culture. However, because of the various ways in which civil-­military relations might be conditioned, I argue there is no one cross-­national variable that systematically produces civil-­military relations; in some countries it could be strategic culture, in others it could be external security environment, in most cases it may be a product of historical peculiarities that have simply ossified over time. I leave open the possibility—­and fully expect—­ that strategic culture may be doing some work in some cases at this level in my framework, but that the relevant variable to measure with respect to choices about nuclear posture is civil-­military arrangements. In other words, in my framework, strategic culture operates through, and only through, civil-­military relations. The critical tests for discriminating between my theory and a strategic culture explanation will be if there is evidence of deep-­rooted strategic cultural beliefs in the writings of a state’s strategic elite over time independent of, and that may thwart, structural pressures, civil-­military arrangements, and resource constraint predictions. So I test whether there are systematic cross-­national or subnational effects of strategic culture that might influence a state’s choice of posture beyond their expression in some cases as civil-­military relations’ organizational culture. That is, I look for evidence that there are persistent and systematic belief structures about conflict, opponents, and the role of nuclear weapons that might condition a state’s choice of nuclear posture independent of the pressures I identify in my framework.

Empirical Strategy In this chapter, I have formulated a theory for the sources of regional power nuclear posture, posture optimization theory. Past efforts to systematically explain the nuclear posture choices of regional nuclear powers either have failed to treat the full universe of cases or have dealt with a small subset of

Sources of Regional Power 53 Regional Nuclear Power

Availability of Reliable Third-Party Patron? Yes

Catalytic Pakistan (1986–1991) Israel (1966–1990) South Africa (1979–1991)

No

Facing Conventionally-Superior Proximate Offensive Threat? Yes

No

Asymmetric Escalation France (1960–1990) Pakistan (1991–present)

Civil-Military Arrangements Assertive

Delegative

Assured Retaliation India (1974–present) China (1964–present)

Resource Constraints? Yes

Assured Retaliation

No

Asymmetric Escalation France (1991–present) Israel (1991–present)

Figure 2.2. Optimization theory’s predictions for regional power nuclear posture

issues associated with nuclear posture. As such, drawing on the insights of the neoclassical realist school, I developed a testable and falsifiable framework for regional power nuclear postures that privileges security environment but that also includes several unit-­level variables where security environment is indeterminate. States that lack a third-­party patron and are facing a conventionally superior proximate threat should have no choice but to adopt an asymmetric escalation posture; in this case, the motivations for nuclearization and posture choice are all structurally conditioned. However, in cases where such a severe and immediate security threat does not exist, several considerations—­the availability of a third-­party patron, civil-­military relations, and relative resource constraints—­should regulate the choice of posture. The following chapters test this framework against all six regional nuclear powers, comprising nine different choices of nuclear postures (Israel and Pakistan switch postures, and France explicitly chose to maintain an asymmetric escalation posture after the dissipation of the Soviet Union). I perform this empirical analysis in two steps. First, I apply the coding rules to establish the empirical variation in regional power nuclear postures. I describe, in substantial detail, the nuclear postures of each of the six regional powers and why I code them as either catalytic, assured retaliation,

54

Chapter 2

or asymmetric escalation at each point in time. For each regional nuclear power, I examine the state’s nuclear capabilities (e.g., survivable forces, tactical nuclear weapons?), its command-­and-­control arrangements (pre-­ delegated or assertive centralized control?), the level of transparency regarding capabilities and deployment patterns, and finally the envisioned employment mode of a state’s nuclear arsenal, in order to identify its nuclear posture. The mutual exclusivity of the postures means that there is a single coding for any state at a particular time and because the postures are sticky—­states can operationalize a posture in a crisis or conflict but it is difficult to change postures except over long time frames—­the confidence in these codings is quite high. Second, after I code which posture a state has adopted, I test that selection against my theoretical prediction as well as the three alternative explanations to see which theories explain the fact of the posture as well as the reasons for its selection. Figure 2.2 gives a preview of the predicted—­ not necessarily actual—­nuclear postures for each regional power based on optimization theory presented in this chapter. The four theoretical frameworks have different expectations about the patterns of postures that should emerge and when, allowing me to set up critical tests to discriminate between my framework and the three main alternative explanations. As will become evident, optimization theory performs best “in sample,” and also suggests that it may provide some predictive leverage to “out of sample” (i.e., future) regional nuclear powers.

Chapter Three

Pakistan

The next three chapters describe, and test the sources of, Asia’s regional power nuclear postures: Pakistan, India, and China. While India and China have persistently adopted assured retaliation postures, Pakistan switched from an explicitly catalytic posture between 1986 and 1998 to an aggressive asymmetric escalation posture that threatens the first use of nuclear weapons against Indian forces in order to deter its conventionally superior neighbor. Pakistan chose a catalytic nuclear posture in the late 1980s, when it employed its nascent nuclear weapons capability to compel the United States to assist it in crises with India. It was then forced to start shifting to an asymmetric escalation nuclear posture after the United States largely abandoned Pakistan following the Soviet withdrawal from Afghanistan, leaving Pakistan alone to face India’s conventionally superior forces. Pakistan finally achieved an explicit asymmetric escalation posture when it tested nuclear weapons in 1998. I show that these choices, and the timing of the shift from a catalytic to an asymmetric escalation posture, are best captured by optimization theory, illustrating how exogenous changes to a state’s security environment and alliance options can trigger a shift in nuclear postures. Many mistakenly view Pakistan’s nuclear weapons program as a natural response to India’s nuclear weapons capability, and thus date the genesis of Pakistan’s nuclear weapons program to 1974 after India’s so-­called peaceful nuclear explosion (PNE). However, by 1974, Zulfikar Ali Bhutto’s Pakistan was already pursuing nuclear weapons. Rather than 1974, it was Pakistan’s searing memory and humiliation from the 1971 war against India that marked its determined march toward nuclearization. Before 55

56

Chapter 3

1971, India and Pakistan had fought two conventional wars to relative stalemates in both 1948 and again in 1965. In both cases, Pakistan could claim that it had not yet lost a war to its conventionally superior neighbor. The 1971 war shattered any such pretensions. Pakistan was cut in two, birthing Bangladesh out of the amputated eastern flank of Pakistan. The war was a comprehensive defeat for Pakistan and resulted in the fall of its military dictator, General Yahya Khan in favor of Zulfikar Ali Bhutto, who had famously remarked as early as 1965 that “If India builds the bomb, we will eat grass or leaves—­even go hungry—­but we will get one of our own.” When he came to power in December 1971 in the aftermath of Pakistan’s humiliating defeat to India, Bhutto set out to make that pledge a reality to insure the territorial integrity of Pakistan’s remaining western wing. India’s test of a peaceful nuclear explosion in May 1974 only hardened Pakistan’s quest for nuclear weapons. The genesis of the Pakistani nuclear weapons program after the 1971 war provides deep insight into its ultimate aim: to avoid another massive conventional defeat at the hands of the Indians. This chapter will explore Pakistan’s evolution as a nuclear weapons state and describe the two distinct nuclear postures Pakistan has adopted. When Pakistan was believed to be able to weaponize a nuclear capability in the mid-­1980s but largely refrained from doing so, it initially adopted a catalytic posture that attempted to compel its then-­patron state, the United States, to intervene on its behalf in crises with India. While Pakistan’s ambiguous nuclear capabilities were developed with India in mind, the United States was the key target of Pakistani nuclear signaling. This signaling was designed to catalyze American intervention on Pakistan’s behalf to cap escalation in crises with India, lest Washington risk the crisis escalating to the nuclear level, something it could ill afford since Pakistan was a lynchpin for its covert war in Afghanistan against the Soviets. This posture was successfully employed to some degree in the 1986–­87 Brasstacks Crisis, but most notably in the 1990 Kashmir Compound Crisis. Following the loss of its American patron in the early 1990s, Pakistan began preparing for a shift to an asymmetric escalation posture, as predicted by optimization theory. Pakistan believed it had little choice but to follow suit when India tested nuclear weapons in 1998. However, in doing so, it was able to credibly establish the functionality and reliability of its deterrent after claiming to test five uranium fission devices of varying yields. Since these tests, and as a result of acquisitions and advances in its delivery capabilities and the development of explicitly tactical nuclear weapons, Pakistan has fully integrated nuclear weapons into its military doctrine and adopted an asymmetric escalation posture that attempts to credibly deter conventional attack by threatening the first use of nuclear weapons against a large-­scale Indian conventional thrust through

Pakistan 57

Pakistan’s vulnerable desert and plains corridor in Sindh and Punjab. As the conventionally weaker power, especially in a long conflict when India’s aggregate power advantage could be brought to bear against it, Pakistan’s army—­the organization responsible for stewarding and managing Pakistan’s nuclear weapons—­has made it clear that it cannot forswear the option of first use of nuclear weapons.1 While Islamabad’s official declaratory nuclear doctrine is “credible minimum deterrence,” Pakistan has, in fact, adopted an operational doctrine, or nuclear posture, resembling NATO’s first-­use posture against the Soviets in the European theater. So while rhetorically, the Pakistan Army clings to the characterization of its nuclear posture as “minimal,” in actuality Pakistan operates an asymmetric escalation posture that envisions the early use of nuclear weapons in a conventional conflict as a deterrent to its outbreak.

Pakistani Nuclear Posture Phase I: Catalytic States that have acquired nuclear weapons have done so for a variety of reasons, from security to domestic political motives to desire for international prestige.2 An active debate continues about why France or India acquired and tested nuclear weapons, even decades after their nuclearization. However, in the case of Pakistan, there is little debate about the motivations for nuclearization: nuclear weapons offered Pakistan the promise of security against a conventionally superior threat on its vulnerable eastern border with India. The summary judgment on Pakistan’s motivations is captured by Lowell Ditmer: “Pakistan’s motive for the acquisition of nuclear weapons is . . . far less complex and more conventional: national security. As India’s weaker rival, defeated in nearly all their military encounters and dismembered in the third, Pakistan has reason for concern.”3 Sumit Ganguly and Devin Hagerty concur, arguing that the “core aim of Pakistan’s nuclear weapons program is to prevent a repetition of 1971 (when Bangladesh was born out of East Pakistan) . . . to deter an Indian attack that might reduce Pakistan’s size even further, or perhaps even put the country out of 1. In November 2008, Pakistan President Asif Ali Zardari, speaking on an Indian television program, seemed to suggest that Pakistan would revise its first-­use doctrine. However, the army made it clear that Zardari had no authority to make such a revision to Pakistani doctrine. See Jane Perlez, “India’s Suspicion of Pakistan Clouds US Strategy,” New York Times, November 27, 2008, p. A1. 2. On motivations for nuclearization, see for example Sagan, “Why Do States Build Nuclear Weapons?”; Paul, Power versus Prudence; Solingen, Nuclear Logics; and Hymans, Psychology of Nuclear Proliferation. 3. Lowell Ditmer, “South Asia’s Security Dilemma,” Asian Survey, vol. 41, no. 6 (November/December 2001), p. 900.

58

Chapter 3

existence entirely.”4 Though Bhutto made his remarks about “eating grass” in 1965, it was only after the humiliation of the 1971 war that Pakistan earnestly pursued a nuclear weapons program. In January 1972, Zulfikar Ali Bhutto, who had replaced the military leader Yahya Khan, launched Pakistan’s nuclear weapons program, employing parallel tracks of plutonium reprocessing and uranium enrichment.5 The Pakistan Atomic Energy Commission (PAEC) spearheaded the plutonium route, under the direction of Munir Ahmad Khan from 1972 to 1991. Given tighter nuclear export controls following the Indian PNE, Pakistan’s plutonium route largely floundered. PAEC’s efforts to acquire a reprocessing facility from France and a suitably scaled reactor stalled in the mid-­1970s. Luckily for Pakistan, A. Q. Khan returned to Pakistan armed with a complete gas centrifuge design which he had stolen while working at the European consortium URENCO. With generous Saudi and Libyan financing, Bhutto supported Khan’s efforts to provide Pakistan with an indigenous uranium enrichment capability, which continued and was accelerated after Zia ul-­Haq deposed Bhutto in 1977. Khan spent the latter part of the 1970s laying the foundation for Pakistan’s uranium enrichment facility at Kahuta. It seems that the choice to initially weaponize via the uranium enrichment pathway was simply a choice of expediency—­the enrichment program was cheaper, more viable given the technology then available to Pakistan, and easier to obfuscate, thereby giving Pakistan plausible deniability that it was pursuing a nuclear weapons capability. But Pakistan never fully abandoned the plutonium route. It built a reprocessing facility in the 1980s, and Pakistan sought and received Chinese assistance to build several nuclear reactors; the first two, a 50MW reactor at Khushab and a 300MW reactor at Chashma, are already operational, though the latter is under safeguards.6 A plutonium pathway will be critical for Pakistan as it attempts to modernize and miniaturize its warheads, since plutonium which generally require greater design sophistication than warheads—­ uranium fission devices—­have better yield-­to-­weight ratios at higher targeted yields and are thus the desired fissile material for missile-­deliverable warheads.

4. Sumit Ganguly and Devin Hagerty, Fearful Symmetry: India-­Pakistan Crises in the Shadow of Nuclear Weapons (Seattle: University of Washington Press, 2005), p. 123. Parenthetical comment added. 5. Khan, Eating Grass, chapters 7 and 10; George Perkovich, “Could Anything Be Done to Stop Them? Lessons from Pakistan’s Proliferating Past” in Henry D. Sokolski, ed., Pa­ kistan’s Nuclear Future: Worries Beyond War (Carlisle: Strategic Studies Institute, 2008), pp. 59–­84; also see Samina Ahmed, “Pakistan’s Nuclear Weapons Program,” International Security, vol. 23, no. 4 (Spring 1999), pp. 178–­204. 6. Bokhari and Moss, “Pakistan Strikes Deal for Chinese Nuclear Reactors.”

Pakistan 59

Nevertheless, the initial material for Pakistan’s nuclear weapons capability was produced through the uranium enrichment facility located at Kahuta facility and under Khan’s direction. By 1981, U.S. intelligence was convinced that Kahuta was operational and that uranium was being enriched at the facility with the aim of “develop[ing] a nuclear explosives capability” and that it had made advances in the trigger package.7 By 1983, it was clear that Pakistan was receiving significant weapons-­design assistance from China, which had incentives to assist Pakistan as a means to balance against India. Indeed, it is widely believed that China directly passed Pakistan a blueprint of a uranium fission design it tested in 1966 at its Lop Nur test site, known as CHIC-­4, China’s fourth known nuclear weapons test.8 These designs would have significantly accelerated Pakistan’s march toward nuclearization, as Pakistani scientists could conduct cold tests—­testing the explosives package without the physics package/fissile material—­with the near certainty that a fully assembled device would work when enough uranium had been enriched to weapons-­grade level.9 In fact, according to Feroz Hassan Khan, cold tests and work on bomb designs, as well as thinking about how Pakistan could deliver these early designs, continued apace in this period: “In March 1984, exactly one year after the PAEC announced its first successful cold test, KRL conducted its first cold test in the Kirana Hills near Sargodha. By December . . . KRL was ready for further presidential orders to begin the hot tests. The product, however, was still a large bomb that could be delivered only by a C-­130 cargo aircraft with no assurance of delivery accuracy.”10 However, when confronted by the United States about operations at Kahuta, Zia continued to claim that Kahuta “can’t be a nuclear installation. Maybe it is a goat shed.”11 Given the critical role Pakistan played in supporting covert U.S. operations in Afghanistan, however, there was a limit to how far the United States was willing to pressure Pakistan to stop its nuclear weapons program. By late 1986, the “goat shed” at Kahuta had enriched enough uranium to put Pakistan, perhaps literally, “two screwdiver turns” away from a fully assembled nuclear weapon.12 By March 1987, Zia publicly declared that Pakistan was a nuclear-­capable, if not a fully weaponized nuclear state 7. Jeffrey T. Richelson, Spying on the Bomb: American Nuclear Intelligence from Nazi Germany to Iran and North Korea, (New York: Norton, 2007), p. 342. 8. Ibid.; also see Thomas C. Reed and Danny B. Stillman, The Nuclear Express: A Political History of the Bomb and Its Proliferation (Minneapolis: Zenith Press, 2009), pp. 249–­50. 9. See Khan, Eating Grass, p. 185. 10. Ibid., p. 189. 11. Zia quoted in Richelson, Spying on the Bomb, p. 342. 12. U.S. intelligence official quoted in Bob Woodward, “Pakistan Reported Near Atom Arms Production,” Washington Post, November 4, 1986, p. A1.

60

Chapter 3

when he said: “Pakistan has the capability of building the Bomb. You can write today that Pakistan can build a bomb whenever it wishes. Once you have acquired the technology, which Pakistan has, you can do whatever you like.”13 It is unclear just how far advanced the pit-­fabrication capability was at this stage and how long it would take Pakistan to fully assemble a nuclear weapon from a decision to do so, but neither the United States nor India doubted that Pakistan had become a nuclear-­capable state by this time. Indeed, by the end of 1987, the United States concluded that “Pakistan had produced enough fissionable weapons-­grade uranium for four to six atomic bombs.”14 U.S. Congressman Stephen Solarz quipped that Pakistan had “the nuclear equivalent of a Saturday night special. It may not be technically elegant, but it’s capable of doing the job.”15 Given that the bomb design was a proven Chinese configuration, Islamabad was able to achieve a credible capability in this phase without necessarily conducting a full-­blown test. Between this time and May 1998, Pakistan was a de facto nuclear state facing a conventionally superior state, India, which had a recessed nuclear capability. Why did Pakistan not immediately become an overt nuclear power and adopt an asymmetric escalation posture to deter India? Most importantly, during the 1980s, Pakistan had an extremely generous American patron that viewed Pakistan as a crucial conduit for its covert war against the Soviets in Afghanistan, and was concerned to avoid Pakistan becoming an overt nuclear state. The availability of this third-­party patron enabled Pakistan to select a catalytic nuclear posture that could be brandished to compel American intercession on Pakistan’s behalf in the event of a crisis with India, since Washington could not allow Islamabad to be distracted by Delhi when it needed it to be fully focused on Afghanistan. In particular, due to American nonproliferation laws, any evidence that Pakistan had crossed certain thresholds in its nuclear program would immediately trigger a cutoff to Pakistan and thereby threaten U.S. strategy in Afghanistan. This provided the Pakistani state with leverage against the United States, and enabled the selection of a catalytic strategy. As I show, a catalytic strategy was an optimal choice since Pakistan could maintain an ambiguous nuclear capability which would avoid provoking India and avoid potentially crippling international sanctions, yet was still sufficient to catalyze American intervention in the event of a crisis. The enabling fea­

13. Pakistan President Zia ul-­Haq quoted in William Doerner and Ross Munro, “Pakistan Knocking at the Nuclear Door,” Time, March 30, 1987. 14. Hedrick Smith, “A Bomb Ticks in Pakistan,” New York Times Magazine, March 6, 1988. 15. Solarz quoted in ibid.

Pakistan 61

ture of this posture was America’s client-­state relationship with Pakistan through the end of the Cold War. As noted earlier, a catalytic nuclear posture is designed to draw in third-­ party—­often superpower, and more often U.S.—­assistance to defend the state under the threat of unleashing nuclear weapons. This nuclear posture calls for the use of nuclear weapons—­first-­use if necessary—­only as a last resort in the event of a mortal conventional or nuclear threat to the state’s existence. It relies on only a few weapons (not at the level of second-­strike forces), perhaps not even fully assembled, adopts a largely demonstrative (or possibly a limited countervalue) targeting strategy since a counterforce option is precluded by the small size of the arsenal, does not pre-­delegate authority to the armed forces, and keeps the weapons in a “recessed” or disassembled posture during peacetime. It also, critically, depends on there being at least one third party who could potentially be compelled to intervene on the state’s behalf in the event it threatened to use nuclear weapons to preserve its existence. Its primary purpose is to draw in international intervention before the point of state death is reached. This posture does not attempt to deter conventional conflict in toto, but it does attempt to deter the conventional extinction of the state and is designed primarily to signal to a more powerful patron the state’s willingness to preserve its survival, through nuclear weapons if necessary in the event of a fatal conventional confrontation. The key distinguishing feature of the posture is that the threat to unsheathe nuclear weapons is sent primarily to a third party, not only to the adversary in the crisis. Scholars, notably P. R. Chari, Zafar Cheema, and Stephen Cohen, have written about the critical role the United States played in resolving the periodic India-Pakistan crises during this phase. However, I argue that U.S. intervention was deliberately catalyzed by a Pakistani nuclear posture that sent specific signals that it knew would be detected by Washington—­and not necessarily in Delhi—­knowing full well that Washington had every incentive to prevent escalation due to Pakistan’s centrality in the war in Afghanistan through the end of the Cold War. Pakistan had the necessary capability—­an ambiguous but credibly functional nuclear capability—­and the availability of a third-­party patron through at least 1990 that enabled it to successfully implement a catalytic posture. This posture allowed Pakistan to more aggressively execute a long-­standing revisionist preference to support insurgent groups—­for example, in Punjab and Kashmir—­with the probabilistic calculation that the United States would intercede to prevent significant Indian retaliation that might result in escalation to the nuclear level. Unfortunately, without access to Pakistani leaders at the time, it is impossible to conclusively determine whether Pakistan’s official nuclear posture accorded with a catalytic posture. The best one can do is to look at

62

Chapter 3

crisis signaling to determine where Pakistan’s primary nuclear signaling was directed to infer Pakistan’s nuclear posture. Evidence for Pakistan employing a catalytic posture during this period can be found in the two India-­Pakistan crises that erupted: the 1986–­87 Brasstacks crisis and, more definitively, the 1990 Kashmir Compound Crisis. In chapter 10, I focus on deterrent successes and failures in these crises. Here, I simply use them to illustrate Pakistan’s posture, since a catalytic posture is most visible and distinctive in the context of a crisis. In both cases, Pakistan’s signals to the United States—­most notably in the Compound Crisis—­and not directly to India, resulted in U.S. diplomatic interventions to defuse the crises.

Brasstacks The Brasstacks crisis took place while Pakistan was on the cusp of achieving a nuclear weapons capability. Indeed, by November 1986, U.S. intelligence had judged that Pakistan and India were both effectively nuclear-­capable—­ but not legally nuclear weapons—­states.16 The Brasstacks crisis was not a direct nuclear crisis between India and Pakistan, but rather a crisis in which the United States had a keen interest in preventing any escalation that might cause Pakistan to cross certain thresholds in its nuclear weapons program (e.g., fully assembling warheads). Had those thresholds been crossed, the United States would have been forced to sanction Pakistan. The so-­called Pressler Amendment banned economic and military assistance to Pakistan unless the president certified annually that Pakistan did not possess nuclear weapons (a determination that in fact gave Pakistan and the Executive substantial interpretive latitude). Such a cutoff of assistance would have jeopardized Pakistan’s critical role in the Afghanistan war. Thus, faced with Pakistani nuclear saber-­rattling, the United States mobilized to ensure that Brasstacks was defused before Pakistan came close to crossing key nuclear weapons lines.17 The Brasstacks crisis started with Indian military exercises that quickly spiraled into a militarized crisis.18 In the mid-­1980s, India was battling a

16. Woodward, “Pakistan Reported Near Atom Arms Production,” p. A1. 17. See P. R. Chari, Pervaiz Iqbal Cheema, and Stephen Cohen, Four Crises and a Peace Process: American Engagement in South Asia (Washington, DC: Brookings Institution Press, 2007), pp. 74–­75. 18. For descriptions and treatments of the Brasstacks crisis, see Kanti Bajpai, P. R. Chari, Pervaiz Iqbal Cheema, Stephen Cohen, and Sumit Ganguly, Brasstacks and Beyond: Perception and Management of Crisis in South Asia (Delhi: Manohar, 1995). Also see Ganguly and Hagerty, Fearful Symmetry, chapter 4; Kapur, Dangerous Deterrent, chapter 4; Chari, Cheema, and Cohen, Four Crises and a Peace Process, chapter 3; T. V. Paul, ed., The

Pakistan 63

Sikh insurgency in the key border state of Punjab. Pakistan, seizing on the Sikhs’ battle with the Indian state, provided substantial material support to the Khalistani militants in the hopes of tying down and weakening India’s army in a protracted domestic counterinsurgency campaign. Punjab is not only India’s wealthiest state, but it is also one of the vulnerable thrust points for any Pakistani incursion into Indian territory. Thus, piqued by Pakistan’s meddling, India’s political leadership led by Prime Minister Rajiv Gandhi and Chief of the Army Staff Gen. K. Sundarji, concluded that “a military exercise aimed at Pakistan’s own weak point—­the province of Sindh—­ would be a fitting riposte to Pakistan and a threat (with echoes of 1971) that there might be more to come.”19 Brasstacks was a large-­scale live-­fire exercise that amassed almost 250,000 Indian troops—­including Reorganized Army Plains Infantry Divisions (RAPIDs) designed to be more mobile than traditional Indian infantry divisions20—­and simulated combined arms operations in the most likely theater for an India-­Pakistan conventional war. The proximity of the exercise to Punjab and Sindh concerned Pakistan because a deep Indian thrust at that point could sever the links between North and South Pakistan. With the memory of the 1971 war still fresh, this was deeply troubling to the Pakistani military.21 The sheer and unprecedented scale of the exercises, combined with reports that the Indians were using live ammunition, raised concerns that India might be preparing for a military strike on Pakistan. In response, Pakistan moved its two large corps to the border, positioned for a potential pincer move against northern and southern Punjab in January 1987. The crisis peaked on January 23, when India positioned defensive deployments in Punjab to block a possible Pakistani offensive and began making noises about preemptive military strikes, particularly on Pakistan’s Kahuta facil­ ity.22 Pakistan also feared that India was preparing for an offensive in its now-­exposed Sindh region. In a January 28 interview with Kuldip Nayar, a journalist for the London Times and the London Observer, A. Q. Khan, who headed Kahuta, threatened, “Nobody can undo Pakistan or take us for granted . . . let it be clear that we shall use the bomb if our existence is

India-­Pakistan Conflict: An Enduring Rivalry (Cambridge: Cambridge University Press, 2005), especially chapter 7. 19. See Bajpai et al., Brasstacks and Beyond, p. 23. 20. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 44. 21. The Indian military evidently did not believe the exercise would cause alarm for three reasons: (1) Exercise Digvijay had been held several years prior without issue; (2) the exercises were held on a North-­South rather than an East-­West axis; and (3) they were being conducted east of the Indira Gandhi canal and it would require a large logistical effort to cross the canal to threaten Pakistan. See ibid., pp. 48–­49. 22. See Perkovich, India’s Nuclear Bomb, p. 13.

64

Chapter 3

threatened.”23 Shuja Nawaz claims that “Pakistan orchestrated [this] leak” through Khan.24 At this point, under American pressure and through diplomacy at cricket matches, both states took rapid action to de-­escalate the crisis. By February 19, both armies had withdrawn from the border and India subsequently continued with Phase IV of Brasstacks which was scheduled to commence a month later. There is little evidence that Brasstacks was perceived to be a serious nuclear crisis between India and Pakistan, though there are several pieces of evidence that suggest awareness of a nuclear backdrop.25 But more than that, it was American concern—­perhaps prompted by Khan’s repeated threats over the years—­that, as noted by Chari, Cheema, and Cohen, “Pa­ kistan would move across several red-­lines that had been mutually agreed upon” that aroused American concern and mobilized U.S. action to ensure that Pakistan could continue to serve as a conduit for U.S. materiel to Afghanistan.26 Chari, Cheema, and Cohen argue that American officials acted because they were acutely worried “about changes in Pakistan’s nuclear status that would lead to termination of American military sales and other forms of aid, directly endangering the war effort in Afghanistan.”27 Thus, if Pakistan felt sufficiently threatened by Indian moves, particularly a possible preventive strike on Kahuta, to cross certain nuclear thresholds, the United States had an overriding incentive to dampen the crisis. Indeed, after Pakistan began mobilizing its forces—­even though the prospect of an armed conflict was still low—­the United States, led by Ambassador John Gunther Dean in New Delhi, made a concerted effort to confirm and accelerate the defusing of the crisis. Chari, Cheema, and Cohen conclude that Pakistan’s nuclear threats “fit into a larger Pakistani strategy: that of linking its own nuclear program (at the highest policy level) with an American commitment to defend Islamabad from an Indian attack.”28 Though the American role was limited, Pakistan was able to exploit American fears of the advancing Pakistani nuclear program to ensure an orderly end to the 23. Quoted in Bajpai et al., Brasstacks and Beyond, p. 39. The interview was with journalist Kuldip Nayar of the London Times and, though it was not published for several weeks, Nayar reportedly passed the threat on to relevant Indian and American officials in Delhi, according to former PMO officials in the Rajiv Gandhi government. Khan made a similar threat in 1984. see Gordon Corera, Shopping for Bombs: Nuclear Proliferation, Global Insecurity, and the Rise and Fall of the AQ Khan Network (Oxford: Oxford University Press, 2006), p. 47. 24. Shuja Nawaz, Crossed Swords: Pakistan, Its Army, and the Wars Within (Karachi: Oxford University Press, 2008), p. 392. 25. See John H. Gill, “Brasstacks: Prudently Pessimistic,” in Ganguly and Kapur, eds., Nuclear Proliferation, chapter 3. 26. Chari, Cheema, and Cohen, Four Crises and a Peace Process, pp. 74–­75. 27. Ibid., p. 75. 28. Ibid., p. 27.

Pakistan 65

crisis. Though the Brasstacks crisis provides an example of the catalytic posture in action, Pakistan’s nuclear posture is most strongly confirmed by the 1990 Compound Crisis over Kashmir.

1990 Kashmir Compound Crisis The most explicit example of Pakistan employing a catalytic posture is the 1990 Compound Crisis over Kashmir, where Pakistan deliberately signaled to the United States—­but not directly to India—­that it was prepared to brandish its nuclear capabilities in an attempt to mobilize American intervention to restrain Indian escalation. By 1990, India was believed to have readied “at least two dozen nuclear weapons for quick assembly and potential dispersal to airbases for delivery by aircraft for retaliatory attacks against Pakistan,” and Pakistan was similarly believed to be nuclear-­ weapons capable.29 However, neither side had established a sufficient ballistic missile capability to deliver nuclear devices, so both India and Pakistan would have had to rely on aircraft—­likely cargo aircraft—­to deliver any weapons.30 In this context, India and Pakistan again nearly came to blows over the insurgency in Kashmir in the winter and spring of 1990. Just as it did with the Khalistani movement in Punjab, Pakistan aggressively supported and supplied Kashmiri militant groups operating in Indian territory, consistent with its long-­standing revisionist preferences on the issue.31 In response, the Indian state deployed infantry units in Punjab and Kashmir to protect transportation and communications lines. Pakistan, in turn, deployed its II Corps across Punjab’s southern border with Rajasthan and deployed elements of the I Corps across Punjab’s northern border with Kashmir.32 These Pakistani deployments came on the heels of Pakistan’s largest ever military exercise (Zarb-­e-­Momin) which tested a new “offensive-­defense” doctrine that planned “to take the war into India, launching a sizeable offensive on Indian territory” according to Pakistani Chief of Staff Gen. Aslam Beg.33 Pakistan believed that India was counter-­escalating when the Indian Army 29. Perkovich, India’s Nuclear Bomb, p. 293, sourced by K. Subrahmanyam. 30. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 101; also see Chari, Cheema, and Cohen, Perception, Politics and Security in South Asia: The Compound Crisis of 1990 (London: Routledge Curzon, 2003), chapter 6. 31. Ambassador Robert Oakley in Ganguly and Hagerty, Fearful Symmetry, p. 88; also see Devin Hagerty, “Nuclear Deterrence in South Asia: The 1990 Indo-­Pakistani Crisis,” International Security, vol. 20, no. 3 (Winter 1995/1996), pp. 79–­114. 32. According to Indian Army Chief of Staff V. N. Sharma in Ganguly and Hagerty, Fearful Symmetry, p. 90. 33. General Mirza Aslam Beg quoted in ibid., p. 91.

66

Chapter 3

conducted armored unit exercises in the Rajasthani desert precisely where the Brasstacks exercises had taken place, though neither India nor the U.S. claim that this was anything more than a routine exercise. After this series of military deployments, the crisis escalated, with both Pakistani and Indian leaders hurling threats at each other about the consequences should the adversary breach the other’s territorial integrity. By the end of March, with 200,000 Indian troops assembled in Kashmir facing off against 100,000 Pakistani Army troops, and corps-­size forces arrayed against each other in both Punjab and Rajasthan, the 1990 crisis seemed to be quickly hurtling toward armed conflict. Though both the Indian and Pakistani armies were deliberately careful not to move their Strike Corps elements too close to the border, the political leadership in both countries continued to escalate the war of words.34 In April, the United States awoke to the potential consequences of the conflict escalating and in May dispatched then-­Deputy NSA Robert Gates to the region to de-­escalate the crisis. Gates notes that, “the analogy I used at the time was that the environment reminded me of something out of August 1914.”35 He leaned on both Indians and Pakistanis to withdraw troops from the border and to de-­escalate the crisis. In India, he warned that any potential conflict “might go nuclear,” though the Indians discounted that warning as American hysteria.36 Nevertheless, the Gates mission succeeded in, if nothing else, opening lines of direct communication between India and Pakistan, ultimately resulting in both states crawling away from potential conflict. The key issue for determining Pakistan’s nuclear posture is to establish whether Pakistan deliberately made any nuclear movements that prompted U.S. intervention as a mechanism to pull India away from Pakistan’s borders. Any such behavior would be consistent with the catalytic posture. There were certainly alarmist assessments in the United States, notably catalogued by Seymour Hersh, who quotes Deputy CIA Director Richard Kerr as saying the crisis was “as close as we’ve come to a nuclear exchange. It was far more frightening than the Cuban crisis.”37 What generated such fears in Washington, but not necessarily in Delhi? The evidence suggests that the Pakistanis deliberately signaled to the United States—­but not necessarily to India—­that it was prepared to escalate the crisis to the nuclear level. Hersh reports that: 34. See Ganguly and Hagerty, Fearful Symmetry, pp. 94–­96; Chari, Cheema, and Cohen, Four Crises and a Peace Process, pp. 94–­95. 35. Robert Gates quoted in Ganguly and Hagerty, Fearful Symmetry, p. 97. 36. Ibid., p. 98. 37. See Seymour Hersh, “On the Nuclear Edge,” New Yorker, March 29, 1993.

Pakistan 67 Sometime in the early spring of 1990, intelligence that was described as a hundred percent reliable—­perhaps an NSA intercept—­reached Washington with the ominous news that General Beg had authorized the technicians at Kahuta to put together nuclear weapons. Such intelligence, of “smoking gun” significance, was too precise to be ignored or shunted aside.38

In addition, American imagery intelligence reportedly spotted the evacuation of Kahuta in May, perhaps in anticipation of a retaliatory strike if Pakistan used nuclear weapons first, and observed activity between a supposed nuclear storage facility and Sargodha Air Base outside Lahore where Pakistan’s nuclear-­rigged F-­16s were believed to be located.39 Though there were questions about whether the F-­16s could deliver a Pakistani nuclear weapon, there were legitimate concerns that a Pakistani C-­130 could still nevertheless drop a nuclear weapon out of its back door.40 Ambassador Oakley revealed that though the United States never had any “hard indications that any nuclear warheads had been delivered to an airbase . . . you could guess that.”41 Regardless of whether these accounts are entirely accurate in hindsight, at the time there seemed to be enough circumstantial evidence of possible nuclear movement in Pakistan to catalyze American intervention in the crisis to prevent the Indian Strike Corps from attacking across the border. Certainly, some of Hersh’s details have been questioned, and others have been flat-­out denied, by scholars and participants in the crisis. But there is nevertheless substantial evidence independent of the Hersh account that Pakistan either did—­or pretended to—­move and threaten to use nuclear assets, thereby compelling the United States to intervene in the crisis.42 That is, whether they “fak[ed] nuclear delivery preparations  .  .  . [to] spur the United States into action” or actually put nuclear assets on alert, there was credible signals intelligence that Gen. Beg had indeed ordered some 38. Ibid. 39. Ibid. 40. Robert Burrows and Robert Windrem, Critical Mass: The Dangerous Race for Superweapons in a Fragmenting World (New York: Simon & Schuster, 1994), p. 82. Shuja Nawaz claims that in “July 1990 even a ‘nuclear aerial device’ was cold tested dropping it from a modified F-­16 aircraft.” See Nawaz, Crossed Swords, p. 491. 41. Oakley quoted in interview with Hersh. Indeed, the F-­16 account has been vehemently disputed by U.S. officials; nevertheless, Oakley has conceded that there were enough indications of unusual nuclear-­related activity in Pakistan to trigger U.S. concern. Emphasis added. 42. See Michael Krepon and Mishi Faruqee, eds., Conflict Prevention and Confidence Building Measures in South Asia: The 1990 Crisis, Henry L. Stimson Center Occasional Paper 17, April 1994, pp. 45–­46; also see Chari, Cheema, and Cohen, Four Crises and a Peace Process, pp. 103–­8.

68

Chapter 3

movement with respect to nuclear weapons.43 William Burrows and Robert Windrem argue that Pakistan might have been engaged in a “colossal bluff of the sort Israel had concocted in 1973” in order to trigger U.S. intervention.44 Indeed, Oakley and Colonel Don Jones concede that U.S. intelligence observed large crates being moved to airfields on trucks, with Oakley only somewhat jokingly saying “on top of each crate it said, ‘Pakistan nuclear devices’.”45 The fact that the “data collected by US intelligence systems, far from being ambiguous, were almost unbelievably explicit” according to Burrows and Windrem, lends weight to the hypothesis that Pakistan took deliberate and clear action whose patterns would unambiguously suggest nuclear movement to U.S. intelligence, thereby prompting American intervention in the crisis to protect Pakistan.46 In the region, a senior American government official present at the time noted that there was a very serious general concern about war between India and Pakistan, due to evidence that trains with conventional capabilities were being deployed close to the border and indications that Pakistan had moved F-­16s “close to the border” which carried with it the threat that these assets were to be used for potential nuclear missions.47 Whether it was actually a nuclear alert is not as relevant as the fact that Pakistan seemed to intentionally use its nuclear capability as a signal to the United States to come to its aid. Was this an intentional signaling gambit by Pakistan? New evidence from the then-­Chief of Army Staff, Gen. Beg, suggests that it was absolutely deliberate and that catalyzing American intervention on Pakistan’s behalf was its intended strategy. General Beg told Brig. Gen. (Ret.) Feroz Hassan Khan that, at the behest of Prime Minister Bhutto, “A squadron of F-­16s was moved to Mauripur [an Air Force Base in Karachi] and we pulled out our [nuclear] devices and all to arm the aircraft, [which carried out] movement from Kahuta, movement from other places, which were picked up by the American satellites . . . all movement was made in a way that is visible . . . It was therefore necessary to convey deterrence signaling by letting the Americans pick up Pakistani preparations and convey it to both India (and Israel) about the consequences.”48 Decades after the crisis, we now have first-­hand definitive confirmation from the then-­Chief of Army

43. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 106. 44. Burrows and Windrem, Critical Mass, p. 85. I describe Israel’s 1973 execution of the catalytic posture in chapter 7. 45. Oakley quoted in Krepon and Faruqee, Conflict Prevention, p. 18. 46. Burrows and Windrem, Critical Mass, p. 85. 47. Author interview with senior U.S. government official in the region at the time, November 25, 2011. 48. General Mirza Aslam Beg quoted in Feroz Hassan Khan, Eating Grass, pp. 230–­31. Emphasis added.

Pakistan 69

Staff that Pakistan indeed engaged in a deliberate catalytic posture whose primary intended signal was Washington. These signals were not picked up by the Indians, who, according to George Perkovich, were not at this point “worrying explicitly about a nuclear threat from Pakistan. The Indians did not know of the activity detected by American intelligence and Gates did not tell them about it . . . Pakistan’s nuclear capability [was] not as an acute threat at this moment” to Delhi.49 Pakistan shrewdly employed nuclear signaling to the United States, compelling Washington to intervene and prevent India from initiating a conventional assault on Pakistan.50 When Ambassador Oakley was asked in an interview whether the Pakistani nuclear program in this period between 1986 and 1990 was primarily directed at the United States, he recalled: “I think that is right. The Pakistanis manipulated very skillfully their nuclear progress [and] used it to obtain concessions from the United States.”51 The 1990 Compound Crisis is the strongest example of Pakistani behavior aimed at catalyzing U.S. intervention on its behalf. Islamabad exploited superpower incentives to compel the United States to intervene on its behalf when Pakistan’s support for insurgencies in India triggered periodic crises. Most scholars treat Pakistan as a closet nuclear state that relied on existential deterrence during this phase.52 However, Pakistan deliberately employed its nuclear weapons in a fashion distinct from simple existential deterrence, which attempts to directly deter aggression from an adversary. Instead, Pakistan employed a catalytic nuclear posture, whose primary signal was to the United States, not India, to catalyze Washington’s intervention on Pakistan’s behalf in a crisis with its larger neighbor. This drawing in of U.S. support was at the heart of Pakistan’s nuclear posture during this period.

Explaining Pakistan’s Nuclear Posture Phase I: Catalytic The selection of a catalytic nuclear posture for such a long period of Pakistan’s nuclear history is captured only by optimization theory. The critical 49. Perkovich, India’s Nuclear Bomb, p. 310; also see K. Subrahmanyam, “Indian Nuclear Policy—­1964–­98 (A Personal Recollection,” in Singh, ed., Nuclear India, p. 45. 50. This is largely supported by the argument made in Karthika Sasikumar, “Crisis and Opportunity: The 1990 Nuclear Crisis in South Asia” in Ganguly and Kapur, eds., Nuclear Proliferation in South Asia, pp. 76–­99, esp. p. 89. 51. Ambassador Robert Oakley phone interview with author, Washington, DC, November 8, 2011. 52. See, for example, Ganguly and S. Paul Kapur; Bajpai et al., Brasstacks and Beyond; Ganguly and Hagerty, Fearful Symmetry; Kapur, Dangerous Deterrent; Chari, Cheema, and Cohen, Four Crises and a Peace Process; Paul, The India-­Pakistan Conflict.

70

Chapter 3

node is whether Pakistan had an incentivized third-­party patron upon whom it could rely to intervene on its behalf by using the risk of nuclear escalation to compel its action. In the late 1980s, Pakistan’s security environment was certainly severe but, as noted, if it could avail itself of American protection against a conventionally superior India, backstopped by an ambiguous nuclear capability, it would be advantageous to do so because Pakistan could not afford a cutoff of American aid, nor the international sanctions that would result if it tested nuclear weapons. Furthermore, while India was certainly a conventionally superior threat, with a quantitative advantage ranging from two to three times the size of Pakistan’s forces, Pakistan’s defensive scheme in the 1980s emphasized the “offensive defense” that (a) was quicker to mobilize than India’s because Pakistan’s forces were closer to the border and operated on interior lines, and (b) could threaten India’s “chicken neck” at the vulnerable Punjab-­Jammu border. Therefore, while a nuclear weapons capability was considered nec­ essary from a security perspective to offset India’s conventional superiority, it could have been operationalized in several ways. One way would be to move as quickly as possible toward an asymmetric escalation posture that envisioned using a limited nuclear capability in tactical theaters against onrushing Indian forces to deter an Indian onslaught. An asymmetric escalation posture would be predicted by both a strictly realist hypothesis and an explanation based on strategic culture. But this is not what Pakistan chose. Instead, with the availability of a U.S. patron motivated by high incentives to keep Pakistan focused on Afghanistan in the late 1980s, Pakistan had the option of a less costly and provocative catalytic option in which it could manipulate the risk of nuclear escalation toward Washington in order to augment its conventional deterrence against India by compelling the United States to keep Pakistan’s conventional military strong and also to blackmail its intervention in crises with India to restrain Delhi. The relationship between the United States and Pakistan in this period was precisely of the character that would enable a catalytic posture: dependency but not alliance. Bruce Riedel writes, “With the Soviet invasion [of Afghanistan], Washington renewed its Cold War love affair with the Pakistani Army and the ISI,” and at the height of the war, America was funneling upwards of $400 million annually to the Pakistani security services.53 As long as the Afghan conflict was simmering, the United States had to rely extensively on the Inter Service Intelligence (ISI) to funnel guns to the jihadists in Afghanistan. Thus, any threats to that supply and logistics chain—­such as a crisis with India—­would have been detrimental to American interests.

53. Bruce Riedel, Deadly Embrace: Pakistan, America, and the Future of the Global Jihad (Washington, DC: Brookings Press, 2011), pp. 27–­28.

Pakistan 71

There is evidence the Pakistanis believed and were aware of this and sought to exploit this leverage, as described in the book The Bear Trap by retired ISI officer Mohammad Yousaf, who was chief of operations for Pa­­ kistan’s covert Afghan program.54 The United States was critically dependent on Pakistan for its arms supply chain to the Afghan jihadis in the 1980s and, according to Steve Coll, Zia “milked Washington for all he could  .  .  . he loaded up his shopping cart, [but] Zia kept his cool and his distance.”55 Zia recognized that he could exploit Washington’s largesse and protection given its interest in defeating the USSR in Afghanistan. The United States could ill afford for Pakistan to be distracted by an Indian threat on its western border, with Secretary of State George Schultz writing: “We must remember that without Zia’s support, the Afghan resistance, key to making the Soviets pay a heavy price for their Afghan adventure, is effectively dead.”56 Shuja Nawaz’s authoritative book on the Pakistan Army claims that the U.S. effort in Afghanistan meant that “Pakistan was now suddenly in the catbird seat, calling the shots on other issues too, such as the return to democracy . . . or Pakistan’s nuclear ambitions.”57 The leverage Pakistan believed it derived with respect to its nuclear program was that any evidence that it had assembled nuclear weapons would trigger congressional legislation that required the United States to terminate most military aid and assistance to Pakistan, since it was outside the Non-­Proliferation Treaty. This threat of cutoff was further augmented by the Pressler Amendment, which required annual presidential certification that Pakistan did not possess nuclear weapons devices in order to continue particularly military aid. Any such cutoff would threaten the American strategy in Afghanistan. American Secretary of State George Schultz wrote in an insightful 1982 memo that: “The Pakistanis probably believe that because of the strategic considerations in the region, the USG will seek to protect the U.S.-­Pakistan security relationship against Congressional moves prompted by Pakistan’s nuclear weapons activities.”58 Thus, any risk that Pakistan might move toward assembling nuclear weapons while the United States was waging war in Afghanistan posed significant problems for U.S. policy and gave Pakistan—­it believed—­the leverage to catalyze 54. See Brigadier Mohammad Yousaf and Major Mark Adkin, The Bear Trap: Afghanistan’s Untold Story (London: Leo Cooper, 1992), especially pp. 81–­96. 55. Steve Coll, Ghost Wars: The Secret History of the CIA, Afghanistan, and bin Laden from the Soviet Invasion to September 10, 2001 (New York: Penguin, 2004), p. 62. 56. Quoted in ibid., p. 62. 57. Nawaz, Crossed Swords, p. 376. 58. George P. Schultz, “How Do We Make Use of the Zia Visit to Protect Our Strategic Interests in the Face of Pakistan’s Nuclear Weapons Activities,” Memorandum to the President, November 26, 1982. Available at http://digitalarchive.wilsoncenter.org /document/114311.

72

Chapter 3

American assistance to diffuse whatever crisis with India might compel nuclear weapons assembly. One episode in 1985 convinced the Pakistanis of the lengths to which the U.S. government, particularly the CIA, would go to convince Congress’s nonproliferation stalwarts that Pakistan was not a nuclear weapons state, thus allowing the aid to the mujahideen to continue to flow. As recorded in George Crile’s Charlie Wilson’s War: “The dirty little secret of the Afghan war was that Zia had extracted a concession early on from Reagan: Pakistan would work with the CIA against the Soviets in Afghanistan, and in return the United States would not only provide massive aid but would agree to look the other way on the question of the bomb.”59 In 1985, a Pakistani agent was caught in the United States attempting to buy kryton switches for nuclear devices, which led Congressman Stephen Solarz to launch hearings in which an aid cutoff seemed imminent. Crile writes that the CIA was “alarmed at what might happen to the Afghan program in the event of a cutoff.”60 The White House and Congressman Charlie Wilson ran interference on Solarz in Congress, which seemed to convince Zia that the United States would go to extraordinary lengths to keep Pakistan from crossing certain thresholds in its nuclear program that might cause Congress to automatically cut off aid and threaten the American effort in Afghanistan.61 In another case, Dutch legal action against A. Q. Khan for allegedly stealing the URENCO designs was directly thwarted by the CIA. Then-­Dutch Prime Minister, Ruud Lubbers, recounted in a 2005 radio interview that in 1985 the U.S. government had directly pressured the Netherlands to not prosecute Khan for his illegal nuclear activities: “Considerable pressure was applied [by the CIA] and we eventually agreed . . . [the pressure] was being used to ensure that nothing untoward in the Pakistani nuclear program jeopardized the aid being poured into Pakistan and hence Afghanistan. You could say we were duped by the [U.S.] administration that allowed Khan to get off.”62 These episodes suggested to Zia the leverage Pakistan had in nuclear “risk taking” and over the United States.63 Ambassador Oakley thought that the Pakistanis “manipulated very skillfully” their nuclear program in order to largely “obtain concessions [or protection] from the United 59. George Crile, Charlie Wilson’s War: The Extraordinary Story of How the Wildest Man in Congress and a Rogue CIA Agent Changed the History of Our Times (New York: Grove Press, 2003), p. 463. 60. Ibid., p. 463. 61. Ibid., pp. 463–­64. 62. Ruud Lubbers interview quoted in Adrian Levy and Catherine Scott-­Clark, Deception: Pakistan, the United States, and the Secret Trade in Nuclear Weapons (New York: Walker & Company, 2007), pp. 117–­18. 63. Crile, Charlie Wilson’s War, pp. 463–­64.

Pakistan 73

States.”64 It further appears that Pakistan viewed the fact that the United States worked very “aggressively [to convince Pakistan] not to put nuclear weapons together,” according to a senior U.S. government official in the region at the time, as a source of leverage against Washington in potential crises, even if an aid cutoff would hurt Pakistan immensely more than it would hurt America.65 This same official does note that the Pakistan military might have believed that it could exploit U.S. fears of it crossing the nuclear assembly threshold—­which would trigger the Pressler Amendment and an aid cutoff to Pakistan—­but it was ultimately predicated on the “self-­ delusion” that the Pakistan military was the lynchpin to the U.S. strategy in Afghanistan, and that on this count the Pakistanis were telling themselves a “bullshit story.”66 Nevertheless, American behavior in trying to aggressively stay Pakistan’s nuclear hand short of weapon assembly seemed to convince the Pakistan military that this story was true and that they indeed had a point of leverage with Washington. The United States and Pakistan in the 1980s were allies, but not friends, and had a set of overlapping interests that Pakistan believed it could potentially exploit to its benefit in other areas—­particularly with respect to nuclear weapons and nuclear posture. After Zia’s death in a 1988 plane crash, the United States directly sent protective warnings worldwide which Ambassador Robert Oakley claimed effectively read as follows: “Don’t mess with the Paks, or the United States is going to be on your ass.”67 The patron-­client dynamic continued even as the Afghan jihad wound down with the Soviet exit from Afghanistan, with the CIA continuing its relationship with the ISI to establish some semblance of acceptable order in Afghanistan while also attempting, according to Oakley, to “shore up” Benazir Bhutto against internal enemies within Pakistan bent on taking her down.68 Indeed, during early 1990 and even subsequently in 1991, the CIA was still intensely involved with the ISI, preparing for the Afghan fighting seasons to install their chosen government in Kabul, with $280 million still flowing to the program.69 Indeed, in 1991, even after the Pressler Amendment had been invoked in October 1990 suspending military assistance to Pakistan, the CIA involvement in Afghanistan surprised even President Bush, who asked Milton Bearden, “Is

64. Ambassador Robert Oakley phone interview with author, Washington, DC, November 8, 2011. 65. Author interview with senior U.S. government official in the region at the time, November 25, 2011. 66. Ibid. 67. Oakley quoted in Coll, Ghost Wars, p. 179. 68. Coll, Ghost Wars, p. 192. 69. Ibid., pp. 207–­10.

74

Chapter 3

that thing still going on?”70 The Soviet withdrawal did not end America’s reliance on the Pakistan military establishment or, accordingly, Pakistan’s ability to exploit Washington’s interest in regional stability, particularly on Pakistan’s eastern border with India. The marriage of convenience between the United States and Pakistan ended in an abrupt separation on January 1, 1992, when the CIA’s legal authority to conduct covert operations in Afghanistan ended. After this, Coll writes, the “CIA had little to offer [Pakistan] anymore.”71 Nevertheless, from the late 1980s through the early 1990s, the complicated U.S.-­Pakistan relationship was a key enabling feature for the catalytic posture. The catalytic posture is precisely what optimization theory would predict. The catalytic posture involved risks for Pakistan: it depended on the actions of a third party and on indirect deterrence. Nonetheless, it avoided extremely costly sanctions and aid cutoffs, and also avoided an unnecessary provocation to India to go overtly nuclear itself. The catalytic posture therefore had significant net advantages for Pakistan. Given the security environment and the nature of Pakistan’s relationship with the United States in this period, my theory predicts that Pakistan should select a catalytic nuclear posture which, although providing only indirect deterrence against India, was less costly and provided significant security advantages over the alternatives. The realist, technological determinism, and strategic culture lenses diverge from my prediction. All three would suggest that as soon as Pakistan developed a nuclear weapons capability, it should have directly and immediately incorporated nuclear weapons into its military doctrine in order to deter its conventionally superior proximate neighbor, India. A strict realist analysis would suggest that, in order to maximize the deterrent value of nuclear weapons against India—­a state against whom it had suffered a cata­ strophic defeat—­Pakistan should have demonstrated the reliability of its deterrent as soon as possible and threatened the first use of nuclear weapons against Indian conventional breaches. That is, a narrow materialist and security analysis would predict that Pakistan should not have relied on probabilistic and whimsical third-­party intervention to secure its territorial integrity but instead that it should have adopted a transparent asymmetric escalation nuclear posture as soon as it was capable of doing so. According to the realist explanation, Pakistan should not have outsourced its security to the United States if it had the capability to demonstrate a direct nuclear deterrent to India. The evidence suggests, however, that Islamabad did not consider testing nuclear weapons in this period—­even though India had

70. President George H. W. Bush quoted in Coll, Ghost Wars, p. 228. 71. Coll, Ghost Wars, p. 234.

Pakistan 75

conducted a peaceful nuclear explosion in 1974—­and there was no integration or training of nuclear assets into military doctrine. The technological determinism theory of nuclear posture might capture the composition and fact of Pakistan’s nuclear program during the late 1980s, but not the form or character of the posture. Why did Pakistan not test when it had sufficient fissile material for a small nuclear arsenal? Why did it adopt a catalytic posture rather than rely on existential deterrence? Why did Pakistan not seek to develop or acquire more robust delivery vehicles for its small nuclear arsenal? These are all questions left unanswered by the theory of technological determinism. Yes, Pakistan had limited technical capabilities by the late 1980s, but it still could have tested a nuclear device and become an overt nuclear power and it could have attempted to develop a more sophisticated delivery capability. But it chose not to do so, opting instead for a catalytic posture that relied heavily on American incentives to keep Pakistan’s nuclear program limited and Islamabad focused on the conflict in Afghanistan. An explanation of Pakistan’s posture based on strategic culture would emphasize the role of the army in managing the Pakistani state and security affairs. As the cliché famously goes, “While most states have an army, in Pakistan the army has a state.”72 As such, with a largely autonomous army that penetrates almost every organ of the state and which has historically viewed itself as the guardian of the state since independence, Pakistan’s strategic culture at this time was largely the strategic culture of the army. As Stephen Cohen argues in his seminal work on the Pakistan Army, two core beliefs underlay the army’s worldview: first, that India posed an existential threat to Pakistan and that war with India was a constant possibility; and, second, that only the army—­not the civilians—­could save Pakistan from such a fate.73 Further, throughout the late 1980s, the Pakistan Army was developing a conventional doctrine of offensive-­defense as a response to India’s large-­scale maneuver doctrine known as the Sundarji doctrine. This was designed to allow the Pakistan Army to retain the initiative and execute counteroffensives against an incoming Indian onslaught. Given that Gen. 72. See Stephen P. Cohen, The Idea of Pakistan (Washington, DC: Brookings Institution Press, 2004); Ayesha Siddiqa, Military Inc: Inside Pakistan’s Military Economy (London: Pluto, 2007); Ayesha Jalal, The State of Martial Rule: The Origins of Pakistan’s Political Economy of Defence (Cambridge: Cambridge University Press, 1990); Ayesha Jalal, Authoritarianism and Democracy in South Asia: A Comparative and Historical Analysis (Cambridge: Cambridge University Press, 1995); Hasan Askari Rizvi, Military, State, and Society in Pakistan (London: St. Martin’s Press, 2000); Nawaz, Crossed Swords; Stephen P. Cohen, The Pakistan Army (Berkeley: University of California Press, 1984); Brian Cloughley, A History of the Pakistan Army: Wars and Insurrections, 2nd ed. (London: Oxford University Press, 2000). 73. See Cohen, The Pakistan Army.

76

Chapter 3

Zia held the reigns of the Pakistani state, its army, and its nuclear program at the point when Pakistan weaponized its capability, a strategic culture analysis would suggest that the Pakistan Army—­to retain initiative against what it viewed as a mortal threat to its existence—­would have selected an asymmetric escalation posture that integrated nuclear weapons into the army doctrine. It did not. Only optimization theory captures Pakistan’s choice of a catalytic nuclear posture from the late 1980s through 1998.

Pakistani Nuclear Posture Phase II: Asymmetric Escalation There were no significant military crises between India and Pakistan between 1990 and 1998, but American abandonment of Pakistan forced it to make a decision to start shifting its posture. Steady uranium enrichment was actively coupled with the search for delivery vehicles in this period, most notably Chinese export-­class and North Korean No Dong missiles. However, without testing nuclear weapons, Pakistan could not credibly threaten the first use of lower yield nuclear weapons in a tactical environment. Thus, in this period, Pakistan had perhaps a tenuous catalytic posture that would attempt to draw the United States back into South Asia on its behalf if necessary—­it did not yet have the capabilities for an assured retaliation or asymmetric escalation posture—­but it appears to have made a decision to shift postures. The decision to shift appears to have occurred in the early 1990s, when Pakistan began in earnest to seek a missile-­delivery capability and devoted significant effort to developing a more reliable air delivery vector. However, the lag time of this development means that Pakistan’s actual posture would only become operational several years later. Indeed, for Pakistan to move credibly to a more robust nuclear posture, it would almost certainly have to conduct tests to demonstrate the reliability and functionality of its designs. But Pakistan probably could not weather the international fallout from a unilateral test. It would have to wait until India tested and then follow suit, exploiting the political cover provided by an Indian test.74 Thus, although Pakistan could not fully switch nuclear postures between 1991 and 1998, it was forced to seek capabilities—­particularly attack aircraft and dual-­use ballistic missiles (especially the Shaheen family based on China’s M-­class export missiles and the Ghauri family based on North Korea’s No-­Dong missile)—­to prepare for a potential shift. With the Bharatiya Janata Party’s (BJP) tests of India’s nuclear weapons capability on May 11 and 13, 1998, however, Pakistan found itself facing a conventional existential threat that was now overtly nuclear. Pakistan quickly 74. Ahmed, “Pakistan’s Nuclear Weapons Program,” pp. 194–­96.

Pakistan 77

took the opportunity to follow suit on May 28 and 30, testing six fission devices. These tests demonstrated the functionality and reliability of Pakistan’s nuclear designs. Afterward, Pakistan consummated its shift to an asymmetric escalation posture that fully integrated nuclear weapons into its military forces and doctrine, credibly threatening the first use of nuclear weapons against Indian conventional forces in the event that they breached Pakistan’s territorial integrity.75 India’s nuclear tests, conducted largely for domestic political reasons, were the final enabler permitting Pakistan’s switch to an asymmetric escalation posture. Indeed, given that Pakistan’s ballistic missile capabilities were thin even in 1998, Pakistan may have even preferred to allow its warhead and delivery technology to mature before fully moving to an asymmetric escalation posture. India’s tests in May 1998, however, gave Pakistan the political cover to test but also left it with little option but to do so, so that it could finally fully adopt a posture that sought to directly deter nuclear India’s conventional military power. It is worth noting that the international sanctions on Pakistan that followed its tests proved correct its earlier calculations about the risks associated with an overt nuclear status: they were crippling to Pakistan’s fragile economy. After testing nuclear weapons in May 1998, I find that Pakistan credibly moved to an asymmetric escalation posture that threatened the first use of nuclear weapons against opposing conventional forces in the event they crossed the international border. Asymmetric escalation postures are aimed at deterring both conventional and nuclear conflict against a state. Unlike the catalytic posture in which the use of nuclear weapons is only contemplated in the face of a significant conventional threat to a state and whose signal is directed toward a third-­party patron, an asymmetric escalation posture attempts to directly deter conventional conflict by another nuclear or non-­nuclear state in toto by threatening the first use of nuclear weapons in either a tactical or strategic strike. The deterrent threat is clearly communicated to the adversary in question, not a third-­party patron. While the third party may still have incentives to cap escalation, the primary goal of the posture is no longer to compel that patron’s intervention but rather to directly deter an opponent. The operationalization of an asymmetric escalation posture can vary (e.g., massive retaliation or flexible response), but the Pakistani posture is explicitly modeled on NATO’s flexible response

75. See Aga Shahi, Zulfiqar Khan, and Abdul Sattar, “Securing Nuclear Peace,” The News International, October 5, 1999 reprinted in P. R. Chari, Sonika Gupta, and Arpit Rajan, eds., Nuclear Stability in Southern Asia (Delhi: Manohar Books, 2003), pp. 189–­94; and Smruti S. Pattanaik, “Pakistan’s Nuclear Strategy,” Strategic Analysis, vol. 27, no. 1 (January–­March 2003), p. 13.

78

Chapter 3

posture which threatened the first use of nuclear weapons in theater should conventional deterrence fail. An asymmetric escalation posture therefore calls for the first use of nuclear weapons and requires sufficient tactical and potentially survivable second-­strike strategic weapons to absorb retaliation from an opponent. While peacetime deployments can be centralized, an asymmetric escalator must have the ability to disperse and deploy assets in the field quickly in the event of a crisis in order to credibly threaten first use against a conventional breach of its territory. As such, the asymmetric escalation posture is the most aggressive option available to nuclear states and involves fielding both tactical and strategic capabilities that can be mobilized and launched quickly, often through pre-­delegated authority in the event of conventional conflict. According to my theory, states that select asymmetric escalation postures are often those that face a conventionally superior proximate threat without the help of a third-­party patron. In the early 1990s, Pakistan found itself in this situation. However, the fear of international backlash prevented it from fully operationalizing an asymmetric escalation posture by testing nuclear weapons. After May 1998, India’s tests provided the political cover that allowed Pakistan to establish a fully operational asymmetric escalation posture. While Pakistan has compensated for its conventional inferiority by relying on high-­quality materiel and operating on interior lines of communication, these features only prevent India from overrunning Pakistan in a relatively short conflict. In a longer conventional conflict, India’s overwhelming aggregate quantitative and qualitative power advantage would likely be decisive.76 From a security perspective then, Pakistan must therefore take all necessary measures to ensure that any conventional conflict with India is blunted early. The optimal nuclear posture for Pakistan to achieve such an outcome is threatening the first use of nuclear weapons on advancing Indian forces after they have crossed the international border on Pakistani soil. Such a posture would blunt the Indian conventional assault and, critically, would give India little justification for a disproportionate nuclear strike on Pakistan’s strategic centers, since Pakistan would not have targeted India’s cities itself.77 In this scenario, Pakistan’s second-­strike 76. See Gill, “Brasstacks: Prudently Pessimistic,” pp. 44–­45. 77. Indeed, Pakistani nuclear planners seem to believe that use of nuclear weapons in a demonstration detonation or actually on Indian forces to halt an Indian conventional thrust across the international border would be a war-­terminating event. This analysis presumes that India would not have the justification to retaliate in this scenario, even though Indian doctrine claims it would, because (a) Indian forces would have been operating offensively on Pakistan’s territory and (b) India is only oriented for strategic retaliation, and Indian nuclear retaliation on a Pakistani city would be a disproportionate response. In fact, Vice Admiral (Ret.) Vijay Shankar, a former head of India’s Strategic Force Command, believes

Pakistan 79

capabilities would not have to be terribly robust, since even the small threat of a counterstrike could deter India from retaliating. Rather, the onus of the Pakistani posture would be on first-­strike capabilities in a limited theater setting in which Pakistan would retain the initiative to escalate the conflict, thus mitigating the numerical and procedural requirements to have a significant survivable second-­strike capability. Pakistan has made it clear that the first use of nuclear weapons is in fact its operational nuclear doctrine.78 There have been several authoritative statements on the conditions under which Pakistan would envision nuclear use. The most important of these was given by Lt. Gen. Khalid Kidwai, the first head of the Strategic Plans Division (SPD) which is responsible for stewarding Pakistan’s nuclear assets. In January 2002, Kidwai gave an oft-­cited interview in which he outlined the conditions for Pakistani nuclear use: Nuclear weapons are aimed solely at India. In case that deterrence fails, they will be used if: (a) India attacks Pakistan and conquers a large part of its territory (space threshold); (b) India destroys a large part of either its land or air forces (military threshold);

that a demonstration detonation would probably be a war-­terminating event and allow Pakistan to “sue for peace,” but of course he refused to comment about whether India would certainly retaliate following use on Indian forces, saying, “it is in our doctrine, one can go read it in black and white. But the ultimate decision would be left to the Prime Minister and may depend on the circumstances.” Author interview with Vice Admiral (Ret.) Vijay Shankar, New York City, July 15, 2011. 78. Existing work on Pakistan’s nuclear doctrine includes Scott D. Sagan, “The Evolution of Pakistani and Indian Nuclear Doctrine” in Scott D. Sagan, ed., Inside Nuclear South Asia (Stanford: Stanford University Press, 2009); Peter R. Lavoy, “Pakistan’s Nuclear Doctrine,” in Rafiq Dossani and Henry S. Rowen, Prospects for Peace in South Asia (Stanford: Stanford University Press, 2005), pp. 280–­300; Zafar Iqbal Cheema, “Pakistan’s Nuclear Use Doctrine and Command and Control,” in Lavoy, Sagan, and Wirtz, eds., Planning the Unthinkable, pp. 158–­181; Lavoy, “Pakistan’s Nuclear Posture”; Maj. Gen. (Ret.) Mahmud Ali Durrani, “Pakistan’s Strategic Thinking and the Role of Nuclear Weapons,” Cooperative Monitoring Center Occasional Paper 37 (Albuquerque: Sandia National Laboratories, 2004); IISS, Nuclear Black Markets: Pakistan, AQ Khan, and the Rise of Proliferation Networks (London: IISS, 2007), chapters 1 and 5; Hoyt, “Pakistani Nuclear Doctrine,” pp. 956–­77; Zafar Iqbal Cheema, “The Role of Nuclear Weapons in Pakistan’s Defense Strategy,” Islamabad Policy Research Institute Journal, vol. 4, no. 2 (Summer 2004) pp. 59–­ 80; Raja Menon, “Nuclear Doctrine in South Asia,” in Chari, Gupta, and Rajan, Nuclear Stability in Southern Asia, pp. 99–­118; Neil Joeck, “Maintaining Nuclear Stability in South Asia,” International Institute for Strategic Studies, Adelphi Paper 312 (Oxford: Oxford University Press, 1997); Feroz Hassan Khan and Peter R. Lavoy, “Pakistan: The Dilemma of Nuclear Deterrence,” in Muthiah Alagappa, ed., The Long Shadow: Nuclear Weapons and Security in 21st Century Asia (Stanford: Stanford University Press, 2008), pp. 215–­40.

80

Chapter 3 (c) India proceeds to the economic strangling of Pakistan (economic strangling); (d) India pushes Pakistan into political destabilization or creates a large scale internal subversion in Pakistan (domestic destabilization).79

Kidwai’s exposition of when Pakistani nuclear weapons might be used was illuminating, yet also intentionally vague for obvious reasons. Brig. Gen. (Ret.) Feroz Hassan Khan, who served under Kidwai in the SPD, notes the reason for Pakistan’s ambiguity about precise red-­lines: “By declaring the red-­line, what you are . . . indicating to them is that ‘up to this point it is fair; you can come and beat me up,’ you know. That is not acceptable . . . By declaring red lines, Pakistan erodes its deterrent value of the nuclear weapons . . . India thinks they know the threshold, but I don’t think so . . . They think that . . . they can go for a limited strike . . . They are not calculating that there could be a tit-­for-­tat response and there is going to be a chain-­reaction effect which will slip out of their control.”80 Thus, the territory that would have to be conquered or the quantity of forces depleted required to reach Pakistan’s red-­lines (or nuclear threshold) are left ambiguous, but it is generally believed that if the Pakistan Army were severely crippled or Indian forces thrust to Pakistan’s so-­called green-­belt roughly 30–­50 miles past the international border, nuclear use might be triggered.81 The key feature of the Pakistani declaratory doctrine is that Pakistan would rely on the first use of nuclear weapons but only once its conventional deterrent or forces were significantly degraded. In short, the military threshold appears to be the critical red-­line since the space threshold could only be violated if the Pakistan Army were already overrun. The other semi-­authoritative writing on Pakistan’s nuclear doctrine, by several former Pakistani foreign ministers, corroborates the relatively early red-­line in Pakistani nuclear thinking. They write that the “assumption has been that if the enemy launches a general war and undertakes a piercing attack threatening to occupy large territory or communication junctions, the ‘weapon of last resort’ would have to be invoked.”82 Invoking the explicit comparison to NATO’s doctrine against the Warsaw Pact, they write that Pakistan’s first-­use posture is critical to deterring India from “exploit[ing]

79. Kidwai quoted in Nuclear Safety, Nuclear Stability, and Nuclear Strategy in Pakistan, A Concise Report of a Visit by Landau Network Centro Volto, January 2002. Available at http://www.pugwash.org/september11/pakistan-nuclear.htm. 80. Feroz Hassan Khan quoted in Martin Schram, Avoiding Armageddon: Our Future. Our Choice (New York: Basic Books, 2003), p. 53. 81. Author interview with anonymous knowledgeable Pakistani scholar, February 23, 2009. 82. Shahi, Khan, and Sattar, “Securing Nuclear Peace,” p. 191.

Pakistan 81

theatre superiority in conventional forces.”83 Brig. Gen. (Ret.) Feroz Hassan Khan has also written: The Pakistani situation is akin to NATO’s position in the Cold War. There are geographic gaps and corridors similar to those that existed in Europe (such as the erstwhile “Fulda gap”) that are vulnerable to exploitation by mechanized Indian forces. In general, the situation is as it was with East and West Germany. With its relatively smaller conventional force, and lacking adequate technical means, especially in early warning and surveillance, Pakistan relies on a more proactive nuclear defensive policy.84

Even though NATO was a status quo power in Europe, while Pakistan is a more revisionist actor, the explicit invocation of a deterrence posture modeled on NATO’s suggests that the Pakistan Army views the threshold for nuclear first use as relatively low in a conventional conflict with India. Zafar Iqbal Cheema, a knowledgeable and well-­connected Pakistani scholar, has also written that the threshold may be very low. While an Indian Strike Corps thrust through Pakistan’s vulnerable plains sector (from Sukkur to Bahawalpur to Lahore) might cross Pakistan’s red-­lines, Cheema writes that even “a repeat of the Brasstacks exercises might compel Pakistan to consider the use of nuclear weapons.”85 While such a threat is almost certainly not credible, since it would involve the use of nuclear weapons on Indian forces on Indian soil prior to military hostilities, and would thus almost certainly invite Indian retaliation, it is not necessarily the case that Pakistan’s territorial red-­line lies as deep as the green-­belt. Indeed, it is likely that the military threshold would be reached first and that the depletion of some significant level of Pakistan’s Army and Air Force would first trigger nuclear use. Regardless of where the precise red-­line lies, Pakistan Army writing on nuclear posture has focused on how to make the enforcement of such a red-­ line credible. Lt. Gen. (Ret.) Sardar F. S. Lodi writes that Pakistan would likely employ a graduated response akin to NATO doctrine in the face of advancing Indian conventional forces: [W]e will use nuclear weapons if attacked by India even if the attack is with conventional weapons.  .  .  . This would entail a stage-­by-­stage approach in 83. Ibid., p. 194. A critical distinction, however, is that, unlike NATO, which was a largely status quo power in the European Theater, Pakistan is a relatively revisionist power in the South Asian theater. 84. Feroz Hassan Khan, “Challenges to Nuclear Stability in South Asia,” Nonproliferation Review, Spring 2003, p. 65. 85. Zafar Iqbal Cheema, “Pakistan’s Nuclear Use Doctrine and Command and Control” in Lavoy, Sagan, and Wirtz, Planning the Unthinkable:, p. 177. Emphasis added.

82

Chapter 3

which the nuclear threat is increased at each step to deter India from attack. The first step could be a public or private warning, the second a demonstration explosion of a small nuclear weapon on its own soil, the third step would be the use of a few nuclear weapons on its own soil against Indian attacking forces. The fourth stage would be used against critical but purely military targets in India across the border from Pakistan. Probably in thinly populated areas in the desert or semi-­desert, causing least collateral damage [as] this may prevent Indian retaliation against cities in Pakistan. Some weapon systems would be in reserve for the counter-­value role. These weapons would be safe from Indian attack as some would be airborne while the ground based ones are mobile and could be moved around the country.86

Others such as Brig. Gen. (Ret.) Feroz Hassan Khan have argued that Pakistan cannot afford to rely on a graduated strategy but must instead use nuclear forces early and massively in a conflict against India, against both tactical and strategic targets.87 The debate has not been publicly resolved, and the expansion of Pakistan’s nuclear capabilities across the spectrum of use options is consistent with both strategies, but there is little doubt that after testing nuclear weapons in 1998, Pakistan moved to incorporate nuclear weapons into its larger military doctrine, with a goal of making their weaponized deterrent credible and potentially usable to achieve Pakistan’s overriding goal of deterring an Indian conventional war. Peter Lavoy writes that, “Before [1998], nuclear weapons had not been integrated into Pakistani military plans; the armed forces had no nuclear employment doctrine to speak of, and command and control over the nuclear arsenal and delivery systems was only vaguely defined and loosely organized.”88 This started to change after May 1998.

Capabilities and Command and Control Though Pakistani rhetoric and declaratory doctrine has articulated an asymmetric escalation posture, has Pakistan developed the capabilities and command-­and-­control procedures to translate that declaratory doctrine into a credible posture? I argue that it has. In terms of capabilities, Pakistan could adopt an asymmetric escalation posture suited to its particular 86. Sardar F. S. Lodi, “Pakistan’s Nuclear Doctrine,” Defence Journal (Pakistan), vol. 3, no. 4 (April 1999), http://www.defencejournal.com/apr99/pak-nuclear-doctrine.htm. This idea was outlined in Syed Anwar Mehdi, “Nuclear Ambivalence versus a Well-­Defined Policy Involving Miximum Political Fallout,” The Citadel (Pakistan), no. 2 (1994), pp. 55–­67. See also Stephen P. Cohen, The Pakistan Army, pp. 177–­78. 87. Feroz Hassan Khan, interview with author, March 19, 2009. 88. Lavoy, “Pakistan’s Nuclear Posture: Security and Survivability,” p. 1.

Pakistan 83

context with even a relatively small arsenal, especially if the first use is envisioned under conditions in which the opponent would have little justification to retaliate against Pakistan’s strategic centers. Pakistan was believed to have several dozen nuclear weapons in 1998, and is largely believed to have somewhere in the neighborhood of 90–­110 warheads as of 2012, with steady uranium enrichment capability and an increasing plutonium production capability.89 For Pakistan’s purposes, these numbers would allow it to—­though perhaps barely—­achieve the posture described. In terms of delivery capabilities, Pakistan has both short-­range and medium-­range ballistic missiles that place India’s major strategic centers within range, as well as nuclear-­capable aircraft, most notably the F-­16 and the 60-­km nuclear-­capable Hatf-­9 Nasr missile which “carries nuclear warheads of appropriate yield with high accuracy, shoot and scoot attributes,” which could be used to deliver a tactical nuclear warhead on advancing Indian forces.90 The development of battlefield nuclear capabilities, such as the Nasr and a growing array of cruise missile capabilities, represents Pakistan’s efforts to lower the nuclear threshold in response to Indian conventional power—­exactly as the asymmetric escalation posture would demand. Pakistan’s ballistic missile families cover a variety of ranges from the 85-­km Hatf I to the 2,500-­km Ghauri. The Shaheen and Ghauri families are road-­mobile, which enhances their survivability, while the Shaheen family uses solid fuel which reduces launch times. The CHIC-­4 uranium fission design is believed to be for a missile-­mateable warhead, and as Pakistan’s plutonium production capability increases to allow the development of warheads with greater yield-­to-­weight ratios, the ballistic missile arm of its delivery capabilities will become increasingly powerful. In short, Pakistan possesses the capabilities to make an asymmetric escalation posture credible. What about command-­and-­control procedures? In order to credibly achieve an asymmetric escalation posture, Pakistan would have to be able to develop command-­and-­control procedures and architectures to allow for the rapid deployment and use of nuclear weapons. The most critical feature that enables this posture is that Pakistan’s nuclear arsenal has always been under strict military control. Under Gen. Zia’s tenure, the nuclear weapons program was firmly under army direction and remained so even during the Benazir Bhutto and Nawaz Sharif years even though the military did not have possession of nuclear assets until after 1998.91 The custody 89. “Pakistan’s Nuclear Forces 2011,” Bulletin of the Atomic Scientists, vol. 67, no. 4 (July/August 2011), pp. 91–­99. 90. See Inter Services Public Relations Press Release (Pakistan), No PR94/2011-­ISPR, April 19, 2011. Available at http://www.ispr.gov.pk/front/main.asp?o=t-press_release&id=1721. 91. Ahmed, “Pakistan’s Nuclear Weapons Program,” p. 188.

84

Chapter 3

of, and command-­and-­control arrangements for, nuclear weapons were a de facto part of the national military command structure until Gen. Pervez Musharraf created the official National Command Authority (NCA) in February 2000, which he chaired.92 The structure of the NCA was initially created when Musharraf was simultaneously the Chief of Army Staff as well as President; he chaired the NCA, although now it is the purview of the Prime Minister, while operational control for Pakistan’s nuclear assets are coordinated by the Strategic Plans Division (SPD) which is led by a Lieutenant General-­rank officer. Even though political power was transferred to a civilian government in 2008, there is no evidence that the Chief of Army Staff (or Chair of the Joint Chiefs of Staff Committee) or SPD has relinquished de facto control over Pakistan’s nuclear assets to civilian authorities. Indeed, though the peculiarities of Musharraf’s constitutional changes continue to keep a civilian as the de jure head of the NCA, the military—­and the army in continues to retain operational control of the arsenal and particular—­ would not, in practice, be subject to any civilian oversight on either the negative or positive control side.93 Various organizations, such as PAEC, the National Defence Complex (NDC), and Khan Research Laboratories (KRL), are responsible for production of nuclear components. Nonetheless, the army and air force have custody of all the necessary components for useable weapons—­the delivery vehicles and the explosives and physics package that constitute a functional warhead.94 Thus, Pakistan’s nuclear command-­and-­control architecture and decision making occurs within a clearly praetorian structure. This is further supported by the fact that Pakistan’s command-­and-­control structures for conventional and nuclear operations are fully integrated. That is, the decision-­making nerve center during a crisis or conflict for both conventional and nuclear operations would be the Joint Services Command Center, in which the Chief of Army Staff is the dominant authority. So while security considerations may have propelled Pakistan toward nuclear weapons and a first-­use asymmetric escalation posture, it is primarily the Pakistan Army—­with its own preferences and pathologies—­that is charged with implementing this posture and ensuring the security and the safety of Pakistan’s nuclear weapons. As Scott Sagan and Barry Posen have argued in different ways, organizations, especially military organizations, tend to undertake procedures that underestimate the probability of 92. Lavoy, “Pakistan’s Nuclear Posture: Security and Survivability,” p. 12. 93. See Zeeshan Haider, “Pakistan’s Nuclear Command Stays Unchanged,” Reuters, April 8, 2008. 94. Author interview with former high-­level U.S. intelligence official, February 28, 2009; confirmed by author discussion with Feroz Hassan Khan, March 29, 2009.

Pakistan 85

accidents or unauthorized use, favor offensive strategies and procedures that allow the retention of the initiative and independence, and take steps to minimize civilian interference.95 This suggests that the Pakistan Army would likely undertake procedures that overestimate its own professionalization and that aim to generate an offensively oriented doctrine to ensure that Pakistan’s nuclear weapons would not be preemptively destroyed. The implication is that the Pakistan Army would adopt largely delegative command-­and-­control procedures that place nuclear weapons in the hands of theater commanders with few physical impediments to their release. Indeed, Tim Hoyt has written, “It is apparent that Pakistan’s C2 procedures are delegative, lean heavily toward the always side of the always/never divide, and probably include both devolution and possibly pre-­delegation in order to ensure the use of weapons.”96 This largely delegative command-­ and-­control structure is tinged with much uncertainty in the open-­source literature, but probably includes three key features that enable rapid assembly, movement, and delivery even under a potentially chaotic crisis situation. First, although the Pakistan Army claims to store the bulk of its components for a deliverable nuclear asset—­the pits, explosives packages, and delivery vehicles—­separately from each other during peacetime to enhance security, it is believed that single or proximate military bases store all the necessary components for rapid assembly and deployment in a crisis.97 General Kidwai has noted that nuclear weapons can be assembled, moved, and mated with delivery vehicles “very quickly,”98 suggesting that all the components needed to assemble and deliver a functional nuclear asset are close to each other: “Whether separated by a yards [sic] or miles the weapons will be ready to go in no time.”99 Pakistan Army officials have further said “that in emergency conditions . . . equipment is repositioned to allow for rapid assembly.”100 Indeed, according to Maurizio Martellini of the Landau Network Centro Volto who interviewed the SPD’s Kidwai in 2008: “Kidwai stated very clearly that the SPD never officially declared that the nuclear cores of the bombs are split from their detonators and that the 95. See Posen, Sources of Military Doctrine, especially chapters 1 and 2; and Sagan, Limits of Safety. 96. Hoyt, “Pakistani Nuclear Doctrine,” p. 966. 97. Some small number of fully assembled warheads may even be maintained for emergency deployment. But even if the entire force is kept in component form, Pakistan appears to have taken measures to ensure rapid assembly and deployment if necessary. 98. Kidwai quoted in P. Cotta-­Ramusino and M. Martellini, Nuclear Safety, Nuclear Stability, and Nuclear Strategy in Pakistan: A Concise Report of a Visit by Landau Network-­ Centro Volto (Como, Italy: Landau Network-­Centro Volto), January 14, 2002. 99. Kidwai quoted in Ron Moreau, “Pakistan’s Nukes,” Newsweek, January 26, 2008. 100. Pakistan military officials cited in Molly Moore and Kamran Khan, “Pakistan Moves Nuclear Weapons: Musharraf Says Arsenal is Now Secure,” Washington Post, November 11, 2001, p. A1.

86

Chapter 3

warheads are kept separate from their delivery system. Instead, without getting into operational specifics, the official position of the SPD is that the weapons [according directly to Kidwai] ‘will be ready when required, at the shortest notice . . . distance is not the issue, the issue is timing. Separation is more linked to time rather than to space’.”101 The present disposition of assets may result from technical features of Pakistan’s weapons designs, wherein weapons technicians may prefer that they not be stored as fully assembled warheads for stability or safety reasons.102 But as its nuclear weapons designs and technology become more sophisticated,103 there is growing evidence that Pakistan may move closer to a ready deterrent to enhance the credibility of the asymmetric escalation posture. This is particularly true of “encapsulated” or “canisterized” systems such as the Nasr which are kept in sealed tubes, where the warhead would likely have to be premated, or mated to the missile immediately before it is deployed. Components are believed to be stored at several military locations (the existence of 6–­12 secret fixed locations has been reported), though it is likely that there are dummy sites and an elaborate tunnel network to enhance security and survivability.104 Specifically, bases proximate to—­but in the rear of—­sectors where Indian forces would advance are thought to store weapons that can be constituted and deployed rapidly in this key the­ater of operations. Logic dictates that Pakistan most likely stores its nuclear assets at military locations in a tight elongated band toward the rear of Punjab and Sindh, south of Islamabad; assets too close to the Indian border (e.g., Lahore) might be susceptible to preemption or capture, while assets too far in the rear would be in the chaotic Northwest Frontier Province and Federally Administered Tribal Areas, and quite distant from likely conflict theaters.105 These constraints narrow the likely locations of Paki101. Maurizio Martellini, “Security and Safety Issues about the Nuclear Complex: Pakistan’s Standpoints. A Concise Report of a Visit to Islamabad by Landau Network Centro Volta (LNCV) Mission Carried out on February 9–­13 2008,” p. 2. 102. This is a challenge most new nuclear powers must face, see, e.g., Lewis and Xue, China Builds the Bomb, pp. 156–­58; also see David Albright and Mark Hibbs, “Pakistan’s Bomb: Out of the Closet,” Bulletin of the Atomic Scientists, vol. 48, no. 6 (July/August 1992), pp. 38–­43. 103. David Albright and Paul Brannan, “Pakistan Expanding Plutonium Separation Facility near Rawalpindi,” Institute for Science and International Security, May 19, 2009. available at http://www.isis-online.org/publications/southasia/PakistanExpandingNewlabs .pdf. 104. Moore and Khan, “Pakistan Moves Nuclear Weapons,” p. A1; and Kenneth Luongo and Naeem Salik, “Building Confidence in Pakistan’s Nuclear Security,” Arms Control Today, December 2007. Available at http://www.armscontrol.org/act/2007_12/Luongo. 105. Author interview with Brig. Gen. (Ret.) Khan, March 19, 2009. He stated that “there are no nuclear weapons in Lahore” since they would be too vulnerable to Indian seizure.

