Forecasting Nuclear Proliferation in the 21st Century: Volume 1 The Role of Theory 9781503627437

Citation preview

Forecasting Nuclear Proliferation in the 21st Century: Volume I. The Role of Theory

Forecasting Nuclear Proliferation in the 21st Century VOLUME 1: THE ROLE OF THEORY Edited by William C. Potter with Gaukhar Mukhatzhanova

Stanford Security Studies An Imprint of Stanford University Press Stanford, California

Stanford University Press Stanford, California © 2010 by the Board of Trustees of the Leland Stanford Junior University. All rights reserved.

No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or in any information storage or retrieval system without the prior written permission of Stanford University Press. Library of Congress Cataloging-in-Publication Data Forecasting nuclear proliferation in the 21st century I edited by William C. Potter with Gaukhar Mukhatzhanova. v. em. Includes bibliographical references and index. Contents: v. 1. The role of theory. ISBN 978-o-8047-6972-3 (cloth: alk. paper) ISBN 978-o-8047-6973-0 (pbk. : alk. paper) 1. Nuclear nonproliferation- Forecasting. I. Potter, William C. II. Mukhatzhanova, Gaukhar. JZs66s.F68 2010 327.1747-dc22 2010011336

Typeset by Thompson Type in 10/14 Minion Special discounts for bulk quantities of Stanford Security Studies are available to corporations, professional associations, and other organizations. For details and discount information, contact the special sales department of Stanford University Press. Tel: (6so) 736-1782, Fax: (6so) 736-1784

CONTENTS

Acknowledgments 1

Forecasting Proliferation: The Role of Theory, an Introduction William C. Potter and Gaukhar Mukhatzhanova

2

vii

1

The Study of Nuclear Proliferation and Nonproliferation: Toward a New Consensus? Jacques E. C. Hymans

13

3 Domestic Models of Political Survival: Why Some Do and Others Don't (Proliferate) Etel Solingen

4 Beyond the Security Model:

Assessing the Capacity of Neoclassical Realism for Forecasting Nuclear Proliferation Natasha E. Bajema

ss

5 Nuclear Latency and Nuclear Proliferation Scott D. Sagan

So

6 When Does a State Become a "Nuclear Weapons State"? An Exercise in Measurement Validation Jacques E. C. Hymans

102

7 The Little-Known Story of Deproliferation: Why States Give Up Nuclear Weapons Activities Harald Muller and Andreas Schmidt

124 v

vi

CONTENTS

8 Why Do States Proliferate? Quantitative Analysis of the Exploration, Pursuit, and Acquisition of Nuclear Weapons Philipp C. Bleek

159

9 A World of Nuclear Powers: A Gedanken Experiment Lewis A. Dunn

193

Notes

231

Contributors

277

Index

281

ACKNOWLEDGMENTS

are the product of a multiyear effort by the James Martin Center for Nonproliferation Studies (CNS) at the Monterey Institute of International Studies to explore nuclear proliferation dynamics. Although the orientation of the project evolved considerably over time-away from an initial embrace of the concept of proliferation chains and reactive proliferation-the research retained a consistent focus on harnessing the insights from both theory and comparative country studies to better understand past nuclear proliferation decisions and to anticipate proliferation trends. Many individuals, in addition to the chapter authors, contributed to this volume. They include Monterey Institute students in my graduate research seminar on "Proliferation Trends and Trigger Events," anonymous reviewers, commentators at project workshops in Washington, D.C. in the fall of 2007 and Monterey in the summer of 2008, and CNS research assistants Liviu Horovitz, Kate Amlin, Sean Dunlop, and Daniel Cunningham. My CNS colleague Gaukhar Mukhatzhanova, however, deserves special mention. This book would not have been possible without her exceptional editorial assistance and tireless efforts to help me navigate the treacherous waters of IR theory and quantitative methods. I also am grateful for the generous support provided to CNS for research related to forecasting proliferation developments by the Carnegie Corporation of New York and the Advanced Concepts and Systems Office of the Defense Threat Reduction Agency. The views expressed in this volume, however, are those of the authors alone and do not necessarily reflect the official policy or THIS BOOK Ar\D ITS COMPANION VOLUME

vii

viii

ACKNOWLEDGMENTS

position of the Defense Threat Reduction Agency, the Department of Defense, or the U.S. government. Last, but not least, I wish to dedicate this book to my wife Anna Vassilieva, who frequently complains-not without some justification-that I am married first and foremost to nonproliferation. I am not sure which of us is more thankful that the study is at last completed. William C. Potter

Forecasting Nuclear Proliferation in the 21st Century: Volume I. The Role of Theory

1

FORECASTING PROLIFERATION The Role of Theory, an Introduction

William C. Potter and Gaukhar Mukhatzhanova

The premise . . . that if one country goes nuclear it may induce another country to do the same seems to rely on the notion that the interstate system is a society offools where an example is contagious. -Ashok Kapur'

about the spread of nuclear weapons has fluctuated widely over the past six decades.2 Periods of alarm, verging on panic, have been followed by times of relative tranquility and optimism that proliferation is neither inevitable nor irreversible. At the time of this writing, the pendulum has swung far in the direction of alarmism, due in part to the fact that the international nuclear nonproliferation regime is under severe stress as a consequence of serious proliferation breaches and disarmament failures. These very real challenges have led to a chorus of proliferation doom and gloom by a large and growing body of policy makers and analysts who employ cataclysmic metaphors forecasting that the emergence of another nuclear-armed state will set in motion nuclear chain reactions, cascades, or dominoes. The current alarm about life in a nuclear-armed crowd has many historical antecedents and can be found in classified National Intelligence Estimates (NIEs) as well as in scholarly analyses. The 1957 NIE, for example, identified a list of ten leading nuclear weapons candidates, including Canada, Japan, and Sweden.3 Sweden, it predicted, was "likely to produce its first weapons in about 1961," while it was estimated that Japan would "probably seek to develop weapons production programs within the next decade."4 In 1963, President John Kennedy expressed the memorable and nightmarish vision of a future world with fifteen, twenty, or twenty-five nuclear weapons powers. 5 In yet another example, the 1965 Gilpatric Committee Report, in one of the earliest usages of proliferation chains terminology, asserted that "an Indian or Japanese decision to build nuclear weapons would probably produce a chain INTERNATIONAL CONCERN

1

2

POTTER AND MUKHATZHANOVA

reaction of similar decisions by other countries."6 A number of the early scholarly projections of proliferation also exaggerated the pace of nuclear weapons spread. A flurry of studies between 1958 and 1962, for example, focused on the "Nth country problem" and identified as many as twelve candidates for "going nuclear" in the near future. 7 As Moeed Yusuf observes in his overview of the history of proliferation forecasts, a concern with the emergence of "nuclear dominoes" has remained high since the end of the 1960s and further intensified after the end of the Cold War. 8 Although proliferation pessimism definitely is not a new refrain, there appears to be a growing consensus today that the world again is approaching a tipping point that could lead to a nuclear proliferation epidemic. 9 This tendency to view nuclear weapons diffusion in terms of automaticity and contagion is not confined to the United States or to a particular political or professional orientation. Indeed, it is equally visible among U.S. officials in past and current administrations, international organizations, scholars, nongovernmental analysts, and media pundits. In contrast to Ashok Kapur's dismissive observation thirty years ago about the illogic of proliferation chains, variants of the domino thesis currently dominate proliferation prognoses. 10 In part, this conventional wisdom may be explained by the prevailing international relations school of thought-political realism and its variantswhich has long dominated scholarly, diplomatic, and popular discourse about international security affairs in general and nuclear weapons in particular. From a classical realist perspective, the quest for nuclear weapons is a rational form of self-help designed to maximize power. 11 Neorealism embraces the same basic assumptions as classical realism but is more attentive to the impact of structural differences in the international system on the occurrence of war and peaceY For much of the nuclear age, a realist or neorealist portrayal of the pursuit of nuclear weapons as an appropriate response to an existential threat has been taken as a given by those who make or follow international security policy. Because self-reliance is the only viable way to secure survival in an anarchical international system, balancing against a powerful adversary or rival necessitates the possession of the "ultimate weapon" or at least a credible security guarantee from a nuclear ally. Employing this logic, John Mearsheimer, Stephen Van Evera, and Benjamin Frankel, among others, thought it likely that the decline of bipolarity after the end of the Cold War would generate a new spate of proliferators, including countries such as Germany, Japan, and

FORECASTING PROLIFERATION

3

UkraineP From a realist perspective, it is neither surprising nor illogical to anticipate the spread of nuclear weapons in a chain or cascadelike fashion, as more and more states balance against competitors far and near. "Every time one state develops nuclear weapons to balance out against its main rival," the argument goes, "it also creates a nuclear threat to another state in the region, which then has to initiate its own nuclear weapons program to maintain its national security." 14 Paradoxically, Kenneth Waltz-the patriarch of neorealism-was himself rather sanguine about the pace of the spread of nuclear weapons in his iconoclastic 1981 Adelphi Paper. "Horizontally, they [nuclear weapons] have spread slowly across countries, and the pace is not likely to change much," he wrote. "Short-term candidates for the nuclear club are not very numerous, and they are not likely to rush into the nuclear military business." 15 To their credit, the authors of the seminal work on proliferation chains, Trends in Nuclear Proliferation, 1975-1995, did not base their analysis exclusively on a neorealist-security argument. 16 Lewis Dunn and Herman Kahn's design and examination of fifteen scenarios of possible nuclear weapons spread over a twenty-year period were in part informed by the recognition that status and prestige also shape states' nuclear choices. Still, their underlying assumption-that one state's acquisition of nuclear weapons would prompt at least several others to follow suit-conforms closely to the neorealist line of thinking. In a companion article to the Trends in Nuclear Proliferation study, Dunn and William Overholt conceptualized the "next phase in nuclear proliferation research" in terms of proliferation chains and argued that, through such an approach, "the dynamics of future proliferation as an overall process are captured more readily." 17 They believed that reactive proliferation described the nature of future spread of nuclear weapons more accurately by highlighting the linkages between countries' decisions. They did not, however, provide a very convincing discussion of the nature and strength of such linkages and simply maintained that "the driving forces of future proliferation" will include not only security and prestige considerations but also a "proliferation momentum." 18 1t was thus assumed that at some future point nuclear proliferation "may become a self-reinforcing process." 19 Although this forecast has yet to be borne out, chain-style metaphors have become the terminological norm, all too often serving as the unquestioned baseline assumption for discussions of future proliferation.

4

POTTER AND MUKHATZHANOVA

Even more detrimental than reactive proliferation metaphors in framing discussions about nuclear futures has been the dearth of theoretically or empirically grounded research on nuclear decision making generally and forecasting nuclear proliferation in particular. Instead, most of what constitutes conventional wisdom about nuclear dynamics is either speculative or based on single cases and therefore also is often contradictory. Besides a broad, but not always explicit, reliance on realist premises (for example, the force of security drivers), most proliferation forecasts are largely devoid of a specific theoretical basis. Commenting on this methodological deficit, Yusuf contends that almost no forecasting work he reviewed "explicitly stated the basis for its projections," while "broad overarching claims were made in highly deterministic tones." 20 As a consequence, past predictions often amounted to speculation about alternative futures, where more attention was paid to designing possible scenarios than explaining why one or the other scenario may come to pass. One problem with the lack of attentiveness to theoretical assumptionseven those of realism-is confusion between probabilistic statements about proliferation tendencies and predictions about individual state behavior. In other words, what may have been reasonable as a probabilistic or "instructive" claim (what states should do given systemic constraints21 ) becomes the basis for deterministic predictions of what states will do, leading to overly simplistic and pessimistic predictions. Attempts to accommodate anomalies such as proliferation restraint have motivated some scholars to modify realism and account for additional factors influencing proliferation. Variables that are not inherently or specifically "realist" (such as prestige aspirations or perceptions of threat and alliance credibility) are sometimes subsumed under a realist framework, and yet realism is not equipped to effectively operationalize such variables in a way that assists forecasting. 22 Jeffrey Legro and Andrew Moravcsik discuss this methodological problem more generally in their critique of latter-day branches of realism, pointing out that attempts to incorporate domestic or institutional factors into realist theory "with no effort to reconcile the resulting contradictions" render the "new" realisms vague and indeterminate. 23 This is not to say that attempts to refine realism or any other theory are unwelcome, but it is evident that efforts to enhance realism's explanatory power reduce the framework's proverbial elegance and parsimony-at least in relation to nuclear proliferation. The absence of observable chain dynamics to date may be explained in part by the intervention of the two superpowers and the presence of their

FORECASTING PROLIFERATION

5

respective alliance structures and security guarantees. Both the United States and the Soviet Union shared an aversion to nuclear proliferation during much of the Cold War period -especially following the Chinese nuclear test in 1964. And in the aftermath of the first Indian test a decade later, the two superpowers often pursued parallel although not identical nuclear export and nonproliferation policies. 24 However, other factors, initially discounted by neorealists, may have played an equally important role. They include the growth of nonproliferation norms (reflected in the rapid expansion in membership of the Treaty on the Non-Proliferation of Nuclear Weapons [NPT] and the diffusion of nuclear weapons-free zones internationally) and the increased impact of international economic considerations on the direction of a number of technologically advanced states that nevertheless chose to foreswear nuclear weapons programs. The end of the Cold War and the diminution of superpower competition reinforced an existing trend in the scholarly analysis of international affairs away from a fixation on international systemic determinants to more complex models of state behavior that sought to capture the role of various domestic sources. This tendency to devote greater attention to subsystemic determinants also is apparent in the burgeoning field of nonproliferation studies, although that literature typically has focused on explanations of past behavior rather than forecasts of nuclear futures. 25 Neoliberal institutionalism-a major alternative paradigm to neorealismshares many basic neorealist assumptions about the pursuit by state actors of rational self-interest in an anarchic international system. Neoliberal institutionalists, however, ascribe a much greater role to economics and are far more optimistic about the possibilities for mitigating security dilemmas and achieving long-term cooperation among states. 26 The key to achieving cooperative outcomes, they argue, is the creation of international institutions that facilitate information sharing about others' capabilities and intentions, mediate conflict, and monitor (and on occasion even enforce) regime compliance. 27 By extension, a neoliberal institutionalist might be expected to argue that countries join nonproliferation institutions to address immediate and projected security concerns or to derive economic benefits such as access to peaceful nuclear energy. Although intuitively plausible, in practice this proposition tends to be more implied than rigorously tested, and relatively few studies have sought to demonstrate the influence of nonproliferation institutions on nuclear weapons restraint. 28

6

POTTER AND MUKHATZHANOVA

Constructivist theory, which emphasizes the evolution and impact of international norms on state behavior, represents a further progression away from realist assumptions. Although accepting the existence of anarchy and the pivotal role of states in the international system, constructivists view anarchy in cultural rather than materialist terms. As such, even "power politics" can be tempered by human practice. 29 Under appropriate conditions, adherents to constructivism maintain, institutions and norms may evolve that are hospitable to the emergence of normative prohibitions against nuclear weapons possession and use. 30 By directing attention to considerations of social reality other than those of a purely materialistic nature, constructivists expand the range of explanations for nuclear weapons abstinence. Most importantly, they demonstrate the potential impact of the international social environment in depressing demand for nuclear weapons and, accordingly, preventing the emergence of proliferation chains. 31 Typically, however, constructivists provide little guidance about when and where to expect normative factors to prevail. A growing body of research suggests that one cannot properly understand nuclear weapons (non)proliferation without reference to the domestic context in which nuclear decisions are made. Indeed, many of the more carefully crafted and thoroughly researched case studies, including those in the companion volume to this book, demonstrate that the interplay of bureaucratic and domestic politics, organizational processes, and individual personalities may be more consequential in shaping proliferation outcomes than the threats emanating from the international security environment. Two recent books by Jacques Hymans and Etel Solingen are especially noteworthy in this regard, and they offer compelling-if also competing-explanations about the force of subnational dynamics in explaining nuclear outcomes. 32 Although the growth of alternative models of nuclear choice is a welcome development, there remains a paucity of theoretically informed research on the dynamics of nuclear proliferation. While this deficit has not prevented a flood of speculative prognoses, most projections about proliferation tendencies and trends display little understanding of the processes by which nuclear decisions in different countries are made. Furthermore, disagreements persist about what constitutes nuclear proliferation, the appropriate level of analysis for study, the importance to be attached to a multitude of plausible proliferation determinants, and how best to penetrate the veil of secrecy surrounding nuclear decisions. 33 As a consequence, although it is relatively easy to demonstrate the deficiencies of conventional neorealist wisdom in terms of predictive power in the proliferation realm, it is less obvious that most alter-

FORECASTING PROLIFERATION

7

native theoretical perspectives such as institutionalism, constructivism, and domestic politics greatly advance our ability to forecast proliferation developments (Jacques Hymans's work on the role of"oppositional nationalist" leaders is the exception to this rule). 34 It is, of course, much more challenging and problematic to employ theory for point predictions than to explain past behavior. It is relatively easy, for example, to find evidence after the fact to support the assertion that a state acquired nuclear weapons due to loosely defined security concerns or, alternatively, to make the case that it exercised nuclear restraint because of-again, loosely defined-effects of the "nonproliferation norm." In contrast, making proliferation forecasts on the basis of theoretical constructs requires a much stricter definition and operationalization of key variables. Among the noteworthy aspects of the recent books by Hymans and Solingen is their attentiveness to more careful definition of variables, emphasis on empirically grounded research, and focus on hypothesis testing. 35 But, even so, several of the case study authors in the companion volume to this book were not entirely satisfied with the precision of Solingen's key variables, and many found Hymans's focus on individual decision makers to be insufficient for the purpose of anticipating future proliferation developments. 36 What then does international relations theory have to offer with respect to nuclear futures? Where are the major gaps in knowledge regarding proliferation dynamics, and how can we better use theory for the purpose of making more informed proliferation forecasts? Does it make sense to search for a single theoretical touchstone, or should one be content to explain much of the variance in nuclear behavior with a combination of theoretical insights? If one opts for a pluralistic approach, how should one proceed to build a more complex model? And, with respect to the concept of reactive proliferation more specifically, can we dispense with the notion of nuclear chains given the failure of many dominoes to fall in chainlike fashion, or should we anticipate that the future may yet correspond to a world in which security and power balancing best explain the proliferation puzzle? The aforementioned questions about forecasting and the potential for the emergence of new proliferation chains were the impetus for the two-year research project, "Forecasting Nuclear Proliferation Developments in the 21st Century," launched by the James Martin Center for Nonproliferation Studies (CNS) at the Monterey Institute of International Studies in 2007. The study, supported primarily by the Carnegie Corporation of New York and the Advanced Systems and Concepts Office of the Defense Threat Reduction Agency,

8

POTTER AND MUKHATZHANOVA

was designed to draw on a combination of cutting-edge social science theory and country or area studies expertise for the purpose of forecasting proliferation developments in twelve key states. As the project evolved, it became apparent that, in addition to producing a significant set of comparative country studies covering both "the usual suspects" and some states under the nonproliferation radar, it also was generating new thinking about how to conceptualize nuclear proliferation, assess the explanatory and predictive power of alternative theoretical approaches, and explore the processes by which a world of more (or fewer) nuclear powers might come about. The project's comparative case studies constitute a companion volume, Forecasting Nuclear Proliferation in the 21st Century: A Comparative Perspective. The present volume is devoted to key theoretical issues identified in the

nuclear proliferation forecasting project. The authors, many of whom are leading figures in the field internationally, probe the broader questions of why states pursue or abstain from nuclear weapons, as well as finer methodological issues involving concept definition and development, hypothesis testing, and generalization of findings. They also draw on both the extensive body of qualitative analysis and the inchoate but important work of a quantitative nature. Also informing this volume is the view that advances in nonproliferation forecasting are well served by challenging old preconceptions-regardless of the support they enjoy-especially if they have not previously been subject to much scrutiny. Although the chapters in this volume represent a diverse set of perspectives, theoretical preferences, and methodological approaches, they share a readiness to look anew at the dynamics of nuclear proliferation and the best means by which to describe, explain, and predict nuclear weapons spread. To be sure, not all of the chapters place primary emphasis on forecasting, and some do not dwell specifically on the relationship between one state's nuclear behavior and that of another. Collectively, however, the essays provide a better understanding of the limits of reactive proliferation as well as the circumstances under which weapons diffusion is most likely to occur. The initial chapter by Hymans provides a critical review of selected landmarks in the evolution of the nonproliferation field. His analysis covers the territory from William Epstein's prescient study of threats to the nonproliferation regime and Stephen Meyer's path-breaking analysis of proliferation dynamics to more contemporary efforts by T. V. Paul to salvage realism (of a "prudential" variety). His critical survey also examines more recent works by Solingen, his own psychological-constructivist contribution, and the im-