Pakistan 87

stani nuclear assets to army and air force installations 50–­150 miles from the international border. While some assets may be pre-­positioned closer to the most likely theaters, the bulk of the assets are likely located around or behind the Indus River or its rearward tributaries (e.g., the Chenab). This deployment pattern would optimize credibility and survivability. Second, the Pakistan Army likely institutes positive control procedures to rapidly deploy these assets in the event of a crisis with India (or external threats to the integrity of the nuclear arsenal). This may include pre-­ delegating some authority to end users in its chain of command to move and release nuclear weapons under the plausible scenario that communication was to break down in the midst of a crisis. Indeed, the so-­called two-­ or-­three-­man rule that Pakistan appears to employ involves codes split at lower levels of military command; codes and warheads are also presently believed to be co-­located. Maj. Gen. (Ret.) Mahmud Durrani, who would subsequently serve as Pakistan’s national security advisor, claimed in 2004 that: “For example, at an air force base the code may be divided between the base commander and the unit commander. In the army, the code may be divided between the group commander and the unit commander. This rule also applies to a launch site. The only exception is a single pilot who will receive the full code during flight.”106 It may also be that physical de-­mating stewardship procedures gave rise to the two-­or-­three-­man rule: separate units may have custody of respective components and all parties’ consent would be required to assemble a functional weapon. In short, it appears that lower-­level officers may be ceded some authority, particularly as a crisis unfolds, to assemble and release nuclear weapons should circumstances demand it. Third, there may be few negative controls to inhibit the release of nuclear weapons. Although stated procedures nominally call for the NCA to authorize each step—­assembly, mating, movement, and release of nuclear weapons—­there may be few safeguards that physically prevent lower-­level officers from taking each step without authorization if they deem it necessary, as a hedge against the possibility that the NCA is out of reach or decapitated. Brig. Gen. (Ret.) Khan concedes that “[A] theater commander would probably [be able to] take matters into his own hands . . . should a trade-­off be required, battle effectiveness of the nuclear force will trump centralized control.”107 For an army persistently concerned with its lack of strategic depth, the absence of robust negative controls may be designed to physically enable quick release of nuclear weapons in the event

106. Durrani, “Pakistan’s Strategic Thinking and the Role of Nuclear Weapons,” p. 33. 107. Feroz Hassan Khan, “Nuclear Command and Control in South Asia during Peace, Crisis, and War,” Contemporary South Asia, vol. 14, no. 2 (June 2005), p. 169.

88

Chapter 3

of a surprise Indian attack or NCA decapitation. With the development of battlefield nuclear capabilities such as the Nasr and cruise missiles, the lack of negative controls as the delivery systems are fielded may be deliberately designed to enhance deterrence. This enhanced deterrent power, however, comes at the cost of the elevated risk of inadvertent use. More specifically, because of the emphasis on rapid assembly, deployment, and potential use under chaotic conditions, it is unlikely that Pakistan has the desire to develop or deploy robust digital Permissive Action Link (PAL) capabilities designed to prevent unauthorized or accidental use.108 In 2002, Kidwai claimed that Pakistan’s nuclear weapons did not have PALs, instead relying on procedural negative controls and component separation to ensure safety.109 But in 2004, Dr. Samar Mubarakmand, one of Pakistan’s top nuclear scientists, incredibly declared that Pakistan had fully embedded PALs into its nuclear weapons at the manufacturing stage, rendering unauthorized or accidental use impossible.110 Kidwai now insists that Pakistan employs “ ‘Pak-­PALs,’ a domestic version of the American system” comprised of a twelve-­digit alphanumeric code.111 Does Pakistan employ PALs and, if so, what strength would be consistent with their known capabilities and posture? Given Pakistan’s disassembled maintenance procedures and the compulsions of its asymmetric escalation posture, all the publicly available evidence suggests that “Pak-­ PALs” are likely weak, bypassable controls that allow for the rapid assembly and release of nuclear weapons if necessary. First, it is unclear what a PAL for a disassembled warhead might look like, because modern PALs are integral to the design of, and digitally integrated into, fully assembled nuclear weapons with limited-­try features. Older U.S. PALs (Category A/B) were heavy-­duty combination or electromechanical locks on some warhead component; these safeguards are most likely given the technical state of the Pakistani arsenal.112 These rudimentary PALs could perform a variety of 108. The United States has chosen not to transfer modern PAL technology to Pakistan, and Pakistan would likely never accept such technology for fear of “kill switches” that might enable the United States to neutralize Pakistan’s nuclear weapons. 109. Cotta-­Ramusino and Martellini, Nuclear Safety, Nuclear Stability, and Nuclear Strategy in Pakistan. 110. Dr. Samar Mubarakmand Interview, “Capital Talk Special,” Geo T.V. (Pakistan), May 3, 2004. Transcript available at http://www.pakdef.info/forum/showthread.php?t=9214. 111. David E. Sanger, “Obama’s Worst Pakistan Nightmare,” New York Times Magazine, January 11, 2009, p. 32; Kidwai in Moreau, “Pakistan’s Nukes.” Kidwai is sufficiently vague to allow for several possibilities, e.g., three components of a weapon could be secured by 4-­digit codes, so all 12 digits would be necessary to assemble a fully functional weapon, or one critical component (i.e., 5 the pit) could be secured by a 12-­digit code. 112. Ashley Tellis, “US-­Pakistan Relations: Assassination, Instability, and the Future of US Policy,” Hearing of the Middle East and South Asia Subcommittee of the House Foreign Affairs Committee, Federal News Service, January 16, 2008; for a primer on PAL

Pakistan 89

functions, such as locking one or more subcomponents of the disassembled warhead, physically blocking the fusing space or warhead assembly, or preventing the arming circuit from closing. All, however, “could theoretically be bypassed.”113 Thus, if Pakistan does indeed maintain warheads in a disassembled state, it is unlikely to have weapons equipped with fully embedded modern PALs. Pakistan most likely employs only weak negative controls on some warhead subcomponent(s), circumventable by design for swift assembly. If and when Pakistan develops closed-­pit designs and stores fully assembled warheads, it is unclear how PALs would be employed and whether they would be removed before, for example, a battlefield system such as Nasr was put into the field. Second, the credibility of the asymmetric escalation posture depends on theater commanders’ ability to quickly release nuclear weapons, especially its battlefield weapons such as the Nasr, in potentially chaotic circumstances, when communication with the NCA might be impossible. Brig. Gen. (Ret.) Khan writes that Pakistan’s “negative control technology provides only a marginal degree of safety if the troops managing a particular weapon are determined to act before they are destroyed.”114 A weak, bypassable PAL that would not definitively prohibit commanders from releasing nuclear weapons without authorization is thus also most consistent with the credibility requirements of Pakistan’s posture. It is therefore likely that Pakistan’s negative controls are lax enough to allow a technical team on a base or in theater to jury-­rig a nuclear weapon for use even without appropriate authorization from the NCA. Whatever arrangement Pakistan employs, Vice Admiral (Ret.) Vijay Shankar, a former Indian Strategic Force Command head, states that India’s security managers “assume there is nothing preventing these chaps from releasing the weapons once they take custody of them.”115 In order to make its asymmetric escalation posture credible, Pakistan seems to be augmenting positive control over its nuclear weapons by developing weak negative controls, or removing them altogether. Officers therefore have the physical ability to assemble and release nuclear weapons

technology, see Peter Stein and Peter Feaver, Assuring Control of Nuclear Weapons: The Evolution of Permissive Action Links, CSIA Occasional Paper No. 2 (Cambridge: CSIA Publications, 1987); and Donald R. Cotter, “Peacetime Operations: Safety and Security” in Ashton B. Carter, John D. Steinbruner, and Charles A. Zraket, eds., Managing Nuclear Operations (Washington, DC: Brookings Press, 1987) pp. 46–­51. 113. Cotter, “Peacetime Operations: Safety and Security,” p. 49. 114. Khan, “Challenges to Nuclear Stability in South Asia,” p. 68. 115. Author interview with Vice Admiral (Ret.) Vijay Shankar, New York City, July 15, 2011.

90

Chapter 3

should they deem it necessary, regardless of whether they are vested with the authority to do so. Brig. Gen. (Ret.) Khan concedes that “although the chain of command is clearly spelled out under all military contingencies, in the event of a command breakdown, a theater commander, seeing the opponent’s forces marching into his area of responsibility, would be hard-­ pressed to stand by and take no action.”116 Whatever negative controls exist to ensure the security and safety of the arsenal may be sufficient during peacetime, but are likely circumventable, by design, for deterrence purposes in a crisis or conflict situation with India. The key point is that, since 1998, Pakistan has been able to achieve a direct deterrent capability against India—­without reliance on the United States—­with the adoption of an asymmetric escalation posture. Unlike during the Brasstacks and Compound Crisis episodes where India may have been an ancillary target of Pakistani nuclear signaling, Delhi has been the direct target of Pakistani nuclear threats since 1998.

Explaining Pakistan’s Nuclear Posture Phase II: Asymmetric Escalation My theoretical framework suggests that, after American abandonment around 1991, Pakistan no longer had the option to maintain a catalytic posture. Instead, left alone to face a conventionally superior proximate offensive threat—­India—­Pakistan would have no choice but to adopt a first-­use asymmetric escalation posture to offset India’s conventional superiority. Indeed, finding itself in one of the most binding security situations possible, the theory anticipates that Pakistan would be forced by security pressures to adopt this posture regardless of the costs associated with it (financial, procedural, civil-­military, etc.). There is little dispute that Pakistan finds itself facing a primary adversary that can pose an existential proximate offensive threat to Pakistan’s survival. India’s conventional superiority improved in the 1990s, with crude quantitative force ratios varying between 2.00:1 to 3.33:1 in India’s favor,117 and materiel of roughly equivalent quality. Further, the likely terrain over which a major conventional war would be fought, the plains and deserts of Punjab, Rajasthan, and Sindh, is easily passable and conducive to the rapid maneuver warfare that India’s Strike Corps are designed to execute. Indeed, Pakistan found itself—­and believes itself to be—­in the same strategic position as NATO forces on the Central Front in Europe. Pakistan faces a nuclear-­armed convention116. Khan, “Challenges to Nuclear Stability in South Asia,” pp. 67–­68. 117. See Kapur, Dangerous Deterrent, pp. 24–­25.

Pakistan 91

ally superior adversary that poses an existential land threat to the state, particularly since the “core” of Pakistan, Punjab, is extremely vulnerable to Indian land conventional power. As such, facing a conventionally superior proximate offensive threat on its own, my theoretical framework predicts that Pakistan would be locked into selecting an asymmetric escalation posture to offset India’s power advantages. This shift should have begun as soon as Pakistan perceived American abandonment, even though the adoption of the full posture might have taken some time to occur due to the lag in developing the requisite capabilities to do so. This decision calculus is straightforward and demonstrates the power of extremely binding security constraints in international politics. In this case, and as noted earlier, the realist, strategic culture, and technological determinism approaches, would also predict that Pakistan should adopt an asymmetric escalation posture as well. If ever there was a case of overdetermination, it is of an army-­led Pakistan that perceives a persistent existential threat from a conventionally superior nuclear-­armed India adopting an asymmetric escalation nuclear posture. However, only optimization theory predicts a shift in posture at this particular point and not earlier, since America’s abandonment of Pakistan created the conditions where Pakistan would have to face India’s conventional superiority alone, triggering a shift to an asymmetric escalation posture. That is, whereas the realist and strategic culture theories would predict that Pakistan should have adopted an asymmetric escalation posture throughout, optimization theory suggests that only after American abandonment would Pakistan even though it took some consider an asymmetric escalation posture—­ time to develop and was midwifed by India’s own nuclear tests in 1998, which freed Pakistan to test as well. The technological determinism theory would also predict an asymmetric escalation posture evolving over time, but through an evolutionary process, not in the abrupt manner in which Pakistan shifted from a catalytic posture to an asymmetric escalation posture without seeking assured retaliatory capabilities first. Indeed, even in 1998, Pakistan did not have the capabilities to credibly and fully adopt an asymmetric escalation posture but it took advantage of the political cover imparted by India’s test to effect the shift, calculating that it would likely never have such a window again. From a technological perspective, Pakistan might have preferred to wait until its delivery capabilities were more mature before shifting, but the BJP’s tests in May 1998 left Pakistan with little choice but to array whatever capabilities it had available toward an asymmetric escalation posture. Nevertheless, the technological determinism theory is not wholly inconsistent with Pakistan’s trajectory, since Pakistan fields whatever it can to deter India’s conventional superiority now that it lacks an American patron.

92

Chapter 3

Summary Pakistan presently finds itself in one of the most precarious security situations on earth. Born into and out of war, it has always perceived an existential threat from its larger neighbor India. Those fears were crystallized in the 1971 war which amputated Bangladesh from post-­independence Pakistan. Since 1971, Pakistan has been on a desperate quest to acquire and operationalize a nuclear weapons capability to deter Indian conventional power. As its uranium enrichment program was reaching critical thresholds to enable the weaponization of the program, Pakistan relied on a catalytic nuclear posture which used the credible threat of nuclear escalation to compel its then-­patron, the United States, to intercede on its behalf in crises with India. I argue that this was a deliberate political nuclear posture implemented by Pakistan when both it and India were covert nuclear states. It was a calculated selection optimized for the constraints and opportunities besetting Pakistan—­it could exploit American interests in the region while at the same time avoid provoking India from becoming an overt nuclear power. Only optimization theory predicts that Pakistan would have selected a catalytic posture in this period, since both a strictly realist and strategic culture approach would suggest that Pakistan—­and the army in particular—­would have adopted an asymmetric escalation posture to directly deter a perceived existential threat. A catalytic posture leaves something to chance since it relies on probabilistic third-­party intervention. Realists, and most army cultures, are disinclined to leave the survival of a nation to chance or to a third party. Yet, that is precisely what Pakistan did in this period. American behavior conformed to Pakistani expectations and the United States did intervene on Pakistan’s behalf on two important occasions. Pakistan’s case demonstrates how the structural variables shape a state’s choice of nuclear posture and how an exogenous tightening of those constraints—­in this case the loss of an American patron and India’s nuclear tests—­resulted in a shift in strategies from a catalytic posture to a more direct and aggressive asymmetric escalation posture. After American abandonment in the early 1990s, Pakistan was forced to prepare for a shift in nuclear postures, seeking missile delivery capabilities to adopt an asymmetric escalation posture to deter Indian conventional superiority. When the BJP tested in May 1998, Pakistan could exploit the political cover imparted by India’s tests to follow suit and make a significant advance in its quest for a credible asymmetric escalation posture. This left Nawaz Sharif little choice but to bring Pakistan’s nuclear weapons out of the closet to complement the delivery capabilities it was already operationalizing. Upon doing so, Pakistan began incorporating nuclear weapons into its military posture and doctrine and adopted a nuclear posture modeled on NATO’s

Pakistan 93

during the Cold War—­an asymmetric escalation posture that threatens the first use of nuclear weapons in the event of a conventional attack by India. The exact threshold that would trigger the release of nuclear weapons on advancing Indian forces is left intentionally ambiguous, but Pakistan attempts to credibly enforce a line several tens of kilometers across the international border. In order to credibly enforce an early red-­line and the first use of nuclear weapons, even under potentially chaotic communication scenarios and the inevitable fog of war, Pakistan has to adopt stewardship procedures that allow for the quick assembly, deployment, and release of nuclear weapons without significant physical impediments for the officers who would be charged with each task in case higher-­level authorities are unreachable. This, of course, generates serious risks of unauthorized or accidental use of nuclear weapons and raises the worrying specter of nuclear theft or terrorism as Pakistan faces the challenge of securing an arsenal that must be deployed in such a way as to be quickly usable by its armed forces. However, my theory suggests that the compulsions of Pakistan’s nuclear posture mean that the security of the arsenal will always be secondary to the usability of the arsenal in an India contingency. Nevertheless, as this chapter shows, the asymmetric escalation posture can be a double-­edged sword: while states may believe it is necessary to deter a conventionally superior adversary, it can pose daunting management challenges that increase the risk of inadvertent nuclear use or nuclear theft.

Chapter Four

India

In this chapter, I explore India’s nuclear posture. Like China’s, which I discuss in the next chapter, I classify India’s nuclear posture as one of assured retaliation. There have been various dramatic moments in India’s nuclear weapons history that were often driven by domestic political considerations, most notably its nuclear tests in May of 1974 and 1998.1 Nonetheless, the capabilities, envisioned use, and command-­and-­control apparatus that Delhi has erected for its nuclear forces have been persistent and consistent with an assured retaliation posture since 1974. Indeed, a former commander-­in-­ chief of India’s Strategic Force Command, Vice Admiral (Ret.) Vijay Shankar, notes that the three pillars of India’s nuclear posture are and always have been: no first use, “assured massive retaliation” and, third, “under no condition would the weapon be conventionalized.”2 India’s posture has been, and will likely continue to be, entirely oriented for retaliation. The reason for this, as illustrated by optimization theory, is that India is in a relatively secure position but with highly assertive civil-­military relations, driving it toward an assured retaliation nuclear posture that emphasizes firm civilian control over the arsenal.

1. See Perkovich, India’s Nuclear Bomb; Sagan, Inside Nuclear South Asia. 2. Author interview with Vijay Shankar, New York City, July 15, 2011; also see Vijay Shankar, “A Covenant Sans Sword,” unpublished paper, July 2011.

94

india 95

Consistency of Assured Retaliation The full history of India’s nuclear weapons program is beyond the scope of this book, but I classify India as having adopted a nuclear posture of assured retaliation against its adversaries, Pakistan and China, from 1974 through the present.3 In terms of capabilities, although India presently lacks true strategic reach against China’s eastern seaboard, it nevertheless has and will increasingly have the ability to inflict significant damage against Chinese targets. There have certainly been variations in the credibility and speed with which India could retaliate with nuclear weapons over that time period, but India possessed the technical capacity to develop and deliver a plutonium fission weapon from its sizeable reactor capacity after May 18, 1974. On that day, India’s then-­Prime Minister Indira Gandhi approved the test of a “peaceful” nuclear explosive device (PNE), with a claimed yield of 12kT.4 Though the actual yield may have been considerably smaller, perhaps 2–­4kT,5 India had demonstrated competency in the scientific and technical skills necessary to produce nuclear fission for explosive purposes. This device itself was too big and unstable to be used as a deliverable nuclear weapon.6 Indeed, the decision to test the PNE involved only a handful of people, none of whom were in the military or the defense ministry. This suggests that there were no immediate plans to fully weaponize India’s nuclear capability. But India’s capacity to detonate a fission device was now transparent to the world, though everything else about India’s nuclear program remained opaque. 3. For works on India’s nuclear weapons history see Perkovich, India’s Nuclear Bomb; Tellis, India’s Emerging Nuclear Posture; Sumit Ganguly, “The Pathway to Pokhran II: The Prospects and Sources of New Delhi’s Nuclear Weapons Program,” International Security, vol. 23, no. 4 (Spring 1999), pp. 148–­77; Singh, Nuclear India; K. Subrahmanyam, “India and the International Nuclear Order,” in SarDesai and Thomas, eds., Nuclear India in the Twenty-­First Century, pp. 63–­84; Paul, “India, the International System and Nuclear Weapons,” in SarDesai and Thomas, Nuclear India, pp. 85–­104; Jaswant Singh, “Against Nuclear Apartheid,” Foreign Affairs, vol. 77, no. 5 (September/October 1998), pp. 41–­52; Stephen Cohen, “Nuclear Weapons and Conflict in South Asia,” Paper Presented at Harvard/MIT Transnational Security Project, November 23, 1998, available at http://www.brook.edu /views/cohens19981123.html; Raj Chengappa, Weapons of Peace (Delhi: Harper Collins 2000); Jacques Hymans, “Why Do States Acquire Nuclear Weapons?” in SarDesai and Thomas, Nuclear India, pp. 139–­60; Hymans, Psychology of Nuclear Proliferation, chap­ ter 7; Rajesh Rajagopalan, “India: The Logic of Assured Retaliation,” in Alagappa, ed., The Long Shadow, pp. 188–­214. 4. The PNE used twelve lenses to compress roughly 6kg of plutonium in a device that was 1.25m in diameter and weighed in at 1400kg. See Chengappa, Weapons of Peace, p. 182. 5. Tellis, India’s Emerging Nuclear Posture, pp. 196–­98. 6. For example, India’s scientists apparently had trouble stabilizing the polonium-­ beryllium neutron initiator.

96

Chapter 4

Nevertheless, the test provided India with critical design-­level data that could allow it to assemble rudimentary nuclear weapons if necessary. A U.S. National Security Council memo dated May 30, 1974 judged at the time that India could “use its 38 light Canberra bombers as delivery vehicles  .  .  . the IAF has been experimenting recently with heavy payloads akin to those of a crude nuclear bomb” and could reach critical targets in all of Pakistan and parts of China.7 The test thus made India a latent nuclear power that could constitute a nuclear weapon relatively quickly from a decision to do so. George Perkovich, in his authoritative history of India’s nuclear weapons history, writes that after the PNE, India “possessed the technological capacity to develop nuclear weapons if security interests required it, but it would refrain from exercising this military option” unless absolutely necessary.8 But in demonstrating the capability to assemble a nuclear fission device that could be used as a weapon and delivered by modified aircraft in its inventory, India demonstrated the ability to potentially retaliate against China (and Pakistan) in the event of a nuclear attack, albeit with a time delay. India would play fast and loose over whether it technically possessed a nuclear weapon—­since it did not keep any fully assembled or any stockpile—­but other states believed it had the capability to assemble and deliver one within weeks of a decision to do so after the PNE.9 I therefore classify India as having adopted an assured retaliation posture since May 1974. Even though its delay in retaliating may have been significant, it had the necessary infrastructure and capability to ensure survivability and guarantee a retaliatory strike against both Pakistan and parts of China if necessary. Critically, its adversaries believed this to be the case. Rajiv Gandhi continued India’s march toward an overt nuclear capability in 1988–­89. In response to reports that Pakistan was close to a weapons capability, Gandhi discretely ordered India’s Atomic Energy Commission (AEC) and Defence Research & Development Organization (DRDO) to “weaponize” India’s nuclear capability so that India would have the components necessary to assemble nuclear weapons more readily available.10 Jacques Hymans reports that the order was given directly to AEC and DRDO chairmen P. K. Iyengar and V. S. Arunachalam to begin “design, testing and production of advanced detonators, ruggedized high volt trigger systems, interface engineering, systems engineering and systems integration

7. Sidney Sober, Acting Chairman NSC Interdepartmental Group for the Near East and South Asia, “Indian Nuclear Development,” Declassified National Security Study Memorandum 156, May 30, 1974, p. 18. 8. Perkovich, India’s Nuclear Bomb, p. 189. 9. Ibid., pp. 267–­68. 10. Ibid.

india 97

to military specifications.”11 With respect to delivery vehicles, in 1983 Indira Gandhi revived and then Rajiv Gandhi accelerated India’s Integrated Guided Missile Development Program (IGMDP) which designed India’s indigenous ballistic missile families Prithvi and Agni, which began flight-­ testing in 1988.12 India’s nuclear-­capable aircraft by this point included the Jaguar as well as French Mirages. Studies of nuclear proliferation often date Indian weaponization to 1988,13 the year in which India had readily assembleable nuclear weapons—­but still kept them in a de-­mated and dispersed state out of the hands of the military—­thereby compressing India’s time to retaliate against a nuclear strike to hours or days rather than weeks. The targeted time to retaliation from an order to do so was reported to be 72 hours. But the basic posture of assured, if delayed and ragged, retaliation persisted. This state of affairs continued with India gradually improving its delivery capabilities through the 1990s. India developed a dedicated aircraft vector, gradually operationalized the short-­range liquid fuel nuclear-­capable Prithvi missile by raising its first strategic missile group in 1995 (based in Secunderabad in South India, some distance from the actual missiles which were kept unfueled in storage bunkers near the India-­Pakistan border14), and continued to test the Agni missile family.15 Although India theoretically could have delivered a nuclear device with a transport aircraft (unreliably and probably only in a one-­way mission) prior to 1995, Gaurav Kampani argues that the air force only reliably developed the ability to deliver a nuclear weapon with Mirage fighter aircraft in the mid-­1990s, around the same time that the land-­based vectors were coming online.16 The lack of military integration in India’s nuclear posture through this period was intentional on behalf of the civilian political leadership. As one member of Rajiv Gandhi’s PMO notes, civilian leaders believed that India reaped a 11. See Hymans, Psychology of Nuclear Proliferation, p. 190; and “Joint Statement by Department of Atomic Energy and Defence Research and Development Organisation,” New Delhi, May 17, 1998. Available at http://www.fas.org/news/india/1998/05/980500 -drdo.htm.; also Chengappa, Weapons of Peace, pp. 334–­35. This point was confirmed by author discussion with a member of Rajiv Gandhi’s PMO staff who claimed that the decision to move toward weaponization was a gradual one through the 1980s but required authorization of very specific steps by the prime minister in order to proceed, which only occurred in 1988/1989. 12. See Vipin Narang, “Pride and Prejudice and Prithvis: Strategic Weapons Behavior in South Asia” in Sagan, ed., Inside Nuclear South Asia. 13. Alexander H. Montgomery and Scott D. Sagan, “The Perils of Predicting Proliferation,” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009), esp. pp. 307–­10. 14. See Tellis, India’s Emerging Nuclear Posture, pp. 371–­72. 15. See Narang, “Pride and Prejudice and Prithvis.” 16. Gaurav Kampani, “When Does a Nuclear Possessor State Become a Nuclear Weapons Power?: Lifting India’s Veil of Nuclear Ambiguity (1989–­1998),” unpublished manuscript.

98

Chapter 4

broad deterrent benefit from a highly recessed nuclear capability, or “force in being,” without having to entrust military end users, of which India’s political leaders were always wary, with nuclear devices.17 Domestic political pressure from the right-­wing Bharatiya Janata Party (BJP) continued to mount on Congress, with one of the key pressure points being “exercising India’s nuclear option,” that is, overtly testing nuclear weapons. In 1995, Congress Prime Minister Narasimha Rao came within a hairs-­breadth of testing before American pressure forced him to walk away from the precipice.18 When the BJP ephemerally formed a thirteen-­day government in 1996, its Prime Minister Atal Bihari Vajpayee also ordered tests but had to pull back when his government collapsed. In March 1998, the BJP established a relatively stable coalition government and, in May, Vajpayee took India across the explicit nuclear weapons threshold, citing everything from China to Pakistan to a hypocritical international community by way of justification.19 All of these domestic political machinations made for high international drama, with the American intelligence community lambasted for missing India’s impending nuclear tests. However, in terms of capabilities, the tests represented nothing more than a victory shot for a nuclear weapons program that was already relatively advanced. Indeed, India’s scientists had little doubt that the four fission devices would successfully detonate, which they did, but they were unable to successfully detonate the device that would have dramatically improved India’s retaliatory power: a boosted fission device,20 let alone a true thermonuclear weapon. Rather than marking a discontinuity, India’s May 1998 tests represented substantial continuity in India’s nuclear posture. Since India’s May 1998 tests, India’s capabilities have continued to evolve. India has raised additional Prithvi21 and Agni missile groups as well as tested nuclear-­capable cruise missiles (BrahMos) and a short-­range 17. Author interview with former member of Rajiv Gandhi’s Prime Minister’s Office, New Delhi, December 2012. 18. Hymans, Psychology of Nuclear Proliferation, p. 193; This was confirmed by K. Santhanam, who was in charge of the aborted test, in an interview “India Aborted N-­test Last Minute in 1994,” Timesnow, April 1, 2008, available at http://timesnow.tv/NewsDtls .aspx?NewsID=6911; also see K. Subrahmanyam, “Narasimha Rao and the Bomb,” Strategic Analysis, vol. 28, no. 4 (2004), pp. 593–­95; K. Subrahmanyam, “From Indira to Gowda It Was Bomb All the Way,” The Times of India, April 17, 2000; Saba Naqvi Bhaumik, “Blast from a Far Past,” Outlook India, May 19, 2008. 19. Prime Minister A. B. Vajpayee, “Indian’s Letter to Clinton on the Nuclear Testing,” New York Times, May 13, 1998, p.A14. 20. Data gathered from the test may nevertheless have been sufficient to allow India to develop and field boosted fission devices. 21. The liquid-­fueled Prithvi missiles (at least the Prithvi I) are now to be replaced by solid fuel 150km Prahaar missiles. See “Prithvi Missiles to be Replaced by More Capable Prahaar: DRDO,” Times of India, July 1, 2013.

india 99

nuclear-­ capable tactical missile, Prahaar, and is aiming for a ship and submarine-­launched capability in the future to enhance survivability. Vice Admiral (Ret.) Shankar, who commanded India’s Strategic Force Command from 2008 to 2009, has additionally noted, however, that the BrahMos has now been assigned an entirely conventional role and that “the Prithvi will go . . . no one wants it. At least I did not want it . . . the range makes no sense, especially with the solid-­fuel Agni family in the arsenal.”22 The Agni will ultimately form the “backbone” of India’s strategic deterrent capability on the land-­based vector since it can hit a range of targets from central India without having to be deployed close to the border as with the land-­based Prithvi.23 Indeed, the extended-­range flight tests of the Agni III through V, and potentially an ICBM-­range Agni VI, will ultimately equip India with strategic reach against high-­value targets on China’s east coast, including Beijing and Shanghai. As of 2010, Robert Norris and Hans Kristensen judged: “Fighter bombers constitute the only fully operational leg, backed by short range ballistic missiles. But significant additions to the arsenal are around the corner: Two ballistic missiles are being deployed with the country’s army, two longer-­range types are under development, [and] the first of a class of nuclear-­powered ballistic missile submarines has been launched.”24 The Strategic Force Command has presently judged the Agni I to be a fully operational system in addition to the aircraft vector.25 Norris and Kristensen further estimated in 2012 India had produced 80–­ 100 plutonium warheads, with sufficient weapons-­grade plutonium for a slightly larger arsenal that they assess has not yet been converted into warheads.26 India’s DRDO claims to be in the development phase for a host of capabilities, including MIRVs and missile defenses.27 However, at the time of writing, there has been no political sanction for the incorporation of any of these capabilities into India’s nuclear posture, which remains one of increasingly sturdy but clearly assured retaliation. After testing nuclear weapons, India’s BJP government decided that it needed to develop a nuclear doctrine, what it terms “credible minimum 22. Author interview with Vijay Shankar, New York City, July 15, 2011. The de-­emphasis of the Prithvi and cruise missiles for nuclear roles has the significant benefit of stabilizing crisis signaling since all land capabilities would be assigned single roles—­either conventional or nuclear—­thereby removing a significant source of misperception or miscalculation in crises with Pakistan. 23. Author interview with Vijay Shankar, New York City, July 15, 2011. 24. See Robert Norris and Hans Kristensen, “India’s Nuclear Forces, 2012,” Bulletin of the Atomic Scientists, , vol. 68, no. 4 (2012), p. 96. 25. Author interview with Vijay Shankar, New York City, July 15, 2011. 26. Norris and Kristensen, “India’s Nuclear Forces, 2010,” Bulletin of the Atomic Scientists, vol. 66, no. 5 (2010), p. 76. 27. See Christopher Clary and Vipin Narang, “Capability without Strategy,” Indian Express, May 22, 2012.

100

Chapter 4

deterrence,” which it drafted in 1999 and formally adopted in 2003.28 The salient features of India’s declaratory doctrine, which was released in draft form in 1999 and then as a series of eight points in 2003, are (1) its overriding objective to “deter the use and threat of use of nuclear weapons” by maintaining an “adequate retaliatory capability should deterrence fail,” and (2) a declared commitment to the no-­first-­use of nuclear weapons.29 The goal of India’s declaratory doctrine is thus an assured retaliation capability with the explicit corollary that it will not be the first to initiate a nuclear strike. Brajesh Mishra, who drove India’s nuclear posture developments from 1998 to 2004 as National Security Advisor, states that India’s nuclear posture is “fundamentally and primarily based on no first use” and that “we aim to have enough to absorb an attack and strike back.”30 Vice Admiral (Ret.) Shankar concurred that India does not have the capabilities to have a launch on warning (LoW) or launch under attack (LuA) posture; instead it has a “launch after hit” posture, meaning it will absorb a nuclear hit before the release of nuclear weapons in retaliation would be authorized.31 This is an explicitly assured retaliation doctrine. Do the stewardship and command-­and-­control features of India’s nuclear arsenal since 1998 align with India’s capabilities and declaratory doctrine? I show that India’s operational doctrine is largely consistent with its declaratory doctrine, geared entirely toward assured retaliation, though certainly with less time delay for retaliation than was the case prior to 1998. Ashley Tellis, in his authoritative 2001 study on Indian nuclear doctrine, described 28. Government of India, “Draft Report of National Security Advisory Board on Indian Nuclear Doctrine,” August 17, 1999. Available at http://www.indianembassy.org/policy /CTBT/nuclear_doctrine_aug_17_1999.html. The contours and motivations of key components of the doctrine—­partly modeled on Western doctrines, with bureaucratic pressures to appease all three service arms, etc.—­have been discussed elsewhere. See most notably Waheguru Pal Singh Sidhu, “India’s Nuclear Use Doctrine,” in Lavoy, Sagan, and Wirtz, Planning the Unthinkable, pp. 125–­57. For my purposes, declaratory doctrine is mostly irrelevant except as a baseline to determine whether it is consistent with operational doctrine, or posture. So though the declaratory doctrine aspires for a triad, India is only now embarking on developing a submarine leg for its nuclear weapons; and though the Indian doctrine models the United States by leaving open the possibility of retaliating with nuclear weapons against a chemical or biological attack and by claiming that India might retaliate with nuclear weapons if its armed forces were attacked by nuclear weapons on foreign soil, neither are believed to be terribly credible threats. See Scott D. Sagan, “The Evolution of Pakistani and Indian Nuclear Doctrines” in Sagan, ed., Inside Nuclear South Asia. Discussions with high-­ level Indian government officials in Delhi in July 2010 confirmed that it would be difficult for India to justify retaliating with nuclear weapons in either of the scenarios listed here; the only clear-­cut case where India would use nuclear weapons is in retaliation for a nuclear attack on Indian soil. 29. “Draft Report of National Security Advisory Board on Indian Nuclear Doctrine.” 30. Author interview with Brajesh Mishra, New Delhi, January 6, 2011. 31. Author interview with Vijay Shankar, New York City, July 15, 2011.

india 101

the three key features of India’s posture: limited in size, separated in disposition, and centralized in control.32 His description of this arrangement at the time was: The weapons and delivery systems are developed and produced, with key subcomponents maintained under civilian custody, but these assets are not deployed in any way that enables the prompt conduct of nuclear operations. Such assets are, in fact, sequestered and covertly maintained in distributed form, with different custodians exercising strict stewardship over the components entrusted to them for safekeeping.33

While there has certainly been evolution in the disposition of the arsenal since Tellis’s study, especially since the 2003 raising of an official Strategic Forces Command and with some systems now likely minutes away from being readied,34 the pillars of an assured retaliation posture still persist: no first use, centralized control, and no conventionalization of nuclear weapons.35 In short, India has maintained an assured retaliation posture for decades with increasingly compressed time to launch. India continues to maintain its land-­and air-­based nuclear assets in various forms of a disassembly and firmly under civilian custody and control.36 In practice, this meant that at least through the mid-­2000s the Department of Atomic Energy would steward the fissile pits while DRDO stewarded the non-­fissile components of the physics package; delivery vehicles such as nuclear-­capable but technically dual-­use aircraft and ballistic missiles are maintained in various operational forms by the military end users (after 2003, army, air force, and eventually navy assets that would be tasked to the Strategic Forces Command when necessary). To enhance survivability, each of the civilian agencies was—­and is still—­believed to disperse reserves of their respective subcomponents over several secret locations to ensure that a fully constitutable nuclear capability would survive a first strike as well as “iterative” attempts to disarm India’s capability.37 For a nation that otherwise leaks state secrets like a sieve, India has seemingly been able to keep the locations and procedures of how it administers its dispersed assets extremely well-­guarded. There are several possible ways 32. Tellis, India’s Emerging Nuclear Posture, esp. pp. 374–­466. 33. Ibid., p. 367. 34. See Vipin Narang, “Five Myths about India’s Nuclear Posture,” Washington Quarterly, vol. 36, no. 3 (Summer 2013), pp. 143–­57. 35. Author interviews with high-­level PMO official, New Delhi, January 3, 2011, and Vijay Shankar, New York City, July 15, 2011. 36. See Tellis, India’s Emerging Nuclear Posture, chapter 4. 37. Author discussions with Rear Admiral (Ret.) Raja Menon, Bharat Karnad, Manoj Joshi, Brig. Gen. (Ret.) Gurmeet Kanwal in July 2006 and August 2007.

102

Chapter 4

this arrangement could obtain, with weapons subcomponents in various states of disassembly and co-­location to optimize survivability and rapidity of deployment if necessary.38 And as noted earlier, in some cases, there was further geographic separation of the military’s missile brigade from the actual missiles themselves, which increased the time required to fully deploy a useable nuclear asset, particularly if a crisis unfolded rapidly. Referring to the arrangement of separating and concealing subcomponents of the nu­ clear arsenal, Gen. (Ret.) K. Sundarji, one of the few Indian Army officers to think seriously about nuclear doctrine, wrote in 1997: The entire sequence of pre-­launch movements will take place after receiving the first strike . . . This process is not merely a question of needles in haystacks. It involves parts of many needles in many haystacks which may be brought together, within hours or days, to form full needles in many more different haystacks. Each haystack will be tenanted for a brief period before the actual launch. This will ensure an acceptable level of survival for the Indian missiles, even in the face of an all-­out nuclear first strike.39

This concept instituted a costly signal of a time lag—­believed to be anywhere from several minutes to a day depending on the scenario and the specific delivery vehicle in question40—­lending credibility to the posture’s retaliatory orientation. Through the early 2000s, India’s nuclear weapons simply were not in an operational state to be used either tactically or in a bolt-­from-­the-­blue strategic strike. More recently, former Indian Chief of Army Staff, Gen. S. Padmanabhan stated: “I do not see the weapon as a war-­fighting one. It’s like you make an expensive ornament for your wife. 38. See Tellis, pp. 409–­410 and Karnad, India’s Nuclear Policy, p. 99. The fundamental character of a de-­mated and disassembled arsenal with reserve components stewarded by DAE, DRDO, and the military was confirmed in discussions with high-­level PMO officials in India in July 2010, though there is increasing variation in this disposition across the arsenal. In the past, there appeared to be several alert levels that would have to be ordered by the PMO to first assemble the non-­fissile explosives package (though some may be preassembled), insert the pit (which would require interfacing DAE with DRDO officials), transfer the assembled warhead to the end users, mate the warhead, and then release the warhead. This appears to be changing as India moves to a more ready and canisterized force. 39. General Krishnaswamy Sundarji, “Prithvi in the Haystack,” India Today, June 30, 1997, p. 73. 40. See Karnad, India’s Nuclear Policy, p. 99. Also see Harsh V. Pant, “India’s Nuclear Doctrine and Command Structure: Implications for India and the World,” Comparative Strategy, vol. 24, no. 3 (2005), p. 286; and Harsh V. Pant, “India’s Nuclear Doctrine and Command Structure: Implications for Civil-­Military Relations in India,” Armed Forces & Society, vol. 33, no. 2 (January 2007), pp. 238–­64. A former Indian National Security Advisor cautioned against believing anything in the public domain about specific operationalizations, since there has been no authoritative discussion of these issues and they remain classified at India’s highest levels; personal correspondence with author August 12, 2010.

india 103

It may have taken years to make it, but she wears it once and keeps it in the vault. The nuclear weapon is also to keep in the basement.”41 But though there was a time lag, these stewardship procedures—­even with capabilities of limited numbers—­enhanced survivability and therefore provided sufficient guarantees that India could assure nuclear retaliation against an opponent that uses nuclear weapons against India.42 The tradeoff India confronts is between assertive control and rapidity of constitution of a useable nuclear asset. Throughout much of its nuclear history, India has chosen to privilege assertive control at the expense of the ability to swiftly constitute the bulk of its nuclear weapons. Tellis referred to component and organization dispersion as a “super-­PAL,” which was India’s cost-­effective solution to ensuring that nuclear weapons are not used except when so ordered by the prime minister.43 This arrangement seems to have evolved in the late 2000s under SFC coordination, however, with co-­location of assets and users being the standard operating procedure, but with significant procedural controls that continue to assert maximal civilian control over the arsenal. Since then, India’s force disposition seems to have undergone a significant evolution, with some portion of the nuclear arsenal managed at a higher state of readiness than the rest. However, they remain technically either de-­mated or disassembled and subject to centralized and assertive civilian control.44 All observable indicators are that India, while adhering to its posture of assured retaliation, has now increased the baseline readiness of at least a subset of its nuclear forces, if not all of them—­particularly as it marches toward a force consisting largely of so-­called encapsulated or canisterized systems in which the warhead is likely pre-­mated to the delivery vehicle and kept hermetically sealed for storage and transport (though there may be procedures that allow for the warhead to be mated in the field). This encapsulation is made possible by the fact that India’s ballistic missiles are essentially all now solid fuel. Encapsulation enhances missile longevity by protecting the solid fuel

41. General S. Padmanabhan quoted in “Pak Nuclear Threat Worried Only Me, says Ex-­ Army Chief,” Indian Express, February 3, 2004. 42. Of course, these procedures and civilian custody of nuclear weapons will be challenged if India ever deploys a ready sea deterrent. That capability is several years away, but if my theory is correct, India will devise a method to keep sea forces under firm civilian control (e.g., civilian representatives on SSBNs, etc.). 43. Tellis, India’s Emerging Nuclear Posture, p. 433. 44. In my series of interviews, it was clear that officials took the term “de-­mating” to have various meanings. It is possible that a subset of the Indian force, e.g., the Agni I, is effectively mated and ready excepting physical completion of warhead assembly with the rapid insertion of some subcomponent and the supplying of digital inputs that would complete the arming of the warhead. Author interview with high level PMO advisor, New Delhi, January 3, 2011.

104

Chapter 4

from the elements, but it also complicates our picture of India as managing its nuclear forces in a disassembled or de-­mated state. Though reserve components are likely stewarded by their respective civilian agencies—­nuclear pits with the Department of Atomic Energy, the explosives package with DRDO, the delivery vehicle with the SFC—­it seems increasingly likely that India has some subset of the force within minutes of readiness. Indeed, former senior civilian security officials and former SFC officers themselves suggest that some portion of the nuclear force is now kept at a much higher state of readiness. Some systems are capable of being operationalized and released within seconds or minutes once alerted—­not hours, as has been previously assumed. This is particularly true of those weapons designed for retaliation against Pakistan. For example, former officers in the Strategic Forces Command confirmed that a portion of the land-­based missiles are now maintained at very high levels of readiness in peacetime, and that at least some nuclear bombs for aircraft are co-­located with aircraft on bases and stored in underground bunkers for rapid mating if necessary. Other systems that would be necessary only in contingencies with longer lead times may be kept at lower states of readiness. Instead of maintaining its entire nuclear force in a disassembled and de-­mated state, India now appears to maintain its force at a range of readiness levels, with some systems almost fully ready during peacetime. Specifically, Brajesh Mishra notes that because India has a “greater ap­ prehension that Pakistan might use nuclear weapons [against us], not China . . . one can be sure that the 800-­km [Agni I ] missile [designed specifically for nuclear retaliation against Pakistan] would not take as much time . . . to operationalize.”45 Vice Admiral (Ret.) Shankar also indicates that India’s nuclear forces are now kept in a range of alert levels—­some that can be readied in minutes in order to retaliate against a “bolt out of the blue” attack (again, likely the solid-­fuel Agni I ), and others kept at lower states of readiness for contingencies for which there would be longer lead times.46 The higher readiness systems could be kept at a variety of states just short of a fully armed nuclear system, such as final warhead assembly requiring civilian representatives to quickly insert some component co-­located with the system, or some digital code being required to arm the warhead, or some combination of similar steps (e.g., “mated but unassembled or unarmed” systems). Because these warheads would be mated with solid-­fuel delivery systems, the land-­based ballistic missiles no longer require long fueling times and could thus be readied much more quickly than old liquid-­fuel systems. These could still meet the strict definition of

45. Author interview with Brajesh Mishra, New Delhi, January 6, 2011. 46. Author interview with Vijay Shankar, New York City, July 15, 2011.

india 105

disassembled and/or de-­mated systems, which most officials with whom I spoke confirmed India still relies upon, but would obviously be quicker to operationalize than was reported by Tellis in 2001. Dr. Avinash Chander, as head of the DRDO, publicly stated that it is India’s ultimate aim to deploy its entire strategic nuclear missile force in an “anytime-­anywhere” state: “We are working on canisterised systems that can launch from anywhere at any time . . . We are making much more agile, fast-­reacting, stable missiles so response can be within minutes.”47 The canisterization of India’s long-­ range missiles likely means that India will no longer keep its land-­based nuclear forces in disassembled or de-­mated states and would significantly compress the time to launch once appropriate alert levels were ordered. It is important to note that this evolution concerns India’s strategic missile force: there is no evidence that India is developing tactical nuclear weapons or has ever contemplated nuclear warfighting capabilities. However, even higher-­readiness strategic systems are and would still be subject to significant centralized control that would require direct communication from the Prime Minister’s Office (PMO) to the end users in order to release any weapon. Although India’s force disposition and stewardship procedures have evolved over the decades, the key permanent feature of India’s assured retaliation posture is that civilians not only maintain control over India’s nuclear forces, but they maintain custody of it. Nuclear assets would only be constituted, operationalized, and transferred to military end users on the orders of the Prime Minister, the de jure head of the Nuclear Command Authority (NCA). Thus, in peacetime and even in relatively intense crises, India’s nuclear arsenal is kept under firm civilian control, which minimizes the risk of unintentional use. This is markedly more assertive than other modern nuclear states, since in no other current nuclear regional power do civilians physically steward the nuclear arsenal (though in the early parts of the U.S. nuclear program, nuclear pits were stewarded by the Department of Energy; and Israel’s procedures are presently unknown). Only the Prime Minister “or the designated successor(s)” (in the event of decapitation) is vested with the authority to order component re-­constitution, movement, and release of nuclear weapons.48 Organizationally this is achieved by having the PMO be the only node at which the civilian and military end users, institutionalized in the Strategic Forces Command (SFC), charged with stewarding India’s nuclear subcomponents meet—­that is, DAE and DRDO cannot interface with the SFC without going through the PMO. Since the early 2000s, to ensure certain retaliation

47. Avinash Chander quoted in Shiv Aroor, “New Chief of India’s Military Research Complex Reveals Brave New Mandate,” India Today, July 3, 2013. 48. See “Draft Report of National Security Advisory Board on Indian Nuclear Doctrine.”

106

Chapter 4

more thought appears to have been given to developing alternate civilian chains of command that can function even after nuclear first strike.49 Vice Admiral (Ret.) Shankar argues that the cornerstone of India’s present command-­and-­control structure is “a clear and inviolable demarcation between custodian and controller.”50 The custodians are civilians represented by DRDO and DAE who could not assemble or arm a full nuclear system without express orders from the NCA; only once those orders are given would the custodians transfer nuclear assets to the SFC “controllers” to ready a nuclear system and prepare release. In fact, discussions with high-­level Indian government officials revealed that the PMO would have to directly order each stage of India’s alert ladder to the relevant agencies, of which there are multiple escalatory rungs—­probably four based on various sources at the time of writing—­before a release order would be issued to the military.51 Shankar further writes explicitly that there is “multiple redundancy and dual-­rule release authorization at every level,” which involves both “procedural and technical” negative controls that require express authority from the NCA.52 This highly assertive structure maximizes centralized civilian control over India’s nuclear assets.53 This stewardship arrangement also serves as an inherent negative control on the arsenal. It prevents unautho­ rized and accidental release of nuclear weapons under most circumstances because it necessarily involves the “two-­man-­plus” rule to deploy and release a fully usable asset. Therefore, regardless of their state of readiness, all nuclear systems are still presently subject to the two-­man-­plus rule at every alert and the release stage and would require direct inputs from the NCA; no SFC commander could release a nuclear system without input from the highest political authority, even if it is fully physically mated, according to Shankar.54 Former Chief of the Naval Staff, Admiral Arun Prakash, notes that the fundamental character of the three-­agency rule at every stage of nuclear operations has, 49. Prime Minister’s Office, “Cabinet Committee on Security Reviews Progress in Operationalizing India’s Nuclear Doctrine,” January 4, 2003. Available at http://pib.nic.in /archieve/lreleng/lyr2003/rjan2003/04012003/r040120033.html; Shyam Saran, “Is India’s Nuclear Deterrent Credible?” speech to the India Habitat Center, New Delhi, April 24, 2013. 50. Author interview with Vijay Shankar, New York City, July 15, 2011. 51. Discussions took place on author trip to India in July 2010. 52. Shankar, “A Covenant Sans Sword,” p. 5; and author interview, New York City, July 15, 2011. 53. Bharat Karnad in India’s Nuclear Policy suggests a more devolved command-­and-­ control structure, but provides few citations; in my discussions with Indian Army and political officials, I have been unable to corroborate the details Karnad suggests. 54. Shankar, “A Covenant Sans Sword,” p. 5; and author interview, New York City, July 15, 2011

india 107

to his knowledge, persisted even though some warheads may be managed in the digital equivalent of the “last screw” (or “last code”) state.55 Both Brajesh Mishra and PMO officials in the Manmohan Singh government have repeatedly emphasized the degree of negative control—­largely procedural, but increasingly technical—­on the arsenal exerted by the civilian authorities, even for systems that may be fully “canisterized” (encapsulated and deployed at a higher state of readiness) and that would only take several minutes to release. It is argued by all parties involved that, irrespective of the state of readiness of its arsenal, India’s political leadership aims to maintain maximal control over the nuclear arsenal even after it had absorbed a first strike.56 This is done for two reasons: (1) so that India’s political leadership would not cede authority and ability to release nuclear weapons to the military in a tense crisis situation; and (2) so that India’s limited number of assets would not be vulnerable to a disarming first strike once fully constituted. In fact, because the conventional military does not have possession of nuclear weapons, they have not regularly trained for the release of nuclear weapons and thus nuclear weapons “could not be used in the conduct of routine military operations” according to Tellis, but only in situations of extreme national emergency that required their retaliatory release after absorbing a first strike.57 Organizationally, the Strategic Forces Command is under the PMO’s chain of command and not integrated with the conventional military chain of command—­that is, an order to release nuclear weapons from any of the service chiefs or even the Chairman of the Chiefs of Staff Committee would have no authority. This is consistent with the principle of “no conventionalization” of the Indian nuclear force; the chains of command are separate and operations would be completely siloed.58 Admiral Prakash, who served as Chief of the Navy Staff and the Chairman of the Chiefs of Staff Committee, writes that: “The isolation of the armed forces from India’s strategic programmes has been so complete . . . [it is] a policy of segregation.”59 Thus, in India’s current command-­and-­control arrangement, it is likely that the SFC would only be given control of a nuclear system and its target in the ultimate step of a retaliatory strike, after the PMO had ordered their release following nuclear use against India or its forces.60 The targets 55. Author phone interview with Admiral Prakash, New Delhi, June 10, 2011. 56. Also, because of dual-­use capabilities in the region and India’s probable inability to discriminate between a nuclear and conventional launch in real time, Delhi could only confirm a nuclear launch after the warhead detonated. 57. Tellis, India’s Emerging Nuclear Posture, p. 370. 58. Author interview with Vijay Shankar, New York City, July 15, 2011. 59. Arun Prakash, “9 Minutes to Midnight,” Force Magazine, July 2012, p. 4. 60. Tellis, India’s Emerging Nuclear Posture, p. 452.

108

Chapter 4

that have been worked up with a dedicated Nuclear Planning Group in conjunction with the NCA reportedly are all strategic targets—­population centers and cities—­and India’s retaliation is entirely oriented for a punitive counterstrike following nuclear use on India.61 But Shankar indicates that the specific targeting package would be transmitted to the SFC commander only along the release order.62 Thus, the arrangement described by V. S. Arunachalam to Stephen Rosen in the late 1990s seems to be still largely accurate in spirit: “If New Delhi goes up in a mushroom cloud, a certain theater commander will go to a safe, open his book, and begin reading on page one, paragraph one, and will act step by step on the basis of what he reads.”63 Shankar elaborates that although this arrangement has evolved, it is in principle still true.64 Importantly though, speaking as a former SFC commander, he continues that this “stabilizing  .  .  . centralized control is the cornerstone of our command and control and this is a virtue. The idea is to not use these things.”65 As I will show in the next chapter, India’s nuclear posture resembles that of pre-­1990s China, where custody of nuclear weapons and the chain of command to move and release them ran through a dedicated service whose apex authority was the highest political authority in the nation.66 The critical test for the current management of India’s assured retaliation posture, which relies on civilian custody of subcomponents to maximize survivability and civilian control, will be how India manages the sea-­ based leg of its envisioned triad, which is aimed at imparting India a truly assured retaliation capability against both Pakistan and China. Though the sea leg is currently non-­operational, how India manages nuclear weapons for it will reveal just how strong its commitment to civilian control is. For surface ship-­launched ballistic missiles, the challenge is less daunting since components could be physically separated on the ship and stewarded by civilian representatives until authorization from the PMO was received. For a nuclear ballistic missile submarine (SSBN), however, India would have to devise procedures to either mate and arm its submarine launched ballistic missiles (SLBMs) while submerged on deterrent patrol with appropriate communications technologies and redundancies that prevented unauthorized or inadvertent launch, or mate the SLBMs in port at some predetermined alert level before sending the SSBN out on deterrent patrol, but

61. Author interview with Vijay Shankar, New York City, July 15, 2011. 62. Ibid. 63. V. S. Arunachalam quoted in Stephen P. Rosen, Societies and Military Power: India and its Armies (Ithaca: Cornell University Press, 1996), p. 252. 64. Author interview with Vijay Shankar, New York City, July 15, 2011. 65. Ibid. 66. Author interview with high level PMO advisor, New Delhi, January 3, 2011.

india 109

again with sufficient negative controls to prevent unintentional launch.67 Rear Admiral Raja Menon is confident that Delhi’s political masters will devise “bizarre suggestions”—­such as potentially limited deterrent patrols, or the use of multiple civilian representatives to release nuclear assets—­to maintain centralized control over nuclear assets, over the navy’s strong objections.68 However, Admiral Prakash flatly states that the Indian Navy will not allow civilians onboard submarines, raising questions about how this leg will be managed in what is inherently an encapsulated or canisterized nuclear delivery system (though there are ways to complete, for example, warhead assembly in the SLBM tubes while an SSBN is on a deterrent pa­ trol).69 As a navy commander of the Strategic Force Command, however, Shankar indicates that even if India is able to deploy SSBNs on continuous deterrent patrol, command and control would run through the SFC, not the navy, and therefore would be subject to the same centralized civilian authority as any of India’s currently encapsulated systems (e.g., the Agni family). That is, civilian and SFC inputs into arming an SLBM could be performed remotely, such that the captain of an SSBN has no physical ability to launch an SLBM absent centralized SFC inputs.70 But, even though the civilian relationship with the navy has always been better than with the army, how India manages its nuclear weapons at sea—­even though it will still only impart India with an assured retaliation posture and not be suitable for an asymmetric escalation posture—­will ultimately be a critical test of its current assured retaliation posture that overridingly emphasizes assertive civilian control.

Explaining the Endurance of Assured Retaliation in India In this section, I analyze the drivers of India’s nuclear posture. Specifically, I ask why has India chosen an assured retaliation posture and why has that posture endured? Despite the voluminous scholarly attention paid to India’s nuclear tests in 1974 and 1998, currently there is no satisfactory explanation for why India has consistently chosen an assured retaliation nuclear posture since 1974. I argue that this consistent and persistent choice of strategy is the product of three variables according to optimization theory. First, India did not believe it had a reliable third-­party patron—­not the Soviet Union, and certainly not the United States—­to avail itself of a catalytic posture. Second, while its security environment may have pushed India to 67. Tellis, India’s Emerging Nuclear Posture, p. 459. 68. Raja Menon in personal correspondence with author, August 3, 2009. 69. Author phone interview with Admiral Arun Prakash, New Delhi, June 10, 2011. 70. Author interview with Vijay Shankar, New York City, July 15, 2011.

110

Chapter 4

go nuclear, particularly after China’s nuclearization in 1964, that environment is sufficiently lax—­India has conventional superiority over Pakistan and natural geographic barriers buffering it against China’s land forces—­ that India does not have to adopt a first-­use posture to deter the conventional power of any adversary. Third, its civil-­military structures since independence have been highly assertive and strongly privilege centralized civilian control. As such, optimization theory predicts that India would adopt an assured retaliation posture to deter nuclear attack and coercion, which enables strong centralized control over nuclear assets. I demonstrate that optimization theory better captures the history and decisions of India’s nuclear posture than explanations based on structural realism, strategic culture, or technological determinism. The debate about India’s motivations for testing nuclear weapons in May 1998 continues to this day, with scholars emphasizing a combination of security pressures, the desire for international prestige, and domestic politics. None of these explanations, however, satisfactorily explains India’s choice of nuclear posture. For example, even if India has a preference for prestige, perhaps furthered by the overt testing of nuclear weapons, this does not lead to any determinate predications about how India might operationalize its nuclear weapons capability. In this section I test optimization theory against the three alternative explanations for Indian nuclear posture. I find that only optimization theory explains the choice, persistence, and character of India’s assured retaliation posture. First, India has never had a reliable third-­party patron that might intervene on its behalf in a conventional conflict with either Pakistan or China. The United States never offered India that level of commitment against China in the 1960s or 1970s and often intervened on Pakistan’s side against India, while the USSR was never willing to extend a conventional or nuclear umbrella to India.71 Furthermore, although India has had close ties to the USSR, particularly after 1971, the level of commitment between the two nations was constrained by India’s insistence on nonalignment and India’s reluctance to become overly dependent on what it viewed as a capricious Soviet Union. India never gambled, and the Soviet Union never indicated, that it would intervene on India’s behalf in any conflict with either Pakistan or China.72 Thus, in a potential India-­China conflict, the USSR officially professed neutrality. India has thus always lacked a third-­party 71. See A. G. Noorani, “India’s Quest for a Nuclear Guarantee,” Asian Survey, vol. 7, no. 7 (July 1967), pp. 490–­502; also see Andrew B. Kennedy, “India’s Nuclear Odyssey: Implicit Umbrellas, Diplomatic Disappointments, and the Bomb,” International Security, vol. 36, no. 2 (Fall 2011), pp. 120–­53. 72. Even during the Sino-­Indian war, the Soviet Union offered tepid diplomatic support for India, and both the United States and USSR were preoccupied with each other in the Cuban Missile Crisis at the time.

india 111

patron with incentives to intervene on its behalf. Furthermore, any explicit reliance on either superpower risked undermining India’s position in the nonaligned movement. So, although there was a period when India searched for a reliable third-­party patron as optimization theory would predict, India’s leaders ultimately failed in that quest for an umbrella. India’s leaders thus concluded that they clearly lacked a reliable third-­party patron that it could compel to intervene on its behalf, thereby rendering a catalytic posture infeasible.73 Second, optimization theory anticipates that India’s security situation is lax enough to allow it to select out of an asymmetric escalation posture should it wish to do so. India has conventional superiority over Pakistan74 and natural barriers provide India with protection against an existential threat from China. Against Pakistan, India has historically enjoyed a significant quantitative advantage in land forces, a 2:1 advantage in air forces, and a dominant navy. There is no question that India could blunt a Pakistani conventional offensive, even if its ability to achieve rapid decisive victories on Pakistani territory is questionable.75 Though Pakistan is a nuclear weapons state, India’s nuclear capabilities simply need to deter the use of Pakistan’s nuclear weapons, since India’s conventional capabilities are sufficient to meet likely Pakistani conventional threats. As a result, the choice of an assured retaliation or an asymmetric escalation posture could both be consistent with strictly realist prescription. With respect to China, the mountainous terrain of the Himalayas gives India a natural defensive advantage. The inhospitable terrain and high altitudes go a long way in blunting an existential land threat involving an armored punch from China over the passes of Arunachal Pradesh or Aksai Chin, even though the terrain favors China in theater around the line of actual control (LAC) since Chinese forces have the advantage of operating on the plateau, while Indian forces have to scale uphill.76 The force balances along the India-­China border are difficult to measure over time, but both countries since 1962 have deployed significant mountain division assets in the terrain to deter movement along the disputed border regions. For its part, India has raised roughly ten mountain divisions since the 1980s 73. See Andrew B. Kennedy’s excellent article on this, “India’s Nuclear Odyssey,” especially pp. 134–­46. 74. John H. Gill, “Brasstacks: Prudently Pessimistic,” in Sumit Ganguly and S. Paul Kapur, eds., Nuclear Proliferation in South Asia, chapter 3; Ashley Tellis, Strategic Stability in South Asia, Arroyo Center Document Briefing (Santa Monica: RAND Corporation, 1997), pp. 12–­33. 75. See Tellis, Strategic Stability in South Asia, pp. 12–­33. 76. For an excellent treatment of border disputes between India and China, see M. Taylor Fravel, Strong Borders, Secure Nation: Cooperation and Conflict in China’s Territorial Disputes (Princeton: Princeton University Press, 2008), especially chapter 4.

112

Chapter 4

solely to deploy along the India-­China border region, and is currently in the process of raising several more. This force strength, and an increasingly augmented air capability in theater, is aimed at providing India the capability to hold the LAC, even during a potential conventional war with Pakistan. Even if China could make limited gains in the northeast theater, it is unlikely to be able to pose an existential threat to the Indian state with conventional forces. Though border disputes between the two nations persist, given the terrain and the peripheral nature of the disputes from either nation’s strategic centers, India does not face an existential land threat—­or proximate offensive threat—­from China. India can therefore employ a “defense-­in-­depth” conventional strategy to meet any Chinese land threat. Though China is a nuclear weapons state, India is again not forced to offset its larger nuclear neighbor’s conventional power with a first-­use posture. Therefore, India’s security environment and alliance patterns are indeterminate in driving its nuclear posture. India thus has sufficient latitude to select either an asymmetric escalation posture or an assured retaliation posture. Both postures are consistent with the pressures of India’s security environment, suggesting that a strictly realist theory of Indian nuclear posture is incapable of adjudicating between an asymmetric escalation posture and an assured retaliation posture. Optimization theory, however, recognizes this indeterminacy and moves to the next node in its analysis: civil-­military relations. In India’s case, civil-­ military relations proves to be the decisive variable (conditional on the security environment described previously) pushing India squarely in the direction of an assured retaliation posture whose character overwhelmingly privileges assertive civilian control over nuclear assets. This is because, since Nehru and independence, India has maintained a strong—­ arguably pathological—­emphasis on civilian dominance of an intentionally defanged military that is kept out of both high-­level and even low-­level decision-­making processes.77 On India’s Independence Day, Nehru abolished the position of the military commander in chief, which traditionally had been India’s highest ranking army officer, to render impotent “the only challenge to civilian authority . . . [and to make] it easier to balance off the army with the other two services.”78 A series of organizational changes in 1947 significantly undermined the decision-­making capacity and autonomy of the military in favor of 77. See for example Stephen Cohen, The Indian Army: Its Contribution to the Development of a Nation (Delhi: Oxford University Press, 1990), chapter 7; Stephen Cohen, “The Military and Indian Democracy,” in Atul Kohli, ed., India’s Democracy: An Analysis of Changing State-­Society Relations (Princeton: Princeton University Press, 1988, pp. 99–­143; and Stephen P. Rosen, Societies and Military Power, chapter 6; Anit Mukherjee, “Absent Dialogue,” Seminar Magazine (India), July 2009. 78. Cohen, The Indian Army, pp. 171–­72.

india 113

civilian authorities in the PMO, the Ministry of Defence, and the Ministry of Finance. This was an intentional strategy designed to keep India’s military peripheral, and wholly subservient, to civilian political authorities. One of India’s most powerful defense ministers, V. K. Krishna Menon, who held charge during the Sino-­Indian war in 1962, summarizes the persistent condition of Indian civil-­military relations: “It is wrong for the army to try to make policy  .  .  . Military planning and arrangements and things of that kind must remain in the hands of the Government.”79 Though it expresses the principle of civilian control, the assertion that the civilian bureaucracy should be engaging in military planning and arrangements suggests a level of civilian intrusion that goes beyond simple hierarchy. Civilian interference in the strategic, operational, and even tactical details of the 1962 war is largely credited with India’s humiliating debacle in that conflict. Menon and Nehru interfered so deeply in the army’s affairs that they were managing the placement of individual platoons in the war.80 The civilian defense that the army was incapable of theater innovation following the Chinese invasion obscures the fact that Nehru and Menon had so defanged the military—­through irregular appointments and rotations and Menon’s intentionally abrasive behavior toward senior army officers—­that army commanders were incapable of, and proscribed from, making operational or tactical decisions on their own. Thereafter, the civilian leadership has tried to refrain from meddling with tactical and operational matters during conflicts, but it persists in erecting significant firewalls between itself and the military. Noting the praetorian trajectory of Pakistan, India’s civilian authorities have always been wary of its armed guardians.81 As Stephen Cohen points out, “no military of equivalent importance or size has less influence. This situation was not inherited from the British but was assembled, piece by piece, over the years, and is now enshrined in various constitutional and bureaucratic structures.”82 The initial defanging of the military in the decision-­making process at independence has ossified over time and created a rigidly poor civil-­military relationship that is very resistant to change.83 Even after the decisive victory 79. V. K. Krishna Menon quoted in Michael Brecher, India and World Politics: Krishna Menon’s View of the World (New York: Praeger Books, 1968), p. 260. 80. See Srinath Raghavan, “Civil-­Military Relations in India: The China Crisis and After,” Journal of Strategic Studies, vol. 32, no. 1 (February 2009), pp. 149–­75. 81. In 1984, for example, following Operation Bluestar against the Sikh Golden Temple, there were various Sikh mutinies across the Indian Army in protest; and despite their overrepresentation in the officer and generals corps, it took almost 50 years for a Sikh to be made Chief of Army Staff due possibly to fears of Sikh separatism. See Cohen, “The Military and Indian Democracy,” pp. 134–­37. 82. Ibid., p. 114. 83. Ibid., pp. 114–­21.

114

Chapter 4

in the 1971 war, Indira Gandhi lowered the army’s status in the Warrant of Precedence due to fears, according to George Tanham, that they would “become too popular or too powerful.”84 Though the Indian military is extremely professional and has been used extensively for both internal and external security operations, its budget remains tightly controlled by civilian masters who often do not spend the entire appropriation despite significant materiel deficiencies. Everything from the acquisition process to recruitment and training is subject to political interference, whether from the Ministry of Defence, Finance, or DRDO.85 Tanham notes that fears of coups mean that “control of major troop movements or other military forces in peacetime resides with the [defense] minister. The Ministry of Defence in essence serves as a higher headquarters that makes all decisions; the service chiefs have no statutory power to make government decisions.”86 This condition persists even today. A recent Chief of Naval Staff, Admiral Arun Prakash, quipped that even on military affairs, “Between a scientist and a soldier, the politicians . . . believe the scientist.”87 Rear Admiral Raja Menon, one of India’s preeminent strategic thinkers, has gone even further, writing: “New Delhi is one of the few capitals of the world where turf battles don’t just end in bloodshed but in the annihilation of an entire group—­the military, the only people with operational staff knowledge, annihilated by the victors, the N-­physicists, bureaucrats, intelligence-­wallahs and DRDO scientists.”88 The fact that this view is held by two retired senior officers from the navy, which has historically enjoyed a better relationship with the political leadership in Delhi than the army, suggests that assertive control over all the military services—­not just the army—­continues to be a pervasive feature of India’s civil-­military relations. Further, the political leadership employs “divide and rule” tactics to keep the military—­and the army in particular—­weak. There is little military input in the Ministry of Defence, which is populated almost entirely by civil servants who rotate frequently. For example, it is not unusual for a member of the Indian Administrative Services who has just spent years in the Ministry of Culture and Tourism to find him-­or herself deputed to the Ministry of Defence, with no knowledge whatsoever of military affairs, in

84. George K. Tanham, Indian Strategic Thought: An Interpretive Essay (Santa Monica: RAND Publications, 1992), p. 73. 85. The extent of this civilian penetration was revealed in a seminar entitled “Modernization of Infantry in India,” Centre for Land Warfare Studies, New Delhi, India, May 25, 2009. 86. Tanham, Indian Strategic Thought, p. 74. 87. Admiral Arun Prakash, quoted in Karnad, India’s Nuclear Policy, p. 69. 88. Raja Menon, “Just One Shark in the Deep Blue Ocean,” Outlook, August 10, 2009.

india 115

charge of sophisticated procurement or planning decisions.89 Indeed, this state of affairs has led the Indian military to claim that they are not only subservient to civilian political authority but under an impenetrable layer of intervening “bureaucratic control” that has neither expertise nor much interest in military affairs.90 The service chiefs lack consistent access to the prime minister, and the position of Chief of Defence Staff, aimed at integrating military input into the policy-­making process, continues to remain absent due to lack of political will. In fact, prime ministers and defense ministers have often interfered—­by denying promotions at lower levels—­in the selection of even the Chief of Army Staff based on who they favor and who is politically acceptable, even though the selection is supposed to be made by the army on the basis of seniority. For example, wary of the Sikhs’ power and overrepresentation in the military, several Sikh generals were eliminated at lower level promotions in anticipation of their ranking seniority for the Chief of Army Staff position. Not until 2005 did India have a Sikh Chief of Army Staff, Gen. J. J. Singh, despite the overrepresentation of Sikhs in the officer corps. The result of India’s civil-­military architecture is a military firmly subject to the political leadership, with little ability to seek its own organizational preferences. This assertive civil-­ military arrangement generates a prediction that India would have a similarly assertive nuclear command-­and-­control structure. This eliminates the option of a first-­use asymmetric escalation posture which requires the incorporation of ready nuclear forces into the military and integration of civil-­military structures to manage the arsenal. Furthermore, beyond simply civilian authority over the movement and release of nuclear assets—­as even the United States nominally has—­India’s extreme civilian distrust of the military would suggest that nuclear assets would actually be under civilian custody and that those custodial arrangement would persist until the point at which the military’s role in releasing nuclear weapons is absolutely unavoidable. This is precisely the character of India’s assured retaliation posture that is observed over time. Not only is there a persistence of civilian control over the direction of the arsenal with little military input, but civilian authorities continue to steward the arsenal amongst several civilian agencies, keeping nuclear assets out of the hands of the military. According to optimization theory, India’s assured retaliation posture is driven primarily by a deep civilian distrust of the military, not because of 89. See Rosen, Societies and Military Power, pp. 230–­31; and Stephen Cohen, “The Military and Indian Democracy,” p. 119. 90. Ali Ahmed, “The Central Debate in India’s Civil-­Military Relations,” IDSA (New Delhi) Strategic Comments, July 6, 2009.

116

Chapter 4

oft-­cited financial constraints that limit India’s ability to develop a more robust arsenal. The critical indicator is the inviolable distinction between “custodian and controller” in India’s nuclear stewardship procedures: civilians continue to maintain custody of India’s nuclear assets and would not transfer them to the military until after it had absorbed a nuclear strike. India has already overcome the biggest financial hurdle in developing a nuclear arsenal—­indigenous fissile material production. The marginal cost of a larger arsenal, especially with operationalized indigenous ballistic missile families, would be wholly manageable for a country that is growing at India’s present rate, particularly since the budget for the nuclear program is spread across both civilian and military organizations. Indeed, India already has sufficient capabilities—­both nuclear weapons and delivery vehicles—­to achieve an asymmetric escalation posture that envisions nuclear use on the battlefield. In short, anything Pakistan can afford, India can surely also afford. But such a posture would require a fundamental shift in civil-­military relations and organizations. This is currently politically unacceptable to India’s civilian leadership. Certainly, India’s security environment is lax enough to permit India to have this choice. But the determinate variable is India’s extremely assertive civil-­military relationship, which not only predicts an assured retaliation posture, but an assured retaliation posture in which nuclear assets are physically administered by civilians. The implication of my argument for India is that even if it were unconstrained by resources, it would still adopt an assured retaliation posture because of its civil-­military structures. As noted earlier, a strict structural realist explanation cannot account for India’s choice, or the persistence, of the assured retaliation posture. India’s security environment is not pacific, but neither is it so hostile that its vital interests are persistently threatened. India possesses conventional superiority over Pakistan, while natural geographic barriers and the distance of India’s strategic centers from the Chinese border enable India to adopt a defense-­in-­depth strategy against any Chinese incursion that provides India sufficient security on its frontiers. India’s security environment is thus permissive enough to allow for a choice between a first-­use asymmetric escalation posture and an assured retaliation posture. Both could be consistent with India’s security position, but a strictly realist account cannot adjudicate between them. If anything, a realist prediction for Indian nuclear posture would lean toward an asymmetric escalation posture so that India could deter a more powerful Chinese land army and avoid a repeat of the 1962 war by credibly threatening tactical nuclear first use against Chinese forces. But there is no evidence that India has ever contemplated adopting such a nuclear posture. At best, structural realism cannot generate a specific prediction for Indian nuclear posture. At worst, the choice of an assured retaliation posture is an anomaly for structural realism.

india 117

Similarly, a technological determinist explanation inadequately captures India’s nuclear trajectory. This explanation would posit that India would gradually evolve a more robust nuclear capability over the years as its technical capabilities and fissile material stockpile evolved. It suggests that although India may have started with an assured retaliation posture in 1974, after 1998 its technological evolution with short-­and medium-­range ballistic missiles should have effected a shift to a ready deterrent that enables the first use of nuclear weapons if necessary. The key prediction is that India’s nuclear capabilities would determine India’s choice of posture, rather than the other way around (choice of posture determining the capabilities India would need to develop). Clearly this theory flunks the test with respect to India, which retarded its nuclear developments for political reasons. After the 1974 PNE, even though India’s atomic energy bodies and DRDO had the technical capability to begin perfecting weapons designs, they were prevented from doing so by Indira Gandhi, who wanted to keep India’s nuclear weapons capability dormant. It was not until nearly fourteen years later, at Rajiv Gandhi’s personal behest based on reports of Pakistan’s advancing program, that India’s scientists were authorized to begin weaponizing India’s capabilities with extensive work on weapons designs and ballistic missile production. By the early 1990s, India had the technological foundation to move to a more robust nuclear posture, but refrained from doing so. Even after 1998, India had the technical prowess and required fissile material production to adopt a first-­use asymmetric escalation posture if it so desired, relying on aircraft and short-­range Prithvi missiles for deterrence by denial missions. India’s political leadership chose not to do so, instead persisting with a widely dispersed assured retaliation posture for political reasons. India has the technical capacity to integrate nuclear weapons into its armed forces to orient them for first use to deter, say, Chinese border incursions in Arunachal Pradesh. But it lacks the organizational and political will to make that shift. In India’s case, then, political choices about what posture to adopt have determined its technical developments, not the other way around. There is little question that DRDO’s scientists dream of building everything from ICBMs tipped with multiple warheads, ballistic missile defenses, and an array of cruise and ballistic missile capabilities. But each deployment ultimately requires political clearance, which often comes slowly if at all, regardless of DRDO’s research activities. The more compelling alternative explanation for India’s assured retaliation posture, most notably advanced by Rajesh Basrur, is a strategic culture explanation.91 Basrur and I agree on the character of India’s nuclear posture—­assured retaliation, or what he terms minimum deterrence—­but 91. Rajesh Basrur, Minimum Deterrence and India’s Nuclear Security (Stanford: Stanford University Press, 2006).

118

Chapter 4

we disagree on the sources of that choice. Basrur attributes India’s minimum deterrence posture to a uniquely Indian strategic culture, deploying the term as Colin Gray did in reference to “national styles”: “the socially constructed and transmitted assumptions, habits of mind, traditions, and preferred methods of operation—­that is, behavior—­that are more or less specific to a particular geographically based security community.”92 Though Basrur uses this formulation to posit that strategic culture is a “set of structured preferences . . . about the nature of the strategic threat, the role of force in that environment, perceptions of threat, and the framing of responses” to those threats, as does Johnston, his unit of analysis is the Indian state.93 Exploring India’s basic beliefs and the operationalization of those beliefs with respect to nuclear weapons, Basrur argues that India’s strategic culture therein is based on three characteristics: nuclear weapons have limited utility for national security, they are primarily political weapons, and India has restrained responses to external stimuli.94 The implication, according to Basrur, is that India since Nehru has been ambivalent about nuclear weapons, because Nehru was a reluctant realist but “consider[ed] nuclear weapons both morally unacceptable and detrimental to security because of the risks associated with it.”95 For Basrur, it is this latter feature of Indian strategic culture—­an intense abhorrence of nuclear weapons as military weapons from Nehru on—­that has caused India to adopt a dispersed assured retaliation posture. Basrur’s strategic culture explanation, however, does not adequately explain India’s nuclear posture over time for several reasons. First, while many of India’s Congress prime ministers—­Nehru, Indira Gandhi, Morarji Desai, and perhaps Manmohan Singh—­have been deeply ambivalent about nuclear weapons qua weapons, India’s BJP government was much more hawkish in its view of nuclear weapons.96 Neither BJP Prime Minister Vajpayee nor his closest advisors such as L. K. Advani, Brajesh Mishra, or Jaswant Singh held the moral aversion to nuclear weapons that their Congress predecessors did (if they even did)—­indeed, all believed in their utility for India’s security and to give India shakti (power) and self-­confidence.97 Basrur’s strategic culture explanation does not adequately capture the persistence of the assured retaliation posture through the BJP years, when one of the supposed drivers of Indian strategic culture varies by party. Since it was the BJP that took India over the overt nuclear threshold and that 92. Ibid., p. 54 quoting Colin S. Gray, Modern Strategy (Oxford: Oxford University Press, 1999), p. 28. 93. Basrur, Minimum Deterrence, p. 54. 94. Ibid., p. 58. 95. Ibid. 96. See Hymans, Psychology of Nuclear Proliferation, chapter 7. 97. A. B. Vajpayee quoted in Chengappa, Weapons of Peace, p. 36.

india 119

drafted India’s declaratory nuclear doctrine, this is a significant failing in Basrur’s explanation. Second, Basrur’s explanation does not clarify the char­acter of India’s assured retaliation posture. More specifically, Basrur cannot explain why the military is so deeply distrusted in the stewarding of nuclear weapons. There are various stewardship procedures that would be consistent with “nuclear minimalism,” including entrusting the military with warheads but instituting negative control procedures that prevent unauthorized use. However, India has chosen a very specific method of administering its nuclear posture, relying heavily on civilian agencies and sidelining the military almost entirely. I argue that optimization theory in which a combination of India’s security environment and poor civil-­military relations, which have ossified over time since independence, offers a stronger explanation for the persistence of India’s assured retaliation posture and the specific method of administering India’s nuclear assets. Although Congress and the BJP have different worldviews and thus may be expected to favor different operationalizations of India’s nuclear posture, they inherited the same firewalled civil-­military structures that pushed them toward an assured retaliation posture that relied almost entirely on civilian custodial arrangements and dispersion. Thus, across political parties and variable strategic cultures, India has evinced remarkable continuity in its nuclear posture. However, it is worth noting that the civil-­military relations variable in my theory is not incompatible with a strategic culture explanation. India’s unique civil-­ military structures have evolved since Nehru as a result of deep, and not unfounded, initial fears of a military coup—­as a newly independent state, it was a perfectly reasonable fear that India’s military might take the reins of political power if the writ of the state were insufficiently strong to keep the nation together. A distrust of India’s armed forces and particular organizational features Nehru instituted to defang the military hardened over time, producing a civil-­military relationship in which India’s political leadership is patently unwilling to entrust any dedicated nuclear subcomponents to the armed forces. Though my reading of Nehru’s civil-­military decisions—­ and the subsequent institutionalization of those decisions—­takes a more instrumental view, a strategic culture explanation that focuses on this feature of India’s security structures could be wholly compatible with my explanation (which takes civil-­military structures as an exogenous variable). But Basrur’s explanation focusing on India’s ambivalence toward nuclear weapons themselves—­when it is, in fact, an ambivalence toward the armed forces—­is not a convincing account for why India has opted for an assured retaliation posture and persisted with that choice even as India’s domestic political configuration and external security stimuli have varied. Thus in India’s case, optimization theory offers the most compelling explanation for the sources of India’s assured retaliation. The theory

120

Chapter 4

illustrates that even in outcomes such as nuclear strategy, domestic-­level variables can play a crucial role. In India, it is civil-­military relations—­not its security environment—­that provides the driving and trumping variable in this choice. There are some features of India’s assured retaliation posture that this theory cannot explain—­notably the decision regarding creeping weaponization and testing. Rajiv Gandhi’s decision to move from a deeply recessed assured retaliation posture to a more readily available one, Rao’s and Vajpayee’s aborted decisions to test nuclear weapons in 1995 and 1996, and Vajpayee’s successful decision to test in 1998 all require deeper structural and domestic political explanations as each of these decisions seems to have been triggered by changes in India’s security environment and domestic politics. But these decisions should not obfuscate the continuity India has evinced in its choice of an assured retaliation posture since 1974.

Chapter Five

China

China was the last of the five recognized nuclear weapons states to develop nuclear capabilities, testing its first fission device, a 20–­22 kT uranium device, on October 16, 1964. Unlike the superpowers before it, China has always maintained a limited nuclear arsenal, small numbers of intercontinental ballistic missiles (ICBMs), and hardly any support architecture. In this way, the Chinese nuclear posture has more closely resembled that of other regional nuclear powers than those of the superpowers. This chapter relies on the best available English language sources (of which there are actually more than, say, for India) to detail Chinese nuclear posture over time. China, like India, has persistently adopted an assured retaliation posture against its potential adversaries. Though Chinese strategists have long debated the merits of shifting from what they term a “minimum deterrence” doctrine to a more aggressive “limited deterrence” doctrine that fields theater nuclear capabilities for possible first use, there has in fact been remarkable continuity in China’s nuclear posture. China’s posture enables its nuclear weapons to survive a first strike and effectively retaliate with nuclear weapons. As in India, China faces no existential land threats and also displays a highly assertive civil-­military, or more accurately, party-­military relationship that privileges tight centralized control over China’s nuclear assets. These variables have pushed China toward an assured retaliation nuclear posture. The capabilities required to achieve that posture have evolved over time as China has had to dynamically respond to growing American conventional and nuclear capabilities. In particular, China faces more accurate U.S. missiles that could potentially disarm China’s nuclear forces and American 121

122

Chapter 5

missile defenses that might intercept any Chinese forces that survived. Nonetheless, I demonstrate that China has shown remarkable consistency in its posture of assured retaliation. The Chinese effort to improve capabilities over the past several decades has focused on survivability and penetration in response to U.S. capabilities to ensure that China maintains the forces required to guarantee nuclear retaliation following a first-­strike effort by an adversary. By contrast, China has not invested in the systems, command-­ and-­control architecture, and early warning capabilities that would enable China to make a shift in its posture away from assured retaliation and toward asymmetric escalation.

China’s Nuclear Posture: Assured Retaliation by Many Names China’s motivations for acquiring a nuclear weapons capability are described in detail by John Lewis and Xue Litai.1 A series of crises between China and the United States made it clear to Mao and China’s leadership that China faced the threat of repeated American efforts at nuclear coercion. At the end of the Korean War, the Eisenhower administration’s nuclear threats to Beijing “forced the Chinese to compromise” on an armistice agreement.2 In 1954, the crisis over Quemoy and Matsu took place under the shadow of Eisenhower’s “massive retaliation” nuclear posture which lowered the threshold for use of nuclear weapons, further highlighting China’s vulnerability to American nuclear threats. Hence, Lewis and Xue conclude that “the events in Korea, Indochina, and the Taiwan Strait constituted the proximate cause of the Chinese decision to build a national strategic force,” though they also note that Mao’s view of China’s national interest may have led to the Chinese pursuit of nuclear weapons even in the absence of these crises.3 Nevertheless, the initiation of China’s nuclear weapons program followed a series of militarized crises against the United States in which China’s leadership believed it was held hostage by American nuclear blackmail. Though Mao famously derided nuclear weapons as “paper tigers,” he directed China to develop them as a “defense against nuclear blackmail and nuclear war” against the United States and, ultimately, the USSR following China’s subsequent split with Moscow.4 Unlike other regional nuclear powers—­India and Pakistan for example—­ China’s decision to develop nuclear weapons did not arise from a defeat in

1. Lewis and Xue, China Builds the Bomb. 2. Ibid., p. 14. 3. Ibid., p. 35. 4. Mao Zedong quoted in ibid., p. 36.

china 123

conventional war but rather to balance against the threat of future nuclear coercion and potential nuclear use. In January 1955, Mao initiated China’s nuclear weapons program at a meeting of the Central Secretariat. The details of China’s quest for nuclear weapons are documented in detail elsewhere, but after Soviet assistance ceased, China was able to develop a sufficiently large indigenous base of uranium enrichment and eventually plutonium production, engineers, and physicists to progress with minimal Soviet help.5 Less than ten years after the decision to weaponize nuclear capabilities, China successfully detonated its first uranium fission device on October 16, 1964 from a 400-­foot tower. On May 14, 1965, China tested its second nuclear device by air-­ dropping a gravity bomb from a Hong-­6 bomber, demonstrating its ability to deliver a nuclear payload. The next year, after several trials of multiple ballistic missile variants, on October 25, 1966 China tested a fully operational DF-­2 ballistic missile with a live nuclear warhead mated atop, demonstrating that in addition to bomber delivery, China had mastered missile warhead designs and now had a credible ballistic missile delivery capability as well.6 On June 17, 1967, less than three years after testing a uranium fission device—­faster than the United States and USSR—­China successfully tested a large three-­megaton thermonuclear device, giving it a true weapon of mass destruction against potential adversaries. China was a credible, independent nuclear weapons power with an indigenous development capacity at this point. But how did it choose to operationalize its nuclear weapons capability? Did it choose to follow the superpowers and develop a massive arsenal of nuclear weapons across a range of yields and delivery methods to deter conflict across the spectrum of intensities? It did not. Instead, China chose to rely on ambiguous numbers and deployment procedures under Mao’s direct control to adopt an assured retaliation posture to deter nuclear use and nuclear coercion.7 The 5. See Lewis and Xue, China Builds the Bomb; also see Nie Rongzhen, Inside the Red Star: The Memoirs of Marshal Nie Rongzhen, translated by Zhong Renyi (Beijing: New World Press, 1988), chapter 24; also see Robert S. Norris, Andrew S. Burrows, and Richard W. Fieldhouse, Nuclear Weapons Databook, Volume V: British, French, and Chinese Nuclear Forces (Boulder: Westview Press, 1994), chapters 6 and 7. 6. See Inside the Red Star, p. 711. 7. The English language literature on Chinese nuclear posture, both past and present, is surprisingly deep. See Lewis and Xue, China Builds the Bomb; Lewis and Xue, China’s Strategic Seapower; Lewis and Xue, Imagined Enemies; Lewis and Xue, “Strategic Weapons and Chinese Power,” pp. 541–­ 54; Johnston, “China’s New ‘Old Thinking’,” pp. 5–­42; Alastair Iain Johnston, “Prospects for Chinese Nuclear Force Modernization: Limited Deterrence versus Multilateral Arms Control,” China Quarterly, no. 146 (June 1996), pp. 548–­76; Paul J. Bolt and Albert S. Willner, eds., China’s Nuclear Future (Boulder: Lynne Rienner, 2006); Goldstein, Deterrence and Security in the 21st Century; Chong–­pin Lin, China’s Nuclear Weapons Strategy; Lewis, Minimum Means of Reprisal; Bates Gill, James

124

Chapter 5

first phase of China’s nuclear posture is from 1964 through the late 1980s when China relied on what Marshal Nie Ronghzen once termed “the minimal means to stage a counterattack in case [China] suffered a surprise nuclear attack by the imperialists”8—­what some, such as Jeffrey Lewis, have translated as “the minimum means of reprisal.” During this period, China sought a sufficient capability to absorb a nuclear first strike yet retain the ability to retaliate with nuclear forces against any adversary. While many scholars refer to this posture as minimum deterrence (zuidi xiandu he weishe), Iain Johnston has pointed out that the term itself does not appear very often in Chinese writings to describe China’s own posture.9 I identify the Chinese posture in this phase as one of assured retaliation (which might be loosely translated to the Chinese equivalent of “effective counter-­attack capability” [ youxiao he baofu]) since it captures the key features of China’s posture in this period: no-­first-­use made credible by strong centralized and negative controls, survivability through dispersed and concealed stewardship procedures and numerical ambiguity, and punitive retaliatory strikes Mulvenon, and Mark Stokes, “The Chinese Second Artillery Corps: Transition to Credible Deterrence” in James Mulvenon and Andrew N, D, Yang, The People’s Liberation Army as Organization (Santa Monica: RAND Conference Proceedings, 2002), chapter 11; Evan S. Medeiros, “Minding the Gap: Assessing the Trajectory of the Second Artillery,” in Roy Kamphausen and Andrew Scobell, eds., Right Sizing the People’s Liberation Army: Exploring the Contours of China’s Military (Carlisle: Strategic Studies Institute, U.S. Army War College), September 2007, chapter 4. See M. Taylor Fravel and Evan S. Medeiros, “China’s Search for Assured Retaliation: The Evolution of Chinese Nuclear Strategy and Force Structure,” International Security, vol. 35, no. 2 (Fall 2010), pp. 48–­87; Larry M. Wortzel, “PLA Command, Control, and Targeting Architectures: Theory, Doctrine, and Warfighting Applications,” in Kamphausen and Scobell, Right Sizing the People’s Liberation Army, chap­ ter 6; Michael S. Chase and Evan Medeiros, “China’s Evolving Nuclear Calculus: Modernization and Doctrinal Debate,” in James Mulvenon and David Finkelstein, eds., China’s Revolution in Doctrinal Affairs: Emerging Trends in the Operational Art of the People’s Liberation Army (Alexandria: CNA Corporation, December 2005), chapter 5; Kenneth Allen and Maryanne Kivlehan-­Wise, “Implementing PLA Second Artillery Doctrinal Reforms,” in Mulvenon and Finkelstein, China’s Revolution in Doctrinal Affairs, chapter 6; John Lewis and Hua Di, “China’s Ballistic Missile Programs: Technology, Strategies, Goals,” International Security, vol. 17, no. 2 (Fall 1992), pp. 5–­40; Michael S. Chase, Andrew S. Erickson, and Christopher Yeaw, “Chinese Theater and Strategic Force Modernization and its Implications for the United States,” Journal of Strategic Studies, vol. 32, no. 1 (February 2009), pp. 67–­114; Jing-­dong Yuan, “Effective, Reliable and Credible: China’s Nuclear Modernization,” The Nonproliferation Review, vol. 14, no. 2 (July 2007), pp. 275–­301; Chu Shulong and Rong Yu, “China: Dynamic Minimum Deterrence,” in Alagappa, ed., The Long Shadow, pp. 161–­87. 8. Nie Rongzhen quoted in Litai Xue, “Evolution of China’s Nuclear Strategy” in John C. Hopkins and Weixing Hu, eds., Strategic Views from the Second Tier: The Nuclear Weapons Policies of France, Britain, and China (New Brunswick: Transaction Publishers, 1995), p. 171. 9. See Johnston, “China’s New ‘Old Thinking’,” pp. 17–­18.

china 125

against key countervalue targets and soft counterforce targets to degrade the adversary’s ability to follow-­on with a second wave of nuclear forces.10 Though there was not an articulated nuclear doctrine during Mao’s tenure, there were sufficient declarations of China’s nuclear strategy, coupled with operational details that gave insights into China’s nuclear posture. Indeed, based on a series of Mao’s famous one-­sentence maxims, Lewis and Xue aggregate seven broad key principles that they argue characterized China’s nuclear posture since the 1960s: no-­first-­use, no tactical nuclear weapons, “small but better” (shao er jing) forces that were reliable and credible, “small but inclusive” (xiao er quan) forces, “minimum retaliation” (qima de huanji), quick recovery from nuclear attack, and “soft-­target kill capability” in addition to large strategic centers.11 Another phrase that was often used to describe China’s posture in this period was “to gain mastery by striking only after the enemy has struck” (hou fa zhi ren), which encapsulates the aim of the assured retaliation posture.12 Xue Litai also writes that China’s nuclear weapons were designed to deter strictly nuclear coercion and nuclear attack: “PLA strategists long maintained that a balance must exist between the People’s War to deter or repel conventional attacks and small nuclear forces to deter a nuclear attack.”13 In 1983, Deng Xiaoping reaffirmed this basic aim of China’s nuclear forces. These limited aims distinguish China’s posture from American, French, and Pakistani postures that have aimed to deter conventional attacks by credibly threatening the asymmetric first use of nuclear weapons against conventional forces. Here I focus on the key characteristics of China’s posture that define it as an assured retaliation posture, as opposed to an asymmetric escalation posture. First, with respect to capabilities, China developed credible but limited nuclear forces deliverable by both bomber and land-­based missiles. China did not develop sea-­based capabilities until 1988 when the Type-­ 092 submarine (Xia class) was deployed with the Julang-­1 ( JL-­1) SLBM, though neither is believed to have ever been operational. China’s initial fission devices in 1964–­65 were deliverable by the indigenous Hong-­6 (3000km range) and Soviet-­made Tu-­4 bombers (6000km range), as well as the Dongfeng-­2 (DF-­2) ballistic missile (1250km range with a 1550kg warhead) which gave China theater reach against U.S. bases and forces in the region, as well as Soviet targets.14 China began development on other variants of its DF family in the mid-­1960s, each with a different envisioned 10. See Lewis and Xue, China’s Strategic Seapower, chapter 10. Evan Medeiros also uses this term to describe China’s doctrine. 11. Lewis and Xue, China’s Strategic Seapower, pp. 232–­33. 12. Xue, “Evolution of China’s Nuclear Strategy,” p. 181. 13. Ibid., p. 171. 14. See Norris, Burrows, and Fieldhouse, chapter 7; and Lewis and Hua, “China’s Ballistic Missile Programs.”

126

Chapter 5

target: DF-­2 for Japan, DF-­3 for the Philippines, DF-­4 for Guam, and DF-­5 for the continental United States.15 Though China could reach U.S. targets in the East Asia theater, it was not able to operationalize the DF-­5, which is capable of delivering a single 4–­5 megaton warhead 12,000 km, until 1981. Prior to that relatively late date, China had no credible capability to deliver nuclear weapons against the American homeland. All of these ballistic missiles are liquid-­fueled, and therefore require several hours to fuel—­which can only occur once the missile is erect on its launchpad—­and ready for launch (2–­3 hours was the target launch preparation time).16 Nevertheless, these capabilities and a warhead production rate that averaged about ten per year over twenty years gave China sufficient warheads and delivery capabilities to retaliate against key U.S. theater targets as early as 1965. It was for this technical reason that soft counterforce targets were part of China’s retaliatory target package. China simply did not have the reach to employ countervalue targeting until the DF-­5 became operational. A critical feature of China’s capabilities, however, is that by developing primarily massive megaton-­yield warheads and inaccurate liquid-­fuel ballistic missiles, China did not emphasize the development of lower-­yield weapons or delivery forces on alert necessary to make the first use of nuclear weapons credible. The types of capabilities that China developed are consistent with a retaliatory posture aimed at deterring nuclear coercion and nuclear use. It is difficult to imagine Chinese capabilities being in response to conventional threats. It would simply be too disproportionate to respond to a conventional threat with the use of a several-­megaton nuclear warhead. The technical features of these capabilities therefore precluded a war-­fighting posture. Rather, they could only be oriented for a retaliatory strike. Since these capabilities were fielded in an era without effective missile defenses, their operational effectiveness as a deterrent could be achieved with small numbers and single-­warhead systems provided they were survivable. While the primary utility of these forces is for punitive countervalue strikes, soft counterforce targets—­key military and command-­and-­control nodes—­were (and continue to be) included in China’s target package to degrade an opponent’s ability to launch follow-­on nuclear strikes for “de-­ escalatory and possibly even war-­termination purposes.”17 In order to credibly threaten retaliation, China undertook stewardship and deployment procedures to ensure survivability against a potential nuclear first strike. Even in the early years of China’s nuclear program, poor U.S. intelligence and the sheer number of targets American bombers and 15. Lewis and Xue, China Builds the Bomb, p. 212. 16. Lewis and Hua, “China’s Ballistic Missile Programs,” p. 23. 17. Medeiros, “Evolving Nuclear Doctrine,” in Bolt and Willner, China’s Nuclear Future, p. 64.

china 127

missiles would have to hit in order to neutralize China’s nuclear deterrent was judged to be prohibitively high—­thereby imparting China, even by 1965, an assured (if delayed) nuclear retaliatory capability.18 The organization charged with strategic missions was the Second Artillery, reportedly created by Premier Zhou Enlai on July 1, 1966.19 To ensure that China had a second-­strike capability that survived an opponent’s attempt at a disarming first strike, China and the Second Artillery relied on numerical ambiguity, the dispersion of assets, concealment, and basing modes to render an adversary sufficiently uncertain about its ability to fully destroy China’s nuclear forces and thereby render it vulnerable to a retaliatory strike with megaton-­yield warheads. In terms of ambiguity, while China’s overall capabilities were largely transparent as a result of its testing sequence, the numbers and deployment of missiles and warheads was intentionally left vague to instill uncertainty about whether an opponent could fully disarm China. For example, although open-­source literature suggests that China currently has roughly 20–­24 DF-­5s, the risk of an accounting error of even a single DF-­5 could leave, say, Los Angeles vulnerable to a retaliatory strike with an enormous 5 MT warhead (absent missile defenses). Similarly, the Second Artillery relied upon multiple layers of dispersion and de-­mating to enhance survivability of a full system. Warheads are kept at warhead bases which are separate from missile brigades, which themselves are widely dispersed and deployed in silos or caves for survivability by increasing the number of hard targets that would have to be destroyed in order to fully disarm China’s nuclear capabilities.20 Indeed, though the DF-­5 was deployed in silos of increasingly questionable survivability in the 1980s, the DF-­4 was made road-­mobile and deployed in high mountain caves to enhance survivability, for what the Central Military Commission (CMC) termed “in-­cave storage/preparation and out-­cave erection/filling(fueling)/firing” or what Zhang Aiping colorfully called “shooting a firecracker outside the front door.”21 To enhance the survivability of the DF-­5, the Chinese built a significant number of dummy silos, only a small fraction of which contained actual DF-­5 missiles, thereby increasing the requirements for a successful U.S. first strike.22 Warheads were stored separately from these missiles, which enhanced the survivability 18. Richelson, Spying on the Bomb, chapter 4; also see Walt W. Rostow, “The Bases for Direct Action against Chinese Nuclear Facilities,” Department of State Policy Planning Memo, April 22, 1964 (Declassified) available at http://www.gwu.edu/~nsarchiv/nukevault /special/doc09.pdf. 19. Gill, Mulvenon, and Stokes, “The Chinese Second Artillery Corps,” p. 517. 20. Ibid., pp. 544–­45. 21. Lewis and Hua, “China’s Ballistic Missile Programs,” p. 24; Zhang quoted in ibid., p. 24. 22. Lewis and Hua, “China’s Ballistic Missile Programs,” p. 25.

128

Chapter 5

of a fully assembled delivery vehicle (a strike that hit a ballistic missile would not take out a warhead with it), but it also required the missiles to be serviced by warhead units in the field. Nevertheless, these dispersion and deception procedures enabled China’s Second Artillery to establish a plausibly survivable retaliatory capability that would be difficult for an adversary to disarm with confidence. This was true even as early as the 1969 Zhenbao Island crisis, which escalated after China ambushed Soviet forces. Although the USSR contemplated striking a Chinese nuclear facility in Xinjiang, it possessed neither the intelligence nor the forces necessary to reliably execute a disarming preemptive strike against all Chinese nuclear forces, particularly an unknown number of dispersed warheads and DF-­2 missiles which might survive any Soviet preemptive attempt.23 China’s command-­and-­control procedures during the early phases were highly centralized, with the Chairman of the CMC having direct control over Second Artillery launch brigades and battalions. Lewis and Xue write: All launch battalions of the Second Artillery are under the ultimate direction of the CMC and subject to its decisions on procurements, launch-­site construction, deployment and employment, and exercises. Those decisions pass through the base headquarters, but never originate in those headquarters . . . In some emergencies, the CMC and missile headquarters may bypass the base and brigade headquarters and issue launch orders directly to those battalions . . . The center has devised special codes for preventing unauthorized or accidental launches and multistep procedures to transmit and verify orders. A launch will automatically be aborted if any step violates the verification requirements, and several steps depend on the coordinated action of at least two authorized officers.24

Only the Chair of the CMC (to date Mao Zedong, Deng Xiaoping, Jiang Zemin, Hu Jintao, and Xi Jinping) is believed to be able to authorize any nuclear activity, let alone the release of nuclear weapons. Any such authorization would be transmitted to the appropriate bases or directly to the launch battalions following a first use of nuclear weapons against China. This highly centralized control over nuclear weapons is magnified by the additional constraint that missile bases and warhead bases reportedly cannot communicate with each other, which enhances negative controls over 23. See Yang Kuisong, “The Sino-­Soviet Border Clash of 1969: From Zhenbao Island to Sino-­American Rapprochement,” Cold War History, vol. 1, no. 1 (August 2000), pp. 31–­41. Though Mao was sufficiently afraid of a Soviet nuclear strike, going so far as to evacuate the leadership from Beijing into bunkers in August 1969, he claimed that if the USSR were to use nuclear weapons, “Moscow also needed to remember that China possessed its own nuclear bombs and could take terrible revenge.” See Yang, p. 36. 24. Lewis and Xue, Imagined Enemies, pp. 198–­99.

china 129

China’s arsenal, preventing warhead bases from servicing anything but the units they are supposed to.25 These institutional structures limit the risk of unauthorized or accidental launch (given the steps required to prepare an unmated liquid-­fueled ballistic missile for launch, an accidental launch is of minimal concern during most contingencies). It is unknown whether China ever employed PALs to prevent unauthorized use. Nonetheless, given the highly professional nature of the Second Artillery, a 1999 U.S. National Intelligence Council report assessed “that an unauthorized launch of a Chinese strategic missile is highly unlikely,” though how they arrived at that assessment is unclear.26 To establish positive control over nuclear weapons, it is believed that the CMC and the Second Artillery Headquarters in Xishan, outside Beijing, have preset “plays” with a set of numbered cards at each of the reportedly six Second Artillery bases corresponding to a particular launch sequence and target set.27 The Second Artillery Headquarters and CMC had hardened bunkers to be able to ride out a nuclear attack, but it is unclear how robust and hardened the communication capabilities were and whether they would have been able to communicate with bases and launch battalions in a scintillated nuclear environment. There are some suggestions that if a crisis were to develop, the plays and authorization codes would be delegated to the appropriate units and they would be ordered to retaliate after absorbing a first-­wave nuclear attack, thereby assuring retaliation while the number of people required to authorize each step in the field would prevent unauthorized launch. China’s no-­first-­use (NFU) claim following its first nuclear test in 1964 is often derided as a rhetorical canard. However, China’s capabilities and procedures make its no-­first-­use policy not only credible but the central feature of its assured retaliation posture. The time it would take after CMC Chair authorization to arm and prepare a ballistic missile for launch could be detectable by an opponent. Given the far quicker U.S. and Soviet launch timelines, any attempt at nuclear first use by China would expose the locations of Chinese nuclear forces and enable a potentially disarming strike. The slow Chinese timelines only make sense in a retaliatory mission. Additionally, launch under attack (LuA) or launch on warning (LoW) postures are not possible for China, given the absence of sophisticated early warning systems. China could not launch on warning and it did not have the capability to discriminate between a conventional and nuclear missile attack in flight, so a LuA posture was precluded. As a result, as Lewis and 25. Ibid., p. 19. 26. National Intelligence Council, Foreign Missile Developments and the Ballistic Missile Threat Through 2015, September 1999, p. 11. 27. Lewis and Xue, Imagined Enemies, p. 206.

130

Chapter 5

Xue write, “the CMC was forced to choose between launch on warning or waiting out the actual hit. It chose the latter. Missile units would only be allowed to ‘launch the first counteroffensive after the first nuclear explosions but before the second-­wave nuclear strike’.”28 Thus, in this period, Mao’s original NFU posture is credibly reflected in the capabilities China chose to develop. By choosing to develop large city-­busting warheads delivered by a few inaccurate but sufficiently survivable ballistic missiles which required a significant time to prepare for launch; by choosing not to develop meaningful early warning architectures; and by instituting highly centralized command-­and-­control procedures which additionally increased the time to retaliate but which served as powerful negative controls on the nuclear arsenal, China was in no position to adopt anything but an assured retaliation posture. It could not threaten nuclear use in most conventional contingencies, but by credibly threatening to eliminate soft counterforce and eventually major strategic targets should an opponent use nuclear weapons first, China aimed to deter nuclear coercion and nuclear use. These capabilities and command-­and-­control procedures suggest that Mao’s aphorisms which, in toto, describe an assured retaliation posture, were at the heart of China’s operationalization of its nuclear weapons capability after 1964. Whether China’s forces were completely invulnerable to a disarming first strike from either the United States or USSR is impossible to test, but the CMC and the Second Artillery went to great lengths to try to ensure the survivability of its limited land-­based nuclear forces by relying on ambiguity, concealment, deception, and increasing mobility. All of this—­an assured, but delayed, retaliatory capability—­was achieved without a clear articulation of China’s nuclear doctrine through the mid-­1980s or without systematic thinking on the requirements to achieve a credible second-­strike capability, beyond “general notions of survivability.”29 Evan Medeiros writes that, in this period, “China’s nuclear doctrine was largely defined by the general beliefs of Mao Zedong and Deng Xiaoping about the need to possess a basic [assured] retaliatory capability to deter an adversary from using or threatening to use nuclear weapons against China.”30 Furthermore, until the 1980s, Chinese leaders placed “multiple organizational and political constraints on the ability of the People’s Liberation Army to develop nuclear strategy and an associated operational doctrine.”31 By the mid-­1980s, after Deng’s effort at greater openness and his revision of the PLA’s missions,32 PLA writings on nuclear doctrine began to develop. 28. Ibid., p. 210. 29. Medeiros, “Minding the Gap,” p. 146. 30. Ibid.; also see Fravel and Medeiros, “China’s Search for Assured Retaliation.” 31. Fravel and Medeiros, “China’s Search for Assured Retaliation,” pp. 51–­52. 32. Medeiros, “Evolving Nuclear Doctrine,” p. 51.

china 131

Debates emerged about what Johnston identifies as “limited deterrence” (youxian heweishe), as distinct from “minimum deterrence,” as early as 1987.33 This coincided with Chinese advances with solid-­fuel missiles such as the medium-­range DF-­21 and sea-­launched JL-­1, as well as the initiation of the DF-­31 project which aimed to give China a solid-­fuel ICBM to shorten launch times against U.S. homeland targets. Johnston identified the limited deterrence posture as requiring “enough capabilities to deter conventional, theater, and strategic nuclear war, and to control and suppress escalation during a nuclear war”—­essentially an asymmetric escalation posture that envisions both first-­use capabilities to deter conventional war as well as a second-­strike reserve to maintain escalation dominance in the event of general nuclear war.34 Abandoning, or at least conditioning, China’s no-­first-­use policy would be necessary to move from a minimum to a limited deterrence posture. Though Johnston identified strong threads of this debate in Chinese military and strategic writings from the mid-­1980s through mid-­1990s, he concludes that the gap between such a doctrine and China’s capabilities at the time was so large (few tactical nuclear weapons, solid-­fuel theater capabilities, early warning architecture), that a limited deterrence posture at the time was little more than a wish list of capabilities. Since Johnston’s article, new PLA writings have emerged that suggest that what was termed the limited deterrence strategy was actually aimed at systematically establishing an assured retaliation capability against a U.S. strategic force that included increasingly accurate ballistic and cruise missiles (which could potentially disarm China in a first strike) and proposed strategic missile defense systems (which could intercept the few Chinese nuclear forces that might survive an attempted first strike).35 That is, in this period, Chinese strategists recognized that they had to improve their posture from plausible retaliation to assured retaliation given the evolution of American forces. Some strategists may have advocated an evolution of Chinese posture away from a second-­strike capability toward a more robust “superpower-­style” posture that aimed to deter a range of conflicts from the conventional to the nuclear. Nevertheless, the capabilities and command-­and-­control architecture China has developed in the past twenty years—­and more importantly what it has not developed—­suggest that China’s strategists and political leaders have systematically and dynamically developed what they believe are the required forces to assure retaliation 33. See Johnston, “China’s New ‘Old Thinking’,” pp. 18–­19. 34. Ibid., pp. 19–­20. 35. See Medeiros, “Minding the Gap”; Chase and Medeiros, “China’s Evolving Nuclear Calculus: Modernization and Doctrinal Debate”; Allen and Kivlehan-­Wise, “Implementing PLA Second Artillery Doctrinal Reforms”; Yuan, “Effective, Reliable, and Credible: China’s Nuclear Modernization”; Gill, Mulvenon, and Stokes, “The Chinese Second Artillery Corps.”

132

Chapter 5

against a more capable U.S. strategic force. The basic posture, according to almost all China scholars, has not changed. Indeed, Medeiros studied almost a dozen authoritative PLA writings since 1996 including several “campaign outlines” (zhanyi gangyao) specific to the Second Artillery and concludes: “Chinese military writings have identified only one mission for Second Artillery nuclear forces, i.e., ‘a nuclear counterstrike campaign,’ in connection with which they discuss conducting retaliatory nuclear strike operations . . . China does not seek a capability for assured destruction, but rather assured retaliation.”36 The term that is now emerging to describe this posture, which Medeiros and Johnston identify as relatively new in the Chinese strategic lexicon and which first appeared in the 1987 PLA publication Zhanluexue, is “effective nuclear deterrence” (youxiao heweishe). Medeiros defines this term as a nuclear force that is “reliable, survivable, and can penetrate an adversary’s missile defenses. As long as the PLA is assured that it can retaliate and impose unacceptable damage on an adversary following a nuclear attack then its deterrent is assessed to be effective.”37 In 2006, China’s National Defense White Paper noted that China’s aim was to have a “lean and effective nuclear deterrent capability” (jinggan youxiao he liliang) while the 2008 White Paper specified that the fundamental mission of China’s nuclear forces is to protect “China from any nuclear attack.”38 The key Second Artillery campaign guiding thoughts that underlie this posture are “close defense” (yanmi fanghu) which focuses on survivability of forces so that a first strike cannot disarm China and “key point counterstrikes” (zhongdian fanji) which focuses on the subsequent retaliatory strikes against both strategic centers and soft-­counterforce targets to degrade the adversary’s ability to follow with a second wave of nuclear strikes.39 These terms are endur­ing operational principles that have been constant since the Mao years and were reaffirmed by Jiang Zemin’s “five musts” for China’s nuclear arsenal.40 In fact, since Johnston first identified the limited deterrence debate, the scholarly consensus is that the definition of limited deterrence in Chinese writings may deviate significantly from American understandings of the term which centers on nuclear weapons as war-­fighting, not war-­deterring, instruments. This illustrates the dangers of superimposing Cold War lexicon 36. Medeiros, “Minding the Gap,” p. 150. Emphasis in original. 37. Ibid., pp. 153–­54. 38. Ibid., p. 151; and Information Office of the State Council of the People’s Republic of China, China’s National Defense in 2008, p. 29. Available at http://english.gov.cn /official/2009‐01/20/content_1210227.htm. 39. Medeiros, “Minding the Gap,” pp. 152–­53; also see Medeiros, “Evolving Nuclear Doctrine,” p. 63. 40. See a discussion of the five-­musts in Gill, Mulvenon, and Stokes, “The Chinese Second Artillery Corps,” p. 548.

china 133

upon regional power nuclear strategies. Chase and Medeiros conclude that the writings in Zhanluexue, the Academy of Military Science’s publication on strategy, indicate: The concept of . . . limited deterrence demands only that the country maintain retaliatory forces sufficient to inflict unacceptable damage (wufa chenghsou de pohuai) on its adversary. This characterization of the concept of limited deterrence appears to depart from the standard Western definition of limited deterrence. Zhanluexue’s characterization of the concept of limited deterrence is seemingly more consistent with what Gill et al term “credible minimum deterrence.” . . . Turning to the practice of nuclear deterrence, the editors write that the most important objective is using deterrence to avoid nuclear war.41

Thus, the debate about limited versus minimum deterrence in the Chinese literature may have resolved itself to be less about a shift in nuclear posture than about how to make China’s assured retaliation posture credible against increasingly capable opponents (first the Soviet Union, then the United States after the end of the Cold War). Since the end of the Cold War, China has not become more reliant on its nuclear weapons force to deter conflict, nor has it developed tactical nuclear capabilities that could credibly threaten the first use of nuclear weapons against adversaries. Indeed, according to Medeiros, there is “no evidence of a Chinese discussion [about] using nuclear weapons as a cost-­saver or substitute for conventional weapons; indeed China’s intense conventional build-­up in recent years belies the very notion. There is very little, if any, substantive discus­sion about using tactical nuclear weapons to deter major conventional aggression against China . . . there appears to be little interest in Cold War-­style ‘nuclear war­ fighting’ strategies in which nuclear exchanges can be calibrated and finely managed.”42 Though the terminology used to describe China’s nuclear posture from the mid-­1980s on is certainly plagued by inconsistency and ambiguity particularly when translated into English, the posture itself has consistently been one of assured retaliation. Improvements in capabilities and command and control have focused on two levels. First, at the nuclear level, the PLA and Second Artillery seem to have become acutely aware of their nuclear forces’ vulnerability to the twin threats of accurate American first-­ strike capabilities, which forced them to think harder about survivability, and American missile defense architectures, which has forced them to think about effective penetrability and warhead/MIRV modernization. Second, 41. Chase and Medeiros, “China’s Evolving Nuclear Calculus,” p. 138; also see Yuan, “Effective, Reliable, and Credible,” p. 275. 42. Medeiros, “Minding the Gap,” p. 158.

134

Chapter 5

at the conventional level, the Second Artillery has taken custody of a large number of conventional ballistic missiles with non-­nuclear strategic roles to be used as war-­fighting instruments in a local theater conflict. This conventional modernization significantly affects China’s security environment, imparting to it the capability to engage conventional conflicts with conventional power and neutralizing any potential existential threats (to the mainland or Taiwan), and thereby obviating the need to rely on nuclear weapons to deter a conventional attack. I focus here on the nuclear forces and why they still constitute an assured retaliation posture. The capabilities China has attempted to evolve since the 1990s all conform to the broad pattern of assuring that a retaliatory nuclear force survives an attempted first strike, particularly from the United States. First, China is increasing the survivability and reliability of its delivery vehicles. This has focused on two key legs: moving away from corrosive liquid-­fuel MRBMs and ICBMs, which are difficult to move since the necessary logistics complement is large and detectable, toward solid-­fuel MRBMs and ICBMs that are more mobile and can be launched without a 2–­3 hour lead time; and making advances in sea-­based capabilities by developing a less problem-­ridden SSBN (the Type 094 Jin-­class submarine with the upgraded JL-­2 SLBM) that would assure survivable forces assuming a sufficient number were available and quiet enough to have at least one on continuous patrol, particularly during a crisis.43 These developments, according to Chinese analysts, will “fundamentally ensure the reliability and credibility of China’s nuclear force.”44 The mobility of its land-­based forces will enable further concealment and deception in rotating silos and caves, while an operational sea leg could potentially be highly survivable depending on how quiet Chinese SSBNs become. However, as of 2012, it is estimated that only a handful of road-­mobile solid-­fuel DF-­31s and only about 75–­100 DF-­21 MRBMs have been operationalized. The Jin-­class submarine is still not operational—­though it eventually should be, and will impart China a truly survivable force—­and the lone Xia SSBN is not believed to be reliably operable and, indeed, is judged to be a “colossal failure.”45 It is not clear how many DF-­21s/DF-­31s and Jin-­class submarines China intends to ultimately deploy, but U.S. analysts 43. See Hans Kristensen, Robert S. Norris, and Matthew G. McKinzie, Chinese Nuclear Forces and US Nuclear War Planning (Washington, DC: Federation of American Scien­ tists  & National Resources Defense Council, November 2006), chapter 2; and Office of the Secretary of Defense, Annual Report to Congress: Military Power of the People’s Republic of China, 2009. 44. Wang Zhongchun quoted in Chase, Erickson, and Yeaw, “Chinese Theater and Strategic Force Modernization,” p. 71. 45. Ibid.; also see Robert S. Norris and Hans Kristensen, “Chinese Nuclear Forces 2011,” Bulletin of the Atomic Scientists, vol. 67, no. 6 (November/December 2011), pp. 81–­87.

china 135

have consistently claimed that China only deploys about 240 warheads to complement a force structure of roughly 140 nuclear-­capable ballistic missiles and keeps the rest of its stockpile in reserve.46 There is no evidence that the 100 or so lower yield nuclear weapons in China’s arsenal are deployed for tactical purposes, and the SRBMs in its inventory are all judged to be strictly for conventional roles. Even if the DF-­31 ICBM augments, rather than replaces, the old silo-­based DF-­5, China’s forces would only be capable of executing retaliatory strikes. China’s efforts at improving the survivability of its nuclear forces have not included building the same number of systems as the Cold War superpowers. The second component of China’s so-­called modernization program focuses on effectively penetrating missile defenses. The United States has judged that China has had Multiple Re-­entry Vehicle (MRV) and Multiple Independently Targetable Re-­entry Vehicle (MIRV) technology for more than twenty years, but has chosen not to convert its DF-­5s to multiple warheads.47 In fact, it is not even clear that the DF-­31 will be MIRV’ed, though China may have the technical capability to do so. Instead, China relies on the ambiguity of its MIRV capabilities and the use of “decoys and penetration aids aimed at confusing defensive systems” to induce uncertainty that a missile defense architecture could blunt China’s retaliatory forces.48 Li Bin and colleagues have written that the focus will likely be on penetration aids and decoys since they “are inexpensive and they have a low political cost. Further, it is technically unlikely that a U.S. defense system would ever work so well that it could sort out penetration aids from warheads. That would ensure that China’s retaliatory force would remain viable.”49 According to Michael Chase and colleagues, the aim of these capability improvements is to give China “perhaps for the first time, a highly survivable strategic nuclear force.”50 In terms of command and control, Chinese leaders and the PLA have invested significant resources to harden their communications systems, developing redundancy and alternate command centers, and instituting stronger positive controls so that the CMC “could survive an attack and respond” with an assured, but perhaps delayed, retaliatory strike.51 But the central features of Chinese command and control appear to persist: assertive control under the direct authority of the Chairman of the CMC, dispersion and 46. Norris and Kristensen, “Chinese Nuclear Forces 2011.” 47. Ibid., pp. 60–­61. 48. Ibid., p. 61. 49. Li Bin, Zhou Baogen, and Liu Zhiwei, “China Will Have to Respond,” Bulletin of the Atomic Scientists, vol. 57, no. 6 (November/December 2001), p. 28. 50. Chase, Erickson, and Yeaw, “Chinese Theater and Strategic Force Modernization,” p. 85. 51. Gill, Mulvenon, and Stokes, “The Chinese Second Artillery Corps,” p. 552.

136

Chapter 5

de-­mating of nuclear system subcomponents, multiple steps to constitute a fully ready system with multiple people requiring authenticated orders as the central negative controls, and the lack of horizontal communication between Second Artillery subunits (i.e., a base cannot launch the missiles of another base). Indeed, a 2010 report by Mark Stokes suggests that China has a system of centralized warhead depots that services the other six or so warhead bases in order to maximize assertive control over the arsenal: The Chinese Communist Party’s Central Military Commission (CMC) maintains strict control over China’s operational nuclear warheads through a centralized storage and handling system managed by the People’s Liberation Army (PLA) Second Artillery. Nuclear warheads are granted special consideration due to their political significance and potential consequences of an accident, incident, or unauthorized use. As a result, warheads are managed in peacetime through a system that is separate and distinct from Second Artillery missile bases and subordinate launch brigades. Second Artillery nuclear warheads also appear to be managed separately from China’s civilian nuclear industry, and fissile material control and accounting system. In addition, the Second Artillery appears to control and manage nuclear warheads that could be delivered by other services, such as the PLA Air Force and Navy.52

Robert Norris and Hans Kristensen further report that “under normal circumstances, China keeps its warheads in storage rather than mated with delivery systems.”53 They further note that these warheads are “under the control of the Central Military Commission and would [only] be released to the Second Artillery Corps ‘if China comes under nuclear threat.’ ”54 Communications technology between the Beijing-­based leadership and an eventual SSBN deployed in blue waters poses a challenge, since China is not yet believed to have hardened VLF (very low frequency) and ELF (extremely low frequency) communications capabilities required to communicate with a submerged submarine.55 How Chinese authorities handle command and control with the Jin-­class submarines will be a key test of its commitment to centralized control—­either they will have to cede authority and assets to the PLA Navy or they may be forced to limit deterrent patrols to keep the SSBNs closer to shore and under firm centralized command (as the Soviets were believed to do as well). If China’s commitment 52. Mark Stokes, China’s Nuclear Warhead Storage and Handling System, Project 2049 Institute Monograph, March 12, 2010, p. 2. 53. Norris and Kristensen, “Chinese Nuclear Forces 2010,” Bulletin of the Atomic Scientists, vol. 66, no. 6 (2010), p. 135. 54. Ibid., p. 138. 55. Gill, Mulvenon, and Stokes, “The Chinese Second Artillery Corps,” pp. 34–­36.

china 137

to centralized control overrides all other considerations, the latter method of command and control would seem to be more likely. Most critically, however, the architectural features that China would need in order to move away from a retaliatory posture do not appear to be in development. In particular, China has not yet invested in space-­based early warning capabilities that would enable it to move to a launch on warning (LoW) or launch under attack (LuA) posture. While there was a vigorous debate about conditioning or altogether abandoning China’s no-­first-­use policy, as Johnston detected in Second Artillery writings, that debate was, according to Medeiros “resolved in favor of not altering it” for several reasons.56 First, Liu Huaqiu, a Chinese strategist, writes that, given China’s inferior nuclear forces, the first use of nuclear weapons against the United States would be tantamount to suicide: “China has developed a very small nuclear arsenal, and this is a reflection of China’s no-­first-­use policy. If China should use nuclear weapons first, the adversaries would retaliate with many more nuclear weapons, which could in effect be self-­destruction for China.”57 Second, China simply does not have the early warning capability to distinguish between an incoming conventional versus nuclear attack, so China would only be able to determine that it was under nuclear attack after the first warheads detonated. Hence, Jing-­Dong Yuan writes that the Second Artillery is strictly ordered to “delay a counterattack until after it had absorbed the incoming nuclear strike.”58 Indeed, the Second Artillery simulates its counter­ attack exercises in harsh environments to come as close to replicating the conditions—­without an actual nuclear detonation of course—­under which it would have to launch retaliatory strikes.59 This suggests a persistent, and costly, commitment to a strict no-­first-­use policy in which Chinese nuclear forces would only be released after physically absorbing an opponent’s nuclear attack. But there is a strong emphasis in Zhanluexue, according to Chase and Medeiros, on the ability to retaliate quickly after this first strike and against both countervalue (punitive strikes) and soft counterforce targets as a damage-­limitation measure to degrade the opponent’s ability to follow with a second wave of nuclear attacks.60 This timing would call for immediate retaliatory strikes and require higher readiness forces than a response that allows for a several-­hour delay, but it need not result in a capability that abandons the no-­first-­use policy. 56. Medeiros, “Minding the Gap,” p. 156. 57. Liu Huaqiu, “No-­First-­Use and China’s Security,” Eliminating Weapons of Mass Destruction: Electronic Essays, Stimson Center, October 1998, reprinted in Chase and Medeiros, “China’s Evolving Nuclear Calculus,” p. 141. 58. Yuan, “Effective, Reliable, and Credible,” p. 288. Emphasis added. 59. Chase and Medeiros, “China’s Evolving Nuclear Calculus,” pp. 144–­45. 60. Ibid., p. 146.

138

Chapter 5

My characterization of Chinese nuclear posture emphasizes the endurance of an assured retaliation posture. Since 1964, China has aimed to possess the capabilities and procedures required to survive an opponent’s first strike and be able to retaliate with terrifyingly large-­yield nuclear forces as a means to deter nuclear coercion and nuclear attack. The capabilities China developed from 1964 through the mid-­1980s may have barely met the threshold for a fully credible assured retaliation capability, but by relying on numerical ambiguity, deception, and concealment, a Soviet or American opponent could not be confident that it could achieve a disarming first strike. By the mid-­1980s, as the PLA began to think more systematically about the requirements for survivability and effectiveness against American forces especially, the capabilities China calculated it required in order to achieve an assured retaliation capability have increased. In particular, facing a United States whose surveillance and intelligence assets were increasingly capable of locating Chinese nuclear forces, whose conventional and strategic forces were increasingly capable of striking those forces, and the specter of strategic missile defenses which might be able to intercept any surviving forces, Chinese military writings emphasized the growing vulnerability of their nuclear forces to a disarming first strike. Though there seems to have been an increased call for what Chinese strategists called limited deterrence, the notion as it evolved looked less like U.S. nuclear warfighting doctrine and more like acquiring the necessary capabilities to assure a survivable and effective retaliatory limited-­quantity force against the United States to deter conflict at strictly the nuclear level. Indeed, the slow pace of Chinese nuclear force improvements and the capabilities and command-­and-­control architectures that China has both invested and not invested in gives deep insights into its ultimate aim. By emphasizing mobility for survivability and countermeasures for penetration, China today seems to still be seeking the minimum means to assure the ability to absorb a nuclear attack from the United States and still be able to retaliate with ferocious punitive strikes against cities and degrading attacks against U.S. military nodes to prevent further nuclear attacks. In doing so, China’s assured retaliation posture from Mao through the present and at least the foreseeable future has simply aimed to deter nuclear coercion and nuclear attack by the USSR and United States.

Explaining the Endurance of Assured Retaliation in China Like India, China has also adopted an assured retaliation posture, and largely for the same broad reason. China priveleges a highly assertive civil-­ military structure that is resistant to ceding assets and authority to the military. After the break with the USSR in the late 1950s, China lacked a

china 139

third-­party patron that might enable the selection of a catalytic posture in the event of conflict. Second, China has not faced an existential proximate offensive threat, relying on its territorial and manpower advantage to adopt a conventional “defense-­in-­depth” strategy that could meet even a large Soviet land threat; and relying on the Pacific Ocean to blunt any serious existential threat from the United States. Thus, relative security has granted China the latitude to choose between the assured retaliation and asymmetric escalation postures. While much scholarship emphasizes China’s resource limitations in its development of a small nuclear force, I show that just as important—­if not more so—­was Mao Zedong’s (and subsequent General Secretaries’) emphasis on assertive centralized control over the military, which precluded the selection of an asymmetric escalation posture. China did not use nuclear forces as a means to reduce its reliance on conventional forces, as proponents of the resource limitation argument would have to sustain. Instead, after nuclearization, China continued to maintain large conventional forces and has invested tremendous amounts in improving its conventional military capabilities. As in the case of India, it is an assertive civil-­military relationship that has driven China’s choice of an assured retaliation posture. This understanding of what motivates China’s choice of nuclear posture illuminates why even a limited deterrence posture is still essentially a posture of assured retaliation. I show that optimization theory better captures China’s choice of posture than structural realism, technological determinism, or strategic culture theories. The puzzle of why China adopted a characteristically different nuclear force posture from its primary adversaries—­the USSR and then the United States—­has attracted much attention. Some scholars, such as Lewis and Xue, have averred that China’s nuclear choices were a function of technological determinism, that China adopted the posture that its technical capabilities enabled and articulated post hoc rationalizations for it. Others have argued that China’s minimalism is a result of the enduring legacy of Maoist strategic culture that viewed China as a resource-­constrained state and that only a few nuclear weapons were sufficient to achieve China’s security aims.61 In this section I lay out my predictions from the more general optimization theory. I argue that China’s security environment, and particularly its massive conventional force and vast territory gave China the latitude to opt out of a demanding first-­use asymmetric escalation posture if it so desired. According to my theory, the feature that pushed China toward an assured retaliation posture then was an assertive civil-­military relationship that drove Mao and subsequent General Secretaries to assert strong centralized control over China’s nuclear forces, which precluded a first-­use posture that would require high levels of delegation to military end 61. See Fravel and Medeiros, “China’s Search for Assured Retaliation.”

140

Chapter 5

users. Thus, an assured retaliation posture optimized China’s nuclear force structure given its external and internal security constraints. The prediction of optimization theory requires an evaluation of several variables. First, did China have a third-­party patron that could be compelled to intervene in the event of a conventional attack, possibly enabling China to select a catalytic nuclear posture? For a while it seemed as if the Soviet Union might be a potentially reliable third-­party patron against American threats. However, before China even got its nuclear program off the ground, the Sino-­Soviet split was well under way. Indeed, the retraction of Soviet assistance for the Chinese nuclear program itself was a key indicator to Mao that the ideological and geopolitical tensions that arose between him and Khrushchev were reaching irresolvable levels. Lorenz Luthi writes that “Mao eventually pushed for the collapse [of the alliance] after 1959, when he decided that it had run the full course of what he considered its usefulness to the country.”62 After the split, China was left to face the United States and USSR by itself and therefore lacked a third-­party patron that could enable it to implement a catalytic nuclear posture. Even after the United States normalized relations with China, the United States could not have been characterized as a reliable third-­party patron. At the second node, was China ever facing a conventionally superior proximate offensive threat? The obvious threat of this type against China was posed by the USSR. The USSR had clear nuclear superiority over China from 1950 on, but did the USSR pose an existential conventional threat to China? Against almost every other country on its border, the USSR was an imposing behemoth that could effectively exercise its will without much resistance. But against China, which had a larger land army and plenty of strategic depth along its border with the USSR, the Soviet Union did not pose an existential threat.63 Though China had to split its 2.5 million strong PLA between fronts in the Far East and Vietnam in the 1960s–­ 1980s, China always had a 2:1 quantitative theater advantage over the Soviet Union even after the USSR repositioned forces to the Far East before and subsequent to the Zhenbao Island clashes in 1969.64 Given long-­ standing evidence that “the defender can inflict more casualties than the

62. Lorenz M. Luthi, The Sino-­Soviet Split: Cold War in the Communist World (Prince­ ton: Princeton University Press, 2008), p. 2. 63. See Fravel, Strong Borders, Secure Nation, pp. 204–­208; Lyle J. Goldstein, “Return to Zhenbao Island: Who Started Shooting and Why It Matters,” China Quarterly, no. 168 (December 2001), pp. 991–­94; and Thomas W. Robinson, “China Confronts the Soviet Union: Warfare and Diplomacy on China’s Inner Asian Frontiers” in Roderick MacFarquhar and John K. Fairbank, eds., The Cambridge History of China, Volume 15 (Cambridge: Cambridge University Press, 1992), pp. 297–­99. 64. Fravel, Strong Borders, Secure Nation, pp. 204–­8.

china 141

Year

Soviet Divisions in the Far East

Chinese Divisions on Sino-­Soviet Border

1964

14

47

1965

15

47

1966

21

47

1967

26

47

1968

29

47

1969

31

47

1970

37

47

1971

40

47

1972

44

49

1973

45

65

1974

45

70

1975

43

78

1976

46

78

Table 5.1. Chinese and Soviet land force deployments on the Sino-­Soviet border in the Far East

attacker,”65 that 2:1 force advantage was more than sufficient to thwart a Soviet invasion, even if the quantitative differential was offset by superior quality Soviet materiel. Table 5.1 shows the number of Soviet and Chinese divisions deployed in the Far East (these divisions are of equivalent size, enabling direct comparison).66 The Chinese responded to increased Soviet deployments in the Far East by shifting their own forces to the theater, but China was never in danger of being overrun by Soviet land forces. Certainly, Soviet forces had superior equipment. In particular, Soviet attack aircraft were of significantly higher quality than their Chinese equivalents. Nonetheless, the sheer Chinese man­ power advantage combined with a sufficient degree of strategic depth posed a daunting conventional deterrent to a Soviet invasion. In­deed, even at the height of tensions during the Zhenbao clashes, Zhou Enlai argued that “it was impossible for the Soviets to launch a large-­scale invasion of China in the near future, as they still had a long way to go before they could turn the relatively undeveloped Soviet Far East into the bases for [successfully]

65. John J. Mearsheimer, “Assessing the Conventional Balance: The 3:1 Rule and Its Critics,” International Security 13, no. 4 (Spring 1989), p. 57. 66. Data derived from Fravel, Strong Borders, Secure Nation, p. 205, and Robinson, “China Confronts the Soviet Union,” p. 299.

142

Chapter 5

attacking China.”67 Mao “felt that China had the advantage. China had many more soldiers, [such that] if they had broken through, the Russians would not have been able to rectify the situation.”68 The Soviets themselves, according to former Soviet military intelligence officers, believed that the “Chinese possessed colossal manpower superiority” and the Soviets were only able to achieve balance against the Chinese in the 1970s as the relative gap in force deployments narrowed.69 But in the 1970s, as the U.S. war in Vietnam drew to a close and the U.S.-­China rapprochement unfurled, China was able to further reposition quality divisions and materiel along the Soviet border to further enhance its conventional deterrent against the USSR.70 Thus, even against the bordering Soviet Union, China did not objectively face—­nor believe it was facing—­an existential conventional threat. Other possible threats to China—­ India since 1962, and the United States since 1989—­do not come close to posing existential conventional threats to the Chinese state. India and China are buffered by inhospitable terrain in their disputed territories, which are 3,000 miles from Beijing and Shanghai, and China has significant force advantages over Indian forces in theater. Against the United States, an American invasion of mainland China during the Cold War would have been unthinkable. Though China presently lacks the blue-­water navy force to compete with the American Navy in the East Asia Pacific theater, China’s coastal defenses and increasingly lethal short-­range conventional ballistic and anti-­ship missile capabilities can impose tremendous costs on American forces in a Taiwan contingency, let alone any U.S. attempt to invade China.71 The issue of whether Taiwan is an integral part of China is relevant here because if the United States could pose a threat to Taiwan, or midwife its independence, that could be construed by Beijing as an existential threat to the People’s Republic of China and push China toward an asymmetric escalation posture. However, for now China seems to have opted for a significant conventional modernization campaign to offset American capabilities in a potential Taiwan contingency, obviating the need to rely on nuclear weapons as a deterrent. Furthermore, U.S. forces in East Asia are not capable of mounting a land invasion of mainland China. China’s conventional forces, the distances over which the United States would have to project power, and China’s

67. Yang, “The Sino-­Soviet Border Clash of 1969,” p. 31. 68. Goldstein interview with Yang quoted in Goldstein, “Return to Zhenbao Island,” p. 993. 69. Colonel (Ret.) Victor Gobarev quoted in Goldstein, “Return to Zhenbao Island,” p. 994. 70. Fravel, Strong Borders, Secure Nation, pp. 216–­17. 71. See Office of Naval Intelligence, China’s Navy 2007; and Military Power of the People’s Republic of China, 2009.

china 143

territorial depth preclude a serious conventional existential threat from the United States. Thus, never facing a conventionally superior proximate threat, China was and is in a security environment that is insufficiently severe to force it to adopt a first-­use asymmetric escalation nuclear posture to deter conventional attacks against it. In my theory, when states have no third-­party patron to enable a catalytic posture and possess some security latitude, they can either choose an asymmetric escalation or assured retaliation posture based on additional variables. The next variable to consider in sequence is the nature of civil-­ military relations (or in China’s case, more accurately, party-­military relations). With China, there is little dispute that Mao initially developed, and subsequent General Secretaries have institutionalized, highly assertive military structures that placed China’s military under exceedingly centralized—­ effectively  per­sonal—­political control at all levels of warfare: strategic, operational, and even tactical.72 Here, I measure the character of China’s conventional civil-­military structures to predict what its nuclear command-­ and-­control structure might be. The prediction of optimization theory is that states with a highly assertive conventional civil-­military structure will neither entrust the military with nuclear weapons nor cede delegative authority over their release; that is, an asymmetric escalation posture is precluded in such cases. At the apex of China’s conventional military forces pyramid is the Cen­ tral Military Commission (CMC).Though its size has fluctuated dramatically over the years, the CMC has generally had about a dozen or so core members and is chaired by the General Secretary of the Communist Party of China. The CMC is the “unified command authority for all of China’s armed forces; it determines operational policy for military strategy and armed force; it leads and manages the PLA’s ‘army building’; develops plans; approves weapons development and purchases; determines the PLA’s organizational structure, missions and responsibilities; approves promotions and awards for senior officers; and coordinates the PLA’s budget with the State Council.”73 In short, the General Secretary of the Communist Party exercises firm and direct control over all of China’s conventional 72. See Harvey Nelson, The Chinese Military System: An Organizational Study of the Chinese People’s Liberation Army (Boulder: Westview Press, 1981), Chapter 1; David Shambaugh, “China’s post-­Deng Military Leadership” in James R. Lilley and David Shambaugh, eds., China’s Military Faces the Future (New York: ME Sharpe, 1999), pp. 11–­38; Ken Allen, “Introduction the PLA’s Administrative and Operational Structure,” in Mulvenon and Yang, eds., People’s Liberation Army, chapter 1; Nan Li, “The Central Military Commission and Military Policy in China,” in Mulvenon and Yang, eds., People’s Liberation Army, chapter 2; David Shambaugh, “The Pinnacle of the Pyramid: The Central Military Commission” in Mulvenon and Yang, eds., People’s Liberation Army, chapter 3. 73. Allen, “Introduction to the PLA’s Administrative and Operational Structure,” p. 7.

144

Chapter 5

forces and ensures loyalty by basing promotions on personal and party loyalty, monitoring activities with political officers at every level of the force structure, maintaining a dedicated elite security force to protect party members, regular rotations, and occasional purges. The CMC is not a “service representative” institution—­it is a service shaping and direction institution which exercises direct control over both high-­level state strategy and low-­level tactical decisions, as evidenced in its conduct during actual wars, including Korea and later the Zhenbao Island clashes. Indeed, in the China-­ Soviet dispute, Mao and Zhou Enlai personally directed the ambushes on March 15, 1969, from a suite in the Jingxi Hotel in Beijing with a dedicated phone line which established “direct communication with troops on the front.”74 Since Mao’s death, subsequent chairmen, have all asserted control over both the CMC and PLA by promoting only loyal officers and personnel.75 The consensus in the scholarship on Chinese party-­military relations overwhelmingly suggests that China’s conventional military forces—­at all levels—­are centrally directed by a small coterie of senior party leaders all loyal to the Chairman of the CMC. These structures arise from the nature of the Chinese political system, which views the PLA as an arm of the Communist Party, not the state, and needs to ensure the armed forces’ unwavering loyalty to the regime lest it poses a coup threat. Not facing a conventionally superior existential threat, China’s highly assertive conventional civil-­military relations suggest that the CMC Chairman would also assert extremely high levels of centralized control over China’s nuclear assets, pushing it toward an assured retaliation posture. In my theory, it is the unwillingness to pre-­delegate nuclear assets and authority to even elite Second Artillery forces, let alone to conventional PLA forces, that explains China’s persistent assured retaliation posture. Further, if correct, my theory would suggest that the Second Artillery would be a dedicated stovepiped strategic arm under the sole control of the CMC, and its Chairman in particular, to maximize assertive control over the arsenal. Further, the theory would anticipate substantial negative controls on nuclear assets. These would include asset de-­mating, a lack of horizontal communication between Second Artillery brigades and battalions, and political officers attached to each unit. It also provides a compelling political reason as to why Chinese leaders have chosen not to abandon the no-­first-­use policy. Moving to a launch on warning (LoW) or launch under attack (LuA) posture would likely force the Chairman to pre-­delegate or altogether emplace launch capability to the field. My theory predicts that the sacrifice of negative control that such a move would demand would be unacceptable to the Chinese party leadership given the nature of Chinese civil-­military 74. See Yang, People’s Liberation Army, pp. 29–­30. 75. Shambaugh, “The Pinnacle of the Pyramid,” pp. 109–­11.

china 145

relations. Finally, optimization theory suggests why China’s political leadership may have opted to augment its conventional capabilities, perhaps at greater financial cost, in order to deter American capabilities, rather than make the necessary nuclear command-­and-­control adjustments that moving to an asymmetric escalation nuclear posture would require. Moving to an asymmetric escalation posture that includes tactical or early use of nuclear weapons in a conventional conflict would require the cession of assets and launch authority to lower-­level officers. China’s conventional military structures suggest it would be wholly unwilling to pass over control or physical nuclear assets in this way. In this view, the assured retaliation posture is a function of China’s highly assertive command structure which overwhelmingly privileges centralized control and tight procedural negative controls on the stewardship and release of nuclear weapons. The assured retaliation posture is selected because it is easier to centrally manage, not because of resource constraints. In fact, as Medeiros notes, China’s massive conventional buildup belies the notion of resources significantly constraining China’s military program, and particularly its nuclear weapons program. China has the necessary infrastructure and existing capabilities to adopt a more aggressive first-­use posture if it chose to do so—­it keeps hundreds of warheads and significant amounts of weapons grade plutonium and uranium in reserve and the marginal cost of expanding its arsenal and erecting a limited early warning system would be expensive but not prohibitive given China’s economic power.76 Rather, if my theory is correct, China has chosen an assured retaliation posture because its security environment demands nothing more—­ with large conventional forces, China has been able to meet the array of conventional threats it faces—­and the centralized and assertive management advantages of an assured retaliation posture that deters threats and attacks at strictly the nuclear level are of primary importance to China’s leaders. That is, small nuclear forces are simply easier to control from Beijing. Whatever marginal security benefit that might be derived from an asymmetric escalation posture is outweighed by the management costs and cession of control that would be required by China’s leaders in order to move to such a posture. There are several alternative explanations for China’s assured retaliation posture, some prominently advanced in the scholarly literature. But none of them fully account for the nature and persistence of China’s assured 76. See Johnston, “Prospects for Chinese Nuclear Force Modernization,” p. 564. Based on NRDC estimates that China’s nuclear forces absorb 3–­5 percent of China’s military expenditures, Johnston concludes that it would not be prohibitively expensive for China to shift to a limited war-­fighting nuclear posture given that much of the investments have already been made and even if China needed to develop early warning systems.

146

Chapter 5

retaliation posture. The first alternative explanation is a strictly structural realist theory of posture selection. As my theory suggests, China’s security environment alone does not yield a determinate prediction for the type of nuclear posture it might have ultimately developed. Facing both the USSR and the United States through the Cold War and after, both superpowers with nuclear superiority, China could have selected several nuclear postures that would have been consistent with its external security environment. The only posture that China’s security environment excluded was the catalytic posture, since China had competitive relations with both the USSR and the United States and could rely on neither to assist it using the threat of nuclear escalation as a catalyst. But beyond that, its external security environment allowed for the selection of either an asymmetric escalation posture or an assured retaliation posture—­both postures would have been consistent with China’s security predicament. Though China did not face an existential Soviet threat, Soviet forces were sufficiently powerful that China could have chosen to increase its reliance on nuclear weapons to deter Soviet conventional threats. Such a posture might even have been more cost-­effective and perfectly consistent with China’s security environment. Instead, China chose to deter Soviet conventional aggression by augmenting its own conventional forces but deter Soviet and American nuclear coercion and attack by developing a small but effective assured retaliation nuclear posture, which is also consistent with its security environment. Thus, a strictly realist explanation cannot predict which one China would ultimately select. As a result, optimization theory, which then examines unit-­level intervening variables—­in this case assertive civil-­military relations which pushed China away from an increasing reliance on nuclear weapons as war-­fighting tools—­better captures the rationale and character of China’s ultimate choice of an assured retaliation posture. A second alternative explanation is provided by technological determinism. This argument would be that China developed capabilities first and then a post hoc strategic rationale for those capabilities and expanded the ambition of its doctrine as its capabilities developed and enabled it to do so. This explanation is most notably advanced by John Lewis, Xue Litai, and Hua Di in their works on China’s strategic weapons programs.77 As Lewis and Xue write, Mao and his successors adopted a de facto nuclear doctrine driven by the technologies available to them: “[The] emerging view decisively endorsed the military-­technical side of doctrine and posture to deter a nuclear attack. One might say the weapons, once deployed, spoke for themselves. In their silent vigils, megaton warheads could proclaim a 77. See Lewis and Xue, China Builds the Bomb; Lewis and Xue, China’s Strategic Seapower; Lewis and Xue, Imagined Enemies; Lewis and Xue, “Strategic Weapons and Chinese Power”; and Lewis and Hua, “China’s Ballistic Missile Programs.”

china 147

powerful new doctrine of assured, though limited, retaliation.”78 Lewis and Hua, writing about China’s ballistic missile programs, argue that “there is no evidence that any overarching strategic doctrine informed [Mao’s] decision to proceed with the strategic missile program . . . in the early years these programs were essentially technology-­driven and that only in the early 1980s did Beijing develop relevant strategic and tactical doctrines for its deployed and planned missile forces.”79 This hypothesis avers that China’s scientists and engineers developed the capabilities that they could and China’s posture flowed from that, not the reverse—­that Mao and subsequent Chinese leaders developed postural guidelines based on sound strategic logic and directed China’s military-­industrial complex to develop suitable capabilities. There are two testable implications of the technological determinism hypothesis as applied to the case of China. The first is that China’s initial forces were not guided by any strategic logic. The second is that, if China’s nuclear posture was indeed primarily technology-­driven, as China’s capability base expanded, its posture should have correspondingly broadened as well. Both are false. China’s nuclear posture was a function of strategic choice and was in fact top-­down driven, not dictated bottom-­up by technological advances. On the first implication, Lewis and Xue themselves note that Mao provided ample broad guidelines for China’s strategic force development, distilling seven key parameters that shaped China’s nuclear forces. Furthermore, the development of China’s first four ballistic missiles (DF-­2 through 5) was guided by clear strategic imperatives: the ranges of these missiles were dictated by Nie Rongzhen and Mao to target specific locations that China might have to ultimately retaliate against (Japan/USSR, Philippines, Guam, and the Continental United States). Gill and colleagues further report that “central authorities . . . later change[d] the range and payload requirements for individual missiles to reflect new strategic goals. For example  .  .  . the military commission in 1970 commanded that the range of the DF-­4 be increased from 4,000-­km to 4,500-­km” to put Moscow in range of their deployed positions.80 This suggests that the technology was directed by lucid strategic goals, and not the other way around. The second testable implication of the technological determinism hypothesis is that as China’s technical capabilities expanded, its nuclear posture would also expand. But this too is demonstrably false. First, China limited the deployed capabilities of the technologies it had already mastered. That is, the technological determinism hypothesis cannot explain why China only deployed 20–­24 DF-­5s even though it certainly had the 78. Lewis and Xue, “Strategic Weapons and Chinese Power,” pp. 547–­48. 79. Lewis and Hua, “China’s Ballistic Missile Programs,” pp. 5–­7. 80. Gill, Mulvenon, and Stokes, “The Chinese Second Artillery Corps,” p. 543.

148

Chapter 5

capacity to produce more. Nor can an argument based on technological determinism provide traction on why China continues to deploy only around 200 nuclear warheads and keeps a significant amount in reserve. The choices for China’s deployed capabilities clearly were not driven by technological considerations or constraints, but by strategic logic. Second, since the mid-­1990s, as Johnston declared, “China has the technical capacity to increase the size of its nuclear forces by about two to three times and to improve its operational flexibility.”81 He goes on to write that “as one U.S. government analyst put it, economically and technologically there are no ‘insurmountable obstacles’ to the development of a limited deterrent if the political decision to do so were made.”82 But over the past decade, that political decision does not seem to have been exercised, since China has not quantitatively augmented its nuclear posture or developed a range of tactical warheads, which it certainly has the capacity to do. Rather, though it has improved its capabilities, it has done so to maintain the quality of its assured retaliation posture against particularly the United States. Even though China has had the capacity to shift nuclear postures, for clear strategic reasons—­that a small, effective, survivable force is sufficient to deter nuclear coercion and attack—­it has continued to choose to adhere to its basic enduring nuclear posture of assured retaliation. Thus, the technological determinism hypothesis does not offer a compelling explanation for the character and trajectory of China’s nuclear posture. The final alternative explanation for the choice and persistence of an assured retaliation posture in China is provided by the strategic culture account. The primary problem with such an explanation is that there is no agreement on what constitutes Mao’s, let alone China’s, strategic preferences. Such preferences are not monolithic, but instead vary across actors and time, and are at times also somewhat contradictory. One strand of the strategic culture literature focuses specifically on Mao’s beliefs about the role of nuclear weapons and force, and Chinese leaders’ continued adherence to those beliefs, to account for China’s nuclear posture. Though it is not articulated systematically as such, this explanation seems to be part of Jeffrey Lewis’s account for China’s minimalism with respect to nuclear weapons, as well as Chong-­Pin Lin’s.83 In particular, these works emphasize that Chinese leaders believed that only a few nuclear weapons were sufficient to deter opponents’ threats and that their value was primarily psychological. Mao, in particular, believed that nuclear weapons did not essentially alter the nature of conflict and thus, “[China] should not fear 81. Johnston, “Prospects for Chinese Nuclear Force Modernization,” p. 548. 82. Ibid., p. 564. 83. See Lewis, Minimum Means of Reprisal; Lin, China’s Nuclear Weapons Strategy, especially chapter 2; also see Medeiros, “Evolving Nuclear Doctrine,” pp. 47–­48.

china 149

them because they cannot fundamentally decide the outcome of a war.”84 As such, Mao did not abandon the size and readiness of the PLA for “active defense” and believed that with respect to nuclear weapons, “a few atomic bombs are enough. Six are enough” to “scare others” and that China merely had to “build a few, keep the number small, make the quality high (you yidian, shao yidian, hao yidian).”85 Others such as Nie Rongzhen and Deng Xiaoping made statements along similar lines which implied the belief that it did not take much to deter nuclear use and that Chinese leaders believed that their primary value was psychological—­the mere ability to detonate a terrifyingly large megaton-­yield nuclear weapon on an adversary’s head was sufficient to prevent an opponent from making nuclear threats and attacks. In this view, China’s minimalism and focus on defense with respect to nuclear posture might endure because of the beliefs of the role and nature of nuclear weapons and the enduring legacy of Mao. This strategic culture explanation is also one of the key explanations in the Fravel and Medeiros theory for China’s often-­vulnerable assured retaliation posture.86 According to Fravel and Medeiros, “Since the founding of the People’s Republic in 1949, China’s senior political and military leaders have emphasized that nuclear weapons are principally useful for two reasons: deterring a nuclear attack and countering nuclear coercion.”87 China’s leaders, according to primary sources accessed by Fravel and Medeiros, “agreed that nuclear weapons lacked any meaningful war-­fighting or war-­winning function.”88 Chinese leaders and Mao in particular thus believed that a conservative force structure was sufficient for China’s needs—­ explicitly arguing, “What would we do with so many? To have a few is just fine.”89 Fravel and Medeiros argue that this belief constrained the PLA’s work and thinking on nuclear weapons, generating twin pathways to China’s limited nuclear force. It is important to note that this explanation is not inconsistent with mine. Indeed, optimization theory would anticipate that Chinese leaders would have precisely these views. However, optimization theory attempts to specify why Chinese leaders might have held the ideas and beliefs that they did based on structural and domestic political conditions. That is, Mao and subsequent leaders had the luxury of holding conservative beliefs on nuclear force structure because their security environment demanded nothing more, and certainly did not demand an asymmetric escalation posture. I further argue that the constraints placed on the military on nuclear affairs were also because of the highly assertive 84. Mao quoted and translated in Lewis and Xue, China’s Strategic Seapower, p. 231. 85. Mao quoted and translated in ibid., p. 232. 86. Fravel and Medeiros, “China’s Search for Assured Retaliation,” pp. 48–­87. 87. Ibid., p. 58. 88. Ibid., p. 61. 89. Ibid., p. 63.

150

Chapter 5

civil-­military structures that Mao instituted since the nation’s founding. So my argument is not incompatible with Fravel and Medeiros’s explanation, but it further specifies the possible structural sources of Chinese leaders’ ideas with respect to nuclear weapons and the institutional constraints they placed on their operationalization. A larger problem for the strategic culture explanation is that other work on Chinese strategic culture, notably Johnston’s, disputes this minimalism and defensiveness in Mao’s strategic preferences. Indeed, Johnston characterizes Mao’s strong realpolitik tendencies which emphasized the need for superior forces and the “annihilation, as opposed to the mere routing of forces” in warfare once China is in an advantageous material position.90 He notes that, for Mao, defensive tactics were just a preliminary stage prior to a powerful strategic offensive, which required China to possess superior capabilities to annihilate the enemy in a retaliatory strike. With respect to nuclear weapons, Mao was “no strategic Luddite” and he certainly had no qualms in viewing nuclear weapons as important military tools that could be used both against and by China if necessary—­he viewed them as true “weapons of mass destruction . . . [to be handled] with a scientific attitude.”91 This reading of Mao’s strategic preferences leads to a very different prediction with respect to Chinese nuclear posture—­it implies at least an assured destruction if not a more aggressive war-­fighting, rather than an assured retaliation, posture. It certainly does not imply the degree of vulnerability in nuclear forces, both quantitatively and qualitatively, that Mao’s China and subsequent leaders were willing to accept with respect to both the Soviet Union and the United States. Even allowing for an offensive strategy conditional on material advantages, the key puzzle is why China did not build a more robust nuclear force. The fact that China did not develop nuclear forces on par with, let alone superior to, the United States and USSR, and why China never aimed to develop an assured destruction or offensive nuclear war-­fighting capability, poses problems for this strand of the strategic culture explanation. Thus, one concern with the strategic culture explanations is that there is no determinate prediction for Chinese nuclear posture. One line of thought predicts an assured retaliation posture, while another predicts either an assured destruction posture or a more offensive war-­fighting posture. Taken as a whole, one could find a strategic culture explanation based on Mao’s strategic preferences for virtually any nuclear posture China adopted. Given 90. In particular see Alastair Iain Johnston, “Cultural Realism and Strategy in Maoist China” in Peter Katzenstein, ed., The Culture of National Security: Norms and Identity in World Politics (New York: Columbia University Press, 1996), pp. 236–­37; also see Alastair Iain Johnston, Cultural Realism: Strategic Culture and Grand Strategy in Chinese History (Princeton: Princeton University Press, 1995), especially pp. 255–­58. 91. Mao quoted and translated in Lewis and Xue, China’s Strategic Seapower, p. 232.

china 151

the somewhat contradictory preferences highlighted by these two research threads, it is difficult to conclusively sustain that China’s assured retaliation posture was caused by Mao’s beliefs about minimalism and defense, since Johnston’s work suggests that such an inference would be based on a selective interpretation of Mao’s views. My theory departs from a purely strategic culture account by arguing that beliefs about China’s appropriate nuclear force structure derive from a lucid analysis of external and internal security constraints by Chinese leaders. In this view, Mao and top Chinese leaders were fully capable of carefully optimizing China’s nuclear force structure given China’s capabilities and constraints, and their beliefs about nuclear weapons were not somehow independent of China’s strategic environment.92 Given China’s conventional strength, at least in terms of manpower and increasingly in modern capabilities, China only had to deter nuclear coercion and attack. Mao and subsequent Chinese leaders keenly understood that in order to do that, China merely had to develop an assured—­even if only plausible or delayed—­nuclear retaliatory capability. Further, given the internal tumult unleashed by the 1949 Civil War and later the Cultural Revolution, Chinese leaders put overwhelming priority on centralized control of nuclear weapons. Thus, a few large nuclear weapons not only deterred China’s adversaries to the requisite extent, but were also easier to manage. As noted earlier, this optimization explanation is not inconsistent with some versions of a strategic culture explanation, but it more clearly specifies the constraints China confronted and how its leaders optimized Chinese nuclear posture in response to them. It also leads to a determinate prediction for Chinese nuclear posture, driven by a very strong emphasis on centralized and negative control of China’s nuclear assets which was maximized by an assured retaliation posture. While data on China’s nuclear posture is notoriously difficult to find and corroborate, especially when limited to English-­language sources, key indicators of Chinese posture—­in particular what we can glean about capabilities and command and control—­suggest that not only has China persisted with an assured retaliation posture, but the best explanation of that choice seems to be the rational optimization of China’s external and internal security pressures. Much like India, China does not face an existential threat from a nuclear armed conventionally superior threat; and much like India, there are serious civil-­military pathologies that privilege centralized political control, though for very different reasons in the two states. However, the nuclear posture “output” is the same: an assured retaliation posture that is sufficient to meet these states’ security needs but emphasizes strong negative controls that preclude a first use or nuclear warfighting posture. In 92. A phrase borrowed from Johnston, “China’s New ‘Old Thinking’,” p. 7.

152

Chapter 5

the case of China’s enduring choice of an assured retaliation posture, neither technology nor economics nor strategic cultural beliefs explains why China continues to adopt a largely defensive retaliatory nuclear capability. Instead, the most compelling explanation is Chinese leaders’ emphasis on highly centralized control over China’s nuclear assets. This is not an immutable variable, but since these structures are quite sticky, there would be clear indicators if China were to move to a more delegative command-­ and-­control structure. Since China will likely never face an existential conventional threat, revisions to its assertive civil-­military structures is the factor most likely to change China’s nuclear posture away from assured retaliation. Thus, in the Asian nuclear landscape, what becomes clear is that Pakistan’s nuclear requirements, strategy, and posture are driven primarily by India’s conventional developments, and changes to India’s nuclear arsenal may have only marginal effects on Pakistan’s strategic calculations. Changes in South Asia’s nuclear balance will therefore likely be driven by India’s conventional modernization program, which might push Pakistan to adopt more, and more ready, nuclear weapons for tactical use on Indian forces. India’s nuclear strategy and posture is largely decoupled from developments in Pakistan or China, once it obtains a sufficient assured retaliation capability against especially the latter. Should China develop missile defenses or counterforce capabilities that might threaten the assurance of India’s retaliatory capability, India might have to make corresponding adjustments. But otherwise, India’s nuclear requirements are not maximalist. Similarly, China’s nuclear modernization is presently largely driven by American counterforce and missile defense capabilities. However, once the assurance of its retaliatory capability is achieved, there is little reason to believe that China will embark on a fundamentally different nuclear arsenal or posture. Together, these three cases highlight the different nodes at which my theory can drive choices about nuclear posture, from external security environment in Pakistan’s case to internal considerations about civil-­military relations in the cases of India and China.

Chapter Six

France

The next three chapters describe and code the nuclear postures of the three remaining regional nuclear powers, unified by the cooperation they shared in the nuclear domain: France, Israel, and South Africa. Although there was substantial nuclear cooperation between France and Israel, they operation­ alized their nuclear forces in diametrically opposed ways, with France opt­ ing for an explicit asymmetric escalation posture against the Soviet Union during the Cold War, and then continuing with a tous azimuts asymmetric escalation strategy afterward. Israel, on the other hand, consciously selected a catalytic nuclear posture as a lever to compel the United States to assist and protect Israel in the event of an Arab assault, before later shifting to an assured retaliation posture after the first Gulf War. South Africa and Israel shared substantial nuclear material with each other, as well as a nuclear posture. South Africa, when it had nuclear weapons between 1979 and 1991, explicitly adopted a catalytic posture, in order to compel Washington to assist Pretoria in the event of a “total communist onslaught” sponsored by the Soviet Union through Southern Africa. This chapter traces the contours of France’s nuclear postures over time. During the Cold War and after, France adopted an asymmetric escalation posture—­though for very different reasons in the two periods. During the Cold War, France’s geopolitical position and its fear of American aban­ donment in the event of a conflict pushed France toward an asymmetric escalation posture for strictly security reasons: to deter a conventionally superior nuclear-­armed Soviet Union that could march all the way to the Atlantic. France had little option but to array its nuclear forces toward asymmetric nuclear escalation. After the dissolution of the Soviet Union, 153

154

Chapter 6

however, France enjoyed a more benign security environment. It was faced with a choice: it could scale back its nuclear force posture toward an as­ sured retaliation posture to deter conflict and coercion at the nuclear level, or it could continue to adopt an asymmetric escalation posture as a hedge to deter future potential conventional threats, such as a resurgent Russia. While the United Kingdom decided to remain dependent on the United States through the lease of U.S. technology and equipment, France has opted to maintain an independent asymmetric escalation posture. Opti­ mization theory accounts for both the form and the rationale for France’s asymmetric escalation posture over time.

France during the Cold War: Force de Frappe The first half of the twentieth century was catastrophic for France. Twice invaded by German forces, and having much of its territory occupied, it is no surprise that France sought the ultimate independent deterrent against the Soviet conventional threat during the Cold War: nuclear weapons. Conditioned by the perceived abandonment of its allies—­particularly the United Kingdom and United States—­prior to World War II, and by the ex­ periences of Dien Bien Phu in 1954 and the Suez Crisis in 1956, in which allied support failed to come to France’s aid, France carved out a unique position within NATO with respect to its nuclear forces. Unlike the United Kingdom, which effectively subordinated its nuclear forces to NATO—­and thus American—­command, France retained sole responsibility for control­ ling its nuclear forces and making decisions about employment, targets, and use. Against a massive Soviet conventional force, against which France could not hope to present a serious conventional deterrent, and concerned that the United States would never wage a war—­let alone a nuclear war—­on behalf of another state, France chose to adopt an asymmetric escalation nuclear posture threatening the first use of nuclear weapons against a So­ viet conventional breach into Europe to deter the USSR from attempting to control the European landmass up to the English Channel and Atlantic. Given its experience during World War II and the rapid emergence of the Soviet threat, as well as fears of a potential nuclear-­armed Germany, France’s Fourth Republic government very quickly established an atomic energy commission, known by its French acronym CEA, to develop an in­ digenous nuclear infrastructure with the clear capacity for military appli­ cations. Three key events led France to the conclusion that it required an independent nuclear force to deter potential Soviet conventional aggression in Europe. First, U.S. reluctance to actively assist France’s atomic energy program, according to Avery Goldstein, “could not help but contribute to French

France 155

doubts about the dependability of security based on a deterrent to be pro­ vided by this same ally.”1 In particular, though the U.S. Congress in 1954 and 1958 agreed to some nuclear cooperation with its allies, it required that the country have “made substantial progress in the development of atomic weapons” itself which de facto meant that U.S. cooperation was restricted to only the United Kingdom.2 The “freezing out of France” gen­ erated a two-­tier relationship within NATO’s key power centers in which America and Britain enjoyed tight technological cooperation (indeed the UK’s strategic delivery systems were all American) but France was left on her own, at least through the Nixon administration. The second event that caused France to doubt the reliability of the United States and to seek an independent nuclear force was its 1954 expe­ rience at Dien Bien Phu in Indochina, one of France’s colonies over which it was rapidly losing control. With French forces besieged at Dien Bien Phu, Paris appealed to President Eisenhower to intervene with American airpower. Eisenhower ultimately declined, since there was little American interest in preventing the Vietnamese from expelling the French from a co­ lonial periphery.3 Even though France perceived a critical national security interest in the preservation of its position in Indochina, Avery Goldstein notes that its “two most important allies had demonstrated that there were limits to the circumstances in which they would provide support.”4 France was reminded that its view of what constituted a vital national security interest diverged from what its allies considered a vital threat to French national security interests. As such, following its defeat at Dien Bien Phu, Prime Minister Pierre Mendès-­France’s government approved increasing research on the military application of France’s nuclear program. Though Mendès-­France’s motivations may have been myriad, and likely included the risks of German rearmament and the Soviet threat, France was increas­ ingly dubious of relying on allies for its security.5 The third and most critical trigger for France’s pursuit of an indepen­ dent nuclear force was the 1956 Suez Crisis in which, again, the failure of the United States to support perceived French security interests led Prime Minister Guy Mollet, and subsequently Charles de Gaulle, to seek a nuclear 1. Avery Goldstein, Deterrence and Security in the 21st Century, pp. 192–­93. France received little positive assistance from the United States in the 1950s for its atomic energy program; in the 1970s, once it already had nuclear weapons, the United States provided negative guidance in the form of “20 Questions.” 2. Quoted in Richard H. Ullman, “The Covert French Connection,” Foreign Policy, no. 75 (Summer 1989), p. 5. 3. See Goldstein, Deterrence and Security in the 21st Century, pp. 187–­88. Also see Hy­ mans, Psychology of Nuclear Proliferation, pp. 93–­95. 4. Goldstein, Deterrence and Security in the 21st Century, p. 189. 5. Hymans, Psychology of Nuclear Proliferation pp. 97–­109.

156

Chapter 6

force over which France would possess full operational and decision con­ trol. President Eisenhower’s coercion against British and French military action in Suez to secure what they believed was a critical economic and national security objective was the most crucial event that shaped France’s decision to develop a nuclear deterrent force independent of NATO and the United States.6 Even though Washington clearly viewed joint British / French control of the Suez as less important an objective than London and Paris, the worry in France was that the United States (and UK) might view a war on Continental Europe as a less severe threat than France would. However, while the United Kingdom was also troubled by the American reaction to the Suez operation, London—­already dependent on the United States—­ decided that their answer would not be further distance from Washington, but tighter cooperation through the policy of “interdependence.”7 France lacked two things that pushed it toward an opposite policy of indepen­ dence and self-­reliance against the Soviet threat: an historical special rela­ tionship with Washington, and the English Channel. French fears of abandonment were further amplified by the fact that, unlike Nazi Germany, the USSR not only possessed superior conventional forces but also a full array of tactical and strategic nuclear weapons that could target the U.S. homeland as a deterrent to American intervention.8 Charles de Gaulle explicitly laid out the rationale for an independent French nuclear force, saying, “American nuclear power does not necessarily and immediately meet all the eventualities concerning France and Europe” and thus France must “equip herself with an atomic force of her own” to meet its own security needs, particularly against the USSR.9 President de Gaulle told Kennedy that no country will wage nuclear war for another country and that “the United States would use nuclear weapons only if its own territory was directly threatened.”10 Indeed, de Gaulle went so far as to reject any American assistance with nuclear designs or technology that car­ ried with it any conditions, writing to President Kennedy directly that: “I cannot see how France could receive in this field the assistance of another state without conditions which would limit the right to the use of these

6. Ibid., pp. 163–­64; David S. Yost, France’s Deterrent Posture and Security in Europe, Part I: Capabilities and Doctrine, Adelphi Papers, no. 194, Winter 1984/1985, p. 4. 7. W. Taylor Fain, “John F. Kennedy and Harold Macmillan: Managing the ‘Special Re­ lationship’ in the Persian Gulf Region, 1961–­1963,” Middle Eastern Studies 38, no. 4 (Oc­ tober 2002), pp. 101–­2. 8. See Mendl, Deterrence and Persuasion (New York: Praeger, 1970), p. 103; also see Kohl, French Nuclear Diplomacy, p. 36. 9. Quoted in Kohl, French Nuclear Diplomacy, p. 234. 10. De Gaulle paraphrased in General Pierre Gallois, “French Defense Planning—­The Future in the Past,” International Security, vol. 1, no. 2 (Autumn 1976), p. 17.

France 157

weapons. This France would regard as incompatible with its sovereignty.”11 George Ball concludes that de Gaulle’s position was “crystal clear. It is that he would regard even the most flexible and tenuous conditions as unacceptable if they operated to limit in any way France’s right to com­ mand, use and dispose of its nuclear forces, at will and purely for national purposes.”12 With an independent nuclear force, in the event that France decided its interests were sufficiently threatened, but the United States did not, the French government could unilaterally decide whether and how to use its nuclear weapons to achieve its security objectives. France’s fear of abandonment was not without justification. Recent Cold War histories have shown that even American officials were tepid about maintaining a permanent military presence in the defense of Western Eu­ rope.13 John Foster Dulles wrote that the United States should not promote “the illusion that we can be relied upon indefinitely to rescue [Europeans] from their own errors.”14 Historian Francis Gavin writes that President Eisenhower even as late as around 1960 “strongly believed that America’s conventional force commitment was a temporary remedy to Western Eu­ rope’s weakened state after the Second World War” and constantly looked for ways to “ ‘redeploy,’ or withdraw, American troops stationed in Western Europe.”15 Thus, not only was France able to point to specific episodes of perceived American abandonment, but the most senior U.S. officials sent clear messages that America and its force commitment to NATO was not a reliable patron for the defense of France. French leaders therefore decided to develop an independent nuclear deterrent. Thus, at the end of 1956, Prime Minister Guy Mollet directed the CEA and the armed forces to begin preparing for the militarization of France’s nuclear capability. In April 1958, Prime Minister Felix Gaillard decided to order a nuclear test for 1960.16 On February 13, 1960, in the Algerian 11. De Gaulle letter to President Johnson, 5 August 1963 quoted in “Memorandum from Under Secretary of State George W. Ball to President Kennedy, ‘A Further Nuclear Offer to General De Gaulle’,” National Archives, Record Group 59, Records of Undersecretary of State George Ball, box 21, France, August 8, 1963, pp. 3–­4. 12. Ibid., p. 4. 13. See Sebastian Rosato, “Europe’s Troubles: Power Politics and the State of the Euro­ pean Project,” International Security, vol. 35, no. 4 (Spring 2011), pp. 45–­86; Mark Sheetz, “Exit Strategies: American Grand Designs for Postwar European Security,” Security Studies, vol. 8, no. 4 (Summer 1999), pp. 1–­43; Francis J. Gavin, Gold, Dollars, and Power: The Politics of International Monetary Relations 1958–­1971 (Chapel Hill: University of North Carolina Press, 2007); Marc Trachtenberg, A Constructed Peace: The Making of the European Settlement 1945–­1963 (Princeton: Princeton University Press, 1999). 14. John Foster Dulles in Sheetz, “Exit Strategies,” p. 3. 15. Gavin, Gold, Dollars, and Power, p. 40. Emphasis in original. 16. Yost, France’s Deterrent Posture and Security in Europe, Part I, p. 4. Also see Norris, Burrows, and Fieldhouse, Nuclear Weapons Databook, , chapter 4.

158

Chapter 6

desert, France successfully tested a 60–­70kT plutonium fission device, the gravity bomb AN11. France’s strategic nuclear forces, like America’s, re­ lied on a triad of delivery capabilities: aircraft/bomber, land-­based ballistic missiles, and eventually a sea leg deployed on ballistic missile submarines. The size of the French arsenal at the height of the Cold War was judged to be between 350–­450 total strategic and tactical warheads. The first gen­ eration of French nuclear weapons, the AN11 and AN22 gravity bombs, were pure plutonium fission devices deliverable by Mirage IV aircraft. On August 24, 1968, France tested a 2.6MT thermonuclear device, giving it the capability to develop city-­destroying strategic nuclear warheads. These thermonuclear weapons were deployed atop the intermediate-­range SSBS S2 solid fuel silo-­based missiles in the custody of the Forces Aériennes Stra­ tégiques (FAS), and based in the Albion plateau; these eighteen ballistic missiles could be launched within 1 minute if France was on a crisis foot­ ing, and in roughly 5 minutes during peacetime.17 In 1980, this leg of the triad was replaced on a one-­for-­one basis with the upgraded S3 missile. The virtual invulnerability of France’s strategic forces was ensured with the addition of a sea leg to its nuclear triad—­operationalized during the Cold War with the Redoubtable-­class SSBN first launched in 1967, of which five were ultimately built. They were initially armed with the MSBS M1 submarine-­launched ballistic missile (SLBM) with a 500kT boosted fission warhead, and later with upgraded SLBMs (MIRVed SLBMs were deployed in the late 1980s). In addition to its strategic nuclear forces, France developed aircraft and missile-­deliverable tactical nuclear forces. The AN52s in the mid-­1960s were tactical devices deliverable by a variety of aircraft, including Mirage III’s. From 1961 through 1966, the French Army employed American-­made 25km Honest John missiles which were suitable for tactical delivery of a nuclear weapon. After retiring the Honest Johns, the French Army relied on the 120km Pluton missile, a mobile short-­range surface-­to-­surface tactical ballistic missile, for its tactical nuclear arm.18 The Plutons were to be mated with the AN51 tactical nuclear weapon, which entered service in 1971. Thirty-­five launchers organized into five Pluton regiments all stationed on French territory, as well as Mirage III-­deliverable gravity bombs, consti­ tuted the bulk of France’s tactical nuclear weapons capability in the early phases of its nuclear program. Toward the end of the Cold War, France de­ veloped the Air-­Sol Moyenne Portée (ASMP) missile for the Mirage 2000N which is, according to the U.S. Defense Intelligence Agency, a “tactical nuclear weapon for use against hardened targets, large naval targets, and

17. Norris, Burrows, and Fieldhouse, Nuclear Weapons Databook, p. 236. 18. See ibid., pp. 266–­68.

France 159

targets vulnerable to nuclear strike,” imparting France a “significant capa­ bility for deep theater and possibly Soviet interdiction strikes.”19 In terms of command and control, all of these forces were directly in­ tegrated into France’s armed forces, and the military end users had full operational custody of nuclear assets. These forces have always been under strictly sovereign French control, formally institutionalized by de Gaulle’s withdrawal from NATO’s defense command structure in 1966. France’s nuclear command-­and-­control architecture is more similar to the United States than to, say, China—­with a variety of positive and negative controls that balance the risk of unauthorized use against the need for a potentially immediately releasable nuclear weapon against Soviet forces and targets.20 The French president has ultimate authority in nuclear employment de­ cisions and is accompanied at all times by an aide carrying the arming codes for France’s strategic nuclear weapons. France’s strategic forces are equipped with PAL-­equivalents that require the insertion of a code to arm the weapon. A hardened command-­and-­control architecture linking the Élysée Palace, the National Military Command Center at Taverny, and stra­ tegic nuclear operators has been erected to ensure rapid devolution of the arming codes when required, even during chaotic and nuclearized environ­ ments.21 Taverny is the nerve center for strategic nuclear operations and, in a crisis, would direct nuclear employment options and ultimately the release of France’s strategic assets. According to Shaun Gregory, to hedge against the possibility of a decapitation strike, in a crisis, it appears that procedures have been instituted to “distribute the critical Presidential code to a number of senior nuclear commanders,” or as far down the chain of command as the president sees fit.22 The peacetime status of France’s nu­ clear forces is a fully mated, fully operational, though not necessarily alert, force. But strategic forces, as noted earlier, can be released within minutes from this state of readiness and less than a minute in the event that the alert status is heightened. France’s tactical nuclear forces, which are stewarded directly by the ser­ vices, have a more delegative authority structure even though presidential authority is nominally required for their use. In the case of tactical forces, the commanders of the three armed services have operational control and seem to be able to authorize use.23 Gregory notes that the commander of 19. Defense Intelligence Agency analysis quoted in ibid., p. 230. 20. See Shaun R. Gregory, Nuclear Command and Control in NATO: Nuclear Weapons Operations and the Strategy of Flexible Response (New York: St. Martin’s Press, 1996), chapter 5. Also see Yost, France’s Deterrent Posture and Security in Europe, Part I, pp. 26–­28. 21. Gregory, Nuclear Command and Control in NATO, pp. 132–­33. 22. Ibid., p. 132. 23. Ibid., p. 137.