FORECASTING PROLIFERATION

9

portant but sometimes neglected studies by Itty Abraham, Benjamin Frankel, and Xinyuan Dai. While these authors differ in their views about the nature of the proliferation puzzle and the determinants of nuclear choice, Hymans discerns a number of common features relevant to the issue of forecasting. They include acceptance by most of the authors of "the basic fact that nonproliferation has been the norm" and that nuclear weapons spread is neither normal nor inevitable. 37 Hymans observes that the authors on his landmark list also generally agree that nuclear choices cannot be understood without close examination of domestic factors and that to "go nuclear" is potentially very costly for most states. In addition, he finds a convergence of views about the way forward for the nonproliferation field, most importantly the need to combine theory development with rigorous testing by means of historical research. What is required, he believes, is to move beyond the "folk knowledge of the nonproliferation community" and old laundry lists of proliferation determinants and to concentrate instead "on the political processes that generate specific kinds of nuclear choices." 38 Chapter 3, by Solingen, refines her pioneering model of domestic political survival, which emphasizes the role of the regime's economic orientation in shaping nuclear weapons decision making. Building on her award-winning book Nuclear Logics, she explores the implications of her approach for reactive proliferation, especially as it pertains to potential regional proliferation chains. 39 Solingen concludes that, while states ruled by outward-looking coalitions are less likely to pursue nuclear weapons, regional dynamics-in areas dominated by inward-looking states-may increase the likelihood of economically liberalizing coalitions deciding in favor of nuclear weapons. In Chapter 4, Natasha Bajema, like Paul, attempts to instill realism with renewed life in the proliferation arena. Unlike Paul, however, she employs the relatively new framework of neoclassical realism. This approach accepts the premise that states pursue nuclear weapons because of security challenges but seeks to factor in how perceptions of threat and the ability of states to mobilize internal resources influence their balancing (proliferation) behavior. Bajema argues that, when faced with an external threat, stronger states capable of mobilizing internal resources more quickly and efficiently are more likely to exercise balancing through the acquisition of nuclear weapons. Given the introduction of the intervening variables of threat perception and state capacity, this modification of realism downplays the likelihood of proliferation chain dynamics, though it does not discount them entirely. "Sudden

10

POTTER AND MUKHATZHANOVA

changes in the threat context may trigger nuclear proliferation among strong states," Bajema writes, while weaker states will face much greater initial hurdles to going nuclear. Therefore, "it is unlikely that a decision by one country to develop nuclear weapons will immediately result in a nuclear chain reaction or cascade." 40 One can identify a number of factors that have contributed to unduly pessimistic forecasts about nuclear proliferation. In addition to a fixation on security-centric proliferation drivers, among the more important has been an excessive focus on states' capabilities. This tendency, sometimes referred to as "technological determinism," is reflected both in early U.S. NIEs and in more academic analysesY Although long out of fashion, at least in its most primitive form, reliance on estimates of technological capability for the purpose of forecasting proliferation has reappeared under the increasingly popular guise of nuclear latency. This development, fueled in part by concerns about how soon Iran might acquire a nuclear weapon, has highlighted a number of serious definitional and methodological problems relevant to forecasting. Confusion over what constitutes nuclear weapons "capability," "threshold," or "latent capability," for example, led media pundits and policy makers to announce at various times between 2005 and 2009 that Iran was either nearing or moving beyond the nuclear point of "no return." 42 Scott Sagan, in Chapters, and Hymans, in Chapter 6, seek to penetrate this definitional fog and provide greater clarity regarding what constitutes nuclear latency and nuclear weapons status. Sagan's probing essay points out the frequent misinterpretations and misuses of the concept of nuclear latency and the danger of either conflating capability with intent or completely divorcing one from the otherY He suggests that no "objective" estimate of nuclear latency can fit every case and argues persuasively for the need to better integrate technological and political factors in defining what constitutes a latent nuclear weapons state. Hymans focuses on the related question of when a state becomes a nuclear weapons state. 44 He reviews the definitional evolution of this threshold, questions the validity of reducing the threshold from nuclear testing to accumulation of a significant quantity of fissile material, and proposes an alternative approach for operationalizing "weapons stateness." He then tests in a very preliminary fashion the hypothesis that military-run nuclear weapons programs are more likely than those under civilian control to result in the induction of weapons into the state's arsenal without prior nuclear testing. Both authors identify policy difficulties that flow from relying

FORECASTING PROLIFERATION

11

too heavily on purely technological indicators in assessing a state's weapons status, including a tendency to make faulty-and most often, exaggeratedproliferation predictions. Unlike the broader domain of political science, the emerging subfield of nonproliferation studies has been largely devoid of quantitative research. This condition is especially pronounced with respect to the body of literature focused on proliferation determinants (as opposed to consequences). 45 Two of the contributions to this volume provide a useful corrective to this imbalance. In Chapter 7, Harald Muller and Andreas Schmidt point out significant methodological shortcomings in prior quantitative research, some of which they attribute to the authors' greater familiarity with quantitative tools than the subject matter under analysis. 46 In seeking to redress this deficit, Muller and Schmidt employ relatively simple statistical tests to determine what factors best account for decisions by states to foreswear nuclear weapons. They find that the emergence of a nonproliferation norm and its codification in the NPT, along with the nature of domestic political structure, played a decisive role in this respect. The NPT, they argue, fundamentally altered states' expectations of the costs and benefits of nuclear weapons and what constitutes appropriate state behavior. As long as this new norm persists, they suggest, going nuclear is not what most states will do regardless of changes in their security environment. This optimistic prognosis, however, is conditioned on the persistence of the nonproliferation norm, and the authors worry that failure of NPT parties to fulfill their treaty obligations could lead to the norm's erosion. In Chapter 8, Philipp Bleek, like Muller and Schmidt, regards quantitative analysis as a valuable tool for understanding why states proliferateY More so than his German colleagues, he also makes the case for employing quantitative analyses explicitly for forecasting purposes. After critically reviewing much of the relevant quantitative literature on the subject, he attempts to synthesize the best elements from prior work, develops an improved data set on global proliferation behavior for the 1945-2000 period, and employs an event history/hazard model to explore the correlates of nuclear weapons exploration, pursuit, and acquisition. His findings are often consistent with conventional wisdom on the subject but at odds with some of the conclusions from the comparative case studies in this book's companion volume. For example, Bleek finds most security variables he tests to be significantly correlated with exploration, pursuit, and acquisition of nuclear weapons. At the same time, he concludes that neither major nor regional powers are any more or less likely

12

POTTER AND MUKHATZHANOVA

to explore and pursue nuclear weapons. Most interestingly for the purpose of our volume, he finds no correlation between "rival proliferation" and acquisition of nuclear weapons-that is, no reactive proliferation. Perhaps more than anything else, the two quantitative essays illustrate the importance of the data set and parameters one employs for statistical analysis, considering how much variation there is among existing quantitative studies. Appropriately, this volume concludes with a chapter by Lewis Dunn, the coauthor of the 1976 chains study that served as a backdrop for the current nuclear forecasting project. Creative and provocative as usual, Dunn adopts a contrarian perspective, begins with a world of many nuclear powers, and then resorts to a form of reverse engineering to explore how one might reach that end state. 48 In this "Gedanken experiment," he divides states into twelve categories based on their political-military orientation and attitudes toward the nonproliferation regime for the purpose of exploring the likelihood and means by which these different groups of states might one day find themselves in the possession of nuclear weapons. As such, Dunn's chapter provides a useful baseline against which to compare the findings of the case studies in the companion volume. The chapters in this book do not yield a definitive answer to the question of how best to forecast nuclear proliferation. Indeed, a compelling, simple, and elegant theory of nuclear choice appears beyond reach, and it is not evident that the prospect for the emergence of a single complex one is any more realistic. This compendium of essays, however, does provide compelling arguments about what should be avoided as well as what must be done to improve the proliferation prognosis scorecard for the next decade. A number of the chapters constitute what Scott Sagan refers to as "intellectual minesweeping exercises," which highlight common but often unrecognized assumptions that can lead to faulty conclusions and predictions. Others offer specific suggestions for filling existing methodological gaps and theoretical shortcomings, as well as charting new areas for further research and analysis. If this collection of theoretically oriented chapters stimulates new thinking about the dynamics of nuclear proliferation and helps to inform more country-specific research in the future, it will satisfy a major objective of the forecasting project. Ideally, it also will prove useful for policy makers who often have struggled to make sense of a proliferation puzzle whose pieces do not correspond neatly to conventional wisdom about self-help in the nuclear age.

2

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION Toward a New Consensus?

Jacques E. C. Hymans

facing the study of nuclear proliferation is why there is such a wide and persistent gap between the large number of technically nuclear weapons-capable states-at present, there are perhaps as many as fifty-and the small number of actual nuclear weapons states-eight or nine, depending on how one counts North Korea. This gap has endured for many years, overcoming multiple and major shocks to the international nonproliferation norm. Figure 2.1 presents the basic picture. The top line represents states that are estimated to have the basic indigenous technical capacity to launch a successful nuclear weapons drive, and the bottom line represents states that actually have nuclear weapons. The top line would be much higher if one assumed an active black market in nuclear materials. The slow pace of proliferation has been a major surprise for international relations (IR) scholars and policy makers alike. 1 Why have many states with the technical wherewithal to acquire the so-called absolute weapon been so slow to do so? In addition, what explains the determination of the few states that have gone against the general nonproliferation tide? And under what conditions can we expect the general pattern of abstention to reverse or continue? These are the fundamental questions being tackled in this field of study. Of course, the field tackles other questions as well, including proliferation's potential consequences for strategic stability, 2 what policy tools may be most useful for promoting nonproliferation/ and the possibility of nuclear terrorism.4 Without denying the importance of such matters, this chapter focuses on the evolution of thinking about the basic proliferation puzzle. THE BASIC PUZZLE

13

14 JACQUES E. C. HYMANS

....

50

_,.

45 ~

40

--...

35

..0

E

20

z

15

Ill

I I

G)

(11

I ll

0

30

I

25

G)

)

::::1

10 5 0

~

...

...

I

/ ./

-

1~1~1~1~1~1~

_..

1 ~1~1~

2000 Year .....,. Number of nuclearcapable states Figure 2.1.

-

Number of actual nuclear weapons states

Potential versus actual nuclear proliferation.

It has not been easy for the proliferation field to home in on the basic proliferation puzzle. There were four principal factors that long held it back: (a) its strong predilection, due to its dominant policy orientation, for predicting the future rather than dwelling on the past; (b) its standard realist bias, which led to the belief that proliferation could not remain so rare for so long; (c) its vaguely defined typical dependent variable of "going nuclear," which allowed analysts to see what they wanted to see; and (d) the secrecy with which many states conduct their nuclear affairs, which also gave analysts great freedom of interpretation. Since the mid-1990s the literature has made major strides toward surmounting these problems, but old habits die hard. The first section of this chapter focuses closely on a handful of early works by William Epstein, Stephen M. Meyer, and Benjamin Frankel that represented crucial conceptual breakthroughs in the struggle toward a proper descriptive inference of the dynamics of proliferation. 5 Although not the most-cited contributions to the proliferation literature, even in their own day, they are

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

15

nonetheless milestones in the history of proliferation studies, harbingers of important turns in the basic assumptions and descriptive inferences made by most of the scholars and many of the analysts working in the area. 6 The second section of the chapter shifts to more contemporary political science work on the by-now widely acknowledged central proliferation puzzle of the small number of nuclear weapons states. In particular, it reviews attempts by scholars working inside each of the major IR paradigms to tackle this puzzle: realism/ neoliberal institutionalism/ liberalism, 9 psychological constructivism, 10 and sociological constructivism.U The chapter focuses on these books as strong representatives of each of the major IR paradigms, but it must be added that many other works have made valuable contributions as well. Having conducted this tour d'horizon, the chapter concludes by pointing out signs of an emerging scholarly consensus on five fundamental points: (1) that proliferation has been historically rare; (2) that we cannot take the demand for nuclear weapons for granted; that (3) domestic politics and (4) identity considerations play crucial roles in shaping proliferation choices; and (5) that theory-guided, in-depth comparative case studies are the most appropriate means of advancing the state of our knowledge at this time. EARLY MILESTONES IN THE STUDY OF PROLIFERATION William Epstein, The Last Chance

Concern about proliferation began rising steeply after France and then China joined the nuclear "club" in the 196os, 12 but it was only with the Indian "peaceful nuclear explosion" of 1974 that a real sense of potential worldwide crisis emerged-and, with it, a critical mass of serious writing on the issue. Epstein's The Last Chance was one of the most original contributions to the post1974 proliferation literature_B Epstein's first sentence set the tone: "For the first time in a quarter of a century of working with the problems of the arms race and arms control, I am beginning to get scared." 14 The superpower arms race was bad enough, Epstein wrote, but India's test threatened to open up a catastrophic series of regional nuclear arms races the world over. The rise of nuclear terrorism also was plainly foreseeable. However, Epstein contended, all was not lost. The NPT represented a good basis for nonproliferation diplomacy, and it was still alive-barely. Although the superpowers had essentially neglected the treaty ever since its signing in 1968, Epstein argued that if they were to reinvest themselves in it, the world might still have one "last chance" to save itself from nuclear bedlam.

16

JACQUES E. C. HYMANS

According to Epstein, the NPT had been negotiated within the following broad parameters. There were three points on which essentially all states agreed: First, the spread of nuclear weapons-both "horizontally" to new countries and "vertically" among the existing nuclear weapons states (NWS)-was a bad thing; second, the spread of civilian uses of nuclear energy was a good thing; and third, it was devilishly difficult to separate civilian and military uses of the atom. In addition, there were two more points on which at least most states could agree: First, codifying the existence of two different classes of states was extremely problematic under international law; but, second, realistically the members of the "nuclear club" were unlikely to give up their weapons anytime soon. These basic parameters provided enough room for a broad, multilateral agreement on how to tackle the proliferation issue. At its heart, Epstein wrote, the NPT was an exchange of promises: on the one hand, a promise by the non-nuclear-weapons states (NNWS) not to acquire nuclear arms and, on the other hand, a promise by the NWS to minimize the negative security, political, and economic consequences of that nuclear abstention. Thus, while Article II forbade NNWS from getting the bomb, in Article VI the NWS promised to work toward nuclear and general disarmament. Moreover, while Article III commanded the NNWS to accept technical safeguards against proliferation instituted by the International Atomic Energy Agency (IAEA), Article IV commanded the NWS to assist NNWS's legitimate desires to acquire civilian nuclear technology and materials. Indeed, Article IV protected the right of the NNWS to conduct scientific research on any subject in the nuclear field-even sensitive, "dual-use" research on the nuclear fuel cycle. Only "peaceful nuclear explosions" (PNEs) for engineering and mining purposes were deemed off limits for NNWS to dabble in, but in Article V the NWS even agreed to conduct PNEs on behalf of NNWS who expressed an interest. Finally, there were two escape clauses: Article X allowed any state to unilaterally withdraw from the treaty on just three months' notice, and it also limited the treaty's duration to twenty-five years, unless a majority of NPT parties voted at that time to extend it. Epstein argued that because the NNWS were generally satisfied by this bargain, they agreed to join the treaty in relatively large numbers. By the time of the first NPT Review Conference in 1975, the treaty boasted ninety-six signatories, including West Germany, the state whose nonproliferation commitment the two superpowers had been most anxious to lock in. But Epstein also argued that since the negotiation of the treaty, the NWS had not lived up to

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

17

their side of the bargain. Their provision of aid and technical assistance for civilian nuclear energy in developing countries was "pitifully small"; 15 they quickly took off the table any notion of actually conducting a PNE on behalf of a NNWS; the security guarantees they were willing to extend to non allied NNWS were paper thin; and, worst of all, soon after the NPT was signed, the superpower arms race truly kicked into high gear. In short, quoting one of the American negotiators, the NWS treated the NPT as "one of the greatest con games of modern times." 16 The NNWS' only consolation was that the NWS (along with the advanced NNWS like West Germany and Switzerland) were still too interested in promoting the commercial possibilities of nuclear power to impose an onerous system of export controls, technical safeguards, and international inspections on their customers. This was cause for even more lamentation by Epstein, as, paradoxically, having first provided the NNWS substantial motive for breaking out of the NPT, the NWS were also proving more than willing to sell them the means to do so. Such was the international context of India's 1974 test, the spark that could ignite a global proliferation wildfire. With the benefit of hindsight, we now know that the Indian test did not ignite a proliferation wildfire. Why not? Some authors have argued that the major states responded to the Indian test just enough to head off a proliferation catastrophe. While they still failed to provide the NNWS with the carrots promised in the NPT, at least their newly formed "nuclear suppliers' club" did start wielding real sticksY For instance, in 1978 the United States started to demand adherence to countrywide ("full-scope") safeguards in exchange for provisioning NNWS reactors with highly enriched uranium (HEU) fuel, and it convinced other nuclear suppliers to abandon sales of fuel reprocessing facilities to Pakistan, South Korea, and Taiwan. Clearly, the 1970s did witness some supply-side nonproliferation successes. But, on the other hand, some nuclear suppliers strenuously resisted abandoning prized clients; for instance, West Germany was willing to endure the worst turn in its relations with the United States since 1945 to follow through on providing a complete nuclear fuel cycle to non-NPT member BraziJ.1 8 And despite the gradual tightening of formal nonproliferation rules over the 1970s and 1980s, a good deal of sensitive nuclear exporting persisted, even to states such as Iraq and Pakistan that were actively seeking the bomb. The persistence of this trade reflected not only the commercial interests of unscrupulous individuals and companies but also a willful negligence on the part of numerous

18

JACQUES E. C. HYMANS

Western states. 19 Thus, traditional supply-side explanations for slowness of proliferation after 1974 seem at best incomplete. Epstein himself actually did not place the emphasis on supply-side controls. Rather, his principal claim was that the NWS' continuing failure to deal with the real political gripes of the NNWS would inexorably lead to the regime's breakdown. As he put it, "The credibility and viability of such a system of safeguards depends on the moral climate and on the strength of the entire nonproliferation regime." 20 It was impossible to completely prevent cheating; what was necessary was to remove the incentives to cheat. The lack of respect the NWS showed for their NPT commitments hardly removed those incentives; indeed, "the developing NNWS, on the whole, felt that they had been cheated." 21 Epstein thought that, for the NPT to survive, it at least had to seem to be a fair deal. But although in subsequent years the NNWS did not stop feeling cheated, the general pattern of nuclear abstinence remained. Indeed, as Stephen M. Meyer showed, the choice for nuclear abstinence was even more widespread than people like Epstein believed. Stephen M. Meyer, The Dynamics of Nuclear Proliferation

Meyer's book was the first sustained effort to apply modern political science analysis to the subject of proliferation. As he wrote in the preface, "The objective of this book is to take one step back-ignoring the policy disputes-and undertake a rigorous and systematic examination of the assumptions and contending hypotheses that constitute contemporary thinking on nuclear proliferation." 22 The result of taking this step back was a giant leap forward, although the nonacademic policy analysts who still dominated the field generally did not recognize the full import of Meyer's contribution. Meyer's argument began with some careful distinctions between possible dependent variables in the proliferation field. He argued that we should analytically separate four such variables: (a) the development of a latent nuclear weapons capacity, whether this happened on purpose or as an unintended byproduct of other choices; (b) the conscious "capability decision" to generate a certain level of nuclear weapons breakout capacity; (c) the "proliferation decision" to try to get a functional nuclear arsenal; and (d) the actual acquisition of a functional arsenal. He further argued that variable (c) generally represented the most important bottleneck on proliferation. 23 To explain proliferation decisions, Meyer developed an explicit test of "technological imperative" hypotheses versus "motivational" hypotheses. "Technological imperative" hypotheses contend that "governments 'decide' to

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION 19

go nuclear because the technology is available, thereby making the technical/ financial costs manageable and the opportunity irresistible." 24 By contrast, "motivational" hypotheses counter that "nuclear weapons do not generate spontaneously from stockpiles of fissile material."25 To go nuclear requires making a tough political choice, and that choice will reflect the top leadership's assessment of its likely costs and benefits. Moreover, there is no reason to suppose a priori, as "technological imperative" analyses do, that the benefits of having the bomb outweigh the costs. Meyer's empirical test of the "technological imperative" hypotheses was groundbreaking. Previous work on proliferation had focused mostly on "near-nuclear" states like Germany, Israel, or India, whose advanced nuclear infrastructures placed them on the cusp of a nuclear weapons arsenal if they desired one. But Meyer pointed out that such analyses unfairly tipped the scales in favor of "technological imperative" hypotheses because many states developed their nuclear infrastructures precisely to have that nuclear weapons capability. Instead, Meyer tested the "technological imperative" hypothesis using a new model of "latent capacity" to build the bomb within six years or less, which did not rely primarily on estimating a state's present level of nuclear technology. Rather, Meyer's "latent capacity" model placed most emphasis on broad indicators of scientific and industrial accomplishment-things that could be placed at the service of a nuclear program, if the state so chose (see Table 2.1). Meyer's solution to the endogeneity problem that had dogged the literature up to that point is now the standard starting point for analyses of proliferation capacity. Not surprisingly, Meyer's broader approach to the technical capacity question added several new states to the list of potential proliferants, and it also generally advanced the dates on which states were estimated to have attained that capacity. Thus, although President Kennedy ominously warned in 1963 that at least fifteen states would have the capacity to build the bomb by 1975, actually, at the time Kennedy made that claim, according to Meyer's calculations eighteen states already were in a position to build the bomb by 1970! The "technological imperative" camp might be frightened by this news, but Meyer argued that it paradoxically showed the low likelihood of a proliferation wildfire. If so many states had been capable of going nuclear for such a long time, and yet so few had done so, the proper conclusion to draw was that the "technological imperative" was no imperative at all. Meyer's careful study of the relationship between technology and proliferation thus pointed to the primacy of politics, or, as he called it, the "motivational

20

JACQUES E. C. HYMANS

Table 2.1.