160

Chapter 6

the First Army controlled France’s Pluton forces while the Force Aérienne Tactique controlled the air-­based tactical nuclear forces; the naval leg of France’s tactical nuclear weapons, aboard the aircraft carriers Clemenceau and Foch, were under the command of the Naval Mediterranean Com­ mand based in Toulon.24 It is not clear whether France’s tactical forces re­ quire arming codes or whether they are equipped with PALs. Thus these forces are not only physically integrated into the French armed forces, but authority for their release is also vested with the military for theater use. During the Cold War, France’s nuclear capabilities were therefore suited, and indeed arrayed, for an asymmetric escalation posture. What was the envisioned employment doctrine for these forces? Since de Gaulle, France has envisioned the first use of both tactical and strategic nuclear forces against even a conventional threat to inflict damage on the adversary that was proportional to the value of France itself. In the late 1970s, President Giscard d’Estaing clearly and publicly articulated the doctrine of propor­ tional deterrence that France had followed since de Gaulle, stating: “French nuclear forces have been calculated to permit reaching a population of the adversary of the same order as that of our own country. If France were destroyed, our adversary would lose the equivalent of France.”25 The ad­ versary remained unnamed, but France’s forces were clearly primarily de­ veloped to deter the Soviet Union during the Cold War. France, lacking the nuclear force structure to inflict assured destruction against the Soviet Union, could not threaten total destruction as the United States could. It could, however, inflict a level of damage so punishing that it would be, ac­ cording to former Chief of Staff of the Armed Forces Gen. Lacaze, “judged superior to the demographic and economic” gains that could be achieved by conquering France.26 This was the whole concept behind the force de frappe which aimed at allowing a weaker state like France to deter a much stronger—­both conventionally and nuclear—­state like the USSR. France operationalizes this proportional deterrence in a specific way. France relies on its tactical nuclear forces to achieve deterrence by de­ nial—­a so-­called pre-­strategic strike—­followed by an all-­out strategic nu­ clear strike to achieve deterrence by punishment should the adversary not cease its advances. This concept was first articulated by Chief of Staff Gen.

24. Ibid., pp. 137–­38. 25. Quoted in Yost, France’s Deterrent Posture and Security in Europe, Part I, p. 15. 26. Quoted in ibid.; also see Bruno Tertrais, “ ‘Destruction Assuree’: The Origins and Development of French Nuclear Strategy 1945–­1981,” in Henry D. Sokolski, ed., Getting MAD: Nuclear Mutual Assured Destruction, Its Origins and Practice (Carlisle, PA: Strate­ gic Studies Institute, November 2004), pp. 51–­122; David S. Yost, “France’s Nuclear Deter­ rence Strategy: Concepts and Operational Implementation,” in Sokolski, Getting MAD, pp. 197–­237.

France 161

Fourquet in 1969.27 France aims to effect war-­termination through the first use of nuclear weapons if necessary, and to manage conflict (de)es­calation through the threat of a massive follow-­on strategic strike. Although there was some debate about whether France’s “warning shot” before resorting to a strategic strike would be aimed at military formations or a soft coun­ terforce target—­as opposed to a demonstrative airburst that did not target a specific asset—­Charles Hernu, Defense Minister from 1981 to 1985, has stated that France’s tactical forces would be aimed at a military target, “dealing a militarily significant blow to halt an aggressor,” or deterrence by denial.28 That is, the concept of a “shot across the bow” before a massive strategic strike, according to France’s defense white papers, “presupposes a certain military effectiveness; our pre-­strategic capacity must be suffi­ cient to deliver a blow which stops the enemy’s progression.”29 When he was prime minister in 1975, Chirac declared that tactical nuclear weapons “are at the same time weapons of deterrence and weapons on the battle field.”30 Norris, Burrows, and Fieldhouse write that “the single and mas­ sive ‘final warning’ was to be launched against military targets, not urban concentrations.”31 Gen. Lacaze describes the envisioned employment of tactical nuclear weapons “before the use of strategic weapons, in order to lead [the enemy] to renouncing his enterprise . . . [they] must have a military effect, which is to say that it must be effective and brutal, which means a relatively mas­ sive employment and therefore limited in time and space. But above all this warning must be well integrated in the general deterrent maneuver.”32 The operational role assigned to French tactical forces was primarily theater-­ oriented, with former Chief of Staff of the Armed Forces, Gen. Francois Maurin, in 1974 identifying “valid targets . . . of two types . . . on the one hand, so called targets of opportunity that concentrations of enemy units would form . . . and, on the other hand, interdiction targets which are fixed infrastructure targets: crossing points, airfields, radars, etc. This second cat­ egory of targets would be aimed at by our air-­delivered tactical nuclear weapons, whose range clearly surpasses the territory of the Federal Repub­ lic [of Germany].”33 Although forces were arrayed in French territory, they 27. Sten Rynning, Changing Military Doctrine: Presidents and Military Power in Fifth Republic France (Westport Praeger, 2002), pp. 26–­27. 28. Quoted in Norris, Burrows, and Fieldhouse, Nuclear Weapons Databook, pp. 266–­67. 29. SIRPA, “The French Military Five Year Plan, 1987–­1991,” quoted in ibid., p. 261. 30. Quoted in Rynning, Changing Military Doctrine, p. 80. 31. Norris, Burrows, and Fieldhouse, Nuclear Weapons Databook, p. 261. 32. Quoted in Yost, France’s Deterrent Posture and Security in Europe, Part I, p. 52. Emphasis added. 33. Quoted in ibid., p. 53.

162

Chapter 6

could be employed against Soviet forces that threatened France directly—­ enhancing the credibility of the threat—­as well as against rearward targets. This also meant that the precise red-­line that would trigger France’s resort to nuclear weapons was intentionally ambiguous: France had the capabil­ ity to use the Plutons and air-­based tactical nuclear weapons on the front­ lines in Germany if the United States failed to commit to stopping Soviet forces (either conventionally or with nuclear weapons), but the red-­line could also have been as late as a Soviet approach toward French territory, which was more credible. The use of tactical nuclear weapons thus had both a military and a political logic. The military logic was to deny the Soviets a military aim by blunting Soviet conventional advances through the use of nuclear weapons, since French conventional forces would be insufficiently capable of meet­ ing the Soviet assault. The political logic was to signal to the adversary that France was prepared to massively escalate the conflict to the strategic nuclear level. Because the Soviets understood that France could not survive a Soviet nuclear strike, France had to release its strategic nuclear forces first and massively. The tactical use of nuclear weapons on military targets would provide a costly signal that French interests were sufficiently threat­ ened that France would contemplate escalation to the strategic nuclear level, giving the adversary time to pause to determine whether the expected benefits of continuing the attack were worth the strategic nuclear strike that was about to be launched.34 The overwhelming logic of the politi­ cal rationale led Defense Minister Hernu in 1982 to term France’s tactical nuclear weapons “pre-­strategic nuclear weapons” to explicitly link the use of tactical nuclear weapons to the threat of a follow-­on massive strategic strike. This asymmetric escalation posture was primarily aimed directly at the adversary. Though France’s nuclear forces and posture was complicated by the fact that it was parallel to, but independent of, NATO’s, there is no evidence that it was designed to have a catalytic or trigger effect to ensure that the United States used its nuclear or conventional forces first, to avoid ceding conflict-­escalation control to France. For the same reason that France desired an independent nuclear force in the first place—­the perceived unreliability of the United States and United Kingdom to wage war, let alone nuclear war, on behalf of French interests—France’s nuclear posture needed to be credible. France therefore had to take steps to make the first use of tactical and strategic weapons against the USSR fully op­ erational. David Yost notes that while the “trigger” or catalytic posture was discussed by theorists such as Gen. Andre Beaufre, it is largely a myth

34. See ibid., pp. 54–­55.

France 163

since “there is no evidence of its ever having been adopted as operational doctrine.”35 Bruno Tertrais concurs, arguing that even French strategic theo­ rists like Raymond Aron, Beaufre, and Gallois “did not see the employment of French nuclear weapons as a means to force the United States into using nuclear weapons for the defense of Europe,” but rather their “existence . . . as a contribution to overall Western deterrence.”36 In fact, in 1972 Defense Minister Michel Debré went out of his way to deny that any such concept was part of France’s planning, or could be credible.37 Indeed, the 1972 De­ fense White Paper explicitly stressed that national nuclear forces defended national “sanctuaries” alone, rendering any U.S. guarantees as incredible—­ encapsulating the French belief that no nation will wage a nuclear war for another nation—­and requiring an independent and credible French nuclear deterrent, operationalized as an asymmetric escalation posture. Thus, during the Cold War, France’s nuclear posture was easily iden­ tifiable as an asymmetric escalation posture. Although there might have been an ancillary effect of forcing the United States to heavily weigh its commitment to the nuclear defense of Europe, all the evidence suggests that France’s independent nuclear forces were developed with the aim of directly deterring the USSR should America abdicate that defense—­and there were very real concerns that Washington might, particularly since the USSR could hold American strategic targets hostage as a counter-­deterrent. Faced with the gripping fear of being left alone against the conventionally far superior Red Army, France adopted an asymmetric escalation nuclear posture to deter the Red Army, first by denial, and then with the threat of a punishing strategic strike that would outweigh Soviet gains.

Explaining France during the Cold War: Asymmetric Escalation France’s selection of an asymmetric escalation nuclear posture during the Cold War is relatively straightforward to explain. After 1960, when the fear of a German revival dissipated, France faced the Soviet threat, which outnumbered combined NATO forces in manpower, armored capabilities, and artillery at the German frontier. If France were left to face the So­ viet threat alone with the abdication of the American commitment—­a fear amplified by the fact that Soviet strategic capabilities might dissuade U.S. intervention—­its survival itself would once again be imperiled. Thus, doubting the reliability of American protection and facing a conventionally superior threat that also possessed nuclear superiority, France had little 35. Ibid., p. 17. 36. Tertrais, “Destruction Assuree,” p. 73. Emphasis in original. 37. Yost, France’s Deterrent Posture and Security in Europe, Part I, p. 17.

164

Chapter 6

option but to threaten the first use of nuclear weapons in the event of a conventional attack from the Warsaw Pact to deter its outbreak. Even given potentially problematic civil-­military relations after the de Gaulle ascendancy at the height of the Algerian war, France quickly oriented its military structures and organs to meet the Soviet threat, fully integrating nuclear weapons into its force structure and doctrine to establish a credible first-­use threat. The three key events noted earlier led France to the conclusion that it could not rely on the United States to be a third-­party patron to protect French national interests. The combination of French exclusion from U.S. nuclear cooperation, which Washington granted to London but not Paris, the perceived abandonment at Dien Bien Phu, and the perceived American coercion against France in the Suez Crisis to halt military operations in de­ fense of its perceived national interests convinced leaders in Paris that if the United States could abandon it under these conditions, France could not rely on the United States in the event of Soviet aggression. These episodes were bolstered by statements even at the U.S. presidential level that Amer­ ica might not permanently defend Western Europe. In short, France did not believe it had a reliable third-­party patron that could be compelled to in­ tervene on its behalf should the Red Army assault Western Europe. Charles de Gaulle explicitly spelled out this logic when he stated that France must be self-­reliant because an American extended deterrent under American control would never cover all of France’s perceived vital interests.38 In par­ ticular, with the U.S. homeland increasingly within reach of Soviet strategic capabilities, de Gaulle in particular did not believe that the United States would risk Washington to save Paris. A catalytic posture would require the United States to risk direct nuclear attack from the Soviet Union in order to secure French survival, and was therefore simply out of the question. Thus, without a perceived third-­party patron but facing the proximate and conventionally and nuclear superior threat posed by the Soviet Union, my theory predicts that France would adopt an asymmetric escalation pos­ ture. This largely dovetails with the structural realist prediction of nuclear posture. Strategic culture explanations that focus on de Gaulle’s vision for an independent France are plausible, but they too coincide with the struc­ tural realist explanation, since de Gaulle’s emphasis on an independent nuclear force was entirely consistent with his fear of abdication by the American commitment to Europe, even if nuclear weapons had the ancil­ lary benefit of also boosting French prestige. Technological determinism does not capture France’s posture, since it was forced to adopt an asymmet­ ric escalation posture before it was fully capable of doing so, forcing France to array the capabilities it had for both tactical and strategic first use even 38. See Kohl, French Nuclear Diplomacy, p. 234.

France 165

though its technology was not as mature as it might have preferred it to be. Thus, although the French case does not allow for differentiation between the theory of posture optimization, structural realism, and some strate­ gic culture explanations since they all make the same prediction, though through slightly different pathways, the technological determinism expla­ nation once again makes an incorrect prediction. Optimization theory is consistent with the structural realist prediction with respect to Cold War France’s nuclear posture, since France was fac­ ing a conventionally superior proximate offensive existential threat in the Soviet Union (and Warsaw Pact forces). It was generally believed by NATO command that it was outgunned and outmanned on the German frontier and that, though the Warsaw Pact might not have been able to win quickly or easily, it posed a severe existential threat to Germany, and potentially France.39 NATO’s net assessment in 1984—­the peak of NATO’s relative ad­ vantage owing to Soviet force depletion and engagements in Afghanistan—­ still placed the ratio of armored division equivalents at 1.60:1, the ratio of main battle tanks at 2.06:1, and the ratio of artillery at 2.33:1, all in favor of the Warsaw Pact (though this does not include qualitative differences). Though both Barry Posen and John Mearsheimer raised doubts about the conventional wisdom that the Warsaw Pact could quickly overrun NATO forces,40 there was little doubt that a massive Soviet thrust through the North German Plain, the Gottingen Corridor, and the Fulda Gap would significantly attrite NATO forces and carried with it the risks of inadver­ tent nuclear escalation if theater nuclear forces were themselves targeted. Mearsheimer writes that “it is virtually impossible to defeat the Soviets decisively in a land war in Europe. NATO simply cannot escape a war of attrition with the Soviets, unless the Pact wins a quick and decisive vic­ tory.”41 The implication of this is that either way, NATO was going to be at a serious disadvantage—­particularly if the Soviets were able to mobilize their massive reserves in a war of attrition. This conventional superior­ ity was magnified by the Soviet nuclear arsenal, which fielded somewhere between 8,800 and 11,300 land-­based tactical nuclear weapons as well as 12,300 strategic nuclear weapons—­a substantial nuclear shield behind

39. NATO and the Warsaw Pact: Force Comparisons (Brussels: NATO Information Ser­ vice, 1984), especially pp. 8–­9 and pp. 19–­20. 40. See Barry R. Posen, “Is NATO Decisively Outnumbered?” International Security, vol. 12, no. 4 (Spring 1988), pp. 186–­202; John J. Mearsheimer, “Why the Soviets Can’t Win Quickly in Central Europe,” International Security, vol. 7, no. 1 (Summer 1982), pp. 3–­39; Mearsheimer, “Maneuver, Mobile Defense, and the NATO Central Front,” International Security, vol. 6, no. 3 (Winter 1981/1982), pp. 104–­22; and Mearsheimer, Conventional Deterrence (Ithaca: Cornell University Press, 1983). 41. Mearsheimer, “Maneuver, Mobile Defense, and the NATO Central Front,” p. 120.

166

Chapter 6

which it could achieve conventional war aims.42 The situation was still dire in the 1980s, and in the earlier periods of the Cold War, the Warsaw Pact conventional advantage was even starker. The fact that the United States moved to a policy of first use of tactical nuclear weapons early in the Cold War to deter a Soviet conventional at­ tack in Germany suggests that it severely doubted its ability to thwart War­ saw Pact offensives with solely conventional forces. The problem was even more acute for France.43 The scenario that gripped de Gaulle and subse­ quent French leaders was a Warsaw Pact attack in which the United States and United Kingdom failed to use nuclear weapons to stop the attack owing to the threat of Soviet nuclear retaliation against the U.S. homeland. Should the United States decide that the defense of Germany and France was not worth risking New York and Washington, DC, France would be left to face the Warsaw Pact assault alone—­just as it had had to face German Panzers alone in World War II as external powers, buffered by water, temporar­ ily ceded Continental Europe to Germany. And if aggregate NATO forces doubted their ability to defend against a conventional Warsaw Pact assault, French forces alone did not stand a chance. The British had the luxury of the English Channel; France had no such buffers given the passable terrain of West Germany and the Low Countries. As such, to ensure the territo­ rial integrity of France against the nuclear-­armed conventionally superior Soviet threat during the Cold War, French leaders had little option but to adopt an independent nuclear force that they could call upon as an insur­ ance policy in the event that they were left alone to face Soviet conventional forces at or approaching the French border. Given Soviet conventional and nuclear superiority against the French, estimated to be at least 10:1 quanti­ tatively, and closer to 100:1 on the nuclear balance,44 France’s only option would be to threaten the first use of nuclear weapons in an opening salvo—­ which they hoped to achieve with a pre-­strategic strike followed by the threat of a massive strategic nuclear attack—­that inflicted more damage on the Soviets than they could hope to gain by defeating France. Therefore, optimization theory’s prediction collapses to the structural realist prediction that, in such a dire security situation, France would gen­ erate whatever forces and structures necessary to adopt this asymmetric escalation nuclear posture. Even given historically rocky civil-­military rela­ tions and the circumstances of Gen. de Gaulle’s founding of the Fifth Re­ public in 1958 amidst deep political and military instability due to France’s 42. See “Nuclear Notebook: Estimated Soviet Nuclear Stockpile,” Bulletin of the Atomic Scientists, vol. 47, no. 6 (July 1991), p. 48; “Nuclear Notebook: Estimated Soviet Nuclear Stockpile,” Bulletin of the Atomic Scientists, vol. 45, no. 6 (July/August 1989), p. 56. 43. See Rosato, “Europe’s Troubles,” p. 55. 44. See Posen, “Is NATO Decisively Outnumbered?” p. 192.

France 167

deteriorated position in both Vietnam and Algeria, de Gaulle’s military background led to little hesitation in fully integrating nuclear weapons into the armed forces. Nuclear weapons, in particular, were delegated to all three branches of the armed services, and authority for the release of tacti­ cal nuclear weapons rested with the commanders of the respective services that maintained custody over them. These procedures and capabilities gen­ erated a credible threat of first use should France’s interests be perceived to be severely threatened and NATO forces abdicated their defense. Optimization theory and structural realism thus do both accurately capture the selection of the asymmetric escalation posture and its particu­ lar character in France during the Cold War. However, the evidence that French leaders even considered the role of the United States as a potential defender—­something to which strict structural realism should be aller­ gic—­is suggestive that the optimization theory is broadly more powerful even in this case. Only once de Gaulle concluded that the United States could not be relied upon as a third-­party patron did he direct France to­ ward an asymmetric escalation posture. Thus, although both optimization theory and structural realism predict the final outcome, an asymmetric es­ calation posture for Cold War France, optimization theory more accurately captures how French leaders arrived at that choice. By contrast, the technological determinism explanation does not fully capture France’s choices. Most notably, de Gaulle’s France adopted an asymmetric escalation posture before it had fully developed the technolo­ gies to do so—­the intense security environment swamped France’s technical means and forced it to array whatever forces it had available in an asym­ metric escalation posture. France quickly developed a tactical nuclear capa­ bility, relying on U.S.-­made Honest Johns while the Plutons and air-­based tactical missile were being developed. Technological determinism would have predicted that, as France’s delivery capabilities were immature and its nuclear warhead numbers were relatively low, it would have attempted to either employ a catalytic posture or take steps to ensure survivability of its limited forces and aim for an assured retaliation posture. It did neither. Instead, France immediately selected an asymmetric escalation posture in accordance with its severe security environment and enhanced its credibil­ ity as its technological developments allowed. At considerable costs, France made investments in order to develop the technology that would permit a credible asymmetric escalation posture. This political decision drove tech­ nological developments, rather than the reverse. A compelling strategic culture explanation for France’s Cold War nu­ clear posture might focus on de Gaulle’s belief in an independent French foreign and security policy, particularly for the mantle of leadership in Eu­ rope. Wolf Mendl’s classic work on France’s nuclear program notes that “France’s search for independence  .  .  . was principally inspired by the

168

Chapter 6

search for leadership in Europe—­at least the Western half of it.”45 De Gaulle in particular, and Gaullism subsequently, emphasized French national independence—­perhaps partly as a function of its declining relative power, but more acutely because of the memories of World War II, Dien Bien Phu, and Suez—­and that France should be self-­reliant in terms of its security. In that sense it dovetails with a core structural realist tenet that a state in the international system must be self-­reliant lest it risk its survival on the hope that allies may come to its assistance when necessary. Another thread within de Gaulle’s approach to world affairs was the prestige imparted by an independent nuclear capability. De Gaulle himself stated to John Foster Dulles in 1958: “A France without world responsibility would be unworthy of herself, especially in the eyes of Frenchmen . . . It is for this reason too that she intends to provide herself with an atomic armament. Only in this way can our defence and foreign policy be independent, which is something we prize above everything else.”46 It is this thread that has led scholars such as Scott Sagan to argue that de Gaulle and France pursued nuclear weap­ ons as a result of a strong preference for grandeur, or prestige.47 What would this strategic culture explanation, emphasizing French pres­ tige and independence, predict? It would primarily predict the fact of the French nuclear force and its retention under sovereign control. However, it neither offers a clear prediction of the form of France’s nuclear posture—­ why France chose asymmetric escalation and not assured retaliation—­nor does it deny the potential source for France’s emphasis on independence: security concerns. Indeed, although de Gaulle did reference the utility of France’s nuclear capabilities in enhancing its prestige, he also subsequently gave a core realist rationale for an independent French nuclear force to President Eisenhower a year later: “Will [you] take the risk of devastating American cities so that Berlin, Brussels, and Paris might remain free?”48 Clearly de Gaulle did not believe that the United States would, necessitat­ ing an independent nuclear force under sovereign French control to en­ sure France’s survival. The optimization and structural realist explanations also capture the form of France’s nuclear posture better than this strategic culture explanation. If Gaullism required nuclear forces as a symbol of France’s prestige and independence, an assured retaliation posture would achieve the same end—­and be less costly and risky—­and this explanation is therefore indeterminate. However, only the optimization and structural realist accounts offer an explanation for why an assured retaliation posture was out of the question—­it would not only fail to deter Soviet conventional 45. Mendl, Deterrence and Persuasion, p. 47. 46. Quoted in Yost, “France’s Deterrent Posture and Security in Europe,” pp. 13–­14. 47. Sagan, “Why Do States Build Nuclear Weapons?” pp. 76–­80. 48. Quoted in ibid., p. 77.

France 169

advances, but France would probably not survive a Soviet nuclear strike. Given its security situation, only an asymmetric escalation posture made sense.

France after the Cold War: Asymmetric Escalation Tous Azimuts After the dissolution of the Soviet Union, France was faced with a choice of how to operationalize its nuclear forces. The United Kingdom—­save the tactical forces under NATO control—­decided to shift to a sea-­based nuclear deterrent as a cost-­saving measure, even though all forces were effectively leased from the United States and were still coordinated with NATO forces and American deterrent patrols. On the one hand, the primary threat that pushed France toward an asymmetric escalation posture—­the convention­ ally superior, nuclear-­armed Soviet Union—­largely disappeared. This gave France the latitude to shift to an assured retaliation posture, since it no longer faced an acute conventional existential threat. On the other hand, given that France possessed the technology, structures, organizational ca­ pacity, and finances to maintain an asymmetric escalation posture, it could also choose to hedge against the emergence of future threats by retaining its asymmetric escalation posture. It chose the latter, retiring some of the Soviet-­specific tactical nuclear delivery vehicles but maintaining a robust air-­deliverable tactical capability in addition to a robust sea-­based strategic deterrent to deter future regional and major power threats, from wherever they might emerge (tous azimuts).49 France’s post–­Cold War capability shift included several features, de­ signed to consolidate France’s strategic forces at sea to make them virtu­ ally invulnerable to a potential first strike and to retire land-­based tactical forces and rely entirely on an air-­based tactical nuclear capability with the Mirage and Rafale deliverable Air-­Sol Moyenne Portee (ASMP). That is, the shift recognized that the Soviet Union was no longer an acute threat, and so systems that were largely Soviet-­specific (the Plutons, S2/S3s, etc., whose ranges were suitable largely for the USSR) were retired. The SLBM capability was enhanced to provide strategic reach against emerging poten­ tial adversaries, and the aircraft-­deliverable ASMP was the primary tacti­ cal nuclear option that could also be used against a variety of potential threats.50 In 1996, following much debate about whether to replace the 49. Tous azimuts is a phrase from the de Gaulle era meaning “omni-­directional” which has been employed to describe French nuclear posture at various points throughout its nuclear history. 50. See Robert S. Norris and Hans M. Kirstensen, “Nuclear Notebook: French Nuclear Forces, 2008,” Bulletin of the Atomic Scientists, vol. 64, no. 4 (September/October 2008), pp. 52–­54.

170

Chapter 6

S3D silo-­based ballistic missiles deployed at the Plateau d’Albion, President Chirac’s government decided to retire the land-­based leg of France’s triad and not to replace them with the S45. Instead, France would rely on its new class of SSBNs for strategic deterrence, the Triomphant class ballistic mis­ sile submarines, which began replacing the Redoubtable class submarines in the mid-­1990s. France will ultimately operate four Triomphant class submarines, allowing it to maintain a submarine continuously on deterrent patrol, each equipped with 16 M45 SLBMs with a range of 4,000km, each with up to 6 MIRVed warheads. In the near future, France will replace the M45 with the upgraded M51 SLBM which has a range of 6,000km with a full payload of 6 MIRVed warheads and accompanying penetration aids to defeat missile defenses.51 Thus, France’s strategic forces have now been consolidated into a single sea leg which, since deterrent patrol routes can shift according to the threat environment, is expected to provide sufficient strategic capability against future potential threats globally. However, France made the critical decision to not eliminate its tacti­ cal nuclear forces after the Cold War. It did eliminate the Soviet-­specific land-­based mobile tactical nuclear missiles—­the Plutons, and the Hadès missiles that were to replace the Plutons in 1992, were placed in storage. In 1992, President Mitterand announced the de-­alerting of the Pluton force and began the gradual retirement of the Pluton regiments over 1992 and 1993.52 The mobile solid-­fuel tactical nuclear Hadès missiles, with a 500km range, which had been designed to replace the Plutons as France’s weapon “of final warning,” according to Mitterand, were developed in the early 1990s but ultimately were placed in storage for reserve and upload capa­ bility rather than being operationally deployed.53 With the retiring of the Plutons, and the Hadès placed in reserve, France consolidated its tactical nuclear forces onto aircraft-­based platforms—­deployed on land as well as on aircraft carriers—­armed with the ASMP missile, which France now re­ lies on for its pre-­strategic nuclear strikes.54 Three land-­based squadrons of Mirage 2000N (to be replaced by the Rafale) and one sea-­based comple­ ment of Super Ètendards (also to be replaced by the Rafale) carry a total of 50 ASMP missiles which are mated with variable yield. The ASMP is being upgraded to the ASMP-­A which will be deployed on the Rafales when they become operational; the ASMP-­A will be mated with a modernized war­ head, the TN81, which has dial-­variable yields, giving the ASMP-­A both a tactical as well as possibly a strategic role. 51. Ibid. 52. Norris, Burrows, and Fieldhouse, Nuclear Weapons Databook, p. 270. 53. Quoted in ibid., p. 271. 54. See David S. Yost, “Nuclear Weapons Issues in France,” in Hopkins and Weixing Hu, Strategic Views from the Second Tier, pp. 19–­104, especially p. 23.

France 171

France’s command-­and-­control architecture has also remained relatively consistent (although the SSBN command center has shifted bases) and is still independent of NATO and American control.55 Even though President Sarkozy announced the reintegration of France into NATO’s defense com­ mand, France’s nuclear weapons will remain independent of NATO nu­ clear forces and strictly under sovereign French control. The president still retains ultimate authority to employ France’s strategic nuclear weapons, and the French armed services still retain operational custody over nuclear assets in fully assembled and mated forms. President Sarkozy recently im­ plied that the time to release nuclear weapons remains very short, stating that France’s president “must be able to count on them at all times in order to respond to any surprise.”56 France’s nuclear weapons are believed to be controlled by modern positive and negative control technologies, ensuring that the risk of accidental and unauthorized use is minimized but that the ability to employ nuclear options when necessary is maximized. France’s nuclear posture is still presently oriented toward asymmetric escalation, reverting to a Gaullist concept of tous azimuts to deter threats from wherever they might emerge. Sarkozy noted in 2008 that the purpose of the French nuclear deterrent was to counter state aggression “wherever it may come from and whatever form it may take,” which very carefully implied that France may continue to counter conventional aggression, as well as a full range of weapons of mass destruction (i.e., nuclear, chemi­ cal, biological), with nuclear weapons.57 Sarkozy’s position articulated a remarkable post–­Cold War continuity in French nuclear posture. In 1994, the Defense White Paper explicitly noted that conventional forces were no substitute for France’s nuclear deterrent, particularly in the uncertain post–­ Cold War threat environment where new threats could emerge with little warning, and old threats could yet re-­emerge.58 In 2006, President Chirac gave a wide-­ranging and forthright speech on the rationale for France’s nuclear forces in the post–­Cold War world, making several critical state­ ments: (1) States that sponsor terrorism against France, particularly in­ cidents involving possible weapons of mass destruction use, “risk a firm and adapted response from us. And this response can be of a conventional nature. It can also be of another nature;”59 and (2) France is “not shielded 55. Rachel Williams and Richard Norton-­Taylor, “Nuclear submarines collide in Atlan­ tic,” The Guardian, February 16, 2009. 56. President Nicolas Sarkozy, “Speech: Presentation of Le Terrible in Cherbourg,” March 21, 2008. Available at http://www.carnegieendowment.org/publications/index.cfm ?fa=view&id=20001&prog=zgp&proj=znpp. Emphasis added. 57. Ibid. Emphasis added. 58. Bruno Tertrais, “The Last to Disarm? The Future of France’s Nuclear Weapons,” Nonproliferation Review, vol. 14, no. 2 (July 2007), p. 252. 59. Chirac quoted in ibid., p. 253.

172

Chapter 6

from an unforeseen reversal in the international system, nor from strategic surprise,” and must thus retain sufficient nuclear forces to meet both novel and resurgent threats.60 Thus, France’s nuclear posture is no longer nec­ essarily adversary-­specific (although earlier formulations of posture also avoided naming the Soviet Union), but is capability-­specific, aimed at coun­ tering a spectrum of potential threats from wherever they may arise—­the classic tous azimuts strategy. Sarkozy, in releasing the 2008 Defense White Paper, stated that although “the nature of the threats has changed, they are diverse and evolving. They seem distant to us, but, let us make no mistake, our country . . . could be struck tomorrow.”61 French posture still relies on the “final warning” pre-­strategic strike, though the nature of this strike, and the threat of a subsequent strategic nuclear attack, may vary and has less emphasis on battlefield utility—­ though the targets are still military—­and a greater emphasis on signaling resolve.62 Bruno Tertrais carefully describes France’s nuclear strike options in the post–­Cold War era: To deter a major power, France would rely on the threat of “unacceptable damage of any kind.” To deter a regional power, it would rely on the threat of destroying “centers of power.” An original French concept is that of “final warning” (ultime avertissement), the idea of threatening an adversary (who has misjudged French resolve or miscalculated the limits of French vital inter­ ests) with a single limited strike aimed at “restoring deterrence” . . . the final warning would be just that: a non-­repeatable strike followed by an “unaccept­ able damage” strike if the adversary persisted. French military authorities let it be known in 2006 that the final warning could take the form of a high-­altitude electromagnetic strike. Although specific weapons adaptations have not been made public, it is widely believed that France has diversified its nuclear yield options in recent years. The option of exploding only the first stage primary [for tactical use] may have been exploited, since it is known to be an easy ad­ aptation from a technical point of view.63

Thus, France’s reliance on asymmetrically escalating a conflict to the nu­ clear level persists in the post–­Cold War world. The nature of the initial tac­ tical use may vary—­instead of a direct counterforce or interdiction strike, France may use a nuclear weapon as an electromagnetic pulse to cripple 60. Ibid. 61. President Sarkozy, “Speech on Defense and National Security,” June 17, 2008. Avail­ able at http://www.ambafrance-ca.org/spip.php?article2269. 62. See The French White Paper on Defence and National Security (New York: Odile Jacob Publishers, 2008), chapter 10. 63. Tertrais, “The Last to Disarm?” pp. 255–­56.

France 173

opposing forces, which still has the effect of deterring by denial. However, the calibrated use of nuclear weapons is aimed at signaling French resolve to further escalate the conflict to the strategic nuclear level if the adversary does not desist; Mitterand and subsequent presidents have all consistently stated that “the targets of what is called the ‘final warning’ are military targets.”64 Sarkozy emphasized that French nuclear forces must provide the president with “a wide range of options to face threats” so that France’s strike options—­both conventional and nuclear—­could be both flexible and carefully calibrated.65 One lingering question is: what are France’s red-­lines in the post–­Cold War era? That is, what does France define as the threshold for nuclear es­calation? During the Cold War, it was clear that a Soviet conventional breach was the explicit trigger for French employment of nuclear forces, though whether it would be closer to the German frontier or the French bor­ der was left intentionally ambiguous. The 1994 Defense White Paper and Chirac’s 2006 speech identified several thresholds for nuclear escalation: (1) threats to the “integrity of the national territory, including the mainland as well as the overseas departments and territories,” whether conventional or non-­conventional;66 (2) “the defense of allied countries,” without the usual qualifier of European allies, leaving open a variety of other possible allies;67 and (3) the “safeguard of strategic supplies” which includes vital resources such as oil and gas, should they be strangulated by an adversary.68 To secure its perceived vital interests in the post–­Cold War world, France still relies heavily on nuclear forces, and particularly the threat of first use, as the ultimate guarantor and insurer of France’s security. Indeed, France explicitly rejects the formulation of the use of nuclear weapons as a “last resort.”69 Although some countries such as the United Kingdom moved to explicitly de-­emphasize nuclear weapons in their security strategies, the French emphasis on the importance of nuclear weapons has been consis­ tent and persistent in the post–­Cold War era in order to meet emerging, unforeseen, or resurgent traditional threats whenever and wherever they may arise. And France still orients its nuclear posture for first use in a pos­ sible deterrence by denial, as well as deterrence by punishment, mission.

64. President Mitterand quoted in Yost, “Nuclear Weapons Issues in France,” p. 27. 65. Sarkozy Le Terrible Speech. 66. 1994 Defense White Paper quoted in Tertrais, “The Last to Disarm?” p. 252. Also see David S. Yost, “France’s Nuclear Dilemmas,” Foreign Affairs, vol. 75, no.1 (January/ February 1996), pp. 114–­17. 67. Chirac quoted in Tertrais, “The Last to Disarm?” p. 252. 68. Ibid., p. 253. 69. Yost, “Nuclear Weapons Issues in France,” p. 27.

174

Chapter 6

Explaining Post–­C old War France’s Asymmetric Escalation Posture During the Cold War, France’s security position and its fear of American abdication in the event of a conflict that did not directly threaten the U.S. homeland pushed France toward an asymmetric escalation posture for strictly security reasons: to deter a conventionally superior nuclear-­armed Soviet Union that could potentially march all the way to the Atlantic, France was forced by a binding security environment to array its nuclear forces toward asymmetric nuclear escalation. After the dissolution of the Soviet Union, however, France had significantly more security room to maneuver. It was faced with a choice: it could scale back its nuclear force posture toward an assured retaliation posture to deter conflict and coercion at the nuclear level, or it could continue to adopt an asymmetric escalation pos­ ture as a hedge to deter future potential conventional threats, such as a re­ surgent Russia. With ample technological resources and a largely delegative and professional civil-­military relationship, France has opted to maintain an asymmetric escalation posture—­unlike the United Kingdom, which has effectively moved entirely to a sea-­based posture as a cost-­saving measure. Though France, since 1960, has thus adopted an asymmetric escalation posture, the motivations for doing so shifted dramatically with the exog­ enous shock of the Soviet Union’s dissolution. France adopted an asym­ metric escalation posture initially because it was severely threatened; after the USSR’s collapse, it adopted the posture because it had the luxury to do so. In one fell swoop in 1989, France’s primary existential threat—­the So­ viet Union—­disappeared. The security threat that forced France to adopt an asymmetric escalation posture no longer existed. Why then did France continue to adopt an asymmetric escalation posture in the post–­Cold War period? The simple answer is: because it could afford to do so and because the posture provided a hedge against emerging or re-­emergent threats. In optimization theory, the asymmetric escalation nuclear posture is both the curse of the severely threatened and the luxury of the rich and stable. In one massive shift in the international system in 1989, France saw both sides of that coin. After the fall of the Soviet Union, France had the option of adjusting its nuclear posture toward an assured retaliation orientation by consolidating its nuclear force onto a strategic submarine force. Instead, as a hedge against the resurgence of old threats (Russia, a nuclear Germany) and evolving protean threats (e.g., terrorism, Iran, China), France opted to maintain its asymmetric escalation posture because the marginal deterrent benefits of this posture against the possible spectrum of threats outweighed the cost of continuing the posture, particularly since France possesses “nor­ mal” delegative civil-­military structures and a large comparative techno­ logical advantage over its adversaries.

France 175

My theory of posture optimization fares best in accounting for post–­ Cold War France’s nuclear posture, generating a specific prediction for an asymmetric escalation posture as a hedge against future threats given France’s now stable and delegative civil-­military structures and its relative lack of resource constraints, with a significant comparative technological and financial advantage over its potential adversaries. At the end of the Cold War, France still did not believe it had a reliable third-­party patron given its independence from NATO and the United States, while the dis­ solution of the Soviet Union further reduced American incentives to assist Western Europe militarily. Furthermore, the data presented earlier demon­ strates that France has delegative civil-­military structures, which enables an asymmetric escalation posture if it so desires. The determinate variable in the post–­Cold War French case is whether it has resource constraints that would preclude the selection of an asymmetric escalation posture which, in France’s configuration, would be more costly to maintain than an entirely sea-­based assured retaliation posture. This is precisely the node of the deci­ sion tree at which France presently finds itself. Post–­Cold War France is the only empirical case of a regional nuclear power that goes down the entire decision pathway in the rational opti­ mization framework, facing permissive security, civil-­military, and relative resource advantages that enable an asymmetric escalation nuclear posture without generating prohibitive organizational or financial costs. Thus, post–­Cold War France is rich and secure enough such that the marginal de­ terrent benefits of the asymmetric escalation posture against present and fu­ ture threats outweighed the organizational and economic marginal costs of maintaining the posture. No longer facing a nuclear-­armed conventionally superior existential threat—­the dissipation of the Red Army and the emer­ gence of huge buffers between France/Germany and Russia dramatically reduced France’s security anxiety after the Soviet collapse—­and equipped with “normal” delegative civil-­military structures, France could choose be­ tween an assured retaliation and an asymmetric escalation nuclear posture. Endowed with a financial and technological advantage over its potential threats, France’s rational option was to continue to adopt an asymmetric escalation posture to hedge against an uncertain security landscape. Only the rational optimization theory offers a determinate prediction for post–­ Cold War France’s nuclear posture. Structural realism fails to offer a specific prediction of what post–­Cold War French nuclear posture might look like. With its primary existential threat crippled, France could have adopted several nuclear postures per­ fectly consistent with its new threat environment. In particular, either an as­ sured retaliation posture or an asymmetric escalation posture would have been consistent with France’s position in the international system. Indeed, with powerful conventional forces to meet most realistic post–­Cold War

176

Chapter 6

contingencies in its neighborhood, France could have shifted to an assured retaliation posture and retired its tactical nuclear forces or put them in reserve to hedge against future threats. This is precisely what the United Kingdom has chosen to do, moving entirely to a sea-­based Trident deter­ rent (albeit one largely leased from the United States) and abandoning its tactical nuclear weapons after the Cold War, other than those under NATO control. But France did not do this, instead choosing to field substantial numbers of air-­based tactical nuclear weapons for pre-­strategic use, in ad­ dition to retaining a robust and modernized submarine-­based strategic nuclear force. The structural realist explanation therefore cannot explain why France chose the latter posture over the former, since both are largely consistent with France’s relative power position and security environment. The technological determinism explanation generates the same predic­ tion as optimization theory for post–­Cold War French nuclear posture: asymmetric escalation. Given thirty years of experience in indigenously developing nuclear weapons and delivery vehicles, technological determin­ ism would argue, “Why stop now?” Indeed, the technological determinism explanation would predict that France should replace its silo-­based bal­ listic missiles with modern forces and that the Hadès should have replaced the Plutons as France’s operational tactical nuclear option. But instead, France chose to retire the Plutons and place the Hadès in storage, opting to scale back its tactical nuclear forces to rely entirely on the ASMP for pre-­ strategic/final warning use and upgraded SSBN and SLBM capabilities for its strategic capabilities. So, although the technological determinism expla­ nation captures the fact of the asymmetric escalation posture in this period, it would predict that France’s force posture would expand, rather than contract, as its planned capabilities became operational. Instead, France chose continuity in posture, but with fewer forces than it was capable of fielding in light of the altered security environment. Optimization theory fares slightly better than technological determinism in explaining both the fact and the form of France’s post–­Cold War nuclear posture. However, to be fair to the technological determinism theory, it does largely capture this decision to allow France’s most advanced technologies to remain deployed and operational as a hedge against future threats. The final alternative account for France’s post–­Cold War nuclear pos­ ture is the strategic culture explanation. This explanation centers on the endurance of a “Gaullist tradition” emphasizing France’s independence in foreign and security policy, and nuclear weapons’ centrality to that end. In particular, the Gaullist tradition views nuclear weapons not only as the backstop of French security, but as a frontline defense. The strategic cul­ ture explanation, however, is dealt a blow by the fact that Francois Mit­ terand, de Gaulle’s Socialist archrival, accepted de Gaulle’s formulation of French deterrence when he became president in 1981 and, critically, as he

France 177

guided France in his second term as the Soviet Union collapsed. According to Bruno Tertrais, France’s nuclear posture became an “enduring consen­ sus” with Mitterand’s continuity of de Gaulle’s formulation despite deep disagreements in other areas, and France’s nuclear posture transcended both domestic politics and competing strategic cultures within France.70 Socialist opposition to the Chirac/Sarkozy UMP governments continued to largely support France’s operationalization of nuclear weapons as a robust and independent deterrent that requires a wide range of options, includ­ ing first use if necessary.71 It could be that this consensus is itself a stra­ tegic culture, but it is not a Gaullist tradition and there are no competing cultures internally; it is thus not separable from the rational optimization prediction if one takes such a broad view of strategic culture. The narrow Gaullist strategic culture explanation for the post–­Cold War continuity of French nuclear posture, however, is impossible to sustain since it was the Socialist Mitterand who guided France’s post–­Cold War posture for the first half decade after the collapse of the USSR, with no major changes to the operational doctrine of France’s nuclear forces (he did prefer the term “final warning” to “pre-­strategic” strike since he viewed the first use of nuclear weapons as still achieving strategic ends even if it was in a tactical setting).72 As Pierre Gallois sardonically noted, “Gaullism ended with Presi­ dent de Gaulle’s political retirement in 1969, the rest is politics.”73 Thus, the alternative explanations fail to offer satisfactory predictions for post–­Cold War French nuclear strategy. The structural realist explana­ tion is indeterminate—­given its security environment, France could have justifiably shifted back toward an assured retaliation posture like the UK or maintained an asymmetric escalation posture. The technological deter­ minism explanation might predict the fact of the asymmetric escalation posture, but not its form. France did retire some systems and, though it is upgrading its ASMP and SLBM capabilities, it is fielding fewer forces than it could. The strategic culture explanation—­an enduring Gaullist in­ dependent streak—­cannot be excluded. However, as during the Cold War, if French nuclear capabilities are aimed at enabling an independent foreign and security policy, they can do so equally well with an assured retalia­ tion posture. Indeed, a strategic culture explanation focusing on the en­ durance of Gaullism has a difficult time accounting for France’s continu­ ity in nuclear posture since it was, after all, de Gaulle’s nemesis Francois Mitterand who guided France’s post–­Cold War nuclear decisions. The en­ durance of the asymmetric escalation posture, adopted as a hedge against 70. Tertrais, “The Last to Disarm?” p. 260. 71. Ibid., p. 261. 72. See Yost, “Nuclear Weapons Issues in France,” p. 69. 73. Quoted in Rynning, Changing Military Doctrine, p. 175.

178

Chapter 6

the re-­emergence of old threats and the evolution of new threats, is most consistent with optimization theory. To hedge against past and future po­ tential threats, France was free to maintain an asymmetric escalation pos­ ture, though in a reduced form given the collapse of its primary existential threat. France’s post–­Cold War asymmetric escalation posture is not simply a product of posture inertia. As evidenced by the retiring of the Soviet-­ specific capabilities, France consciously reevaluated its posture and chose to continue to adopt an asymmetric escalation posture. Optimization the­ ory is the only explanation that predicts both that France would explicitly reassess its posture with the dissipation of its existential threat and that it would nevertheless continue to adopt an asymmetric escalation posture, albeit of different character than before. Thus, in explaining both Cold War and post–­Cold War nuclear posture, my theoretical explanation fares best, accounting for both periods. That is, although France adopts an asymmetric escalation nuclear posture during both the Cold War and afterward, it does so for different reasons in the two periods: it had no choice but to adopt an asymmetric escalation pos­ ture if left by itself to face the Warsaw Pact’s overwhelming conventional superiority during the Cold War, and then it was blessed with security and resources that enabled it to hedge with the same posture once its existential security threat disappeared. Only optimization theory captures these two different rationales for France’s asymmetric escalation posture.

Chapter Seven

Israel

Israel is the world’s oldest closet nuclear state. For more than forty years it has neither confirmed nor denied its possession of nuclear weapons, and has vowed not to be the first to “introduce” nuclear weapons into the Middle East (with the definition of “introduce” being left intentionally vague). But it has circulated enough credible rumors and hints that it does possess a nuclear weapons capability to lead most of the world to believe that Israel was the world’s sixth nuclear power. Indeed, not until December 2006—­almost exactly forty years since Israel is believed to have been capable of assembling nuclear weapons—­did an Israeli prime minister come close to conceding that Israel was a nuclear weapons state, when Prime Minister Ehud Olmert included Israel in a list of nuclear weapons states, stating that Iran was “aspiring to have nuclear weapons, as France, America, Russia and Israel.”1 The statement was quickly disavowed as a misquotation, but there are few in the world who believe that Israel does not possess nuclear weapons. Indeed, the Bulletin of the Atomic Scientists has included Israeli forces in their regular “Nuclear Notebook” summaries of worldwide nuclear forces since 2002, providing details—­albeit sketchy ones—­about a nuclear force whose existence has never been openly admitted or declared. How, then, should we treat Israel’s nuclear posture, a posture comprising capabilities that have never been confirmed? This chapter is based on 1. Quoted in an English language interview to the German television station N24 Sat 1. See “Olmert’s Nuclear Slip-­up Sparks Outrage in Israel,” Times Online (UK), December 12, 2006.

179

180

Chapter 7

the best, albeit limited, open-­source literature on Israel’s nuclear program, which has been most notably advanced by Avner Cohen.2 Based on incredible access to Israeli officials and declassified U.S. documents, Cohen has pieced together the most authoritative history of Israel’s nuclear program. Drawing on the work of Cohen and others, I fit Israel’s nuclear posture into my broader comparative typology. I identify two distinct phases in Israel’s nuclear history. The first phase from 1967 through 1991 involved Israel’s use of “opacity” and “ambiguity” to establish a catalytic nuclear posture, deliberately using the threat of unsheathing its undeclared nuclear capabilities to elicit American materiel and diplomatic support for Israel. Exploiting both the clear American interest in Middle East stability and its growing special relationship with Washington, Israeli leaders adopted a concerted two-­ pronged strategy that relied on conventional capabilities to deter Israel’s Arab threats and manipulated the risk of nuclear escalation to compel its American patron state to ensure provision of military materiel. This was the strategy intended prior to the 1967 Six Day War, during which Israel had the capability to assemble two to three rudimentary fission devices. But Israel’s lightning quick successes obviated the need to manipulate nuclear risk. During the 1973 Yom Kippur War, however, the combined Syrian-­ Egyptian surprise assaults against Israel, and the prospect that Syria might cross the Jordan river and threaten Israeli population centers, led Israeli leaders to send specific nuclear signals that would only be detected by the United States—­not Syria—­to catalyze the American airlift of conventional supplies. Ultimately, superior conventional forces provided the bedrock for Israeli security while opaque nuclear weapons ensured America’s support if necessary. An open declaration of nuclear capabilities might have risked devastating sanctions on Israel, which continues to remain outside the NPT, or a significant proliferation cascade effect within the Middle East. The catalytic posture thus optimized Israel’s broader security constraints more 2. Cohen, Israel and the Bomb; “Nuclear Arms in Crisis under Secrecy: Israel and the Lessons of the 1967 and 1973 Wars,” in Lavoy, Sagan, and Wirtz, pp. 104–­24; Cohen, “Israel and the Bomb Revisited,” in Cohen and McNamee, Why Do States Want Nuclear Weapons: The Cases of Israel and South Africa (Oslo: Norwegian Institute for Defence Studies, 2005), pp. 5–­12; Cohen and Burr, “Israel Crosses the Threshold,” Bulletin of the Atomic Scientists, vol. 62, no. 3 (May/June 2006), pp. 22–­30; Cohen, “The Last Nuclear Moment,” New York Times, October 3, 2003; Cohen, “Israel: A Sui Generis Proliferator,” in Alagappa, The Long Shadow, pp. 241–­68; Cohen, The Worst Kept Secret: Israel’s Bargain with the Bomb (New York: Columbia University Press, 2010); Shai Feldman, Israeli Nuclear Deterrence: A Strategy for the 1980s (New York: Columbia University Press, 1983); also see Frank Barnaby, The Invisible Bomb: The Nuclear Arms Race in the Middle East (London: I.B. Tauris, 1989); Michael Karpin, The Bomb in the Basement: How Israel Went Nuclear and What that Means for the World (New York: Simon & Schuster, 2006); Evron, Israel’s Nuclear Dilemma; and Hersh, The Samson Option.

israel 181

effectively than an overt assured retaliation or asymmetric escalation posture would have been able to. After the 1991 Gulf War against Iraq, though the United States attempted to assist Israel against Saddam Hussein’s Scud attacks, Israeli leaders began to question the United States’ commitment to defend Israel. The confluence of Washington’s perceived tepid support for Israel during the first Gulf War, the end of the Cold War which caused Israel to fear that American interests in the Middle East would shift, and the Madrid conference which seemed to confirm those worries, triggered significant concerns of American abandonment of Israel. These concerns about American abandonment grew while Israeli concerns about future nuclear acquisition by Iraq and Iran were also becoming increasingly severe. As a result, a decision was made shortly after the Gulf War to acquire capabilities to augment Israel’s nuclear deterrent by incorporating submarine-­based nuclear cruise missiles into Israel’s force structure. The decision to switch to an assured retaliation posture was therefore made in 1991, but there was necessarily a lag time as Israel developed and acquired the capabilities to support the posture. By 1997, after Israel had taken delivery of its first Dolphin-­class diesel submarine, I classify Israel as having fully achieved an assured retaliatory nuclear capability. Though the contours of this posture are vague, there is no evidence that Israel possesses tactical nuclear weapons for battlefield use—­and has not made explicit nuclear deterrent threats against conventional aggression—­ and thus cannot credibly adopt an asymmetric escalation posture. Instead, to reduce the vulnerability of its nuclear-­capable aircraft and its Jericho surface-­to-­surface ballistic missile capabilities, particularly given Israel’s dearth of territory, Israel appears to have developed a survivable sea-­based deterrent to assure retaliation against a nuclear strike or existential threat to the Israeli state. While optimization theory accurately predicts and explains Israel’s catalytic posture in the first phase, it—­along with each of the alternative explanations—­fails to explain why Israel since 1991 has adopted an assured retaliation posture and not an asymmetric escalation posture.

Israel’s Nuclear Posture Phase I (1966–­1 990): Catalytic The Israeli nuclear program was conceived as an insurance policy for the survival of the Jewish state.3 David Ben Gurion believed Israel required nuclear weapons because “for a state born out of the Holocaust and surrounded by the hostile Arab world, not [to acquire them] would have been irresponsible.”4 When Ben Gurion returned to the prime ministership in 3. Cohen, Israel and the Bomb, especially chapters 3–­10. 4. Ibid., p. 9.

182

Chapter 7

1955, he actively pursued the development of Israel’s nuclear deterrent as a means toward self-­reliance, stating: “the future of Israel was not dependent on what the gentiles would say, but on what the Jews would do.”5 Israel’s nuclear program owes its origins to French assistance beginning around 1955. French support aided Israel’s development of the Dimona reactor and associated reprocessing facilities. In exchange, Israel provided support to the French in North Africa, most notably by participating in the 1956 military intervention during the Suez Crisis.6 It was after the Sèvres conference in October 1956, where Israel, Britain, and France formalized their plans for Suez, that Shimon Peres, then Defense Minister, says that he “finalized . . . an agreement for the building of a nuclear reactor at Dimona, in southern Israel, and the supply of natural uranium to fuel it.”7 The nuclear agreement was not explicitly a quid pro quo for Israeli participation in the Suez operation, but both nations embarked on a nuclear weapons program after being abandoned by the United States in Suez—­France in pursuit of strategic independence, Israel seeking insurance. While France had a young but sophisticated indigenous nuclear energy sector, Israel had to look to France for assistance in building its plutonium reactor. Peres negotiated for a Marcoule-­type reactor to be built at Dimona,8 along with a chemical reprocessing plant, the latter being critical for the military application of nuclear energy. The agreement to build Dimona was signed on October 3, 1957, and construction began the following year, which also marked the year in which Israel’s armament development authority (RAFAEL) was established, giving organizational teeth to the nuclear weap­ ons development program. Although de Gaulle terminated French nuclear assistance to Israel, he was not able to do so until May 13, 1960. By then, Dimona was well under construction and Israel was developing the in-­house expertise required to finish building and eventually operate the reactor. In addition, a compromise of sorts was reached such that, even though the French government would terminate assistance, French companies that had already begun their work could continue to do so until completion. Around this time, in June 1960, the Eisenhower administration in the United States became aware of Israel’s nuclear intentions and sought answers as to what was being built at the “textile plant” at Dimona.9 Over the course of the next several months, the “textile plant” became a “metallurgical research laboratory” and by December 1960, the United States openly suspected that Israel was devel­ 5. Quoted in ibid., p. 43. 6. Ibid., pp. 52–­55; also see Richelson, Spying on the Bomb, pp. 236–­43. 7. Peres quoted in Cohen, Israel and the Bomb, p. 54. 8. The power output is still ambiguous, estimated to be between 24 and 40 MWe. 9. Cohen, Israel and the Bomb, p. 85.

israel 183

oping a “capacity to produce nuclear weapons.”10 Ben Gurion responded with a public announcement that Dimona was only for peaceful purposes, specifically for water desalination. Israeli leaders, Ben Gurion and then Levi Eshkol, would play fast and loose about Dimona’s purpose for the next decade, even under sometimes significant pressure from the Kennedy and Johnson administrations, though U.S. officials would also obfuscate over Dimona’s purpose as well. Dimona went critical around late 1963 or early 1964. By 1966, Avner Cohen reports that “the physical infrastructure of the project was completed, including the capability to produce weapon-­grade fissile material, weapon design, and the testing of delivery means. According to Pierre Péan, “the first plutonium extraction tests took place during the second half of 1965,” and Israel had enough plutonium to “manufacture the bomb during 1966, or at the latest early 1967.”11 It is unclear whether Israel received external assistance for bomb designs, but Munya Mardor, who was involved in the development of these first designs, wrote in his memoirs, “On November 2, 1966, a test with special significance was conducted. It meant an end of an era of development, and a step that brought one of our primary weapons-­systems to its final phases of development and production in RAFAEL. . . . The test was completely successful . . . We have waited for that result for many years.”12 Thomas Reed and Danny Stillman believe that the test was either a hydronuclear test or a cold test of a “supercritical prototype bomb.”13 Either test would have given Israel enhanced confidence in its designs, though certainly not as much as an actual fission test. The American intelligence community estimated at this point that Israel could assemble a nuclear weapon in 6–­8 weeks from a decision to do so.14 On the eve of the 1967 Six Days War in June, Israel therefore had the capacity to assemble nuclear weapons, as well as potentially nuclear-­ capable A-­4 Skyhawk aircraft and Jericho surface-­to-­surface ballistic missiles, which were operationalized in 1966. As hostilities with Egypt and Syria reached a fever pitch in June 1967, unbeknownst to the United States, USSR, Egypt, or Syria, Munya Mardor reveals that RAFAEL, the operational arm of Israel’s nuclear weapons capability, had teams “assembling and testing the weapon system, the development and production of which was completed before the war  .  .  . they fully recognized the enormous, perhaps fateful, value of the weapons system that they [had] brought to 10. John McCone, chairman of the U.S. Atomic Energy Commission, cited in ibid., p. 89. 11. Cohen, Israel and the Bomb, p. 231. 12. Mardor quoted in ibid., p. 232. 13. Reed and Stillman, The Nuclear Express, p. 119. 14. Cohen, Israel and the Bomb, p. 232.

184

Chapter 7

operational alert.”15 Cohen writes that this episode involved the improvised assembly of two to three nuclear weapons, but stopped short of fully arming them.16 These were apparently taken as precautionary steps by Israel’s scientific personnel, without any explicit political guidance to do so. Given this capability, how did Israel intend to use it in the event that the Six Day War turned unfavorably for Israel? That is, what was Israel’s primary envisioned employment mode? I find that Israeli posture during the Six Day War was explicitly catalytic and, though it did not require execution because of the speed and ferocity of Israeli victories, the intention would have been to compel the United States to assist Israel. The deterrent signal would not have been sent to Egypt or Syria directly, but to the United States. There are several features of Israel posture at this point that classify it as catalytic. First, although the United States was becoming an increasingly reliable patron of Israel by 1967, its primary interest was to avoid the nuclearization of the Middle East because of the significant problems that would create for American freedom of action in the region.17 That is, so long as nuclear ambiguity did not trigger a nuclear cascade in the Middle East, the United States could look the other way—­but if Israel were to publicly brandish its nuclear capabilities, let alone use them, it would be disastrous for U.S. nonproliferation goals and interests in the region. Second, nuclear weapons were not integrated into the IDF’s force structure and doctrine—­Israel’s primary deterrent power was still conventional, and nuclear operations were completely decoupled from those forces. Cohen writes that, in the mid-­1960s, before leaving office, Ben Gurion “decided not to restructure the IDF and its military doctrine so as to base it on nuclear weapons. Rather, he would continue to develop a nuclear option without changing the IDF doctrine and basic organization.”18 Similarly, Esh­ kol later persisted with this decision, determining that Israel should not reorient the IDF toward a nuclear strategy.19 As such, Israel did not have the operational capacity to use nuclear weapons in a military sense. And the military viewed the nuclear debate as “irrelevant to their own military mission . . . they were committed to the notion that the IDF mission was to prevent these scenarios from ever coming to pass” by maintaining a powerful conventional force structure capable of meeting any coalition of threats.20 Indeed, Avner Cohen concludes that Israel’s nuclear posture over time—­even today—­is “not about providing another advanced weapon sys­ 15. Mardor quoted in ibid., p. 273. 16. Ibid., p. 274; Cohen, The Worst Kept Secret, p. 73; also Richelson, Spying on the Bomb, p. 242. 17. See Cohen, Israel and the Bomb, chapter 11. 18. See Cohen, Israel and the Bomb, pp. 137–­38. 19. Ibid., p. 231. 20. Cohen, “Nuclear Arms in Crisis under Secrecy,” p. 111.

israel 185

tem to the IDF; it is not about Israel’s military doctrine . . . [it is] insulated from the mainstream military . . . and its liaison with the military must be limited and strict. By and large, it is a political matter, not a military one.”21 As such, Shimon Peres and Levi Eshkol devised Israeli nuclear strategy on the eve of the Six Day War to be two-­tiered. According to Cohen, the primary aim was that “the specter of Israeli nuclear weapons serves as insurance vis-­à-­vis the United States, a strong political incentive for America to keep Israel conventionally armed, believing that a sufficiently armed Israel would not have to use its nuclear option. [Although] this was not among the primary rationales that had led Ben Gurion to the nuclear project, under Eshkol, the insurance component became the central aspect of Israel’s national security strategy.”22 Without a formal commitment from the United States—­or even with it for that matter—­Israeli leaders calculated that the risk of nuclear escalation was a strategic tool to compel the United States to assist Israel if necessary. Of course, the secondary role of its nuclear weapons capability, should the United States fail to intervene, was an instrument of last resort in a supreme national emergency. In other words, though it would not be a trivial logistical matter, Israel would be able to use its nuclear capabilities if its survival were imminently threatened. The red-­lines at which this might be contemplated were defined as Arab military penetration into Israeli population centers, the destruction of the Israeli Air Force, and nuclear use against Israel.23 But the primary signal to be sent was to the United States to compel its support of Israeli conventional materiel as well as ensure its diplomatic and potentially military support in the event that a significant threat materialized against Israel. Before the Six Day War, then, Israeli leaders had already conceived of the catalytic nuclear posture. On its eve, Israel had made no direct deterrent threats to Egypt or Syria, nor any indirect deterrent threats to the United States—­all of which assumed Israel was not yet a nuclear-­weapons state. Shimon Peres, at one point in the run-­up to the war, suggested “reveal[ing] to the United States that it had a nuclear capability, possibly making an oblique declaration or even conducting a nuclear test.”24 There did not appear to be any suggestions to reveal the capability to Egypt, Syria, or the Soviet Union. Nevertheless, Cohen writes that “although contingency planning for a demonstration test did occur, those who participated in it understood clearly that only a truly ‘existential threat’ could provoke the leadership to consider such a step.”25 It is important to note that the first 21. Cohen, “Israel: A Sui Generis Proliferator,” pp. 250–­51. 22. Ibid., p. 236. Emphasis added. 23. Ibid., pp. 236–­37. 24. Cited in ibid., p. 275. 25. Cohen, “Nuclear Arms in Crisis under Secrecy,” p. 113.

186

Chapter 7

declaration of Israel’s nuclear capability was to be a demonstration test “over some remote desert area as a political signal, not to actually use the devices militarily,” in order to compel the United States to assist Israel and perhaps to have an ancillary war-­termination effect.26 Indeed, consistent with the catalytic posture, Cohen has found recent evidence that “preliminary contingency plans were even drawn up for how such improvised devices could be used in a manner that would demonstrate nuclear capability short of a military use. An unpopulated site was even chosen.”27 The idea was that, in a doomsday scenario, Israel would threaten to demonstrate its nuclear capability to compel U.S. intervention to ensure its survival. Never­ theless, Peres’s suggestion was rejected “outright” by Eshkol and, with Israel’s rapid and decisive victory, rendered moot.28 Prior to the war, during the weeks in which the crisis unfolded, Israel made no nuclear deterrent threats against the Arab states and did not attempt to “coerce” the United States by revealing its nuclear weapons status.29 However, had the course of the war gone differently, the available evidence suggests that the latter likely would have been the Israeli strategy. Instead of an existential deterrent strategy that would attempt to primarily deter the Arab states, Israel’s strategy seems to have been—­as early as 1967—­to direct its primary deterrent signal to the United States to compel its intervention on Israel’s behalf. With Israel’s decisive victory against the Arab coalition in the 1967 war, Israel’s conventional superiority was established as its primary deterrent. With respect to nuclear weapons, President Nixon and Prime Minister Meir reached an understanding that Israel would conduct “no test, no declaration and publicity, no joining the NPT, and possibl[y reached] a related understanding about what the commitment to ‘nonintroduction’ meant operationally (i.e., keeping the weapons’ cores unassembled and under the custodianship of the prime minister).”30 Israel’s nuclear capability could not be oriented for military use at this point; it was an entirely political capability designed to keep Washington in the game on Israel’s behalf against its Arab adversaries. Although the events between 1967 and 1973 are beyond the scope of this inquiry, by 1970, Israel’s nuclear weapons capability was an open secret, when the New York Times revealed that: “For at least two years, the United States Government has conducted its Middle East policy on the assumption that Israel either possesses an atomic bomb or has component parts available for quick assembly.”31 Israeli leaders, of course, 26. Cohen, The Worst Kept Secret, p. 72. 27. Cohen, “Crossing the Threshold.” 28. Cohen, The Worst Kept Secret, p. 68. 29. Cohen, “Nuclear Arms in Crisis under Secrecy,” p. 114. 30. Cohen, The Worst Kept Secret, p. 27. 31. Hedrick Smith, “US Assumes the Israelis Have A-­Bomb or Its Parts,” New York Times, July 18, 1970, 1.

israel 187

neither confirmed nor denied the report. Instead, they stuck to the formulation that Israel would not be the first to introduce nuclear weapons into the region, although Israeli leaders have also persistently insisted that neither would they be second. Israel’s neighboring Arab states have been largely unperturbed by Israel’s nuclear capability and have instead focused on how to redress the conventional balance of power to regain the territories lost in the Six Day War.32 Israeli complacency following the Six Day War and after the 1970 War of Attrition laid the groundwork for the surprise Arab assaults in the 1973 Yom Kippur War. The combined Arab offensives were executed simultaneously on October 6, 1973, on Yom Kippur, the most solemn high holiday in the Jewish faith. Israeli intelligence learned of the impending attack only several hours before it began. Israel was thus caught flat-­footed when Egyptian and Syrian forces launched massive offensives at Israel. Because of the sheer mass of armored columns breaking for the Israeli front lines, the Israeli Air Force (IAF) was relegated to close-­air support missions in the first phase of the war and was thus sidelined as an offensive arm. Several days into the war, the possibility of an Israeli defeat—­previously considered highly improbable—­ grew alarmingly possible. The Israeli armed forces lost roughly 500 tanks and almost 50 aircraft, including 14 F-­4 Phantoms, in the first three days of the war.33 Under the cover of Soviet-­supplied SAM batteries, the Egyptians broke through the Bar-­Lev line on the Suez Canal and pushed back the Israeli defenses, while the Syrians punched through Israeli deployments on the Golan Heights with almost 1,400 tanks. Moshe Dayan, Israel’s Minister of Defense, panicked and famously feared that “this is the end of the Third Temple” and that Israel’s very survival was threatened.34 Though Egyptian war aims may have been limited to the recapture of the Sinai, there was no way for Israeli leaders to know this at the time. Indeed, given Israel’s thin geography, there were acute fears, particularly by Dayan, that the Syrian offensives in particular could cross the Jordan River and pose an existential threat to Israeli population centers. There was thus similarly no way for Israeli leaders on October 8 to be confident that Syrian objectives were limited. Unlike the Six Day War, when Israeli nuclear capabilities had been irrelevant, the Arabs were quickly approaching Israel’s red-­lines in the Yom Kippur War. Had Syrian tank units crossed the Jordan River, one of Israel’s critical red-­lines would have been crossed. It was at this point that Israel executed its catalytic nuclear posture. At War Cabinet meetings, Dayan 32. See, for example, Zeev Maoz, “The Mixed Blessing of Israel’s Nuclear Policy,” International Security, vol. 28, no. 2 (Fall 2003), pp. 44–­77. 33. See Hersh, The Samson Option, 222. 34. Ibid., 223.

188

Chapter 7

agonized over the deteriorating Israeli situation. Memoirs from the meeting suggest that Dayan feared that Israeli survival might be threatened in a matter of hours, not days, and thus he may have feared that there was only time to make a direct deterrent threat to Syria. Ora Armoni, aide to Yisraeli Galili who was Prime Minister Golda Meir’s closest advisor, writes that: “There was one moment in the Yom Kippur War, when Moshe Dayan panicked, when he thought the Syrians are breaking through to the Golan, and their tanks are heading to Haifa. He proposed to threaten them with extreme measures. Yigal [Alon] and Galili influenced Golda in her decision to reject this idea. Chief of Staff Elazar also opposed. Since that episode we have never made similar threats.”35 Dayan’s proposal to threaten escalation with nuclear weapons was confirmed by Avner Cohen’s own interview with Naftali Lavie, Dayan’s spokesman during the war.36 Critically though, Meir rejected Dayan’s proposal, taking a more sober approach and refusing the idea of a test or a direct deterrent threat to the Syrians. This suggests that Israel did not adopt an existential deterrence strategy since it refused to send a direct deterrent signal to its adversaries. As in 1967, Israel’s nuclear weapons were not integrated into its armed forces and it certainly did not have tactical nuclear weapons suitable for a deterrence by denial mission; but in 1973, these outcomes were the result of conscious strategic choices the Israeli leadership made in its nuclear posture and its operationalization of its nuclear capabilities.37 According to Cohen, Dayan then conceived of an alternative: signaling to the United States that Israel was contemplating breaking out its nuclear capabilities as a mechanism to jolt Washington into action on Israel’s behalf.38 Instead of making a direct deterrent threat to the Syrians, Dayan ordered—­whether on Meir’s direct orders or on his own is still unclear—­an increase in Israel’s nuclear alert status in a fashion that would have been detectable only by the United States (and perhaps the USSR), given their sophisticated imagery and signals intelligence capacity. Cohen writes that, after the initial proposal of sending a direct deterrent signal to Syria was rejected, “Dayan wanted the United States to take notice that things had reached [a point of “last resort”]. That he meant using nuclear weapons (albeit in coded language, as at the time nobody dared call them by name) was confirmed in an interview . . . by Naftali Lavie, who was Mr. Dayan’s spokesman during the war.”39 In order to send this signal, Dayan ordered 35. Ora Armoni, A Friend and Confidant [Hebrew] (Tel Aviv: Ha’Kibutz Ha’Meuhad, 2008), p. 206. Translated by Ehud Eiran on November 4, 2009. 36. Cohen, “The Last Nuclear Moment.” 37. Cohen, “Nuclear Arms in Crisis under Secrecy,” pp. 121–­22. 38. Cohen, The Worst Kept Secret, p. 80. 39. Cohen, “The Last Nuclear Moment.”

israel 189

“operational checks” of some sort on Jericho missiles—­Israel’s nuclear delivery vehicle—­with the silo hatches open, which means the activity would be detected by U.S. imagery intelligence. Cohen confirms that “Israel had put its Jericho missiles, which could be fitted with nuclear warheads, on high alert (the Israelis had done so in an easily detectible way, probably to sway the Americans into preventive action).”40 According to Cohen, Israel’s policy of nuclear opacity meant that the only way “for Israel to signal its distress to the United States, its closest ally, was to display its strategic systems. This evidently was the meaning behind the missile activity.”41 There were also some reports that Israel placed nuclear-­capable F-­4 Phantoms on ready-­alert, but this has not been confirmed. Regardless, the Jericho checks therefore seem to have been a deliberate strategic move by the Israelis, sending an indirect deterrent signal to the United States and not directly to Syria that Israel felt imperiled enough to contemplate revealing, or potentially using, its nuclear weapons capabilities.42 The goal of the operational checks was to compel the Nixon administration to resupply Israel with the necessary conventional materiel to thwart the combined Arab assaults and enable counteroffensives. Until October 8, the United States was dragging its feet on resupplying Israel with this materiel. Indeed, Secretary of State Henry Kissinger was effectively pursuing a strategy of allowing Israel to get its nose bloodied before the United States would intervene.43 However, Hermann Eilts, a member of Kissinger’s staff, notes that the United States detected that Israel was “opening the hatches” and conducting checks on the Jerichos, and he “doubts the [Egyptians or Syrians] were aware of it. I think that it was generally expected that it was to influence to get off our tails and to get some equipment to Israel, and that we did.”44 In an exchange, Eilts and other participants confirm that they 40. Ibid. 41. Cohen, “Nuclear Arms in Crisis under Secrecy,” p. 121. 42. A 2013 report, Elbridge Colby et al., The Israeli “Nuclear Alert” of 1973: Deterrence and Signaling in Crisis (Washington, DC: Center for Naval Analyses, April 30, 2013), questions whether it was intentional “blackmail” or just simply an ambiguous operational check of Jerichos. This report is mostly concerned with whether the Israeli ambassador to the United States explicitly threatened nuclear blackmail to Kissinger. There is little evidence that this was the case. The report is based almost entirely on American sources and provides little insight into Israeli intentions or thinking. With respect to the Jericho checks, the evidence marshaled above based on Israeli officials is stronger in my view that this was a deliberate move by the Israelis and fits with the broader pattern of their posture in this period. Furthermore, the CNA report does not address the statements by Eilts on the American side, or Quandt’s writings and statements to Cohen discussed later that Kissinger initiated the airlift to Israel based on these nuclear movements. 43. See Hersh, The Samson Option, pp. 227–­30. 44. Eilts quoted in Richard Parker, ed., The October War: A Retrospective (Gainesville: University of Florida Press, 2001), p. 121.

190

Chapter 7

believed the checks were aimed at influencing “American policy . . . [because] the others couldn’t see it.”45 William Quandt, who was on the National Security Council at the time, later recounted “how profound the impact of American intelligence reports of these [movements] was on Henry Kissinger,” and that they were “critical to the American decision to initiate the airlift to Israel.”46 Quandt elsewhere elaborates: “We did know around this time, however, that Israel had placed its Jericho missiles on alert. I did not know what kind of warheads they had, but it did not make much sense to me that they would be equipped with conventional ordnance . . . Without being told in so many words, we knew that a desperate Israel might activate its nuclear option. This situation, by itself, created a kind of blackmail potential. ‘Help us, or else . . .’ But no one had to [explicitly] say it.”47 These visible operational checks on the Jerichos were sufficient to catalyze Kissinger and the Nixon administration to resupply Israel with the necessary conventional arms required to preserve the Israeli state. This episode in the 1973 Yom Kippur War strongly confirms Israel’s catalytic nuclear posture. As with the Six Day War, there was an explicit decision not to send direct deterrent threats to Egypt and/or Syria. Instead, in both cases, the intended strategy was primarily to send a signal to the United States to coerce American intervention on Israel’s behalf through the resupply of conventional arms. In 1967, Israeli victories made the signal unnecessary, but in 1973 it was sent several days into the war and helped successfully compel the United States to intervene, facing the risk of Israeli nuclear breakout, or worse, escalation. According to Cohen, the primary aim of the Israeli nuclear posture since the mid-­1960s was “directed at the United States: to remind the Americans that Israeli needs for conventional weaponry must be met.”48 While Avner Cohen describes this posture as existential deterrence, or sometimes as nuclear opacity, the evidence he and others present clearly demonstrates that Israel’s posture was explicitly catalytic in nature—­the primary signal was designed to be sent to the United States to compel its support, not as a deterrent to its Arab opponents.49 And as far as the United States was concerned, according to Quandt: “There has long been a sense among American policy makers that providing Israel with conventional weapons was justified, in part, by the concern that Israel would otherwise feel compelled to rely excessively on a nuclear defense.

45. Ibid., p. 122. 46. Quandt paraphrased in Cohen, The Worst Kept Secret, pp. 80–­81. 47. William B. Quandt, “How Far will Israel Go?” Washington Post Book World, November 20, 1991, p. X7. 48. Cohen, “Nuclear Arms in Crisis under Secrecy,” p. 121. 49. Ibid., pp. 121–­22.

israel 191

This widespread view is rarely mentioned in policy deliberations, but I am convinced that it has had an impact on decisions.”50 Since 1973, and more particularly after the Camp David Accords that established a separate peace with Egypt, Israel’s primary conventional threat prior to that point, Israel has relied on a deterrence doctrine focused on conventional superiority against all regional threats. This goal has been supported by the United States which, since the Johnson administration, has pledged to assist Israel in maintaining a “qualitative military edge” over its opponents. With Egypt neutralized and Iran and Iraq bleeding each other during the 1980s, Israeli conventional forces faced no serious challenges, and certainly nothing that would have required movements in its nuclear alert status. By the summer of 1974, the U.S. intelligence community, in its National Intelligence Estimate, formally concluded that “Israel already has produced and stockpiled a small number of fission weapons.”51 In 1976, Carl Druckett, the CIA’s deputy director for science and technology, estimated that Israel had 10–­20 nuclear weapons “available for use.”52 Since the early 1970s, Israel’s production of plutonium fission, and boosted fission (with tritium), weapons would have continued with steady operation at Dimona. There is speculation that a 1979 “double flash” detected by U.S. Vela satellites in the South Pacific was an Israeli nuclear or neutron bomb test, conducted with South African assistance, though this is presently unconfirmed. Estimates of the size of the Israeli nuclear arsenal vary wildly, but SIPRI and Bulletin of the Atomic Scientists estimates place the arsenal size between 100 and 200 warheads, based on Mordechai Vanunu’s revelations that Dimona produces about 30kg of fissionable plutonium per year and that Israeli weapons designs require 4kg of plutonium per weapon.53 Israel also developed a two-­stage solid-­fuel Jericho II ballistic missile to give its nuclear deterrent greater strategic reach in the late 1980s, in addition to acquiring nuclear-­capable F-­16 and F-­15 aircraft from the United States. The Jericho II’s range is estimated at 1450–­1800km, enabling it to strike targets as far away as western Iran.54 In terms of nuclear infrastructure, Dimona produces and reprocesses the fissile material, while the 50. Quandt, “How Far will Israel Go?” 51. Prospects for Further Proliferation of Nuclear Weapons, A Special National Intelligence Estimate, Director of United States Central Intelligence Agency, August 23, 1974, p. 20. 52. Cited in Richelson, Spying on the Bomb, p. 272. 53. Robert Norris and William M. Arkin, “Israeli Nuclear Forces, January 2000,” SIPRI Yearbook 2000 (Stockholm: Almquist & Wiksell, 2000), pp. 494–­95; NRDC Nuclear Notebook, “Israeli Nuclear Forces 2002,” Bulletin of the Atomic Scientists, September/October 2002, pp. 73–­75; Harold Hough, “Israel’s Nuclear Infrastructure,” Jane’s Intelligence Review, vol. 6, no. 11 (November 1994), p. 508. 54. NRDC Nuclear Notebook, “Israeli Nuclear Forces 2002,” p. 74.

192

Chapter 7

weapons are believed to be designed and assembled by RAFAEL outside Haifa, at a facility known as Division 20. Fifty Jericho IIs are believed to be stored in caves at the Zekharyah missile base.55 It is believed that Israel’s arsenal is recessed during peacetime and thus requires several hours to operationalize and mate with delivery vehicles, though during a crisis, Israel could heighten its alert status.56 In addition, the Jericho deployment pattern occurs in the center of Israel, which would be the last to fall in a conventional attack. This, according to Jane’s Intelligence Review, “clearly signals that Israel does not consider the nuclear option to be a first-­strike weapon but a last resort device that would only be used if the state of Israel is threatened with annihilation.”57 Indeed, in the 1970s, Prime Minister Yitzhak Rabin evidently “vehemently opposed” the development of tactical nuclear weapons or any move to operationalize Israel’s strategic assets toward militarily useable capabilities—­IDF doctrine was still entirely conventional.58 Avner Cohen writes that, “In the period between the Yom Kippur War and the first Gulf War (1991) Israel significantly expanded its nuclear capability, but it did not move to establish secured, second-­strike, capability,” and indeed was judged to be vulnerable to a first strike.59 In fact, between 1967 and the first Gulf War, Cohen notes that “little [had] changed” in Israel’s nuclear posture: “nuclear weapons may have important symbolic and political value but lack genuine military value and should not be recognized as military weapons systems.”60 Cohen further writes that Israel’s catalytic (Cohen refers to the posture using the Hebrew word for opacity: amimut) nuclear posture persisted and there has been no shift to an asymmetric escalation posture: “It became an Israeli article of faith that as long as the conflict remained conventional, Israel’s military doctrine must remain conventional as well.”61 Israeli deliberations about nuclear red-­lines ran into the following dilemma: nuclear use once Arab armies reached Israeli population centers would be “too late to be militarily effective,” but using them “to preempt Arab army troops on their way to the border was deemed too early to justify use.”62 Israel’s nuclear posture has therefore always carried with it the implicit pledge of “no first use.”63 The consensus was therefore 55. Ibid. 56. See Shannon N. Kile, Vitaly Fedchenko, and Hans M. Kristensen, “Israeli Nuclear Forces, January 2008,” SIPRI Yearbook 2008 (Stockholm: Almquist & Wiksell, 2009), p. 376. 57. Hough, “Israel’s Nuclear Infrastructure,” p. 508. 58. Cohen, “Israel and the Bomb Revisited,” p. 9. 59. Ibid. On vulnerability to a first strike, see Harold Hough, “Could Israel’s Nuclear Assets Survive a First Strike?” Jane’s Intelligence Review, vol. 9, no. 9 (September 1997). 60. Cohen, The Worst Kept Secret, p. 81. Emphasis added. 61. Ibid., p. 67. 62. Ibid., pp. 77–­78. 63. Ibid., p. 78.

israel 193

that Israeli use of nuclear weapons in this phase would always have to be demonstrative and employed as a political, rather than military, signal—­ particularly to the United States to convey the seriousness with which Israel viewed an unfolding crisis.

Explaining Israel’s Nuclear Posture Phase I: Catalytic Why did Israel gamble on American intervention and select a catalytic nu­ clear posture in this period, when almost every theory of international security would predict that it would adopt a much more explicit deterrent against a hostile Arab coalition intent on its destruction? Israel’s initial quest for nuclear weapons was clearly security-­driven. Zeev Maoz identifies the three primary concerns facing the Israeli state in the late 1950s and early 1960s when the decision to acquire nuclear weapons was made: “(1) the prospect of a unified Arab coalition starting an all-­out war aimed at the total destruction of the Jewish state; (2) the military advantage in both quantitative terms  .  .  . and qualitative terms  .  .  . that such a coalition would enjoy; and (3) the widespread international support (including from the Soviet Union) that this coalition would face—­compared with the political isolation that Israel could anticipate.”64 In particular, the Arab arms relationship with the USSR gave the Egyptian and Syrian coalition roughly a 2:1 quantitative interceptor, bomber, and armored tank advantage over Israel.65 Given that extreme security situation, one might anticipate that Israel would adopt an asymmetric escalation nuclear posture to blunt a conventional attack against her. However, by 1967, when Israeli nuclear posture was being formulated, conditions (2) and (3) were becoming invalid, partly because of Israel’s burgeoning arms relationship with the United States, which reoriented Israel’s deterrent posture toward conventional superiority to meet combined Arab threats (a condition probably only achieved around 1973) but also provided Israel, critically, with a possible third-­party patron to enable a catalytic posture. First and most important, Israeli leaders believed that they had a U.S. commitment against the Soviet-­backed Arab states owing to American interests in the region.66 So long as Israel abided by nuclear ambiguity—­a “ ‘don’t ask, don’t tell’ stance”—­the United States “gave Israel a degree of 64. Maoz, “The Mixed Blessing,” p. 46. 65. See, for example, Jon Glassman, Arms for the Arabs: The Soviet Union and War in the Middle East (Baltimore: Johns Hopkins University Press, 1975); SIPRI, Arms Trade Register: Arms Transfers to the Third World (Cambridge, MA: MIT Press, 1975); International Institute for Strategic Studies, The Military Balance 1964–­7 (London: IISS 1964–­67). 66. See Abraham Ben-­Zvi, Decade of Transition: Eisenhower, Kennedy, and the Origins of the American–­Israeli Alliance (New York: Columbia University Press, 1998); and

194

Chapter 7

political cover” as well as substantial frontline arms (e.g., F-­4 Phantoms and later F-­16s and F-­15s).67 Avner Cohen and William Burr situate U.S. policy after the discovery of Dimona in 1960: “From the U.S. perspective, Israel was a small, friendly state, albeit one outside the boundaries of formal U.S. alliance or security guarantees . . . Most significantly, Israel enjoyed unique domestic support in America,” support that would make it virtually impossible for the United States to stand idly by if Israel came under threat.68 Given that Israel was a democratic and friendly state to the United States in a region of vital strategic importance both during the Cold War and after, Washington could ill afford to see Israel placed under significant threat. But precisely because Israel was outside the United States’ formal alliance structure, Israeli leaders seem to have calculated that they might require some leverage to ensure that the United States did indeed act when they felt under threat, not when American leaders calculated that that was the case. This became the primary utility of Israel’s nuclear capability by 1967. While this would be sufficient to predict a catalytic posture, since availing itself of American protection could keep Israel’s nuclear capabilities opaque and thereby avoid triggering a regional nuclear arms race, the availability of the United States as a potential patron also provided Israel with the necessary conventional arms to offset the Arab coalition. This would further enhance the sufficiency of a catalytic posture by mitigating the need for a first-­use nuclear posture against conventionally superior enemies. While the Arab states were making their unsuccessful bid for regional unity, Israel was slowly expanding its offensive military forces during the 1960s, taking advantage of inter-­Arab feuding and the relative calm on the Arab-­Israeli front. With the aid of France and, increasingly, the United States, Israel augmented its air force with 100 French Magister fighters and 72 highly capable Mystere IIICs which were comparable to contemporary Soviet MiGs.69 In addition, the Kennedy administration slowly opened the spigot of U.S. arms transfers to Israel in what would later prove to be Israel’s most fruitful relationship. Under the cover of West Germany, the United States transferred 200 M-­48 Pattons—­a tank that finally gave the Israelis some legitimate ground punch against the Soviet T-­55s, especially once the Israelis extended the range on its 105mm Howitzer gun—­and 100 M-­113 armored personnel carriers for infantry maneuver warfare.70 Although it

Michael Brecher, Decisions in Israel’s Foreign Policy (London: Oxford University Press, 1975), pp. 378–­81. 67. Avner Cohen and William Burr, “Israel Crosses the Threshold,” Bulletin of the Atomic Scientists, vol. 62, no. 3 (May/June 2006), pp. 29–­30. 68. Ibid., p. 24. 69. SIPRI, Arms Trade Register: Arms Transfers to the Third World, p. 53. 70. Ibid., pp. 53, 55.

israel 195

would not be revealed until 1966, the United States also agreed to sell Israel 50 highly potent A-­4 Skyhawk fighter/bombers which, once incorporated into the IAF, would significantly enhance their offensive power against the Arab states. The IDF was thus developing its “mailed fist” while the IAF was training to deal with an Arab force flying qualitatively comparable planes with a twofold quantitative advantage.71 But even though Israel had about 300 combat aircraft, piloted by highly skilled and experienced IAF pilots, 900 medium and heavy tanks, and 200 long-­range artillery pieces, both Syria’s and Egypt’s expanding relationship with the Soviet Union in the 1960s gave them a significant quantitative advantage over the Israelis, especially in the air.72 The Egyptians were taking delivery of and training on 150 MiG 15s/17s, 80 MiG-­19s, and 130 MiG-­21s, the fastest and most maneuverable plane manufactured at the time, and hundreds of T-­54/T-­55 tanks which, except for the M-­48s, the Israelis could not match.73 In the air, even though IAF pilots closed the qualitative differential between the Arabs and Israelis, the MiG-­21 was judged to be a superior interceptor to the A-­4 owing to its greater range, speed, and maneuverability.74 In addition, a thousand Soviet advisors had been dispatched to help mechanize and armorize Egyptian ground divisions so that they could be decisive as a ground blitz.75 But most worrying for the Israelis was the public transfer of surface-­to-­surface missiles from the USSR to Egypt which “could hit any target south of Beirut” to supplement the Egyptian bomber force which consisted of 20 Tu-­16 heavy bombers armed with air-­to-­surface missiles that could inflict substantial damage on Israel’s highly concentrated strategic centers.76

71. See Michael B. Oren, Six Days of War: June 1967 and the Making of the Modern Middle East (Oxford: Oxford University Press, 2002), p. 17. 72. Anthony Cordesman, After the Storm: The Changing Military Balance in the Middle East (Boulder: Westview Press), p. 182; IISS, The Military Balance 1964–­7, London: IISS, 1964–­7; SIPRI, Arms Trade Register: Arms Transfers to the Third World, 43–­46, 63–­65; Glassman, Arms for the Arabs, pp. 24–­26. 73. Cordesman, After the Storm; IISS, The Military Balance 1964–­7; SIPRI, Arms Trade Register; Glassman, Arms for the Arabs. 74. Edward Laurence and Ronald Sherwin, “Understanding Arms Transfers through Data Analysis,” in Uri Ra’anan, Robert Pfaltzgraff, and Geoffrey Kemp, eds., Arms Transfers to the Third World: The Military Buildup in Less Industrial Countries (Boulder: Westview Press, 1978), pp. 102–­4. Even allowing for IAF pilot advantages, the MiG-­21 was assigned a “platform utility” 1.3 times higher than the A-­4 Skyhawk and Mirage IIIs. Not until Israel would take receipt of the F-­4 Phantom would they enjoy a qualitative advantage over the MiG-­21. 75. Glassman, Arms for the Arabs, p. 24. 76. Oren, Six Days of War, p. 16; IISS, The Military Balance, 1965; SIPRI, Arms Trade Register: Arms Transfers to the Third World, p. 44; Nadav Safran, From War to War (New York: Pegasus, 1969), p. 215.

196

Chapter 7

Israel’s calculations before the Six Days War was that the Arab advantages could only be neutralized with a preemptive strike—­a feat that the IAF achieved while Egyptian pilots were eating breakfast on the morning of June 5, 1967. The qualitative and quantitative advantage the Arab coalition enjoyed lay burning on Egyptian airfields. With command of the skies, the Israelis methodically decimated the Egyptian ground forces and then the Syrians and Jordanians in turn. After Israel captured the Sinai, the West Bank, and the Gaza Strip, the superpowers called for a ceasefire. The landscape of the Middle East had been violently disrupted in six days and a new status quo was born. Israel was no longer plagued by fears of an existential threat and, indeed, Yair Evron writes that “the balance of military power was found to be in Israel’s favor to a previously unimagined extent. Moreover, this Israeli superiority was perceived as such by the Arab world. In the second place, American political and military backing for Israel during and following the war, seemed to promise that the arms race ensuing after the war would not undermine the demonstrated Israeli military superiority.”77 Given this emerging conventional military advantage, by 1967, when its nuclear posture was being formulated, Israel no longer feared an existential threat from its Arab neighbors. Zeev Maoz writes that by this point, “the gravity of the threat [the Israeli nuclear weapons capability] was supposed to address had significantly receded  .  .  . Most important, the balance of conventional military forces had started to tilt heavily in Israel’s favor.”78 Thus, Israel, just as it was reaching critical thresholds in its nuclear weapons capabilities, had the availability of a highly incentivized U.S. patron and faced no conventionally superior proximate offensive threats, even in aggregate. Optimization theory would therefore predict that Israel would adopt a catalytic posture—­so long as it believed in the reliability of the U.S. commitment and Israel’s ability to calibrate nuclear signals to compel American intervention when required. Israel’s conventional forces could thwart even combined Arab assaults and its ambiguous nuclear weapons could thus be reserved as a backstop to guarantee Israeli survival, and as a means to compel the United States to maintain Israeli conventional superiority. Israel’s catalytic posture and its associated ambiguity had the additional security benefit of decreasing—­but not eliminating—­the pressures for reactive nuclear proliferation in the region. Avoiding a cascade of regional proliferation was critical to Israel’s security because nuclearization by the Arab states could (1) completely neutralize Israel’s conventional superiority or (2) provide Arab states with a nuclear shield behind which to aggress

77. Evron, Israel’s Nuclear Dilemma, pp. 49–­50. 78. Maoz, “The Mixed Blessing,” p. 52.

israel 197

against Israel. The ambiguous catalytic posture also helped Israel avoid costly international sanctions should it break out its nuclear capability. Thus, optimization theory provides a strong explanation for Israel’s catalytic nuclear posture, which optimized the security constraints Israel faced and the availability of an increasingly close American patron. A structural realist explanation has a more difficult time generating a determinate prediction for Israeli nuclear posture between 1967 and 1991. Given historical concerns about a combined Arab conventional threat, and having largely status quo preferences, structural realism would argue that Israel should have opted for self-­reliance and an asymmetric escalation nuclear posture to deter Arab conventional assaults which were supported by proxy from a nuclear-­armed Soviet Union. That Israel would employ nuclear weapons to compel the United States to keep Israel conventionally armed, and not as a direct deterrent against its Arab adversaries, flies in the face of self-­help structural realist expectations. Even if one takes a broader view of security environment and alliance patterns, at the very least, structural realism would predict that Israel would employ its ambiguous arsenal to directly deter its Arab opponents with veiled hints of a nuclear threat should certain predefined red-­lines be crossed, not indirectly in a catalytic posture. That is, even if Israel wanted to avoid provoking the wider nuclearization of the Middle East, structural realism would have a difficult time explaining why it would develop nuclear weapons as the ultimate self-­help tool and then obscure their existence, revealing them only to the United States, rather than directly to its adversaries. For structural realists, Israel’s hostile threat environment ought to have demanded self-­help behavior. Structural realism cannot explain why Israel developed the ultimate insurance tool—­a nuclear arsenal—­and then chose to utilize it to outsource its protection to the United States rather than establishing a robust direct deterrent capability against its adversaries. Technological determinism, similarly, has difficulties explaining the persistence of Israel’s catalytic posture for more than twenty years when Israel certainly had the capability and technological sophistication to move toward a more robust assured retaliation or asymmetric escalation posture. In the early period of Israel’s nuclear weapons program, technological determinism would predict something akin to nuclear ambiguity when its fissile stockpile, weapons designs, and delivery capabilities were limited so that Israel did not render itself vulnerable to a preventive strike. However, the technological determinism hypothesis makes no prediction about whether Israel would use that ambiguous capability to directly deter Arab opponents or indirectly in a catalytic posture. But as Dimona produced a steady supply of fissile plutonium, and as Israel had the scientific, financial, and technical base to expand its nuclear posture toward even an asymmetric escalation posture, the technological determinism explanation cannot

198

Chapter 7

explain Israel’s persistent adherence to a catalytic posture for more than twenty years. That is, when Israel had a stockpile of perhaps 100 warheads in the 1980s and had operational Jericho I and II ballistic missiles as well as a potent conventional military with largely normal civil-­military structures, why did it not abandon the catalytic posture in favor of a direct deterrent posture that did not rely on American patronage? The technological determinism explanation cannot offer a compelling account. Israeli strategic culture is difficult to measure, but for the purposes of this case, I find that the expectations of an account based on Israeli strategic culture collapses to the expectations of structural realism. Since Ben Gurion, Israeli leaders have had an acute sense of their state’s vulnerability to a hostile Arab coalition and the need for self-­reliance given the history of persecution faced by the Jewish people. Ben Gurion’s speech to RAFAEL in 1963 outlined his worldview and laid the basis for Israel’s largely realpolitik strategic culture over time: I do not know of any other nation whose neighbors declare that they wish to terminate it, and not only declare, but prepare for it by all means available to them. We must have no illusions that what is declared every day in Cairo, Damascus, Iraq are just words. This is the thought that guides the Arab leaders . . . Our numbers are small . . . There is one thing, however, in which we are not inferior to any other people in the world—­this is the Jewish brain . . . And the Jewish brain does not disappoint; Jewish science does not disappoint . . . I am confident . . . that our science can provide us with the weapons that are needed to deter our enemies from waging war against us. I am confident that science is able to provide us with the weapon that will secure the peace, and deter our enemies.79

Ben Gurion’s belief in the power of the atom to ensure Israel’s survival in the face of its hostile security environment is consistent with a realpolitik strategic culture. As such, however, the choice to eventually rely on indirect rather than direct deterrence is puzzling, as noted earlier. Even though Ben Gurion, Eshkol, Meir, and prime ministers since have made conventional military superiority the central pillar of Israeli security, given the historical memories of World War II and the repeated attacks against the Israeli state, it is puzzling from a realpolitik lens as to why Israeli leaders would rely on nuclear ambiguity and not develop a full-­blown asymmetric escalation nuclear posture to directly deter Arab threats against Israel.

79. Ben Gurion quoted in Cohen, Israel and the Bomb, p. 13.

israel 199

In the catalytic phase, optimization theory offers the most compelling explanation for Israel’s choice of nuclear posture. Balancing the availability of its American patron against the risk of triggering a cascade of nuclear proliferation in the region, Israel chose one primary route to ensuring its security: the development of conventionally superior ground and air forces to meet any possible coalition of Arab military threats and an ambiguous nuclear posture to catalyze American support and supply of those conventional forces if necessary. Nuclear weapons were not used to deter Arab opponents directly. Rather, they were operationalized toward a catalytic posture that sought to compel the United States to honor its support for Israel in the way that most mattered for Israeli leaders: the supply and replenishment of conventional arms. Deterrence was indirect, but in operationalizing its nuclear arsenal in such a fashion, Israel avoided provoking its Arab opponents to seek nuclear arms themselves. Explanations based on structural realism, technological determinism, and strategic culture thus all struggle to explain Israel’s selection of a catalytic posture from 1967 through the first Gulf War, since all three would predict that Israel would adopt the most aggressive nuclear posture it could—­an asymmetric escalation posture—­as a self-­help mechanism to robustly deter conventional Arab attacks.

Israel’s Nuclear Posture Phase II: Assured Retaliation The 1991 Gulf War marked a watershed in Israel’s broader security calculations, and its effects reverberated into Israel’s decades-­old catalytic nuclear posture, triggering a shift to an assured retaliation posture that reduced Israel’s reliance on an American patron. In 1991, Israel was threatened by Iraqi Scud attacks on its population centers, as Saddam Hussein attempted to drag Israel into the Gulf War in order to fracture the U.S. coalition. Faced with repeated Scud launches, Israeli leaders attempted to make veiled deterrent threats toward Iraq, with the potential aim of compelling the United States to rapidly augment the Patriot units deployed in Israel to defend it. Foreign Minister David Levy warned that “whoever attacks Israel won’t live to remember it” and Prime Minister Yitzhak Shamir stated that Israel “would strike back in a most forceful manner.”80 These threats, and the American desire to keep Israel out of the war, did indeed compel Washington to deploy Patriot missile defense systems in Israel, in addition

80. David Levy and Yitzhak Shamir quoted in Shai Feldman, “Israeli Deterrence and the Gulf War” in J. Alpher, ed., War in the Gulf: Implications for Israel (Boulder: Westview Press, 1993), p. 198.

200

Chapter 7

to ordering large numbers of coalition forces to attempt (unsuccessfully) to locate and identify Scud launchers in western Iraq. Iraq was able to successfully use mobile SRBM launchers and “shoot and scoot” techniques to evade detection and destruction in order to launch Scuds at Allied forces and Israel.81 An estimated 38–­44 conventionally tipped Scuds were fired at Israeli population centers and at Dimona. Although Dimona was not directly hit, had it been, it could have resulted in a substantial release of radioactive material. The Israeli experience during the Gulf War generated two lessons.82 First, American pressure on Israel for restraint and what some believed was a tepid defense of Israel against the Scud attacks demonstrated that Israel’s ability to catalyze American efforts on its behalf had limits, particularly when the United States was dependent on other Arab states for its own operations.83 Second, though relatedly, the international landscape at the end of the Cold War meant that relying upon the United States for strategic—­ rather than conventional—­deterrence could be risky for Israel. Without the ability to exploit American fears of Soviet penetration in the Middle East, Israel’s perceived room to compel American intervention was lower. As such, after the Gulf War, Israeli leaders made a strategic decision to rely less upon the United States as a patron, whose reliability in the post–­ Cold War era—­as perceived during the Gulf War and the Madrid peace process—­could not be taken for granted any more. As such, instead of relying on a catalytic posture, Israel appears to be evolving toward an assured retaliation nuclear posture with the addition of a sea-­based capability centered upon German-­made Dolphin-­class diesel submarines armed with nuclear-­capable submarine-­launched cruise missiles (SLCMs) known as the Popeye Turbo.84 Avner Cohen writes: “Lessons from [the Gulf War], compounded with new strategic developments, led to Israel’s decision to establish a new sea-­based strategic nuclear arsenal. By July 2000 Israel completed taking delivery of all three Dolphin-­class submarines it ordered after the Gulf War. It is believed that Israel is on its way to restructure its 81. Michael Gordon and General Bernard Trainor, The Generals’ War: The Inside Story of the Conflict in the Gulf (New York: Little, Brown, 1995), 228–­48; also see Chaim Herzog, The Arab–­Israeli Wars 2nd ed. (New York: Vintage, 2004), 407–­14. 82. See Alpher, ed., War in the Gulf; also Laura Zittrain Eisenberg, “Passive Belligerency: Israel and the 1991 Gulf War,” Journal of Strategic Studies, vol. 15, no. 3 (September 1992), pp. 304–­29. 83. See Dore Gold, “The Gulf Crisis and US-­Israel Relations,” in Alpher, ed., War in the Gulf, pp. 73–­95; Abraham Ben-­Zvi, “The Prospects of American Pressure on Israel,” in Alpher, ed., War in the Gulf, pp. 96–­106. 84. The origins of the nuclear-­capable SLCMs are unclear. Some claim that they are mod­ ified Harpoons, while others say that is unlikely and that the SLCMs are indigenously developed to carry a nuclear payload.

israel 201

nuclear forces into a triad form, with a second-­strike capability.”85 Israel has now ordered two additional submarines, for a total of five, so that it could possibly maintain a continuous patrol if it chose to do so.86 In ad­ dition, a Jericho III with a purported range of 4000km is reportedly in de­ velopment to extend Israel’s land-­based reach to include strategic targets in Iran.87 The shift to sea-­based capabilities is the only logical step if Israel wishes to move to an assured retaliation posture to directly deter nuclear use and coercion against the state. A sea-­based deterrent is understandable given Israel’s lack of territorial depth. With continuous patrolling, it is nearly completely survivable and thus assures Israel a nuclear retaliatory capability, particularly given the limited antisubmarine capabilities of likely Israeli foes today and into the indefinite future. Given the delegative—­in some periods, some might characterize it as nearly praetorian—­nature of Israeli civil-­military relations and Israel’s technical capacity, the command-­and-­ control issues involved with a sea-­based deterrent (hardened communications, negative controls under central authorities) would not be a difficult challenge. There is some speculation that Israel may have developed non-­ strategic nuclear weapons, particularly nuclear artillery shells and atomic demolition munitions, but this has never been corroborated and there is no evidence that Israel has developed tactical nuclear weapons to be integrated into the IDF’s military doctrine.88 In 2000, when a report surfaced that Israel was interested in atomic demolition munitions in the Golan, then-­Deputy Defense Minister Ephraim Sneh denied the report by saying, “[T]his report is truly stupid. The person who wrote it not only doesn’t know, but also doesn’t understand anything.”89 Avner Cohen writes that, as of at least 2010, Israel has “decided against developing tactical battlefield nuclear weapons.”90 As such, there is no evidence that Israel has shifted to an asymmetric escalation posture—­even though it almost certainly has the technical, organizational, and financial capacity to do so. The posture is not currently oriented for a tactical first use of nuclear weapons, but rather almost entirely toward an assured retaliation, second-­strike, capability with the addition of a truly survivable platform: submarines.

85. Cohen, “Israel and the Bomb Revisited,” p. 9. 86. Alon Ben-­David, “Israel Looks to Acquire more German Submarines,” Jane’s Defence Weekly, November 30, 2005. 87. “2009 Annual Defence Report: The Middle East and Africa,” Jane’s Defence Weekly, December 14, 2009. 88. See Kile, Fedchenko, and Kristensen, “Israeli Nuclear Forces, January 2008,” p. 399. 89. Arieh O’Sullivan and Douglas Davis, “Sneh Denies Report Israel Will Seed Golan with Neutron Mines,” Jerusalem Post, March 27, 2000. 90. Cohen, The Worst Kept Secret, p. 82.

202

Chapter 7

Explaining Israel’s Nuclear Posture Phase II: Assured Retaliation It was thus after perceived abandonment by the United States in the Gulf War that Israel began to shift to an assured retaliation posture. As noted previously, although the contours of this posture are extremely vague, there is no evidence that Israel possesses tactical nuclear weapons for battlefield use. Further, Israel has not made explicit nuclear deterrent threats against conventional aggression—­and thus cannot credibly adopt an asymmetric escalation posture. Instead, to reduce the vulnerability of its nuclear-­ capable aircraft and its Jericho surface-­to-­surface ballistic missile capabilities, particularly given Israel’s dearth of territory, Israel appears to have decided to rely upon a survivable sea-­based deterrent to assure retaliation against a nuclear strike or existential threat to the Israeli state. Israel’s selection of an assured retaliation posture is difficult to explain. Optimization theory and, indeed, all the other theories of nuclear posture uniformly predict that Israel should adopt an asymmetric escalation posture as a security hedge, as France presently does. With superior conventional forces against the “inner circle” of Arab threats, and not facing an existential threat, Israel has the choice between an asymmetric escalation and assured retaliation posture. However, endowed with relatively delegative civil-­military relations and a technical and financial advantage against its opponents—­and a severe dearth of territory which precludes a defense in depth strategy—­optimization theory would predict that Israel should select an asymmetric escalation posture. It may yet do so but, presently, the selection of an assured retaliation posture is difficult to explain with my theory. Indeed, this choice is difficult to explain with any of the alternative explanations as well, which all similarly predict an asymmetric escalation posture. Why did Israel shift postures in 1991? Israel still maintained conventional superiority over its primary proximate threats, including Iraq and Iran against which Israel has sufficient geographic buffers and which had been bled by their own nearly-­decade-­long war. And in 1991, there was no acute belief that any regional actor was close to acquiring a nuclear capability—­Iraq’s program was, at the time, believed to have been set back by Israel’s preemptive strike in 1981 as well as by the Gulf War and postwar sanctions, while Iran in 1991 was not judged to be close to a nuclear weapons capability. Instead the biggest shift, as noted earlier, was a realization of the limits of America’s commitment to defend Israel during the Gulf War. In particular, the perceived lack of American protection of Israel against the Scud attacks, combined with the end of the Cold War and the subsequent Madrid Peace Conference, suggested that the United States might reorient its strategic priorities, and that Israel might not be able to rely

israel 203

upon American intercession on its behalf as reliably as before. Avraham Ben-­Zvi, an Israeli academic, argues that the end of the Cold War and the Gulf War illuminated the “vastly reduced . . . importance of Israel as a ‘strategic asset’ to the United States.”91 A year after the end of the Gulf War, Clyde Haberman of the New York Times wrote that relations between Israel and the United States were an “unquiet mess” and that the Bush administration was being accused of “bone-­rattling coldness toward a democratic ally.”92 He further cites an Haaretz editorial by Gidon Samet around the same time which argued that Israel’s security “always depended on our alliance with a strong and supportive superpower,” but “now it is gradually withdrawing that support. In the long run, the end result will be to reduce not only Israel’s ability to honorably support itself financially but also to defend itself.”93 America’s perceived lack of support during the Gulf War rattled Israeli leaders. Then-­Prime Minister Shamir writes in his memoirs about the disquiet created by American behavior during the Gulf War: “But what happened to our usefulness [to the United States]? We seemed not to be in the picture at all except as objects of Saddam Hussein’s fury and perhaps also as its victims. When the storm passed, where would it find Israel? I could offer myself no reply to this question.”94 Israeli leaders were concerned that what they viewed as a vital threat to its national interests—­ballistic missile strikes on their population centers—­was viewed with little concern in Washington. Indeed, Shamir complained about a lack of U.S. sympathy for Israeli suffering under Scud bombardment: “How astounded they [the Israeli public] would have been in those hellish days to learn, as we later did, that Desert Storm’s commander, Gen. Norman Schwartzkopf, dismissed those attacks as little more than lightning storms, or that Defense Minister Cheney had compared them to ‘mere’ terrorist attacks—­whatever that means.”95 Moshe Arens, who was the Israeli Defense Minister during the war, wrote that the shifting geopolitical forces meant Israeli leaders no longer believed they would be able to rely on the United States to intervene on their behalf: “if Saddam Hussein had attacked Israel rather than Kuwait, would the United States and its allies have rushed to Israel’s defense? I posed this question to some of the visiting politicians from abroad during

91. Avraham Ben-­Zvi, “The Prospects of American Pressure on Israel,” in J. Alpher, ed., War in the Gulf, p. 98. 92. Clyde Haberman, “Israel Confronts the Idea of Life without Washington,” New York Times, March 22, 1992. 93. Gidon Samet quoted in ibid. 94. Yitzhak Shamir, Summing Up: An Autobiography (New York: Little, Brown, 1994), p. 219. 95. Ibid., p. 224.

204

Chapter 7

the war, and was invariably told Israel would have received the same assistance as Kuwait, but I remained unconvinced . . . the war served as a reminder that if Israel came under sudden attack, it would have to rely solely on itself for its defense.”96 Avi Shlaim writes that: Israel had traditionally been regarded, not least by herself, as a strategic partner and a strategic asset to the United States in the Middle East. The Gulf conflict was a real eye-­opener in this respect. Here was a conflict which threatened America’s most vital interests in the region and the best service that Israel could render to her senior partner was to refrain from doing anything. Far from being a strategic asset, Israel was widely perceived as an embarrassment and a liability.97

After the Gulf War, Shamir was further concerned by what he perceived to be American hostility toward Israel during the Madrid peace process, in which he detected a further shift in sympathy toward the Arab states. After returning from a trip to Washington, he writes: “In the weeks that followed my return to Jerusalem, there was, I felt, a resurgence of Administration hostility to us which found its expression in a series of ‘misunderstandings’ and ‘unsubstantiated claims.’ ”98 This fear of American abandonment, or in the waning reliability of the U.S. commitment to intercede on Israel’s behalf in favor of “a new Arab-­based strategy” rendered a catalytic posture unsustainable for Israel in the post-­1991 period.99 If Israeli leaders believed that the United States could not be counted on as reliably as a third-­party patron, what nuclear posture would optimization theory predict? With conventional superiority against its primary proximate threats, and with Egypt effectively neutralized since Camp David, if Israel believed that it could no longer adopt a catalytic posture, it faced the choice between an assured retaliation and an asymmetric escalation posture. Israel has largely delegative civil-­military relations, with high degrees of confidence and trust in the conventional military organs and structures that submit largely to civilian control. The line between civilians and military leaders is often somewhat blurry in Israel given the frequency with which Israeli military officers subsequently become civilian political leaders. Nevertheless, the norm in Israel is one of high levels of tactical 96. Moshe Arens, Broken Covenant: American Foreign Policy and the Crisis between the U.S. and Israel (New York: Simon & Schuster, 1995), pp. 215–­17. 97. Avi Shlaim, “Israel and the Conflict” in Alex Danchev and Dan Keohane, eds., International Perspectives on the Gulf Conflict, 1990–­1991 (London: Macmillan, 1993), pp. 77–­78. 98. Shamir, Summing Up, p. 248. 99. Laura Zittrain Eisenberg, “Passive Belligerency: Israel and the 1991 Gulf War,” Journal of Strategic Studies, vol. 15, no. 3 (September 1992), p. 319.

israel 205

delegation to a military that has significant input into defense formulation, perhaps more than in other Western democracies, but that ultimately subordinates itself to Israel’s civilian leadership.100 The IDF has exercised tremendous amounts of tactical independence, but with respect to broader policy, according to Yehuda Ben Meir, “the military has complied, without exception, with the operational directives of its civilian superiors.”101 This professional military structure is ideally suited for an asymmetric escalation posture. Indeed, if anything, given the political power of retired IDF generals, Israel leans closer to a praetorian structure, which—­like Pakistan—­is perfectly suited for an asymmetric escalation posture since the chain-­of-­command between the political and military leadership is virtually seamless. There would thus be few organizational and financial costs to integrating nuclear weapons into Israel’s conventional military forces toward an asymmetric escalation posture. Furthermore, relative to its adversaries, Israel has a significant technical and financial comparative advantage that is precisely the condition under which I would expect a state to adopt an asymmetric escalation posture—­ indeed, it is massively on the short end of manpower and territory and is therefore ill-­placed to adopt a defense in depth strategy coupled with an assured retaliation posture as China and India do. In fact, according to optimization theory, Israel after the Gulf War should follow the same path as France, all the way down the decision tree, adopting an asymmetric escalation posture as a hedge against conventional and nuclear threats against it. Yet, there is no evidence that Israel has integrated nuclear weapons into its conventional force posture—­there is nothing to suggest that the IDF trains for nuclear first use in a conventional conflict. The erroneous prediction of optimization theory is an asymmetric escalation posture. It is important to note that optimization theory is unique in predicting that Israel should have shifted postures at the end of the Gulf War given the perceived abandonment of the United States, even if it fails to predict which posture Israel would ultimately adopt. The structural realist, technological determinism, and strategic culture explanations would also predict an asymmetric escalation posture—­the same as during Israel’s catalytic phase, for the same reasons discussed above. Given the confluence of security concerns Israel faced around this period—­the perceived reduction in American support due to the end of the Cold War and manifested in the Gulf War and the Madrid conference, and the looming specter of Iraqi or Iranian 100. See the excellent Yehuda Ben Meir, Civil-­Military Relations in Israel (New York: Columbia University Press, 1995). The Israeli military has a larger role in strategic planning than in other democracies perhaps, but policy is ultimately set by the civilian leadership; see pp. 144–­45. 101. Ibid., p. 128.

206

Chapter 7

nuclearization—­structural realism would certainly again predict that Israel should adopt an asymmetric escalation posture given such a precarious security position. Technological determinism and strategic culture would predict the same, as discussed earlier in the catalytic section. All of these explanations would predict that Israel should have adopted an asymmetric escalation from the beginning and would have little to say about why Israel might shift postures after the Gulf War, let alone to an assured retaliation posture. Israel’s post-­1991 decision to shift to an assured retaliation posture is therefore not predicted by any of the explanations for states’ nuclear posture. So, either I have misclassified Israel’s current nuclear posture, or all four explanations fail to explain this case. Israel’s decisions regarding its nuclear posture in 1991 thus seem difficult to explain. In hindsight, one could plausibly argue that Israel is so confident in its conventional superiority to deter all conventional attacks and so shifted to the minimum required posture to deter nuclear use against the Israeli state (hedging against the emergence of e.g., a nuclear Iran), an assured retaliation posture, even though it has had the capacity to shift to an asymmetric escalation posture and could do so with little marginal cost. Such an explanation is plausible, but given that there were no immediately emergent nuclear threats at the time, Israel’s decision not to shift further to an asymmetric escalation posture when it had both the time and capacity to do so is indeed difficult to explain with any of the theories advanced here. It is also possible that Israel still has largely a catalytic posture that is augmented with sea-­based capabilities. But for reasons noted earlier in the book, my coding scheme for nuclear postures classifies a state as having an assured retaliation posture once it acquires survivable forces, as now possessed by Israel with the integration of the Dolphin-­class submarines. In Israel’s case, optimization theory accurately captures the catalytic posture that it adopted during the Cold War, but fails to explain Israel’s apparent post–­Gulf War assured retaliation nuclear posture, although it does uniquely predict that Israel should shift postures at that point. All of the theories treated here fail to predict Israel’s assured retaliation posture, with all specifying that Israel would likely adopt an asymmetric escalation posture. This anomalous prediction could be due to a variety of factors that are not generalizable to other cases: Israel’s unique security predicament, its belief in the total dominance of its conventional forces obviating the need for an asymmetric escalation nuclear posture under any scenario, or simply no desire by the conventional forces to involve itself in nuclear missions. Nevertheless, it is important to note that Israel’s post-­1991 nuclear posture, as I code it, is an anomaly for optimization theory.

Chapter Eight

South Africa

South Africa’s experience with nuclear weapons remains unprecedented with regard to both nuclear proliferation and nonproliferation. In the 1970s, South Africa began a highly secret nuclear weapons program under Prime Minister John Vorster and subsequently Prime Minister P. W. Botha, taking advantage of its vast supplies of natural uranium. It ultimately built six nuclear devices without ever acknowledging or confirming its possession of nuclear weapons. A seventh nuclear weapon was under construction when President F. W. de Klerk halted production, shuttered South Africa’s uranium enrichment plant, and ultimately acceded to the NPT in 1991. In a remarkably candid speech to South Africa’s parliament in March 1993, de Klerk outlined South Africa’s nuclear weapons history, motivations, and its intended nuclear strategy. Though much about South Africa’s nuclear program remains classified, the rollback of the program generated the opportunity for South Africa’s leaders and those involved at the highest levels of the program to openly discuss, for the first time, the rise and fall of South Africa’s nuclear weapons program. This chapter traces South Africa’s nuclear posture—­how it intended to operationalize and use its six nuclear devices—­and explores the sources of that particular strategy. Since 1978, in a very explicit strategy statement outlined by Prime Minister Botha who held the reins of the South African program for its duration (though he inherited it from Vorster), South Africa clearly envisioned and operationalized a catalytic nuclear posture designed to draw in Western—­particularly American—­assistance in the event of an overwhelming Soviet or Cuban-­backed conventional threat to South Africa through Angola, Namibia, or Mozambique. Given the risk of additional 207

208

Chapter 8

sanctions and isolation if South Africa became an open nuclear power, optimization theory predicts that South Africa would adopt a catalytic posture if it believed it could successfully compel the United States to intervene on its behalf in the face of a severe threat.

South Africa’s Catalytic Posture 1979–­1 991 Prime Minister de Klerk’s 1993 remarkable nuclear rollback announcement resulted in significant transparency about South Africa’s nuclear history.1 South Africa’s vast stores of natural uranium made it a key supplier to the United States and United Kingdom in the 1940s and 1950s, and sixteen commercial uranium production plants were overseen by the Atomic Energy Board (AEB). In 1965, with U.S. assistance, South Africa built a research reactor, SAFARI, outside Pretoria at the Pelindaba site.2 For 10 years, the United States supplied the reactor with uranium reactor fuel until apart­ heid sanctions suspended nuclear cooperation. As a result of increasing international isolation in the late 1960s, South Africa began to explore in­ 1. See Hannes Steyn, Richardt van der Walt, and Jan van Loggerenberg, Nuclear Armament and Disarmament: South Africa’s Nuclear Experience (Lincoln: iUniverse, 2007); Liberman, “The Rise and Fall of the South African Bomb,” pp. 45–­86; David Albright, “South Africa and the Affordable Bomb,” Bulletin of the Atomic Scientists, vol. 50, no. 4 (July/August 1994), pp. 37–­47; Waldo Stumpf, “South Africa’s Nuclear Weapons Program: From Deterrence to Dismantlement,” Arms Control Today, vol. 25 (December 1995/January 1996), pp. 3–­8; Waldo Stumpf, “South Africa’s Nuclear Weapons Programme,” in Kathleen C. Bailey, ed., Weapons of Mass Destruction: Costs versus Benefits (Delhi: Manohar Books, 1994), pp. 63–­81; Mitchell Reiss, Bridled Ambition: Why Countries Constrain their Nuclear Capabilities (Baltimore: Johns Hopkins University Press, 1995), pp. 7– ­44; David Albright and Mark Hibbs, “South Africa: The ANC and the Atom Bomb,” Bulletin of the Atomic Scientists, vol. 49, no. 3 (April 1993), pp. 32–­37; David Albright and Tom Zamora, “South Africa Flirts with the NPT,” Bulletin of the Atomic Scientists, vol. 47, no. 1 (January/February 1991), pp. 27–­31; J. W. de Villiers, Roger Jardine, Mitchell Reiss, “Why South Africa Gave up the Bomb,” Foreign Affairs, vol. 72, no. 5 (November/December 1993), pp. 98–­109; Liberman, “Israel and the South African Bomb,” pp. 46–­80; Stephen F. Burgess, “South Africa’s Nuclear Weapons Policies,” Nonproliferation Review, vol. 13, no. 3 (November 2006), pp. 519–­26; Verne Harris, Sello Hatang, and Peter Liberman, “Unveiling South Africa’s Nuclear Past,” Journal of Southern African Studies, vol. 30, no. 3 (September 2004), pp. 457–­75; Terence McNamee, “The Afrikaner Bomb: Nuclear Proliferation and Rollback in South Africa,” in Cohen and McNamee, Why Do States Want Nuclear Weapons?, pp. 13–­23; Helen E. Purkitt and Stephen F. Burgess, South Africa’s Weapons of Mass Destruction (Bloomington: University of Indiana Press, 2005); Jeffrey T. Richelson, Spying on the Bomb: American Nuclear Intelligence from Nazi Germany to Iran and North Korea (New York: Norton, 2007), chapters 6 and 9. 2. Steyn, van der Walt, and van Loggerenberg, Nuclear Armament and Disarmament, chapter 3; Albright, “South Africa and the Affordable Bomb”; Liberman, “The Rise and Fall of the South African Bomb.”

south africa 209

digenous enrichment capabilities that could both fuel its nuclear power industry and provide the latent capacity to produce highly enriched uranium for nuclear weapons. In 1970, South Africa succeeded in developing an enrichment method that used a gas nozzle, an “aerodynamic technique similar to a stationary wall centrifuge,” according to David Albright, and enabled the creation of Y-­Plant at Valindaba, near Pelindaba, which served as South Africa’s primary enrichment facility.3 Y-­Plant began producing highly enriched uranium (HEU) in 1978, along with reactor-­grade LEU for South Africa’s nuclear power plants. Albright estimates that Y-­plant, after 1978, was capable of enriching enough uranium for about one gun-­type uranium nuclear device per year, though it turned out to be a bit slower than that.4 Concurrently, Prime Minister John Vorster personally approved ongoing work related to a nuclear weapons program in 1974, including research related to design and testing, most notably the development of a test site with two testing shafts at the Vastrap base in the Kalahari desert following the successful test of a model gun-­type device with only conventional explosives and no uranium core.5 All of this was technically approved under the auspices of a “peaceful nuclear explosives” program under the aegis of the AEB. By mid-­1977, South African scientists were prepared to conduct a “cold test” of its gun-­type designs (a test without the HEU but with a depleted uranium core) at the Kalahari test sites. This activity was detected by Soviet satellites and subsequently shared with the United States, which put tremendous pressure on the Vorster government to abort the test—­which both the United States and USSR feared was to be a full-­blown nuclear test—­which the Vorster government did.6 Nevertheless, South Africa’s design capabilities and uranium enrichment were fast approaching the point that made South Africa a nuclear-­capable state by the late 1970s.7 All of this activity was occurring in a security context that Waldo Stumpf, head of South Africa’s Atomic Energy Corporation (AEC) at the time of dismantlement and the author of South Africa’s authoritative history on its nuclear weapons program, describes as follows: “During the 1970s, especially the latter half of the decade, the political and military environment around South Africa deteriorated markedly.  .  .  . The large buildup of Cuban military forces in Angola, beginning in 1975, which eventually 3. Albright, “South Africa and the Affordable Bomb,” p. 40. 4. Ibid. 5. Purkitt and Burgess, South Africa’s Weapons of Mass Destruction, pp. 41–­42; Stumpf, “South Africa’s Nuclear Weapons Program,” p. 4. 6. Richelson, Spying on the Bomb, pp. 277–­82. 7. As noted in “New Information on South Africa’s Nuclear Program and South African-­ Israeli Nuclear and Military Cooperation,” Declassified Memo, Directorate of United States Intelligence, March 30, 1983, p. 1.

210

Chapter 8

peaked at 50,000 troops, reinforced a strong perception within the government that it would remain internationally isolated should the country’s territorial integrity be threatened.”8 Given this context, the pressure to abort the 1977 cold test triggered a decision by then-­President P. W. Botha to fully militarize South Africa’s nuclear program the following year. The militarization of South Africa’s nuclear program was ordered in April 1978 and, in July 1979, a committee that included J. W. de Villiers, executive director of the AEB, recommended the development of a nuclear arsenal that included only seven gun-­type weapons.9 Consistent with the decision to militarize South Africa’s nuclear program at that time, South Africa traded natural uranium for 30 grams of tritium from Israel to potentially boost its fission devices.10 Stumpf writes that the first gun-­type uranium device was the one intended for the 1977 test, and that the HEU core for that device was actually produced only in 1979.11 The second device, also intended as a test device, was smaller and its HEU core was only 80 percent enriched (“weapons-­grade” uranium is generally 90 percent enriched, but lower enriched uranium can still be used in nuclear weapons, although more material is then required to sustain a chain reaction). After these two test devices were built, the Botha government transferred production of the non-­fissile packages to the state-­owned ARMSCOR, while AEB would continue to produce the uranium cores. ARMSCOR built South Africa’s first truly aircraft-­deliverable nuclear weapon in 1982. The Y-­plant produced HEU for nuclear weapons at a rate of approximately one per eighteen months. For a variety of design and safety reasons, the next actual nuclear device, the fourth in South Africa’s arsenal, was not built until 1987. Two more were built by 1990, with an additional bomb under construction before de Klerk terminated Y-­Plant’s enrichment activities.12 These devices were designed to be deliverable by Buccaneer bombers, as either gravity bombs, or potentially as air-­to-­surface glide bombs. All told, South Africa developed six nuclear devices with a seventh under construction before production was halted. In terms of command and control, the assets were under the direct control of the head of state and, according to the de Klerk government, “never deployed militarily or integrated into the country’s military doctrine.”13 Although the original number 8. Stumpf, “South Africa’s Nuclear Weapons Program,” p. 4. Also see Prime Minister F. W. de Klerk, Address to Special Joint Session of Parliament reprinted in “De Klerk Discloses Nuclear Capability to Parliament,” FBIS-­AFR-­93–­056, March 25, 1993, p. 5. 9. See Reiss, Bridled Ambition, p. 9. Based on personal correspondence between Reiss and de Villiers. 10. Liberman, “The Rise and Fall of South Africa’s Bomb,” p. 52. 11. Stumpf, “South Africa’s Nuclear Weapons Program,” p. 5. 12. Albright, “South Africa and the Affordable Bomb,” p. 44. 13. See ibid., p. 38.

south africa 211

of seven gun-­type fission devices was determined in 1979, this number was “clearly reconfirmed” in September 1985.14 South Africa’s nuclear weapons were stored in de-­mated component form in highly secure vaults that required the authentication of the nuclear program’s four highest officials to open: the head of state, the Chief of Defense Force, the Chairman of ARMSCOR, and the Chairman of the AEB.15 Each nuclear device, while it was being constructed, was stored in two halves, and required the authentication of two of the aforementioned officials to gain access to one subcritical half stored in one vault, and the authentication of the other two to gain access to the other half stored in a second vault before a weapon could be worked on.16 No single person had the complete code. The procedures to assemble a completed weapon, which was stored in vaults in disassembled form, were similar. Steyn, van der Walt, and van Loggerenberg, all of whom worked on South Africa’s nuclear weapons, note that the Chief of South African Defence Force (SADF) could only physically order movement if he had codes from the president and the heads of ARMSCOR and the AEC to unlock each of the two vaults containing subcritical halves of the weapon.17 Stumpf confirms that “the assembly of the two separate pieces of hardware required a third interlocking code carried only by the head of the South African Defence Force. The detonation code was carried only by the Head of Government.”18 This procedure suggests the use of some sort of rudimentary PAL, though these procedures are admittedly unclear.19 However, according to Stumpf and others involved with South Africa’s nuclear weapons program, no nuclear weapon was ever fully assembled during South Africa’s nuclear period for envisioned use.20 How did South Africa intend to use its nuclear weapons against the perceived communist threat to it emanating from Angola and, potentially, Mozambique? Especially since it neither confirmed nor denied the possession of nuclear weapons—­though the United States and international community had strong suspicions after the aborted 1977 Kalahari test and then following the uncertainty surrounding the 1979 Vela incident in the South Pacific.21 How did South Africa operationalize its nuclear posture? Rather 14. Stumpf, “South Africa’s Nuclear Weapons Programme,” p. 71. 15. Steyn, van der Walt, and van Loggerenberg, Nuclear Armament and Disarmament, pp. 88–­89; Albright, “South Africa and the Affordable Bomb,” p. 44. 16. Steyn, van der Walt, and van Loggerenberg, Nuclear Armament and Disarmament, pp. 88–­89; Albright, “South Africa and the Affordable Bomb,” p. 44. 17. Steyn, van der Walt, and van Loggerenmberg, Nuclear Armament and Disarmament, pp. 88–­89. 18. Stumpf, “South Africa’s Nuclear Weapons Programme,” p. 70. 19. Ibid.; and Purkitt and Burgess, South Africa’s Weapons of Mass Destruction, p. 65. 20. Also see Reiss, Bridled Ambition, p. 11. 21. See Richelson, Spying on the Bomb, for a detailed description of the Vela incident and lingering suspicions that it was either a French test or an Israeli test with South African

212

Chapter 8

than attempt to directly deter conventional aggression against South Africa, Botha, in 1978, confirmed by Stumpf and de Klerk, outlined a very explicit catalytic strategy. Stumpf reveals: In April 1978, President Botha approved a three-­phase deterrent strategy. Phase 1, which was essentially already in effect, was characterized by strategic uncertainty, whereby South Africa’s nuclear capability would be neither acknowledged nor denied. Should South African territory be threatened (for example, by the Warsaw Pact through surrogate Cuban forces in Angola), the government would move to Phase 2, when it would consider the secret acknowledgement of the country’s capability to certain international powers such as the United States. Should this partial disclosure of South Africa’s capability not bring about international intervention to remove the threat to South Africa, the government would, in Phase 3, consider public acknowledgement of its nuclear capability or even a demonstration through an underground nuclear test. No offensive tactical application of nuclear weapons was ever foreseen or intended by the government, as it was fully recognized that such an act would bring about nuclear retaliation on a massive scale.22

This authoritative account of South Africa’s nuclear posture is remarkable in its detail, and made even more remarkable by de Klerk’s public enumeration of the three-­phased catalytic strategy. In his speech to the special joint session of Parliament in March 1993, de Klerk stated: “[I]f the situation in southern Africa were to deteriorate seriously, a confidential indication of the deterrent capability would be given to one or more of the major powers, for example the United States, in an attempt to persuade them to intervene.”23 André Buys, who chaired ARMSCOR’s strategy group, revealed that South Africa would disclose its nuclear capability by inviting “them to send inspectors to view our bombs. If this failed to convince them to intervene on our behalf we would threaten to detonate a nuclear device underground and therefore ‘go nuclear’.”24 The explicit strategy, revealed by South Africa’s highest level officials including its head of state, was to use nuclear weapons as a political gamble, not as military tools, to compel the United States in particular to intervene on South Africa’s behalf in the event of an overwhelming conventional threat to the state. assistance. Given the state of South Africa’s arsenal at the time however, it is improbable that South Africa had, in 1979, the ability to detonate an atmospheric nuclear device in the 2–­3kT range. 22. Stumpf, “South Africa’s Nuclear Weapons Program,” p. 5. 23. De Klerk, Address to Special Joint Session of Parliament, p. 6. Emphasis added. 24. Buys quoted in Liberman, “The Rise and Fall of the South African Bomb,” p. 56.

south africa 213

Indeed, South Africa may have come close to executing Phase Two of its catalytic nuclear posture in 1987, when it reopened the Vastrap test site in the Kalahari Desert. Buys reveals that this occurred because “for the first time the government started considering the possibility that we might lose the war militarily” and that the second phase became more than a remote possibility “once we were confronted by a serious and escalating military threat. We got close to that in 1987  .  .  . in Angola.”25 At the time P. W. Botha reportedly said: “Once we set this thing off, the Yanks will come running.”26 Mitchell Reiss, a senior U.S. diplomat, writes that this activity was designed “to send a political signal to Washington and Moscow as part of the diplomatic endgame in Southern Africa . . . South Africa’s leaders likely concluded that such activity would reinforce their bargaining position by quietly signaling to the United States and the Soviet Union the potential costs of failing to get the Cuban issue resolved.”27 A South African counterintelligence officer noted that: “Of course we knew satellites would see the whole thing . . . that was part of the plan . . . Soviet and Western intelligence were suddenly convinced we were serious about nuclear weapons and the West began to put pressure on the Soviets to get the Cubans to withdraw from Angola.”28 Though a test was not executed, U.S. intelligence detected the overly conspicuous activity at the Valstrap test site, and the following year the United States brokered the Angola-­Namibia accords to draw down the con­ flicts in Southern Africa. Peter Liberman, in his definitive work on South Africa’s nuclear weapons program, argues that “the covert coercion strategy of enticing Western assistance banked on U.S. aversion to the overt spread of nuclear weapons. Washington’s reaction to the discovery of the Kalahari test site apparently encouraged South African strategists to think that a catalytic strategy might succeed in drawing U.S. support in the event of a future crisis.”29 South Africa likely also bet on the United States’ aversion to the spread of communism in Southern Africa. Reiss writes that the strategy was “aimed to provoke Washington, which would be so disturbed at the potential use of nuclear weapons in the region, and so sensitive to the implications for its global nonproliferation agenda, that it would intervene on South Africa’s behalf before the country was overrun . . . the intention was to compel the United States to intervene politically on its behalf.”30

25. Ibid., p. 60. 26. Ibid. 27. Reiss, Bridled Ambition, p. 14. 28. Quoted in Hilton Hamann, Days of the Generals: The Untold Story of South Africa’s Apartheid-­Era Military Generals (Cape Town: Zebra Press, 2001), p. 168. 29. Liberman, “The Rise and Fall of the South African Bomb,” p. 61. 30. Reiss, Bridled Ambition, pp. 15–­16.

214

Chapter 8

The history of South Africa’s nuclear program is striking. The country only ever planned to build seven weapons which, according to de Klerk, “was considered the minimum for testing purposes and for the maintenance, thereafter, of a credible deterrent capability.”31 Moreover, it never had any intention of militarily employing these weapons, reserving them strictly for catalytic political purposes. The weapons were stored in highly secure vaults under firm civilian control, and evidently there were no military use options. In the other cases of catalytic nuclear postures—­Israel and Pakistan in their first phases—­one has to infer the choice of a catalytic posture from crisis behavior owing to the secrecy and ambiguity inherent in the posture. In South Africa, however, officials have revealed their strategy in considerable detail following the dismantling of the program, explicitly indicating that they selected a catalytic posture, and had no plans to develop more robust nuclear use or retaliatory options.

Explaining South Africa’s Cataly tic Posture Between 1979 and 1991, South Africa clearly envisioned and operationalized a catalytic nuclear posture designed to draw in Western assistance on its behalf. While the motivations for South Africa’s nuclear weapons program are, to this day, unclear, the sources of its catalytic posture are easier to isolate. In particular, the idea of using nuclear weapons to “blackmail” the United States to intervene on its behalf was the product of an optimization process that recognized that using nuclear weapons against a Soviet-­ backed threat, over which South Africa believed it had conventional superiority in Southern Africa, would be potentially suicidal. Thus, the political gamble of adopting a catalytic nuclear posture to draw in the United States to at least diplomatically intervene to thwart communist forces was the optimal choice for Pretoria. This is a determinate prediction of my theory that is not generated by the alternative explanations. Structural realism, technological determinism, and strategic culture are indeterminate with respect to the South African case. While several scholars, most notably Peter Liberman and Terence McNamee, have explored the motivations for South Africa acquiring nuclear weapons and then subsequently relinquishing them—­a remarkable story in and of itself—­no one has attempted to explain why South Africa adopted a catalytic nuclear posture during its nuclear weapons period. In this section, I trace the predictions of optimization theory, along with those of the alternative explanations: structural realism, technological determinism, and strategic culture. Only optimization theory of nuclear posture generates 31. De Klerk, Address to Special Joint Session of Parliament, p. 5.

south africa 215

a determinate prediction for South Africa’s nuclear posture; the other explanations predict a range of options and are therefore indeterminate with respect to South Africa’s choices after the decision to develop nuclear weapons. What would optimization theory predict for South Africa’s nuclear posture? The first node in the theory is whether South Africa had a third party that it believed it could compel to intervene on South Africa’s behalf in the event of a conventional conflict, thereby potentially enabling a catalytic nuclear posture. In South Africa’s case, there was a keen belief that, even though the United States was sanctioning the apartheid regime, the United States had a significant interest in preventing communist penetration toward Pretoria. Even the Soviet Union believed that this was the case, with the Soviet news agency TASS consistently arguing that both Israel and South Africa would not behave as aggressively as they did “without the support and encouragement of Washington.”32 The South Africans themselves believed that U.S. interests in preventing the spread of communism through Angola—­as evidenced by covert U.S. arms supplies to Jonas Savimbi’s UNITA rebel group from the mid-­1980s on—­might compel Wash­ ington to intervene in the event communist forces threatened South Africa. This belief, according to Stumpf, was predicated on the assumption that Washington was sufficiently incentivized to check the penetration of communism further south in Africa. Additionally, if South Africa threatened to unsheathe nuclear weapons, it would further compel the United States to intervene on its behalf due to U.S. interests in preventing the further spread of proliferation in the region. P. W. Botha noted that once they threatened nuclearizing a potentially conventional conflict, the “Yanks will come running.” In some sense, the manipulation of nuclear risk was believed to be necessary to blackmail the United States to intervene in Southern Africa, exploiting Washington’s twin interests in thwarting communist advances and its nonproliferation agenda despite South Africa’s apartheid policies. According to Helen Purkitt and Stephen Burgess, South African leaders, particularly Botha, “sensed that the Reagan administration would be more sympathetic [than the Carter administration] to South Africa’s struggle against a ‘total onslaught’ by the Soviet Union” or its allies.33 It is not clear whether American patronage in this contingency was a fanciful delusion or whether the United States actually would have intervened on South Africa’s behalf. On the one hand, the fact that Chester 32. TASS Statement quoted in Christopher Coker, South Africa’s Security Dilemmas (Westport: Praeger, 1987), p. 72. 33. Helen E. Purkitt and Stephen F. Burgess, “South Africa’s Nuclear Strategy,” in Toshi Yoshihara and James R. Holmes, eds. Strategy in the Second Nuclear Age: Power, Ambition, and the Ultimate Weapon (Washington, DC: Georgetown University Press, 2012), p. 44.

216

Chapter 8

Crocker helped negotiate the 1987 ceasefire in Angola and Namibia suggests that the United States did indeed have a significant interest in the region’s stability. In 1979, Crocker wrote that “it may also be unlikely that the Western powers would let matters [with Soviet intervention] get so far out of hand without making a determined effort to limit the conflict.”34 But whether the United States would have actually intervened if a serious conventional threat to South Africa materialized is impossible to determine. Mitchell Reiss concludes that it is “impossible to believe that any American administration would have rushed to extricate the white regime from imminent extinction,” perhaps only in the unlikely event of a direct Soviet invasion.35 Or perhaps if Pretoria threatened to nuclearize the conflict. In fact, it appears that South African leaders explicitly calculated that the probability of American assistance would increase if they had the ability to threaten nuclear escalation, and thereby blackmail the United States to intervene on their behalf. Without a catalytic posture, the probability of American intervention was effectively zero; with nuclear weapons and the ability to potentially blackmail Washington, the probability may have been only slightly higher, but it was certainly higher. Thus, given perceived Amer­ ican interests in preventing the further spread of communism and nuclear proliferation in the region, South African leaders believed they had a third-­ party patron that could be compelled to intervene on their behalf in the event that their “backs were against the wall.” As such, a catalytic nuclear posture was not only possible, but the most rational option available—­at least compared to the alternatives—­to South African leaders during the 1980s. Adopting an assured retaliation or asymmetric escalation posture might have generated security and organizational costs—­particularly possible reactive proliferation or Soviet punishment—­and would not have reaped any additional security benefit, since South Africa’s objectives could be achieved with a catalytic posture, and because South Africa had no ability to target core Soviet interests. This belief in the ability to compel the United States to intervene was based partly on modeling what South Africa believed to be Israel’s nuclear posture, and the latter’s successful execution of a catalytic strategy during the 1973 Yom Kippur War. André Buys, who chaired the ARMSCOR strategy group, states that members of the group “were aware of the alleged use by Israel of its nuclear capability to obtain U.S. assistance during the 1973 war. We had no proof that this was factual . . . The allegation probably subconsciously influenced our thinking. We argued that if we cannot 34. Chester A. Crocker, “Current and Projected Military Balances in Southern Africa,” in Richard E. Bissell and Chester A. Crocker, eds., South Africa into the 1980s (Boulder: Westview Press, 1979), p. 93. 35. Reiss, Bridled Ambition, pp. 28–­29.

south africa 217

use a nuclear weapon on the battlefield (as this would have been suicidal), then the only possible way to use it would be to leverage intervention from the Western Powers by threatening to use it. We thought that this might work and the alleged Israel-­USA case gave some support to our view.”36 Given the close military and political ties between the two countries, and the perception of similarity between South Africa’s international isolation and Israel’s, South African leaders seem to have made the parallel to the Israeli case, and similarly envisioned using nuclear weapons as a catalytic tool to compel western intervention on its behalf in the event of an existential threat to the state. But at the heart of the posture was the belief that South Africa could use—­and needed to use—­the threat of nuclear escalation to blackmail the United States to intervene to thwart communist penetration into South African territory. Unlike the third-­party relationship the United States had with Pakistan and Israel that enabled their selection of a catalytic posture, the South African case is more tenuous since, although Pretoria believed Washington would come to its aid, at least diplomatically as in the Pakistani case, it is less than clear whether it actually would have done so. Like Israel, South Africa’s selection of a catalytic posture might also have proved optimal because it allowed South Africa to keep its nuclear capabilities opaque, thereby avoiding a regional arms race or being targeted by the Soviet Union. From a security perspective, South Africa did not face any direct conventionally superior proximate offensive threats to its survival, though a Soviet threat certainly had the capacity to pose significant problems for the state. The military balances in Southern Africa and the fact that South Africa was operating in Namibia as a buffer against Angola (the second most powerful state in the region) suggests that short of Soviet introduction of forces, South Africa was the superior conventional military in the region. In 1978, when the decision to militarize South Africa’s nuclear program was taken, South Africa had a 2:1 advantage in active duty military manpower, and a 10:1 advantage accounting for South Africa’s capacity that could be quickly mobilized.37 In 1985, the Military Balance judged that: South Africa remains the only African country capable of significant force projection operations against her neighbours. With the possible exception of Angola, none of her neighbours, singly or in concert, is capable of formal conventional operations against her. Angola, with Soviet-­supplied equipment and with detachments of Cubans and other Soviet-­oriented nations, is stretched to 36. Buys quoted in Liberman, The Rise and Fall of the South African Bomb, p. 62. 37. International Institute for Strategic Studies, Military Balance (London: IISS, 1978), pp. 45–­54.

218

Chapter 8

her limits containing the forces of UNITA and could not also defend against a major South African offensive.38

Chester Crocker, the American official who would eventually coordinate the Tripartite Accord to end the war in Angola, judged in 1979 that: “By almost any conventional index of national military power, the Republic of South Africa continues to tower over any current or foreseeable African opponent or coalition. Its list of military assets and advantages is awesome in regional terms”;39 only with a decisive Soviet intervention could South Africa face a conventional challenge.40 This relative force balance obtained throughout the 1970s and 1980s, when South Africa’s decisions regarding nuclear weapons were being made. Robert Rotberg similarly concludes that: South Africa’s position as the dominant power in all of southern Africa increased dramatically in the 1980s. By mid-­decade, South Africa had no local or global rivals for preeminence in the region south of Zaire and Tanzania. . . . Mozambique and Swaziland had become client states in all but name . . . Angola’s independence had been severely limited, Namibia remained a domestic colony, and even Zimbabwe and Zambia became subject to the intruding tentacles of South African influence.41

Indeed, far from facing any significant objective military threat, South African defense and air forces were both quantitatively and qualitatively superior to their primary regional threats, even with the deployment of Soviet surface-­to-­air missile batteries in the region. Although South Africa faced a favorable military balance over its neighbors, its leaders from Vorster on, but particularly Botha, were concerned about the so-­called total onslaught from communist forces emanating from all sides—­Angola, Namibia, Mozambique, and particularly from within. The fear of a transnational communist revolution that threatened the South African Afrikaner government was termed the total onslaught to “describe the Soviet war-­mongers, antiapartheid activists, and anyone else willing to support the African National Congress or other guerrilla movements in

38. International Institute for Strategic Studies, Military Balance (London: IISS, 1985), p. 90. 39. Crocker, “Current and Projected Military Balances in Southern Africa,” p. 71. 40. Ibid., p. 93. 41. Robert I. Rotberg, “Introduction: South Africa in its Region—­Hegemony and Vulnerability,” in Robert I. Rotberg, Henry S. Bienen, Robert Legvold, and Gavin G. Maasdorp, eds., South Africa and Its Neighbors: Regional Security and Self-­Interest (Lexington: Lexington Books, 1985), p. 1.

south africa 219

antagonizing South Africa.”42 So while South Africa had military superiority against its regional adversaries, there was a fear that a combination of communist forces from without and within, backed by the Soviet Union, could pose an overwhelming threat to the regime. This fear provided the motivation for acquiring nuclear weapons. But it was the belief—­irrespective of how delusional—­that the United States would intervene on its behalf if South Africa threatened nuclear breakout that enabled the selection of a catalytic posture. The structural realist explanation, as noted above, is indeterminate with respect to South African nuclear posture. Almost any nuclear posture would have been consistent with South Africa’s security environment, in which South Africa had conventional superiority against its regional adversaries but feared a coalition of forces backed by the Soviet Union. If it was most concerned about an overwhelming conventional attack, an asymmetric escalation posture would have been the most plausible option. Even accounting for alliance patterns, there is nothing in the structural realist explanation that would allow one to discriminate between the various postures to generate a determinate prediction for South Africa in the 1980s. Indeed, the structural realist prediction for an internationally isolated state is pure self-­ help and the most aggressive nuclear posture that the state can sustain—­ which in South Africa’s case would be an asymmetric escalation posture. As such, the strict structural realist explanation does not adequately explain South Africa’s choice of nuclear posture during the late 1970s and 1980s. The second alternative explanation, technological determinism, fares slightly better in the South African case since South Africa produced as many nuclear weapons as it could based on its enrichment capacity. That is, South Africa could not adopt a more robust nuclear posture because it simply did not have the technical capacity to do so—­thus a catalytic posture was all that it could possibly adopt. However, there are two deficiencies in the technological determinism explanation for South Africa’s nuclear choices. The first is that it presupposes that South Africa did not have the capability to expand its enrichment capacity and that limiting its capacity to enrich weapons grade uranium was a technical and not a political choice. This is difficult to sustain, since South Africa had the option of enriching uranium at the Z-­Plant (also at Pelindaba) to weapons grade level but chose not to do so. South Africa did therefore have the technical base to generate more weapons-­grade uranium than it did but, for political reasons, decided not to do so. Thus, the existence of the Z-­Plant and the timing of South Africa’s decisions about nuclear posture—­particularly Botha’s outlining of South Africa’s nuclear posture in 1978 and the ex ante 42. Robert I. Rotberg, “Decision Making and the Military in South Africa,” in Rotberg et al., South Africa and Its Neighbors, p. 17.

220

Chapter 8

decision in 1979 to produce only seven nuclear weapons—­suggest that political decisions drove technical decisions, rather than vice versa. That is, Botha decided to develop a catalytic posture and then allocated the technical resources required to achieve that posture. The second deficiency with the technological determinism explanation is that while it might suggest that South Africa would have a small arsenal, it does not explain how South Africa would operationalize it. Indeed, South Africa could have employed its small arsenal to directly deter its regional adversaries, or it could have employed it in a catalytic fashion to compel a third party to intervene on its behalf to indirectly deter those adversaries. Either of the two choices would be consistent with a technological determinism explanation, and the theory offers no way to distinguish between them. As a result, for these two reasons, the technological determinism explanation does not offer a fully satisfactory account for why South Africa not only adopted a small nuclear arsenal but chose to operationalize it as a catalytic nuclear posture. The final alternative explanation, strategic culture, also fails to offer a satisfactory account for South Africa’s nuclear posture. Terence McNamee has most notably advanced a strategic culture explanation for South Africa’s nuclear program, in which almost all the individuals from the prime minister/president to the nuclear scientists were Afrikaner, calling it the “Afrikaner Bomb.”43 McNamee argues that “Afrikaner security culture” had four main themes that produced South Africa’s nuclear weapons: threat, isolation, survival, and martial spirit.44 The Afrikaner security culture, in brief, focused on the plight of a chosen people—­with “a God-­given task in Africa”—­persistently threatened and under siege “physically and culturally” by non-­Afrikaners and who thus required a fortified martial spirit in which the application of military force could ensure the survival of the Afrikaners.45 McNamee argues that nuclear weapons were thus viewed as the Afrikaner insurance policy: “Widely unexplored . . . are the critical ways in which the narrow interests of the group—­the Afrikaners—­determined the fate of their arsenal, rather than wider geopolitical dynamics or territorial-­ based security considerations.  .  .  . [G]enerally overlooked is the fundamental consistency in the way the nuclear arsenal was perceived—­vaguely, crudely, even illogically—­by its caretakers as a means to safeguard and perpetuate Afrikaner dominance.”46 In this view, South Africa’s nuclear weap­­ons

43. McNamee, “The Afrikaner Bomb: Nuclear Proliferation and Rollback in South Africa.” 44. Ibid., p. 19. 45. Ibid., pp. 20–­21. 46. Ibid., p. 22.

south africa 221

program and its nuclear choices were a product of this Afrikaner security culture which perceived an acute siege mentality. What would the Afrikaner security culture predict with respect to South Africa’s nuclear posture? Although McNamee may have identified a persuasive rationale for the South African nuclear program, the security culture argument also does not offer any guidance as to how South Africa might operationalize its nuclear assets. Indeed, if anything, this security culture explanation would predict that the Afrikaner bomb would be explicit, openly demonstrated, and oriented for first use if necessary to ensure the survival of the Afrikaner people. Nothing about the Afrikaner security culture explanation would lead one to believe that South Africa’s Afrikaner leaders would adopt a catalytic nuclear posture that did not envision the military use of nuclear weapons—­and indeed lacked significant delivery but instead operationalized them politically as a catalytic capabilities—­ posture to compel a third party to intervene to save the Afrikaners. As a result, the strategic culture explanation for South African nuclear posture also fails to offer a satisfactory account of why South Africa specifically chose to adopt a catalytic nuclear posture. The only theoretical explanation that offers a determinate—­though admittedly not entirely satisfactory—­explanation for South Africa’s nuclear posture during the late 1970s and 1980s is optimization theory. It is the only theory that not only leaves open the possibility of, but actually predicts, a catalytic posture for South Africa, though there is some uncertainty about the actual reliability of the United States as a potential intervener on South Africa’s behalf. Against the full universe of nine nuclear postures selected by the six regional nuclear powers, optimization theory, as a whole, offers the most compelling theory for why regional powers select the nuclear postures they do, and under what conditions they might switch strategies. As I summarize in the conclusion, optimization theory consistently proves superior to the alternative explanations, both in predicting which strategy a regional nuclear power might choose and explaining why it does so at a particular point in time. The following two chapters demonstrate that these choices have critical consequences for international conflict and security.

Chapter Nine

Deterring Unequally I: A Large-­n Analysis

The first part of the book established that the regional nuclear powers have adopted different nuclear postures, coding when and why each power has adopted either a catalytic, assured retaliation, or asymmetric escalation nuclear posture. But does this variation matter? That is, do these various nuclear postures have differential deterrent effects on the outbreak of conflict? Are some nuclear postures better able to deter conflict than others? These are the motivating questions for the second part of the book. Over the next two chapters, I explore whether the fact that regional powers have adopted different nuclear postures matters to international relations and conflict. Only if different postures actually affect international conflict dynamics does it matter why states adopt the nuclear postures they do. I find that regional power nuclear postures do vary substantially in their ability to deter conflict. Contrary to the conventional wisdom, only states that adopt an asymmetric escalation posture experience an identifiable and measurable reduction in conventional attacks against them. I establish this finding in two parts. This chapter tests the deterrent effect of regional power nuclear postures in a large-­n statistical design to systematically analyze whether, on average, some nuclear postures deter conflict better than others.1 The subsequent chapter explores the findings from the large-­n analysis in more fine-­ grained crisis settings to ensure that the hypothesized correlations found are reflected in state decisions about conflict in important crises. 1. This chapter is an extended version of Vipin Narang, “What Does It Take to Deter? Regional Power Nuclear Postures and International Conflict,” Journal of Conflict Resolution, vol. 57, no. 3 ( June 2013), pp. 478–­508.

222

Deterring Unequally I 223

During the Cold War, there was an intuition and some critical evidence that superpower nuclear postures affected conflict dynamics.2 The Cuban Missile Crisis is often portrayed as the quintessential episode in which American and Soviet nuclear postures were critical to both the outbreak of the dispute and the eventual capping of dispute escalation. Another, more complicated, superpower nuclear crisis, whose details are still emerging, especially on the Soviet side, is the nuclear alert during the 1973 Yom Kippur War when both the United States and Soviet Union squared off over the possibility that the Israel Defense Forces would encircle the Egyptian Third Army. Near midnight on October 24, 1973, after Soviet Premier Brezhnev threatened unilateral action to prevent the Israeli encirclement of the Third Army, American forces moved to Defense Condition (DEFCON) III, a de facto nuclear alert. Possibly as a result of this alert, the crisis quickly defused and the United States and Soviet Union moved quickly to avoid a superpower war over the Middle East.3 While much historical work has focused on the role of nuclear signaling by the superpowers to deter conflicts and conflict escalation, no work has explored the effects of regional power nuclear postures on the frequency and intensity of conflict at the level below the superpowers. Indeed, prior to the American DEFCON-­III alert, Israel’s nuclear arsenal failed to deter a massive combined Arab attack, forcing Tel Aviv to employ its catalytic posture to compel the United States to resupply it with conventional materiel. In addition, India’s assured retaliation posture failed to deter Pakistan’s aggression in Kargil, as well as its support for subconventional attacks such as the Mumbai attacks in 2008. On the other hand, Pakistan’s asymmetric escalation posture has powerfully inhibited India’s leaders from authorizing significant conventional retaliatory options. Are there systematic differences in these various nuclear postures’ ability to deter the outbreak and escalation of conflict? That is, are some nuclear postures better at deterring conflict, at various levels of intensity, than others? Those that assert the presence of existential deterrence assume that the possession of even a handful of nuclear weapons imparts a state with a “plausible nuclear retaliation” capability that ought to be sufficient to deter conventional conflicts due to even a small risk of escalation to the nuclear level. This logic should be even more binding at the regional level, where conventional breakdowns can occur more quickly than in the superpower balance. But is this empirically true? The next two chapters interrogate this 2. See for example the excellent volume by Richard K. Betts, Nuclear Blackmail and the Nuclear Balance (Washington, DC: Brookings Institution Press, 1987). 3. See, for example, Scott D. Sagan, “Nuclear Alerts and Crisis Management,” International Security, vol. 9, no. 4 (Spring 1985); Barry M. Blechman and Douglas M. Hart, “The Political Utility of Nuclear Weapons: The 1973 Middle East Crisis,” International Security, vol. 7, no. 1 (Summer 1982).

224

Chapter 9

belief with quantitative and qualitative analysis and find that it is not. I find that nuclear weapons deter unequally, as a function of a state’s nuclear posture. States with different nuclear postures reap different deterrent power, because these postures pose a credible threat to adversaries at different points in a potential conflict. In particular, the asymmetric escalation posture is the only one to exert a powerful deterrent effect on the initiation and escalation of armed conflict. Even a large nuclear force, if arrayed in an assured retaliation posture, fails to systematically deter conventional conflict. Possessing nuclear weapons does not by itself generate the ability to deter conventional conflicts. States that wish to deter conventional conflicts with nuclear weapons must explicitly orient their nuclear forces to do so by adopting an asymmetric escalation nuclear posture. While this may seem an intuitive point, it overturns a central tenet of the prevailing literature on nuclear deterrence. This chapter presents the large-­n analysis, examining whether the adoption of a particular nuclear posture on average reduces the outbreak and escalation of crises (general deterrence effect). The following chapter selects a series of crises nested within the large-­n analysis to explore the immediate deterrence effect of postures on decision-­making dynamics regarding escalation and de-­escalation within crises (which is necessarily conditional on a general deterrence failure and thus not amenable to a large-­n analysis because of the resulting selection effects). The two chapters, together, demonstrate the unique and powerful deterrent effect of the asymmetric escalation posture against conventional conflict.

The Existential Bias It is an article of faith among nuclear deterrence theorists that acquiring nuclear weapons fundamentally and positively changes a state’s quest for security. But what kind of nuclear forces deter conflict? The existing literature evinces a pervasive existential bias, arguing that the mere possession of a small nuclear arsenal ought to ensure a state’s ability to deter its adversaries from initiating conflict. Kenneth Waltz famously argued that nothing more than the “credibility of small deterrent forces” is required to establish Thomas Schelling’s “threat that leaves something to chance”—­explicitly contending not just that nuclear possession deters nuclear use, but that it ought to deter conventional attacks also.4 The existing quantitative work 4. See, for example, Waltz and Sagan, The Spread of Nuclear Weapons, chapter 1, p. 23; Schelling, Arms and Influence; John Lewis Gaddis, The Long Peace: Inquiries into the History of the Cold War (New York: Oxford University Press, 1987); Robert Jervis, “The

Deterring Unequally I 225

on nuclear deterrence and conflict very explicitly expresses this bias by treating all nuclear states as equivalent once they acquire a single nuclear weapon. This assumes that a state with one warhead reaps the same deterrent effect as states with mature second-­strike or first-­use capabilities. Not only is this theoretically suspect, is it empirically true? Present scholarship is inconclusive. This is largely due to the focus on the superpower experiences of the United States and the Soviet Union. I argue that the twin problems of the Cold War hangover and the existential bias in theoretical and empirical studies have created the present cloudiness in our understanding of nuclear deterrence. Indeed, the dichotomous focus on whether a state has nuclear weapons or not is a serious conceptual misspecification. This part of the book attempts to advance our theoretical and empirical understanding of deterrence in two ways. First, it focuses on the critical experiences of the regional nuclear powers, which comprise the lion’s share of existing (and all emerging) nuclear powers and provide the most fertile ground for testing theories of deterrence. Second, it shifts the unit of analysis from the mere possession of a nuclear weapon to nuclear posture: the forces, organizational procedures, and doctrines states adopt to operationalize their nuclear capabilities. The next two chapters ask the critical question: are some nuclear postures better than others in deterring the outbreak of conflict? Using this new independent variable, nuclear posture, I test—­both statistically and qualitatively in sets of crises—­whether and how variation in nuclear posture af­ fects a state’s ability to deter conflict at various levels of intensity. This is one of the first attempts to disaggregate nuclear weapons states by their nuclear postures. I find that nuclear weapons deter unequally, as a function of a state’s nuclear posture. In particular, the asymmetric escalation posture is the only one to exert a powerful deterrent effect on the initiation and escalation of armed conflict; it does so against both nuclear and non-­nuclear opponents. An assured retaliation posture has surprisingly little ability to deter even high-­intensity conventional conflict. And the catalytic posture has experienced serious deterrence failures, including several full-­blown wars. Contrary to conventional wisdom, there is little evidence that the mere possession of nuclear weapons—­or even secure second-­strike forces—­ systematically or significantly deters conventional conflict. If a state wants to deter conventional conflict, it must explicitly array its nuclear posture to do so. Political Effects of Nuclear Weapons: A Comment,” International Security, vol. 13, no. 2 (Autumn 1988), pp. 80–­90. For a classic work on specific doctrinal/postural effects, see John J. Mearsheimer, “Nuclear Weapons and Deterrence in Europe,” International Security, vol. 9, no. 3 (Winter 1984/1985), pp. 19–­46.

226

Chapter 9

This is the first attempt, both theoretically and empirically, to treat states with different nuclear strategies as distinct. The implications of these findings for our understanding of what it takes to deter, and for policy debates about proliferation and counterproliferation, are profound. To reap a significant deterrent effect against conventional attacks from their adversaries, states have to do more than simply acquire nuclear weapons. Indeed, even developing secure second-­strike forces is insufficient. For scholars and policy makers alike, this suggests that the key independent variable of interest for stability among nuclear powers should be shifted from considering their acquisition of simply nuclear weapons to focusing on their nuclear postures.

Moving beyond the Existential Bias A dogma in security studies is that the critical threshold in a state’s quest for security is the acquisition of nuclear weapons. In addition to deterring nuclear attack, the mere risk of nuclear use should also deter adversaries from initiating conventional attacks for fear of escalation to the nuclear level.5 Although the Cold War witnessed the development of massive nuclear architectures, many influential deterrence and proliferation theorists ultimately concluded that the basic existence of a nuclear weapons capability ought to provide sufficient deterrence to conventional conflict.6 In fact, although Cold War practitioners assumed and behaved as if nuclear postures mattered, some deterrence theorists and almost all proliferation theorists now seem to believe that the sky will literally fall if a state acquires even a single nuclear weapon.7 The logic behind existential deterrence8 is that the destructive power of a single nuclear weapon is so great that conventional conflict ought to be inhibited even due to a small risk of nuclear escalation. Thomas Schelling suggests that even a plausible threat of nuclear use with a small arsenal 5. See footnote 4. 6. To be sure, there were practitioners and theorists such as Paul Nitze who argued that more maximal postures were required, but most scholarly theory ended up focusing on minimal nuclear forces or basic second-­strike forces. Nuclear weapons may deter nuclear use, but since nuclear and non-­nuclear states have experienced no nuclear use since 1945, this is a difficult proposition to test. I focus here on the role of nuclear weapons in deterring conventional conflict. 7. See, for example, Matthew H. Kroenig, “Time to Attack Iran,” Foreign Affairs, January/February 2012. 8. This term, first coined by McGeorge Bundy, posits according to Marc Trachtenberg that the “mere existence of nuclear forces,” even ambiguous or non-­weaponized, should induce caution in adversaries and deter aggression. See Trachtenberg, “Influence of Nuclear Weapons in the Cuban Missile Crisis,” no. I p. 139.

Deterring Unequally I 227

could inhibit limited wars through the progressive “generation of risk.”9 McGeorge Bundy wrote that the powerful deterrent effect of nuclear weapons “rests on the uncertainty about what could happen . . . deterrent power is unaffected by most changes in the arsenals on both sides.”10 Theorists including Waltz, Jervis, and Mearsheimer argue that this threshold is achieved with just a few nuclear weapons because adversaries can never be certain that a strike will fully disarm or eliminate an adversary’s capacity for nuclear retaliation.11 Waltz writes that once a state “has a small number of deliverable warheads of uncertain location . . .” and can thus plausibly retaliate with nuclear weapons, it should be capable of deterring armed conflict.12 Mearsheimer similarly concurs with this reasoning, concluding that “there is no question . . . the presence of nuclear weapons makes states more cautious about using military force of any kind against each other.”13 In the established logic, the mere acquisition of nuclear weapons is the crucial leap to achieving security, not only against nuclear but also conventional attack. This logic should be even more binding among regional powers. The potential speed of conventional military breakdowns in regional conflict scenarios increases the risk of a rapid escalatory spiral to nuclear use by an imperiled state, and should thus make any opponent facing a regional nuclear power extremely cautious in initiating disputes. Waltz famously wrote that: “Nuclear weapons lessen the intensity as well as the frequency of war among their possessors. For fear of escalation, nuclear states do not want to fight long or hard over important interests—­indeed, they do not want to fight at all. Minor nuclear states have even better reasons than major ones to accommodate one another peacefully and to avoid any fighting.”14 Bernard Brodie similarly argued that a “small menaced nation” required only a single nuclear weapon which “it could certainly deliver on Moscow if attacked”; this ought to be “sufficient to give the Soviet government pause” and achieve protection against invasion.15 Particularly since the end of the Cold War, the efficacy of existential deterrence has been essentially accepted and unchallenged by scholars and practitioners. 9. Thomas Schelling, The Strategy of Conflict (Cambridge: Harvard University Press, 1960), pp. 187–­94. 10. McGeorge Bundy, “Existential Deterrence and its Consequences,” in Douglas Mac­ Lean, ed., The Security Gamble: Deterrence Dilemmas in the Nuclear Age (Totowa: Rowman and Allanheld, 1984), p. 9. Emphasis added. 11. Waltz in Waltz and Sagan, “The Spread of Nuclear Weapons,” pp. 141–­42; Robert Jervis, The Meaning of the Nuclear Revolution (Ithaca: Cornell University Press, 1989), p. 35; Waltz, More May Be Better. 12. Waltz in Waltz and Sagan, “The Spread of Nuclear Weapons,” p. 142. 13. Mearsheimer, Tragedy of Great Power Politics, p. 129. 14. Waltz, The Spread of Nuclear Weapons: More May Be Better, p. 25. Emphasis added. 15. Bernard Brodie, Strategy in the Missile Age (Princeton: Princeton University Press, 1959), p. 275.

228

Chapter 9

As a result, most post–­Cold War theory has focused on how and why states initially acquire nuclear weapons.16 Very little has been written on the deterrence effects of choices states make after they acquire nuclear weapons. This lacuna is reinforced by another mostly unarticulated, untested assumption: that nuclear postures are simply optimized for a state’s security environment—­as was sometimes claimed about the United States and USSR—­and are therefore epiphenomenal, exerting no independent effect on a state’s ability to deter conflict.17 If states optimize their nuclear force postures to their security environments, there should be no resulting difference in deterrence power across states. This hypothesized uniform deterrent effect further justifies treating nuclear weapons states as equivalent and reinforces the existential bias. Existing empirical work, both qualitative and quantitative, expresses the existential bias by assuming that the size, structure, and orientation of nuclear arsenals are all irrelevant. For example, recent qualitative work on crisis-­prone South Asia treats India and Pakistan as nuclear equivalents, even though they operationalize their nuclear forces in very different ways.18 As both states were developing nuclear weapons, then-­Indian Chief of Army Staff K. Sundarji wrote: “If a mutual minimum nuclear deterrent is in place, it will act as a stabilizing factor. Even the chances of conventional war between the two will be less than before . . . Why all this fuss about India and Pakistan?”19 However, the fact that they have achieved such different levels of deterrence success since nuclearization suggests that the assumption of nuclear equivalence may be untenable.20 Qualitative studies of deterrence are additionally plagued by the difficulty of measuring deterrence success, 16. See Sagan, “Why Do States Build Nuclear Weapons?” pp. 54–­86; Eric Gartzke and Dong-­Joon Jo, “Determinants of Nuclear Proliferation,” Journal of Conflict Resolution, vol. 51, no. 1 (2007), pp. 167–­94; Sonali Singh and Christopher Way, “The Correlates of Nuclear Proliferation: A Quantitative Test,” Journal of Conflict Resolution, vol. 48, no. 6 (2004), pp. 859–­85. 17. Glaser, Analyzing Strategic Nuclear Policy, 1990; and Keir Lieber and Daryl Press, “How Much Is Enough? Nuclear Deterrence Then and Now,” Working Paper presented at Wisconsin International Relations Colloquium, University of Wisconsin–­Madison, February 3, 2009. 18. See, for example, Sumit Ganguly, “Nuclear Stability in South Asia,” International Security, vol. 33, no. 2 (Fall 2008), pp. 45–­70; S. Paul Kapur, “Ten Years of Instability in South Asia,” International Security, vol. 33, no. 2 (Fall 2008), pp. 71–­94; E. Sridharan, ed., The India-­Pakistan Nuclear Relationship (Delhi: Routledge, 2007); Russell J. Leng, Bargaining and Learning in Recurring Crises: The Soviet-­American, Egyptian-­Israeli, and Indo-­Pakistani Rivalries (Ann Arbor: University of Michigan Press, 2000); and Kapur, Dangerous Deterrent. 19. K. Sundarji, “India’s Nuclear Weapons Policy,” in Jorn Gjelstad and Olav Njolstad, eds., Nuclear Rivalry and International Order (London: Sage Publications, 1996), p. 179. 20. See Vipin Narang, “Posturing for Peace?: Pakistan’s Nuclear Postures and South Asian Stability,” International Security, vol. 34, no. 3, pp. 38–­78.

Deterring Unequally I 229

or the “crises that don’t bark,” since these successes are largely unobservable. Thus, qualitative studies do tease out mechanisms and evidence for why and when deterrence might fail, but have not been designed to offer a systematic understanding of what kinds of nuclear forces it takes to successfully deter conflict. In order to address this problem, the quantitative dispute literature has analyzed large-­n datasets to estimate the deterrence effects of nuclear weapons by measuring levels of conflict a state experiences pre-­and post-­ nuclearization. But this literature has several problems. First, it explicitly suffers from the existential bias by simply using a dummy variable for whether a state has nuclear weapons or not in a particular year, treating Cold War USSR the same as post-­1967 Israel.21 This assumes that one nuclear weapon has the same deterrent power as ten thousand, irrespective of how they are deployed.22 The aggregation of nuclear weapons states into a single category misses potential variation that different types of nuclear states have experienced in deterring conflict. Second, quantitative literature based on dyad-­year datasets overweights the superpower experience, since they possessed nuclear weapons for the longest period, were involved in the most politically relevant conflict dyads, and had many crises with each other and by proxy. This has thus generated inconclusive results about the role of nuclear weapons in deterring conflict. For example, Bennett and Stam’s analysis of nuclear weapons and the probability of conflict is indeterminate: “While variables are statistically significant, estimated effects 21. See, for example, Paul Huth, “Extended Deterrence and the Outbreak of War,” American Political Science Review, vol. 82, no. 2 (June 1988), 423–­44; Paul Huth and Bruce Russett, “What Makes Deterrence Work? Cases from 1900–­1980,” World Politics, vol. 36, 496–­526; Paul Huth and Bruce Russett, “Deterrence Failure and Crisis Escalation,” International Studies Quarterly, 32:1 (1988) 29–­45; D. Scott Bennett and Alan Stam, The Behavioral Origins of War (Ann Arbor: University of Michigan Press, 2004); Daniel Geller, “Nuclear Weapons, Deterrence, and Crisis Escalation,” Journal of Conflict Resolution, vol. 34, no. 2 (June 1990), 291–­310; Bruce Bueno De Mesquita and William Riker, “An Assessment of the Merits of Selective Nuclear Proliferation,” Journal of Conflict Resolution, vol. 26, no 2 (June 1982), 283–­306; and Curtis Signorino and Ahmer Tarar, “A Unified Theory and Test of Extended Immediate Deterrence,” American Journal of Political Science, vol. 50, no. 3 (July 2006), 586–­605; Kyle Beardsely and Victor Asal, “Proliferation and International Crisis Behavior,” Journal of Peace Research, vol. 44, no. 2 (2007), pp. 139–­55; Kyle Beardsley and Victor Asal, “Winning with the Bomb,” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009), pp. 278–­301; Michael Horowitz, “The Spread of Nuclear Weapons and International Conflict: Does Experience Matter,” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009), pp. 234–­57; Robert Rauchhaus, “Evaluating the Nuclear Peace Hypothesis: A Quantitative Approach,” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009), pp. 258–­77. 22. Horowitz, “The Spread of Nuclear Weapons,” accounts for the age of a nuclear power, which is an important contribution, but still treats nuclear weapons powers as equivalent based on the mere possession of a nuclear weapons capability.

230

Chapter 9

differ in direction across subsets and outcomes.” As a result they are “[u]nable to estimate key effect on war probability.”23 The effect of nuclear weapons on deterring conflict remains fundamentally unclear. Fortunately, two simple correctives can advance our understanding of how nuclear weapons affect international conflict. First, as argued earlier, regional nuclear powers should be analyzed as a separate class of states. For a variety of reasons outlined in earlier chapters—­financial, technical, and through learning from others’ experience—­regional nuclear powers have cho­sen fundamentally different approaches to operationalizing their nu­ clear arsenals than the superpowers.24 These states developed nuclear arse­ nals of similar size and faced similar structural constraints and opportunities operating below the superpower plane. Although substantial variation exists in aggregate power metrics across regional nuclear powers such as China and South Africa, their similarity on key relevant dimensions—geo­ strategic situation, size of nuclear arsenal, and regional conflict landscapes— ­means that we can reasonably view them with a common analytical lens. Whereas the superpowers had tens of thousands of nuclear weapons that could “do it all” with respect to deterrence, regional powers face tighter constraints, they do not have to worry about the problems associated with extended deterrence, and are forced to make critical decisions about how to allocate their much scarcer nuclear forces and toward what end. In addition, their widely different choices about nuclear posture, and their seemingly varying deterrence successes, make them fertile and critical test­ ing ground for the effects of nuclear weapons on conflict. That is, the sim­ ilar power metrics and arsenal sizes but varying deterrence outcomes among regional nuclear states allows for theorizing and isolating the effects of nuclear weapons in deterring armed conflict. The second corrective is that the unit of analysis should not be a nuclear weapons capability, but nuclear posture or strategy. It is nuclear posture, rather than numbers of warheads or declaratory doctrine, that should theoretically be expected to generate deterrent power against an opponent. Numerical estimates are fraught with uncertainty, especially in real time. Moreover, absent changes in nuclear posture or strategy, additional nuclear weapons do not alter the probability and credibility of use, and therefore should not affect an adversary’s calculations about initiating conflict (particularly since states largely moved away from counterforce nuclear strategies in the middle of the Cold War). Furthermore, because posture focuses on observable capabilities, organizational procedures and interests, and 23. Bennett and Stam, Behavioral Origins of War, p. 112. 24. The term “regional power” is also intended to appeal to previous work that argued for the unique structural situation of so-­called middle or semi-­major powers. See Peter Katzenstein, Small States in World Markets (Ithaca: Cornell University Press, 1985).

Deterring Unequally I 231

patterns of behavior that are measurable both to adversaries and analysts, posture is a more consistent indicator than declaratory doctrine (though the two may often be consistent). States care more about what an adversary does with its nuclear weapons than what it says about them. As such, I hypothesize that it is differences in nuclear posture that generate variation in states’ ability to deter different types and intensities of conflict. The United States’ and Soviet Union’s various nuclear postures had differential deterrent effects during the Cold War.25 Similarly, regional nuclear powers have adopted identifiable nuclear postures across a spectrum of capabilities, management procedures, and transparency, with each having different deterrent effects. By unpacking regional nuclear powers, I move beyond the existential bias to isolate what types of nuclear forces are required to deter conflict. In doing so, I hope to resolve the literature’s present ambiguity by specifying which types of nuclear states, if any, are best able to deter conflict at various levels of intensity. This provides an answer to the critical question: what types of nuclear forces are required to deter conflict? I now unpack the hypothesized deterrence effects of the regional power nuclear postures.