Meyer's latent nuclear capacity indicators.

Previous national mining activity Indigenous uranium deposits Metallurgists Steel Construction workforce Cement Chemical engineers Nitric acid Electricity production capacity Nuclear engineers, physicists, chemists Nuclear graphite production capacity Electronics specialists Explosives specialists SouRcE: From Table 26 in Meyer, The Dynamics of Nuclear Proliferation, 192.

hypothesis." The question then became, What specific factors might produce motivations for or against going nuclear? Few had attempted to think the question through systematically. Meyer combed the existing case-study literature to produce a list of general factors that should be either "motive" or "dissuasive" for most states when confronting the question of whether or not to get the bomb. He came up with a long list of potential independent variables (see Table 2.2). Meyer's list has two interesting features that are particularly striking in light of the fact that he wrote his book at a time of overwhelming neorealist dominance in the IR field. First, Meyer found no clear theoretical valence for the hard-core "realist" variables that are still the most commonly cited reasons for proliferation. Actual or latent nuclear threats or overwhelming conventional threats might cause states to respond with a nuclear drive of their own, but they might also cause states to lie low. Meanwhile, having a nuclear ally might calm states down about needing to deter others themselves, but it might also cause states to seek the bomb to escape sinking into the status of a semiindependent protectorate. In short, for Meyer, neorealist variables were theoretically indeterminate. The second interesting feature of this list is that most of what Meyer listed as theoretically clear motive factors for proliferationregional and great power status or pretensions, pariah status, and the need to save face after loss in war-relate more to what we would today term

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

Table 2.2.

21

Meyer's nuclear predictor variables. Variable

Type

Threatening nuclear arsenal

Both?

Threatening latent nuclear capacity

Both?

Overwhelming conventional threat

Both?

Nuclear ally

Both?

Domestic turmoil

Motive

Intolerable defense expenditure burden

Motive

National self-image (regional or global status or pretensions, or desire to recover self-esteem after loss of a war)

Motive

Pariah status

Motive

Regional nuclear proliferation

Motive

Legal treaties in force

Dissuasive

Peaceful reputation

Dissuasive

Possible nuclear preemption

Dissuasive

Risk of unauthorized seizure

Dissuasive

SoURCE: Adapted from Table 8 in Meyer, The Dynamics of Nuclear Proliferation, 74.

constructivist, "identity" variables than they do to the materialist calculus still favored by most analysts. More recent contributions to the proliferation literature suggest that Meyer's instincts on these issues were very good. Meyer's empirical testing suggested the importance of all of the hypothesized motivational factors. From the perspective of 2010, however, it must be said that the book's statistical methodology leaves much to be desired. In recent years, other quantitative analysts with vastly enhanced computing power at their disposal have done analyses similar in spirit to Meyer's. For instance, Dong-Joon Jo and Erik Gartzke largely seconded Meyer's conclusions. 26 But Sonali Singh and Christopher Way found overwhelming support for the variable of "enduring rivalries"-which neither Meyer nor Jo and Gartzke had tested-as the key motive factor. 27 These very sophisticated quantitative analyses are a clear step beyond Meyer's initial forays. But the differences in their results may reflect a more basic problem for quantitative attacks on this question: the lack of a reliable data set on which to base worldwide statistical tests. For instance, Jo and Gartzke do not code Libya as ever having had a nuclear weapons program, whereas Singh and Way code it as having had one consistently since 1970, and Meyer codes it as definitely having had one only from 1970 to 1975. The disagreements over the Libya case reflect a more general

22

JACQUES E. C. HYMANS

uncertainty about the historical record. Indeed, the coded start or end dates of most country programs are different across the data sets. 28 In light of this confusion about the basic historical facts, it may be premature to attempt large-N analyses of the proliferation phenomenon. 29 Meyer's clear identification of the basic proliferation puzzle was an essential contribution to the field. The next step for the literature might logically have been to dig deeper into the question of motivations to go nuclear and to develop a genuine explanatory theory of nuclear proliferation; 30 but, before that could happen, there was a theoretically sophisticated assault on the relevance of the basic proliferation puzzle that Meyer had identified. Benjamin Frankel, ed., Opaque Nuclear Proliferation

Contrary to Meyer's generally positive story of general proliferation restraint, Frankel and his coauthors painted a much darker picture. 31 They argued that the Israeli case of "opaque proliferation" was ideal-typical of an ongoing second generation of nuclear proliferation, whose magnitude was being sorely underestimated due to its opacity. The implication was that, like Hegel's Owl of Minerva, scholars of proliferation had arrived at the basic puzzle of why there were so few nuclear weapons states at the very moment that the historical pattern was becoming obsolete. The lead piece in the volume, by Frankel and Avner Cohen, argued that opaque proliferation was a response to the NPT's success in delegitimizing nuclear weapons desires. 32 Quite surprisingly, given their neorealist moorings, they accepted that the nonproliferation norm existed and that it was powerful. But, they argued, it would be overly optimistic to think that states' recognition of the norm meant real respect for it. As Cohen and Frankel wrote, "This does not mean that nations have lost interest in the acquisition of nuclear weapons, but rather that nations cannot voice this interest publicly in the international arena." 33 The norm had simply pushed nuclear weapons activities underground. (Cohen and Frankel did not specify exactly when or how the international nonproliferation norm had become entrenched, and therefore it is difficult to evaluate their hypothesis that the norm caused Israeli nuclear opacity, the center of their empirical concerns. The book's chapters specifically focusing on Israel actually largely dropped this international norms hypothesis in favor of an explanation focusing on Israeli domestic political culture and U.S. diplomatic pressure.) The most notable implication of the logic of opaque proliferation was the demise of the testing of nuclear explosive devices as "a clear-cut and visible

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

23

criterion for recognizing when and how the nuclear threshold had been crossed." 34 In addition to the rule of "no tests," Cohen and Frankel argued that opaque proliferation also involved denial of possession, no direct threats, no military doctrine, no military deployment, no open debate, and organizational insulation. 35 Cohen and Frankel argued that these behavioral changes had important epistemological implications. The international community thought it "knew" that, thanks to the NPT, proliferation was under control: There were the five NWS on the one hand and 175-plus NNWS on the other. But, in fact, this "knowledge" was based on who had or hadn't tested; yet, in the new world of opaque proliferation, testing was an unnecessary luxury that most proliferants would simply bypass. True, there were still not many "opaque" casesIsrael, Pakistan, India, South Africa, and perhaps a handful of even more secret programs-but, Frankel argued, this general pattern of abstention was because of the superpower nuclear arms race, not despite it as NPT advocates believed. 36 Now that the superpower arms race was over, a new era of horizontal proliferation was beginning. In this new era, while not directly challenging the NPT, states would simply submerge it in a sea of new de facto nuclear arsenals. The opaque proliferation idea wormed its way into U.S. policy after the shocking 1991 discovery of Iraq's gigantic and formerly unknown effort to enrich uranium. In response to the Iraq revelations, Washington did not merely devote more resources to proliferation intelligence and diplomacy; it also changed its methodology of proliferation assessment. Traditional proliferation assessment had used such inductive techniques as visual identification to judge the progress of nuclear weapons efforts. But, over the course of the 1990s, the intelligence community gradually developed a "capabilities-based" approach. The "capabilities-based" approach, evident in extreme form in the influential Rumsfeld Commission report on ballistic missile threats, took for granted that proliferation would remain opaque; therefore, to compensate for the lack of reliable information, it extrapolated estimates of states' strategic weapons from Meyer-type latent capacities. 37 The result of this assessment methodology was a "technological imperative" hypothesis on steroids. Worse yet, "capabilities-based" assessments were unfalsifiable-or at least, unfalsifiable until after a country had been invaded, and every stone from the rubble of its real or imaginary nuclear program carefully examined. Washington's deadly mistakes notwithstanding, the opaque proliferation concept could still be a valid one. To evaluate this, we need to answer the

24

JACQUES E. C. HYMANS

following question: Is the Israeli experience indeed the ideal type of a much broader category of cases, or is it in fact still an outlier? Actually, the past fifteen years have witnessed Cohen and Frankel's other signal cases of opacity falling by the wayside. Almost as soon as their book was published, South Africa destroyed its nuclear devices and went openly nonnuclear. A few years later, both India and Pakistan tested bombs and declared their nuclear weapons status. Meanwhile, some of the other suspected cases of "opaque proliferation," such as Libya and Iraq, turned out to have simply failed to reach the finish line; while others, such as Argentina, turned out never to have made the crucial "proliferation decision" in the first place. 38 Moreover, after the breakdown of the Agreed Framework, North Korea adopted precisely the opposite strategy, loudly proclaiming its nuclear capability several years before it finally conducted a first test in 2006-whose paltry results hardly justified U.S. estimates that the country had become a nuclear power already in the early 1990s. 39 In sum, although we cannot say for sure what various states may be doing behind the back of the IAEA, it would seem that Israel's path to the bomb remains a sui generis one-a point increasingly recognized even by Cohen himself. 40 Thus the NPT's traditional bright-line distinction between NWS and NNWS appears still relevant to today's worldY With Cohen and Frankel's challenge blunted, Meyer's fundamental puzzle of the gap between the nuclear capable and the nuclear armed returns to the fore. And, not surprisingly, over the past decade a number of works have taken up Meyer's question. RECENT STUDIES OF PROLIFERATION

What causes states to seek nuclear weapons? In the mid-1990s, Scott Sagan published a now-classic article outlining three basic "models" of proliferation intentions that could be discerned from a close reading of the idiographic case study literature. 42 Sagan's "security model" followed the typical "defensive realist" point of view that nuclear proliferation occurs in response to significant external threats; his "domestic politics" model saw proliferation as a political payoff to powerful domestic electoral or bureaucratic constituencies; and his "norms model" explained it as the product of a state's quest for acceptance as a "legitimate, modern" member of international society. Sagan's article was a halfway house: It went beyond the old-style literature's lists of potentially motive or dissuasive variables, but it was not yet a genuine theory of proliferation with clear microfoundations linking cause to effect and with multiple,

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

25

testable hypotheses. Still, Sagan (in addition to others such as Davis and Frankel, Flank, Lavoy, Reiss, and Solingen) had pointed the way forward.H Scholars took up the challenge and strove to build and test genuine theories of proliferation. The new literature on proliferation reflects the diversity of theoretical paradigms available to scholars working in IR today. By now we have strong examples of realist, neoliberal institutionalist, liberal, psychological-constructivist, and sociological-constructivist theories of proliferation to choose from. While these works diverge on many points, in many respects they actually dovetail quite nicely together. This section of the chapter summarizes and critiques recent contributions from each of the five paradigms listed above. The Realist Take: T. V. Paul, Power Versus Prudence

Breaking with standard realist perspectives on proliferation (for example, Kenneth Waltz or Colin Gray), 44 T.V. Paul depicts the question of going nuclear as a great dilemma for states. On the one hand, Paul accepts the standard realist point of view that acquisition of nuclear weapons provides a state with more military and political power; but, on the other hand, he points out that it also threatens to destabilize relations with the state's neighbors. 45 Therefore, Paul contends that the cardinal realist value of prudence actually generally commands nuclear abstinence. However, some states can be expected not to act "prudently." First, great, or aspiring great, powers simply cannot forfeit the chance to own a weapon with such a fundamental impact on the structure of international polarity. Second, nongreat powers engaged in "enduring rivalries" and not protected by a superpower guarantee or by the knowledge that their enemy is incapable of going nuclear will go for the bomb, because such states cannot assume that any nuclear restraint they show will be reciprocated.46 Having developed these hypotheses on proliferation choices, Paul then tests them on a wide range of fourteen historical cases of both nuclear and nonnuclear weapons states, with cases selected to ensure variation on their placement within the international power structure and the Cold War alliance system. He claims that the test overwhelmingly confirms the model. Paul's work represents a major step forward for realist approaches to proliferation. Some of its most important contributions are, first, the frank acknowledgment that proliferation has been rare and that this is a puzzle with which realism needs to come to grips; second, the refusal to fall back on the typical realist defense that, although proliferation has been rare in the past, it is about to explode tomorrow; and third, the clear exposition of why, even

26

JACQUES E. C. HYMANS

within the confines of realist theory, going nuclear may be a very unattractive option for most states. But, on the other hand, while the book's explanation of nonproliferation is convincing enough, its explanation of why some states actually do decide to acquire the bomb could be further developed. First, Paul emphasizes the importance of "enduring rivalries" in sparking proliferation drives. But the mechanisms via which "enduring rivals" turn into nuclear rivals are not precisely specified. Although "enduring rivalries" often exhibit high tendencies toward conflict spirals, it is not clear why such tendencies must also extend to the nuclear level. Indeed, from a "prudential realist" perspective, the fact that such rivals know that their relationship has this tendency might be expected to make them even more prudent than normal states before launching a nuclear program. In short, although "enduring rivalry" probably does contribute to proliferation intentions, recognizing this fact should point us away from realism, not toward it. Second, Paul's expectation that "states with intense aspirations to become great powers" will seek the bomb also strains his realist framework to the breaking point. This exception to the rule of prudence does not emerge out of the logic of a well-specified theory but rather seems tacked on to capture some of the unexplained variance. Moreover, Paul neither explains nor measures these "intense aspirations." Therefore, although Paul applies this exception only to the case of India, one wonders why many other states that have apparently chosen not to go nuclear, including today's Brazil or even South Korea, could not be categorized similarly. In sum, Paul's "prudential realism" is a good start for understanding why many states have abstained from going nuclear, and the variables he points to as causing exceptions to the general rule of prudence seem plausible. A fully developed theory of proliferation, however, would have to genuinely integrate Paul's basic rule of prudence with the empirical exceptions he notes to that rule. Moreover, because of the complex sources of the "enduring rivalry" and "great power aspiration" variables in domestic as well as international political dynamics, it seems hard to imagine that such a theoretical integration could be carried out while remaining within the confines of realism. The Neoliberal Institutionalist Take: Xinyuan Dal,

International Institutions and National Policies

It is quite surprising that the renowned NPT regime has not been subjected to close scrutiny by the neoliberal institutionalist literature that has become

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

27

so dominant in the IR field over the last two decades. Dai makes a small step toward filling that gap, although in her work the NPT is introduced only as a shadow case to contrast with the weaker international institutions whose success she explores much more fullyY Dai begins her analysis with the assumption that the NPT, like many treaties, is a response to a "prisoner's dilemma" (PD)-type situation. 48 In the PD, while neither party wants to defect-meaning, here, to go nuclear-each fears that the other may well do so, and in that case it would have to defect as well. The well-known solution to this dilemma is an institution that permits each side to have confidence that the other is not defecting. However, as Dai notes, such an institution is only as good as its system for monitoring compliance; yet the neoliberal institutionalist literature has strangely shown little curiosity about how such monitoring is actually done. 49 Dai's book tries to answer this question. Dai's general answer to this question focuses, first, on states' basic level of interest in ensuring compliance by their peers and, second, on their desire to do so in a cost-effective manner. 5°For the case of nonproliferation, she contends that states view potential proliferation as a top concern and therefore want high-quality information on it; moreover, given the secrecy that enshrouds the issue, they know they are unlikely to get the information they seek from a low-cost, decentralized process. Therefore, in this case they have built a very strong, centralized information system centered on the IAEA safeguards and inspections process. This system, Dai implies, has reassured states sufficiently of others' intentions to be willing to keep their own nuclear powder dry. 51 Dai's book is a model of clarity and precision, and the brief application of her model to the nonproliferation monitoring system has a great deal of plausibility. At least in the post-Cold War era there have been considerable efforts to strengthen the IAEA safeguards and inspections regime. Indeed, in light of the second Bush administration's general disregard for international institutions, the IAEA's steady progress since the late 1990s is remarkable. But, on the other hand, the nonproliferation regime is actually not an ideal-typical case of a centralized monitoring system. In fact, the environmental and peace movements have also played a crucial role in providing information on dangerous nuclear activities, particularly in Western countries but also in many developing countriesY The IAEA has often found itself following up on leads developed by such "low-cost monitors," in Dai's parlance. Thus, rather than the pure case of monitoring by international governmental organizations (IGOs)

28

JACQUES E. C. HYMANS

as Dai portrays it, the nonproliferation regime would instead appear to be yet one more mixed case of monitoring by both IGOs and nongovernmental organizations (NGOs), not unlike the other regimes to which Dai devotes the bulk of her text. Dai 's assessment of this mixed sort of monitoring system is very optimistic; her book strongly emphasizes "the power of weak international institutions." 53 But her optimism can be questioned. Dai shares with most neoliberal institutionalists the basic assumption that more information availability automatically translates into more regime credibility. Yet what is important for regime credibility is actually not simply the raw amount of information available but rather states' confidence that the information is accurate and complete. Thus, by bringing to light facts missed by the centralized proliferation monitoring system, NGOs can actually undermine the overall credibility of the nonproliferation regime. Indeed, undermining that credibility is often the NGOs' explicit purpose because their ultimate goal in many cases is to demonstrate that the centralized monitoring system cannot be counted on-and therefore that the only solution to proliferation is to curtail the use of nuclear power altogether. So, these "low-cost" monitors, far from being complementary to the regime, may actually exact a very heavy price on it. And, indeed, in spite of the vastly increased resources that states have poured into the centralized monitoring system since the early 1990s, due to the activities of such groups as well as the advent of globalized proliferation networks and the widespread acceptance of the Cohen-Frankel "opaque proliferation" concept, international confidence in the nonproliferation regime today is at an all-time low. 54 One might retort that states nevertheless appear to have retained confidence in the NPT regime because, after all, they have not abandoned it. But this pattern of continued adhesion may in fact simply reveal the inadequacy of the most basic assumption on which Dai rests her analysis: that the proliferation "game" is a PD in the first place. This point returns us to the domestic level, for nuclear choices inevitably reflect how particular state elites perceive the bomb's value, or lack thereof. The Liberal Take: Etel Sollngen, Nuclear Logics

Solingen offers a liberal, that is, domestic politics-driven, theory of proliferation and nonproliferation. 55 The book, which elaborates on an argument originally made a decade earlier in an International Security article, 56 has been rewarded with the prestigious Woodrow Wilson Prize of the American Political Science Association. Solingen's "political survival" model focuses on

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

29

how ruling coalitions inside states have, since the institutionalization of the NPT regime, pursued their material self-interest via what she terms policies of "nuclearization" or "denuclearization." Solingen argues, Inward-looking models [of economic development] approximate necessary if not sufficient conditions for nuclear weapons programs. Internationalizing models are not necessary but likely to be sufficient for denuclearization except under two circumstances: (a) when neighboring inward-looking regimes seek nuclear weapons (or other WMD); and (b) when nuclear weapons were acquired prior to the inception of internationalizing models. 57 Solingen also notes the possibility of"compromise coalitions" between inwardoriented and internationalizing forces, which try to have it both ways, outwardly respecting NPT commitments while building up their potential for a nuclear breakout that only some coalition members truly want. 58 The bulk of Solingen's book contains nine thoroughly researched country case studies from two regions, East Asia and the Middle East. The case studies strongly undermine the realist and neoliberal institutionalist assumption that proliferation (or nonproliferation) choices are the product of unitary, rational states pursuing objective national interests. Rather, they show that different groups within states can actually have very different opinions on the value of "nuclearization." The case studies also clearly demonstrate that, when deciding where they stand on the question of nuclear weapons, political actors consider a much wider array of values than simply the state's international security and power position. Along these lines, Solingen argues that, although alliance relationships with the United States may have solidified key East Asian states' nuclear abstention, the fact of alliance should be seen as endogenous to their other political choices. In other words, East Asian states such as Japan did not stand down from nuclear weapons ambitions because the United States placated them with a credible nuclear guarantee, as most accounts suggest; rather, they actively sought that guarantee because they wanted to beat back calls for an indigenous nuclear weapons program that they viewed as potentially damaging to their broader political interests. 59 Solingen's spin on the conventional alliance hypothesis is a fine example of intellectual judo. There is much to praise in Solingen's work but also room for improvement. First, more attention could have been given to the conceptualization, operationalization, and measurement of the book's key dependent variables. In a footnote, "nuclearization" is defined as "movement toward nuclear weapons

30 JACQUES E. C. HYMANS

acquisition, even if it does not result in actual acquisition." 60 This definition is a start, but it is rather general, for instance not clearly distinguishing as Meyer did between the two types of movement "toward" the bomb-political and technological. 61 Reflecting this ambiguity, in Solingen's book the relative weight assigned to these two types of indicators varies from chapter to chapter. Meanwhile, denuclearization is defined simply as "renunciation," without further elaboration.62 The word renunciation suggests a very strong, indeed irreversible commitment to remaining nonnuclear, but elsewhere in the book "denuclearization" seems to be simply the decision not to get nuclear weapons for the time being. It is certainly true that operationalization and measurement are very difficult issues in the study of proliferation, and it would be impossible to resolve them to everyone's satisfaction. Nevertheless, without more precision on these issues it is difficult to determine whether the empirical record really confirms Solingen's theory or not. Second, Solingen actually makes rather modest claims about the power of her basic theoretical model for explaining "nuclearization." The asymmetrical nature of the book's theoretical contribution can be seen in the title of the theory chapter: "Alternative Logics on Denuclearization." In that chapter, as noted above, the presence of an inward-oriented coalition in power is merely said to be one "necessary" condition for nuclearization; 63 and, though we are not provided with a clear list of the additional conditions that are required to fully explain nuclearization, the case study chapters imply that they may be numerous. The book's suggestion that inward-oriented coalitions may, depending on other variables, adopt a wide variety of nuclear postures ranging from denuclearization to nuclearization starkly contrasts with the book's seemingly much stronger claim about the causal link between internationalizing coalitions and denuclearization. There is absolutely nothing wrong with such theoretical modesty; indeed, one of the real strengths of Solingen's work is how carefully she points out her own theory's limits. But of course the field as a whole needs to overcome those limits, particularly when it comes to such a crucial question as why some states choose to build nuclear weapons. Third, Solingen does not problematize the relationship between importsubstitution industrialization and nationalism. Therefore, she tends to take the fact that nationalist sentiments in countries like North Korea led to "nuclearization" as further indication of the fundamental impact of inward-looking economic models on states' nuclear choices. But actually nationalism and inward economic orientations are clearly separable analytically. And this leads to the question: What if they were also separated empirically? 64 Would it turn

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

31

out that tendencies toward "nuclearization" and "denuclearization" more strongly followed the pull of economic interests or the tug of identity? The case of late-1990s India-in which a strongly nationalist and economically internationalizing coalition chose to risk its economic and political survival for an overt, sustained nuclear breakout-suggests that, when push comes to shove, nationalist urges may at least sometimes get the better of economic selfinterest. Solingen properly acknowledges that the Indian case demonstrates her theory's "refutability."65 But in fact the 1998 Indian decision may well be more than a mere anomaly to be brushed off in a footnote. Given the reality of globalization, it has been hard to be anything but an economic internationalizer in recent years-Solingen suggests that even Kim Jong II may be one-yet we know that, despite this trend, nationalism and "nuclearization" are still very much with us. 66 In sum, while we should certainly take note of the material interests Solingen emphasizes, we must also follow Max Weber and attempt to identify the fundamental ideas that give rise to those perceived interests. 67 The Psychological-Constructivist Take: Jacques E. C. Hymans, The Psychology of Nuclear Proliferation

In his psychological constructivist6R approach to the proliferation puzzle, Hymans seeks to identify the basic ideational foundations for some leaders' desire for the bomb.69 Hymans begins with the contention that the decision to go nuclear is nothing less than a leap in the dark, a revolutionary decision with consequences that are wildly uncertain. 70 It is the tremendous uncertainty about what might follow from the creation of a new nuclear arsenal, Hymans contends, that explains the general historical pattern of nuclear restraint. Moreover, the existence of such a high level of uncertainty undermines the standard notion that the choice to go nuclear could follow from a typical costbenefit calculation. Thus, Hymans argues, to understand the decision to go nuclear, we must look beyond standard structural variables and instead to the identity-driven perceptions and emotions that guide leaders to make a choice in the absence of more solid indicators. In particular, Hymans argues that leaders who hold a "national identity conception" of "oppositional nationalism" -that is, who define their nation as being both naturally at odds with and naturally equal (if not superior) to a particular external "key comparison other"-are likely to express that identity by deciding in favor of nuclear weapons. By contrast, leaders holding other types of national identity conception will tend to shy away from taking that ultimate' step, probably preferring instead to engage in some level of nuclear

32

JACQUES E. C. HYMANS

hedging. 71 Hymans argues that oppositional nationalists' opposition-induced fear and nationalism-induced pride combine to produce a very strong tendency to reach for the bomb. Therefore, the arrival in power of such a leader is almost sufficient in itself to spark a clear proliferation decision, assuming that the state is engaged in reasonably intense interactions with the key comparison other, has at least some measure of experience in the nuclear field, and has placed centralized control of its nuclear establishment into the hands of the top leader. 72 Hymans further contends that once the nuclear choice has been made, powerful bureaucratic and psychological forces tend to make it very hard for the state to turn back on that original commitment.73 Hymans tests his theory against four country cases: France, Australia, Argentina, and India. The first step in the test is a qualitative and quantitative measurement of the national identity conceptions of dozens of top leaders from the four country cases. On the basis of these measurements, Hymans offers a set of hypotheses on these specific leaders' likely nuclear preferences. He then tests these hypotheses with detailed process tracing, using extensive primary source documentation. The case studies sometimes demonstrate the inadequacy of received historical wisdom about these states' basic nuclear trajectories. For instance, the book's analysis of Argentina uses archival materials to show that, contrary to conventional accounts, that state never had a nuclear weapons program. Hymans claims that all four cases highlight the importance of different individual leaders' conceptions of the national identity. When oppositional nationalist leaders came to power, they hastened to seek nuclear weapons; by contrast, leaders holding other national identity conceptions avoided making that decision. The importance of the individual level of analysis is underscored by the radically different positions on nuclear armament that were held at a given point in time by different political elites, sometimes even from within the same political party. Hymans's book has some clear limitations. First, the book tests the theory against only four country cases, hardly a definitive test. 74 Second and more problematically, although Hymans (following Meyer75 ) may be right to focus on the leader's "proliferation decision" as the key proliferation bottleneck, it surely is not the only bottleneck. Indeed, it is an established historical fact that not all states have seen such decisions through to the end. Moreover, Hymans notes that in some cases nonproliferation attitudes and commitments may become entrenched in state institutions, making the mere arrival in power of a new leader insufficient to turn policy around. Indeed, one of Hymans's key prescriptions for nonproliferation policy is that

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

33

states' nuclear decision making structures should be opened up to increase the number of veto points.76 Clearly, more analysis is needed not only of proliferation decisions themselves but also of the domestic institutional settings with which leaders must deal. 77 Third, Hymans takes state leaders' identities as given and therefore does not tackle the important further questions of how people become oppositional nationalists in the first place and why oppositional nationalists may rise to power at certain moments in history. The decision to bracket these questions may be legitimate from the perspective of theory testing-after all, there is no "independent variable" in social science that cannot also be viewed as a "dependent variable" if one so chooses-but nonetheless it seems to short-circuit consideration of some of the deeper forces that may be at play here. The Sociological-Constructivist Take: /tty Abraham,

The Making of the Indian Atomic Bomb

Abraham boldly goes where Hymans fears to tread; he reaches beyond the level of decision makers to the wider cultural context that shapes them. 78 Although Abraham's book focuses only on the single case of India (and indeed only on some aspects of that case), it is highly theorized and has clear implications for the interpretation of proliferation around the world. He has spelled out some of those broader implications in an important article published in 2006? 9 Abraham's basic claim is that at the particular world-historical moment of high modernity in the post-World War II era, the "nuclear" became a "fetish" object connoting both progress and power-the core values embraced by the nation-states that dominated that era. Therefore, it was only natural that highmodern states would seek to master the atom and in more ways than one: by utilizing its potential energy, of course, but also by claiming a state monopoly over all things atomic and by establishing atomic energy commissions with extraordinary powers. 80 Abraham further argues that so-called developing states actually bought into the basic tenets of high modernity even more fully than the metropolitan states that had originally spawned them. This was because the liminal, "postcolonial" status of most developing states made them greatly eager to retain their newly won membership in the interstate club. 81 The result was that they felt an even greater urge than other, more established states to master the atom and indeed to go all the way and build nuclear weapons. Abraham sees India, a postcolonial, developing state par excellence, as having been quite prone to those urges. What is more, unlike many other developing states, India also had the resources to implement them. Jawaharlal Nehru,

34

JACQUES E. C. HYMANS

the dominating figure in the first decade and a half of India's independence, completely embraced the ideology of modernity-and placed the atom at the center of it all. 82 The Indian nuclear estate therefore grew by leaps and bounds under his watch. Abraham admits that Nehru's nuclear dreams may have been born of good, peaceful intentions, but he insists that they nevertheless paved his country's road to hell. The initial Indian commitment to making a better world slowly receded, and as it did the atom started to be viewed more explicitly in military terms. But this was a gradual transformation, observable in the different rhetorics surrounding India's two nuclear tests-the almost apologetic half-truths about the "peaceful explosion" of 1974, followed by the crude militarism of 1998. 83 Ironically, however, Abraham writes that by the time India had finally completely resolved its internal ideational contradictions in favor of embracing high modernity, the world-historical moment of high modernity was gone; the era of postmodernity had arrived. As a result, Abraham writes, "India has demanded its right to become a nuclear power just when the atomic age has come to an end, and thus remains an outsider, a spoiler, but for reasons completely opposed to its original purpose."84 Abraham's book is bursting with insights great and small. This chapter will simply mention two that seem fundamental. First, he challenges us to recognize not only that the meaning and value of the bomb depend on the eye of the beholder but also that the supposedly core state goals of power and plenty, which the acquisition of the bomb may contribute to, are themselves not fixed in stone. Indeed, his book suggests that the really important international norms that are holding back proliferation around the world today are not norms related to the nonproliferation regime per se, but rather the global tarnishing of the high-modern value of unlimited "progress" that has been achieved by the transnational environmental movement and other bearers of the "new politics." In other words, many reject the bomb today not just because they think it would retard their national development but rather because they think that the people who want it are retarded. A second and equally fundamental insight of Abraham's is his recognition (in contrast to many constructivists) that the global level is not supreme. Politics-debates about the bomb per se and also about those broader questions of state purpose-still occur largely within rather circumscribed, national boundaries. Global cultural trends affect those debates, but domestic arenas also have their own particular dynamics. Abraham warns us that it is therefore only through profound engagement with specific national contexts

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

35

that we can hope to understand, let alone predict, the course of nuclear decision making. This epistemological stance justifies Abraham's narrow consideration of just a single country case. On the other hand, due to the alternative, postpositivist nature of Abraham's epistemology, in the end whether one believes Abraham's story about India-let alone his wider story about world-historical moments-is largely dependent on whether one is predisposed to believe it. For instance, Abraham concludes that "the 1974 explosion was the demonstration for which 'India' had unconsciously been waiting since 1948 and the formal announcement of the creation of an atomic energy program."H 5 But unconscious urges are, shall we say, difficult to assess. Moreover, the above quotation also reflects a decided tendency Abraham has to rub out the real differences between the ideas and mentalities of different Indian political actors in favor of broad-brush depictions of the seemingly inexorable flow of history. Abraham's insistence on going down deep minimizes the impact of human agency and the general unpredictability of politics. Yet throughout Abraham's tale itself we can actually find numerous examples of people who went against the grain: the parliamentarian S. V. Krishnamurthy Rao, the atomic scientist Vikram Sarabhai, and scholars such as Achin Vanaik and Partha Chatterjee all raised profound questions about the direction of the nuclear program and of the Indian state as a whole. Perhaps Abraham is right to imply that in the long run these critics had no chance of winning the argument; but the fact that India resisted the temptation to build a genuine nuclear arsenal all the way up to 1998 would seem to indicate that they were not entirely ineffective, either. THE FUTURE OF PROLIFERATION STUDIES: AN EMERGING CONSENSUS?

The state of the proliferation literature is strong today. In the past decade or so, the topic has come to the forefront of the IR debate, and the result has been a great deal of productive scholarship from all of the major paradigms in the literature. The different authors covered in this chapter of course have many disagreements-for instance, over the proper weight to assign to individual, domestic political, or international systemic variables. However, one can also perceive the faint outlines of a consensus in the literature on the dynamics of nuclear proliferation. In particular, there are five points on which most recent works converge. The first point of agreement is that proliferation has historically been a rare phenomenon. Despite acknowledging that opaque proliferation can happen,

36

JACQUES E. C. HYMANS

academic analyses today are increasingly starting from the basic premise that there are few nuclear weapons states and moreover that this is not a fluke or an aberration but rather a battle-tested empirical reality that needs to be explained. 86 Underscoring the historical rarity of proliferation is one of the most important contributions that the literature can make to the often hysterical debates of the policy world. The second point of agreement is that, while proliferation is undoubtedly a difficult technical challenge that has been made more difficult still by greater international export controls and surveillance, 87 the most significant bottleneck to the widespread development of nuclear arsenals has actually been most states' belief that attempting to build the bomb would be unnecessary or even counterproductive. To explain why the bomb has proven less attractive to states than was once feared, Paul and Dai primarily stress states' anticipation of a proliferation domino effect; Solingen primarily stresses their desire to advance their economic interests in a globalized world; Abraham primarily stresses the gradual international normative delegitimization of the pursuit of nuclear weapons; and Hymans stresses the paralyzing degree of uncertainty that is produced by all of these potential downsides and by the potential upsides as well. While the literature's disagreements over which disincentives are more important need to be resolved, all the authors agree that, for most states, the idea of building an indigenous nuclear arsenal is simply not very attractive. The third point of agreement is that the analysis of states' nuclear choices cannot bypass the domestic level of analysis. Domestic factors are central to the work of Abraham, Solingen, and Hymans, in particular. But the other strands in the literature reviewed here are also compatible with a focus on domestic factors. For instance, Dai's overall theoretical framework demonstrates a strong commitment to analyzing the interaction between international institutions and domestic-level variables, so reconsidering the NPT along those lines could be a "friendly amendment" to her book. The fourth point of agreement is the close connection that most of the authors suggest exists between national identity and the desire for nuclear weapons. Hymans's "oppositional nationalist" leaders, Abraham's "modern" states, Paul's "enduring rivals" and "great power aspirants," and even Solingen's "inward-oriented coalitions" are all doing more than simply developing a military response to a material security threat. They are also reaching for a powerful symbol whose meaning is hardly limited to its practical consequences.

THE STUDY OF NUCLEAR PROLIFERATION AND NONPROLIFERATION

37

The fifth point of agreement is that the way forward for the proliferation literature is to further develop our theories and to rigorously test new theoretical developments by using systematic process-tracing grounded in strong familiarity with the subtleties of this unique issue area. The cutting edge of the proliferation literature today tries to explicate in detail the political processes that generate different nuclear choices. The field's commitment to understanding processes also inevitably means conducting careful, systematic, comparative historical case studies that aim to increase what is still a very limited empirical foundation for political science research. Most scholars working in this field recognize that its past overreliance on declassified American intelligence or on the folk knowledge of the nonproliferation policy community long constrained its progress. 88 At the same time, the field's strong commitment to detailed case-study analysis has had the additional benefit of making this small part of the IR literature unusually comprehensible and instructive to our colleagues in the policy world.

3

DOMESTIC MODELS OF POLITICAL SURVIVAL Why Some Do and Others Don't (Proliferate)

Etel Solingen

of the International Commission on Nuclear Nonproliferation and Disarmament in October 2008, the cochairman-former Australian Foreign Minister Gareth Evans-warned that nuclear weapons could be on the verge of spreading to many new states. "We are on the brink," he said, "of ... an avalanche or a cascade of proliferation unless we are very, very careful indeed and find ways collectively to hold the line." 1 At the same meeting, former U.S. Defense Secretary William Perry argued that "if we fail to deal effectively with [Iran and North Korea] I think we are facing a veritable cascade of nuclear proliferation." Statements of this sort have become common across the U.S. domestic and international political spectra. Indeed, a September 2008 bipartisan report maintains that "given historical instability in the region, the prospects of a nuclear Middle East-possibly Iran, Saudi Arabia, Egypt, and Turkey-are worrying enough, even before the proliferation cascade continues across North Africa and into Southern Europe." 2 Increased concern with a proliferation wave, contagion, or epidemic compels a proper understanding of what exactly drives states to develop or acquire nuclear weapons. Various logics can explain the range of decisions associated with acquiring or abstaining from nuclear weapons. 3 As an early study by Meyer suggested, it is quite likely that some assumptions from different logics are valid; the task is identifying when and why different logics apply. 4 However, the same study reminded that all motives of nuclear behavior are, in the end, filtered through the domestic politics within which decisions are made. Yet, a systematic understanding of domestic effects has eluded most work in the area of nuclear proliferation in the last three decades. Country studies AT THE FIRST MEETING

38

DOMESTIC MODELS OF POLITICAL SURVIVAL

39

of nuclear aspirants often provide ad hoc depictions of the domestic scene, but these are often drawn without reference to an overarching comparative framework. Cleavages between "moderates" and "hard-liners" are often identified on the basis of inductive "who's who" analyses, lacking an underlying logic for what makes leaders, institutions, and ruling coalitions (or their opponents) "moderates" or "hard-liners." 5 Efforts to understand the deep personal, philosophical, or normative sources for moderate or hard-line nuclear choices may sometimes be helpfuland even represent progressive moves relative to earlier fixations with structural neorealism-but also entail an open-ended and protracted enterprise that must confront severe methodological obstacles and holds unknown universal applicability. 6 The pressing policy relevance of nuclear proliferation highlights the value of more discrete markers, shortcuts, or rules of thumb that might help identify the motivations ofleaders, their ruling coalitions and opponents, institutions, and relevant constituencies. This chapter focuses on a particular conceptual framework that provides, among other things, one way of making the question of "Who is likely to be a moderate or a hardliner and why?" an integral part of the explanation. That is, those proclivities are not explained by some extraneous theory but are, instead, endogenous to the general argument that I develop in the first section. This argument is designed to address the broader issue of which states are more likely to pursue nuclear weapons and why. At its heart is the contention that different models of domestic political survival-how leaders seek to gain and maintain powerprovide important information regarding nuclear decisions. The second section elaborates the scope and conditions under which the argument is expected to hold. The third section applies the general argument to the issue of proliferation chains and the conditions under which they might be more or less likely. This section thus builds on the proposed framework outlined in the preceding two sections to identify a coherent set of scenarios. The final section outlines how evolving models of political survival may portend continuity and change in nuclear trajectories. THE ANALYTICAL ADVANTAGES OF DOMESTIC MODELS OF POLITICAL SURVIVAL

The argument that leaders and ruling coalitions adopt different domestic models of political survival that are consequential for nuclear decisions was designed to address second tier, or second nuclear age, nuclear aspirants, whose

40

ETEL SOLING EN

decisions to launch or abandon quests for nuclear weapons were shaped under a particular global "world time" in place since the 196os.7 Negotiations that crystallized in the Nuclear Non-Proliferation Treaty (NPT) in the 1960s and developments toward a globalizing political economy created this new "world time" under which decisions to consider, pursue, or abandon nuclear weapons were made. 8 These decisions were inextricably linked to the models of political survival adopted by leaders and ruling coalitions to gain and maintain power. In particular, these models entailed different orientations to the global political economy and its associated economic, political, and security institutions and had different implications for nuclear choices. On the one hand, leaders advocating economic growth through integration in the global economy ("internationalizing" models henceforth) had incentives to avoid the costs of embarking on nuclear weapons programs. Leaders vary in their tolerance for domestic, international, political, and economic (including opportunity) costs entailed by nuclear weapons. What specific aspects of models emphasizing economic growth and openness to the global economy as tools of political survival made leaders more receptive to denuclearization than others?9 The answer lies in a range of incentives and disincentives, including: (a) the need to appeal to foreign investors with an interest in domestic economic growth and stability; (b) the related need to reassure neighbors to preserve regional cooperation, stability, and attractiveness to international economic actors; (c) the requirement of securing access to international markets for exports, capital, technology, and raw materials; (d) the related aversion to risking reputationallosses at home and abroad for uncertain nuclear gains; and (e) the costs of alienating domestic agents of internationalizationboth within and outside state structures-that would be adversely affected by nuclear weapons development. Clearly, there are several causal pathways linking nuclear weapons' renunciation to models emphasizing economic growth through global integration. Nuclearization burdens efforts to enhance exports, economic competitiveness, macroeconomic and political stability, and global access-all objectives of internationalizing models-while strengthening state bureaucracies and industrial complexes opposed to economic transformation. Denuclearization thus often took place as part of a broader program of internationalization designed to strengthen market-oriented forces, leaders, and institutions-state and private-favoring export-led growth. On the other hand, leaders relying on inward-looking bases of support had greater tolerance-and in some cases strong incentives-for developing

DOMESTIC MODELS OF POLITICAL SURVIVAL

41

nuclear weapons. Nuclearization entailed fewer costs for inward-looking leaders and coalitions whose political platforms were rooted in mistrust for international markets, investment, technology, and institutions. Such coalitions protected uncompetitive national industries, sprawling state enterprises, and ancillary military-industrial and nuclear complexes. 10 Nuclearization entailed considerable domestic advantages for foes of internationalizing models in inward-looking, import-substituting regimes favoring extreme nationalism, religious radicalism, or other forms of autarky, such as North Korea's juche (national self-sufficiency). Such leaders and their political allies often relied on extreme language to compel and threaten regional adversaries, wielding potential nuclear and other weapons of mass destruction (WMDs) as means to coerce and intimidate. Statements such as North Korea's repeated threats to turn Seoul and Tokyo into a "sea of fire"; Saddam Hussein's threats to incinerate Iran, Saudi Arabia, and Israel; and similar Iranian threats to Israel have certainly been rarer in domestic political contexts driven by internationalizing objectives. Most inward-looking nuclear aspirants were NPT members who not only misled the IAEA or violated their nonproliferation commitments but were also more promiscuous regarding state-directed or state-endorsed exports of sensitive nuclear technologies, fueled largely by the very structure of the domestic models that sustained them in power. Militarily sensitive exports have generally been a source of income for those inward-looking nuclear aspirants that were ill suited to otherwise accumulate resources from competitive civilian exports to the rest of the world. Thus, whereas inward-looking models might have regarded nuclear weapons programs as assets in the arsenal of building a regime's legitimacy and prestige, outward-oriented ones thwarted such latent utility. 11 From this point of view, Middle East leaders faced lower barriers to, and stronger incentives for, the pursuit of nuclear weapons than did East Asian ones. As a region, the Middle East gravitated toward the inward-looking end of the spectrum for decades, accounting for most cases of nuclearization. Most, though not all, Middle East leaders relied on models of self-sufficiency and nationalism for their political survival, and they had stronger domestic incentives to seek nuclearization. Egypt, a leader of the Arab world, can be considered an anomaly, arguably suppressing nuclearization since President Anwar Sadat launched injitah, an opening to the global economy. In contrast to the Middle East, nuclearization has been much less attractive and far more costly for most East Asian leaders since the 1970s, except for North Korea, the autarky-seeking anomaly.