Hypotheses: Effects of Posture on Deterrence This chapter takes nuclear posture as an independent variable, testing whether these postures have differential effects on states’ ability to deter the outbreak of conventional conflict. The fortunate fact that nuclear weapons have not been used since 1945 means that there is no way to test whether these postures deter nuclear use differently, so I leave that question aside. Instead, I ask: compared to when the state did not have nuclear weapons, and compared to other states, what effect should we expect these different postures to have on a state’s ability to deter the eruption of conventional conflict (general deterrence) against both nuclear and non-­nuclear opponents?26 In theory, different nuclear postures and strategies should result in different likelihoods of conflict initiation and escalation because they generate different spaces for conflict. Extrapolating from the literature on costly signaling and the credibility of deterrent threats, different nuclear postures create different thresholds at which the threat to use nuclear weapons 25. See Freedman, Evolution of Nuclear Strategy; Betts, Nuclear Blackmail and Nuclear Balance; Glaser, Analyzing Strategic Nuclear Policy; Lieber and Press, “How Much Is Enough?” 26. See James Fearon, “Selection Effects and Deterrence,” International Interactions, vol. 28, no. 1 (January 2002), pp. 5–­29.

232

Chapter 9

becomes credible, thereby establishing the point at which a given intensity of conflict should be deterred. 27 Differences in this “enforced” conflict space should therefore have a measureable effect on these postures’ ability to deter the frequency and intensity of conflict. That is, these nuclear postures should have different effects on adversaries’ calculations about initiating conflict and how far they are willing to press a conventional conflict before nuclear weapons become relevant to their decisions. Building on Schelling’s exposition about how to make deterrent threats credible, James Fearon identifies the role of costly signals in crisis or conflict diplomacy that seek to convince an adversary that one is capable of and willing to execute a particular threat at a particular threshold. One way to do this according to Fearon is to sink costs into actions, for example military mobilization, that are financially and organizationally costly ex ante in order to signal resolve.28 Fearon writes that “sunk-­cost signals are actions that are costly for the state to take in the first place but do not affect the relative value of fighting versus acquiescing in a challenge.” The case of nuclear posture is not a pure case of this type of signaling, but a state undertakes enormous costs ex ante to operationalize its nuclear forces in a particular way to make deterrent threats credible at dif­ ferent thresholds, and those choices do not necessarily affect the relative value of fighting versus acquiescing in any specific challenge. All states with nuclear forces have already sunk substantial costs into developing their first nuclear weapon. After that point, however, the marginal financial and organizational costs of each posture vary significantly—­and each posture creates different types of credible threats that “leave something to chance.” It is from this variation that each posture becomes credible at a different threshold, or red-­line. Developing tactical nuclear weapons and the command-­and-­control structures to release them early in a conventional conflict is a very costly signal that is observably different from the signal sent by a state that invests in survivable forces that are recessed but capable of reaching an adversary’s population centers; these are both distinct from ambiguous forces that serve as a vague signal to a third party or adversary. In this way, states incur different sunk capital and organizational costs in operationalizing their nuclear forces ex ante. Consequently, their deterrent threats become credible at different thresholds. The principal source of variation in the costly signaling of nuclear postures is thus generated by differences in the types of forces, command-­and-­ control procedures, and deployment patterns of a state’s nuclear forces. 27. Ibid.; Schelling, Arms and Influence, especially chapter 2. 28. In addition to the above, see also James D. Fearon, “Signaling Foreign Policy Interest: Tying Hands versus Sinking Costs,” Journal of Conflict Resolution, vol. 41, no. 1 (February 1997), pp. 68–­90.

Deterring Unequally I 233

How a state invests in its nuclear posture is at least partially, if not fully, observable to its adversaries. Because these postures are distinct and require different types of forces and architectures, they ought to generate different spaces for conflict. Thus, the literature on costly signaling and the credibility of threats provides theoretical reasons to believe that different nuclear postures should result in differential abilities to deter conflict. Because catalytic postures render the threat of nuclear use very low in most realistic conventional scenarios, and because third-­party intervention is uncertain, there exists a large space for conventional conflict and escalation before the threat to use nuclear weapons becomes relevant to adver­ saries. States with catalytic postures are therefore unlikely to reap a deterrent benefit, even against high-­intensity conventional conflicts. Likewise, because assured retaliation forces are primarily oriented for strategic retaliation rather than tactical use, this posture may also be incapable of deterring limited conventional attacks. The asymmetric escalation posture, however, compresses the space for conflict, since even limited conventional breaches early in a conflict can credibly trigger nuclear use. Asymmetric escalation postures should thus have a significant effect in deterring armed conflict, more so than the other two postures, because the investments necessary to make this posture credible advance the point in the conflict at which the threat to use nuclear weapons becomes salient. The hypothesized effect ought to vary by whether the adversary has nuclear weapons or not as well, since whether the adversary in question has the ability to retaliate with nuclear weapons might affect the ability of a regional power to credibly enforce a particular conflict space.29 Indeed, since the effects might theoretically be expected to differ against nuclear and non-­nuclear opponents, the following section derives separate hypotheses for the effects of nuclear posture on conflict outcomes against both types of adversary. The empirical tests that follow also derive separate estimates of deterrence success against nuclear and non-­nuclear opponents.

Hypotheses by Posture Catalytic postures are designed primarily to draw international intervention when a state’s existence is imperiled by conventional attack. Because third-­ party intervention is indirect and probabilistic, the catalytic posture may not have a strong deterrent effect on conventional conflict. The 29. Robert Powell notes the distinction between the “spectrum of risk” versus the “spectrum of violence,” or “irrational” versus “rational” escalation when an adversary has survivable nuclear forces. See Robert Powell, “The Theoretical Foundations of Strategic Nuclear Deterrence,” Political Science Quarterly, vol. 100, no. 1 (Spring 1985), pp. 75–­96.

234

Chapter 9

ambiguous manner in which nuclear forces are managed makes nuclear use noncredible in conventional contingencies short of total war. Adversaries may therefore calculate that they can achieve limited conventional war aims before the catalytic posture’s nuclear weapons are operationalized and before third-­party intervention is triggered. Adversaries may not even know of the existence of the catalytic state’s nuclear weapons program at the onset of hostilities. Since its primary utility is to compel third-­party intervention rather than to directly deter conflict against particular adversaries, there is no reason to expect the catalytic posture to successfully deter even significant conventional conflict. Non-­nuclear states are unlikely to be deterred from initiating substantial armed conflict because they do not fear credible nuclear retaliation, and nuclear opponents with their own retaliatory capability are even less likely to be deterred. Hcatalytic: A catalytic posture should have little effect in deterring low-­or high-­ intensity conflict initiated by either nuclear or non-­nuclear opponents.

According to the prevailing wisdom, assured retaliation postures should deter any kind of conventional attack because conventional conflict risks nuclear escalation. I argue, however, that the effect of an assured retaliation posture on conflicts of various intensities differs depending on whether the adversary is nuclear or non-­nuclear. Two nuclear-­armed adversaries with retaliatory capabilities may experience a stability-­instability paradox, in which both parties believe that stable mutual nuclearization caps escalation and frees them to engage in more frequent low-­intensity conventional conflict. Robert Jervis, following Glenn Snyder, argues that, “to the extent that the military balance is stable at the level of all-­out nuclear war, it will become less stable at lower levels of violence.”30 The lack of total war but increase in militarized crises between India and Pakistan since 1998 is often cited as evidence of this hypothesis. When it comes to attacks from nuclear opponents, then, I hypothesize that an assured retaliator will experience an increased frequency of low-­intensity attacks, but a reduced frequency of high-­intensity attacks.31

30. See, for example, Robert Jervis, The Illogic of American Nuclear Strategy (Ithaca: Cornell University Press, 1984), p. 31. 31. Depending on one’s reading of the stability-­instability paradox, the level of escalation that can occur can nevertheless include even tactical nuclear use, since mutual stable nuclearization ought to cap escalation to the full strategic level. See Glenn Snyder, “The Balance of Power and the Balance of Terror,” in Paul Seabury, ed., The Balance of Power (San Francisco: Chandler Books, 1965), p. 199. Also see Mark S. Bell and Nicholas L. Miller, “Questioning the Effect of Nuclear Weapons on Conflict,” Journal of Conflict Resolution, forthcoming.

Deterring Unequally I 235

Even against non-­nuclear opponents, however, there may still be a wide space for conventional conflict. Assured retaliators aim to deter conflict at the nuclear level, and their resulting orientation of capabilities and procedures—­particularly a reliance on recessed capabilities under centralized control—­renders the use of nuclear weapons noncredible in limited, low-­ intensity conflicts and contingencies. Because nuclear retaliation is unlikely, non-­nuclear opponents will likely feel free to continue the same frequency of low-­intensity conventional conflict against assured retaliators as before. Non-­nuclear opponents should, however, fear the risk of nuclear escalation if they engage in high-­intensity conventional attacks, so adopting an assured retaliation posture should make this type of attack less frequent. HAssuredRet: States with an assured retaliation posture should face a reduced frequency of high-­intensity attacks from both nuclear and non-­nuclear opponents. They should face an increased frequency of low-­intensity conventional attacks initiated by nuclear opponents, and an unchanged frequency of low-­intensity conventional attacks from non-­nuclear opponents.

Note that these theoretical predictions suggest that with an assured retalia­ tion posture, nuclear use only becomes credible at a very high level of conventional conflict, where a state’s existence is fatally threatened. In that case, even full-­blown conventional wars against assured retaliators may not be deterred by nuclear weapons, and the assured retaliation posture would actually have no measureable deterrent effect on conventional conflict. If true, this would be a substantial blow to the conventional wisdom, which holds that assured retaliatory nuclear capabilities pose a sufficient threat of escalation to dampen conventional wars. Asymmetric escalation postures should have the most significant effect on a state’s ability to deter conventional conflict, due to two complementary mechanisms. First, the asymmetric escalator threatens to use nuclear weapons against significant conventional breaches, taking costly measures to make the threat of nuclear use at a lower threshold credible and salient. Second, the devolution of nuclear assets and authority to the military required to make this posture credible creates a “mad man” deterrent: the fear that rogue military officers could take matters into their own hands and release nuclear weapons in the midst of a low-­level conventional conflict. Because the asymmetric escalator manipulates the risk of escalation to the nuclear level very early, these two mechanisms should reinforce each other, generating a substantial deterrent to even limited conventional conflict. I thus hypothesize that after adoption of this posture, there should be a substantial decrease in the frequency of armed conventional attacks across all intensities. Marked reduction in conflict should occur against both nuclear opponents (whose retaliatory strike should be deterred by the

236

Chapter 9

asymmetric escalator’s ability to respond in kind) and non-­nuclear opponents (who have no defense against nuclear use). In short, by credibly threatening rapid escalation to the nuclear level, states with asymmetric escalation postures should experience a decrease in the frequency of armed conflict across all intensities and against all types of opponents. Because states that adopt this posture are those likely to be already facing a high threat of substantial conventional conflict, this is a hard test for the hypothesis. Any reduction in conflict we observe would provide extremely strong evidence for the potent deterrent benefits of an asymmetric escalation posture. HAsymmetricEscalation: States with an asymmetric escalation nuclear posture should face fewer attacks from both nuclear and non-­nuclear opponents across all measurable intensities of armed conflict.

These hypotheses compete with the overall null hypothesis suggested by the extant deterrence literature: that there should be no variation in the deterrent effect of regional power nuclear postures. This hypothesis might obtain because even small nuclear forces serve as a basic existential deterrent against adversaries irrespective of nuclear posture, or because postures are already strictly optimized for a state’s security environment and should therefore deter uniformly well. In contrast, my hypotheses predict that there should be a differential impact of regional nuclear posture.

Research Design: Data and Methods I now turn to systematically testing whether a regional state’s adoption of a particular nuclear posture dampens the average frequency of conventional attacks across various levels of intensity. This is a critical test of what types of nuclear forces are required to deter various levels of conventional conflict. The most appropriate methodological avenue of investigation is a large-­n analysis that examines all states in the system across a large swath of time to systematically isolate the average effects of adopting a specific nuclear posture on the frequency and intensity of conflict. This method estimates the reduction in conflict initiation and escalation against a state both after it has adopted a particular nuclear posture (by including observations prior to posture adoption), and as compared to other states with similar covariates but with a different (or no) nuclear posture. By including non-­dispute observations, and thus measuring the “crises that don’t bark,” it has the advantage of circumventing one set of sample bias that has plagued deterrence studies. Examining the broad universe of dyadic-­ interaction observations minimizes selection bias and enables the estima-

Deterring Unequally I 237

tion of a relatively unbiased effect of adopting a particular nuclear posture on deterrence success. In order to test the effects of these various nuclear postures on general deterrence, I begin by following Bennett and Stam, which developed a directed-­dyad dataset consisting of more than 1 million dyads, coded as Initiator (State A) vs. Target (State B), in the international system between 1816 and 1992.32 That is, the dataset is structured such that, in every year, there is an observation for whether each state initiated a militarized dispute against every other state in the system and, if so, how far the conflict escalated. I use a more relevant and manageable politically relevant subset of this dataset, substantively similar to the full sample but containing fewer irrelevant observations (e.g., Pakistan-­Uruguay). Politically relevant dyads are those that include interactions with major powers and geographically contiguous states (and those separated by some set distance of water). I first analyze this dataset, which contains 116,057 observations from 1816 to 1992, using the full complement of controls available in Bennett and Stam. However, because the 1992 cutoff truncates almost 20 percent of the nuclear era, I also use the EUGene software package to construct a second directed-­dyad dataset that includes dyads between 1816 and 2001 and contains key conflict variables and available controls (though fewer controls than those available for the 1816–­1992 time period) in order to test the effect of nuclear posture over a more extended time period. I primarily choose to follow Bennett and Stam because its algorithm generates a dependent variable that is appropriate for measuring conflict deterrence. While the Militarized Interstate Dispute (MID) dataset has variables for dispute initiation as well as level of dispute escalation, it has no variable that combines both initiation and escalation. Bennett and Stam, however, calculates a dependent variable that identifies both which side initiated a dispute and how high that dispute ultimately escalated.33 I employ this dependent variable because it directly measures the frequency and intensity of attacks a state faces over time. This ordinal dependent variable ranges from 0 to 4. Table 9.1 summarizes the variable and the empirical frequency of each level of conflict. The structure of this dependent variable therefore provides a direct measure for the question of interest: once a state adopts a particular nuclear posture, do the frequency and intensity of disputes initiated against it decline and, if so, by how much?34

32. The original Bennet and Stam dataset consisting of data structure and controls is available at: http://www.personal.psu.edu/dsb10/data/ BOWReplication.zip. 33. Bennett and Stam, Behavioral Origins of War, p. 63. 34. Ibid., p. 65.

238 DV Level

Chapter 9 Description

Empirical Frequency

0

State A does not initiate a dispute against State B in year t

0.988

1

State A initiates a dispute against State B in year t but it does not escalate to the use of force

0.0033

2

State A initiates a dispute against State B in year t and dispute escalates to use of force by one state

0.0044

3

State A initiates a dispute against State B in year t and both states use force but it does not escalate to war

0.0032

4

State A initiates a dispute against State B in year t and the dispute escalates to full scale war

0.00064

Table 9.1. Description and empirical frequencies for large-n dependent variable (DV)

My independent variable of interest is nuclear posture. Because I am interested in the deterrent effect of various regional nuclear postures, I disaggregate the nuclear status of the target states in the directed dyad (State B35) into their different nuclear postures. In practice, this means creating a set of binary variables for target states with nuclear weapons that includes catalytic, assured retaliation, asymmetric escalation, and superpower posture. These variables are empirically coded as in table 2.1, with the United States (and the UK) and USSR taking the value of superpower posture in the period after acquiring nuclear weapons. This isolates the deterrent effects of the regional power nuclear postures, but avoids selection bias by retaining all of the superpower observations. Table 9.2 depicts descriptive statistics for the frequency of attacks each nuclear posture has experienced at various levels of intensity. Table 9.2 yields two striking observations. First, disputes are—­thank­fully—­relatively rare. Second, regional power nuclear postures have not fared equally well in preventing the outbreak of higher levels of conflict. In particular, compared to the other postures, asymmetric escalators experience far fewer disputes and far fewer that escalate to high levels of intensity. Furthermore, the catalytic posture has experienced a disproportionate number of conflicts that have escalated to full-­blown wars, suggesting some serious deterrence failures. We cannot infer anything definitive from these descriptive statistics, however, without examining the broader universe of observations to determine the change in conflict probabilities after adopting a particular posture and compared to non-­nuclear states.

35. I also ran analyses that disaggregated State A in the analysis, to see if these postures had any effect on emboldenment, that is, a state’s willingness to initiate disputes. The results were not systematic and not robustly significant.

Deterring Unequally I 239 Frequency Count for Each Target Nuclear Posture 1945–­2001

Level of Escalation

Catalytic

Assured Retaliation

Asymmetric Escalation

Superpowers ( plus UK ) 21667

7141

9152

6744

Dispute, No Force

3

12

4

36

One Side Uses Force

6

14

2

111

Reciprocal Use of Force

11

22

2

28

War

6

1*

0

4

No Dispute

Table 9.2. Summary statistics for nuclear posture and frequency of attack *This does not count the Kargil War, which was originally coded as “reciprocal use of force” even though the fatality count has since been revised up such that it should be classified as a war.

I thus proceed with a formal statistical estimation. The most appropriate statistical model given the structure of the dependent variable is a multinomial logit model which estimates the probability that the outcome (DV) will occur for each discrete value of the dependent variable given the set of independent variables. For example, it estimates the probabilities that the United States in 1991 might initiate, say against Iraq, no dispute (DV=0), a dispute that does not involve the use of force (DV=1), and so on through war (DV=4).36 This modeling choice, however, requires a different assessment of the statistical and substantive impacts of the independent variables of interest. Multinomial logit estimations produce multiple coefficients per independent variable, estimating the effect that an asymmetric escalation posture, for example, has on each level of the dependent variable. Since a variable can exert an overall statistically significant effect on the dependent variable even though it may only be statistically significant at some levels of the dependent variable and not others, the variable’s significance must be assessed by block likelihood ratio tests, to determine whether each posture has a statistically significant and systematic effect on the frequency and in­ tensity of conflict outbreak. To assess the substantive results, I present the relative risk ratio for each posture at each level of conflict intensity.37 This is a clean measure that interprets how many times more or less likely, on average, conflict is at each 36. The probabilities must sum to 1.0 across the five levels of the dependent variable. 37. This risk ratio is calculated using the actual population frequencies of conflict, not the odds ratio. It is calculated by taking the model’s predicted probabilities and simulating on average if all target states in the dataset shifted from not having nuclear weapons to adopting a particular nuclear posture against a specific type of opponent.

240

Chapter 9

level of intensity after adopting a given nuclear posture (compared to having no nuclear weapons). Since the probability of conflict between any two states in any two years is so low, this measure provides a straightforward way to interpret how nuclear posture affects those already small probabilities, while also normalizing for the level of preexisting conflict each state faces. Relative risk ratios should be interpreted as follows: a value of 1.00 means that adopting a particular posture has no effect on a state’s ability to deter conflict at that level; the probability of conflict at a particular level of intensity before and after posture adoption is statistically equivalent. A risk ratio of, for example 3.00, however, indicates that the probability of conflict at that level after adopting a particular posture is three times higher than before. And, finally, a risk ratio of, for example 0.33, indicates that the probability of conflict at that level after adopting a particular posture is three times lower than before. Thus, deterrence success is indicated by relative risk ratios less than 1.00.

Results: General Deterrence Tests I first present results that analyze the politically relevant set of directed dyads in the Bennett and Stam dataset from 1816–­1992, using the full set of available control variables at the monadic, dyadic, and system levels—­ such as regime types, transitions, conventional balance of power, revisionist intentions, and so on.38 Inclusion of these various controls, almost all of which have been shown to have a statistically significant effect on the outbreak of conflict and on escalation, is designed to isolate the independent effects of nuclear posture on dispute outbreak and escalation.39 In 38. The controls included in the Bennett and Stam (Behavioral Origins of War) dataset include, at the monadic level: measures for GDP, whether the initiator has nuclear weapons, regime type, democratization, variance of democratization, and polity change all from Polity IV. The dyadic level controls include dyadic defense pact (one dummy variable for defense pact between dyad members, one dummy variable if one or both states were members of NATO), arms race following Diehl (1983), balance of conventional capabilities (the ratio of State A’s CINC score from COW to the total CINC capability within the dyad), and a distance-­discounted measure of power following Bueno de Mesquita. Several measures of dyadic power transition are included: a five-­year moving average of the differential growth rates from COW, dyadic satisfaction with the status quo based on the Taub as well as S-­scores which measure alliance portfolio and foreign policy similarity within a dyad, and dyadic trade as a proportion of GDP for both initiator and target. The system-­level controls include measures for hegemonic stability and concentration of capability following Singer, Bremer, and Stuckey (1972). 39. I made the following coding corrections in the Bennett and Stam (Behavioral Origins of War) dataset: I code the 1967 and 1973 Arab-­Israeli wars as full wars, since they are coded as such in the Correlates of War Dataset.

Deterring Unequally I 241

Catalytic Posture

Assured Retaliation Posture

Asymmetric Escalation Posture

Non-­ Nuclear Initiator

Nuclear Initiator

Non-­ Nuclear Initiator

Nuclear Initiator

Non-­ Nuclear Initiator

Nuclear Initiator

No Dispute

1.00

1.00

1.00

1.00

1.00

1.00

Dispute, No Force

1.31

3.91

0.69

2.02

0.23

0.69

Unilateral Force

0.78

1.02

0.35

0.44

0.11

0.15

0.87

1.19

0.70

0.93

0.14

0.20

17.71

0.00

0.00

0.00

0.00

0.00

Conflict Level

Reciprocated Force War

Table 9.3. Multinomial logit results for 1816 –­1992 (relative risk ratios) Catalytic is significant at the p < 0.001 level, with the confidence intervals wider at lower levels of conflict but highly significant at the level of war. Assured Retaliation is not significant at the p < 0.10 level. Asymmetric Escalation is significant at the p < 0.001 level and the variable is significant for every level of the dependent variable.

other specifications, I use fewer controls to avoid overfitting or erroneous estimates, but the substantive results remain unchanged. To account for the time-­series structure of the dataset, following Beck, Katz, and Tucker, I employ cubic spline corrections based on peace years to correct for temporal autocorrelation. I employ dyad-­clustered robust standard errors. To show the substantive effects of these postures on each level of conflict, I present the relative risk ratios for each posture against both nuclear and non-­nuclear initiators. These represent the average relative risk of conflict after a target state has adopted a given nuclear posture. The results of this analysis are shown in table 9.3. The block significance tests for the catalytic and asymmetric escalation postures is p < 0.001, which suggests their inclusion in the model improves its predictive power by a statistically significant level against all types of opponents.40 The assured retaliation posture is not significant in this specification (or many alternative specifications) at the p < 0.10 level, suggesting that this posture has little systematic deterrent effect against either nuclear or non-­nuclear opponents.41 The statistical significance of the asymmetric escalation posture is the most robust to model specification, whereas assured retaliation is the least robust. The effect of the catalytic posture is

40. These are pooled block likelihood ratio tests. 41. The results for superpower posture were mixed and not robustly significant.

242

Chapter 9

Catalytic Posture

Conflict Level

Non-­ Nuclear Initiator

Assured Retaliation Posture

Asymmetric Escalation Posture

Nuclear Initiator

Non-­ Nuclear Initiator

Nuclear Initiator

Non-­ Nuclear Initiator

Nuclear Initiator

No Dispute

1.00

1.00

1.00

1.00

1.00

1.00

Dispute, No Force

0.56

1.17

0.83

1.38

0.62

1.08

Unilateral Force

0.35

1.31

0.43

1.29

0.18

0.60

0.85

1.18

0.92

1.01

0.28

0.33

15.15

0.00

1.53

0.00*

0.00

0.00

Reciprocated Force War

Table 9.4. Multinomial logit results for 1816–­2001 (relative risk ratios) *This again classifies the Kargil conflict as shy of full war; if coded as a war, this is greater than zero but the assured retaliation variable is not significant and still has “no effect.”

relatively robust as well, particularly for its surprising breakdown in high-­ intensity conflict. I next present the results from the expanded dataset of politically relevant dyads from 1816 to 2001 using a truncated set of theoretically important control variables.42 This dataset covers an additional decade of nuclear deterrence interactions, but by necessity includes fewer control variables. Results from the EUGene-­constructed dataset that comport with the dataset through 1992 would offer a robustness check on the results obtained using the original data. The coding rules for political relevance, for the dependent variable, and for nuclear postures are the same as above, extended through 2001. Temporal autocorrelation is again corrected by inclusion of peace-­years cubic splines; dyad-­clustered robust standard errors are again employed. All else is calculated as above. The relative risk ratios for the various postures on the extended dataset are shown in table 9.4. The block significance tests for the model on this dataset again reveal that the catalytic and asymmetric escalation postures are significant at the p < 0.001 level, while the assured retaliation posture is not significant at the p < 0.10 level. In addition to the lack of statistical significance of assured retaliation, the relative risk ratios at higher levels of armed conflict are close to 1.00, further suggesting that this posture does not systematically

42. The controls included in this dataset are the conventional balance of power, whether the initiator state is a nuclear weapons state, a cold war dummy, revisionist intentions, and the Polity IV regime type measures.

Deterring Unequally I 243

deter armed attacks. The overall results are substantively similar to those in table 9.3, and confirm the important trend it identified: the unique deterrent power of the asymmetric escalation posture, the persistent irrelevance of the assured retaliation postures in deterring even intense conventional conflict, and the riskiness and deterrence suboptimality of the catalytic posture.43 Indeed, on average, states adopting a catalytic posture have a significantly higher risk of being attacked by a non-­nuclear state in a conflict that escalates to war. Not only is the space for conflict against catalytic states wide, but adoption of this posture seems to make high-­intensity conflict even more likely. On the other hand, states that adopt an asymmetric escalation posture see a marked reduction in the risk of armed conflict at the most important levels of intensity—­three to four times lower probability of conflict where one or more parties uses force. In combination, tables 9.3 and 9.4 suggest that the asymmetric escalation posture is deterrence optimal. Against non-­nuclear initiators, the asymmetric escalation posture is unique in deterring conflict at every intensity level, whereas states adopting the catalytic and assured retaliation postures have suffered significant deterrence failures, including full-­scale wars. Against nuclear opponents, all three postures experience roughly the same level of probing as before by other nuclear powers at low-­level disputes (at DV = 1), but asymmetric escalation is unique in keeping these probes from escalating beyond a war of words. This finding comports with the hypotheses for asymmetric escalation against both non-­nuclear and nuclear opponents. The catalytic and assured retaliation postures do not fare nearly as well, with neither able to provide a systematic deterrent to even full-­blown conventional wars.44 Thus, the most important finding is that adopting an asymmetric escalation posture uniquely and significantly deters conflict at every level of armed intensity. Not only is there little evidence for an existential deterrent effect of nuclear weapons but, surprisingly, even the assured retaliation posture generates little systematic deterrent power against conventional conflict. Moreover, the catalytic posture is significant, but for the wrong reasons: it has had some spectacular deterrence failures at the level of full war, meaning that this nuclear posture is not, on average, deterring even high-­ intensity conventional conflict. Taken together, these findings suggest that the extant deterrence literature is mistaken about the ability of 43. The Bennett and Stam (Behavioral Origins of War) dataset does not code Vietnam’s invasion of China as a war. The EUGene-­generated dataset does, which accounts for the discrepancy in the estimates for non-­nuclear initiators against assured retaliation powers at DV=4. In both analyses, however, assured retaliation is not significant. 44. Based on these results at least, there is little evidence for a stability-­instability paradox for assured retaliation.

244

Chapter 9

states with small or even assured nuclear retaliatory capabilities to deter high-­intensity conventional conflicts. The former breaks down disturbingly regularly, while the latter seems to have no measurable effect in deterring opponents from attacking with armed force. Just how powerful is the asymmetric escalation nuclear posture? That is, if a state has no nuclear weapons, how much deterrence benefit would it reap if it adopted an asymmetric escalation posture? Similarly, if a state has a catalytic or an assured retaliation posture, how much deterrence benefit would it reap if it shifted to asymmetric escalation? The answer is: quite a bit. In order to illustrate this, I present first-­difference plots: the change in probability for each level of the dependent variable if the average state shifted to an asymmetric escalation posture (all control variables set at the mean, or median for variables with discrete levels). These first differences were calculated using Clarify and illustrate the absolute change in probability for each level of the dependent variable from the baselines in table 9.1.45 Figure 9.1 depicts the first differences against nuclear and non-­ nuclear opponents, with 95 percent confidence intervals, if a state were to shift to an asymmetric escalation posture (a) from not having nuclear weapons; (b) from a catalytic posture; and (c) from an assured retaliation posture.46 If the confidence interval does not cross zero, then the first-­ difference is significant at the p=0.05 level. Figure 9.1(a) shows that for each level of the dependent variable, a state is better off having an asymmetric escalation posture than it is not having nuclear weapons, against both nuclear and non-­nuclear opponents. Though the absolute change in probability at the war intensity is low, since the baseline rate is 0.00064, the first difference suggests a 3 to 4-­fold reduction in the probability of war if a state shifts to asymmetric escalation. Perhaps more importantly, 9.1(b) and 9.1(c) suggest that shifting from catalytic and assured retaliation to asymmetric escalation also yields deterrence benefits. A state with a catalytic posture would reap a measurable deterrence benefit, particularly at high-­intensity conflict levels, if it shifted to an asymmetric escalation posture. An assured retaliator would also experience a substantial and statistically significant deterrence benefit at high-­intensity conventional conflict by shifting to an asymmetric escalation posture, though the benefit is less pronounced at lower levels of conflict. Because the confidence intervals at the high intensities of conflict do not cross zero, this represents a measurable and statistically significant reduction in conflict against both nuclear and non-­nuclear opponents. This 45. Michael Tomz, Jason Wittenberg, and Gary King, “CLARIFY: Software for Interpreting and Presenting Statistical Results,” Journal of Statistical Software 8, no. 1 (2003). 46. All first differences were calculated using Clarify in Stata, based on the model in table 9.4.

Deterring Unequally I 245

0.002

0.003

0.004

Asymmetric Escalation vs. Nuclear Opponent

0.001 –0.002 –0.001 0.000

Change in probability of escalation

Asymmetric Escalation vs. Non-Nuclear Opponent

0

1

2

3

4

0

2

3

4

0.004 0.002 0.000 –0.002 0

1

2

3

4

0

1

2

3

4

Assured Retaliation Shifts to Asymmetric Escalation vs. Nuclear Opponent

0

0

0.002

0.003

0.004

Assured Retaliation Shifts to Asymmetric Escalation vs. Non-Nuclear Opponent

0.001 –0.002 –0.001 0.000

Change in probability of escalation

1

Catalytic Shifts to Asymmetric Escalation vs. Nuclear Opponent

–0.004

Change in probability of escalation

Catalytic Shifts to Asymmetric Escalation vs. Non-Nuclear Opponent

1

2

3

Level of escalation

4

1

2

3

Level of escalation

4

Figure 9.1. The deterrent power of asymmetric escalation nuclear posture, as calculated by the first differences for each level of the dependent variable for a shift (a) from no nuclear weapons to an asymmetric escalation posture; (b) from a catalytic posture to an asymmetric escalation posture; and (c) from an assured retaliation posture to an asymmetric escalation posture against both non-­nuclear and nuclear opponents.

246

Chapter 9

demonstrates that the asymmetric escalation posture is deterrence optimal compared to the other two postures. These results demonstrate that, on average, the asymmetric escalation posture is the most powerful nuclear posture available to states to deter conventional conflict. States without nuclear weapons, with a catalytic posture, and with an assured retaliation posture would all experience measurable reductions in high-­intensity armed conflicts if they adopted asymmetric escalation postures. This suggests that the deterrence benefits of nuclear weapons are not evenly distributed. Indeed, orienting nuclear forces toward an asymmetric escalation posture seems to be the only way for a state to generate a substantial deterrent against high-­intensity conventional conflicts with nuclear weapons. I performed a battery of robustness checks, none of which had any major impact on the substantive conclusions presented earlier. Although point estimates change based on specification, the fundamental findings remain largely robust. For example, controlling for the size of a state’s nuclear arsenal, or its log transformation, does not affect the findings.47 To ensure that no single state’s posture was driving the results, such as France’s effect on asymmetric escalation or Israel’s effect on catalytic, I tested country-­ specific fixed effects. Though the point estimate for the catalytic posture’s failure at the “war” level is dampened when including an Israel dummy, there is in total little substantive difference in the results. Similarly, even with country fixed effects, the asymmetric escalation posture still has a significant and marked substantive effect in reducing conflict across all levels of armed intensity. I also performed an array of sensitivity analyses. For example, tests restricted to the nuclear era (1945–­2001) were substantively equivalent to the results provided in table 9.4. In addition, I tested whether the results were sensitive to disagreements about when a regional power obtained nuclear weapons or adopted a particular posture (i.e., 1974 vs. 1989 in the case of India or 1967 vs. 1970 for Israel).48 Though some point estimates 47. I coded, numnukes, the estimated number of nuclear weapons each target state had in each year, based on best-­estimate data from the Federation of American Scientists Nuclear Weapons Databooks and Nuclear Notebooks. The results were actually slightly stronger when controlling for number of nuclear weapons, perhaps because the size of the superpower arsenals simply swamps those of the regional powers and magnifies the importance of posture; the inclusion of a log transformation for numnukes also does not affect the findings. The uncertainties surrounding the estimates for the number of nuclear weapons in each state’s arsenal provides theoretical justification for treating nuclear posture rather than numbers as the unit of analysis. That is, a shift from a catalytic to an asymmetric escalation posture theoretically should have more impact on deterrence than a shift from 5,000 to 5,001 nuclear weapons. 48. A similar sensitivity test is proposed by Montgomery and Sagan, “The Perils of Predicting Proliferation,” p. 308.

Deterring Unequally I 247

change, the substantive findings remain. Other sensitivity analyses also revealed few substantive differences.49 The results in this chapter are strongly suggestive. They are, however, by no means the definitive or final word on the differential effects of regional power nuclear postures on the outbreak and escalation of conflict. In particular, there are several methodological and data issues worth highlighting. First, these results do not provide valid causal inferences, nor are they designed to. There is no denying that states non-­randomly acquire both nuclear weapons and specific nuclear postures, so there is unavoidable “selection into treatment.” Nevertheless, except in the extreme cases of an existential security threat, the causes of nuclearization do not strongly correlate with a state’s ultimate choice of nuclear posture. States that have developed nuclear weapons for largely security reasons—­Israel, China, and Pakistan for example—­have adopted all three different nuclear postures. This research design therefore takes nuclear posture as a quasi sui generis variable in an attempt to isolate the correlation with conflict outcomes. It is not ideal, but given the empirical realities, I believe it is the soundest available approach. Second, it is widely established that the decision for a state to initiate and escalate a dispute against a nuclear state is strategic, in that the knowledge of that state’s nuclear weapons factors into its initial decision to start the conflict.50 If this is true, it is unlikely that nuclear weapons will deter conflict once a dispute breaks out—­a selection effect that may dampen the ability to study the true deterrent effect of nuclear weapons and that is particularly acute in intra-­crisis tests that select on the outbreak of a dispute. Three features of my research design mitigate this potential issue. First, the structure of the chosen dependent variable measures both whether a state initiated a dispute at all (i.e., including nondispute observations) and how high that dispute eventually escalated. This dependent variable is therefore a good measure of whether the adoption of a particular posture dampens dispute initiation as well as dispute escalation. Second, because the dataset

49. For example, I tested specifications that included the UK as an asymmetric escalation regional posture rather than dummying it out as superpower posture. This did not significantly affect the substantive results, though the attack the UK experienced in the Falklands War does attenuate the results. I also tested the United States, USSR, and UK in the early phases of their nuclear programs as asymmetric escalation, when they might have been in the same class as some of the regional powers. Even after trying various cutoff years, there was little change in the substantive results for the various postures. But, the substantive effect of asymmetric escalation is stronger when the superpowers are dummied out as superpower posture because of the number of disputes with each other. 50. Alastair Smith, “To Intervene or Not Intervene: A Biased Decision,” Journal of Conflict Research, 40, no. 1 (1996), pp. 16–­  40.

248

Chapter 9

includes both cross-­dyadic observations of similar characteristics where nuclear weapons are not present, and observations within dyads over time, substantial controls effectively isolate the effect of a nuclear posture on a given dispute’s dynamics. Third, the findings mitigate the reverse causality concern that frequency of conflict drives a state’s nuclear posture since, if this were true, we would expect states with the asymmetric escalation posture to have the greatest amount of conflict and states with the catalytic posture to have the least. I find exactly the opposite, suggesting an independent role of nuclear posture in driving conflict outcomes. If anything, these results are probably underestimating the true effect of the asymmetric escalation posture. Finally, just as the quantitative dispute literature’s tests on the effects of nuclear weapons are driven by the United States and USSR because of their overrepresentation in dyad-­year datasets, the paucity of states for each of the nuclear postures means that a small number of states drives the effects of each posture.51,52 However, including individual country fixed effects does not largely affect the significance or substantive effect of the postures because there is enough temporal and cross-­dyadic variation to establish their effects. The dataset that spans through 2001 contains more observations per posture and captures the Israeli and Pakistani changes in posture. Nonetheless, there is no avoiding the fact that this statistical analysis is critically dependent on a handful of countries. However, I believe that I have performed as many of the technically necessary adjustments and tests as possible given the state of the empirical data and quantitative techniques available to analyze them.

Importance of These Findings The most important finding in the preceding analysis is the degree to which the deterrence dividend is unevenly distributed across regional nuclear powers. Not all nuclear postures deter equally well. Contrary to the expectations of deterrence scholars, not to mention policy makers, nuclear weapons do not ipso facto deter conventional conflict. For better or worse, my analysis suggests that the answer to the fundamental question “what does

51. Though these postures are rare, they represent substantial n in the dataset; rare events techniques are not appropriate here because the dependent variable, MIDs, is not rare. 52. It would be interesting to test the dynamics when these various nuclear postures of the target states interact with the various nuclear postures of the initiator state (i.e., asymmetric escalation versus assured retaliation). As one separates the initiator out by posture, however, the strain on the statistical models due to the small n of each possible combina­ tion makes it difficult to draw systematic inferences.

Deterring Unequally I 249

it take to deter conventional conflict?” appears to be: “a nuclear posture explicitly oriented to do so.” There is thus both theoretical and empirical value in disaggregating nuclear powers by their postures. The most striking result is the significant reduction in conflict frequency and escalation experienced by states that adopt the asymmetric escalation posture. These states enjoy a unique reduction in armed conflict at every level of intensity, compared to both non-­nuclear states and states that adopt other nuclear postures. In contrast, states with assured retaliation postures enjoy no measurable deterrence improvement—­even against non-­nuclear opponents. This finding alone overturns a central argument advanced by deterrence theorists. And, strikingly, states employing a catalytic posture face a significant increase in high-­intensity attacks, as exem­ plified by the combined Egyptian and Syrian attacks on Israel in 1973, when two non-­nuclear states risked full-­blown war with a known nuclear-­ armed adversary. Indeed, the catalytic posture may be especially deterrence suboptimal. Despite some dampening of low-­intensity disputes, states with a catalytic posture have experienced a marked increase in the likelihood of high-­ intensity conflict. This increased level of conflict may be due to the posture’s low credibility of nuclear use, coupled with incentives for adversaries to preventively strike catalytic states’ nuclear programs while they are still inchoate. Indeed, some of the most spectacular deterrence failures seem at least partly attributable to the creation of preventive war incentives to strangle the nuclear baby in its cradle: in 1967, Israel’s nuclear capabilities were a primary target of the preempted Arab assault and, in 1984 and 1986, crises erupted as India considered preventive strikes on Pakistan’s nuclear infrastructure.53 Coupled with the fact that their arsenals are smaller for a longer period of time by design, these incentives partly explain why catalytic states face a substantially increased likelihood of being targeted in a conflict that escalates to war. For these states, pursuing nuclear weapons writes a security check that their catalytic deterrents cannot cash. In fact, the catalytic posture seems to backfire by creating incentives for preventive military action without an attendant ability to deter conflict. There may be good reasons why states prefer a catalytic posture, but they face a systematically higher risk of armed conventional attacks as a consequence. Just as surprising perhaps, and contrary to conventional wisdom, an assured retaliation posture has little systematic ability to deter conflict initiation and escalation—­even to full-­blown war, and even against non-­nuclear

53. Matthew Fuhrmann and Sarah Kreps, “Targeting Nuclear Programs in War and Peace: A Quantitative Empirical Analysis,” Journal of Conflict Resolution, 54, no. 6 (2010): 831–­59; Narang, “Posturing for Peace?” pp. 38–­78.

250

Chapter 9

opponents who have no ability to retaliate with nuclear weapons. This par­ticular posture has little systematic deterrent effect on armed conventional conflict against nuclear opponents. Indeed, the only two instances of near-­war or war between nuclear powers targeted assured retaliators: the USSR escalated the 1969 Ussuri River conflict against China and Pakistan targeted India in the 1999 Kargil War. Both of these attacks involved substantial armed conflict and nuclear alerts, and risked significant escalation. Assured retaliation has not had regular and serious deterrence failures like the catalytic posture, but it also seems to have little net effect on opponents’ decisions to initiate or escalate armed conflict. Armed conflicts, even serious ones, simply do not disappear once a state acquires the ability to assure nuclear retaliation. This suggests that the risk of escalation to the use of nuclear weapons, even in full-­blown wars, is deemed to be noncredible by adversaries. This finding damages the conventional wisdom that secure second-­strike forces are sufficient to deter high-­intensity conventional conflict. Empirically, they have not been. The asymmetric escalation posture, on the other hand, deters conflict at both low and high levels of violence, against both nuclear and non-­nuclear states. The sheer drop in frequency of armed conflict—­at all levels of intensity, against any type of initiator—­suggests that asymmetric escalation is uniquely deterrence-­optimal, and that the manipulation of escalation risk to the nuclear level appears to be credible enough to deter even limited armed attacks. After adopting this posture, states face an average of three to four times fewer attacks at the war and sub-­war levels of intensity. As I show in the next chapter, after Pakistan switched to asymmetric escalation, it was able to deter Indian conventional attacks in a way that it had failed to do with a catalytic posture. France similarly experienced fewer disputes initiated against it once it adopted an asymmetric escalation posture. No other posture experiences such a significant reduction in armed conflict outbreak. Especially since a state’s decision to acquire nuclear weapons in the first place is typically correlated with high preexisting levels of conflict, the fact that only the adoption of an asymmetric escalation posture significantly reduces the level of violence experienced by a state suggests that this posture has powerful and independent deterrent effects. These findings have important implications for our understanding of nuclear deterrence and nuclear proliferation. First, they overturn a central dogma in international relations and nuclear deterrence theory: that the acquisition of even a minimal nuclear capability radically improves a state’s ability to deter conventional conflict. Empirically, regional powers that merely possess nuclear weapons simply do not experience any identifiable existential deterrent effect. Second, even states with secure second-­ strike forces may not succeed in deterring substantial conventional attacks, whether from nuclear or non-­nuclear powers. There is no magical deterrent

Deterring Unequally I 251

benefit against conventional conflict generated by “existential,” catalytic, or even assured retaliation forces. To reap a significant deterrent effect against conventional conflict, regional states must—­for better or worse—­explicitly orient their nuclear forces to do so by adopting an asymmetric escalation posture. This posture undoubtedly carries with it other significant risks, such as severe command-­ and-­control pressures and an attendant increase in the risk of inadvertent nuclear use. Furthermore, states with this posture have strong incentives to undermine the so-­called nuclear taboo in order to keep their nuclear threats credible, and may do so in ways that risk their own, or international, security.54 However, the findings in this chapter provide a strong clue as to why states may be willing to run all these risks: the significant deterrence benefit that this posture provides. All of this suggests that, theoretically, scholars should cease treating nuclear weapons states as equivalent. The fact that nuclear powers have adopted widely varying nuclear postures that have radically different effects on international conflict calls for a revision to our thinking about how conflict is successfully deterred with nuclear weapons. The fundamental point is that nuclear postures matter. How a state op­ erationalizes its nuclear arsenal has a critical effect on the type of conflict it is likely to experience. Nuclear weapons may deter, but they deter unequally. This finding ought to color how we think about the emerging nuclear landscape, and about what it means for international conflict.

Nuclear Strategy in Crises In addition to the general deterrence effects of these postures, we also want to know how these postures fare in high-­stakes international crises, when nuclear strategy becomes particularly salient. Whereas the full directed dyad-­year dataset provides significant leverage on the general deterrent effect of nuclear postures, the immediate deterrence cases—­when a crisis erupts and there has already been a failure of general deterrence—­are more problematic. It is obviously important to know how different nuclear postures affect state decisions about escalating or de-­escalating a dispute once a crisis has broken out. But precisely because some nuclear strategies dampen the frequency of crisis initiation and escalation, measuring the intra-­crisis role of nuclear weapons after a state has initiated a dispute—­ with full knowledge of the target state’s nuclear status—­becomes difficult. That is, because opponents only strategically initiate or escalate disputes against a nuclear state, nuclear weapons become endogenous to the analysis, 54. Nina Tannenwald, The Nuclear Taboo: The United States and the Non-­Use of Nu­ clear Weapons since 1945 (Cambridge: Cambridge University Press, 2007).

252

Chapter 9

making their impact on intra-­crisis behavior difficult to isolate in a large-­n analysis. For example, after acquiring nuclear capabilities, only six times does a state initiate a dispute with France, and two of these are initiated by Canada (!) over the maritime boundary and fishing rights therein around St. Pierre and Miquelon. The asymmetric escalators especially (France and Pakistan) have faced few serious deterrence failures precisely because of the powerful general deterrence effect of their postures. This powerful general deterrent makes studying the effect of nuclear posture on dispute de-­ escalation very difficult in these cases—­there are very few disputes that require de-­escalation. The quantitative dispute literature often tests intra-­crisis behavior by measuring whether the target state (B) reciprocated the initiator’s (A) threat.55 Several measures of reciprocation exist in various datasets (COW, ICB, and Schultz) based on detailed coding rules for each crisis. However, because of the selection effects induced by nuclear weapons and the strategic entry into crises, there is both a sheer paucity of crises and virtually no relationship between regional power nuclear type and crisis reciprocation. The finding in Horowitz and Beardsley and Asal that nuclear weapons reduce the probability that an adversary will reciprocate crisis escalation is almost entirely driven by the superpowers. For the regional nuclear powers, however, there are much fewer crises and although I find a slight coercive effect for asymmetric escalation postures, nothing reaches standard conventions of statistical significance. It is also unclear why reciprocation should be the dependent variable of interest, as opposed to ultimate level of escalation. Quantitative crisis tests are thus inherently limited in their ability to illuminate the intra-­crisis effect of nuclear postures. Therefore, the next chapter complements the large-­n research design with sets of detailed cases chosen on the basis of these quantitative results, illustrating how these various postures become salient in particular crises. By tracing the process of a series of crises, it is possible to observe not only whether crises were less likely to escalate with the adoption of an asymmetric escalation posture, but whether the mechanism and reason was because adversaries were specifically deterred by the fear of nuclear use in ways that they were not previously. 55. Horowitz, “The Spread of Nuclear Weapons and International Conflict,” pp. 234–­57; Beardsley and Asal, “Winning with the Bomb,” pp. 278–­301.

Chapter Ten

Deterring Unequally II: Regional Power Nuclear Postures and Crisis Behavior

On November 26, 2008, terrorists from Lashkar-­e-­Taiba—­a group historically supported by the Pakistani state—­launched a daring sea assault from Karachi, Pakistan, and laid siege to India’s economic hub, Mumbai, crippling the city for three days and taking at least 163 lives. The world sat on edge as yet another crisis between South Asia’s two nuclear-­armed states erupted with the risk of armed conflict. But India’s response was restrained. It did not mobilize military forces to retaliate against Pakistan or Lashkar camps operating there. A former Indian Chief of Army Staff, Gen. Shankar Roychowdhury, bluntly stated that Pakistan’s threat of early nuclear first use deterred India from seriously considering conventional military strikes.1 India’s restraint has been evident in several crises since Pakistan adopted an asymmetric escalation nuclear posture—­in Kargil, Operation Parakram, and after the Mumbai attacks—­in ways that Delhi was not previously restrained when Pakistan adopted a catalytic nuclear posture or when it had no nuclear weapons. Yet, India’s nuclear weapons capability failed to deter subconventional attacks in Mumbai and Delhi, as well as Pakistan’s conventional aggression in the 1999 Kargil War. That is, even with an assured retaliation nuclear posture, India was unable to deter Pakistan from launching a relatively aggressive conventional attack against India and persisting with the infiltration even after substantial Indian and international mobilization against Pakistan. Why are these two neighbors able to achieve 1. Quoted in “Pakistan’s Nuclear Weapons Deterred India,” The Hindu, March 10, 2009.

253

254

Chapter 10

such different levels of deterrence with their nuclear weapons capabilities? I argue that it is because these states have different nuclear postures, and that leaders are more or less inhibited by different nuclear postures when they make decisions whether to escalate crises. This chapter focuses on that and similar experiences, exploring how regional power nuclear postures have fared in specific dispute dynamics in the cases of India-­Pakistan and the Arab-­Israeli conflict. In chapter 9, I analyzed a large-­n directed dyad-­year dataset to gain some purchase on the general deterrent effect of nuclear postures. However, while the large-­n analysis tests the role of nuclear postures in deterring the outbreak of disputes, it is also important to explore the role of postures in the unfolding of disputes in order to test the hypothesized mechanisms through which nuclear posture affects deterrence. Presently, however, we have a very poor understanding of how nuclear weapons or nuclear postures affect crisis dynamics. This chapter probes these questions by examining whether there is variation in states’ decisions to escalate or de-­escalate a crisis as a function of nuclear posture. That is, within a crisis, do some nuclear postures deter states from conflict escalation better than others? In answering this question, this chapter uncovers the mechanisms responsible for the relationship between regional nuclear postures and deterrence outcomes, ensuring that the correlations established in the statistical analysis are not just spurious but are real and causal. My research design for the within-­crisis tests therefore focuses on detailed comparisons within and across dyads both before and after one or both of the states adopted a particular nuclear posture, in an attempt to isolate its effect on crisis dynamics. While the large-­n tests in the previous chapter address the general deterrence effects—­and therefore the “crises that don’t bark”—­these cases are designed to tease out the role that nuclear postures play once a crisis has broken out, in other words, they are conditional on a deterrence failure. It is precisely in militarized interstate disputes that nuclear postures should have their most salient effect, so what variation do we see in deterrence success based on posture? How does a specific regional nuclear posture affect the calculations of an opponent surrounding a crisis? Are some postures irrelevant? Is one posture particularly inhibitory? The most appropriate method to answering such questions is to explore the actual experience of regional nuclear powers in dispute settings. How one selects the particular crises to study, however, is of critical importance. Any case selection procedure necessarily introduces some sample bias because the full universe of crises is by definition truncated in any limited case study exercise. Nevertheless, some procedures for case selection are more systematic than others. My approach is to select crises involving one or more regional nuclear power based on the regression results from the previous chapter employing a nested research design, where

Deterring Unequally II 255

I pick so-­called on-­the-­line as well as off-­the-­line cases to gain some lever­ age on the dynamics surrounding nuclear posture and crisis behavior.2 To do this, I calculate the predicted probabilities for levels of escalation based on the multinomial logit model for dyad-­years in 1816–­2001. Based on these predicted probabilities and conditional on escalation beyond a certain threshold (DV > 1), I identify crises involving one or more regional nuclear powers that fall into one of three categories: (1) crises whose level of escalation is accurately predicted by the model, (2) crises whose level of escalation is underpredicted by the model, and (3) crises whose level of escalation is overpredicted by the model.3 This procedure provides variation on both the independent and dependent variables, and is consistent with recent guidance on how to select cases from a large-­n analysis.4 In particular, cases are selected that are both on and off the regression surface to ensure that the hypothesized causal processes are operating as expected. Cases should also have key variation in nuclear postures (the independent variable of interest) and crisis intensity (the dependent variable of interest). In addition, I select crises from enduring rivalries in which there is variation in nuclear posture, since this allows me to hold many other country-­and rivalry-­specific variables constant. Based on the quantitative tests from chapter 9, I classify historically significant crises that meet the above criteria according to the respective nuclear postures and whether the model accurately predicted, underpredicted, or overpredicted escalation, shown in table 10.1. In general, the model does well in predicting escalation levels across the various postural types. However, because of the way in which many of these crises cluster around specific dyads, no statistically significant relationship emerges in large-­n tests involving intra-­crisis escalation; and such tests are plagued by unavoidable selection effects. Hence, my research strategy is to conduct focused and detailed analysis of some of these dyads over time to see when and why the model succeeds and fails in accurately predicting escalation levels based on nuclear posture. I do not deny that a whole host of variables are responsible for the outbreak and conclusion of 2. See Evan S. Lieberman, “Nested Analysis as a Mixed-­Method Strategy for Comparative Research,” American Political Science Review, vol. 99, no. 3 (August 2005), pp. 435–­52. 3. To classify crises, I identify the highest predicted probability outcome and compare it to the observed level of escalation. Where the two are the same, I classify the crisis as “accurately predicted”; when the mass of the predicted probability is lower than the observed level of escalation, I classify the crisis as “underpredicted”; when the mass of the predicted probability is greater than the observed level of escalation, I classify the crisis as “overpredicted.” 4. See Lieberman, “Nested Analysis,” and also James Vreeland, “Why Do Governments and the IMF Enter into Agreements?: Statistically Selected Case Studies,” International Political Science Review, vol. 24, no. 3 (2003): 321–­43.

256

Chapter 10 Nuclear Posture of Target State

Level of Escalation Accurately Predicted

Catalytic

Asymmetric Escalation

Assured Retaliation

Vietnam-­China 1978–­79 Syria-­Israel 1967 US-­China 1995–­96 Syria-­Israel 1973 US-­China 1999* Syria-­Israel 1982 Pakistan-­India 1999 Egypt-­Israel 1970 Pakistan-­India 2001 Egypt-­Israel 1973 Angola-­South Africa 1987 Bangladesh-­India 1976 Sri Lanka-­India 1991–­92 India-­Pakistan 1984 India-­Pakistan 1990

Underpredicted

India-­Pakistan 1986 Lebanon-­Israel 1980 Iraq-­Israel 1991

USSR-­China 1969* Vietnam-­China 1974 Vietnam-­China 1987 China-­India 1985–­86

Overpredicted

Egypt-­Israel 1979 India-­Pakistan 1987

Vietnam-­China 1984 Pakistan-­India 1997

Pakistan-­India 1999* Pakistan-­India 2001*

Somalia-­France 1976 Tunisia-­France 1961

Table 10.1. Summary of crisis-dyad predictions based on large-­n results *Denotes regional power was the initiator, not the target, in these crises but faced escalatory moves in which their nuclear postures were relevant as a target state.

these crises. However, the goal of this chapter is to explore and isolate—­as best as possible in an observational study—­the role, if any, of nuclear posture in influencing the course of the crisis by analyzing the calculations and decisions of key actors surrounding each crisis. Because of the difficulties of performing controlled comparisons of crises due to the sheer number of differences between dispute dyads, I employ a congruence method to determine whether the hypothesized effects of nuclear postures are congruent with the observed effects in actual crises.5 This method complements the large-­n analysis because it probes whether the effects that we see on average actually occur in particular cases, lending support to the hypothesized mechanisms in operation. In order to isolate the effect of nuclear posture using this method, it is critical to hold as many other variables constant as possible. Ideal sets of cases will therefore involve a series of disputes involving a single dyad over time, that is, an enduring rivalry, so that changes in nuclear posture can be tested against 5. See Alexander L. George, “The Role of the Congruence Method for Case Study Research,” Working Paper presented at MacArthur Program on Case Studies, Georgetown University, March 18–­22, 1997. Available at http://www.ciaonet.org/wps/gea01/.

Deterring Unequally II 257

conflict outcomes involving the same set of two states, thus holding many variables constant, and thereby minimizing the number of potentially confounding omitted variables. To select crisis cases then, I consider (1) enduring rivalries that hold as many fixed effects constant as possible in an observational study of this nature while still providing variation in nuclear postures as well as frequency/intensity of conflict; (2) within enduring rivalries, I select a range of disputes where nuclear posture shifts and there is variation in conflict intensity, some of which are accurately predicted by the statistical analysis and some that are not; and (3) those disputes that are intuitively salient to students of international security, though this is an admittedly subjective criterion. The India-­Pakistan rivalry and Arab-­Israeli disputes are obvious candidates that meet all three criteria. I first analyze the India-­Pakistan crises before and after the overt nuclearization of the subcontinent. I analyze all the militarized crises over this time period to avoid sample bias within the case. There are thus two disputes in which Pakistan adopts a catalytic posture between 1986 and 1998, and then three disputes where it adopts an asymmetric escalation posture after 1998; India has adopted an assured retaliation posture throughout. Second, I examine major Arab-­Israeli crises since 1967 to trace the effectiveness of Israel’s catalytic deterrent (1967–­91) on capping escalation. Focusing on these enduring rivalries allows me to isolate the effect of the nuclear posture variable over time to probe its independent effect on crisis dynamics.

South Asia under the Shadow of Nuclear Weapons Since the mid-­1980s, all India-­Pakistan crises operated under the shadow of nuclear weapons—­opaquely until May 1998 and overtly ever since.6 Currently there is a debate between scholars of South Asia about the effect of nuclear weapons on the Subcontinent’s security dynamics. Deterrence optimists, such as Sumit Ganguly, argue that the mutual acquisition of nuclear weapons has stabilized conflict in the India-­Pakistan dyad by eliminating the prospect of full-­blown conventional war. Deterrence pessimists, such as Paul Kapur, argue that nuclear weapons have actually had destabilizing effects by incentivizing lower-­level conflict which is not only 6. This section is largely derived from Vipin Narang, “Posturing for Peace: Pakistan’s Nuclear Postures and South Asian Stability,” International Security, 34:3 (Winter, 2010), pp. 38–­78. © 2010 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.

258

Chapter 10

costly in its own right but innately carries with it the potential to uncontrollably escalate.7 However, one omitted variable in this debate is the different nuclear postures adopted by India and Pakistan: the two countries are not the same type of nuclear state and, indeed, Pakistan itself has shifted nuclear postures over time. This section therefore tries to move beyond the optimist-­pessimist debate by focusing on India’s and Pakistan’s respective nuclear postures in order to advance our understanding of nuclear dynamics in South Asia. In this section, I trace the role of India’s and Pakistan’s nuclear postures in the crises where both states had the capability to assemble nuclear weapons, which I date to be from 1986.8 India’s nuclear posture has been constant throughout South Asia’s de facto and overt nuclear periods. Facing no existential conventional threat and privileging strong centralized civilian control over its nuclear assets, India has adopted an assured retaliation posture. Although India tested a nuclear device in 1974 and had a rudimentary retaliatory capability deliverable by aircraft, it only fully operationalized an assured retaliation posture under Rajiv Gandhi in the late 1980s, when it perfected weapons designs and developed nuclear-­capable missile systems. Although the May 1998 tests enhanced its credibility, India continued to maintain an assured retaliation posture. India’s capabilities and stewardship procedures promise certain retaliation should nuclear weapons be used against India.9 This posture currently relies on both nuclear-­ capable aircraft and land-­based ballistic missiles to primarily deter nuclear use against India or its forces.10 Pakistan initially adopted a catalytic posture involving ambiguous nuclear capabilities, exploiting U.S. patronage and interests in Afghanistan in the late 1980s to compel U.S. intervention to defuse crises with India. After May 1998, however, Pakistan faced a severe security situation: India continued to pose a perceived existential threat but was now an overt nuclear power, and the United States was no longer a reliable third-­party 7. See Ganguly, “Nuclear Stability in South Asia,” International Security, pp. 45–­70; S. Paul Kapur, “Ten Years of Instability in South Asia,” International Security, vol. 33, no. 2 (Fall 2008). 8. Some scholars date Pakistan’s ability to assemble weapons anywhere from 1984 (e.g., Ganguly and Hagerty) to 1988 (e.g., Way and Singh). Bennett and Stam code the date to be 1986, though the robustness of the quantitative and qualitative results are unchanged if the date is shifted. 9. India’s official 2003 nuclear doctrine expanded the 1999 draft doctrine to include clauses threatening nuclear retaliation for chemical or biological attacks, and for nuclear use on Indian forces anywhere, but these may have been modeled on U.S. doctrine and are not believed to be credible. See K. Subrahmanyam, “No First Use: An Indian View,” Survival, vol. 51, no. 5 (October/ November 2009), p. 35. 10. K. Subrahmanyam and V. S. Arunachalam, “Deterrence and Explosive Yield,” The Hindu, September 20, 2009.

Deterring Unequally II 259

patron—­in fact quite the opposite. As a result, Pakistan took advantage of India’s nuclear tests to finally adopt an explicit asymmetric escalation posture, which fully integrated nuclear weapons into its military forces to credibly and directly deter Indian conventional attacks. This shift in Pakistani nuclear posture allows me to probe the differential deterrent effects of regional power nuclear postures in South Asia. I analyze every India-­Pakistan militarized crisis under the shadow of nuclear weapons since 1986,11 five in total, to determine whether there is any variation in deterrence dynamics as a function of nuclear posture. The shift in Pakistani posture in 1998 provides a critical test for my hypotheses for the intra-­crisis effects of nuclear posture. Many key variables, including the balance of conventional forces and so-­called dyadic fixed effects (e.g., Pakistani revisionist intentions, the political power of the military in Pakistan, regime type, Indian political aims, etc.12), have been relatively constant since 1986. This feature allows me to isolate the effect of a switch from a catalytic to an asymmetric escalation posture against a state that has an assured retaliation posture throughout. I break up my hypotheses by posture-­dyad “type,” so one hypothesis captures the catalytic vs. assured retaliation interaction in the period 1986–­98, while a second hypothesis focuses on the asymmetric escalation vs. assured retaliation interaction from 1998 to the present. HCatalytic-­AssuredRet: Crises are more likely to be initiated by the assured retaliation state, since it is unlikely to be deterred by an ambiguous and unproven capability; there may be preventive war motivations since the catalytic state’s nuclear capabilities are limited enough to theoretically be eliminated. The catalytic state should make veiled nuclear threats when armed action seems imminent in order to try to catalyze the intervention of an external party. There should be little to no nuclear-­induced cap on escalation as far as the assured retaliation state is concerned. HAsymEsc-­AssuredRet: Crises are more likely to be initiated by the asymmetric escalator, which has a credible shield behind which to pursue its political goals. The assured retaliation state should be deterred from taking armed action beyond a red-­line that might invite a nuclear response. Both states should seek to cap escalation short of full-­scale war. 11. I do so to avoid selection bias in the case selection, analyzing every militarized crisis between the two states. There are still unavoidable selection effects, however, since it is difficult to measure the role of nuclear postures in the non-­disputes qualitatively here. 12. Obviously there is some variation in the overt power of the military post-­Zia, but Pakistan has primarily been a praetorian state with the army and its sister organization, the ISI, wielding tremendous power over domestic society and certainly the nuclear and conventional orientation of the state. See, e.g., Cloughley, A History of the Pakistan Army; Siddiqa, Military Inc.

260

Chapter 10

These compete with the overall null hypothesis, which argues that the mere existence of nuclear weapons and the development of a posture appropriate to one’s own security environment should produce little variation in the effect of different nuclear postures within crises. The null hypothesis is the one effectively expressed by nuclear optimists, such as Ganguly, that nuclearization in the subcontinent since 1986 has eliminated the possibility of conventional action on both sides. Hnull: States adopt a posture that is security-­efficient and there should therefore be no difference in deterrence power between postures. The very existence of nuclear capabilities should provide a sufficient deterrent to conventional or strategic aggression.

In order to test these hypotheses, I examine five crises: the Brasstacks Crisis involving unprecedentedly large-­scale Indian military exercises along the Pakistani border in 1986–­87, when Pakistan was on the cusp of weaponizing its nuclear program; the 1990 Kashmir Compound Crisis when both India and Pakistan had the ability to assemble weapons at short notice; the 1999 Kargil War, the first conflict following overt nuclearization; the crisis in 2001–­2 involving Pakistani-­supported attacks on the Indian Parliament followed by India’s ten-­month mobilization in Operation Parakram; and finally, the crisis provoked by the Mumbai terrorist attacks in 2008.13 These crises provide at least two data points for testing each hypothesis, and though there is a degree of selection on the dependent variable (a crisis), the ultimate level of escalation in these crises varies—­from dispute without the use of force (Brasstacks) to reciprocal use of force (1990 Kashmir Crisis and Operation Parakram) to a full-­blown war (Kargil) with more than 1,000 battle deaths.

Brasstacks I earlier described Pakistan’s use of the catalytic posture in the Brasstacks Crisis. Here I show that the role Pakistan’s nuclear posture played in deterring Indian escalation was negligible. In fact, there is tantalizing evidence that India initiated the crisis as a potential pretext to destroy Pakistan’s inchoate nuclear weapons capability, illustrating the risks associated with a 13. While the 2008 Mumbai crisis is outside the time range for the dataset used in the large-­n analysis in chapter 9, it was a serious crisis that provides an out-­of-­sample case against which I can test my theory. While the full corpus of quantitative data necessary to expand my large-­n tests to cover this time period are not available, there is sufficient qualitative data available to examine the effect of postures on immediate deterrence in this case.

Deterring Unequally II 261

catalytic posture. Furthermore, the reason for Indian de-­escalation had little to do with fears of Pakistani nuclear use, but rather because the United States pressured both governments to withdraw their forces and provided a face-­saving mechanism to do so. To briefly review the origins of the crisis, in the mid-­1980s, India was battling an increasingly violent Sikh insurgency in the key border state of Punjab. Pakistan, seizing on the Sikhs’ battle with the Indian state, provided substantial material support to the Khalistani militants in the hopes of tying down and weakening India’s army in a protracted domestic counter­ insurgency campaign. Punjab is not only India’s wealthiest state, but it is also one of the vulnerable thrust points for any Pakistani incursion into Indian territory. Thus, India did not view Pakistani support for the separatist Khalistani movement in Punjab lightly. In this context, India’s political leadership led by Prime Minister Rajiv Gandhi along with Gen. K. Sundarji, India’s charismatic Chief of the Army Staff, believed, according to the most authoritative study of the crisis, that “a military exercise aimed at Pakistan’s own weak point—­the province of Sindh—­would be a fitting riposte to Pakistan and a threat (with echoes of 1971) that there might be more to come.”14 The crisis occurred in the context of a maturing Pakistani nuclear weapons capability and a nascent catalytic posture. The effect of this catalytic posture, combined with India’s known assured retaliation capability at the time, was to incentivize Indian attempts to possibly destroy Pakistan’s nascent nuclear capability before it fully matured, and before Pakistan could mobilize American diplomatic intervention on its behalf.15 Considerable controversy still surrounds the aims of Exercise Brasstacks and the level of autonomous decision making undertaken by Gen. Sundarji, who has been accused of independently attempting to bait a Pakistani attack on Indian forces so as to launch a preventive war in return, with Pakistan’s Kahuta enrichment facility to be a high-­value target. The most comprehensive study on the Brasstacks exercise discounts the preventive war motivation, instead concluding that the exercise was designed to warn Pakistan about India’s conventional military capabilities and willingness to use them if Pakistan failed to cease and desist from fomenting domestic insurgencies within India’s borders.16 However, critical pieces of evidence remain that Indian military officers, and particularly then-­Chief of Staff Gen. Sundarji, fomented the exercises 14. See Bajpai et al., Brasstacks and Beyond, p. 23. Also see Ganguly and Hagerty, Fearful Symmetry, chapter 4; Kapur, Dangerous Deterrent, chapter 4; Chari, Cheema, and Cohen, Four Crises and a Peace Process, chapter 3; and Paul, The India-­Pakistan Conflict, especially chapter 7. 15. Kapur, Dangerous Deterrent, pp. 85–­86. 16. See Bajpai et al., Brasstacks and Beyond.

262

Chapter 10

in order to preventively strike Pakistan and its nuclear capabilities, particularly the Kahuta facility.17 As the crisis developed, George Perkovich reports that Rajiv Gandhi “contemplated beating Pakistan to the draw by launching a preemptive attack on the Army Reserve South. This also would have included automatically an attack on Pakistan’s nuclear facilities to remove the potential for a Pakistani nuclear riposte to India’s attack.”18 Indeed, Lt. Gen. P. N. Hoon, who led India’s Western Command, wrote in his memoirs that “Brasstacks was no military exercise. It was a plan to build up a situation for a fourth war with Pakistan.”19 Not only was India’s military or civilian leadership undeterred by Pakistan’s nuclear posture, some evidently seriously contemplated fomenting a crisis through Brasstacks to initiate a preemptive military strike on Pakistan’s nuclear capabilities as well as its conventional formations. Although India ultimately did not attack Pakistan, it was not because it was deterred by Pakistan’s nuclear capabilities. It is thus possible here to trace the unfolding of the crisis and the role India’s and Pakistan’s nuclear postures played therein. The Indian military had conducted an earlier exercise, Digvijay, in 1983, setting the precedent for corps-­level exercises that would simulate a war against Pakistan across the plains of Punjab and across the Thar desert that lies astride the Indian state of Rajasthan and the Pakistani state of Sindh. In 1986, Gen. Sundarji was preparing the successor to Digvijay, Exercise Brasstacks, which envisioned four phases (see figure 10.1). The third (November–­December 1986) and fourth (February–­March 1987) phases were to be large-­scale live exercises involving two armored divisions, one mechanized division, and six infantry divisions (two of which were the recently developed Reorganized Army Plains Infantry Divisions (RAPID) designed to be, according to Chari, Cheema, and Cohen, “partly mobile but capable of holding territory, which was a uniquely Indian concept suitable for the India-­Pakistan theater”20). The exercise amassed almost 250,000 Indian troops and simulated combined arms operations in the most likely theater for an India-­Pakistan conventional war. The live exercises were to run on a north-­south axis in the Rajasthani desert roughly 60–­80km from Pakistan’s Sindh province. The north-­south axis was designed to be somewhat less provocative than an east-­west axis, since it would require additional movement to enable offensive operations. Even so, with such a rotation of the force, a deep Indian thrust at that particular point along the 17. See in particular Scott D. Sagan, “The Perils of Proliferation in South Asia,” Asian Survey, vol. 41, no. 6 (November/December 2001), especially pp. 1068–­71. 18. Perkovich, India’s Nuclear Bomb, p. 280. 19. P. N. Hoon, Unmasking Secrets of Turbulence (New Delhi: Manas Publications, 1999), p. 102. Also Sagan, “Perils of Proliferation in South Asia,” pp. 1068–­71. 20. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 44.

Deterring Unequally II 263

Figure 10.1. The Brasstacks Crisis

264

Chapter 10

border could pose a threat to the existence of the Pakistani state.21 As described in chapter 3, the crisis reached a peak in January 1987 until, under American pressure, both states took rapid action to de-­escalate the crisis and removed their forces from the border by mid-­February. What role did Pakistan’s catalytic posture have in the de-­escalation of the crisis? Although the United States intervened to prevent Pakistan from crossing critical red-­lines in its nuclear program, at no point does it appear that India was deterred by Pakistan’s nuclear posture. Pakistan had not yet fully weaponized its nuclear capabilities, but consistently signaled an ambiguous capacity to assemble a weapon if necessary. In Brasstacks, Pakistani fears of a preventive Indian attack on the enrichment facility at Kahuta caused Islamabad to reach out to the Americans, precipitating Ambassador John Dean’s intervention as a pseudo-­mediator to confirm the withdrawal of forces from both sides of the border.22 In the Brasstacks crisis, the threat came from A. Q. Khan in an interview with Indian journalist Kuldip Nayar, who worked for the London Times, conducted on January 28. Though it was not published for several weeks, and thus its effect on the crisis is unclear (Nayar did reportedly pass on the quotation to relevant Indian decision makers), Khan noted that Pakistan had enriched weapons-­grade uranium and boldly stated that: “Nobody can undo Pakistan or take us for granted. We are here to stay and let it be clear that we shall use the bomb if our existence is threatened.”23 As noted earlier, Shuja Nawaz claims that “Pakistan orchestrated [this] leak” through Khan.24 These veiled threats had been made before in 1984, when Khan boasted of Pakistan’s capability to enrich weapons-­grade uranium quickly.25 In addition, the 1999 Kargil Review Committee Report also claims that a private nuclear threat was conveyed from the Pakistani Minister of State for Foreign Affairs, presumably Zain Noorani, to Indian Ambassador S. K. Singh around the same date as the Khan interview, but this also did not seem to deter Indian leaders in any way.26 Even if the Nayar interview was not passed along, Indian and American officials were certainly aware of Pakistan’s maturing nuclear 21. The Indian military, for its part, evidently did not believe that the exercises would alarm Pakistan given that (1) Digvijay had been held several years prior without issue, (2) the exercises were held on a North-­South rather than an East-­West axis, and (3) they were being conducted east of the Indira Gandhi canal and the logistical effort required to cross the canal in order to threaten Pakistan would be substantial for the Indian forces. See ibid., pp. 48–­49. 22. Ibid., pp. 75–­76. 23. Quoted in Bajpai et al., Brasstacks and Beyond, p. 39. 24. Nawaz, Crossed Swords, p. 392. 25. See Corera, Shopping for Bombs, p. 47. 26. Kargil Review Committee, From Surprise to Reckoning: The Kargil Review Committee Report (New Delhi: Sage Publications, 2000), pp. 239–­40.

Deterring Unequally II 265

capability, with Pakistan employing the threat of breakout as a catalytic nuclear posture. Pakistan’s ambiguous nuclear threats did compel Washington’s intervention to alleviate the Indian threat. Given Pakistan’s centrality to American strategy in Afghanistan, the catalytic posture worked in this case. Although the U.S. role was limited, Pakistan was aware of U.S. fears of the advancing Pakistani nuclear program and its incentive to ensure an orderly end to the crisis. However, nothing in that process served as a direct deterrent threat to India. India’s political and military leaders were not deterred from initiating the crisis in the first place due to Pakistan’s nuclear posture, and the de-­ escalatory process was not due to Indian fears of Pakistani nuclear use, but because the United States intervened to ensure mutual troop withdrawals. The Brasstacks Crisis thus provides some evidence of how the dynamics of these various nuclear postures affected the course and conclusion of the crisis. It is important to note that the quantitative model slightly overpredicts the ultimate level of escalation in this crisis. The model predicts a higher level of conflict than actually occurred since catalytic postures—­ because of their limited size and vulnerability—­may provide a window for adversaries to strangle the infant nuclear program in its crib with a preventive strike of some sorts. There is evidence that India contemplated this in 1984 and again in 1986, but refrained from escalating the latter crisis due to American intercession.

1990 Kashmir Compound Crisis As discussed in chapter 3, the most explicit example of the catalytic effect of Pakistan’s pre-­1998 nuclear posture can be found in the 1990 Kashmir Compound Crisis. Deputy National Security Advisor Robert Gates was dis­ patched to the region by Washington to defuse tensions, after U.S. intelligence detected indications that Pakistan might have assembled and moved nuclear weapons. This crisis similarly illustrates that India was not deterred by Pakistan’s catalytic nuclear posture from contemplating serious conventional escalation. By this point, India was believed to have an operational assured retaliation posture and, indeed, was believed to have readied “at least two dozen nuclear weapons for quick assembly and potential dispersal to airbases for delivery by aircraft for retaliatory attacks against Pakistan.”27 Pakistan, on the other hand, again had a vague but emerging capability that might have taken several hours to days to assemble but could be functional, what Congressman Stephen Solarz termed a “Saturday

27. Perkovich, India’s Nuclear Bomb, p. 293 attributed to K. Subrahmanyam.

266

Chapter 10

night special.”28 Indeed, there was growing evidence that Pakistan had obtained bomb designs from China, and both India and the United States believed that Pakistan could assemble a weapon in short order if necessary.29 The Compound Crisis again centered on Pakistan’s support for an insurgency in India, this time in the disputed state of Kashmir. Pakistani assistance to the Kashmir insurgency led India to threaten serious military retaliation across both the Line of Control and the international border. At the height of the crisis, U.S. intelligence concluded that Pakistan might have crossed certain thresholds in its nuclear program. This intelligence seemed to catalyze U.S. intervention in the crisis. Pakistan thus intentionally used its nuclear capability as a signal to the United States to come to its aid to successfully de-­escalate the crisis. These indicators appear to have been consciously manipulated by Pakistani leaders, as then Chief of Army Staff Gen. Aslam Beg admitted, knowing that Delhi did not have the means to detect them. Indeed, these signals were not picked up by the Indians who, according to George Perkovich, were unaware of any indications that Pakistan might have moved nuclear assets and did not particularly care if they did.30 Pakistan’s nuclear activity failed to deter India from threatening action across the LoC or threatening to open a second front on the international border, where India’s quantitative military superiority could be brought to bear on Pakistan. India was therefore not necessarily deterred by Pakistan’s limited catalytic nuclear posture but—­just as Israel and South Africa had done before—­Pakistan shrewdly employed nuclear signaling to the United States, compelling Washington to intervene and prevent India from initiating a conventional assault on Pakistan. Any nuclear signaling to Washington, however, did not seem to inhibit Delhi from threatening to escalate the conflict or open additional fronts along the international border if necessary. Prior to, during, and following the 1990 Kashmir crisis, India’s assured retaliation posture has consistently failed to deter Pakistani proxy support for Kashmiri militants. Likewise, Pakistan’s catalytic deterrent did not weigh heavily on Indian political and military leaders in their decision to deploy upward of 300,000 troops across vulnerable points along the Pakistani border. There is little evidence that Pakistani nuclear capabilities had any inhibitory effect on India at all during either Brasstacks or the 1990 Kashmir crisis. In both cases, the Pakistanis pulled the only gambit they possibly could: signaling to the Americans that they were willing to use 28. Quoted in Ganguly and Hagerty, Fearful Symmetry, p. 101. 29. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 101; also see Chari, Cheema, and Cohen, Perception, Politics and Security in South Asia, chapter 6. 30. Perkovich, India’s Nuclear Bomb, p. 310.

Deterring Unequally II 267

nuclear weapons in the event of a potentially mortal Indian assault aimed at Pakistan’s heartland. Although this crisis escalated and both sides technically used force (across the LoC in Kashmir), the use of force was lower than would be expected by the model. However, critically—­and the reason to study this crisis in particular—­the hypothesized dynamics of the respective postures manifested themselves in the course and conclusion of the crisis and conformed to the theoretical expectations.

Kargil 1999 In May 1998, both India and Pakistan became declared nuclear powers. For India, the tests did not represent a sharp discontinuity in Indian capabilities since they had been gradually expanding their arsenal and delivery capabilities for years, notably through the development of ballistic missile capabilities. India tested a total of five weapons: three sub-­kiloton devices, one claimed-­12kT plutonium fission device, and either a boosted fission device which did not fully “boost” or a thermonuclear device whose second stage failed to ignite. The tests did not immediately affect India’s assured retaliation stewardship procedures or custodial arrangements. Pakistan followed India’s tests three weeks later and claimed to have successfully detonated six fission devices. For Pakistan, the tests marked a sharper discontinuity in its posture as it validated weapons designs and demonstrated a credible capability to successfully detonate uranium fission weapons. After May 1998, Pakistan credibly shifted from an ambiguous catalytic posture to an aggressive asymmetric escalation nuclear posture that threatened the first use of nuclear weapons, particularly on Indian forces should they cross the international border between the two countries. In addition to aircraft, Pakistan also improved its ballistic missile delivery capabilities in this period, often through foreign acquisitions (e.g., Chinese M-­9, M-­11, and North Korean No Dongs). Although Gen. Musharraf downplayed the ability of Pakistan to deliver nuclear weapons in Kargil in his memoirs (and had self-­serving reasons to do so), Brig. Gen. (Ret.) Feroz Hassan Khan notes that in May 1995, Pakistan’s Air Force and PAEC had successfully mastered the F-­16 delivery of a nuclear weapon, meaning it was fully militarily operational at that point and could be delivered by a combat aircraft as opposed to a cargo plane.31 All of Pakistan’s nuclear assets were and are under direct military control. The critical discontinuity for Pakistani posture in 1998 is that Pakistani leaders could publicly and credibly threaten that the 31. Khan, Eating Grass, p. 186.

268

Chapter 10

threshold for Pakistani use of nuclear weapons was now quite low. For example, Musharraf publicly stated multiple times, albeit in various formulations that: “Whenever a foreign leader met me, I asked him to convey my message to Mr. Vajpayee that if his troops took even a step across the international border or the LoC . . . it will not remain a conventional war.”32 Though Musharraf equivocated at times by claiming that nuclear use on the Subcontinent would only occur in “last resort” situations, the 1998 tests did enable Pakistan to credibly threaten nuclear use at relatively low thresholds. It was in this nuclear context—­the overt nuclearization of the subcontinent with Pakistan switching from a catalytic to an asymmetric escalation nuclear posture—­that the Kargil War erupted. While India and Pakistan had embarked on a series of confidence-­building measures at the February 1999 Lahore Summit, the Pakistani military led by Gen. Pervez Musharraf was planning an infiltration operation in the Kargil sector across the Line of Control in Kashmir (see figure 10.2). Whether Pakistani Prime Minister Nawaz Sharif was aware of the extent of the plans remains disputed. Nevertheless, in early May 1999, Indian villagers close to the LoC reported Pakistani intrusions into Indian territory. An Indian Army patrol sent to investigate the reports was attacked by Pakistani infiltrators within India.33 By mid-­May, Indian Army intelligence estimated that almost a thousand regular, but non-­uniformed, Pakistani forces from the Northern Light Infantry had crossed the LoC into India, and by the end of May these forces were believed to have captured seventy positions just inside the LoC.34 These Pakistani forces were supported by artillery and helicopters from the Pakistani side of the LoC in the rugged and inhospitable Ladakh range terrain.35 By this time, the Indian Army was awakening to the extent and scope of the Pakistani intrusion and began to countermobilize to thwart the infiltration. The initial Indian Army response was disastrous, as the high passes and rugged cliffs in the Ladakh range made it difficult for the paramilitary Border Security Force, or even the regular Indian Army mountain units, to repel the Pakistani intrusion. Gen. V. P. Malik, Chief of the Army Staff, requested the use of Indian Air Force assets to provide tactical cover for 32. “Warning Forced India to Pull Back Troops, Says President,” Dawn (Islamabad), December 31, 2002. Available at http://www.dawn.com/2002/12/31/top3.htm. 33. See Ganguly and Hagerty, Fearful Symmetry, chapter 7; Kapur, Dangerous Deterrent, chapter 6; and Chari, Cheema, and Cohen, Four Crises and a Peace Process, chapter 5; General Ved Prakash Malik, Kargil: From Surprise to Victory (Delhi: Harper Collins, 2006). 34. Malik, From Surprise to Victory, p. 115; Ganguly and Hagerty, Fearful Symmetry, p. 154. 35. Malik, From Surprise to Victory, p. 117.

Deterring Unequally II 269

Figure 10.2. Alleged Pakistani incursions in Kargil War

army operations. This was granted by Delhi on May 24 with the explicit condition that “the air force refrain from crossing the LoC in pursuit of its goals.”36 It was the first time since the 1971 war that offensive Indian air power was used against Pakistani positions. The altitudes made combined arms operations very difficult and the Indian Army suffered relatively heavy losses, and the IAF lost two planes. But, importantly, the BJP leadership in Delhi refused to allow Indian forces to cross the LoC in hot pursuit of the Pakistani forces or open up another front elsewhere along the LoC or international border. Though the Indian Army had begun counteroffensives to dislodge Pakistani forces from Indian territory in the first week of June, progress was slow, and the conflict continued through July, when President Clinton decided to finally lean on the Pakistanis to withdraw from Indian territory. The United States did not serve as a mediator in this case, however. President Clinton simply made it clear to Nawaz Sharif in no uncertain terms that Pakistan must withdraw from Indian territory. Combined with the emerging Indian battlefield success, U.S. diplomatic pressure ensured the conflict de-­escalated with Pakistani withdrawal.

36. Ganguly and Hagerty, Fearful Symmetry, p. 154.

270

Chapter 10

What was the role of nuclear weapons in triggering the conflict, which was the first crisis after overt nuclearization? There is substantial evidence that it was the balance of nuclear postures that incentivized Pakistan’s bold infiltration in the first place, invading Indian territory with thousands of irregular Pakistani military forces under its new nuclear umbrella, which Indian commentators believed gave Pakistan a “shield to protect them against Indian retaliation” and a license to aggress to achieve limited revisionist aims.37 While India’s post-­1998 posture was not appreciably different from its pre-­1998 posture, Pakistan was emboldened by its explicit capability which, according to Paul Kapur, now served as “an even more powerful deterrent” to Indian conventional action or retaliation.38 Islamabad, and certainly the Pakistan Army, were not deterred by India’s assured retaliation posture—­especially in undertaking limited incursions, for which nuclear retaliation would be a disproportionate response and difficult to justify to both domestic and international audiences. So for the Pakistani Army, the Kargil operation was conceived as essentially a repeat of 1965’s Operation Gibraltar but under the protection of nuclear weapons. In 1965, India retaliated across the LoC and the international border; in 1999 Pakistan believed that India would not. Was India inhibited from crossing the LoC or the international border because of Pakistan’s asymmetric escalation posture? That is, what was the role of nuclear postures during the crisis? There was plenty of nuclear “chest-­ thumping,” with Pakistani Foreign Secretary Shamshad Ahmad publicly and aggressively threatening, “We will not hesitate to use any weapon in our arsenal to defend our territorial integrity.”39 According to Bruce Riedel, the official note-­taker at the Clinton-­Sharif meeting, Clinton pointedly asked whether Sharif knew that “his military was preparing their nuclear tipped missiles? Sharif seemed taken aback and said only that India was probably doing the same.”40 General Malik’s successor as Chief of the Army Staff, Gen. S. Padmanabhan, claims that Pakistan “activated one of its nuclear missile bases and had threatened India with a nuclear attack” during the Kargil conflict.41 General Malik himself discounts the threat of actual Pakistani movement of nuclear weapons, but concedes that India had received “one or two 37. Kanti Bajpai in “Pakistan in Kashmir,” ABC News Transcripts, March 20, 2000, Lexis-­Nexis; also see Kapur, Dangerous Deterrent, pp. 124–­25. 38. Kapur, Dangerous Deterrent, p. 124. 39. Shamshad Ahmad quoted in Celia Dugger, “Atmosphere Is Tense as India and Pakistan Agree to Talks,” New York Times, June 1, 1999. 40. Bruce Riedel, American Diplomacy and 1999 Kargil Summit at Blair House (Philadelphia: University of Pennsylvania Center for Advanced Study of India, 2002), p. 11. 41. General S. Padmanabhan quoted in Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 141. It is unclear how he learned of this threat given that he was leading

Deterring Unequally II 271

intelligence reports indicating . . . the Tilla [ Jogian] Ranges being readied for possible launching of missiles . . . [so] accordingly, some of our missile assets were dispersed and relocated.”42 Raj Chengappa reports that, based on some unspecified nuclear activity that it detected, India moved to “Readiness State 3” which calls for the assembly of nuclear warheads and deployment alongside delivery vehicles for quick mating if necessary to retaliate against a feared Pakistani first strike.43 One possible explanation for Indian caution is that movement at that particular (Tilla Jogian) missile test range and at Sargodha Air Base—­ where Pakistan’s ballistic missiles and aircraft were both deployed and operationally housed—­is consistent with both nuclear and conventional movements. Thus, they could not rule out the possibility that Pakistan was operationalizing nuclear assets and responded in kind. Accounts of nuclear movements are disputed by both sides. Both Musharraf and Indian officials deny having moved nuclear assets; and it could very well have been that the activity in the Tilla Ranges and subsequent Indian dispersion of its missile assets all involved conventional ordnance. Still, Riedel’s account is the most authoritative and he has since defended the accuracy of the intelligence and the meeting, suggesting that the United States had clear evidence that the Pakistani Army was moving nuclear assets—­with or without Sharif’s knowledge. Regardless of the details of this nuclear signaling, Pakistan’s asymmet­ ric escalation posture directly affected political-­military calculations being made in Delhi. In particular, the BJP, fearing Pakistan’s now-­credible nu­ clear threats, curtailed the Indian military’s options to expel Pakistani for­ces and strictly prevented any operations on or above Pakistani soil. This was in striking contrast to the manner in which India had conducted previous engagements with Pakistan, most notably in response to the 1965 infiltration, which provided the blueprint for Kargil. Most critically, according to Chari, Cheema, and Cohen, the BJP was extremely firm about: [N]ot enlarging the theater of operations beyond the Kargil sector or to attack Pakistani forces, staging posts, and lines of communications across the LoC, despite the fact that this defied military logic and entailed the acceptance of heavier casualties. Its air force had strict orders to avoid attacking targets in Pakistan-­administered Kashmir. This restraint was in marked contrast to India’s response in the 1965 and 1971 conflicts, when nuclear weapons had India’s Southern Command at the time, although he may have learned of these reports upon assuming the post of Army Chief in September 2000. 42. Malik, From Surprise to Victory, pp. 259–­60. 43. Chengappa, Weapons of Peace, p. 437.