42

ETEL SOLING EN

Heavy regional concentration ofinternationalizing models-economic growth via integration in the global economy-in East Asia reinforced each state's incentives to avoid nuclearization. Conversely, heavy regional concentration of nationalist, economically protectionist, and militarized models throughout the Middle East exacerbated mutual incentives to develop nuclear weapons. Identifying core models of political survival underlying the domestic politics of nuclear aspirants provides a systematic tool, portable worldwide, with premises backed by important evidenceY Both qualitative and quantitative studies validate the need to pay greater attention to the links between domestic models and nuclear choices. These models are not merely about "domestic politics" but, more broadly, about the way in which leaders define the very nature of their states' place in the global political economy and associated institutions. This very definition provides a filter through which leaders canvass external threats and opportunities, estimate the utility of international regimes such as the NPT, and formulate nuclear policies. Building on preliminary findings in earlier articles13 and subsequent, more detailed case studies along these lines, 14 a quantitative study found support for the propositions that "the process of economic liberalization is associated with a reduced likelihood of exploring nuclear weapons"; that "economic openness has a statistically significant negative effect across all three levels of proliferation" (that is, exploring, pursuing, or acquiring nuclear weapons); and that "economic liberalization dampened the risk" of states deciding "to explore seriously the nuclear option." 15 Another quantitative study found that economic liberalization had a positive and statistically significant effect on nuclear-weapons-free zone (NWFZ) treaty ratification. 16 In-depth comparative study of nine cases in the Middle East and East Asia, in Nuclear Logics,I? finds the nuclear choices of all cases to be compatible with domestic survival models. 18 A recent study refers to the connection between increasing trade openness and reduced incentives to develop nuclear weapons as a "general law." 19 The connection between models of political survival and nuclear policies since the 1960s finds support in systematic observations across different regional security contexts, diverse associations with hegemonic powers, and over successive leaderships within the same state. First, every known case of nuclear renunciation since the 1970s, where a weapons program was either entertained or launched, entailed a domestic evolution toward internationalization. Of all nuclear aspirants (under the world time stipulated earlier), not one endorsed denuclearization-fully and

DOMESTIC MODELS OF POLITICAL SURVIVAL

43

effectively-under domestic regimes that shunned integration in the global political economy. Only leaders and ruling coalitions advancing their political survival through export-led industrialization undertook effective commitments to denuclearize (Japan, Taiwan, South Korea, Egypt under Sadat, South Africa, Brazil and Argentina, Algeria, Libya since 2003). And nuclear decisions were invariably nested in a broader shift toward internationalization in economics and security. Second, where internationalizing leaders and coalitions became politically stronger, as in Japan, South Korea, and Taiwan, the departure from nuclear aspirations was sustained even as their security context deteriorated (as in the Korean peninsula and the Taiwan Straits, intermittently). The relationship between politically stronger internationalizers and the timing of rolling back nuclear ambitions was also evident in Argentina under Carlos Menem, Brazil under Fernando H. Cardoso, Spain's accession to the NPT preceding European Union (EU) membership, and South Africa, among others. Third, where leaders and coalitions favoring internationalization were weaker, as was the case historically in Argentina and Brazil until the early 1990s or in Iran until today, they were more politically constrained in curbing nuclear programs. Fourth, most, albeit not all, defiant nuclear courses have been unmistakably embraced by autarkic or inward-oriented models from Juan D. Peron's Argentina to Getulio Vargas in Brazil, Gamal Abdel Nasser in Egypt, and leaders in North Korea, Iraq, Iran, and pre-2003 Libya. Indonesia's Sukarno and, more recently, the rulers of Syria and Myanmar-all suspected of coveting nuclear weapons-fit this model as well. Fifth, even advocates of internationalizing models may have to contend with dangerous regions where neighboring leaders endorse alternative economic and nuclear policies. This problem is less intractable in East Asia, where export-led industrialization spread from Japan to most neighboring countries except North Korea, than in the Middle East. In other words, whether the regional center of gravity is internationalizing or inward looking matters for the individual calculations of states within that region. As gleaned from the experience of an overwhelming number of cases, domestic models of political survival should be considered not merely as afterthought or residual factors, as has been the case thus far, but as more fundamental-indispensable-considerations in explaining nuclear choices. Given the limits of alternative understandings of nuclear behavior/0 the lack

44

ETEL SOLING EN

of rigorous examination of political survival models as sources of nuclear postures is particularly puzzling. This omission has important implications. A "missing" or "omitted" variable may lead to an overestimation of other causal variables, granting them too large an effect on the outcome while rendering at least some of their effects spurious. 21 Without taking into account political survival models, one may not properly understand nuclear behavior or estimate the actual effects of balance of power, international norms and institutions, or democracy. This is different from arguing that such models are the only relevant variable. Nor does introducing a previously omitted variable imply that other variables are rendered irrelevant, only that we are better able to understand their relative impact on nuclear choices. Thus, political survival models help explain why security dilemmas are sometimes seen as more (or less) obdurate; why some states rank alliance higher than self-reliance while others do not; why nuclear weapons programs surfaced where there was arguably little need for them (Libya, the Southern Cone, and South Africa, among others); and why such programs were obviated where one might have expected them (Vietnam, Singapore, Jordan, and many others). Balance of power, norms, and institutions may be more (or less) significant than political survival in some cases than others; but, in the aggregate, complete explanations of nuclear behavior must include all relevant sources of nuclear behavior for any particular case. In sum, models of political survival and nuclear policies are not merely loosely associated but joined at the hip. Their omission as a significant independent variable may have led to an overestimation of other causal variables and to potential spurious effects. Their inclusion may improve our understanding of the actual effects of security dilemmas, international norms, and institutions when interacting with domestic models. DOMESTIC MODELS: SCOPE CONDITIONS

The proposition that domestic orientations to the global economy and nuclear policy may be linked is probabilistic, bounded, and refutable. It is probabilistic, as are most arguments in the social sciences, because it does not suggest an inevitable or deterministic outcome. 22 It is bounded in three ways. First, resistance to the global economy may provide only necessary but not sufficient conditions for the development of nuclear weapons programs. For instance, there is no confirmed evidence that all Middle East leaders who rejected the global economy as a favored platform of domestic political survival have also tried to develop nuclear weapons. In previous work I had placed

DOMESTIC MODELS OF POLITICAL SURVIVAL

45

Syria in the category of an anomaly given its inward-looking, nationalist, protectionist, and militarized political-economy. I had done so because, despite long-standing suspicions that Syria was interested in developing nuclear weapons, 23 the IAEA had never questioned Syria for possible violations of its NPT commitments. Yet recent evidence suggests that Syria, too-like Iraq, Iran, Algeria, and Libya-had evaded detection of its clandestine nuclear program despite allowing traditional IAEA inspections. 24 Furthermore, as Albright and Scheel suggest, unwillingness to implement the Additional Protocol may be an important indicator of increased risk of proliferation or of intentions to hide secret plutonium separation or enrichment efforts that cannot be detected through traditional safeguards. 25 By this measure, there would be less confidence to completely exclude several inward-looking regimes that have not yet signed or ratified the Additional Protocol from the list of suspect states. 26 Second, the proposition is also bounded by regional considerations. The extent to which a given region shares a congruent orientation toward the global political economy (either positive or negative) modifies domestic preferences on nuclear issues in each country. For instance, in recent decades East Asianincluding Southeast Asian-leaders and ruling coalitions generated one of the most internationalized regions in the world. This fact reinforced the individual incentives of each East Asian leader to maintain a collective trend away from nuclearization for the sake of common regional stability, foreign investment, and domestic growth (even though China had already acquired nuclear weapons in earlier decades). North Korea's economic closure made it more impermeable to the positive regional synergies operating among other East Asian states. By contrast, in the Middle East, most rulers remained committed to relatively closed political economies for many decades. 27 The disincentives to develop nuclear programs that often operate for more open political economies were thus weaker in this region as a whole. Individual leaders and their supportive coalitions who might have otherwise favored global integrationas in Jordan or Lebanon-faced an unwieldy neighborhood that actively discouraged it for many years, often through coercive measures. Third, the argument emphasizing competing political-economy models may be bound by temporal sequences in the acquisition of nuclear weapons. For instance, the incentives of a globally integrated political economy may operate more forcefully on cases where nuclear programs have not yet yielded nuclear weapons, as seems to have been the case in Japan, South Korea, and Taiwan (as well as in Argentina, Brazil, Spain, and others). Such incentives

46

ETEL SOLING EN

may have less impact once nuclear thresholds-often in the form of nuclear tests-have been crossed. 28 China developed nuclear weapons in the early 1960s, many years prior to Deng Xiaoping's decision to integrate China in the global economy. 29 North Korean leaders have yet to take the China road to political survival through export-led economic growth. China and North Korea are the only declared nuclear weapons states in East Asia (unless one includes Russia in that region). The argument regarding China is also applicable to Israel, which arguably began efforts to develop nuclear weapons in the 1950s. In neither case was the subsequent internationalization of the economy accompanied by denuclearization, but China retained a minimal deterrent, and Israel kept its formal policy that "it will not be the first to introduce nuclear weapons into the region." In most cases, including first-tier nuclear states, it seems far more costly politically for leaders and ruling coalitions to eliminate their existing nuclear weapons entirely than it would be for those who have not yet acquired such weapons to abandon steps in that direction. This would be the case even for internationalizers and most particularly for those surrounded by inward-looking neighbors, such as India and Israel. Prospect theory provides an explanation for this difference. Leaders value more what they already have ("endowment effect") than what they might get. Hence they are more averse to losing what they possess than foregoing potential future gains. 30 Accordingly, leaders may be hypothesized to accept higher risks to retain existing nuclear weapons than to retain programs leading to their potential acquisition. Moreover, the disincentives stemming from an internationalizing model may be stronger at deliberative or incipient stages of nuclear weapons consideration than after they have been acquired. In other words, one may conjecture that, when nuclearization precedes the inception of internationalizing models, subsequent denuclearization may be much harder. This expectation can be coupled with an argument related to audience costs, which may also play a role in the consideration to abandon nuclear weaponsas distinct from abandoning programs on the road to yield nuclear weapons. Audience costs, such as removal from office or no-confidence votes, are incurred by leaders when they renege on public commitments they have madeY Because domestic audiences operating in democracies both possess the legal authority to remove leaders from office and face significantly lower hurdles in overcoming collective action problems inherent in removing an incumbent from office, democratic leaders are expected to incur audience costs at a higher rate than nondemocratic leaders. 32 This argument could suggest that the costs

DOMESTIC MODELS OF POLITICAL SURVIVAL

47

of reneging on explicit, or even tacit, commitments to protect the country with a nuclear deterrent could be particularly high for democratic leaders, but they may also be high for authoritarian ones, as Weeks suggests. 33 For instance, audience costs for Chinese leaders may be high due to domestic expectations that China's rising role in world politics must continue to be backed up by nuclear weapons. Combining prospect theory and audience costs arguments yields four kinds of scenarios. The first leads to the argument that backing down from (even implicit) commitments to acquire full nuclear capabilities may be easier for autocratic leaders in countries that have not yet achieved weaponization. Audience costs in such cases are assumed to be lower for leaders who must back down from a program rather than a realized nuclear weapons capability. These conditions might enable would-be internationalizers to step down from a nuclear weapons program before it comes to fruition. Iran under an internationalizing leadership-a configuration that is quite hard but less impossible to envisage after the events ofJune 2009 in the longer run-would match these conditions. Libya may be an example of this scenario as well. Second, and by contrast, leaders of a democracy that is already in possession of nuclear weapons (or assumed to be so) may perceive domestic audience costs to be too high for advancing denuclearization. Even internationalizing Indian and Israeli leaders may find themselves in such a position, given the regional domination of inward-looking models. Domestic audience costs may be further heightened in these two cases because both India and Israel are democracies surrounded by nondemocracies. Following arguments in democratic peace theory, this could suggest that their publics might be particularly distrustful of denuclearizing when facing nondemocratic regional adversaries. Third, internationalizing leaders in possession of nuclear weapons in autocratic contexts may be as constrained as in the previous scenario. Chinese leaders, for instance, must compensate inward-looking constituencies (such as the military and other nationalist forces) that have been less than favorable to China's progressive internationalization. The audience costs that might be incurred by abandoning nuclear weapons-given an existing "endowment"might thus be prohibitive. The fourth scenario entails a democracy that is not yet in possession of nuclear weapons but has been in the process of developing them. Here, the presumably higher audience costs for a democracy might be offset by the fact that it is prepared to abandon only a program rather than actual weapons. 34 Leaders

48

ETEL SOLING EN

of internationalizing democracies may have stronger incentives to abandon such programs than leaders of inward-looking democracies. Historically, this has been the case with Sweden, Switzerland, Italy, Norway, and others. Audience costs arguments can also lead to counterintuitive claims, as analyzed in Schultz, such as the "Nixon goes to China" phenomenonY Under some circumstances, "backing down" or reneging on commitments may be more likely to ensure a leader's political survival than "standing firm." Wolf, for instance, suggests that reneging does not always lead to the imposition of audience costs and that these costs vary depending on whether or not leaders are "hawks" or "doves." 36 On the one hand, backing down may be interpreted by domestic "doves" as a sign of prudence rather than incompetence. On the other hand, because of their greater credibility in matters of national security, hawks may be able to back down without facing severe audience costs. As the discussion in this section makes clear, the domestic political survival argument is only probabilistic. Internationalizing leaders may embrace nuclear weapons, and inward-oriented leaders may decide to abandon them, 37 contingent on the relative incidence of one model or another throughout the region, the presence of a nuclear weapons program as opposed to actual nuclear weapons, and the nature of audience costs in democratic and autocratic regimes. Models of political survival may not capture all the correlates of nuclear preferences (no theory can) and are, after all, only ideal types or conceptual constructs. As such they need not fit every case or indeed any particularly case completeli 8 but rather provide a heuristic, a helpful shortcut, and a comparative, portable framework capable of reducing complex reality-and all cases-down to some fundamentals. Models of political survival can explain a number of things: (a) why different actors within the same state vary in their approaches and preferences regarding nuclear policy; (b) why nuclear policies within states may vary over time as a function of the relative power of particular domestic forces; and (c) why different states vary in their commitments to increase information, transparency, and compliance with the nonproliferation regime even when their external landscapes remain unchanged. This heuristic also provides a different foundation for the design of positive and negative inducements to encourage denuclearization than those conceiving of states as unitary actors. It is of course possible that, even if one finds this (or any other) approach reasonably persuasive in explaining the past, it does not necessarily follow that it will also apply in the future. Different dynamics at work could trigger

DOMESTIC MODELS OF POLITICAL SURVIVAL

49

conditions under which internationalizing models may no longer provide sufficient conditions for continued denuclearization. As Campbell and his coauthors suggested: ... there is widespread concern that the calculus of incentives and disincentives has shifted during the past decade, with incentives increasing and disincentives declining. New threats have arisen while the nuclear taboo has weakened. And it is not just a single factor in this new strategic landscape that gives pause. Rather, it is the accumulation of multiple factors and their interplay and mutual reinforcement that account for many of these new dangers. 39 Nonetheless, the framework proposed here provides a roadmap for considering the conditions under which its expectations might be corroborated or refuted. This in itself is a significant advantage because various extant frameworks suffer from indeterminacy, tautologies, or post hoc-ism. Neorealism's concepts of self-help and relative power, for instance, can drive states to a wide array of nuclear choices, from straightforward acquisition of nuclear weapons to alliances to renunciation and myriad possibilities in between. Its tenets are thus indeterminate, rendering them hardly a reliable guide to anticipate what states might actually do. Anomalies for neorealism habitually require additional information unrelated to international power balances. As Betts argues, insecurity is not a sufficient condition for acquiring nuclear weapons; many insecure states have not. 40 In other words, such theories are afflicted with the problem of multifinality, suggesting many outcomes consistent with a particular value of one variable. 41 To this day, it is unclear what the precise underlying measures of relative power are that should lead to nuclearization or abstentionY Yet such clarity is a sine qua non to avoid circularity and ex post facto rationalizations (such as, "state x went nuclear because of acute insecurity," whereby the acuteness threshold is detected by a nuclear test). Arguments about nuclear chains must be cast in falsifiable terms and enable more clearly defined and testable propositions. Some of neorealism's deficiencies can be illustrated through the case of Egypt. By any (structural) neorealist account-where the domestic nature of states does not matter and only "relative power" between or among adversaries does-Egypt would have been a "most likely case" for going nuclear. First, it was geographically adjacent to a presumed nuclear weapons state, Israel. Second, over the last few decades it also faced other Middle Eastern states seeking nuclear weapons in its neighborhood, such as Libya in the immediate

50

ETEL SOLINGEN

vicinity as well as Iraq and Iran, and the prospects of additional ones. Third, it lacked conventional superiority over some of those potential adversaries. Fourth, it inhabited a multipolar environment that neorealist predictions have unequivocally associated with nuclearization. Fifth, it lacked a hegemon explicitly providing a nuclear umbrella. Yet, changes in relative power do not provide coherent accounts of Egypt's denuclearization. Presumed threats from Israel and other regional rivals with nuclear ambitions arguably remained when Egypt rejected "reactive proliferation" since the 1970s. Egypt under Nasser considered nuclear weapons when its conventional gap with Israel was the narrowest and abandoned nuclear aspirations under Sadat, when the gap widened. Egypt considered nuclear weapons when it enjoyed stronger external (Soviet) security guarantees but abandoned them in their absence, as the United States never provided Egypt with equivalent guarantees. The fact that Egypt did not, thus far, acquire nuclear weapons deals a serious disconfirming blow to neorealism. A test of a particular theory where there is a rather close fit between that theory and the actual outcome-but no corresponding fit between the expectations of other theories and that outcome-provides strong corroboration for that particular theory. Egypt seems to be such a case. Tracing Egypt's nonnuclear status to a domestic political survival strategy seems a risky prediction. Yet no other leading theory predicts that status. Indeed, most predict exactly the opposite. Instead, both the early efforts by Nasser to develop a nuclear weapons program and the subsequent shift by Sadat to renounce nuclear weapons were policies compatible with the respective models each leader relied on to maintain himself in power. Nuclear decisions thus followed changes in domestic political strategies that, in turn, had led to changes in Egypt's strategic alliances and regional policies. Whereas Nasser's Egypt thrived in an aura of inward-looking self-reliance, hypernationalism, and military-technical prowess, Sadat's emphasis on economic growth, foreign investment, exports, military conversion, and a new relationship with international markets and institutions did not leave much room for an expensive nuclear program. Egypt was able to retain its nonnuclear weapons policy despite a strong domestic revisionist current advocating nuclearizationY Alleged signs of possible revisions in Egypt's policy have been strongest as the threat of Iran's nuclearization grew higher. The likelihood of this turn is, once again, hard to estimate given unclear thresholds for what kinds of changes in relative power yield decisions to acquire nuclear weapons. If indeed

DOMESTIC MODELS OF POLITICAL SURVIVAL

51

Egypt embarks on a weapons-related program in the future, neorealist perspectives would need to establish that: (1) Iran was a strategic threat to Egypt as a state rather than a political one to Egypt's leaders; (2) Iran was a more serious threat to Egypt's security than Israel has been (because dramatic steps would have been taken only after the rise of Iran as a threat and not vis-a-vis Israel for decades); and (3) that the causal mechanisms and consequences of changes in relative power are independent from domestic considerations, such as pressures for matching Ahmadinejad's defiance among Egypt's inwardlooking constituencies. This would have far more to do with domestic strains in the existing model of political survival, and perhaps with concerns with Shi'a-Sunni cleavages, than with any fear that Iran might threaten Egypt's territorial integrity. Scenarios

In the spirit of providing hypotheses applicable to the future that are cast in falsifiable terms and offer testable propositions, Table 3.1 suggests four possible scenarios for the application of models of political survival to 21st-century proliferation trends. Two of these scenarios are compatible with the premises of the framework analyzed so far. The other two scenarios falsify those premises. The horizontal axis in Table 3.1 refers to the two basic models, internationalizing and inward-looking. The vertical axis maps two basic trends, toward nuclearization and away from it. A more detailed elaboration of this schema could help assign more specific probabilities to each scenario. Scenario 1 suggests a situation where leaders continue to steer internationalizing models in their respective countries and, at the same time, retain commitments to denuclearization. This joint outcome would be compatible with the framework's expectations. This scenario matches the reality of most of East Asia in the early 21st century and has a reasonable likelihood to persist, provided most enabling central features remain in place, including regional Table 3.1.

Models of political survival and nuclear outcomes: Four scenarios. Model of regime survival

Nuclear outcomes

Internationalizing

Inward-looking

Denuclearization

Compatible (reasonably likely)

Anomaly (reasonably unlikely)

Nuclearization

Anomaly (reasonably unlikely)

Compatible (reasonably likely)

52 ETEL SOLINGEN

and global conditions-economic and political-propitious for these models' survival. This scenario is supported, among many other considerations, by the presence of some 28,ooo Japanese companies employing over a million workers in China as of 2005, double the number merely a decade earlier, and of over a million Taiwanese entrepreneurs operating in the mainland. Former Japanese Ministry of Foreign Affairs official Kaneko Kumao draws attention to another requisite for the continuity of the postwar model of political survival incepted by Yoshida Shigeru: Japan maintains cooperative nuclear agreements with six countries, the United States, Britain, France, Canada, Australia, and China. I personally negotiated ... most of these .... If Japan misuses its civilian nuclear program for military purposes, a set of stringent sanctions will be imposed on it, including the immediate return of all imported materials and equipment to the original exporting country. Should that ever happen, nuclear power plants in Japan [would] come to a grinding halt, crippling economic and industrial activities. It is simply unthinkable that the nation would be willing to make such a heavy sacrifice-unless it [was] really prepared to start a war. In this sense, the bilateral nuclear energy agreements provide a rather effective deterrent, certainly more effective than the NPT:4 Scenario 3 entails the continuity of internationalizing models accompanied by discontinuities in nuclear policies. In other words, internationalizers go nuclear, which would constitute an anomaly for the main argument proposed in this chapter. This may be less likely under the current circumstances of a strongly internationalizing East Asia functioning as the locomotive of an expanding global economy. However, should some leaders backtrack on internationalizing models, such prospects could be higher. For instance, a Chinese leadership that does not cope appropriately with domestic challenges of economic and political transitions could be weakened or replaced by inward-looking opponents, with attending regional consequences, including heightened concerns among its neighbors. 45 Furthermore, internationalizing leaders everywhere are not immune to miscalculations in overplaying nationalist cards or falling victims to "blowback" and entrapment by inwardlooking constituencies more favorable to nuclearization. 46 The 2005 Chinese legislation codifying a declaration of war against Taiwan if the latter declares independence could provide an example of unintended effects of such miscalculations. In the Middle East, some have suggested that Turkey could, under

DOMESTIC MODELS OF POLITICAL SURVIVAL

53

certain circumstances, reconsider its nuclear status, for instance if it were to face a nuclear IranY In the last two decades, Turkish leaders have appeared to have transcended the Middle East's modal inward-looking path, consolidating an internationalizing model and renouncing nuclear weapons. If this choice were reversed while Turkey sustains the current model, the political survival argument would be refuted. If, however, Turkey were to reverse its nuclear commitments in tandem with progressively more inward-looking domestic models-exacerbated by EU exclusion-the argument would be sustained. Such domestic changes could also unleash a deterioration in Turkey's relationship with the North Atlantic Treaty Organization (NATO), highlighting the importance of domestic considerations in shaping security policies. Scenario 2 points to conditions where inward-looking models dominate but nonetheless embrace denuclearization. The past record of nuclear aspirants shows that this joint occurrence has been rare. This scenario would constitute another anomaly for the basic argument and could be illustrated by situations where inward-looking regimes in North Korea and Iran join and implement durable, transparent, and mutually and unconditionally verifiable agreements renouncing nuclear capabilities. The prospects for this outcome do not seem very likely in early 2010. However, if, for instance, Iran's and North Korea's nuclear policies change in tandem with domestic survival models, as they had in Egypt, Libya, and elsewhere, the argument that these models are deeply implicated in shaping nuclear policy would be corroborated. In East Asia, scenario 2 would involve the rise of inward-looking models in pivotal states that nonetheless retain NPT commitments and compliance. This outcome might be explained by path dependency or the increasing returns of a nearly fourdecades-old legacy of shunning nuclear weapons (except for North Korea), a legacy that would have to overwhelm the incentives of inward-looking leaders to transcend it. Scenario 4 suggests resilient inward-looking leaders resistant to internationalization, a defining characteristic of much of the Middle East for many decades, accompanied by intermittent efforts to acquire nuclear weapons. This scenario is compatible with the basic framework advanced here, and its permanence does not bode well for denuclearizing shifts in that region. The dramatic expansion in demand for nuclear technology in the Middle East could well present very real challenges to the nonproliferation regime. Much of it could take place in response to Iran's ability to flout nonproliferation commitments with relative impunity and an increased concern with Iran's hegemonic

54

ETEL SOLING EN

assertion, particularly through support ofShi'a minorities within Gulf states. The outcome of the Iranian nuclear debacle, and of resolutions adopted by the U.N. Security Council, may be the single most important predictor of what happens in the region next, because other inward-looking regimes are bound to learn important lessons about the degree of robustness of the nonproliferation regime. In East Asia, the widespread replacement of internationalizing models of political survival is certainly plausible, although the early signs of recovery from the 2oo8-2009 financial crisis suggest otherwise. The outcome of the 1997 Asian crisis also signaled more resilience than anticipated and, despite some political turnovers, did not lead to significant departures from internationalizing strategies. Such turns remain nonetheless conceivable in conjunction with global recessions or other regional and domestic downward economic protectionist or nationalist spirals. Significant domestic evolutions away from internationalizing trajectories-from China to South Korea, Indonesia, and Japan-might encourage nuclear chains (see the discussion in the following section). This outcome, although unlikely under present conditions in the world's most economically dynamic region, would be compatible with predictions in Cell 4. Thus far, the leaders of China, Japan, and South Korea have responded to the most severe global economic crisis in recent decades with a first-ever joint summit meeting in December 2008, pledging to institutionalize regular three-way summits. As the Japan chapter in the companion volume to this one recounts, even highly conservative leaders such as Abe Shinzo and Taro Aso remained constrained by a significant domestic consensus to stay the internationalizing, nonnuclear course. REGIONAL EFFECTS ON MODELS OF POLITICAL SURVIVAL AND OTHER DIFFUSION MECHANISMS

The scenarios in section II offer some guidance regarding continuity and change in nuclear trajectories on the basis of evolving models of political survival. This is not, however, a purely "domestic" garden-variety argument but one that concentrates attention on the connection between domestic and international politics by definition (ruling coalitions make choices about the kinds oflinks to the outside world that best serve their design to stay in power). Furthermore, these models are deeply interactive with regional circumstances, making the issue of chains and diffusion effects endogenous to the argument. In other words, the relative incidence of alternative models in neighboring

DOMESTIC MODELS OF POLITICAL SURVIVAL

55

states matters. One can think about the regional environment as an aggregate measure of the relative strength of internationalizing or inward-looking models. The extent to which regions share congruent orientations toward internationalization (either positive or negative) modifies domestic preferences on nuclear issues. The collective evolution of East Asia toward internationalization reinforced individual incentives of leaders to avoid nuclearization to preserve regional stability, foreign investment, and domestic economic growth, despite China's 1964 tests. Converging internationalizing models thus appear collectively stable, creating an environment that exhibits stronger immunity to nuclearizing chain reactions, even if juche blocks these regional synergies from influencing North Korea. In the Middle East, many rulers retained relatively closed political economies, facing fewer domestic and international disincentives for nuclearization. A regional environment dominated by inward-looking models is also collectively stable, in the sense that it reinforces individual incentives by leaders and ruling coalitions to retain those models, which expand state entrepreneurship, military expenditures, and baroque weapons. 48 Thus, such models feed on each other's existence, raising immunity against internationalizing models within the region and lowering immunity to nuclear chains. Neighboring leaders who might have otherwise favored internationalization face an unwieldy neighborhood that actively discourages it. Beyond these contextual regional influences, deep local recessions, regional or global economic downturns, and other severe disruptions in the global political-economic system could provide important triggers capable of undermining internationalizing models and buttressing inward-looking ones. Writing in the immediate aftermath of the worst financial and economic crisis since the Great Depression indeed gives rise to concern. In a more globalized economy, the potential for economic crises to diffuse can arguably accelerate the ascendance ofleaders logrolling antiglobalization inward-looking forces, in a coalitional domino of sorts. It is against such background that decisions to proliferate could become more feasible, a possibility entertained in Nuclear Logics, even under the more favorable conditions of 2006. 49 The prospect for increased proliferation under such circumstances is quite different from arguing, as neorealism does, that the nuclearization of one state almost inexorably leads to the nuclearization of its neighbor(s). Nuclear decisions are not the automatic result of shifting international power balances that are very hard to measure accurately a priori but rather remain contingent on

56

ETEL SOLING EN

the changing nature of domestic coalitions. The fact that automatic proliferation has not yet happened as systematically as one might think brings to relief the importance of understanding the conditions (political survival models) that mediate between triggers such as North Korea's nuclear tests and nuclear decisions by other East Asian states. East Asian leaders have adopted and diffused internationalizing models of engagement in the global economy. Some might argue that such models make individual states less secure, presumably because they thwart nuclear weapons programs. Yet many East Asian states lacking natural resources or nuclear weapons seem far less vulnerable or insecure than North Korea, Iran, or pre-2003 Iraq. Internationalizing models have turned them into engines of the 21st-century global economy, with much higher levels of domestic political stability, social and gender equity, human rights, expected life spans, employment, and educational endowments than unstable Middle East nuclear weapons aspirants, despite their rich natural and human resources. The fate of the nonproliferation regime could also influence the domestic balance of power between internationalizing and inward-looking forces. A variety of challenges appears to have placed the regime in a tenuous and fragile predicament. Increased legal and illegal flows of sensitive technologies; the IAEA's inadequate access to information regarding illegal transfers (Iraq, Iran, Libya, North Korea, and Syria); NPT violations by North Korea, Iran, and Syria; and the failed 2005 NPT Review Conference remain important tests of the regime's effectiveness, increasing uncertainty about its ability to endure beyond the 2010 Review Conference. 50 A collapse of the global nonproliferation regime could weaken internationalizers in their efforts to stem pressures for nuclearization from inward-looking (domestic and foreign) adversaries. At the same time, such collapse should not be taken for granted. It is precisely in times of crisis, as Dunn points out, that institutions may develop stronger foundations. 51 In sum, as this chapter's overview suggests, domestic models of political survival are filters through which other considerations operate, tilting the balance of incentives and disincentives to acquire nuclear weapons in one direction or another. This point is often missed by reductionist interpretations of the argument, which falsely assume that taking domestic political survival models into account implies lack of attention to external dimensions, including threats. One might consider these models as something akin to a lens that refracts, enhances, or redirects external stimuli according to the (dis)incentives

DOMESTIC MODELS OF POLITICAL SURVIVAL

57

embedded in each model. Without a proper understanding of the operation of this variable, much of the utility of the argument can inevitably be lost. Properly interpreting these models as filters for a wide-ranging set of domestic, regional, and global opportunities and constraints also requires tolerance for complexity. A state's evolution toward or away from nuclear weapons takes place neither in the soul of a single person nor as an autopilot response to system-level changes in relative power. As Philip Tetlock's masterful treatise on expert political judgment and prediction suggests, parsimony can be the enemy of accuracy, a substantial liability in real-world forecasting. 52

4

BEYOND THE SECURITY MODEL Assessing the Capacity of Neoclassical Realism for Forecasting Nuclear Proliferation

Natasha E. Bajema

of scholarly and diplomatic efforts to solve the nuclear dilemma, the puzzle of what causes nuclear proliferation and what leads to nonproliferation-the dynamics of nuclear proliferation-remains largely unsolved. 1 Policies and instruments for preventing and managing nuclear proliferation-for example, IAEA safeguards, the NPT, the Nuclear Suppliers Group, and the like-are designed and strengthened based on expectations by policy makers about their respective positive effects toward resolving the nuclear dilemma. These expectations are rooted in assumptions and theories about how the world works. In other words, international relations theory undergirds the efforts of policy makers to design policy mechanisms and to forecast trends in nuclear proliferation. Theories of international relations are typically developed to fulfill several functions: to account for the past, to explain the present, and "to provide at least a preview of what is to come."2 Academics are often content with theories that can adequately account for past and explain present trends in world politics. Policy makers, on the other hand, in choosing appropriate courses of action in the present, must be vigilant about the future. To be useful, a theory must go beyond explaining past decisions to acquire or renounce nuclear weapons. For example, to manage the threat of proliferation, policy makers need to be able to determine the propensity of countries to develop nuclear weapons in the future, to predict what events or changes in underlying conditions might trigger further weapons spread, to estimate the negative effects of nuclear weapons acquisition on the nuclear policies of other countries, and to anticipate the impact of alternative preventive measures. Although theories AFTER MORE THAN SIXTY YEARS

58

BEYOND THE SECURITY MODEL

59

of international relations are generally designed to account for "patterns from the past in such a way to make them useable in the present as guides to the future," they are far more proficient at the task of explaining the past and present than the more complex tasks of forecasting and prediction. 3 Nonetheless, without such theories, "all attempts at forecasting and prediction would be reduced to random guessing." 4 Without a better theoretical understanding of nuclear proliferation dynamics, policy makers will be hindered in developing effective international mechanisms for managing, preventing, and reversing nuclear proliferation in the 21st century. Despite a large literature on nuclear proliferation, there is still no single theory that fully accounts for the dynamics of nuclear proliferation-that is, why some states "go nuclear," why other states forgo the development of nuclear weapons, and how the nuclear choices of states change over time. In fact, theory building on the causes of nuclear (non)proliferation remains at a preliminary stage. Although there is a wide variety of proposed models based on competing theories of international relations, "they have not been adequately analyzed, nor placed in a comparative theoretical framework, nor properly evaluated against empirical evidence." 5 As an additional complication, the empirical evidence on nuclear proliferation strongly suggests that multicausality "lies at the heart of the nuclear proliferation problem."6 In spite of such evidence, however, most of the proposed models in the nuclear proliferation literature emphasize a single causal factor for explaining proliferation dynamics. 7 REALISM AND THE DYNAMICS OF NUCLEAR PROLIFERATION

Many scholars-realists and nonrealists alike-in the absence of a better monocausal theory consider the security model to be the standard explanation for nuclear proliferation. 8 The security model asserts that states will "develop nuclear weapons when they face a significant military threat to their security that cannot be met through alternative means." 9 Based on Kenneth Waltz's theory of neorealism-a narrow, albeit influential segment of realist theory-the security model is rooted in the traditional concept of the balance of power. 10 According to Waltz, states engage in competitive strategies to ensure their survival. In the short term, it is predicted that states will balance against changes in relative power that undermine their position in the international system, internally by developing their own nuclear arsenal and/ or externally by aligning themselves with a nuclear power. In the long term,

60

NATASHA E. BAJEMA

states are expected to emulate the successful strategies pursued by other states and, if it is perceived as a successful strategy, to develop nuclear weapons. Against the empirical record of existing nuclear weapons programs, the security model holds up quite well. Most decisions to develop nuclear weapons "appear to be best explained by the security model." 11 Thus, the security model provides a satisfactory explanation for cases of nuclear proliferation. However, the model begins to falter in its explanation for important cases of nuclear nonproliferation, in which technologically capable states have refrained from developing nuclear weapons despite acute threats to their security. The practice of many countries of ensuring their security against nuclear threats by forming alliances in spite of their own capacity for developing nuclear weapons contradicts a key underlying assumption of realism-that states prefer using self-help strategies to ensure their survival in an anarchic international system. For Waltz, this contradiction is not problematic because he did not design his theory to predict what strategies "will be imitated, and how quickly and closely." 12 Rather, his theory is designed to explain how similarities in state behavior (that is, balancing and emulation) lead to continuity in the international system. To understand why states respond differently to similar security contexts, "a theory would have to show how the different internal structures of states affect their external policies and actions." 13 Waltz observes that states may resist the pressure to conform to systemic imperatives.14 The success or failure of states to behave according to constraints of the international system is caused by internal attributes. If states fail to conform to the expectations imposed on them by the structure of the international system, however, Waltz predicts that they will suffer the consequences. 15 The security model faces even greater difficulty explaining the reversal of existing nuclear weapons programs or oscillation in nuclear choices over time. Based on systems-level theory, the security model can explain changes in state behavior only as a response to changes in the distribution of power. Absent a change in the threat context, Waltz's theory is unable to explain why the nuclear decision-making dynamics within a state vary over time. Even in the case of a changing threat context, however, the security model is hindered in predicting specific state behaviors due to the vaguely defined concepts of balancing and emulation. Although Waltz suggests a broad interpretation of the concept of balancing, the notion of internal and external balancing in realist literature is generally restricted to the military realm: "States either balance or they do not." 16 When this logic is applied to the nuclear arena, internal balancing involves developing nuclear weapons, and external balancing, forming an

BEYOND THE SECURITY MODEL

61

alliance with a nuclear power. Beyond the crude binary choice between internal and external balancing, there are a significant number of possible options and nuances in strategy for responding to security threats. However, because neorealist theory is not concerned with predicting foreign policy behavior, "little attention has been given to the critical question of the conditions under which a state pursues one strategy rather than the other." 17 The deficiencies of the security model for explaining the dynamics of nuclear proliferation-and thus for forecasting nuclear proliferation developmentsarise primarily from the neglect of domestic-level variables in neorealist theory. Although neorealists do not deny the importance of domestic variables, they argue that internal attributes are not relevant for explaining and predicting broad and recurring patterns in international relations. At the same time, Waltz admits that "under most circumstances, a theory of international politics is not sufficient ... for the making of unambiguous foreign-policy predictions."18 To enhance our understanding of a state's nuclear choices from any theoretical perspective, the analyst must open up the black box of the state and examine variation in domestic variables. In other words, developing a theory of the state-one that explains how the internal attributes of a state constrain and shape its ability to respond to systemic pressures-is a necessary component of any model assigned with the task of explaining and predicting the foreign policy choices of individual states, including decisions to develop or renounce nuclear weapons. 19 Paying inadequate attention to domestic factors, the security model is unable to predict when states will choose to balance internally or externally, explain why states might choose from many possible combinations of internal and external balancing, or explain why a state's balancing strategy changes over time. As such, the security model is unable to offer the policy maker reliable tools for forecasting nuclear proliferation. Given the shortcomings of the security model, one might conclude that the capacity of realist models for forecasting nuclear proliferation developments is extremely limited-a viewpoint that is often touted by scholars of competing theoretical perspectives. However, the failure of the security model to consistently explain and predict the nuclear choices of states does not reflect negatively on the realist paradigm as a whole. Frequent emphasis on neorealist theory, or even more narrowly on Waltz's theory, as the sine qua non of realism obscures the fact that the realist literature is expansive and offers several potential lenses through which to view the puzzle of nuclear proliferation. 20 In addition to being comprised of three major variants-classical realism, neorealism, and neoclassical realism-the realist literature is further divided

62

NATASHA E. BAJEMA

by two competing perspectives, that is, offensive and defensive. While each major variant of realism shares a basic set of tenets, they differ significantly on "the extent to which unit level causes-independent variables within the state-are thought to play a causal role." 21 Whereas neorealists do not take the effects of unit-level variables into account, neoclassical realists argue that, to have impact on state behavior, "systemic pressures must be translated through intervening variables at the unit level." 22 Neoclassical realism emerged as a formal intellectual movement in the late 1990s following an influential review article by Gideon Rose in which he argues that various works of realist scholars should be viewed as comprising a new realist theory of foreign policy. 23 Notably, the first scholars labeled by Rose as "neoclassical realists" did not perceive themselves as actively contributing to a new intellectual movement. 24 Rather, they simply sought more determinate explanations for foreign policy than pure neorealism could offer. 25 The sudden, peaceful, and unexpected end to the Cold War had not only demonstrated the weak explanatory and predictive power of structural variables but had also raised doubts about core realist assumptions regarding the inherently conflictual nature of international politics. 26 To remain relevant, realist scholars were under increasing pressure to explain significant variations in foreign policies as well as the sources of systemic change. Both of these tasks required realists to move beyond their preference for purely systemic approaches, open up the black box of the unit, and develop a theory of the state. In recent years, a growing number of realist scholars have responded to the challenge of opening the black box, and many are explicitly assuming a neoclassical realist approachY As Rose suggests, these authors have several important things in common. First, neoclassical realists view their enterprise as complementing rather than competing with insights from neorealism. Many neoclassical realists build on Waltz's structural theory by integrating consideration of both external and internal variables into a single theoretical framework. 28 Accordingly, their works "proceed from the same core assumptions" as neorealist theory. 29 To allow for consideration of both external and internal variables, however, neoclassical realists relax the rationality and unitary actor assumptions. Neoclassical realists assert that the state is distinct from society, thereby implying that the leaders of a state enjoy a significant degree of autonomy from society. 30 Second, neoclassical realists argue that system-level variables are filtered through intervening variables at the unit and/or individuallevels to produce different types of foreign policy. 31 In opening up the black