272

Chapter 10

not entered the equation and it had not displayed any inhibitions in invading Pakistan.44

Although Gen. Malik may have discounted the possibility of a nuclear response should Indian armed forces have escalated the war horizontally, the BJP leadership was extremely concerned about Pakistan using nuclear weapons on Indian forces if they crossed into Pakistani territory.45 In particular, Prime Minister Vajpayee “was known to have seriously considered a Pakistani nuclear strike had India escalated the war,” particularly if India threatened to open a second front across the international border, which India had done in every previous war with Pakistan.46 Indeed, in an anecdote related by Brig. Gen. (Ret.) Gurmeet Kanwal, Vajpayee seems to have internalized the risks of opening a second front across the international border: General Malik explained to Vajpayee that the armed forces could abide by the order not to cross the LoC but that, tomorrow, if the armed forces are unable to expel the Pakistani infiltrators or were sustaining heavy losses, the government may need to consider his request to expand the conflict or threaten to open an additional front. General Malik said Vajpayee looked “aghast” and said, “but General Sahib, voh thho atom bomb hai (they have a nuclear bomb)!”47

Brajesh Mishra, the National Security Advisor during that time, further revealed that “we were 95 percent sure that we wouldn’t have to cross the LoC. [The use of n]uclear weapons would have been risked if we did cross it however.”48 Both mechanisms by which the asymmetric escalation posture deters were present. First, the BJP feared intentional nuclear use if India retaliated with substantial strike corps conventional power across the international border. Second, Mishra suggested that the BJP also feared the “mad man” possibility: that a lower-­level commander might use nuclear

44. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 139. 45. The logic being that if Pakistan used nuclear weapons on Indian forces in Pakistani territory, it would make a massive retaliatory attack more difficult to justify. It should be noted that the Kargil War took place before India released its Draft Nuclear Doctrine (released just after the war in August 1999). 46. V. K. Sood and Pravin Sawhney, Operation Parakram: An Unfinished War (Delhi: Sage Publications, 2003), pp. 70–­1; also see Kapur, “Ten Years of Instability in Nuclear South Asia,” pp. 80–­83. 47. Author interview with Brigadier (Ret.) Gurmeet Kanwal in New Delhi, January 5, 2011. 48. Author interview with Brajesh Mishra, New Delhi, January 6, 2011.

Deterring Unequally II 273

weapons if Indian forces even advanced across the LoC.49 Indeed, Ganguly and Hagerty concur that, this time, it was not the threat of American intermediation that stopped India from expanding the conflict, but Pakistan’s nuclear capabilities which created a “fear of generating uncontrollable escalation.”50 In discussions with retired and active Indian Army personnel, there seems to be a widely held belief that the Indian military options were curtailed by Pakistan’s nuclear threats and cost the Indian military lives, even though India was ultimately successful in expelling the infiltration.51 The Kargil War thus provides strong support for the hypothesized effects of the different nuclear postures. One of the actors, India, had a deterrent posture that was perceived to be essentially irrelevant in a limited conventional setting. This created space for Pakistan, with an asymmetric escalation posture, to probe India’s escalation threshold in Kashmir through low-­level but deliberate incursions into Indian territory with limited fear of vertical and horizontal escalation. Indeed, had Pakistan been able to achieve its conventional objectives and presented India with a fait accompli, India may have been restrained from using conventional force to expel Pakistani forces altogether since such a situation would have left India no choice but to cross the LoC. As it was, the Indian conventional response was successful enough to thwart the Pakistani incursion, while international—­especially American—­opinion mobilized against Islamabad and helped end the crisis, though Brajesh Mishra noted that India achieved its military aims before American intervention (suggesting India had no catalytic intentions).52 Despite this positive outcome for India, the Kargil War demonstrates (1) the power of the asymmetric escalation posture to deter even a nuclear state from significant conventional operations against it, and (2) the flip side, that an assured retaliation posture is not able to always successfully deter conventional or subconventional attacks by other nuclear powers.

Operation Parakram 2001–­2 002 The same basic dynamics that operated during the Kargil War would be repeated in 2001–­2. The crisis began after Pakistani-­supported Kashmiri militants, believed to be members of Jaish-­e-­Mohammad with some Lashkar-­ e-­ Taiba involvement, audaciously attacked the Indian Parliament on 49. Ibid. 50. Ganguly and Hagerty, Fearful Symmetry, p. 162. 51. Author interviews with anonymous active and retired Indian military officers, New Delhi, July 2010 and January 2011. 52. Author interview with Brajesh Mishra, New Delhi, January 6, 2011.

274

Chapter 10

December 13, 2001.53 Again, India’s assured retaliation posture—­declared publicly through the August 1999 release of a “Draft” Nuclear Doctrine by a government-­appointed advisory body—­did not deter a direct attack on India’s capital by Pakistani-­trained and -­supported organizations. The BJP was outraged and, after its reluctance to authorize horizontal escalation in Kargil because of the fear of a Pakistani nuclear response, this time it decided to authorize preparations for a more robust response. Indeed, Gen. Malik noted to Paul Kapur that “Kargil showed the way. If Pakistan could do Kargil [without escalation to the strategic level], India could do something similar.”54 India’s response in this crisis thus provides a critical test of the role of postures and whether the asymmetric escalation posture can indeed deter even limited conventional responses against Pakistan across the international border. On December 18, India began its military mobilization, calling it Operation Parakram (“Valor”). The holding formations close to the international border were mobilized in three days, while the process of mobilizing India’s three massive strike corps—­which feature the bulk of India’s armored divisions—­to their border assembly boxes from their peacetime cantonments at Ambala, Bhopal, and Mathura took several weeks. Parakram called for the largest mobilization of Indian forces since the 1971 war—­almost 800,000 troops—­with several infantry and mountain divisions deployed across the LoC in Kashmir, three of which were taken off the border with China and redeployed to the west, and all three strike corps deployed in Rajasthan and Punjab for the first time in Indian history.55 And unlike Brasstacks, there was no doubt that these strike corps elements were oriented on an east-­west axis ready to make limited, as well as potentially deep, penetrations into Pakistan’s Sindh and Punjab provinces. Facing the bulk of the Indian military arrayed against it, Pakistan had no choice but to mobilize its corps-­level forces and reserves (ARN and ARS) across the international border. This was the state of affairs by mid-­January 2002. At this point, India engaged in an exercise in compellence by demanding the unequivocal end of Pakistani support to insurgents in Kashmir and to shut down all terrorist training camps behind the LoC.56 Under heavy U.S. pressure, President Musharraf banned Lashkar-­e-­Taiba (LeT) and Jaish-­e-­Mohammed (JeM) and agreed to cut off material support for Kashmiri terrorist organizations harbored in Pakistani territory, decisions he announced in a much publicized January 12, 2002 television address. 53. The most authoritative study on Operation Parakram is Lt. Gen. V. K. Sood and Pravin Sawhney, Operation Parakram: The War Unfinished (Delhi: Sage Publications, 2003). 54. Gen. V. P. Malik quoted in Kapur, Dangerous Deterrent, p. 133. 55. Sood and Sawhney, Operation Parakram, pp. 77–­78. 56. See Kapur, Dangerous Deterrent, p. 134.

Deterring Unequally II 275

Brajesh Mishra claimed that this first war threat was very “serious” and that “we were ready to authorize operations until [U.S. Deputy Secretary of State Richard] Armitage came running [over the telephone] and said wait for Musharraf’s television address,” temporarily staying Delhi’s hand.57 The crisis momentarily abated—­though Indian forces continued to remain deployed—­until May 14, when JeM and LeT terrorists again struck and killed thirty-­two members of Indian Army families at the Kaluchak Army camp in Jammu. General Padmanabhan and Prime Minister Vajpayee thus prepared for a decisive conventional assault on Pakistan in June (see figure 10.3). The three strike corps were concentrated in their respective assembly boxes, ready to execute deep penetrating maneuvers to engage and destroy Pakistan’s two strike corps and seize the Sindh and Punjab provinces, thus threatening to effectively slice Pakistan in two. Again under heavy American pressure, Musharraf pledged to permanently end Pakistani-­backed infiltration into Kashmir. But as a deterrent signal, Pakistan test-­fired three nuclear-­capable ballistic missiles between May 25 and 28, “warning  .  .  . India to apply brakes on its most ambitious plan ever.”58 Both states also reportedly deployed their short-­range ballistic missile assets close to the border which, at this point, were believed to be capable of delivering both conventional and nuclear warheads.59 Brig. Gen. (Ret.) Feroz Hassan Khan also claims that “dispersal and mating plans were actually tested under [these] extremely trying conditions of a physical threat from an enemy.”60 Whether India was aware of these exercises—­which could easily have been misinterpreted as readying nuclear forces for use—­is unclear. In addition, Pakistani officials sent ominous nuclear threats through very public interviews. In one case, Lt. Gen. Javed Ashraf Qazi, a former Director General of the ISI, went so far as to threaten: “If Pakistan is being destroyed through conventional means, we will destroy them by using the nuclear option . . . as they say, if I am going down the ditch, I will also take my enemy with me.”61 President Pervez Musharraf also claims that he conveyed to Vajpayee that “if Indian 57. Author interview with Brajesh Mishra, New Delhi, January 6, 2011. 58. Sood and Sawhney, Operation Parakram, pp. 82–­83; also see Rahul Roy-­Chaudhury, “Nuclear Doctrine, Declaratory Policy, and Escalation Control” in Michael Krepon, Rodney W. Jones, and Zaid Haider, eds., Escalation Control and the Nuclear Option in South Asia (Washington, DC: Henry L. Stimson Center, 2004), pp. 101–­18. 59. Chari, Cheema, and Cohen, Four Crises and a Peace Process, p. 173. 60. Feroz Hassan Khan, “Pakistan as a Nuclear State,” in Maleeha Lodhi, ed., Pakistan: Beyond the ‘Crisis State’ (New York: Columbia University Press, 2011), p. 280. Emphasis added. Khan confirmed this in an author interview on March 19, 2009. This activity may have been detected by Indian intelligence. See Shishir Gupta, “When India Came Close to War,” India Today, December 19, 2002. 61. “Pak Will not Hesitate to Use Nuke Against India,” Press Trust of India, May 22, 2002.

276

Figure 10.3. Operation Parakram June 2002 planned offensives

Chapter 10

Deterring Unequally II 277

troops moved a single step across the international border or the Line of Control, they should not expect a conventional war from Pakistan.”62 However, during the height of the crisis at the end of May and in early June, the BJP balked and never gave the Indian military the authorization to cross the LoC or the international border along Rajasthan. With the monsoon season quickly approaching, the BJP passed on the window it had to execute offensive operations. Nevertheless, the Indian strike corps remained deployed at tremendous cost to equipment and morale, awaiting Delhi’s authorization until October, when Operation Parakram was officially called off and the strike corps returned to their cantonments. Indian military and strategic analysts describe Parakram as an ill-­conceived mobilization that “ended as an ignominious retreat after having failed to secure even its minimum objectives.”63 Clearly, Pakistan remained undeterred from supporting subconventional activity against India’s metropoles. India’s conventional advantage had been blunted by Pakistan’s nuclear posture, and Islamabad did not fear a nuclear response from Delhi over subconventional attacks. Indeed, as Sood and Sawhney record, “the question that kept the Indian political leadership on tenterhooks was whether Pakistan was deterred by India’s nuclear arsenal. The answer, of course, was an unambiguous negative.”64 But just as in Kargil, India was—­at great cost—­deterred from employing limited, let alone overwhelming, conventional force against Pakistan across the international border or the LoC. Although several factors may have stopped Delhi from executing Parakram, the role of Pakistan’s asymmetric escalation posture in deterring India’s conventional assault was crucial.65 Since the 1980s, Indian Army war planning centered on a large-­scale cross-­border attack in a war with Pakistan, known as the Sundarji Doctrine after one of its creators.66 Pakistan’s asymmetric escalation posture directly and powerfully shaped Indian decisions in one indisputable way: 62. Quoted in Zarar Khan, “Pakistan Was Ready to Wage Nuclear War Against India, President Says,” Associated Press, December 30, 2002. 63. Praveen Swami, “Beating the Retreat,” Frontline Magazine, vol. 19, no. 22 (October 26, 2002). Available at http://frontlineonnet.com/fl1922/stories/20021108007101200.htm. It should also be noted that Indian analysts and military officials do believe that Parakram re-­established a foothold in Kashmir and that resulted in the subsequent decline in insurgent activity there since. However, whether this was due to U.S. pressure on Musharraf to rein in the ISI’s and Army’s support for insurgents or due to Parakram is difficult to determine. 64. Sood and Sawhney, Operation Parakram, p. 179. 65. Praveen Swami, “A War to End a War: The Causes and Outcomes of the 2001–­2 India Pakistan Crisis,” in Ganguly and Kapur, Nuclear Proliferation in South Asia, pp. 144–­61; Kanti Bajpai, “To War or Not to War: The India–­Pakistan crisis of 2001–­2” in Ganguly and Kapur, pp. 162–­82. 66. Walter C. Ladwig III, “A Cold Start for Hot Wars? India’s New Limited War Doctrine,” International Security, vol. 32, no. 3 (Winter 2007/2008), especially pp. 159–­63.

278

Chapter 10

Indian decision makers now concluded that such an attack, as was contemplated in May and June 2002, risked triggering nuclear use and was thus no longer possible. Limited war operations before Pakistan countermobilized conventional and nuclear assets were contemplated, but Gen. Padmanabhan argued that such strikes would be “totally futile” in achieving any reasonable objectives, and the only effective response was to “smash [Pakistan].”67 Though the army did not believe Pakistan would use nuclear weapons in a limited conventional war—­as envisioned in the aborted January offensives—­Lt. Gen. V. K. Sood notes that if India “sever[ed] Punjab and Sindh with its conventional forces  .  .  . Pakistan would use nuclear weapons in that scenario.” This was essentially the planned June offensive.68 Hence, once the window for limited retaliatory options passed by early January 2002, senior Indian officials described that “Vajpayee feared that a full-­scale military response . . . could precipitate a wider conflagration. Although Vajpayee believed that the risk of nuclear war was small, he nonetheless saw no advantage in precipitating a crisis of which it might be an outcome.”69 Most tellingly, more than a decade later, Brajesh Mishra himself explicitly conceded that: “We were pretty sure—­fairly certain—­ that if we crossed the border, Pakistan would threaten the use of nuclear weapons. Actual use is uncertain, perhaps doubtful,” but in his view—­the then-­National Security Advisor to Vajpayee—­the risk of nuclear weapons use increased sharply as soon as Indian forces crossed either the LoC (with Pakistani use of nuclear weapons in a different theater) or more critically the international border.70 Mishra went further to argue that there was no such thing as limited war in the India-­Pakistan context, arguing that “if you cross the Line of Control or the Punjab border there is bound to be an all-­out war,” and that Pakistan would escalate and this would be the mechanism for nuclear use.71 During Parakram, the BJP, according to Sood and Sawhney, “was deterred by Pakistan’s nukes more than Pakistan was by India’s putative nuclear second-­strike capability.”72 Manoj Joshi, a well-­placed Indian defense writer, concluded: “[Indian] doctrine had not catered for the simple contingency—­Indian forces being struck by nuclear weapons in Pakistani territory. It was for this reason that . . . Op[eration] Parakaram was called

67. Quoted in Swami, “A War to End a War,” p. 149. 68. Quoted in Kapur, Dangerous Deterrent, p. 138. 69. Two high-­level Indian officials interviewed by Praveen Swami. Praveen Swami, “A War to End a War,” p. 150. 70. Author interview with Brajesh Mishra, New Delhi, January 6, 2011. 71. Ibid. 72. Sood and Sawhney, Operation Parakram, p. 83; also see Ganguly and Hagerty, Fearful Symmetry, pp. 170–­1.

Deterring Unequally II 279

off.”73 Thus, India again decided against undertaking significant conventional assaults across the international border in great part because of the threat of Pakistani escalation to the nuclear level. Pakistan was not deterred from taking aggressive subconventional action against India while India’s political leadership was consistently deterred from retaliating with even limited—­forget decisive—­conventional force because of the fear of nuclear escalation, a fear that had become acute only post-­1998.74

Mumbai 2008 India’s frustration with Pakistan-­backed terrorism again surged after ten LeT militants executed a commando attack on Mumbai on November 26, 2008.75 From the outset, India’s Congress government conceded that its military options to retaliate against Pakistan were again limited, because any meaningful strikes risked uncontrollable escalation, possibly up to the nuclear level.76 India was once again deterred by Pakistan’s perceived low nuclear threshold from executing retaliatory airstrikes against suspected Lashkar camps in Pakistan for fear of escalation to general war. Former Army Chief of Staff Roychowdhury conceded that India was inhibited from a military response because of Pakistan’s nuclear posture in Mumbai as well as Parakram.77 In short, a former Indian Chief of Army Staff—­ who has every incentive to minimize the inhibitory effects of Pakistan’s nuclear deterrent—­has admitted that since 1998, India was at least twice

73. Manoj Joshi, “Ballistic Missile Nasr: A Bigger Threat from Pakistan,” India Today, June 2, 2011. 74. The civil-­military relationship point was consistently emphasized in author interviews with Brig. Gen. (Ret.) Gurmeet Kanwal, Lt. Gen. (Ret.) RK Nanavatty, Air Commodore (Ret.) Jasjit Singh, as well as with active members of the Director General Military Operations (DGMO). In general, the military is much more sanguine about the space for limited and quick conventional assaults across the international border without crossing the nuclear rubicon. This is part of the reason why there is strong bureaucratic support for so-­called Proactive Strategy Options, popularly termed the Cold Start doctrine, which envisages eliminating the long mobilization times for the strike corps by breaking up the larger corps into division-­sized integrated battle groups. This is believed to allow the Indian military to execute operations along the January 2002 lines more quickly. For one view on this, see Ladwig, “A Cold Start for Hot Wars.” 75. Angel Rabasa, Robert D. Blackwill, Peter Chalk et al., The Lessons of Mumbai (Santa Monica: RAND Publications, 2009). 76. Raj Chengappa and Saurabh Shukla, “Reining in the Rogue,” India Today, December 15, 2008. 77. Quoted in “Pak’s N-­Bomb Prevented India from Attacking it after 26/11,” Press Trust of India, March 9, 2009. Available at http://www.ptinews.com/pti%5Cptisite.nsf/0/44975A B7E2577539652575740022C9B7?OpenDocument.

280

Chapter 10

deterred from conventionally retaliating against Pakistan for fear of nuclear escalation. An investigative report by one of India’s leading English-­language newspapers, the Indian Express, on the two-­year anniversary of the attack explored in detail the decision-­making process by India’s principals in the days and weeks after the attack.78 According to that report, in a meeting at the prime minister’s residence on November 29, 2008 attended by India’s national security principals, the prime minister “asked [the service chiefs] whether they had any options in mind . . . he preemptively made it clear that he did not favour another Operation Parakram. That option was off the table from day one, recall sources.”79 The air force proposed limited airstrike packages, but the political leadership was tepid. The army was pressed on its ability to undertake limited ground options, but “General Deepak Kapoor is said to have responded that an operation was possible but he would need a week’s notice and that it would be a ‘highly risky’ affair  .  .  . In the Army’s assessment, any strike would definitely lead to an escalated military conflict and the government ought to be prepared for it.”80 As the conversations about retaliatory military options evolved among the principals in the early days of December 2008, “the Prime Minister then wanted to know if there was a chance Pakistan could misjudge a conventional strike by India and trigger a nuclear response. There was near silence . . . The larger consensus was that you could not be sure about Pakistan’s response. It is reliably learnt that it was this uncertainty which halted Indian strategists from fully backing any military response.”81 Although the United States, and President Bush personally, tried to defuse the tensions between India and Pakistan, “when the dust settled, all agreed that the unpredictability on the Pakistan side and the fear that its decision makers could opt for a disproportionate response, including the nuclear option, stymied any possible chance of military action on India’s behalf after 26/11.”82 The general thrust of this account was largely confirmed by one of the participants in a personal interview with the author.83 As Roychowdhury explicitly declared: “Do nuclear weapons deter? Of course, they do. Pakistan’s nuclear weapons deterred India from attacking that country after the Mumbai strike.”84 78. Pranab Dhal Samanta, “26/11: How India Debated a War with Pakistan that November,” Indian Express, November 26, 2010. 79. Ibid. 80. Ibid. Emphasis added. 81. Ibid. Emphasis added. 82. Ibid. Emphasis added. 83. Interview with high-­level Government of India official, New Delhi, January 2011. 84. Gen. (Ret.) Shankar Rowchowdhury quoted in “Pak’s N-­Bomb Prevented India from Attacking it after 26/11,” Press Trust of India, March 9, 2009. Available at http://www

Deterring Unequally II 281

Since adopting an asymmetric escalation nuclear posture in 1998, Pakistan has been able to uniquely and directly achieve deterrent success against India. These three militarized crises together reveal that Pakistan’s asymmetric escalation posture means that major conventional war—­even in retaliation—­is no longer a viable option for India.85 There is little evidence that the threat of U.S. intermediation stopped India from escalation: U.S. pressure was focused on Pakistan—­not India—­to rein in militant groups. Further, given that India had already withstood U.S. pressure on nuclear testing in 1998, there is little to suggest that India’s leaders were restrained by Washington in these cases in which Indian national interests were even more severely threatened. The more compelling explanation is that Pakistan’s asymmetric escalation posture inhibited Indian leaders from executing militarily effective retaliatory options that might have otherwise been on the menu for fear of triggering Pakistani nuclear use—­concerns not present during the catalytic phase. In conjunction, Kargil, Parakram, and Mumbai demonstrate the dynamic interaction of the assured retaliation posture with the asymmetric escalation posture. One might have expected to see Indian learning, or adaptation between the crises to devise a solution to subconventional Pakistani attacks. However, there is little evidence that such adaptation occurred. Undeterred by India’s assured retaliation posture, Pakistan was and remains able to execute a strategy of “bleeding India by a thousand cuts,” or what is sometimes termed by Indian strategists as “nuclear weapons enabled terrorism.” India, on the other hand and across political parties, has been consistently deterred by Pakistan’s asymmetric escalation posture from pursuing decisive conventional retaliation. Combined with the Brasstacks and 1990 Kashmir crises, the South Asian cases suggest the dominance of the asymmetric escalation posture in intra-­crisis settings. With a catalytic posture, Pakistan was vulnerable to Indian conventional attacks and was saved twice by decisive American intervention on Islamabad’s behalf. In neither case was the reason for India’s restraint Pakistan’s nuclear capabilities. But once Pakistan explicitly tested nuclear weapons in 1998 and established a credible capability to employ nuclear weapons should India cross pre-­defined red-­lines, Pakistan has been able to pursue its political aims at a conventional and subconventional level with relative impunity as India’s political leadership has thrice been inhibited from authorizing horizontal or vertical escalation beyond India’s borders due to the fear of rapid escalation to the nuclear level. So long as India retains an assured retaliation posture against a Pakistani .ptinews.com/pti%5Cptisite.nsf/0/44975AB7E2577539652575740022C9B7?Open Document. 85. Chengappa and Shukla, “Reining in the Rogue.”

282

Chapter 10

asymmetric escalation posture coupled with Pakistani revisionist intentions, the subcontinent’s security environment should concern India. The South Asian cases—­the only ones, except for a Sino-­Soviet clash in 1969, involving direct mobilization between two nuclear powers—­thus provide compelling evidence for the hypotheses regarding the differentiated effects of nuclear postures within crises and the dominant deterrent power of the asymmetric escalation posture over the alternatives.

Arab-­Israeli Crises The second set of crises I explore focuses on the Arab-­Israeli conflicts and the role of Israel’s opaque nuclear capabilities, which were first oriented as a catalytic posture aimed at compelling the United States to intervene on Israel’s behalf to prevent the destruction of the Israeli state. The question animating this set of crises is: to what extent, if any, have Israel’s undeclared—­but not unknown—­nuclear capabilities deterred aggression against the state? The verdict is relatively clear: not very much, even against non-­nuclear adversaries. Israel was subject to Arab conventional threats in 1967 while on the cusp of a nuclear capability, and then again in 1973 and 1991 when its ability to assemble and deliver nuclear warheads was known by its adversaries. The Israeli case—­especially the 1973 Yom Kippur War—­ demonstrates that while a catalytic posture may sometimes be effective in drawing in superpower assistance in the event of a mortal assault, it has little deterrent effect in preventing attacks from occurring in the first place or in preventing escalation during the crisis. In 1991, when Israel’s nuclear capabilities were widely known and its ability to retaliate against aggressors virtually certain—­its posture being in flux as it was at an inflection point between a catalytic and assured retaliation posture—­Saddam Hussein was undeterred from launching 38–­44 Scud missiles at Tel Aviv and Dimona itself. Indeed, the hypothesis for a catalytic nuclear posture within a crisis is: HCatalytic: There should be limited evidence that a catalytic posture deters crisis initiation or escalation, since there is little credible threat of use except in a total war that threatens the survival of the state. The mechanism of capping escalation should be dependent on the catalytic posture compelling a third party to intervene to assist the state.

This hypothesis again competes with the null existential hypothesis that the existence of even minimal nuclear forces should induce a deterrent effect on conventional and strategic assaults on a state. The Israeli case provides strong evidence against the idea that the mere possession of nuclear weapons results in strong deterrent effects. Arab leaders consistently make

Deterring Unequally II 283

largely rational assessments based on the balance of conventional power and forces. Where they were undeterred by Israel’s nuclear weapons it was because they did not find the possibility of Israeli nuclear first use credible.

The Six Day War Israel’s nuclear program was born out of its participation in the 1956 Suez Crisis, when France agreed to supply Israel with reactor and reprocessing technology in large part for assistance in recapturing the Suez Canal along with the United Kingdom.86 After years spent obfuscating Dimona’s true purpose, the reactor finally went online and critical in December 1964. By 1966, Israel had developed the capability to extract and reprocess weapons-­grade fissile material and, on November 2, 1966, Israel conducted final developmental tests on its nuclear designs and was subsequently in a position to produce them.87 Whether this was an actual low-­yield test or a cold test is unknown, but the significance is that, though other states did not know it, Israel was effectively a closet nuclear power before the 1967 Six Day War. There is little doubt that the United States, the USSR, and Egypt knew Dimona’s true purpose and that Israel was seeking a nuclear weapons capability. However, it is also relatively clear that they did not know quite how advanced Israel’s nuclear weapons program was by June 1967. Just as with Pakistan, this generated preventive war incentives for the Egyptians, who had substantial interest in preventing Israel from realizing its nuclear weapons ambitions. Indeed, in 1966, Egyptian President Gamal Abdel Nasser had made specific threats about the need to “immediately wipe out all that enables Israel to produce an atomic bomb.”88 What Nasser did not know was that Israel was, by that point, creeping beyond the point of prevention, though it still had a small enough arsenal and production capacity that destroying Dimona could have seriously set back Israel’s nuclear weapons capability. Against this backdrop, tensions along the Syrian-­Israeli and Egyptian-­Israeli borders escalated through the spring of 1967, culminating in the expulsion of the UN Emergency Force and setting the stage for the first of three crises I examine: the 1967 Six Day War. The causes of the June 1967 Six Day War are well documented elsewhere.89 But from 1963, the Egyptian Army was progressively weakened 86. See Avner Cohen, Israel and the Bomb, chapter 3. 87. Mardor quoted in ibid., p. 232. 88. Nasser quoted in ibid., p. 257. 89. In particular, see Tom Segev (translated by Jessica Cohen), 1967: Israel, the War, and the Year that Transformed the Middle East (New York: Metropolitan Books, 2007); Oren, Six Days of War.

284

Chapter 10

because of the protracted conflict in Yemen, and Nasser—­whose alliance with Syria was increasingly rocky after the dissolution of the United Arab Republic—­was in no position to go to war with an Israel whose conventional capabilities were being augmented by American military sales. Syria’s fruitful arms relationship with the Soviet Union enabled Damascus to pursue an independent foreign policy, and between 1965 and 1967 Syria continued to launch mortars and fida’iyyun raids to terrorize the Israelis even over Nasser’s objections.90 Growing increasingly frustrated, the Israeli Air Force retaliated decisively on April 7, 1967, after a pitched battle erupted on the Golan Heights, downing several Syrian fighter aircraft.91 This episode tightened the Arab coalition, with Syria, Jordan, and Egypt renewing their military cooperation directed toward Israel. In a cycle of increasingly offensive moves, the Israelis and the Arab coalition barreled toward war. The Egyptians expelled the UN force separating them from the Israelis to better defend the Sinai. The Israelis saw this as de facto mobilization for war and, after receiving ambiguous messages from the United States, initiated plans to ensure that their offensive advantage would not be neutralized. For Israel to survive, Israeli leaders believed the Egyptian air force had to be preemptively struck.92 On May 22, Egypt closed the Straits of Tiran, which Israel had declared would be a casus belli. Therefore, on the morning of June 5, the IAF launched a massive three-­ hour 164 sortie surprise attack against the Egyptian Air Force. Conducted while Egyptian pilots were eating breakfast, which left the Egyptian MiGs and Tupolovs vulnerably exposed on their runways, the Israeli preemptive strike effectively eliminated the Egyptian Air Force as a factor in the war. The qualitative and quantitative advantage the Arab coalition had enjoyed on June 4 was swiftly decimated. With command of the skies, the Israelis methodically defeated the Egyptians and then the Syrians and Jordanians in turn. After Israel captured the Sinai, the West Bank, and the Gaza Strip in six days, the superpowers called for a ceasefire. What was the role, if any, of Israel’s ambiguous nuclear capability—­and Dimona in particular—­in the causes or course of the Six Day War? The crisis and war occurred while Israel was on the cusp of its nuclear program and none of its adversaries believed it was capable of assembling or using 90. Safran, From War to War, p. 273; Herzog, The Arab-­Israeli Wars, pp. 146–­47. 91. Oren, Six Days of War, pp. 45–­47. 92. These events are all described in detail in Segev, chapters 5–­15; Oren, Six Days of War, chapters 1–­5. For other works on the Six Days War, see Safran, From War to War; Avi Shlaim, The Iron Wall; Herzog, The Arab Israeli Wars; Kenneth Pollack, Arabs at War: Military Effectiveness 1948–­1991 (Lincoln: University of Nebraska Press, 2002); Trevor Dupuy, Elusive Victory: The Arab-­Israeli Wars, 1947–­1974 (New York: Harper & Row, 1978); Ibrahim Abu-­Lughod, The Arab-­Israeli Confrontation of June 1967: An Arab Perspective (Evanston: Northwestern University Press, 1970).

Deterring Unequally II 285

nuclear weapons. Nasser clearly understood the need to prevent Israel from acquiring nuclear weapons in 1966, but there is little evidence that that was a primary motivation for the escalating hostilities that led to the Six Day War.93 Indeed, the bulk of the historical evidence suggests that Nasser had no intention of initiating hostilities with Israel, especially since he was well aware of his own army’s deplorable shape. Instead, the summary judgment has been that Nasser simply miscalculated how Israel would perceive moves he intended to be a conventional deterrent and to relieve pressure off the Syrian front.94 However, though Nasser may not have wanted a war at that particular time, there is substantial evidence that, in the event of a conflict, Dimona was a high-­priority target. In fact, on May 17, 1967 at 4 pm, the Egyptian Air Force reportedly, according to Michael Oren, conducted its “first-­ever reconnaissance of the Dimona nuclear reactor. Two MiG-­21 jets cut through Jordanian airspace . . . and swooped low over the top-­secret site . . . The incident touched on one of Israel’s darkest concerns, that its pursuit of nuclear power would impel Egypt to launch a conventional attack while it still had the chance.”95 Another reconnaissance mission, which the Israelis again failed to intercept, occurred on May 26.96 These reconnaissance missions are the basis for a controversial argument by Isabella Ginor and Gideon Remez in Foxbats over Dimona, where they hypothesize that the missions were actually conducted by Soviet pilots in MiG-­25 Fox­ bats and that the entire Six Day War was deftly manufactured by the USSR to preventively destroy Israel’s nuclear capability.97 Although many of the critical pieces of evidence have been disputed, there is no doubt that the Egyptians were, at the very least, keen on targeting Dimona. As Avner Cohen notes, “Egyptian maps and contingency plans for offensive operations, found in air bases in the Sinai, confirmed that aerial bombing of Di­ mona was a primary Egyptian objective if hostilities broke out.”98 Had their air capabilities not lay smoldering on Egyptian tarmacs on June 5, Egyptian MiGs and Tupolovs would have almost certainly targeted Dimona during the course of the Six Day War.99 Oren concludes that “Israel’s fear for the reactor, rather than Egypt’s of it—­was the greater catalyst for war.”100 93. See Avner Cohen, Israel and the Bomb, chapter 14. 94. Ibid., p. 265. 95. Oren, Six Days of War, p. 75. 96. Avner Cohen, “Crossing the Threshold: The Untold Nuclear Dimension of the 1967 Arab-­Israeli War and its Contemporary Lessons,” Arms Control Today, June 2007. 97. Isabella Ginor and Gideon Ramez, Foxbats over Dimona: The Soviets’ Nuclear Gamble in the Six-­Day War (New Haven: Yale University Press, 2007). 98. Cohen, Israel and the Bomb, p. 266. 99. See Oren, Six Days of War, p. 75; Cohen, Israel and the Bomb, p. 266. 100. Oren, Six Days of War, p. 76.

286

Chapter 10

There is also evidence that Israel—­evidently unbeknownst to Egypt, the USSR, or the United States—­assembled several nuclear weapons for potential use during the conflict.101 The MiG-­21 reconnaissance missions, which the Israelis failed to intercept, may have conditioned the urgency with which Israel operationalized its nascent nuclear arsenal.102 Munya Mardor’s May 28, 1967 diary entry notes that after midnight, presumably at Dimona, his “teams were assembling and testing the weapon system, the development and production of which was completed prior to the war . . . they fully recognize the enormous, perhaps fateful, value of the weapons system that they brought to operational alert.”103 There is no confirmation of the number of weapons assembled, but Avner Cohen’s authoritative study notes that credible reports suggest that “on the eve of the war, Israel ‘improvised’ two deliverable nuclear explosive devices.”104 The idea was that, as described in chapter 7, if Israel’s survival were threatened, it would demonstrate its nuclear capability to compel the United States to ensure its survival. The speed and ferocity of Israel’s conventional success in the Six Day War ultimately precluded the need to send nuclear signals. There is no evidence that the United States or Egypt was aware that Israel had been able to weaponize its nuclear assets, so there was no deterrent effect to be had. The role of Israel’s incipient catalytic posture was to incentivize preventive action against Dimona. Otherwise, Israel’s nuclear weapons were largely irrelevant to Arab decision making, since the Israelis chose to keep their nuclear weapons advances secret after their initial victories. The critical point is that, though there was sufficient ambiguity about how close Israel was to achieving a nuclear weapons capability, these capabilities did not deter Syrian conventional and subconventional aggression in the 1960s, nor did it deter Egyptian conventional threats. Zeev Maoz calls 1967 a “colossal deterrence failure” for Israel because the mere possibility that Tel Aviv had the ability to assemble a bomb had little deterrent effect on the Arab coalition.105

The 1973 Yom Kippur War If the Six Day War was a “colossal deterrence failure,” the failure of Israel’s nuclear deterrent in 1973 was even worse. Israel’s limited nuclear 101. Cohen, Israel and the Bomb, p. 273. 102. Cohen, “Crossing the Threshold.” 103. Munya Mardor quoted in Cohen, Israel and the Bomb, p. 273. 104. Cohen, Israel and the Bomb, p. 274. 105. Maoz, “The Mixed Blessing,” p. 53.

Deterring Unequally II 287

capabilities were, by 1970, widely acknowledged. As a front-­page story in the New York Times reported: “For at least two years, the United States Government has conducted its Middle East policy on the assumption that Israel either possesses an atomic bomb or has component parts available for quick assembly.”106 In this period, the Israelis were unable to deter Egyptian shelling—­with Soviet assistance—­across the Sinai in the “War of Attrition.” The Israelis responded to Egypt’s attrition strategy with F-­4 strikes on artillery positions, which impelled the Soviet Union to take a more direct role in the conflict by providing air cover to the Egyptians. In fact, by April 1970, Soviet pilots could be heard on the “ether” of the Middle East as they began piloting MiG-­21Js to deter IAF strikes.107 The Israeli military leadership commented that they were now facing a “Russian fist covered by an Egyptian glove.”108 On July 30, the IAF engaged in a dogfight with Soviet piloted MiG-­21Js and shot down five aircraft, killing three Soviet pilots. At this point, the United States feared further escalation and interceded to effect a cease-­fire. Israel’s known nuclear capabilities did not deter Egyptian/Soviet shelling of Israeli positions, but the fear of escalation involving Israeli and Soviet conventional and/or nuclear capabilities did seem to compel the United States to act before the situation spiraled out of control. After the War of Attrition, Anwar Sadat took the reins of Egyptian leadership upon Nasser’s death. Both Egypt and Syria were beneficiaries of an increasingly generous Soviet patron, with advanced versions of the MiG-­ 21, MiG-­25, and Scud surface-­to-­surface missiles being incorporated into Arab inventories prior to 1973. These were balanced by Israel’s increasing incorporation of American platforms, notably the F-­4.109 Sadat’s aim in initiating a fourth round with the Israelis was not to eliminate the Israeli state—­the Six Day War demonstrated that such an aim was unrealistic—­but rather to launch a limited offensive to precipitate an equitable and enduring peace process with Israel.110 That is, Sadat wanted to inject a shock into a status quo in which Israel was satisfied and unwilling to negotiate with a militarily weak Egypt. He knew, however, that his aims were divergent 106. Hedrick Smith, “US Assumes the Israelis Have A-­Bomb or its Parts,” p. 1. 107. David Korn, Stalemate: The War of Attrition and Great Power Diplomacy in the Middle East, 1969–­1970 (Boulder: Westview, 1992), pp. 191–­92; Herzog, The Arab-­Israeli Wars, p. 216. 108. Quoted in Lawrence Whetten, The Canal War (Cambridge: MIT Press, 1974), p. 110. 109. It is difficult to assess a conventional balance of forces because, though the Arab coalition had a numerical advantage, the IDF and IAF kill ratios in 1967 were so large (3:1 and 10:1 respectively) that Israel’s conventional deterrent was perceived to be quite strong, but not insurmountable if the Syrians and Egyptians acted in concert. 110. See Anwar Sadat, In Search of Identity (London: Collins, 1978), 238; Patrick Seale, Asad: The Struggle for the Middle East (London: IB Tauris, 1988), 194–­201; Dupuy, Elusive Victory , p. 387.

288

Chapter 10

from Assad’s, who still aimed to march all the way to Tel Aviv in a “war of liberation.”111 Both Sadat and Assad determined that they would need to exploit the element of surprise and better coordinate their offensive plans than in 1967. To exploit their offensive capabilities in aggregate, the Arab joint command performed the following analysis: In analyzing the problems facing [the Arabs], General Ismail realized that Israel had four advantages: air superiority, technological ability, a high standard of training and what he considered to be guaranteed supplies from the United States . . . The Arabs had studied in great detail the lessons of 1967 and had analyzed every point of Israeli superiority in order to produce an answer. The first conclusion they reached was that it had been a mistake to bring the Israelis to deliver the first strike in 1967; they would do it themselves in 1973 . . . This time they would throw everything they had into the initial attack. In 1967 they had failed to wage a simultaneous multifront war, thus enabling Israel to deal with the various elements in the war at its leisure; this time they would coordinate major Egyptian and Syrian offensives and use the other Arab forces, including Jordan [and Iraq], as a reserve.112

Israel’s nuclear capabilities were neither mentioned nor considered. The Arabs were considering a surprise joint offensive against the Israeli state—­ which the Israelis could easily view as an existential threat to the state, or which could easily morph into one as the conflict unfolded—­without concern or consideration of Israel’s nuclear deterrent. This largely rational assessment of the strengths and weaknesses of the Arab-­Israeli forces simply deemed the use of Israeli nuclear weapons to not be credible, rendering the arsenal irrelevant in the Arab coalition’s decision to undertake an assault on it. Indeed, the Arab offensives were executed simultaneously on October 6, 1973, on Yom Kippur. Israel was caught flat-­footed when Egyptian and Syrian forces launched massive offensives at Israel. Several days into the war, the previously improbable possibility of an Israeli defeat grew alarmingly plausible. The Israeli armed forces were losing materiel and territory at an alarming rate and faced the prospect of the Syrians potentially crossing the Jordan River and the Egyptian forces seizing the Sinai.113 It was at this point that Moshe Dayan, Israel’s Minister of Defense, feared that “this is the end of the Third Temple” and that Israel’s very survival was threatened.114 Though Egyptian war aims may have been limited to the recapture 111. Seale, Asad, p. 197. 112. Herzog, The War of Atonement, p. 28. 113. See Hersh, The Samson Option, p. 222. 114. Ibid., p. 223.

Deterring Unequally II 289

of the Sinai, there were acute fears, particularly by Dayan, that the Syrian offensives could cross the Jordan River and pose an existential threat to Israel (see figure 10.4). There was no way for Israeli leaders on October 8 to be confident that Syrian objectives were limited. Should Syrian ground forces cross the Jordan, Israel’s population centers would have been vulnerable and the survival of the state would have been at risk. As noted earlier, on October 8, fearful of the imminent collapse of the state, Israel performed operational checks on its nuclear-­capable Jericho missiles, signaling that it was potentially considering unsheathing its nuclear weapons—­this signal was not sent to the Egyptians or Syrians, but to the United States in order to compel Nixon and Kissinger to resupply the Israeli state, which was running dangerously low on materiel.115 Up until then, the United States had resisted resupplying Israel because of fears of an Arab backlash and the consequent effect on the global oil market. Moshe Dayan, certain that Israel was quickly approaching the point of last resort, executed Israel’s catalytic posture in an attempt to mobilize American assistance and intervention. With Kissinger still reluctant to meet with Golda Meir, Israel made concerted efforts to place its nuclear assets—­reportedly both nuclear-­capable Jericho missiles and possibly F-­4s—­on alert, according to Cohen, “in an easily detectible way, probably to sway the Americans into action.”116 As discussed previously, the details of the alert are murky, but it has been confirmed that, at the very least, the Israelis opened the hatches on the Jericho silos and conducted “operational checks” on the delivery vehicles.117 This deliberate move, which was sure to be picked up by the United States and presumably the USSR as well, could have had two potential effects: (1) it could compel the USSR to restrain the Egyptian and Syrian advances because of the fear of nuclear escalation, and (2) it could trigger the United States to come to Israel’s aid, which was considered the only viable mechanism to preserve Israel’s existence at the time.118 With respect to the first effect, although Egypt would not have directly picked up the imagery intelligence regarding operationalizing the Jerichos, the Soviets reportedly informed the Egyptians, according to Mohammed Heikal, a well-­placed Egyptian journalist, that the “Israelis had three warheads assembled and ready.”119 But, if true, even this did not deter the Egyptians from “launching a massive armored attack on October 14, aimed at capturing the Mitla and Gidi Passes in the Sinai. Nor did it deter the Syrians from firing Frog 115. Avner Cohen, “The Last Nuclear Moment.” 116. Ibid. 117. See Richard Parker, The October War, p. 121. 118. Hersh, The Samson Option, p. 227. 119. Mohammed Heikal quoted in Hersh, The Samson Option, p. 227.

290

Chapter 10

Figure 10.4. Maximum Syrian penetration during Yom Kippur War, midnight, Octo­­ ber 7, 1973

Deterring Unequally II 291

missiles at an Israeli air force base in Ramat David that also hit civilian targets.”120 So if the Soviets did detect Israel’s nuclear activity, either, contra Heikal, they did not inform the Egyptians or, as is more likely, that warning and any accompanying pressure from the Soviet Union had little deterrent effect on the Arab coalition. With respect to the United States, as described in chapter 7, Hermann Eilts, who was on Kissinger’s staff, notes that the United States became aware early in the war that the Israelis had opened “some of the nuclear hatches” and the move was clearly designed “to influence us to get off our tails and get some equipment to Israel, and that we did.”121 Avner Cohen writes that: By the afternoon of 7 October, US intelligence was aware of ground preparations associated with activation of the Jericho surface-­to-­surface missiles. That activity was interpreted in Washington to be nuclear-­related. Furthermore it was thought that the preparations were conducted in a manner aimed to be detected. William Quandt, at the time a member of the NSC . . . recalls the profound impact these reports made on National Security Adviser Henry Kissinger.122

Kissinger quickly abandoned his strategy of letting Israel “bleed” and, on the evening of October 9, the United States began a massive resupply of materiel to Israel. The resupply of equipment and logistics was a major factor in enabling the Israelis to regroup, blunt, and then push back the Arab armed forces—­eventually encircling the Egyptian Third Army. This generated a superpower nuclear alert, until all sides backed down from conflict. The 1973 Yom Kippur war represents a serious deterrence failure for the Israeli catalytic posture. There is no evidence that either Sadat or Assad were in any way deterred by Israel’s nuclear capabilities. In the run-­up to the war, both spoke of a “total confrontation” and Israel’s known nuclear weapons did not stop them from planning a massive surprise assault on the state.123 Zeev Maoz finds that “at no time did Israel’s nuclear capability

120. Maoz, “The Mixed Blessing,” p. 55. 121. Hermann Eilts quoted in Parker, The October War, p. 121. This has also been confirmed by Yuval Ne’eman, who was Dayan’s liaison with the Pentagon; see Avner Cohen, “Nuclear Arms in Crisis under Secrecy: Israel and the Lessons of the 1967 and 1973 Wars,” in Lavoy, Sagan, and Wirtz, eds., Planning the Unthinkable, p. 118. 122. Cohen, “Nuclear Arms in Crisis,” p. 119. 123. See, e.g., Speeches and Interviews by Anwar el Sadat, January–­December 1973 (Cairo: Ministry of Information, 1975); Raphael Israeli, The Public Diary of President Sadat: Part One, The Road to War (Leiden: E. J. Brill, 1978).

292

Chapter 10

come up as a factor” for Sadat in planning the October 1973 War.124 And during the war, the Egyptians were not deterred by any potential nuclear alerts about which they were made aware. So not only was there no general deterrent effect of the catalytic posture, there is little evidence that there was any intrawar deterrent effect either. In sum, Israel’s catalytic posture did not deter the Arab onslaught in the first place, nor did Tel Aviv’s nuclear activity cap escalation during the war. Certainly Israel was able to achieve at least one of its possible intended effects in signaling nuclear activity during the war: a massive U.S. resupply which possibly—­if not certainly—­ saved the state from serious territorial losses. However, the delay in the American resupply illustrates the riskiness of the catalytic posture and the potential dangers of outsourcing one’s security to a third-­party patron.

Scuds over Dimona: 1991 The final crisis I examine, one whose level of ultimate escalation is underpredicted by the quantitative model,125 is the 1991 Gulf War in which Saddam Hussein launched between 38 and 44 Scud missiles at Israel during the American-­led Operation Desert Storm to dislodge Iraq from Kuwait. Saddam Hussein’s attacks on Israel were hardly a sideshow during the Gulf War: they were deliberately designed to fracture the American-­led coalition by baiting Israel into a fierce response. Neither America’s conventional or strategic power, nor Israel’s conventional or nuclear capabilities deterred Saddam Hussein from brazenly launching conventionally tipped SRBMs at Israeli strategic centers—­including the Dimona reactor. However, he did refrain from using chemical warheads on the Scuds. As I argue below, this was most likely due to American—­not Israeli—­deterrent threats. Even if this is not the case, Saddam Hussein was still undeterred from launching ballistic missiles at Israeli population centers. This crisis occurred as Israel began doubting the American commitment to protect it, and the beginnings of its shift away from a catalytic posture can be detected. Indeed, even prior to the war, the Israeli leadership had made veiled deterrent threats to Saddam Hussein. Foreign Minister David Levy had warned that “whoever attacks Israel won’t live to remember it” and Prime Minister Yitzhak Shamir stated that Israel “would strike back in a most forceful manner.”126 These threats were made to deter any attack 124. Maoz, “The Mixed Blessing,” p. 61, emphasis added. 125. The model predicts reciprocal use of force—­i.e., that Israel would have retaliated against Iraq, even though Israel is coded as being a superpower/U.S. ally. 126. David Levy and Yitzhak Shamir quoted in Shai Feldman, “Israeli Deterrence and the Gulf War,” in J. Alpher, ed., War in the Gulf: Implications for Israel (Jerusalem: Jerusalem Post Press, 1992), p. 198.

Deterring Unequally II 293

on the Israeli state, not just a chemical attack. Then in December 1990, Dan Shomron, head of the IDF, apparently misspoke when he said that “Israel would not be the first to use nuclear weapons in the Middle East,” rather than using Israel’s long-­standing formulation of not being the first to introduce nuclear weapons to the region.127 Although perhaps unintentional, the phrasing was interpreted gravely in the Arab world, according to Shai Feldman, “as a purposive move by Israel to take its nuclear potential ‘out of the basement’.”128 American Secretary of Defense Cheney also hinted that Saddam Hussein should tread with caution against Israel, lest it respond to a chemical attack with “unconventional weapons.”129 These threats, and the U.S. desire to keep Israel out of the war, compelled Washington to deploy Patriot missile-­defense systems in Israel, consistent with the catalytic posture. Nevertheless, Iraq was able to successfully use mobile SRBM launchers and evasion techniques to avoid detection and destruction, launching tens of Scuds at Allied forces and Israel.130 An estimated 38–­44 conventionally tipped Scuds were fired at Israeli population centers and at the Dimona nuclear reactor. Israel’s nuclear posture is still coded as catalytic at this point but was clearly in flux as Israel began to make direct deterrent threats to its adversaries as it began to doubt America’s commitment to protect it. The posture was designed to compel the United States to defend Israel against Iraqi attacks by manipulating the risk of nuclear retaliation, and clearly failed to deter an attack on its strategic and population centers with conventionally tipped Scuds, though Saddam Hussein did not employ chemical weapons. It should be noted that the Israeli leadership was not confident that it could deter a chemical attack, going so far as to employ an array of civil-­defense and chemical-­preparedness measures. Zeev Maoz’s review of the success of Israel’s deterrent again concludes that “the Gulf War represents a deterrence failure” as far as Israel’s nuclear capabilities are concerned.131 But while it did not stop Saddam Hussein from targeting urban centers, military installations, and Dimona itself, did Israel’s nuclear capabilities play any role in preventing a chemical attack? There are two reasons to think that Saddam’s failure to use chemical weapons had nothing whatsoever to do with Israel’s deterrent. First, it is unclear whether Iraq could actually successfully mate chemical warheads with Scuds in 1991. Iraqi military commanders claim that they had chemical warheads ready to be used but refrained from doing so. However, there 127. Dan Shomron quoted in ibid., p. 201. 128. Feldman, “Israeli Deterrence and the Gulf War,” p. 202. 129. Dick Cheney interview with Evans and Novak, CNN Transcripts, February 2, 1991. 130. Gordon and Trainor, The Generals’ War, pp. 228–­48; also see Herzog, The Arab-­ Israeli Wars, pp. 407–­14. 131. Maoz, “The Mixed Blessing,” p. 56.

294

Chapter 10

is not enough data to evaluate this claim.132 Second, America’s deterrent posture, and in particular its policy of “calculated ambiguity” designed to deter chemical use on coalition forces, seems to have weighed heaviest on Saddam Hussein’s mind.133 This policy entailed a veiled threat to respond to chemical or biological weapon use, primarily against coalition forces, with American nuclear weapons. Although it is possible that America’s calculated ambiguity threat could have been partly catalyzed by a desire to prevent Israel from brandishing its own nuclear weapons if Saddam used chemical weapons against Israel, there is insufficient publicly available evidence to determine whether this was the case. In fact, James Baker notes in his memoirs that the primary aim of the policy was to deter chemical and biological use explicitly on “American forces,” and it is unclear whether Israel fell under the umbrella of the threat.134 Indeed, the evidence suggests that, if anything, it was this ambiguous American threat of nuclear retaliation—­not Israel’s—­that prevented Saddam from crossing the chemical rubicon. Though Scott Sagan notes that the evidence for the effectiveness of American deterrent threats is inconclusive, he does write that “Gen. Wafic al-­Samarrai, the head of Iraqi Military Intelligence, later claimed that Saddam was convinced that the United States would retaliate with nuclear weapons if the Iraqis used CW [chemical weapons] or BW [biological weapons].”135 Al-­Samarrai recounts in a Frontline interview that: I do not think that Saddam was capable of taking a decision to use chemical weapons or biological weapons, or any other type of weapons against the allied troops, because the warning was quite severe, and quite effective. The allied troops were certain to use nuclear arms and the price will be too dear and too high . . . We told him very clearly that should he use chemical weapons they [the Americans] will use their nuclear weapons.136

Though the United States did not seem to intend its threat of nuclear retaliation to extend to Israel, it appeared to sow enough doubt in Saddam’s mind that, assuming the use of chemical weapons was an option, he 132. Interview with Wafic al-­Samarrai, Frontline, available at http://www.pbs.org/wgbh /pages/frontline/gulf/oral/samarrai/1.html. 133. See Scott D. Sagan, “The Commitment Trap: Why the United States should not use Nuclear Threats to Deter Biological and Chemical Weapons Attacks,” International Security, vol. 24, no. 4 (Spring 2000), pp. 85–­115. 134. James A. Baker III (with Thomas M. DeFrank), The Politics of Diplomacy: Revolution, War, and Peace, 1989–­1992 (New York: G.P. Putnam’s Sons, 1995), p. 359. 135. Sagan, “The Commitment Trap,” p. 95. 136. al-­Samarrai Frontline interview. There is some evidence that Saddam thought that his chemical stockpile deterred nuclear use as well. See Hal Brands and David Palkki, “Saddam, Israel, and the Bomb: Nuclear Alarmism Justified?” International Security, vol. 36, no. 1 (Summer 2011), pp. 133–­66.

Deterring Unequally II 295

refrained from using them at all, at least so long as he did not face outright regime change. The most convincing explanation for the observed Scud attacks on Israel is that Israel’s nuclear posture failed to deter conventionally tipped attacks at Israel’s strategic centers and its nuclear reactor, while America’s perceived threat shrouded in calculated ambiguity prevented Saddam Hussein from tipping those Scuds with chemical warheads (if he was in fact able to do so). Nevertheless, as far as the catalytic effect of Israel’s posture, it did compel the United States to take substantial and costly efforts to hunt Scuds and provide Israel with Patriot batteries—­perhaps not primarily to defend Israel but to prevent Israel from entering the war and thereby fracturing the Arab coalition supporting America. Some Israeli political and military leaders believe that the United States did not do enough to defend Israel in the 1991 Gulf War and, in fact, as noted in chapter 7, there is substantial evidence that this episode contributed to post–­Cold War pressure in Israel to improve the credibility and robustness of its nuclear deterrent, fearing American abandonment. But the bottom line is that the posture failed to deter or stop serious strategic attacks on the Israeli homeland. In sum, the Israeli crises demonstrate critical failures of the catalytic posture’s ability to deter conventional conflict. There are also reasons to doubt whether it has deterred strategic attacks on Israel as well, given the United States’ role in the Gulf War. Indeed, the 1991 flux in Israeli posture toward something closer to assured retaliation, which failed to deter almost 40 ballistic missile strikes on the Israeli state, does not bode well for that posture’s ability to deter attacks either. In sum, these crises cast serious doubt on the optimality of the catalytic posture to achieve a state’s security goals. This is because the posture relies on outsourcing a state’s security at the most critical moments to a third party that may or may not intercede effectively: both Pakistan and Israel, when they had catalytic postures, experienced major deterrence failures. In at least one of these cases, for Israel in the first Gulf War, American intervention on its behalf was perceived to be insufficient. One school of thought holds that the mere risk of nuclear capabilities ought to induce caution in one’s adversaries. Both the Pakistani and Israeli crises suggest that this is not the case, since neither nuclear opponents (India) nor non-­nuclear opponents (the Arab states) were deterred from aggressing against the catalytic state with significant violence.

Conclusions These crises from the India-­Pakistan and Arab-­Israeli conflicts allow us to isolate some of the dynamics and mechanisms of various nuclear postures in cases of deterrence failure, examining why a particular posture

296

Chapter 10

failed to prevent the eruption of a dispute and how it fared during the crisis. These crises were complicated, with many variables associated with their onset, conduct, and termination. This inherently limits the conclusiveness of any findings. Hence, I employed a congruence-­method approach to isolate the independent variable of interest, nuclear posture, to see if the hypothesized effects on crisis dynamics actually unfolded and whether decision makers were deterred by particular postures. What this exercise further suggests is that the asymmetric escalation posture may indeed be a deterrence-­optimal posture. Not only is there a severe dearth of disputes initiated against France and Pakistan once they adopt an asymmetric escalation posture, but the other nuclear postures examined here—­catalytic and assured retaliation—­have suffered serious deterrence failures against both nuclear and non-­nuclear opponents. Table 10.2 provides a summary of the crisis cases. Israel’s catalytic nuclear posture did not deter Arab states from launching severe assaults against it. While Israel did succeed in drawing in American assistance during the Yom Kippur War and, to a certain extent, during the 1991 Gulf War, it failed to have any deterrent effect in preventing Arab attacks or in capping escalation. Similarly, when Pakistan adopted a catalytic posture prior to 1998, it too relied on its nascent nuclear arsenal to compel American intervention to prevent Indian conventional strikes. India’s assured retaliation posture failed to prevent Islamabad’s support for domestic insurgencies and subconventional conflict. And post-­1998, India’s assured retaliation posture has faced multiple failures against Pakistan: at Kargil at the conventional level, and in the 2001 Parliament attacks and 2008 Mumbai attacks at the subconventional level. In all cases, India’s nuclear posture has failed to deter Pakistani aggression. These failures stand in sharp contrast to the powerful deterrent effect that Pakistan’s asymmetric escalation posture has had on India’s leaders’ responses, in ways that did not occur prior to Pakistan’s shift to an asymmetric escalation posture. These several sets of cases, selected for a variety of research-­design reasons—­variation in model fit, variation in the independent and dependent variables, holding fixed effects constant, and because of their historical significance—­illuminate the dynamics suggested by the large-­n analysis. The results from the quantitative analysis implied that only a particular nuclear posture, the asymmetric escalation posture, is deterrence optimal, uniquely capable of dampening the frequency of conventional attacks as well as capping any resultant escalation. In contrast, both statistically and in important cases, the assured retaliation nuclear posture seems to have little to no effect on the frequency of conflict initiation or subsequent escalation. Indeed, statistically, a state with an assured retaliation nuclear posture is effectively equivalent to a non-­nuclear state in its ability to deter conventional attack. The fact that such states are unable to reap significant

Deterring Unequally II 297

Target State

Nuclear Posture

Opponent(s) Deterred?

Brasstacks

Pakistan

Catalytic

No

India contemplates preventive strikes

Brasstacks

India

Assured Retaliation

No

Pakistan actively supports Sikh insurgency

1990 Kashmir

Pakistan

Catalytic

No

India threatens conventional action

1990 Kashmir

India

Assured Retaliation

No

Pakistan intensifies Kashmir insurgency

Kargil

Pakistan

Asymmetric Escalation

Yes

India constrains response

Kargil

India

Assured Retaliation

No

Pakistan invades across LoC

2001–­2 Parakram

Pakistan

Asymmetric Escalation

Yes

India fears nuclear escalation if it retaliates

2001–­2 Parakram

India

Assured Retaliation

No

Pakistan sponsors terrorism against Delhi

2008 Mumbai

Pakistan

Asymmetric Escalation

Yes

India fears nuclear escalation if it retaliates

2008 Mumbai

India

Assured Retaliation

No

Pakistan sponsors terrorism against Mumbai

Six Day War

Israel

Catalytic

No

Egypt and Syria consider attacking Dimona

Yom Kippur War

Israel

Catalytic

No

Egypt and Syria invade Israel

1991 Gulf War

Israel

Catalytic/ Assured Retaliation

No

Scud attacks at population centers and Dimona

Crisis

Description

Table 10.2. Summary of crisis cases

deterrent power against even high-­level conventional conflict, such as the Kargil War, suggests that regional states need to do something more, and potentially more aggressive, to reap significant deterrent power from nuclear weapons. Equally, the catalytic posture has had serious deterrence failures, especially in the initial phases of adoption when preventive strike incentives are greatest. These findings fly in the face of the conventional wisdom about the deterrent power of nuclear weapons. Exploring particular crises allows for the examination of these processes in actual cases and exploration of how decision makers in opponent states made calculations about conflict initiation and escalation. They largely show that the asymmetric escalation posture is indeed deterrence optimal.

298

Chapter 10

Any operationalization of nuclear forces short of that, empirically, has resulted in significant deterrence failures and a substantial frequency and intensity of conventional attacks. Certainly an asymmetric escalation posture runs other risks, especially those involving unauthorized or accidental nuclear use. Nonetheless, the powerful deterrent effect it generates suggests why some states might be willing to run these risks. The basic point is that nuclear postures matter to the pattern of conflict a state experiences. Not only do regional powers select different strategies and postures, but those choices have critical implications for their ability to deter armed attacks. Nuclear weapons may deter, but they deter unequally.

Chapter Eleven

Conclusion

This book has developed the first rigorous understanding of the sources and deterrence consequences of regional power nuclear strategies. It was motivated by several observations. First, I focused on the experience of the regional nuclear powers because it is their experiences that are most relevant and crucial to the second nuclear age in which we presently find ourselves. Their nuclear choices and their relative power position are distinct from those of the superpowers, demanding theories and analysis distinct from the Cold War scholarship that dominates our present understanding of nuclear strategy and deterrence. This focus on regional powers is in and of itself important. The emerging nuclear powers in the world will all be regional powers, faced with the challenge of operationalizing limited nuclear forces under potentially unstable domestic and international constraints. They will maneuver below the superpowers, confronting different strategic considerations and choosing among different options that more closely resemble those presented to China, India, Pakistan, France, Israel, and South Africa. Second, the existing literature on nuclear weapons proliferation suffers from a considerable existential deterrence bias, focusing almost entirely on a state’s quest for a nuclear weapons capability without differentiating or exploring the choices states make after they acquire nuclear weapons. I find that these choices about how a state operationalizes its nuclear weapons capability—­choices about nuclear strategy or posture—­vary across states and that they matter deeply to the texture of international politics, producing very different abilities to deter the outbreak and escalation of conflict.

299

300

Chapter 11

The deterrence dividend is not evenly distributed across nuclear weapons states. Some states are much better able to deter conflict than others; I find that this variation is a function of their nuclear postures. Critically, the fact that these regional nuclear powers have similar arsenal sizes but different nuclear strategies allowed me to isolate precisely what kinds of nuclear forces are required to deter conflict. Understanding the drivers, constraints, and effects of regional nuclear postures is of immense importance to both scholars and policy makers. I first developed a novel typology of regional power nuclear postures to systematically categorize and characterize the unique strategic decisions that regional nuclear powers make, moving beyond the Cold War lexicon that previously confined discussion of nuclear posture and strategy. By developing a theory for why regional powers might select a particular posture, the book provides insight into the variables that regulate the choice of nuclear strategy, identifies the conditions under which regional powers might shift nuclear postures, and allows one to draw inferences from a comparable set of cases to make predictions about the future proliferation landscape and identify particular security challenges associated with current nuclear powers. Finally, I established that these choices matter to the pattern of conflict in international relations: some nuclear postures deter conflict better than others. With a rigorous understanding of why regional powers select the nuclear postures they do and what effect that posture is then likely to have on regional and systemic conflict, we have a template for predicting those cases where nuclear proliferation may most contribute to instability. For example, states that adopt catalytic or assured retaliation nuclear postures are unlikely to reap a significant deterrent benefit against high-­intensity conventional conflict. Second, states that select asymmetric escalation postures should experience less conventional conflict at all levels of intensity, but face challenges in successfully managing their arsenals and experience higher risks of unauthorized or accidental use of nuclear weapons owing to the credibility requirements of this particular nuclear posture. This understanding allows the international community to specifically tailor its approach to individual regional powers and regions, depending on which risks are most acute.

The Findings and Their Implications The first part of the book systematically identified the diversity and the sources of regional power nuclear postures. I developed a testable and falsifiable theory for predicting which nuclear posture states will adopt. Based on ex ante and observable indicators, one can predict what type of nuclear

Conclusion 301

posture a state might select before it has acquired nuclear weapons. Optimization theory is the first comparative theory of regional power nuclear postures, drawing on insights from the neoclassical realist school of international relations. Clearly, with respect to nuclear strategy, a state’s security environment matters significantly, and a state that is left alone to face a conventionally superior and proximate existential threat has little choice but to adopt an asymmetric escalation nuclear posture. But that condition, though it does sometimes obtain, is rare. Most nuclear powers have sufficient latitude in their security environment that they are able to choose among different nuclear postures. Those that can manipulate superpower intervention through the risk of nuclear breakout or escalation, for example, may find it rational to do so. Others have a choice between asymmetric escalation and assured retaliation postures. In that choice, I argue that the nature of a state’s civil-­ military relations is a critical consideration. In particular, states with highly assertive civil-­military relations should select an assured retaliation posture to avoid relinquishing nuclear assets and the authority to use them to the military, as would be demanded by an asymmetric escalation posture. States with delegative structures, however, have the option to select an asymmetric escalation posture. In these cases, the determinant variable is a state’s comparative resource advantage against its primary adversaries. A state ought to adopt an asymmetric escalation posture if it can afford to do so, or select an assured retaliation posture if it faces resource constraints. I tested optimization theory against all nine regional power nuclear posture choices. I showed that it performs consistently better than the comparable alternative explanations: those offered by structural realism, technological determinism, and strategic culture. It outperformed the alternatives, largely because the theory carefully specifies when and where intervening variables may mediate a state’s choice of nuclear posture. Figure 11.1 depicts the performance of optimization theory’s predictions for each regional power. The only major misprediction is post-­1991 Israel, which the theory anticipated should have adopted an asymmetric escalation posture. In fact, Israel has adopted an assured retaliation posture since 1991. A summary chart of the relative explanatory power and success of optimization theory and the alternative explanations is depicted in table 11.1. For each case, optimization theory may not be the only explanation that accurately predicts the posture chosen. In aggregate, however, it far outperforms any of the alternative explanations. Optimization theory accurately predicts 7 of the 9 posture choices and partially captures an eighth case (the South African case, which relies on analyzing South Africa’s beliefs about being able to compel the United States to intervene on its behalf, even though it is unclear whether the United States would actually have done so, as it did on behalf of its Israeli and Pakistani clients). The only

302

Chapter 11 Regional Nuclear Power

Availability of Reliable Third-Party Patron? Yes

Catalytic Pakistan (1986–1991) Israel (1966–1990) South Africa (1979–1991)

No

Facing Conventionally-Superior Proximate Offensive Threat? Yes

No

Asymmetric Escalation France (1960–1990) Pakistan (1991–present)

Civil-Military Arrangements Assertive

Delegative

Assured Retaliation India (1974–present) China (1964–present)

Resource Constraints? Yes

Assured Retaliation

No

Asymmetric Escalation France (1991–present) Israel (1991–present)

Figure 11.1. Optimization theory’s performance against empirical record

outcome it fails to accurately predict is Israel’s assured retaliation posture after the 1991 Gulf War. However, none of the theories accurately predicts this outcome, which is either an anomaly or inaccurately classified. Nevertheless, optimization theory achieves at least a 78 percent success rate, and 89 percent if one accommodates South Africa’s unit-­level beliefs about superpower patronage. Structural realism and strategic culture only accurately predict 2 of 9 postures, with each partially predicting three additional cases (narrowing it down to two possible postures, of which the state chose one). This is a success rate of between 22 percent and 55 percent depending on how generous one is in allowing partially correct predictions to count. Technological determinism explanation is the least accurate theory, correctly predicting only 1 of 9 outcomes, and partially capturing three additional cases (and even then not for the correct reasons). The view that nuclear postures are driven bottom-­up by technology and not top-­down by political decisions to develop specific technologies is difficult to sustain empirically. Even in the cases where technological determinism correctly predicts the posture outcome, the outcomes and rationales are more consistent with political decisions to adopt those postures than with technological drivers.

Conclusion 303 Correctly Predicted?

Posture Decisions

Optimization Theory

Structural Realism

Technological Determinism

Strategic Culture

Pakistan pre-­1991 (Catalytic)









Pakistan post-­1991 (Asymmetric Escalation)





?



India (Assured Retaliation)



?





China (Assured Retaliation)



?



?

France pre-­1990 (Asymmetric Escalation)





?

?

France post-­1990 (Asymmetric Escalation)







?

Israel pre-­1991 (Catalytic)









Israel post-­1991 (Assured Retaliation)









South Africa (Catalytic)

?

?

?



7–­8

2–­5

1–­4

2–­5

Successful Predictions

Table 11.1. Explanatory success for optimization theory and alternative explanations = successfully predicted; = unsuccessfully predicted; ? = partially predicted

Furthermore, even where optimization theory and, for example, structural realism, predict the same posture, there is more evidence that the former does so for the correct reasons, and that states consider the variables of optimization theory in the hypothesized sequence—­searching first for a reliable third-­party patron to augment external balancing before assessing whether they face binding security constraints. In the seven accurate predictions optimization theory generates, it does so for identifiable reasons that are largely corroborated by the empirical record. That is, optimization theory generally captures the correct postures, for the correct reasons and correctly identifies the decision-­making processes. Thus, against the full universe of empirical cases involving selection of nuclear posture—­a decision that unfolds deliberately over many years and often over many leaders—­optimization theory is the most valid theory available. It is also the first broadly comparative theory for why states select the nuclear postures they do, suggesting that states may be rational to sacrifice deterrent power in certain security environments and under particular organizational and relative endowment circumstances. I have demonstrated that states do appear to make decisions along these lines. Where an asymmetric escalation posture is not demanded by a state’s security environment and its civil-­military organs are not capable of safely and

304

Chapter 11

trustingly stewarding such a posture, states opt out and adopt an assured retaliation posture that entails substantial procedural and physical negative controls, minimizing the risk of unauthorized and accidental nuclear use and reducing the threat to the state’s and international security. Optimization theory also provides a generalizable framework to predict when regional nuclear powers might shift their postures. If a state adopts a particular posture at a specific node, exogenous events that change the value of that variable can lead to different postures. Understanding that dy­namic and the reasons for why a state adopts a specific nuclear posture provides insights into when and under what conditions that posture might change. The empirical tests validated the internal construction of the theory and therefore the framework provides some testable implications for how future regional nuclear powers might operationalize their nuclear forces, and under what conditions present regional nuclear powers might shift their nuclear postures. It is important to underscore that this is a structural theory that predicts outcome and not necessarily process: states can take a while to fully adopt the posture they are seeking, as in the case of India, and the process by which it does so may be uneven and inefficient. How quickly or efficiently a state adopts its chosen nuclear posture demands a different theory and is an important area for future research. The theoretical contribution of optimization theory is that it provides the first valid and testable theory of regional power nuclear postures. It provides an understanding for why regional powers select the nuclear postures they do, with clearly specified and measurable variables indicating when postures might shift and how to assess what nuclear postures future nuclear powers might adopt. In and of itself, this is an important theoretical advancement. However, in combination with the second part of the project, it also provides some insights into the types of conflict one can expect involving regional nuclear powers—­before these powers have even acquired nuclear weapons. There are also real-­world contributions of this theory. For example, the theory provides a template for how to think about how and why emerging nuclear powers might select their postures, providing testable implications for the type of posture countries such as North Korea or Iran, as well as those that may follow, might select. For example, before Iran potentially acquires nuclear weapons, one can deploy optimization theory to assess which nuclear strategy it would be likely to adopt and then predict how its conflict environment might be likely to change. In Iran’s case, it has neither a reliable third-­party patron nor faces a proximate offensive threat. It does, however, have complicated civil-­military structures. On the one hand, the regular military forces are under highly assertive control of the theocratic leadership; on the other hand, a highly delegative organ, the Iranian

Conclusion 305

Revolutionary Guard Corps (IRGC), is charged with managing strategic weapons. This suggests that Iran is in a position to either go down the path of China and India and select an assured retaliation posture, or—­ depending on how civil-­military relations evolve—­could go down the path of Pakistan and adopt an asymmetric escalation posture. If it heads toward the latter, the region could face a conflict dynamic where the pressing issues are arsenal security and management, and possible Iranian emboldenment as observed in the Pakistani case. In East Asia, the theory helps us think about two critical cases: Japan and North Korea. Although Japan is not a nuclear weapons state in the traditional sense of the definition, its status as a “standby nuclear state” means that it has many of the critical pieces in place: tons of reprocessed plutonium under sovereign control, an advanced technical and manufacturing base, and a sophisticated military that already possesses possible nuclear delivery vehicles. If Japan were to decide to move beyond its standby status toward a nuclear weapons capability, how might it choose to posture its forces? My theory suggests that, thus far, Japan would manipulate the threat of breakout to ensure that the United States meets Tokyo’s perception of Japan’s defense requirements. That is, Japan should have a strategy like the catalytic posture already in mind, and if it moved further toward an ambiguous nuclear capability it would likely do so to keep its superpower patron, the United States, in a hard alliance. If the alliance with the United States were ever to fail to meet Japan’s perceived defense requirements and it chooses to openly nuclearize, optimization theory suggests that Japan ought to adopt an assured retaliation posture. As an island state with maritime buffers that does not face a conventionally superior proximate offensive threat, and with assertive civil-­military structures as a result of its self-­imposed post–­World War II limitations, Japan would be likely to select an assured retaliation posture. Similarly, in North Korea’s case, my theory suggests that Pyongyang is likely to use its small and technically limited nuclear arsenal as a mechanism for keeping its third-­party patron—­China—­ closely attentive to North Korea’s security. Thinking about the North Korean problem this way offers a novel means to analyze its choices and behavior, and also underscores the centrality of China to a resolution of the nuclear crisis on the Korean Peninsula. The theory also suggests that were China to abandon North Korea, Pyongyang’s resultant posture would likely shift toward a more aggressive first-­use nuclear posture. That is, an isolated North Korea facing the conventionally superior U.S.-­South Korea alliance might select an asymmetric escalation posture, with all the attendant challenges of managing such a posture. These are testable propositions as East Asia’s nuclear dynamics become increasingly complicated. Optimization theory, then, is not just an

306

Chapter 11

abstract exercise about the sources of posture, of interest only to academics, but something that provides traction in predicting how nuclear proliferation will affect specific patterns of regional and systemic conflict. The second part of the book demonstrated that the choice of nuclear posture matters to international conflict. A state’s choice of nuclear posture significantly affects its ability to deter armed attacks. Focusing on the regional powers, which have similar size arsenals but different postures, enabled me to answer the critical question: what does it take to deter conflict? Using a nested research design that began with a large-­n analysis of the average effects of nuclear posture over time and space and that concluded with a set of detailed crises to tease out the effects and mechanisms of the postures, I found that the asymmetric escalation posture is deterrence optimal. That is, it uniquely deters the outbreak and escalation of conventional conflict at all levels of measured intensity against both nuclear and non-­ nuclear opponents. The assured retaliation posture has no measureable effect on the incidence and intensity of conventional conflict. From a statistical perspective, it is as if these states had no nuclear weapons at all, since there is no measurable reduction even in intense conventional aggression relative to non-­nuclear states. The proposition that these postures deter nuclear use rather than conventional attack may hold, but it is impossible to test empirically, since all states—­both nuclear and non-­nuclear—­have deterred nuclear use with equal success since 1945. The catalytic posture seems particularly security inefficient; these states have experienced an increased rate of high-­intensity conflict, partly due to preventive war incentives to destroy the catalytic nuclear state’s potentially vulnerable nuclear arsenal. The catalytic nuclear posture seems to gamble on the augmented deterrent effect of being able to exploit a third-­party power to intervene on the state’s behalf—­but, too often, adversaries seem willing to exploit that gamble and attack the catalytic state, including sometimes with high-­ intensity conventional power, as in the 1973 Yom Kippur War. The catalytic and assured retaliation postures likely afford states insurance against outright extinction, but my analysis suggests that these postures buy them little else in terms of deterrent power. The finding that the asymmetric escalation posture is deterrence optimal has several important implications. First, from the standpoint of international relations and proliferation theory, this finding suggests that nuclear weapons do not serve as an existential deterrent, even against high-­level conventional conflict. Prevailing theories of deterrence suggest that even small and non-­operational nuclear forces, and certainly secure second-­ strike forces, ought to induce caution in adversaries even at the level of low-­ intensity conventional conflict, owing to the small but catastrophic risk of escalation to the nuclear level. This assumption turns out to be wrong. Secure second-­strike forces operationalized in the assured retaliation posture

Conclusion 307

have been unable to systematically deter conventional wars, such as the 1999 Kargil War. Only states that operationalize their nuclear forces aggressively toward first use, which deters through a direct deterrent and an indirect “mad-­man” mechanism, seem to be capable of systematically deterring serious conventional attacks. This suggests that the deterrence dividend of nuclear weapons is unequally distributed across regional powers, and that states have to do substantially more than simply acquire nuclear weapons in order to reap deterrence benefits against conventional conflict. Second, this finding provides insights into the types and frequencies of conflicts one can expect to see nuclear powers involved in because these conflicts will be a function of the specific nuclear strategies they adopt. States that adopt catalytic nuclear postures are likely to be particularly vulnerable to preventive attack in order to strangle the nuclear baby in its cradle before the threat grows and becomes explicit. Regional powers with assured retaliation postures are likely to experience the same frequency and intensity of conventional conflicts as before they nuclearized; that is, states that adopt assured retaliation nuclear postures should not expect to suddenly enjoy a significant reduction in conflict. Regional powers that adopt asymmetric escalation postures should experience a significant deterrent effect against both nuclear and non-­nuclear opponents. However, the asymmetric escalation posture is likely to raise acute concerns about the risk of unauthorized and accidental nuclear use because the credibility requirements of the posture demand potentially ceding significant authority and nuclear assets to end users, particularly in chaotic crisis scenarios. Thus, the primary management challenge for these states—­and the regional and international community—­is ensuring the safety and security of the nuclear arsenal. This research all points to the overwhelming conclusion that the existing proliferation literature has missed considerable texture and variation in the sources and consequences of nuclear proliferation by focusing almost exclusively on a state’s acquisition or test of its first nuclear weapon. Obviously this is a critical step in a state’s power position, but I find that it is what comes afterward—­which often has no relationship to why a state pursued or tested nuclear weapons in the first place—­that has a significant impact on regional and international conflict. By stopping at a state’s first nuclear test or acquisition of its first nuclear weapon, the proliferation literature mistakenly treats all nuclear weapons states as equivalent. They are not. Some adopt ambiguous catalytic nuclear postures, others select assured retaliation postures, and still others choose asymmetric escalation postures—­each with differential deterrence success. Understanding why states adopt these particular postures and what the consequent effects on conflict are demands different theories and tests than those presently available. This research fills that lacuna and finds that nuclear postures matter

308

Chapter 11

greatly to international politics. Far more attention should be paid to the challenges and choices states face in operationalizing their nuclear forces.

Avenues for Future Research This book has offered a first cut at developing a rigorous understanding of the sources and deterrence consequences of regional power nuclear postures. Much more research remains to be done to augment these findings. On the deterrence consequences side, the large-­n general deterrence findings for regional nuclear postures were found to be largely robust to specification, and the intra-­crisis tests in the enduring rivalries isolated some of the hypothesized deterrence effects of the various postures. I focused on enduring rivalries—­to hold as many variables constant as possible—­and especially India-­Pakistan and Arab-­Israeli crises for research design purposes. Thus far, in conjunction with the large-­n general deterrence tests, my probe of several enduring rivalries suggests that nuclear postures do vary in their deterrent power. However, more work can be done on the latter set of crisis cases to bolster the robustness of this finding. It would be useful to have a full database exploring all crises involving regional nuclear powers in which nuclear threats played a role. As more data becomes available from China and the former Soviet Union, for example, exploring the role of nuclear postures and China’s assured retaliation posture against the USSR’s superpower posture in the 1969 Ussuri River clashes could provide further corroboration of the weak deterrent effect of assured retaliation nuclear postures. In any case, having a full database of crises involving regional powers and nuclear threats would help further systematically test how these deterrence dynamics appear in actual decision-­making processes. On the sources of nuclear posture, the biggest methodological limitation at present is the small number of regional power nuclear powers—­ something that may be beneficial for international stability, but that complicates the study of regional nuclear proliferation and strategy. Though shifts in nuclear posture enabled me to test the validity of my theory against nine posture observations, rather than just six, the theory may require further refinement as additional regional powers shift postures and as emerging nuclear powers select nuclear postures. These events will provide out-­of-­ sample tests that can validate or further refine the theoretical structure. In particular, there may be certain types of states—­perhaps “revolutionary” states with more than limited revisionist intentions—­that do not adhere to the theoretical predictions of the model. (The model admittedly has a status-­quo-­intentions bias, owing to the characteristics of the majority of empirical regional powers.) Additionally, as more data becomes publicly

Conclusion 309

available about the reasons for why regional powers have selected the nuclear postures they have, the theory can be further tested and refined. Moreover, while the typology of nuclear postures that I generated in this project is broad enough to capture the logical and empirical differences between the three nuclear postures while maintaining analytical clarity within them—­roughly aligned across the spectrum of political use, assured retaliatory, and first-­use postures—­it may not be fully exhaustive. It is thus far, but we could in the future find variations of nuclear postures that do not fit neatly into this typology. The typology would then require further exposition and refinement. For example, with the end of the Cold War, catalytic postures may give way to purely existential deterrent postures—­ unoperationalized or scarce forces without a third-­party patron—­if great powers are no longer deemed to be reliable patrons. The structural conditions that enable a catalytic posture are simply more difficult to obtain in the post–­Cold War world. Indeed, two states in this book with catalytic postures, Pakistan and Israel, shifted postures after perceived abandonment from the United States at the end of the Cold War, for reasons specified by optimization theory. However, both took concerted steps to adopt an asymmetric escalation and assured retaliation posture, respectively, rather than continue to rely on an existential force. Regardless, however, this study provides a first step in moving beyond the Cold War vocabulary to think about the rich variety of nuclear postures and strategies that regional powers have crafted. This work also spawns several additional research questions. First, optimization theory is a theory of outcomes, not process. One rich area for future research is exploring how efficiently states translate a choice of nuclear strategy into an operational posture. India, for example, unbeknownst to its adversaries, took many more years than is traditionally appreciated to develop a reliable delivery capability for its nuclear weapons.1 Jacques Hymans has explored the efficiency with which states acquire nuclear weapons, but it is also critically important to analyze the efficiency with which they operationalize their postures as well. There may be unanticipated windows of vulnerability if states cannot efficiently adopt the postures or strategies they select.2 Second, the fact that some nuclear postures have such powerful deterrent effects stems from the fear that nuclear weapons might actually be used. Does this mean that the so-­called nuclear taboo is not very strong, at 1. See Gaurav Kampani, “The Process of State Agency Formation and India’s Pre-­1998 Operational Nuclear Posture,” unpublished manuscript, 2013. 2. Jacques E .C. Hymans, Achieving Nuclear Ambitions: Scientists, Politicians, and Proliferation (New York: Cambridge University Press, 2012).

310

Chapter 11

least not strong enough that nuclear threats are ignored? Or does it mean that the taboo only exists in some subset of states, such as the United States and Western Europe? The credibility of nuclear threats varies inversely to the power of the taboo, so states that need to rely on nuclear weapons to deter some set of conflicts may have incentives to weaken the taboo. This research suggests that a global taboo may not yet exist, or at least may not apply equally across states. It would be worth investigating where this is the case, and why. Third, this book has focused primarily on the characteristics of the deterrer, and its choice of nuclear posture. One assumption made in this book is that there is common knowledge and full information about a state’s nuclear posture. This was necessary to make the analysis tractable and empirically proves to be a reasonable assumption, but it may not always be the case. Are there features of the intended deterred target that make a state more or less likely to be deterred by nuclear threats or postures? Do states with different types of selectorates or audience costs have differential propensities to be deterred by certain nuclear postures or threats? When do states perceive that an opponent has achieved an “assured retaliation” posture—­is the point of plausibly certain retaliation reached earlier rather than later in a state’s attempt to adopt this posture? Finally, the book also raises interesting questions about the dynamic relationship between nuclear and conventional postures. Although I collapse much of this dynamism into the blunt “security environment” variable, states’ decisions about nuclear and conventional posture are coupled and may be dynamically related to those of its adversaries. Understanding the drivers of both conventional and nuclear posture, and how they relate, may also provide a fuller understanding of deterrence dynamics—­particularly in crises where a state can deter an opponent through manipulations at both the conventional and nuclear levels. Answers to these additional research questions would provide a deeper understanding of the various inputs and outputs of nuclear postures, and how nuclear postures relate to a state’s broader ability to deter conflict against it.