BEYOND THE SECURITY MODEL

63

box, neoclassical realists examine how internal attributes produce behavior either that cannot be explained by neorealism or that contradicts the theory's predictions. Finally, these scholars trace how changes in relative power translate into corresponding shifts in foreign policy and how this process is shaped by variation in domestic variables using detailed historical analysis and case studies. The nuclear proliferation puzzle has yet to be viewed through the neoclassical realist lens. Given its consideration of both external and internal variables, neoclassical realism offers significant potential for supplementing insights from Waltz's balance-of-power theory and thereby enhancing the capacity of the security model for explaining and forecasting nuclear proliferation. The remainder of this chapter evaluates the potential of neoclassical realism for explaining and forecasting nuclear proliferation and proposes how the basic elements of a neoclassical realist approach can be used to gain a better understanding of the dynamics of nuclear proliferation. ASSESSING THE POTENTIAL OF NEOCLASSICAL REALISM FOR EXPLAINING AND FORECASTING NUCLEAR PROLIFERATION

Contrary to what realism's critics would have us believe, the limitations of neoclassical realism do not lie in the alleged inability of realism to examine domestic variables within a coherent theoretical framework. 32 The challenges of using a neoclassical realist approach to explain and forecast nuclear proliferation arise from the fact that it remains a relatively new approach, has not been widely tested, and has yet to evolve into a unified theory. Although Rose deserves credit for his early recognition of a significant and innovative trend among disparate realist works, his declaration about the collective and theoretical nature of that effort was premature. A review of the literature reveals that neoclassical realism, as a whole, fails to systematically identify the factors that are most important for explaining state behavior and demonstrate consistent relationships among them. 33 Neoclassical realism is gradually evolving into a purposeful effort to refine insights from classical realism about the role of domestic variables, to integrate these insights into structural realism, and, most ambitiously, to advance the realist literature toward a comprehensive theory of international relations. 34 Recent works of neoclassical realism have abandoned the ad hoc examination of diverse domestic variables and have begun to unpack the "transmission belt" between systemic pressures and foreign policy choices using a similar set of

64

NATASHA E. BAJEMA

variables and concepts. 35 Neoclassical realists such as Randall Schweller and Jeffrey Taliaferro use Waltz's balance-of-power theory as a starting point and draw on insights from classical realism to investigate the causal mechanism linking changes in relative power to different types of balancing strategies. In doing so, they address the deficiencies of Waltz's theory for explaining and predicting foreign policy choices by broadening the concept of power, specifying a role for perception, and explicating the concepts of balancing and emulation. Each of these innovations-if embedded into the security modeloffers significant potential for expanding its capacity to explain and forecast nuclear proliferation. Broadening the Concept of Power

Like neorealists, neoclassical realists argue that the distribution of power remains the most critical driver of state behavior in that it "establishes the basic parameters of a country's foreign policy."36 At the systems level, neoclassical realists adopt the neorealist concept of relative power, which refers to a state's material capabilities, both military and economic, in relation to other states in the system. States are concerned primarily with the distribution of military capabilities because their survival depends on their ability to defend themselves against an attack. However, because economic capabilities serve as a critical foundation for building military capabilities in the long term, states must also take economic potential into account when assessing relative power. For neoclassical realists, however, a systems-level concept of power by itself is inadequate for predicting state behavior because "national leaders may not have easy access to a country's total material power resources." 37 In other words, material capabilities by themselves cannot be equated with national power. Rather, "effective national power depends on the state's capacity to mobilize, organize and deploy those resources efficiently in the service of its objectives."38 By holding domestic variables constant, neorealists assume "that all states have similar extractive capacity, such that aggregate national resources may be equated with actual state power and global influence." 39 In practice, leaders are constrained by their relative ability to extract and mobilize resources from domestic society to conduct their preferred foreign policies. To make specific predictions about foreign policy choices, neoclassical realists need to define and operationalize the causal mechanism of power in greater detail. Neoclassical realists broaden the concept of power to include variables at both the systems and unit levels of analysis.

BEYOND THE SECURITY MODEL

65

To expand the concept of power beyond material capabilities to include its political dimensions, neoclassical realists draw on insights from classical realism. As Taliaferro suggests, Classical realism is concerned primarily with the sources and uses of national power in international politics and the problems that leaders encounter in conducting foreign policy. These issues lead the analyst to focus on the power distributions among states as well as the character of states and their relation to domestic society.•o Classical realists include both internal and external factors in their definitions of power, many of which are difficult to measure quantitatively (for example, geography, industrial capacity, military preparedness, national character, and the like). To integrate the classical realist concept of national power into Waltz's balance-of-power theory, neoclassical realists draw a distinction between relative power (at the systems level) and state power (at the domestic level). 41 Relative power is comprised of military and economic capabilities (that is, material factors) and represents the potential power of a country relative to other countries. State power, on the other hand, describes the relationship between the state apparatus and society within a country (nonmaterial factors) and represents national effective power-that is, the ability ofleaders to exploit a country's potential powerY Because the resources of a nation-state are located within society "the strength of a country's state apparatus and its relation to the surrounding society" shapes the foreign policy choices available to the state. 43 While a country's relative power establishes the parameters for state behavior, state power shapes how leaders respond to changes in relative power at the systemic level. Many of the nonmaterial elements of power described by classical realists are captured by the neoclassical realist concept of state power. Thomas Christensen defines it "as the ability of state leaders to mobilize their nation's human and material resources behind security policy initiatives .... Without political power, national economic and military potential would never be actualized."44 For Taliaferro, components of state power include the relative strength of existing state institutions, levels of nationalism, and existence of statist or antistatist ideology. 45 Schweller measures "mobilization capacity" by assessing the degree of consensus among elites, the level of cohesion of both central government and society, and the degree of government legitimacy. 46 In Fareed Zakaria's model, state power is measured by examining several characteristics

66

NATASHA E. BAJEMA

of the state apparatus: the scope of governmental responsibilities, the degree of autonomy from society, the level of cohesion among state institutions, and the ability to raise tax revenuesY Neoclassical realists assess a country's power using a two-step approach, at both the systemic and domestic level. Whereas measuring relative power at the systemic level is somewhat straightforward-assuming the availability of necessary quantitative data-measuring power at the domestic level entails a qualitative assessment of various factors that are determined to contribute to state power. According to Stephen Krasner, "the strength of the state in relationship to its own society can be envisioned along a continuum ranging from weak to strong."48 The leaders of weak states have minimal ability to extract resources because they are unable to resist societal pressures and are compelled to "serve specific interests within the country, rather than the general aims of the citizenry as a whole."49 The leaders of strong states-possessing significant ability to extract resources-are not only capable of achieving objectives perceived to be in the national interest, but they are able "to actively reshape aspects of the economy and society."50 Using Zakaria's model, a strong state-for example with a broad scope and high levels of autonomy, coherence, and capacity for raising revenues-will have more flexibility in its choice of balancing strategies. 51 The leaders of a weak state, with low scores on each of these characteristics, will be constrained in defining a balancing strategy according to their preferences. Specifying a Role for Perception

Though the notion of perception is implicit in its theoretical assumptions, Waltz's theory neglects to describe the process through which changes in relative power are interpreted as threats to security and then translated into balancing behavior. 52 In Waltz's theory, the state, abstractly conceptualized as a billiard ball, acts as a frictionless "transmission belt" for systemic pressures. In contrast, neoclassical realists assert that "the notion of a smoothly functioning mechanical transmission belt is inaccurate and misleading." 53 Aaron Friedberg explains that "shifts in the distribution of power within an international system may be 'real' in some sense, but they may fail to have any impact unless and until they are perceived." 54 For this reason, neoclassical realists specify a distinct role for the perception of decision makers. The need to consider the role of perception in a neoclassical realist model is a logical consequence of the assumed distinction between state and society. 55 Because it is the leaders of a state who assess their state's position within the

BEYOND THE SECURITY MODEL

67

international system and calculate the costs and benefits of different courses of action, "material capabilities can influence states' external behavior only through the medium of central decision-makers' perceptions, calculations, and estimates." 56 Hans Morgenthau acknowledges that "the task of assessing the relative power of nations for the present and the future resolves itself into a series of hunches, of which some will certainly turn out to be wrong, while others may be proved by subsequent events to have been correct." 57 In addition to miscalculations, there are often time lags between actual changes in the distribution of power and threat perception. Recognizing the role of decision makers' perceptions and calculations does not mean that "states' foreign policies or international outcomes consistently fail to correspond to material power relationship." 58 Despite short-term discrepancies between material capabilities and state behavior, "over the long run, international outcomes correspond to the relative distribution of material capabilities." 59 In-depth consideration of the perception of decision makers allows neoclassical realists to probe deeper explanations for the incongruous impact of relative power on state behavior-for example, why states perceive certain changes in relative power as more negatively affecting their position in the international system than others. Through the conduit of perception, neoclassical realists examine additional structural factors beyond relative power such as technology and geography that shape how threats are perceived and how leaders will respond to threats. Leaders are sensitive to the strategic consequences of their own actions for achieving greater security and often interpret the actions of other states as indicators of their intentions-that is, whether states are seeking to maintain the status quo or harbor expansionist objectives. However, in certain situations, referred to as acute security dilemmas, "an increase in one state's security decreases the security of others." 60 In these situations, leaders cannot reliably infer the intentions of other states from their action and therefore must prepare for conflict to ensure their state's survival. The leaders of a state with an acute security dilemma-determined by factors such as geographic location, size of territory, the offensive or defensive nature of military capabilities and strategy-are more likely to perceive changes in their state's position within the international system as security threats. The operative security dilemma increases or decreases the likelihood of conflict among states and thus shapes balancing strategies. In a neoclassical model, the perceptions and calculations of decision makers form the critical link between the systems and unit levels of analysis.

68 NATASHA E. BAJEMA

Whereas relative power determines the broad parameters of a state's response to emerging threats, the process of identifying and selecting among policy options for responding to threats is shaped by how leaders perceive international and domestic constraints. In other words, "security policies are not dictated by systemic considerations alone .... National leaders must weigh both the internal and external consequences of their foreign policies. Domestic politics shapes and constrains what options leaders can adopt in pursuit of national security."61 From the perspective of leaders, "the relative ability of the state to extract or mobilize resources from domestic society ... shapes the types of internal balancing strategies a state is likely to pursue." 62 As Michael Barnett suggests, "security policy is itself two-faced: it is concerned with the construction of strategies vis-a-vis foreign threats and with the construction of strategies for mobilizing societal resources as well." 63 In a neoclassical realist approach, the perception of decision makers provides a lens through which to view relative power. Rather than responding to actual changes in relative power, neoclassical realists argue that states engage in balancing behavior when leaders perceive changes in relative power as threats to their security. As such, it is necessary not only to calculate relative power but also to determine when and how changes in relative power are perceived by decision makers. Friedberg suggests that both calculations by themselves are inadequate for understanding state behavior: "Purely quantitative indicators of capabilities simply cannot capture decision-makers' assessments."64 Even if leaders have access to necessary information about resources and policy options, there are aspects of power that cannot be captured quantitatively-in particular, at the domestic level. To properly assess how changes in relative power translate into foreign policy, William Wohlfarth suggests that scholars must achieve "a correct rendering of the perceptions that inform decisions." 65 Whereas calculating relative power is relatively straightforward, accurately measuring the perception of decision makers is complicated. Changes in relative power can be traced using existing quantitative data. The perceptions and calculations of decision makers regarding international and domestic constraints are reflected in their discussions and debate about the policy options available for responding to threats. 66 However, leaders may not reveal their true motivations or perceptions in any public discussion or documents. Neoclassical realists address the problems associated with measuring perception by closely examining the relationship between changes in relative power and perceived threats by decision makers. Taliaferro conceptualizes the relative distribution of power

BEYOND THE SECURITY MODEL

69

and how it is perceived as a level of external vulnerability, which is "a function of the relative distribution of power (both in the international system and in the particular region), the offense-defense balance in military technology, and geographic proximity."67 Thus, in a neoclassical realist model, the threat level-a function of relative power and threat perception-determines the broad parameters of state behavior. Explicating the Concepts of Balancing and Emulation

According to Waltz's balance-of-power theory, states compete with each other in the short term by balancing against changes in relative power that undermine their position in the international system. In the long term, as a result of their interactions over time, states emulate the successful practices of other states. Given Waltz's emphasis on explaining recurring patterns in international politics, the concepts of balancing and emulation are vaguely defined. Neoclassical realists, concerned with explaining specific state behaviors in response to security threats, explicate the concepts of balancing and emulation in more detail. Neoclassical realists agree that states respond to changes in relative power by engaging in balancing behavior. However, neoclassical realists broaden the concept ofbalancing behavior beyond a simple binary choice between arms and alliances. Rather, balancing refers to any action taken by states to increase their material capabilities and ability to respond to external threats and includes the broad spectrum of policies-both internal and external-that a state pursues "to reduce or match the capabilities of a powerful state or a threatening actor."68 External balancing involves any action by states to reduce threats to their security by relying on external resources and includes forming or strengthening an alliance, weakening rival alliance, and engaging in arms control to place limits on rivals or reduce the costs of an arms race. Internal balancing involves any action by states to reduce threats to their security by relying on internal resources and includes actions taken by a state's leaders to enhance their ability to extract resources from society in the future. Although internal balancing is most commonly understood to involve extracting resources to build military capabilities, it also includes actions to mobilize resources for the purposes of generating wealth and other long-term material capabilities. In reality, states do not choose between internal or external balancing. Rather, they engage in both. Thus, neoclassical realists are concerned with how and why states pursue different "balancing strategies" consisting of both internal and external actions. 69 They argue that states tend to place a stronger emphasis

70

NATASHA E. BAJEMA

on either internal or external balancing, depending on the nature of their international and domestic constraints?0 Because anarchy encourages states to pursue self-help strategies, neoclassical realists expect that states-given access to the necessary resources-will prefer internal balancing over external balancing. Whereas the strategies of strong states are expected to emphasize internal balancing, the strategies of weak states will stress external balancing. In the long term, neoclassical realists agree with neorealists that states will emulate the successful strategies of other states over time. Waltz argues that, as states compete with each other, they duplicate "the activities of other states at least to a considerable extent. Each state has its agencies for making, executing, and interpreting laws and regulations, for raising revenues, and for defending itself." 71 This argument draws on the insights of Charles Tilly that "war made the state and the state made war." 72 The origins and ongoing development of the modern state organization have been widely linked to the need to effectively engage in war fighting and preparation for war. According to neoclassical realists, as states emulate the successful strategies of other states for responding to security threats, they engage in state building, which "refers to an increase in the mobilization and extractive capacity of central state institutions relative to other societal actors" (that is, an increase in state power).7 3 In preparing for war, the leaders of a state drive the state building process through their relative emphasis on internal versus external balancing.74 Whereas strategies emphasizing external balancing tend to sustain the prevailing relationship between state and society, strategies that stress internal balancingin particular, resource extraction-lead to increases in state power and thus cultivate the future ability of leaders to respond to systemic imperatives. 75 Increases in state power allow leaders greater ability to extract resources and thus more flexibility to select among policy options and implement their preferred foreign policies. In contrast, a decrease in threat level undermines the ability of leaders to extract resources from society for building military capabilities among both strong and weak states. As threat perception decreases over time, the domestic constraints on resource extraction typically increase, leading to decreases in state power and a reduced ability to expand military capabilities. EXPANDING THE CAPACITY OF THE SECURITY MODEL FOR EXPLAINING AND FORECASTING NUCLEAR PROLIFERATION

Expanding the security model using a neoclassical approach offers significant potential for improving its capacity for forecasting nuclear proliferation. Using Waltz's theory as its starting point, neoclassical realism broadens the

BEYOND THE SECURITY MODEL

71

explanatory and predictive power of realist theory for foreign policy choices by simultaneously investigating the effects of relative power, perception, and state power on state behavior. Before proposing a neoclassical realist approach for explaining, predicting, and forecasting nuclear proliferation, several basic issues need to be briefly discussed. Given the conflicting offensive realist and defensive realist perspectives, it is necessary to make an assumption about the preferences of statesthat is, to assume that states are expansionist and continuously pursue power or that states prefer maintaining the status quo and thus seek to achieve security. The choice to build on and complement Waltz's theory implies that my model proceeds from a defensive perspective and assumes that most states prefer maintaining the status quo and therefore aim to maximize security rather than power. Although assuming a defensive perspective may limit the predictive scope of the theory, there are several factors that speak against drawing this conclusion too hastily. First, the task of determining which states are status quo or expansionist is a highly subjective endeavor. States rarely declare their real intentions. Furthermore, because the status quo behavior of balancing also involves expanding power and influence, it is difficult to distinguish from expansionist behavior. Second, the distinction between expansionist and status quo states is largely based on the notion of a smooth and frictionless transmission belt between systemic pressures and foreign policy. Given presumed constraints at the domestic level, it is possible that the preferences and objectives of states (whether status quo or expansionist) matter less than is often suggested in realist literature. Finally, nuclear weapons appear more suitable for achieving status quo aims rather than expansionist aims. Over time, nuclear weapons, though considered to be offensive weapons, have not proven useful for implementing strategies of compellence. In contrast, nuclear weapons have served as effective tools of deterrence (for example, for deterring conventional and nuclear attacks and expansionist aggression) and support status quo objectives rather than expansionist pursuits. The assumption that states seek security requires a definition for the subjective concept of security. Security means different things to different people: What is protected ultimately depends on what is valued; how much security is enough depends on how acutely threats are perceived and the nature of priorities. Rather than treating the concept of security in terms of outcomes-the subjective status of being or feeling secure-I argue that the "quest for security" involves two types of activities by states, actions that reduce perceived

72

NATASHA E. BAJEMA

threats and actions that minimize the risk of facing future threats.76 In addition to balancing against specific threats to their security in the short term, states attempt to maximize their flexibility or freedom of action in the future. In the following section, I expand the security model in three steps that parallel an "ideal" nuclear decision-making process. The resulting framework offers an analytic model for explaining and forecasting the nuclear choices of states. However, it is a work in progress and has yet to be formally tested. 77 Step 1: Identify a Security Threat That Requires a Robust Response

As the first step, I integrate the perception of decision makers into the security model. The security model argues that states will respond to changes in relative power by balancing but cannot explain why states might fail to balance as expected. Given the intervening effect of perception, neoclassical realists do not expect states to balance automatically or proportionally to changes in relative power. Before a state even considers developing nuclear weapons, the leaders of the state must first recognize that a change in relative power has occurred, interpret that change as negatively affecting the position of their state within the international system, and perceive the change in relative power as an acute threat to their security-that is, one that threatens its survival or vital national interests and requires a robust response. Nuclear weapons are not likely to factor into the policy equation unless leaders perceive an acute threat to their security. Throughout this process, there are several junctures at which leaders may fail to perceive a change in relative power that would normally justify certain balancing actions as a security threat. Leaders may not immediately recognize that an adverse change in relative power has occurred, which explains short-term discrepancies between changes in relative power and shifts in a state's balancing strategy. Alternatively, leaders may miscalculate the impact of changes in relative power on the position of their respective states in the international system. Due to the lack of information and intangible aspects of power, the impact of changes in relative power is often ambiguous. In these situations, states may underbalance or overbalance against changes in relative power. Once a threat is perceived by the leaders of a state, however, states are expected to take a proportional response to the perceived threat level. Throughout the nuclear age, the historical record abounds with examples of how perception has intervened between changes in relative power and the type of balancing behavior predicted by realist theory. For example, immediately following World War II, the swift demobilization of U.S. troops produced a massive conventional imbalance in Europe-by 1950, the United States

BEYOND THE SECURITY MODEL

73

maintained only about 1.5 million troops compared to the Soviet Union's 4.3 million troops. 78 Despite the rapid onset of the Cold War, Western powers were confident that the U.S. nuclear weapons would deter Soviet aggression. It was not until after the outbreak of the Korean War in June of 1950 that Western policy makers were determined to balance against the military threat posed by the Soviet Union. After the Korean War demonstrated to Western policy makers that the Soviet threat to Western Europe was real, NATO allies responded by creating a military command structure for NATO, establishing ambitious goals to increase conventional forces in Europe, and allowing West Germany to regain its full sovereignty in exchange for its rearmament. However, the West's determination to balance against the Soviet threat was short lived. Already by 1953, the perceived threat of the Soviet Union had significantly diminished. Despite the enduring conventional imbalance between the East and the West, the Western powers largely abandoned their ambitious plans to balance against the conventional threat posed by the Soviet Union and returned to their pre-Korea threat perception. Step 2: Consider the Available Policy Options