Final Words This book has moved the proliferation literature forward by focusing on the choices and pressures states face after they acquire nuclear weapons. The finding that the asymmetric escalation posture is uniquely deterrence optimal is likely to be controversial among scholars and practitioners of non-­proliferation since it implies that nuclear weapons remain useful to deter conflict and that that utility is only achieved by states that orient their nuclear forces explicitly toward first use. It may mean that the proliferation

Conclusion 311

problem is harder to stem, since states still view nuclear weapons—­and an asymmetric escalation posture—­as valuable tools to achieve security. The tradeoff and danger is that this particular operationalization is more difficult to manage than the alternatives. It can generate significant risks of unauthorized and accidental uses of nuclear weapons, particularly for states that find themselves forced to adopt this posture for security reasons rather than because of permissive civil-­military organs and resource advantages. By understanding the pressures that states in this position face, however, the international community and regional adversaries have the option to take steps to ameliorate the variables that force states into an asymmetric escalation posture, either by taking steps to reduce the severity of the state’s security environment or by recognizing that these states may operationalize their nuclear forces in potentially risky ways and providing negative control and management assistance to reduce the risks of accidental or unauthorized nuclear use. These findings also provide a cautionary note to states that pursue nuclear weapons for deterrence purposes, suggesting that they have to do a lot more to reap a substantial deterrent benefit from nuclear weapons than simply acquiring and testing them. This ought to be an input into a state’s cost-­benefit calculation for pursuing nuclear weapons in the first place. Leaders should take a hard look at whether they are willing to take the steps and investments necessary to pursue nuclear weapons for deterring conventional conflict. If not, then they may be better off without nuclear weapons, lest they face the increased preventive war conflicts that other nascent nuclear states have experienced, or be forced into an unholy choice between deterrence and arsenal security. In this way, my findings provide an argument against acquiring nuclear weapons for the vast majority of states in the international system that do not face existential conventional threats. Ultimately, this book is realist in the sense that it recognizes the world as it is—­states have nuclear weapons and more states may acquire them in the future—­and provides insights into how and why states manage their nuclear forces the way they do. It shows that nuclear posture and strategy are important and matter deeply to international relations. The world stands on the threshold of a second nuclear age, characterized by a growing number of regional nuclear powers, complicated multipolar nuclear interactions, and unstable conflict dynamics. As it does, understanding the compulsions and effects of these choices about nuclear strategy provides scholars and policy makers alike additional tools to better manage the proliferation problem and to address the likely crises and challenges ahead.

Bibliography

Abraham, Itty. The Making of the Indian Atomic Bomb: Science, Secrecy, and the Postcolonial State (New York: St Martin’s Press, 1998). Abu-­Lughod, Ibrahim. The Arab-­Israeli Confrontation of June 1967: An Arab Perspective (Evanston, IL: Northwestern University Press, 1970). Ahmed, Ali. “The Central Debate in India’s Civil-­Military Relations.” IDSA (New Delhi) Strategic Comments, July 6, 2009. Ahmed, Samina and David Cortwright, eds. Pakistan and the Bomb (South Bend, IN: University of Notre Dame Press, 1999). Ahmed, Samina. “Pakistan’s Nuclear Weapons Program,” International Security, vol. 23, no. 4 (Spring 1999). Alagappa, Muthiah, ed. The Long Shadow: Nuclear Weapons and Security in 21st Century Asia (Stanford, CA: Stanford University Press, 2008). Albright, David. “South Africa and the Affordable Bomb.” Bulletin of the Atomic Scientists, vol. 50, no. 4 (July/August 1994). Albright, David and Paul Brannan. “Pakistan Expanding Plutonium Separation Facility near Rawalpindi.” Institute for Science and International Security, May 19, 2009, available at http://www.isis-online.org/publications/southasia /PakistanExpanding`Newlabs.pdf. Albright, David and Mark Hibbs. “Pakistan’s Bomb: Out of the Closet” Bulletin of the Atomic Scientists, vol. 48, no. 6 (July/August 1992). Albright, David and Mark Hibbs. “South Africa: The ANC and the Atom Bomb.” Bulletin of the Atomic Scientists, vol. 49, no. 3 (April 1993). Albright, David and Tom Zamora. “South Africa Flirts with the NPT,” Bulletin of the Atomic Scientists, vol. 47, no. 1 (January/February 1991). Arens, Moshe. Broken Covenant: American Foreign Policy and the Crisis between the U.S. and Israel (New York: Simon & Schuster, 1995). Armoni, Ora. A Friend and Confidant [Hebrew] (Tel Aviv: Ha’Kibutz Ha’Meuhad, 2008). Bailey, Kathleen C., ed. Weapons of Mass Destruction: Costs versus Benefits (Delhi: Manohar Books, 1994). Bajpai, Kanti, P. R. Chari, Pervaiz Iqbal Cheema, Stephen Cohen, and Sumit Ganguly. Brasstacks and Beyond: Perception and Management of Crisis in South Asia (Delhi: Manohar 1995). 313

314

Bibliography

Baker, James A., III (with Thomas M. DeFrank). The Politics of Diplomacy: Revolution, War, and Peace, 1989–­1992 (New York: G.P. Putnam’s Sons, 1995). Barnaby, Frank. The Invisible Bomb: The Nuclear Arms Race in the Middle East (London: I.B. Tauris, 1989). Basrur, Rajesh. Minimum Deterrence and India’s Nuclear Security (Stanford, CA: Stanford University Press, 2006). Beardsley, Kyle and Victor Asal. “Proliferation and International Crisis Behavior.” Journal of Peace Research, vol. 44, no. 2 (2007). Beardsley, Kyle and Victor Asal. “Winning with the Bomb.” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009). Bell, Mark S. and Nicholas L. Miller. “Questioning the Effect of Nuclear Weapons on Conflict.” Journal of Conflict Resolution, forthcoming. Ben-­David, Alon. “Israel Looks to Acquire more German Submarines.” Jane’s Defence Weekly, November 30, 2005. Ben Meir, Yehuda. Civil-­Military Relations in Israel (New York: Columbia University Press, 1995). Bennett, D. Scott and Alan Stam. The Behavioral Origins of War (Ann Arbor: University of Michigan Press, 2004). Ben-­Zvi, Abraham. Decade of Transition: Eisenhower, Kennedy, and the Origins of the American-­Israeli Alliance (New York: Columbia University Press, 1998). Betts, Richard K. Nuclear Blackmail and Nuclear Blackmail (Washington, DC: Brookings Press, 1987). Bhaumik, Saba Naqvi. “Blast from a Far Past.” Outlook India, May 19, 2008. Bin, Li, Zhou Baogen, and Liu Zhiwei. “China Will Have to Respond.” Bulletin of the Atomic Scientists, vol. 57, no. 6 (November/December 2001). Bissell, Richard E. and Chester A. Crocker, eds. South Africa into the 1980s (Boulder, CO: Westview Press, 1979). Blechman, Barry M. and Douglas M. Hart. “The Political Utility of Nuclear Weapons: The 1973 Middle East Crisis.” International Security, vol. 7, no. 1 (Summer 1982). Bokhari, Farhan and Trefor Moss. “Pakistan Strikes Deal for Chinese Nuclear Reactors.” Jane’s Defence Weekly, October 21, 2008. Bolt, Paul and Albert S. Willner, eds. China’s Nuclear Future (Boulder, CO: Lynne Rienner, 2006). Brands, Hal and David Palkki. “Saddam, Israel, and the Bomb: Nuclear Alarmism Justified?” International Security, vol. 36, no. 1 (Summer 2011). Brecher, Michael. India and World Politics: Krishna Menon’s View of the World (New York: Praeger Books, 1968). Brecher, Michael. Decisions in Israel’s Foreign Policy (London: Oxford University Press, 1975). Brodie, Bernard. Strategy in the Missile Age (Princeton, NJ: Princeton University Press, 1959).

Bibliography 315 Bueno De Mesquita, Bruce and William Riker. “An Assessment of the Merits of Selective Nuclear Proliferation.” Journal of Conflict Resolution, vol. 26, no. 2 (June 1982). Bundy, McGeorge. “Existential Deterrence and Its Consequences.” In Douglas MacLean, ed., The Security Gamble: Deterrence Dilemmas in the Nuclear Age (Totowa, NJ: Rowman and Allanheld, 1984). Burgess, Stephen F. “South Africa’s Nuclear Weapons Policies.” Nonproliferation Review, vol. 13, no. 3 (November 2006). Burrows, Robert and Robert Windrem. Critical Mass: The Dangerous Race for Superweapons in a Fragmenting World (New York: Simon & Schuster, 1994). Buzan, Barry. An Introduction to Strategic Studies: Military Technology and International Relations (London: MacMillan Press, 1987). Carter, Ashton B., John D. Steinbruner, and Charles A. Zraket, eds. Managing Nuclear Operations (Washington, DC: Brookings Press, 1987). Chari, P. R., Pervaiz Iqbal Cheema, and Stephen Cohen. Perception, Politics and Security in South Asia: The Compound Crisis of 1990 (London: Routledge Curzon, 2003). Chari, P. R., Pervaiz Iqbal Cheema, and Stephen Cohen. Four Crises and a Peace Process: American Engagement in South Asia (Washington, DC: Brookings Institution Press, 2007). Chari, P. R., Sonika Gupta, and Arpit Rajan, eds. Nuclear Stability in Southern Asia (Delhi: Manohar, 2003). Chase, Michael S., Andrew S. Erickson, and Christopher Yeaw. “Chinese Theater and Strategic Force Modernization and its Implications for the United States.” Journal of Strategic Studies, vol. 32, no. 1 (February 2009). Cheema, Zafar Iqbal. “The Role of Nuclear Weapons in Pakistan’s Defense Strategy.” Islamabad Policy Research Institute Journal, vol. 4, no. 2 (Summer 2004). Chengappa, Raj. Weapons of Peace (New Delhi: Harper Collins, 2000). Chengappa, Raj and Saurabh Shukla. “Reining in the Rogue.” India Today, December 15, 2008. “Chinese Nuclear Forces 2006.” Bulletin of the Atomic Scientists, vol. 62, no. 3 (May/June 2006). Christensen, Thomas and Jack Snyder. “Chain Gangs and Passed Bucks: Predicting Alliance Patterns in Multipolarity.” International Organization, vol. 40 (1990). Cloughley, Brian. A History of the Pakistan Army (Oxford: Oxford University Press, 2000). Cohen, Avner. Israel and the Bomb (New York: Columbia University Press, 1998). Cohen, Avner. “Crossing the Threshold: The Untold Nuclear Dimension of the 1967 Arab-­Israeli War and Its Contemporary Lessons.” Arms Control Today, June 2007.

316

Bibliography

Cohen, Avner. The Worst Kept Secret: Israel’s Bargain with the Bomb (New York: Columbia University Press, 2010). Cohen, Avner and William Burr. “Israel Crosses the Threshold.” Bulletin of the Atomic Scientists, vol. 62, no. 3 (May/June 2006). Cohen, Avner and Terence McNamee. Why Do States Want Nuclear Weapons: The Cases of Israel and South Africa (Oslo: Norwegian Institute for Defence Studies, 2005). Cohen, Stephen P. The Pakistan Army (Berkeley: University of California Press, 1984). Cohen, Stephen P. The Indian Army: Its Contribution to the Development of a Nation (Delhi: Oxford University Press, 1990). Cohen, Stephen P. “Nuclear Weapons and Conflict in South Asia.” Paper Presented at Harvard/MIT Transnational Security Project, November 23, 1998. Available at http://www.brook.edu/views/cohens19981123.html. Cohen, Stephen P. The Idea of Pakistan (Washington, DC: Brookings Institution Press, 2004). Coker, Christopher. South Africa’s Security Dilemmas (Westport, CT: Praeger, 1987). Colby, Elbridge et al. The Israeli “Nuclear Alert” of 1973: Deterrence and Signaling in Crisis (Washington, DC: Center for Naval Analyses, April 30, 2013). Coll, Steve. Ghost Wars: The Secret History of the CIA, Afghanistan, and bin Laden from the Soviet Invasion to September 10, 2001 (New York: Penguin, 2004). Cordesman, Anthony. After the Storm: The Changing Military Balance in the Middle East (Boulder, CO: Westview Press, 1993). Corera, Gordon. Shopping for Bombs: Nuclear Proliferation, Global Insecurity, and the Rise and Fall of the AQ Khan Network (Oxford: Oxford University Press, 2006). Cotta-­Ramusino, P. and M. Martellini. Nuclear Safety, Nuclear Stability, and Nuclear Strategy in Pakistan: A Concise Report of a Visit by Landau Network-­Centro Volto (Como, Italy: Landau Network-­Centro Volto), January 14, 2002. Crile, George. Charlie Wilson’s War: The Extraordinary Story of How the Wildest Man in Congress and a Rogue CIA Agent Changed the History of Our Times (New York: Grove Press, 2003). Danchev, Alex and Dan Keohane, eds. International Perspectives on the Gulf Conflict, 1990–­1991 (London: Macmillan, 1993). “De Klerk Discloses Nuclear Capability to Parliament.” FBIS-­AFR-­93–­056, March 25, 1993. De Villiers, J.W., Roger Jardine, and Mitchell Reiss. “Why South Africa Gave Up the Bomb.” Foreign Affairs, vol. 72, no. 5 (November/December 1993).

Bibliography 317 Desch, Michael. Civilian Control of the Military: The Changing Security Environment (Baltimore: Johns Hopkins University Press, 1999). Ditmer, Lowell. “South Asia’s Security Dilemma,” Asian Survey, vol. 41, no. 6 (November/December 2001). Dossani, Rafiq and Henry S. Rowen. Prospects for Peace in South Asia (Stanford, CA: Stanford University Press, 2005). Dupuy, Trevor. Elusive Victory: The Arab-­Israeli Wars, 1947–­1974 (New York: Harper & Row, 1978). Durrani, Mahmud Ali. “Pakistan’s Strategic Thinking and the Role of Nuclear Weapons.” Cooperative Monitoring Center Occasional Paper 37 (Albuquerque, NM: Sandia National Laboratories, 2004). Eden, Lynn and Steven Miller, eds. Nuclear Arguments: Understanding the Strategic Nuclear Arms and Arms Control Debates (Ithaca, NY: Cornell University Press, 1989). Eisenberg, Laura Zittrain. “Passive Belligerency: Israel and the 1991 Gulf War.” Journal of Strategic Studies, vol. 15, no. 3 (September 1992). Elman, Colin. “Horses for Courses: Why Not Neorealist Theories of Foreign Policy.” Security Studies, vol. 6, no. 1 (1997). Evron, Yair. Israel’s Nuclear Dilemma (Ithaca, NY: Cornell University Press, 1994). Fain, W. Taylor. “John F. Kennedy and Harold Macmillan: Managing the ‘Special Relationship’ in the Persian Gulf Region, 1961–­1963.” Middle Eastern Studies 38, no. 4 (October 2002). Fearon, James D. “Signaling Foreign Policy Interest: Tying Hands versus Sinking Costs.” Journal of Conflict Resolution, vol. 41, no. 1 (February 1997). Fearon, James D. “Selection Effects and Deterrence.” International Interactions, vol. 28, no. 1 (January 2002). Feaver, Peter D. Guarding the Guardians: Civilian Control of Nuclear Weapons in the United States (Ithaca, NY: Cornell University Press, 1992). Feaver, Peter. “Command and Control in Emerging Nuclear States.” International Security, vol. 17, no. 3 (Winter 1992–­93). Feaver, Peter D. Armed Servants: Agency, Oversight, and Civil-­Military Relations (Cambridge, MA: Harvard University Press, 2003). Feaver, Peter, Scott D. Sagan, and David Karl. “Correspondence: Proliferation Pessimism and Emerging Nuclear Powers.” International Security, vol. 22, no. 2 (Autumn 1997). Feldman, Shai. Israeli Nuclear Deterrence: A Strategy for the 1980s (New York: Columbia University Press, 1983). Feldman, Shai. “Israeli Deterrence and the Gulf War.” In J. Alpher, ed., War in the Gulf: Implications for Israel (Jerusalem: Jerusalem Post Press, 1992). Finer, Samuel. The Man on Horseback: The Role of the Military in Politics (New York: Praeger, 1962).

318

Bibliography

Fravel, M. Taylor. Strong Borders, Secure Nation: Cooperation and Conflict in China’s Territorial Disputes (Princeton, NJ: Princeton University Press, 2008). Fravel, M. Taylor and Evan S. Medeiros. “Living With Vulnerability: Explaining the Evolution of China’s Nuclear Strategy and Force Structure.” Unpublished Working Paper, August 2009. Fravel, M. Taylor and Evan S. Medeiros. “China’s Search for Assured Retaliation: The Evolution of Chinese Nuclear Strategy and Force Structure.” International Security, vol. 35, no. 2 (Fall 2010). Freedman, Lawrence. The Evolution of Nuclear Strategy, 2nd ed. (London: Macmillan, 1989). Friedberg, Aaron. “Why Didn’t the US Become a Garrison State?” International Security, vol. 16, no. 4 (Spring 1992). Fuhrmann, Matthew and Sarah Kreps. “Targeting Nuclear Programs in War and Peace: A Quantitative Empirical Analysis.” Journal of Conflict Resolution, 54, no. 6 (2010). Gaddis, John Lewis. The Long Peace: Inquiries into the History of the Cold War (New York: Oxford University Press, 1987). Gallois, Pierre. “French Defense Planning—­The Future in the Past.” International Security, vol. 1, no. 2 (Autumn 1976). Ganguly, Sumit. “The Pathway to Pokhran II: The Prospects and Sources of New Delhi’s Nuclear Weapons Program.” International Security, vol. 23, no. 4 (Spring 1999). Ganguly, Sumit. “Nuclear Stability in South Asia.” International Security, vol. 33, no. 2 (Fall 2008). Ganguly, Sumit and Devin Hagerty. Fearful Symmetry: India-­Pakistan Crises in the Shadow of Nuclear Weapons (Seattle: University of Washington Press, 2005). Ganguly, Sumit and S. Paul Kapur. Nuclear Proliferation in South Asia: Crisis Behavior and the Bomb (New York: Routledge, 2009). Gartzke, Eric and Dong-­Joon Jo. “Determinants of Nuclear Proliferation.” Journal of Conflict Resolution, vol. 51, no. 1 (2007). Gavin, Francis J. Gold, Dollars, and Power: The Politics of International Monetary Relations 1958–­1971 (Chapel Hill: University of North Carolina Press, 2007). Geller, Daniel. “Nuclear Weapons, Deterrence, and Crisis Escalation.” Journal of Conflict Resolution, vol. 34, no. 2 (June 1990). George, Alexander L. “The Role of the Congruence Method for Case Study Research.” Working Paper presented at MacArthur Program on Case Studies, Georgetown University, March 18–­22, 1997. Available at http://www .ciaonet.org/wps/gea01/. Ginor, Isabella and Gideon Ramez. Foxbats over Dimona: The Soviets’ Nuclear Gamble in the Six-­Day War (New Haven, CT: Yale University Press, 2007).

Bibliography 319 Gjelstad, Jorn and Olav Njolstad, eds. Nuclear Rivalry and International Order (London: Sage Publications, 1996). Glaser, Charles L. Analyzing Strategic Nuclear Policy (Princeton, NJ: Princeton University Press, 1990). Glassman, Jon. Arms for the Arabs: The Soviet Union and War in the Middle East (Baltimore: Johns Hopkins University Press, 1975). Goldstein, Avery. Deterrence and Security in the 21st Century: China, Britain, France, and the Enduring Legacy of the Nuclear Revolution (Stanford, CA: Stanford University Press, 2000). Goldstein, Lyle J. “Return to Zhenbao Island: Who Started Shooting and Why It Matters.” China Quarterly, no. 168 (December 2001). Gordon, Michael and General Bernard Trainor. The Generals’ War: The Inside Story of the Conflict in the Gulf (New York: Little, Brown, 1995). Gourevitch, Peter A. “The Second-­Image Reversed.” International Organization, vol. 32, no. 4 (Autumn 1978). Government of India. “Draft Report of National Security Advisory Board on Indian Nuclear Doctrine,” August 17, 1999. Available at http://www .indianembassy.org/policy/CTBT/nuclear_doctrine_aug_17_1999.html. Gray, Colin. Nuclear Strategy and National Style (Lanham, MD: Hamilton Press, 1986). Gray, Colin S. Modern Strategy (Oxford: Oxford University Press, 1999). Gregory, Shaun R. Nuclear Command and Control in NATO: Nuclear Weapons Operations and the Strategy of Flexible Response (New York: St. Martin’s Press, 1996). Hagerty, Devin. “Nuclear Deterrence in South Asia: The 1990 Indo-­Pakistani Crisis.” International Security, vol. 20, no. 3 (Winter 1995/1996). Haider, Zeeshan. “Pakistan’s Nuclear Command Stays Unchanged.” Reuters, April 8, 2008. Hamann, Hilton. Days of the Generals: The Untold Story of Africa’s ApartheidEra Military Generals (Cape Town: Zebra Press, 2001). Harris, Verne, Sello Hatang, and Peter Liberman. “Unveiling South Africa’s Nu­clear Past.” Journal of Southern African Studies, vol. 30, no. 3 (September 2004). Hersh, Seymour. The Samson Option: Israel’s Nuclear Arsenal and American Foreign Policy (New York: Random House, 1991). Herzog, Chaim. The Arab-­Israeli Wars (New York: Vintage Books, 2005). Hoon, P. N. Unmasking Secrets of Turbulence (New Delhi: Manas Publications, 1999). Hopkins, John C. and Weixing Hu, eds. Strategic Views from the Second Tier: The Nuclear Weapons Policies of France, Britain, and China (New Brunswick, NJ: Transaction Publishers, 1995). Horowitz, Michael. The Diffusion of Military Power: Causes and Consequences for International Politics. (Princeton, NJ: Princeton University Press, 2010).

320

Bibliography

Horowitz, Michael. “The Spread of Nuclear Weapons and International Conflict: Does Experience Matter?” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009). Hough, Harold. “Could Israel’s Nuclear Assets Survive a First Strike?” Jane’s Intelligence Review, vol. 9, no. 9 (September 1997). Hoyt, Timothy D. “Pakistani Nuclear Doctrine and the Dangers of Strategic Myopia.” Asian Survey, vol. 41, no. 6 (November/December 2001). Huntington, Samuel P. Soldier and the State (Cambridge, MA: Harvard University Press, 1957). Huth, Paul. “Extended Deterrence and the Outbreak of War.” American Political Science Review, vol. 82, no. 2 (June 1988). Huth, Paul and Bruce Russett. “Deterrence Failure and Crisis Escalation.” International Studies Quarterly, vol. 32, no. 1 (1988). Huth, Paul and Bruce Russett. “What Makes Deterrence Work? Cases from 1900–­1980.” World Politics, vol. 36, no. 4 (July 1984). Hymans, Jacques E. C. The Psychology of Nuclear Proliferation: Identity, Emotions, and Foreign Policy (Cambridge: Cambridge University Press, 2006). Hymans, Jacques E .C. Achieving Nuclear Ambitions: Scientists, Politicians, and Proliferation (New York: Cambridge University Press, 2012). Information Office of the State Council of the People’s Republic of China. China’s National Defense in 2008. Available at http://english.gov.cn /official/2009‐01/20/content_1210227.htm. Inside the Red Star: The Memoirs of Marshal Nie Rongzhen, translated by Zhong Renyi (Beijing: New World Press, 1988). International Crisis Behavior Dataset, version 9.0, Center for International Development and Conflict Management, University of Maryland, 2009. International Institute for Strategic Studies. Nuclear Black Markets: Pakistan, AQ Khan, and the Rise of Proliferation Networks (London: IISS, 2007). International Institute for Strategic Studies. The Military Balance 1964–­7 (London: IISS 1964–­67). International Institute for Strategic Studies. The Military Balance (London: IISS, 1978). International Institute for Strategic Studies. The Military Balance (London: IISS, 1985). Israeli, Raphael. The Public Diary of President Sadat: Part One, The Road to War (Leiden: E. J. Brill, 1978). Jalal, Ayesha. The State of Martial Rule: The Origins of Pakistan’s Political Economy of Defence (Cambridge: Cambridge University Press, 1990). Jalal, Ayesha. Authoritarianism and Democracy in South Asia: A Comparative and Historical Analysis (Cambridge: Cambridge University Press, 1995). Janowitz, Morris. The Professional Soldier (Glencoe, IL: Free Press, 1960). Jervis, Robert. The Illogic of American Nuclear Strategy (Ithaca, NY: Cornell University Press, 1984).

Bibliography 321 Jervis, Robert. “The Political Effects of Nuclear Weapons: A Comment.” International Security, vol. 13, no. 2 (Autumn 1988). Jervis, Robert. The Meaning of the Nuclear Revolution (Ithaca, NY: Cornell University Press, 1989). Joeck, Neil. “Maintaining Nuclear Stability in South Asia.” International Institute for Strategic Studies, Adelphi Paper 312 (Oxford: Oxford University Press, 1997). Johnston, Alastair Iain. “Thinking about Strategic Culture.” International Security, vol. 19, no. 4 (Spring 1995). Johnston, Alastair Iain. Cultural Realism: Strategic Culture and Grand Strategy in Chinese History (Princeton, NJ: Princeton University Press, 1995). Johnston, Alastair Iain. “China’s New ‘Old Thinking’: The Concept of Limited Deterrence.” International Security, vol. 20, no. 3 (Winter 1995–­96). Johnston, Alastair Iain. “Prospects for Chinese Nuclear Force Modernization: Limited Deterrence versus Multilateral Arms Control,” China Quarterly, no. 146 (June 1996). Joint Statement by Department of Atomic Energy and Defence Research and Development Organisation.” New Delhi, May 17, 1998. Available at http:// www.fas.org/news/india/1998/05/980500-drdo.htm. Kamphausen, Roy and Andrew Scobell, eds. Right Sizing the People’s Liberation Army: Exploring the Contours of China’s Military (Carlisle, PA: Strategic Studies Institute, U.S. Army War College), September 2007. Kapur, S. Paul. “India and Pakistan’s Unstable Peace: Why Nuclear South Asia is not Like Cold War Europe.” International Security, vol. 30, no. 2 (Fall 2005). Kapur, S. Paul. “Ten Years of Instability in South Asia.” International Security, vol. 33, no. 2 (Fall 2008). Kapur, S. Paul. Dangerous Deterrent: Nuclear Weapons Proliferation and Conflict in South Asia (Stanford, CA: Stanford University Press, 2007). Kargil Review Committee. From Surprise to Reckoning: The Kargil Review Committee Report (New Delhi: Sage Publications, 2000). Karnad, Bharat. Nuclear Weapons and Indian Security: The Realist Foundations of Strategy (New Delhi: Macmillan, 2002). Karnad, Bharat. India’s Nuclear Policy (Westport, CT: Praeger, 2008). Karpin, Michael. The Bomb in the Basement (New York: Simon & Schuster, 2006). Karpin, Michael. The Bomb in the Basement: How Israel Went Nuclear and What that Means for the World (New York: Simon & Schuster, 2006). Katzenstein, Peter. Small States in World Markets (Ithaca, NY: Cornell University Press, 1985). Katzenstein, Peter, ed. The Culture of National Security: Norms and Identity in World Politics (New York: Columbia University Press, 1996). Kennedy, Andrew B. “India’s Nuclear Odyssey: Diplomatic Disenchantment, Leaky Umbrellas, and the Bomb.” Paper prepared for the 2009 Annual Meeting of the International Studies Association, February 15–­18, 2009.

322

Bibliography

Khan, Feroz Hassan. “Challenges to Nuclear Stability in South Asia.” Nonproliferation Review (Spring 2003). Khan, Feroz Hassan. “Nuclear Command and Control in South Asia during Peace, Crisis, and War.” Contemporary South Asia, vol. 14, no. 2 (June 2005). Khan, Feroz Hassan. “Pakistan as a Nuclear State.” In Maleeha Lodhi, ed., Pakistan: Beyond the ‘Crisis State’ (New York: Columbia University Press, 2011). Khan, Feroz Hassan. Eating Grass: The Making of the Pakistani Bomb (Stanford, CA: Stanford University Press, 2012). Khanna, V. N. India’s Nuclear Doctrine (New Delhi: Samskriti, 2000). Kier, Elizabeth. “Culture and Military Doctrine: France Between the Wars.” International Security, vol. 19, no. 4 (Spring 1995). Kile, Shannon N., Vitaly Fedchenko, and Hans M. Kristensen. “Israeli Nuclear Forces, January 2008.” SIPRI Yearbook 2008 (Stockholm: Almquist & Wiksell, 2009). Knorr, Klaus, ed. Historical Dimensions of National Security Problems (Lawrence: University Press of Kansas, 1976). Kohl, Wilfred L. French Nuclear Diplomacy (Princeton, NJ: Princeton University Press, 1971). Kohli, Atul, ed. India’s Democracy: An Analysis of Changing State-­Society Relations (Princeton, NJ: Princeton University Press, 1988). Koithara, Verghese. Managing India’s Nuclear Forces (Washington, DC: Brookings Press, 2012). Korn, David. Stalemate: The War of Attrition and Great Power Diplomacy in the Middle East, 1969–­1970 (Boulder, CO: Westview, 1992). Krepon, Michael and Mishi Faruqee, eds. Conflict Prevention and Confidence Building Measures in South Asia: The 1990 Crisis. Henry L. Stimson Center Occasional Paper 17, April 1994. Krepon, Michael, Rodney W. Jones, and Ziad Haider, eds. Escalation Control and the Nuclear Option in South Asia (Washington, DC: Henry L. Stimson Center, 2004). Kristensen, Hans M. and Robert S. Norris. “Nuclear Notebook: French Nuclear Forces, 2000,” Bulletin of the Atomic Scientists, vol. 64, no. 4 (September/ October 2008). Kristensen, Hans, Robert S. Norris, and Matthew G. McKinzie. Chinese Nuclear Forces and US Nuclear War Planning (Washington, DC: Federation of American Scientists & National Resources Defense Council, November 2006). Kroenig, Matthew H. “Time to Attack Iran: Why a Strike Is the Least Bad Option,” Foreign Affairs, January/February 2012. Kuisong, Yang. “The Sino-­Soviet Border Clash of 1969: From Zhenbao Island to Sino-­American Rapprochement.” Cold War History, vol. 1, no. 1 (August 2000).

Bibliography 323 Ladwig III, Walter C. “A Cold Start for Hot Wars? India’s New Limited War Doctrine.” International Security, vol. 32, no. 3 (Winter 2007/2008). Lavoy, Peter R. “Pakistan’s Nuclear Posture: Security and Survivability.” Presentation to the Conference on Pakistan’s Nuclear Future, Nonproliferation Education Center, April 2006. Lavoy, Peter, Scott D. Sagan, and James J. Wirtz, eds. Planning the Unthinkable: How New Powers Will Use Nuclear, Biological, and Chemical Weapons (Ithaca, NY: Cornell University Press, 2000). Leng, Russell J. Bargaining and Learning in Recurring Crises: The Soviet-­ American, Egyptian-­Israeli, and Indo-­Pakistani Rivalries (Ann Arbor: University of Michigan Press, 2000). Levy, Adrian and Catherine Scott-­Clark. Deception: Pakistan, the United States, and the Secret Trade in Nuclear Weapons (New York: Walker & Company, 2007). Lewis, Jeffrey. The Minimum Means of Reprisal: China’s Search for Security in the Nuclear Age (Cambridge, MA: MIT Press, 2007). Lewis, John and Hua Di. “China’s Ballistic Missile Programs: Technology, Strategies, Goals.” International Security, vol. 17, no. 2 (Fall 1992). Lewis, John and Xue Litai. “Strategic Weapons and Chinese Power: The Formative Years.” China Quarterly, vol. 112 (December 1987). Lewis, John and Xue Litai. China Builds the Bomb (Stanford, CA: Stanford University Press, 1988). Lewis, John and Xue Litai. China’s Strategic Seapower: The Politics of Force Modernization in the Nuclear Age (Stanford, CA: Stanford University Press, 1994). Lewis, John and Xue Litai. Imagined Enemies: China Prepares for Uncertain War. (Stanford, CA: Stanford University Press, 2006). Liberman, Peter. “The Rise and Fall of the South African Bomb.” International Security, vol. 26, no. 2 (Autumn 2001). Liberman, Peter. “Israel and the South African Bomb.” Nonproliferation Review, vol. 11, no. 2 (Summer 2004). Lieber, Keir and Daryl Press. “How Much Is Enough? Nuclear Deterrence Then and Now.” Working Paper presented at Wisconsin International Relations Colloquium, University of Wisconsin–­Madison, February 3, 2009. Lieberman, Evan S. “Nested Analysis as a Mixed-­Method Strategy for Comparative Research.” American Political Science Review, vol. 99, no. 3 (August 2005). Lilley, James R. and David Shambaugh, eds. China’s Military Faces the Future (New York: ME Sharpe, 1999). Lin, Chong-­Pin. China’s Nuclear Weapons Strategy: Tradition within Evolution (Lexington, MA: Lexington Books, 1988). Lodi, Sardar F. S. “Pakistan’s Nuclear Doctrine.” Defence Journal (Pakistan), vol. 3, no. 4 (April 1999).

324

Bibliography

Luongo, Kenneth and Brig. Gen. (Ret.) Naeem Salik. “Building Confidence in Pakistan’s Nuclear Security.” Arms Control Today, December 2007. Available at http://www.armscontrol.org/act/2007_12/Luongo. Luthi, Lorenz M. The Sino-­Soviet Split: Cold War in the Communist World (Princeton, NJ: Princeton University Press, 2008). MacFarquhar, Roderick and John K. Fairbank, eds. The Cambridge History of China, Volume 15 (Cambridge: Cambridge University Press, 1992). Malik, Ved Prakash. Kargil: From Surprise to Victory (Delhi: Harper Collins, 2006). Maoz, Zeev. “The Mixed Blessing of Israel’s Nuclear Policy.” International Security, vol. 28, no. 2 (Fall 2003). Mearsheimer, John J. “Maneuver, Mobile Defense, and the NATO Central Front.” International Security, vol. 6, no. 3 (Winter 1981/1982). Mearsheimer, John J. “Why the Soviets Can’t Win Quickly in Central Europe.” International Security, vol. 7, no. 1 (Summer 1982). Mearsheimer, John J. Conventional Deterrence (Ithaca, NY: Cornell University Press, 1983). Mearsheimer, John J. “Nuclear Weapons and Deterrence in Europe.” International Security, vol. 9, no. 3 (Winter 1984/1985). Mearsheimer, John J. “Assessing the Conventional Balance: The 3:1 Rule and Its Critics.” International Security, vol. 13, no. 4 (Spring 1989). Mearsheimer, John J. The Tragedy of Great Power Politics (New York: Norton, 2001). Mendl, Wolf. Deterrence and Persuasion: French Nuclear Armament in the Context of National Policy, 1945–­1969 (New York: Praeger, 1970). Menon, Raja. “Just One Shark in the Deep Blue Ocean.” Outlook, August 10, 2009. Montgomery, Alexander H. and Scott D. Sagan. “The Perils of Predicting Proliferation.” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009). Mulvenon, James and Andrew N. D. Yang. The People’s Liberation Army as Organization (Santa Monica: RAND Conference Proceedings, 2002). Mulvenon, James and David Finkelstein, eds. China’s Revolution in Doctrinal Affairs: Emerging Trends in the Operational Art of the People’s Liberation Army (Alexandria, VA: CNA Corporation, December 2005). Narang, Vipin. “Posturing for Peace? Pakistan’s Nuclear Postures and South Asian Stability.” International Security, vol. 34, no. 3 (Winter 2009/2010). Narang, Vipin. “What Does It Take to Deter? Regional Power Nuclear Postures and International Conflict,” Journal of Conflict Resolution, vol. 57, no. 3 (June 2013). Narang, Vipin. “Five Myths about India’s Nuclear Posture.” Washington Quarterly, vol. 36, no. 3 (Summer 2013). National Intelligence Council. Foreign Missile Developments and the Ballistic Missile Threat Through 2015. September 1999.

Bibliography 325 National Resource and Defense Council Nuclear Notebook. “Israeli Nuclear Forces 2002.” Bulletin of the Atomic Scientists (September/October 2002). NATO and the Warsaw Pact: Force Comparisons (Brussels: NATO Information Service, 1984). Nawaz, Shuja. Crossed Swords: Pakistan, Its Army, and the Wars Within (Karachi: Oxford University Press, 2008). Nelson, Harvey. The Chinese Military System: An Organizational Study of the Chinese People’s Liberation Army (Boulder, CO: Westview Press, 1981). “New Information on South Africa’s Nuclear Program and South African-­Israeli Nuclear and Military Cooperation.” Declassified Memo, Directorate of United States Intelligence, March 30, 1983. Noorani, A. G. “India’s Quest for a Nuclear Guarantee.” Asian Survey, vol. 7, no. 7 (July 1967). Norris, Robert and William M. Arkin. “Israeli Nuclear Forces, January 2000.” SIPRI Yearbook 2000 (Stockholm: Almquist & Wiksell, 2000). Norris, Robert S., Andrew S. Burrows, and Richard W. Fieldhouse. Nuclear Weapons Databook, Volume V: British, French, and Chinese Nuclear Forces (Boulder, CO: Westview Press, 1994). Norris, Robert and Hans Kristensen. “India’s Nuclear Forces, 2008.” Bulletin of the Atomic Scientists (November/December 2008). Norris, Robert and Hans Kristensen. “India’s Nuclear Forces, 2010.” Bulletin of the Atomic Scientists, vol. 66, no. 5 (2010). Norris, Robert and Hans Kristensen. “India’s Nuclear Forces, 2012.” Bulletin of the Atomic Scientists, vol. 68, no. 4 (2012). “Nuclear Notebook: Estimated Soviet Nuclear Stockpile.” Bulletin of the Atomic Scientists, vol. 45, no. 6 (July/August 1989). “Nuclear Notebook: Estimated Soviet Nuclear Stockpile.” Bulletin of the Atomic Scientists, vol. 47, no. 6 (July 1991). Office of Naval Intelligence. China’s Navy 2007; and Military Power of the People’s Republic of China, 2009. Office of the Secretary of Defense. Annual Report to Congress: Military Power of the People’s Republic of China, 2009. “Olmert’s Nuclear Slip-­up Sparks Outrage in Israel.” Times Online (UK), December 12, 2006. Oren, Michael B. Six Days of War: June 1967 and the Making of the Modern Middle East (Oxford: Oxford University Press, 2002). “Pakistan’s Nuclear Forces 2007.” Bulletin of the Atomic Scientists, vol. 63, no. 3 (May/June 2007). Pant, Harsh V. “India’s Nuclear Doctrine and Command Structure: Implications for India and the World.” Comparative Strategy, vol. 24, no. 3 (2005). Pant, Harsh V. “India’s Nuclear Doctrine and Command Structure: Implications for Civil-­Military Relations in India.” Armed Forces & Society, vol. 33, no. 2 (January 2007).

326

Bibliography

Parker, Richard, ed. The October War: A Retrospective (Gainesville: University of Florida Press, 2001). Pattanaik, Smruti S. “Pakistan’s Nuclear Strategy.” Strategic Analysis, vol. 27, no. 1 (January–­March 2003). Paul, T. V. Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge: Cambridge University Press, 2004). Paul, T. V. Power versus Prudence: Why Nations Forego Nuclear Weapons (Montreal: McGill-­Queens University Press, 2000). Paul, T. V., ed. The India-­Pakistan Conflict: An Enduring Rivalry (Cambridge: Cambridge University Press, 2005). Perkovich, George. India’s Nuclear Bomb (Berkeley: University of California Press, 1999). Pollack, Kenneth. Arabs at War: Military Effectiveness 1948–­1991 (Lincoln: University of Nebraska Press, 2002). Posen, Barry R. “Is NATO Decisively Outnumbered?” International Security, vol. 12, no. 4 (Spring 1988). Posen, Barry R. Sources of Military Doctrine: France, Britain, and Germany Between the World Wars (Ithaca, NY: Cornell University Press, 1986). Posen, Barry R. Inadvertent Escalation: Conventional War and Nuclear Risks (Ithaca, NY: Cornell University Press, 1992). Prospects for Further Proliferation of Nuclear Weapons. A Special National Intelligence Estimate, Director of United States Central Intelligence Agency, August 23, 1974. Powell, Robert. “The Theoretical Foundations of Strategic Nuclear Deterrence.” Political Science Quarterly, vol. 100, no. 1 (Spring 1985). Purkitt, Helen E. and Stephen F. Burgess. South Africa’s Weapons of Mass Destruction (Bloomington: University of Indiana Press, 2005). Purkitt, Helen E. and Stephen F. Burgess. “South Africa’s Nuclear Strategy.” In Toshi Yoshihara and James R. Holmes, eds. Strategy in the Second Nuclear Age: Power, Ambition, and the Ultimate Weapon (Washington, DC: Georgetown University Press, 2012). Quandt, William B. “How Far will Israel Go?” Washington Post Book World, November 20, 1991. Ra’anan, Uri, Robert Pfaltzgraff, and Geoffrey Kemp, eds. Arms Transfers to the Third World: The Military Buildup in Less Industrial Countries (Boulder, CO: Westview Press, 1978). Rabasa, Angel, Robert D. Blackwill, Peter Chalk, Kim Cragin, C. Christine Fair, Brian A. Jackson, Brian Michael Jenkins, Seth G. Jones, Nathaniel Shestak, and Ashley J. Tellis. The Lessons of Mumbai (Santa Monica, CA: RAND Publications, 2009). Raghavan, Srinath. “Civil-­Military Relations in India: The China Crisis and After.” Journal of Strategic Studies, vol. 32, no. 1 (February 2009). Rauchhaus, Robert. “Evaluating the Nuclear Peace Hypothesis: A Quantitative Approach.” Journal of Conflict Resolution, vol. 53, no. 2 (April 2009).

Bibliography 327 Reed, Thomas C. and Danny B. Stillman. The Nuclear Express: A Political History of the Bomb and Its Proliferation (Minneapolis, MN: Zenith Press, 2009). Reiss, Mitchell. Bridled Ambition: Why Countries Constrain their Nuclear Capabilities (Baltimore: Johns Hopkins University Press, 1995). Richelson, Jeffrey T. Spying on the Bomb: American Nuclear Intelligence from Nazi Germany to Iran and North Korea (New York: Norton, 2007). Riedel, Bruce. American Diplomacy and 1999 Kargil Summit at Blair House (Philadelphia: University of Pennsylvania Center for Advanced Study of India, 2002). Riedel, Bruce. Deadly Embrace: Pakistan, America, and the Future of the Global Jihad (Washington, DC: Brookings Press, 2011). Rizvi, Hasan Askari. Military, State, and Society in Pakistan (London: St. Martin’s Press, 2000). Rosato, Sebastian. “Europe’s Troubles: Power Politics and the State of the European Project.” International Security, vol. 35, no. 4 (Spring 2011). Rose, Gideon. “Neoclassical Realism and Theories of Foreign Policy.” World Politics, vol. 51, no. 1 (1998). Rosen, Stephen P. Societies and Military Power: India and Its Armies (Ithaca, NY: Cornell University Press, 1996). Rostow, Walt W. “The Bases for Direct Action against Chinese Nuclear Facilities.” Department of State Policy Planning Memo, April 22, 1964 (Declassified). Available at http://www.gwu.edu/~nsarchiv/nukevault/special/doc09 .pdf. Rotberg, Robert I., Henry S. Bienen, Robert Legvold, and Gavin G. Maasdorp, eds. South Africa and Its Neighbors: Regional Security and Self-­Interest (Lexington, MA: Lexington Books, 1985). Rynning, Sten. Changing Military Doctrine: Presidents and Military Power in Fifth Republic France (Westport, CT: Praeger, 2002). Sadat, Anwar. In Search of Identity (London: Collins, 1978). Safran, Nadav. From War to War (New York: Pegasus, 1969). Sagan, Scott D. “Nuclear Alerts and Crisis Management.” International Security, vol. 9, no. 4 (Spring 1985). Sagan, Scott D. Moving Targets: Nuclear Strategy and National Security (Prince­ ton, NJ: Princeton University Press, 1989). Sagan, Scott D. The Limits of Safety: Organizations, Accidents, and Nuclear Weapons (Princeton, NJ: Princeton University Press, 1993). Sagan, Scott D. “Why Do States Build Nuclear Weapons? Three Models in Search of the Bomb.” International Security, vol. 21, no. 3 (Winter 1996–­97). Sagan, Scott D. “The Commitment Trap: Why the United States Should Not Use Nuclear Threats to Deter Biological and Chemical Weapons Attacks.” International Security, vol. 24, no. 4 (Spring 2000). Sagan, Scott D. “The Perils of Proliferation in South Asia.” Asian Survey, vol. 41, no. 6 (November/December 2001).

328

Bibliography

Sagan, Scott D., ed. Inside Nuclear South Asia (Stanford, CA: Stanford University Press, 2009). Samanta, Pranab Dhal. “26/11: How India Debated a War with Pakistan that November.” Indian Express, November 26, 2010. SarDesai, D. R. and Raju G. C. Thomas, eds. Nuclear India in the 21st Century (New York: Palgrave, 2002). Sarkozy, Nicolas. “Speech: Presentation of Le Terrible in Cherbourg.” March 21, 2008. Available at http://www.carnegieendowment.org/publications/index .cfm?fa=view&id=20001&prog=zgp&proj=znpp. Sarkozy, Nicolas. “Speech on Defense and National Security.” June 17, 2008. Available at http://www.ambafrance-ca.org/spip.php?article2269. Schelling, Thomas. The Strategy of Conflict (Cambridge, MA: Harvard University Press, 1960). Schelling, Thomas. Arms and Influence (New Haven, CT: Yale University Press, 1966). Schram, Martin. Avoiding Armageddon: Our Future. Our Choice (New York: Basic Books, 2003). Schweller, Randall. “New Realist Research on Alliances: Refining, not Refuting, Waltz’s Balancing Proposition.” American Political Science Review, vol. 91, no. 4 (December 1997). Schweller, Randall. Deadly Imbalances: Tripolarity and Hitler’s Strategy of World Conquest (New York: Columbia University Press, 1998). Seale, Patrick. Asad: The Struggle for the Middle East (London: I.B. Tauris, 1988). Segev, Tom. Transl.Jessica Cohen. 1967: Israel, the War, and the Year that Transformed the Middle East (New York: Metropolitan Books, 2007). Shahi, Aga, Zulfiqar Khan, and Abdul Sattar. “Securing Nuclear Peace.” News International, October 5, 1999. Reprinted in P. R. Chari, Sonika Gupta, and Arpit Rajan, eds., Nuclear Stability in Southern Asia (Delhi: Manohar Books, 2003). Shamir, Yitzhak. Summing Up: An Autobiography (New York: Little, Brown, 1994). Sheetz, Mark. “Exit Strategies: American Grand Designs for Postwar European Security.” Security Studies, vol. 8, no. 4 (Summer 1999). Shlaim, Avi. The Iron Wall: Israel and the Arab World (New York: Norton, 2000). Siddiqa, Ayesha. Military Inc: Inside Pakistan’s Military Economy (London: Pluto, 2007). Signorino, Curtis and Ahmer Tarar. “A Unified Theory and Test of Extended Immediate Deterrence.” American Journal of Political Science, vol. 50, no. 3 (July 2006). Singh, Jasjit, ed. Nuclear India (New Delhi: IDSA, 1998). Singh, Jaswant. “Against Nuclear Apartheid.” Foreign Affairs, vol. 77, no. 5 (September/October 1998).

Bibliography 329 Singh, Sonali and Christopher Way. “The Correlates of Nuclear Proliferation: A Quantitative Test.” Journal of Conflict Resolution, vol. 48, no. 6 (2004). SIPRI. Arms Trade Register: Arms Transfers to the Third World (Cambridge, MA: MIT Press, 1975). Smith, Alastair. “To Intervene or Not to Intervene: A Biased Decision.” Journal of Conflict Resolution, vol. 40, no.1 (March 1996). Smith, Merritt Roe and Leo Marx, eds. Does Technology Drive History?: The Dilemma of Technological Determinism (Cambridge, MA: MIT Press, 1994). Snyder, Glenn. “The Balance of Power and the Balance of Terror.” In Paul Seabury, ed., The Balance of Power (San Francisco: Chandler Books, 1965). Snyder, Jack. Soviet Strategic Culture: Implications for Limited Nuclear Operations (Santa Monica, CA: RAND Reports, September 1977). Snyder, Jack and Thomas Christensen. “Chain Gangs and Passed Bucks: Predicting Alliance Patterns in Multipolarity.” International Organization, vol. 44, no. 2 (Spring 1990). Sober, Sidney. Acting Chairman NSC Interdepartmental Group for the Near East and South Asia. “Indian Nuclear Development.” Declassified National Security Study Memorandum 156, May 30, 1974. Sokolski, Henry D., ed. Pakistan’s Nuclear Future: Worries Beyond War (Car­ lisle, PA: Strategic Studies Institute, 2008). Solingen, Etel. Nuclear Logics: Contrasting Paths in East Asia and the Middle East (Princeton, NJ: Princeton University Press, 2007). Sood, V. K. and Pravin Sawhney. Operation Parakram: The War Unfinished (Delhi: Sage Publications, 2003). Speeches and Interviews by Anwar el Sadat, January–­December 1973 (Cairo: Ministry of Information, 1975). Sridharan, E., ed. The India-­Pakistan Nuclear Relationship (Delhi: Routledge, 2007). Stein, Peter and Peter Feaver. Assuring Control of Nuclear Weapons: The Evolution of Permissive Action Links. CSIA Occasional Paper No. 2 (Cambridge, MA: CSIA Publications, 1987). Steyn, Hannes, Richardt van der Walt, and Jan van Loggerenberg. Nuclear Armament and Disarmament: South Africa’s Nuclear Experience (Lincoln, NE: iUniverse, 2007). Stokes, Mark. China’s Nuclear Warhead Storage and Handling System. Project 2049 Institute Monograph, March 12, 2010. Stumpf, Waldo. “South Africa’s Nuclear Weapons Program: From Deterrence to Dismantlement.” Arms Control Today, vol. 25 (December 1995/January 1996). Subrahmanyam, K. “Narasimha Rao and the Bomb.” Strategic Analysis, vol. 28, no. 4 (2004). Subrahmanyam, K. “No First Use: An Indian View.” Survival, vol. 51, no. 5 (October/November 2009).

330

Bibliography

Swami, Praveen. “Beating the Retreat.” Frontline Magazine, vol. 19, no. 22 (October 26, 2002). Available at: http://frontlineonnet.com/fl1922/stories /20021108007101200.htm. Tanham, George K. Indian Strategic Thought: An Interpretive Essay (Santa Monica, CA: RAND Publications, 1992). Tannenwald, Nina. The Nuclear Taboo: The United States and the Non-­Use of Nuclear Weapons since 1945 (Cambridge: Cambridge University Press, 2007). Tellis, Ashley J. Strategic Stability in South Asia. Arroyo Center Document Briefing (Santa Monica: RAND Corporation, 1997). Tellis, Ashley J. India’s Emerging Nuclear Posture (Santa Monica: RAND, 2001). Tellis, Ashley J. “US-­Pakistan Relations: Assassination, Instability, and the Future of US Policy.” Hearing of the Middle East and South Asia Subcommittee of the House Foreign Affairs Committee, Federal News Service, January 16, 2008. Tertrais, Bruno. “ ‘Destruction Assuree’: The Origins and Development of French Nuclear Strategy 1945–­1981.” In Henry D. Sokolski, ed., Getting MAD: Nuclear Mutual Assured Destruction, Its Origins and Practice (Carlisle, PA: Strategic Studies Institute, November 2004). Tertrais, Bruno. “The Last to Disarm? The Future of France’s Nuclear Weapons.” Nonproliferation Review, vol. 14, no. 2 (July 2007). The French White Paper on Defence and National Security (New York: Odile Jacob Publishers, 2008). Thomas, Raju G. C. and Amit Gupta, eds. India’s Nuclear Security (Boulder, CO: Lynne Rienner, 2000). Tomz, Michael, Jason Wittenberg, and Gary King. “CLARIFY: Software for Interpreting and Presenting Statistical Results.” Journal of Statistical Software 8, no. 1 (2003). Trachtenberg, Marc. “The Influence of Nuclear Weapons in the Cuban Missile Crisis.” International Security, vol. 10, no. 1 (Summer 1985). Trachtenberg, Marc. A Constructed Peace: The Making of the European Settlement 1945–­1963 (Princeton, NJ: Princeton University Press, 1999). Ullman, Richard H. “The Covert French Connection.” Foreign Policy, no. 75 (Summer 1989). United States National Intelligence Council. Foreign Missile Developments and the Ballistic Missile Threat Through 2015. September 1999. Vasquez, John. “The Realist Paradigm and Degenerative versus Progressive Research Programs: An Appraisal of the Neotraditional Research on Waltz’s Balancing Proposition.” American Political Science Review, vol. 91, no. 4 (December 1997). Vreeland, James. “Why Do Governments and the IMF Enter into Agreements?: Statistically Selected Case Studies.” International Political Science Review, vol. 24, no. 3 (2003).

Bibliography 331 Walt, Stephen M. The Origins of Alliances (Ithaca, NY: Cornell University Press, 1987). Waltz, Kenneth. Theory of International Politics (New York: McGraw Hill, 1979). Waltz, Kenneth. The Spread of Nuclear Weapons: More May Be Better. Adelphi Paper No. 171 (London: IISS, 1981). Waltz, Kenneth and Scott D. Sagan. The Spread of Nuclear Weapons: A Debate Renewed (New York: Norton, 2002). “Warning Forced India to Pull Back Troops, Says President.” Dawn (Islamabad), December 31, 2002. Available at http://www.dawn.com/2002/12/31/top3 .htm. Whetten, Lawrence. The Canal War (Cambridge, MA: MIT Press, 1974). Wohlforth, William. The Elusive Balance: Power and Perceptions During the Cold War (Ithaca, NY: Cornell University Press, 1993). Woodward, Bob. “Pakistan Reported Near Atom Arms Production.” Washington Post, November 4, 1986. Yost, David S. France’s Deterrent Posture and Security in Europe, Part I: Capabilities and Doctrine. Adelphi Papers, no. 194 (Winter 1984/1985). Yost, David S. “France’s Nuclear Dilemmas.” Foreign Affairs, vol. 75, no. 1 (January/February 1996). Yousaf, Brigadier Mohammad and Major Mark Adkin. The Bear Trap: Afghanistan’s Untold Story (London: Leo Cooper, 1992). Yuan, Jing-­dong. “Effective, Reliable and Credible: China’s Nuclear Modernization.” Nonproliferation Review, vol. 14, no. 2 (July 2007). Zhang, Shu Guang. Deterrence and Strategic Culture (Ithaca: Cornell University Press, 1992).

Index

acquisition bias, nuclear weapons scholarship, 3, 6–­7 Advani, L. K., 118 AEB (Atomic Energy Board), South Africa, 208–­9, 210 AEC (Atomic Energy Commission), India, 96 Afghanistan, as Pakistan’s leverage, 59, 60–­61, 70–­74 Afrikaner strategic culture, 220–­21 Agni missile family, India, 97, 98–­99, 104 Ahmad, Shamshad, 270 aircraft delivery capabilities: China, 123, 125; France, 158–­59, 170; India, 96, 97, 98; Israel, 183, 191; Pakistan, 67, 68, 83, 267; South Africa, 210 Albright, David, 209 al-­Samarrai, Wafic, 294 Angola, 209–­10, 212–­13, 215–­16, 217–­18 AN gravity bombs, France, 158 Arens, Moshe, 203–­  4 Armitage, Richard, 275 Armoni, Ora, 188 ARMSCOR, South Africa, 210, 211 Arunachalam, V. S., 96–­97, 108 Asal, Victor, 252 ASMP missiles, France, 159–­60, 169–­70 assembled/disassembled status. See storage arrangements assertive control feature, civil-­military arrangements, 37–­39, 41, 45–­  46 assured retaliation posture: defined, 4, 8; features of, 17–­19, 21–­22; variable priorities, 8–­9, 27, 31–­  42. See also China, assured retaliation posture; conflict deterrence effects, nuclear postures; crisis entries; India, assured retaliation posture; Israel, assured retaliation

posture phase; Posture Optimization Theory asymmetric escalation posture: defined, 4, 8; features of, 19–­23; variable priorities, 8–­9, 27, 31–­  ­42. See also conflict deterrence effects, nuclear postures; crisis entries; France, asymmetric escalation posture; Pakistan, asymmetric escalation posture phase; Posture Optimization Theory Atomic Energy Board (AEB), South Africa, 208–­9, 210 Atomic Energy Commission (AEC), India, 96 Baker, James, 294 Ball, George, 157 Basrur, Rajesh, 117–­19 Beardon, Milton, 73–­74 Beardsley, Kyle, 252 The Bear Trap (Yousaf ), 71 Beaufre, Andre, 162–­63 Beg, Aslam, 65, 67–­68, 266 Belarus, research exclusion decision, 3n3 Ben Gurion, David, 181–­82, 183, 184, 185, 198 Bennett, D. Scott, 229–­30, 237, 240n38 Ben-­Zvi, Avraham, 203 Bharatiya Janata Party (BJP), 76, 98, 118, 271–­73, 274, 277 Bhutto, Benazir, 68, 73, 83 Bhutto, Zulfikar Ali, 58 BJP (Bharatiya Janata Party), 76, 98, 118, 271–­73, 274, 277 Bluestar operation, India, 113n81 Botha, P. W., 207, 210, 212, 213, 215, 218, 219–­20 BrahMos missiles, 98–­99

333

334 Brasstacks crisis, 62–­65, 260–­65 Brodie, Bernard, 227 Buccaneer bombers, South Africa, 210 Bulletin of the Atomic Scientists, 179, 191 Bundy, McGeorge, 6, 226n6, 227 Burgess, Stephen, 215 Burr, William, 194 Burrows, Andrew S., 161 Burrows, William, 68 Bush, George H. W. (and administration), 73–­74, 203 Buys, André, 212, 216 C-­130 aircraft, Pakistan, 59, 67 Camp David Accords, 191 Canada, 252 capabilities feature, nuclear postures, 4, 16–­22 capabilities status, countries: China, 123, 125–­26, 129–­30, 133–­36, 145–­  ­48; France, 156–­59, 167, 169–­70, 176; India, 75, 95–­99, 265; Israel, 182–­84, 191–­92, 197–­98, 200–­201; Pakistan, 57–­60, 67, 68, 76–­77, 82–­83, 90, 265–­ 66; South Africa, 207, 208–­10, 219–­20 Carter, Jimmy (and administration), 215 catalytic posture: defined, 4, 8; features of, 15–­17, 20–­23; variable priorities, 8–­9, 27, 31–­  ­42. See also conflict deterrence effects, nuclear postures; crisis entries; Israel, catalytic posture phase; Pakistan, catalytic posture phase; Posture Optimization Theory; South Africa, catalytic posture CEA, France, 154 Central Military Commission, China, 127–­ 30, 135–­36, 143–­  ­45 Chander, Avinash, 105 Chari, P. R., 61, 64, 262, 271 Charlie Wilson’s War (Crile), 72 Chase, Michael, 133, 135, 137 Chashma facility, Pakistan, 58 Cheema, Zafar, 61, 64, 81, 262, 271 chemical weapons, Iraq, 292–­95 Cheney, Richard (Dick), 203, 293 Chengappa, Raj, 271 CHIC-­4 design, 59, 83 China: Goldstein’s posture selection argument, 25–­26; in India’s posture selection, 110–­12, 116; as North Korea

index patron, 305–­6; Pakistan nuclearization assistance, 58, 59, 76 China, assured retaliation posture: capabilities development, 123, 125–­26, 129–­30, 133–­36; command-­and-­ control arrangements, 126–­30, 133–­34, 135–­37; doctrine declarations, 125, 130–­33; limitations of alternative theory predictions, 145–­51; optimization theory’s predictive value, 138–­  ­45, 151–­52; overview, 21n12, 121–­22, 123–­25, 138, 152; security environment, 122–­23 Chirac, Jacques, 161, 170, 171 Christensen, Thomas, 33 CIA, 72, 73, 74 civil-­military relations feature, nuclear postures, 9, 36–­39, 41, 45–­  ­46, 301. See also command-­and-­control arrangements feature, nuclear postures civil-­military relations status, countries: China, 138, 139–­  ­40, 143–­  ­45; France, 166–­67; India, 39, 97–­98, 105, 110, 112–­16, 119–­20, 279n74; Israel, 204–­ 5; Pakistan, 39. See also command-­and-­ control arrangements status, countries Clinton, Bill (and administration), 269, 270 code controls. See command-­and-­control arrangements entries coding criteria, methodology, 21–­22 Cohen, Avner (writing on Israel): capabilities development, 183, 192, 200–­201; doctrine development, 184–­86, 192, 194; Six Day War, 285–­86; Yom Kippur War, 188–­89, 190, 289, 291 Cohen, Stephen, 61, 64, 75, 113, 262, 271 Coll, Steve, 71, 74 command-­and-­control arrangements feature, nuclear postures, 4, 18–­19, 21–­22, 26 command-­and-­control arrangements status, countries: China, 126–­30, 133–­34, 135–­37, 139, 143–­  ­45, 151; France, 159–­60, 171; India, 100–­109, 112–­13, 115; Israel, 184–­85, 201; Pakistan, 82–­ 90, 93; South Africa, 210–­11. See also civil-­military relations status, countries conflict deterrence effects, nuclear postures: data results, 238, 239t, 240–­  ­48;

index 335 hypotheses about, 231–­36; overview, 222–­24, 225–­26; scholarly treatment, 5–­6, 224–­31; significance of findings, 248–­52, 306–­7; testing methodology, 236–­  40 Congress, India, 98 conventional offensive threat feature, nuclear postures, 34–­36, 44 conventional offensive threat status, countries: China, 140–­  43; France, 153–­54, 165–­66; India, 109–­10, 111–­12; Israel, 184–­85, 186, 191, 194–­96, 202; South Africa, 217–­18 costs. See resource constraints entries Crile, George, 72 crisis analysis, Arab-­Israeli, posture effects: Gulf War, 292–­95; hypotheses, 282–­83; methodology, 254–­57; overview, 253–­ 54, 295–­98; Six Day War, 283–­86; Yom Kippur War, 286–­92 crisis analysis, India-­Pakistan, posture effects: Brasstacks, 260–­65; hypotheses, 259–­60; Kargil War, 267–­73; Kashmir Compound, 265–­67; methodology, 254–­ 57; Mumbai terrorist attack, 279–­81; Operation Parakram, 273–­79; overview, 253–­54, 257–­59, 281–­82, 295–­98 Crocker, Chester, 215–­16, 218 Cuban military, Angola presence, 209–­10, 212–­13 Cuban Missile Crisis, 223 custody arrangements. See command-­and-­ control entries Dayan, Moshe, 187–­89, 288–­89 Dean, John Gunther, 64, 264 Debré, Michel, 163 declaratory nuclear doctrine, defined, 4. See also doctrine development and declarations Defence Research & Development Organization (DRDO), India, 96–­97, 105–­6, 117 Defense White Paper, France, 163, 171, 172, 173 de Gaulle, Charles, 155–­57, 164, 166–­68, 182 de Klerk, F. W., 207, 210, 212, 214 delegative feature, civil-­military arrangements, 37, 38–­39, 41, 46

delivery capabilities: China, 123, 125–­26, 131, 134–­35; France, 158, 169–­70; India, 65, 96, 97, 98–­99, 103–­  4; Israel, 183–­84, 191–­92, 200–­201; Pakistan, 59, 65, 67, 76, 83; United Kingdom, 155 Deng Xiaoping, 125, 149 d’Estaing, Giscard, 160 deterrence power. See conflict deterrence effects, nuclear postures de Villiers, J. W., 210 DF (Dong Feng) missile family, China, 125–­26, 127, 131, 134–­35, 147–­  48 Dien Bien Phu, 155, 164 Digvijay exercise, India, 262, 264n21 Dimona reactor, Israel, 182–­83, 191, 200, 283, 285–­86 disassembled/assembled status. See storage arrangements Ditmer, Lowell, 57 doctrine development and declarations: China, 125, 130–­33; France, 160–­63, 171–­73; India, 99–­101; Israel, 184–­87, 192–­93; Pakistan, 79–­82; South Africa, 211–­13 Dolphin-­class submarines, Israel, 181, 200–­201 Dong Feng (DF) missile family, China, 125–­26, 127, 131, 134–­35, 147–­  48 DRDO (Defence Research & Development Organization), India, 96–­97, 105–­6, 117 Druckett, Carl, 191 Dulles, John Foster, 157, 168 Durrani, Mahmud, 87 Egypt: Camp David Accords, 191; Six Day War, 183–­84, 190, 196, 283–­86; Soviet Union assistance, 193, 195; War of Attrition, 287; Yom Kippur War, 16–­17, 180–­81, 187–­91, 216–­17, 223, 286–­92 Eilts, Hermann, 189–­90, 291 Eisenhower, Dwight D. (and administration), 122, 155, 156, 157, 182 Elazar, Chief of Staff, 188 employment doctrine, defined, 4. See also doctrine development and declarations equivalency problem, nuclear scholarship, 6, 228–­29 Eshkol, Levi, 183, 184, 185 Evron, Yair, 196

336 existential bias, nuclear weapons scholarship, 3, 5, 6–­7, 223–­26, 248, 250–­51 existential deterrence, 6–­7, 16, 34, 69, 188, 223, 226, 227 F-­14 aircraft, Israel, 191 F-­16 aircraft: Israel, 191; Pakistan, 67, 68, 83, 267 Fearon, James, 232 Feaver, Peter, 26, 29, 36, 37, 38 Feldman, Shai, 293 Fieldhouse, Richard W., 161 first-­use doctrine. See asymmetric escalation posture; France, asymmetric escalation posture; Pakistan, asymmetric escalation posture phase Forces Aérienne Tactique, 158, 160 Fourquet, General, 161 Foxbats over Dimona (Ginor and Remez), 285 France: Goldstein’s posture selection argument, 25–­26; Israel assistance, 182, 194, 283 France, asymmetric escalation posture: capabilities development, 156–­59, 169–­70; command-­and-­control arrangements, 159–­60; doctrine articulation, 160–­63, 171–­73; limitations of alternative theory predictions, 164–­65, 175–­77; optimization theory’s predictive value, 163–­69, 174–­75, 177–­78; overview, 34, 153–­54; security environment, 153–­57. See also conflict deterrence effects, nuclear postures Fravel, M. Taylor, 149–­50 Gaillard, Felix, 157 Galili, Yisraeli, 188 Gallois, Pierre, 177 Gandhi, Indira, 95, 96, 114, 117 Gandhi, Rajiv, 63, 96–­97, 117, 120, 261, 262 Ganguly, Sumit, 57–­58, 257, 273 Gates, Robert, 66, 265 Gavin, Francis, 157 geography’s role, security environment, 35, 44, 139, 142 Ghauri missile family, India, 76, 83 Gill, Bates, 133, 147 Ginor, Isabella, 285 Goldstein, Avery, 25–­26, 47, 154–­55

index Gray, Colin, 50, 118 Gregory, Shaun, 159–­60 Guam, as China target, 126, 147 Gulf War (1991), 181, 199–­200, 202–­6, 292–­95 Haaretz, 203 Haberman, Clyde, 203 Hadès missiles, France, 170, 176 Hagerty, Devin, 57–­58, 273 Heikal, Mohammed, 289 Hernu, Charles, 161, 162 Hersh, Seymour, 66–­67 Honest John missiles, 159 Hong bombers, China, 125 Hoon, P. N., 262 Horowitz, Michael, 229n22, 252 Hoyt, Timothy, 85 Hua Di, 146, 147 Huntington, Samuel P., 37 Hymans, Jacques, 23, 96–­97, 309 immediate threat possibility, as security environment condition, 35–­36 India, as China threat, 142 India, assured retaliation posture: capabilities development, 76, 95–­99; command-­and-­control arrangements, 101–­9; doctrine declarations, 99–­101; limitations of alternative theory predictions, 116–­19; optimization theory’s predictive value, 109–­16, 119–­20; overview, 21n12, 35, 94, 152. See also crisis analysis, India-­Pakistan, posture effects; Pakistan entries Indian Express, 280 Inter Services Intelligence (ISI), Pakistan, 70–­71, 73 Iran, 1, 202, 304–­5 Iraq, 181, 199–­200, 202–­3 ISI (Inter Services Intelligence), Pakistan, 70–­71, 73 Israel, assured retaliation posture phase: capabilities development, 200–­201; optimization theory’s predictive value, 202–­6; overview, 153, 179–­80, 181, 199–­200; security environment, 199–­ 200. See also crisis analysis, Arab-­ Israeli, posture effects Israel, catalytic posture phase: capabilities development, 182–­84, 191–­92; doctrine

index 337 development, 184–­87, 192–­93; limitations of alternative theory predictions, 197–­98; optimization theory’s predictive value, 193–­97, 199; overview, 153, 179–­81; security environment, 181–­82; third-­party patron availability, 16–­17, 34, 43, 187–­91. See also crisis analysis, Arab-­Israeli, posture effects Israel, South Africa relationship, 210 Iyengar, P. K., 96–­97 Jaguar aircraft, India, 97 Jaish-­e-­Mohammad, 273–­75 Jane’s Intelligence Review, 192 Japan, 1, 17n3, 126, 147, 305 Jericho missiles, Israel, 183, 189–­90, 191–­ 92, 201, 289 Jervis, Robert, 6, 11, 227, 234 Jiang Zemin, 132 Jin-­class submarines, China, 134–­35, 136–­37 Jing-­Dong Yuan, 137 JL (Julang) missile family, China, 125, 134 Johnston, Iain, 50–­51, 118, 124, 131, 132, 137, 145n76, 148, 150 Jones, Don, 68 Jordan, 284 Joshi, Manoj, 278 Julang (JL) missile family, China, 125, 134 Kahn, Herman, 5 Kahuta facility, Pakistan, 58, 59, 63, 67 Kampani, Gaurav, 97 Kanwal, Gurmeet, 272 Kapoor, Deepak, 280 Kapur, Paul, 257, 270, 274 Kargil War, 11, 250, 253, 267–­74, 281 Kartha, Tara, 4 Kashmir Compound crisis, 62, 65–­68, 265–­67 Kazakhstan, research exclusion decision, 3n3 Kennedy, John F. (and administration), 156–­57, 194 Kerr, Richard, 66 Khan, A. Q., 58, 63–­64, 72, 264 Khan, Feroz Hassan: on delivery capabilities, 267; on NATO-­Pakistan comparison, 81; on negative controls, 87, 90; on nuclear forces timing, 82; Operation Parakram, 275; on red-­lines, 80; on signaling strategy, 68; test events, 59

Khan, Munir Ahmad, 58 Khan, Yahya, 58 Khushab facility, Pakistan, 58 Kidwai, Khalid, 79–­80, 85–­86, 88 Kier, Elizabeth, 51 Kissinger, Henry, 189–­90, 289, 291 Kristensen, Hans, 99, 136 Lacaze, General, 160, 161 Lashkar-­e-­Taiba, 253, 273–­75, 279 Lavie, Naftali, 188 Lavoy, Peter, 82 Levy, David, 199, 292 Lewis, Jeffrey, 124, 148–­  49 Lewis, John, 122, 125, 128, 129–­30, 139, 146–­  47 Liberman, Peter, 213, 214 Li Bin, 135 Libya, 58 Line of Control (LoC). See crisis analysis, India-­Pakistan, posture effects Liu Huaqiu, 137 LoC (Line of Control). See crisis analysis, India-­Pakistan, posture effects locations of weapons. See storage arrangements Lodi, Sardar FS, 81–­82 Lop Nur test site, 59 Lubbers, Ruud, 72 Luthi, Lorenz, 140 Madrid Conference, 181, 200, 202, 204 Magister aircraft, Israel, 194 Malik, V. P., 268–69, 270–­71, 274 Maoz, Zeev, 193, 196, 286, 291–­92, 293 Mao Zedong, 122–­23, 128n23, 139, 140, 142, 144, 146–­  47, 148–­51 Mardor, Munya, 183, 283 Martellini, Maurizio, 85–­86 Maurin, Francois, 161 Mauripur Air Base, Pakistan, 68 McNamee, Terence, 214, 220–­21 Mearsheimer, John, 11, 44, 165, 227 measurement methodology, 43–­  46 Medeiros, Evan S., 130, 132, 133, 137, 145, 149–­50 Meir, Golda, 186, 188 Meir, Yehuda Ben, 205 Mendès-­France, Pierre, 155 Mendl, Wolf, 167–­68 Menon, Raja, 109, 114

338 Menon, V. K. Krishna, 113 MiG aircraft, 195, 285, 286, 287 Military Balance, 217–­18 Ministry of Defence, India, 113–­15 Ministry of Finance, India, 113 Mirage aircraft: France, 158–­59, 170; India, 97 MIRV technology: China, 135; France, 170 Mishra, Brajesh, 100, 104, 107, 118, 272–­ 73, 275, 278 missiles. See delivery capabilities Mitterand, FranÇois, 170, 173, 176–­77 Mollet, Guy, 155–­56, 157 Mozambique, 207, 211, 218 Mubarakmand, Samar, 88 Mumbai, terrorist attack, 223, 253, 260, 279–­82, 296 Musharraf, Pervez, 84, 267–­68, 274–­75, 277 Mystere aircraft, Israel, 194 Namibia, 207, 213, 216, 217, 218 Nasr missiles, Pakistan, 83, 86 Nasser, Gamal Abdel, 283, 285 National Command Authority (NCA), Pakistan, 84 National Defense White Paper, China, 132 National Military Command Center, France, 159 NATO: France’s positions, 154, 155, 171; Pakistan doctrine compared, 77–­78, 80–­81; Warsaw Pact capabilities compared, 165–­66 naval forces. See sea-­based delivery capabilities Nawaz, Shuja, 64, 71, 264 Nayar, Kuldip, 63, 264 NCA (National Command Authority), Pakistan, 84 Ne’eman, Yuval, 291n121 negative controls, in command-­and-­control arrangements: China, 128–­29, 144; France, 159, 160; India, 106–­7; Pakistan, 87–­90; South Africa, 211 Nehru, Jawaharlal, 112, 113, 118, 119 neoclassical realist school, 28–­29 Netherlands, 72 New York Times, 186, 203, 287 Nie Ronghzen, 124, 147, 149 Nitze, Paul, 226n6

index Nixon, Richard M. (and administration), 186, 189–­90, 289, 291–­92 No Dong missiles, North Korea, 76 no-­first-­use doctrine. See assured retaliation posture; China, assured retaliation posture; India, assured retaliation posture Non-­Proliferation Treaty (NPT), 32, 71, 180, 186, 207 Noorani, Zain, 264 Norris, Robert S., 99, 136, 161 North Korea, 1, 3n3, 76, 305–­6 NPT (Non-­Proliferation Treaty), 32, 71, 180, 186, 207 Nuclear Logics (Solingen), 34n35 nuclear opacity. See Israel, catalytic posture phase nuclear postures, overview: future research possibilities, 308–­10; international security implications, 12, 300, 303–­8; scholarly treatment, 2–­8, 23–­26, 299–­ 300; types of, 4–­5, 8, 14–­23. See also specific topics, e.g., conflict deterrence effects, nuclear postures; Posture Optimization Theory; South Africa, catalytic posture Oakley, Robert, 67, 69, 72–­73 offensive threat, measurement methodology, 44 Olmert, Ehud, 179 operational doctrine. See doctrine development and declarations Operation Bluestar, India, 113n81 Operation Parakram, India, 273–­79 optimization theory. See Posture Optimization Theory Oren, Michael, 285 Padmanabhan, S., 102–­3, 270, 275, 278 PAEC (Pakistan Atomic Energy Commission), 58 Pakistan, asymmetric escalation posture phase: capabilities development, 76–­77, 82–­83; command-­and-­control arrangements, 82–­90; limitations of alternative theory explanations, 91; optimization theory’s predictive value, 90–­91; overview, 19n8, 55, 56–­57, 76–­79, 152. See also conflict deterrence effects, nuclear

index 339 postures; crisis analysis, India-­Pakistan, posture effects; India, assured retaliation posture Pakistan, catalytic posture phase: Brasstacks crisis significance, 61–­65, 260–­65; capabilities development, 57–­60; Kashmir Compound crisis significance, 61–­62, 65–­69, 265–­67; limitations of alternative theory explanations, 74–­76; optimization theory’s predictive value, 69–­74; overview, 55–­ 56; third-­party patron availability, 34, 43, 60–­69. See also conflict deterrence effects, nuclear postures; crisis analysis, India-­Pakistan, posture effects; India, assured retaliation posture Pakistan Atomic Energy Commission (PAEC), 58 PALs, 88–­89, 159, 160, 211 Parakram operation, India’s, 273–­79 Patriot missiles, 199–­200, 293 Patton tanks, Israel, 194 Péan, Pierre, 183 People’s Liberation Army, 130–­31, 132, 140–­  41, 143–­  45 Peres, Shimon, 182, 185–­86 Perkovich, George, 69, 96, 262, 266 Permissive Action Link (PAL) capabilities, 88–­89, 159, 160, 211 Philippines, as China target, 126, 147 plutonium pathway: China, 123; France, 158; India, 95, 99; Israel, 183, 191; Pakistan, 58 Pluton missiles, France, 159, 160, 169, 170, 176 Popeye Turbo missiles, Israel, 200–­201 Posen, Barry, 21, 38, 84–­85, 165 Posture Optimization Theory: alternative explanations critiqued, 46–­52; civil-­ military arrangements variable, 36–­39; decision-­making tree, 27–­28, 32f, 53f; measurement methodology, 43–­  46; neoclassical realism’s contributions, 28–­31; overview, 8–­9, 41–­   42, 301–­   4; security environment variable, 31–­36; testability of, 27–­28, 42–­  43, 52–­54. See also specific topics, e.g., China, assured retaliation posture; security environment entries Powell, Robert, 233n29

Prahaar missile, 99 Prakash, Arun, 106–­7, 109, 114 Pressler Amendment, 62, 71, 73 Prithvi missile family, India, 97, 98 proximity threat, measurement methodology, 44 Punjab: Brasstacks crisis, 63–­65, 260–­65; Kashmir Compound crisis, 62, 65–­68, 265–­67 Purkitt, Helen, 215 Qazi, Javed Ashraf, 275 Quandt, William, 190, 291 Rabin, Yitzhak, 192 RAFAEL, Israel, 182, 183–­84, 192 Rao, Narashimha, 98, 120 rationality assumption, 29–­30 Reagan, Ronald (and administration), 72, 215 Reed, Thomas, 183 regional nuclear powers. See nuclear postures, overview; Posture Optimization Theory; specific countries Reiss, Mitchell, 213, 216 Remez, Gideon, 285 resource constraints feature, nuclear postures, 9, 32, 39–­  41, 42, 46 resource constraints status, countries: China, 40, 46, 139, 145; France, 167, 174–­75; India, 115–­16 revisionist intentions variable, exclusion purposes, 30n33 Riedel, Bruce, 70, 270 Rose, Gideon, 28 Rosen, Stephen, 108 Rotberg, Robert, 218 Roychowdhury, R. K., 279, 280 Sadat, Anwar, 287–­88 SAFARI reactor, South Africa, 208–­9 Sagan, Scott, 23, 29, 84–­85 Samet, Gidon, 203 sanctions programs, 77, 208–­9 Sargodha Air Base, Pakistan, 67, 271 Sarkozy, Nicolas, 171, 172, 173 Saudi Arabia, 58 Sawhney, Pravin, 277 Schelling, Thomas, 226–­27 Schultz, George, 71

340 Schwartzkopf, Norman, 203 Scud missiles, 199–­200, 203, 287, 292, 293 sea-­based delivery capabilities: China, 125, 131, 134–­35, 136–­37; France, 158, 160, 169–­70; India, 99, 108–­9; Israel, 200–­201; United Kingdom, 169, 176 Second Artillery, China, 127–­30, 132–­34, 136–­37, 144 second-­strike forces feature, nuclear postures, 17–­19. See also assured retaliation posture Secunderabad, India, 97 security environment feature, nuclear postures, 9, 25–­26, 31–­36, 41–­  42, 301. See also structural realism theory, predictive value security environment status, countries: China, 122–­23, 139, 145, 146, 151; France, 153–­57, 165–­67, 174; India, 35, 109–­10, 111–­12; Israel, 181–­82, 193, 197, 198, 199–­200; Pakistan, 35, 70, 76–­82, 90–­92; South Africa, 209–­ 10, 217–­19 SFC (Strategic Forces Command), India, 99, 103, 105–­9 Shaheen missile family, Pakistan, 76, 83 Shamir, Yitzhak, 199, 203, 204, 292 Shankar, Vijay, 78n77, 89, 94, 99, 100, 104, 106, 108 Sharif, Nawaz, 83, 92, 268, 269, 270 Shlaim, Avi, 204 Shomron, Dan, 293 Sikh conflicts, India, 62–­65, 113n81, 260–­65 Sindh, Brasstacks crisis, 63–­65, 260–­65 Singh, J. J., 115 Singh, Jaswant, 118 Singh, S. K., 264 Six Day War, 183–­84, 190, 196, 283–­86 Skyhawk aircraft, Israel, 183, 195 Sneh, Ephraim, 201 Snelling, Thomas, 224 Snyder, Glenn, 234 Snyder, Jack, 33, 50–­52 Solarz, Stephen, 60, 72, 265 Solingen, Etel, 23, 33n35 Sood, V. I., 278 Sood, V. K., 277 South Africa, catalytic posture: capabilities development, 207, 208–­10; command-­ and-­control arrangements, 210–­11;

index doctrine declarations, 211–­13; limitations of alternative theory predictions, 219–­21; optimization theory’s predictive value, 214–­19, 221; overview, 34, 153, 207–­8, 214; security environment, 209–­10 South African Defence Force, 211 South Pacific, Vela incident, 191, 211 Soviet Union: Afghanistan conflict, 72, 73–­74; China relationship, 128, 140–­   42, 146, 147; Cuban Missile Crisis, 223; India relationship, 110; in South Africa’s decision-­making pathway, 212–­ 13, 215–­16, 218–­19. See also Egypt; France, asymmetric escalation posture SPD (Strategic Plans Division), Pakistan, 79–­80, 84 Stam, Alan, 229–­30, 237, 240n38 status quo preferences variable, exclusion purposes, 30n33 Steyn, Hannes, 211 Stillman, Danny, 183 Stokes, Mark, 136 storage arrangements: China, 127–­28; India, 96, 97, 101–­5, 108–­9; Israel, 192; Pakistan, 85–­89; South Africa, 211 strategic culture theory, predictive value: China’s assured retaliation posture, 139, 148–­51; France’s asymmetric escalation posture, 164, 167–­69, 176–­ 77; India’s assured retaliation posture, 117–­19; Israel’s postures, 197–­98, 205–­ 6; overview, 50–­52, 302–­3; Pakistan’s postures, 75–­76, 91; South Africa’s catalytic posture, 220–­21 Strategic Forces Command (SFC), India, 99, 103, 105–­9 Strategic Plans Division (SPD), Pakistan, 79–­80, 84 structural realism theory, predictive value: China’s assured retaliation posture, 145–­  46; France’s asymmetric escalation posture, 164–­65, 166–­69, 175–­76, 177; India’s assured retaliation posture, 116; Israel’s postures, 197, 205–­6; overview, 28–­30, 47–­  49; Pakistan’s postures, 74–­75, 91; South Africa’s catalytic posture, 219 Stumpf, Waldo, 209–­10, 211, 212, 215 submarines. See sea-­based delivery capabilities

index 341 Suez Crisis, 155–­56, 164, 182 Sundarji, K., 63, 102, 228, 261–­62 Sundarji Doctrine, 277–­78 sunk-­cost signals, Fearon’s theory, 232–­33 superpower model, limitations, 1–­3, 5–­6, 225, 229–­30 survivability feature, nuclear postures, 17–­ 19, 21–­22. See also assured retaliation posture Syria: Six Day War, 183–­84, 190, 196, 283–­86; Soviet Union assistance, 193, 195; Yom Kippur War, 16–­17, 180–­81, 187–­91, 216–­17, 286–­92 Taiwan, in China’s posture decision-­ making, 142 Tanham, George, 114 Taverny, National Military Command Center, 159 technological determinism theory, predictive value: China’s assured retaliation posture, 139, 146–­  48; France’s asymmetric escalation posture, 164–­65, 167, 176, 177; India’s assured retaliation posture, 117; Israel’s postures, 197–­98, 205–­6; overview, 49–­50, 302–­3; Pakistan’s postures, 75, 91; South Africa’s catalytic posture, 219–­20 Tellis, Ashley, 100–­101, 103, 107 Tertrais, Bruno, 163, 172, 177 test events and sites: China, 59, 121, 123; France, 157–­58; India, 76–­77, 94, 95–­ 96, 98, 267; Israel, 183, 185–­86, 191; Pakistan, 59, 76–­77, 275; South Africa, 209, 211, 213 third-­party patron feature, nuclear postures, 15–­17, 31–­36, 43–­   44 third-­party patron status, countries: China, 139, 140; France, 154–­57, 163–­64, 166, 167, 175; India, 109, 110–­11. See also Israel, catalytic posture phase; Pakistan, catalytic posture phase; South Africa, catalytic posture Trachtenberg, Marc, 226n6 Tupolev bombers, China, 125 Ukraine, research exclusion decision, 3n3 unitary actor assumption, 29–­30 United Kingdom: Goldstein’s posture selection argument, 25–­26; in NATO, 154,

155; nuclear posture decisions, 156, 169, 176; research exclusion decision, 3n3; Suez Crisis, 156, 182 United States: as China’s target, 125–­26, 127, 137, 147; as China’s threat, 130, 133, 142–­  43; Cuban Missile Crisis, 223; PAL technology, 88n108. See also France, asymmetric escalation posture; Israel entries; Pakistan entries; South Africa, catalytic posture; third-­party patron entries uranium enrichment pathway: China, 123; Pakistan, 58–­60, 76; South Africa, 207–­9 URENCO designs, 58, 72 Vajpayee, Atal Bihari, 98, 118, 120, 268, 272, 275, 277, 278 van der Walt, Richard, 211 van Loggerenberg, Jan, 211 Vanunu, Mordechai, 191 Vastrap base, South Africa, 209, 213 Vela incident, South Pacific, 191, 211 Vorster, John, 207, 209 Waltz, Kenneth, 6, 11, 28n25, 47, 224, 227 War of Attrition, 187, 287 Warsaw Pact, 20, 80–­81, 164–­66, 212 Wilson, Charlie, 72 Windrem, Robert, 68 Wohlstetter, Albert, 5 Xia class submarines, 125 Xue Litai, 122, 125, 128, 130, 139, 146–­  47 Yom Kippur War, 16–­17, 180–­81, 187–­91, 216–­17, 286–­92 Yost, David, 162–­63 Yousaf, Mohammad, 71 Y-­Plant, South Africa, 209, 210 Zardari, Ali, 57n1 Zekharyah missile field, Israel, 192 Zhang Aiping, 127 Zhanluexue, 132, 133, 137 Zhenbao Island crisis, 128, 140, 144 Zhou Enlai, 127, 141–­  42, 144 Zia ul-­Haq, 58, 59–­60, 71, 73, 83 Z-­plant, South Africa, 219