In the second step, I consider the impact of the security dilemma and expand the range of external factors for consideration beyond changes in relative power to include other structural factors such as geography and technology. Before developing nuclear weapons, leaders consider all of the available policy options for balancing against threats and the impact of various structural factors on the outcomes of their choices. The nature of the security dilemma creates incentives or disincentives for nuclear proliferation. In the nuclear realm, the security dilemma can translate into a reluctance to develop nuclear weapons even in the face of an acute security threat. This reluctance arises from uncertainty about the consequences of nuclear weapons acquisition for achieving security. Even if leaders believe that the development of a single nuclear weapon would increase the security of their state, this action may cause rivals to follow suit. This action-reaction pattern would likely produce a costly arms race that would decrease the security for all states. Despite this dangerous spiral, the state to act first is seen to have the advantage. Thus, if the intensity of the security dilemma and danger of a first strike is acute, state decision makers may feel that they have no choice but to develop nuclear weapons. Leaders understand that certain actions to increase security may actually lead to greater insecurity and take the strategic consequences of their actions into account before they select from available policy options. For my first case

74

NATASHA E. BAJEMA

study, I am looking at West Germany's decision to equip the newly established German army with tactical nuclear weapons in 1958. There is significant evidence of the influence of the security dilemma and structural factors such as technology and geography on this decision. In response to the decreasing credibility of its nuclear guarantee after the Soviet launch of Sputnik, the United States offered to deploy Intermediate-Range Ballistic Missiles (IRBMs) on the territory of its allies in Europe in 1957. Some allies accepted the offer, but West Germany resisted despite its insistence on greater access to U.S. nuclear weapons. Chancellor Adenauer argued that "the Thor and Jupiter systems would be inappropriate for deployment on the territory of West Germany" for several reasons. 79 First, he noted the military vulnerability of stationary missiles located above ground, showing that he was aware of the limits of this technology for reducing security threats. Second, this vulnerability would be heightened by their placement on the front line of NATO defenses, which indicates his sensitivity to the geographical location of West Germany. Third, according to Adenauer, the deployment of such missiles in West Germany could lead to military instability in crises. In other words, such deployment would intensify the security dilemma of the Soviet Union, increasing the likelihood of conflict. The deployment of missiles with ranges capable of reaching Soviet territory would threaten the Soviet Union and endanger progress on the reunification of Germany. In contrast to his rejection of the IRBMs, Adenauer decided to equip the German army with tactical nuclear weapons in 1958, a decision that led to the deployment of thousands of U.S. tactical nuclear weapons on German soil under a dual key arrangement, creating the third-largest-albeit virtual-nuclear weapons state by the 1970s. Step 3: Select from Available Policy Options

In the third and final step, I integrate domestic factors into the security model. When leaders attempt to implement their preferred policy option, they are constrained in their ability to extract and mobilize resources by the domestic structure of their state and may be compelled to make adjustments to their balancing strategy. I conceptualize the impact of domestic structure using the neoclassical realist concept of state power, which is defined as the ability of leaders to extract and/or mobilize resources from domestic society. Because the concept of state power is rather broad for understanding why states choose from a complex array of policy options, I have dis aggregated the state power into four key abilities. To successfully implement any balancing strategy, leaders must have sufficient ability to: (1) make and implement executive decisions on

BEYOND THE SECURITY MODEL

75

defense issues; (2) raise adequate money for defense purposes; (3) raise troop numbers for military forces; and (4) tap into their state's technology and industrial capacity to produce armaments. The ability of leaders to achieve all four of these tasks depends on certain attributes of domestic structure. For example, the ability to make and implement decisions depends on the scope of authority given to the government on defense matters, how the authority for defense policy is distributed among the government institutions, and to what extent the leadership must maintain consensus within its ruling coalition to implement its preferred defense policy. Even if authority for defense policy is highly centralized, the power of a leader to pursue preferred policies may be diluted by the need to maintain a majority coalition. The ability to raise troops, raise money, and produce armaments depends not only on the state's resource capacity but also on the laws governing the relationship between the government and society. These laws determine how the economy is structured, the methods available to the government for increasing the revenue at its disposal, and the ability of the government to tap into its industrial capacity and use its population resources. In my dissertation, I distinguish between two ideal types of states. Whereas a strong state has a high ability to extract and mobilize resources, a weak state has low ability to extract and mobilize resources. I have found that states are not broadly strong or weak across all four of these dimensions of state power. While a state may be characterized as strong or weak in making and implementing decisions on defense policy, states possess a different level of ability for implementing each of the three remaining tasks. Consequently, the relative ability of leaders to raise revenues, raise troops, and produce armaments shapes the decisions that leaders make to balance against security threats. For example, in the late 1950s, Chancellor Adenauer enjoyed substantial autonomy in making decisions on defense policy. In addition, Adenauer had massive economic resources at his disposal to spend on defense-a result of budget surpluses from previous years and growing tax revenues resulting from the rapid growth of the German economy. However, Adenauer still faced serious constraints establishing and equipping the soo,ooo-person German army that had been promised to NATO. Adenauer confronted difficulties in raising adequate troop numbers, partly due to work force shortages resulting from West Germany's booming economy but also as a result of compromises reached on the details of its conscription law. Moreover, because most of Germany's industrial capacity for producing armaments had been dismantled following World War II, West Germany lacked the capacity to produce its own armaments

76

NATASHA E. BAJEMA

in the late 1950s and depended almost entirely on imports of weapons from its allies-in particular the United States. All of these factors undermined Adenauer's ability to maintain a coalition in support of his plan to establish the Bundeswehr. As a preliminary finding from my first case, I argue that Adenauer's inability to raise sufficient troop numbers to meet conventional force goals and West Germany's dependence on exports for equipping its army led to Chancellor's decision to equip the West German army with U.S. tactical nuclear weapons in 1958. Predicted Outcomes and Implications for Nuclear Forecasting

A neoclassical realist model expands the capacity of the security model for explaining and predicting variation in balancing behavior in both the short term and the long term. In the short term, the leaders of a strong state will have greater ability to extract and mobilize domestic resources and thus greater flexibility in their balancing options for responding to external threats. In a weak state, on the other hand, the ability of leaders to extract and mobilize resources for defense purposes is highly constrained. Consequently, leaders of weak states have fewer balancing options and will be forced to rely more on alliances. Thus, whereas strong states will broadly emphasize internal balancing, weaker states, facing greater internal constraints, are more likely to seek alliance protection and emphasize external balancing in their balancing strategies. In the long term, a neoclassical realist model predicts that all states facing acute security threats-both weak and strong-will take internal actions that lead to increases in state power (that is, state building) to minimize risk when facing threats in the future. The type of internal actions pursued by states (for example, shifting from a voluntary force to conscription, broadening the sources of revenue of the state, taking measures to simplify the decision-making process on defense issues, or streamlining access to its technological and industrial capacity for producing armaments) will ultimately depend on both the perceived level of external threats and the preexisting level of state power. The resulting growth in state power gives leaders greater flexibility in implementing their preferred foreign policies and greater capacity for engaging in internal balancing. Facing an acute threat over time, the balancing strategies of weak states will gradually imitate those of strong states. If the threat level is perceived as low over an extended period, however, states are expected to refrain from state building and even reverse the state building process, resulting

BEYOND THE SECURITY MODEL

77

in a decrease in state power. If the threat level decreases significantly, even strong states are expected to become weaker because they can no longer justify the same level of resource extraction to their societies. These predictions about balancing strategies have several implications for forecasting nuclear proliferation more broadly. Strong states-possessing a greater capacity for internal balancing, in particular resource extraction-are more likely to develop nuclear weapons than weak states if the leaders determine that nuclear weapons will improve their security. A growth in state power is expected to increase the propensity of a state to develop nuclear weapons in response to an external threat. However, because state organizations gradually grow stronger when they face acute external threats, I expect there to be a long transition process for a weak state to become a strong one. Sudden changes in the threat context may trigger nuclear proliferation among strong statesespecially those states near the threshold of nuclear capability. However, weak states will be initially hindered from going nuclear due to a lower ability to extract the necessary resources. For this reason, it is unlikely that a decision by one country to develop nuclear weapons will immediately result in a nuclear chain reaction or cascade. This suggests that nuclear proliferation is not inevitable. Widespread nuclear proliferation can be prevented by addressing the security needs of threatened states, especially over the long term. The task of going beyond explaining and predicting internal or external balancing strategies to forecasting the specific nuclear choices of states remains challenging. Using neoclassical realist theory, scholars can broadly predict and explain why states choose among different balancing strategies when faced with similar threat contexts. Explaining, predicting, and forecasting instances of nuclear proliferation, however, involves understanding why states choose from specific categories of weapons for ensuring their security. Although scholars would never expect a theory to explain why a state chooses, for example, to build more artillery over expanding its infantry, we cannot escape the fact that, by using theory to forecast nuclear weapons decisions, we are attempting to explain and predict the choice of a weapon type. Indeed, nuclear weapons are not like other weapons; they are revolutionary and have been referred to as the "absolute weapon." Because there is no defense against nuclear weapons they have a particularly acute impact on a state's security dilemma. A state possessing nuclear weapons does not need to defeat an opponent militarily to cause severe destruction and devastation. In other words, the possession of nuclear weapons by one state directly threatens the survival of another state and demands a response by the threatened state. Compared to

78

NATASHA E. BAJEMA

other decisions regarding the selection of means for ensuring security, states are thus particularly aware of the choice between conventional and nuclear weapons. As such, explaining and forecasting the outcome of such choices are legitimate and essential tasks for scholars and policy makers alike. Acknowledging the nature of this challenge, the model proposed in this chapter offers potential for predicting and forecasting the nuclear choices of states. As discussed earlier, the leaders of states select among a range of internal policy options for responding to threats based on the intensity of perceived threat and their ability to extract resources. In my framework, the concept of state power is disaggregated into four distinct abilities: making and implementing policy decisions, raising troops, raising revenues, and producing armaments. To understand why states choose between conventional and nuclear weapons to balance internally, these four abilities need to be assessed for defense policy and nuclear policy. On comparison of the two issue areas, I expect to find variation in levels across all four of these abilities for both defense and nuclear policy. More importantly, I expect to find that state power will broadly vary depending on the issue area in question-that is, states will be stronger in one of the two areas. For example, although the German chancellor-with the support of his cabinet and his party coalition-held the sole authority to make and implement defense policy in the late 1950s, authority for making and implementing nuclear policy in West Germany was far more decentralized. The lack of executive authority on nuclear policy would have complicated any effort by the West German government to extract resources for the purpose of building nuclear weapons. In other words, irrespective of West Germany's external policy constraints and relative technological capacity for nuclear or conventional weapons, it was relatively easier for German leaders to extract resources to build conventional weapons than to start a nuclear weapons program in the late 1950s. These hypotheses have several implications for nuclear forecasting. In response to an emerging security threat, both strong and weak states are expected to engage in some type of internal balancing and expand their military capabilities. The choice between increasing conventional capabilities and building nuclear weapons will be shaped by their relative ability to extract resources for both categories of weapons. When faced with an acute security threat, states with a stronger ability to extract resources for building conventional weapons than nuclear weapons are predicted to refrain from developing nuclear weapons. These states will balance internally by increasing their

BEYOND THE SECURITY MODEL

79

conventional military capabilities and, if possible, balance externally by forming an alliance with a nuclear power. Conversely, states with a stronger ability to extract resources for building nuclear weapons are predicted to go nuclear when faced with an acute external threat. CONCLUSION

Because the empirical record offers convincing evidence of multicausality, it follows that the task of explaining and predicting nuclear proliferation should involve consideration of several internal and external variables within a logically coherent framework. Yet, much of the theoretical enterprise in the area of nuclear proliferation emphasizes a single independent variable of choice for explaining the dynamics of nuclear proliferation. Though considered the standard explanation for nuclear proliferation, the security model is insufficient for explaining the dynamics of nuclear proliferation-and thus for forecasting nuclear proliferation developments-because neorealism lacks a well-developed theory of the state. Given its consideration of both external and internal variables within a single theoretical perspective, neoclassical realism offers significant potential for expanding the capacity of the security model for explaining and forecasting the nuclear choices of states. Although not yet a full-fledged theory, neoclassical realism represents a powerful analytic approach for understanding variation in foreign policy, including the broad spectrum of nuclear balancing strategies. The security model, expanded using a neoclassical realist approach, not only can account for both cases of nuclear proliferation and nonproliferation but can also explain why technologically capable states form alliances to ensure their security against nuclear threats, despite an assumed preference for self-help strategies. Thus, an expanded security model affords useful insights about future nuclear proliferation developments to guide policy makers in their efforts to prevent, manage, and reverse nuclear proliferation.

5

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION Scott D. Sagan

HOW QUICKLY COULD INDIVIDUAL GOVERNMENTS, starting from different levels of nuclear-related expertise and technology, develop a nuclear weapon if they chose to do so? This question-which I will call the "nuclear latency" question-is both exceedingly important and poorly understood. It is important because an accurate understanding of both underlying state capabilities and the time needed to utilize such capabilities is necessary to analyze a wide set of nuclear policy issues: for example, dealing with the Iran nuclear crisis (how quickly could Tehran make a weapon from its stockpile oflow-enriched uranium?); understanding the relationship between the spread of civilian nuclear power and the spread of nuclear weapons capability (will new civilian programs make breakout to military programs easier and more likely?); evaluating potential NPT reforms (what would be the effects of lengthening the ninety-day notice in the Article X withdrawal clause?); or assessing the stability of a world without nuclear weapons (could disarmed states rearm in five days, five weeks, five months, or five years?). Despite widespread discussion of these policy issues, however, a set of mirror-image analytic failures has limited our ability to make clear predictions about nuclear latency and proliferation: Political scientists working on these subjects have often failed to examine basic technical factors regarding the nuclear fuel cycle that strongly influence how quickly states can get the bomb; the more technical literature about nuclear latency has similarly often failed to examine the political factors that strongly influence the ability of a government to develop nuclear weapons. This chapter is both a conceptual minesweeping exercise and a modest effort to suggest a better way forward. It has four main parts. First, I briefly

80

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

81

present examples of how some journalists, diplomats, and scholars have misunderstood the nature of nuclear latency in their interpretations of IAEA reports, leading to exaggerations concerning the number of states that currently could build nuclear weapons in a short period of time. Second, I review and critique the political science literature on "nuclear capability" and "nuclear latency." Even the most sophisticated political science studies on this subject have too often used misleading measures of the key variables involved in nuclear technology, focusing on broad measures of industrial capability and nuclear research reactor experience and not on the specific fuel cycle technologies and facilities needed to make the fissile materials required for a nuclear weapon. This has led some political scientists, quite mistakenly in my view, to denigrate the NPT regime, arguing that efforts to restrict the spread of sensitive nuclear technology have failed in the past and that further restrictions in the future are likely to be ineffective or even counterproductive. Third, I review and critique leading examples from the technical literature on nuclear latency. These studies have usefully focused on how long it has taken individual states to develop highly enriched uranium (HEU) or weaponsgrade plutonium but have unfortunately too often left out of their analyses the political factors that accelerate or constrain such fissile material development. These technical studies have also usefully included estimates of how long it might take a state to develop one or more nuclear weapons once it has the necessary fissile material and the political leadership has made a decision to seek a nuclear bomb. However, they have ignored the strategic and domestic factors that can influence the urgency of a leader's demand for nuclear weapons and the domestic political and organizational factors that can influence whether a state bureaucracy can successfully implement a nuclear weapons acquisition plan. The conclusions offer suggestions for improved assessments of nuclear latency in the future. Instead of chasing the quixotic goal of a single measure of nuclear latency, scholars should seek to understand how political factors can influence technological developments in this arena and how reaching various thresholds in nuclear power technology can affect the politics of proliferation decisions. Unfortunately, the field's long-standing intellectual tradition of dividing the proliferation puzzle into "supply-side" and" demand-side" factors has reduced attention on the crucial and complex relationship between the supply of nuclear technology and the demand for nuclear weapons. In the conclusion, I therefore outline how one might better conceptualize and study

82

SCOTT D. SAGAN

the nature of nuclear latency and its relationship to the acquisition of nuclear weapons. In addition to sketching out an interdisciplinary research agenda, the chapter ends with a discussion of how an improved understanding of nuclear latency could influence policy debates regarding nuclear disarmament, managing the fuel cycle, and the future of the NPT. (MIS)QUOTING MOHAMED

In his September 2004 address to the IAEA General Conference, Director General Mohamed ElBaradei stated: Some estimates indicate that 40 countries or more now have the know-how to produce nuclear weapons, which means that if they have the required fissile material-high enriched uranium or plutonium-we are relying primarily on the continued good intentions of these countries, intentions which are in turn based on their sense of security or insecurity, and could therefore be subject to rapid change. Clearly, the margin of security this affords is thin, and worrisome.' The subtlety of ElBaradei's argument and conditional nature of this prediction-"ifthey have the required fissile material"-were, however, often ignored. A National Defense University study, for example, cited ElBaradei when it claimed that there now exists "a high degree of nuclear latency that challenges traditional thinking about nuclear threats": "whereas 30 or 40 years ago, only a handful of countries were assumed to know how to acquire nuclear weapons, as many as 35 or 40 nations currently are believed to be in the know." 2 The Austrian government switched "know-how" into "technical capability" in its 2007 NPT Preparatory Committee statement: "Approximately 40 countries are said to have the technical capability to produce nuclear weapons."3 In some cases, the problem went beyond imprecise language to clearly misleading claims, such as the 2005 pronouncement by Green peace: "Through the IAEA's worldwide support of nuclear power, 35-40 countries today have the capability of building atomic weapons in several months, as Dr. ElBaradei recently admitted."4 A second commonly cited statement is ElBaradei's October 2006 comment about the potential for many "virtual nuclear weapons States" around the globe: Verifying enrichment facilities or reprocessing facilities is quite difficult and the so-called conversion time is very short. So we are dealing with what I call

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

83

"virtual nuclear weapon States." One of the issues I have been talking about for a number of years is the need to develop a new international or multinational approach to the fuel cycle so as to avoid ending up with not just nine nuclear weapon States but another 20 or 30 States which have the capacity to develop nuclear weapons in a very short span of time. 5 Again, the conditional nature of this prediction-referring to a future world if more states develop independent enrichment or reprocessing technologywas widely ignored. For example, Zia Mian claimed that "Mohamed ElBaradei of the [IAEA] warned that there are another 20 or 30 'virtual nuclear weapons states' that have the capacity to develop nuclear weapons in a very short time span."6 John Peffer, Marcus Raskin, and Kevin Martin similarly claimed that there are "20-30 virtual nuclear weapons states, which Mohamed ElBaradei of the ... IAEA ... warns have the capacity to develop nuclear weapons in a very short time span." 7 Such statements by journalists, diplomats, and scholars clearly exaggerate the number of states that could build nuclear weapons in a short time period. But they also point to a deeper conceptual confusion about the nature of nuclear latency. How could one measure it? What are the relationships among acquiring research or power reactors, constructing uranium enrichment or reprocessing facilities, developing fissile material, and acquiring nuclear weapons? A logical place to start would be the political science literature on nuclear latency. Unfortunately, such literature on the subject has not been very helpful, in large part because it has failed to focus on the nuclear fuel cycle. THE POVERTY OF POLITICAL SCIENCE ON NUCLEAR LATENCY

This failure is due in part to the tendency in social science scholarship of the past decade to focus on the "demand side" of the nuclear proliferation puzzle. This emphasis on the motives of governments to develop or refrain from developing nuclear weapons was understandable. Much of the earlier political science literature had focused on the effects of arms control treaties, export controls, and other technical constraints on the supply of nuclear materials and technology, and an increase in attention to why some governments wanted nuclear weapons, and why some governments did not, was clearly needed. 8 Moreover, the emergence of nuclear "proliferation rings," such as the A. Q. Khan network and other technology smuggling efforts, encouraged fears that it would be increasingly difficult to prevent weapons proliferation

84

SCOTT D. SAGAN

through "supply-side" constraints. 9 Finally, as will be demonstrated shortly, the quantitative literature on the "correlates of nuclear proliferation" began to present surprising evidence on the rapid growth in the number of states that were "nuclear weapons capable." Such considerations have led some prominent political science scholars to be highly skeptical of efforts to control the supply of nuclear technology for the purposes of nonproliferation. For example, in his innovative book The Psychology of Nuclear Proliferation, Jacques Hymans argues that "given the widespread diffusion of nuclear capacities, supply-side control measures against potential proliferant states are clearly of declining utility": A stricter regime will likely do nothing to change proliferation intentions. It is highly unlikely that more stringent controls will dissuade oppositional nationalist leaders from seeking the bomb and in today's freewheeling global market they probably will be able to obtain the materials they need to build it, albeit perhaps more slowly and with difficulty. More problematically, it is highly unlikely that more stringent controls will dissuade sportsmanlike nationalist leaders from resisting the nonproliferation regime. Indeed, the harsher the regime becomes, the more likely that both types of nationalists will resent and resist it ... In short, the construction of ever-higher supply-side hurdles to civilian nuclear development, far from "strengthening" the non-proliferation regime, is in fact likely to leave the regime even weaker than it is today. 10 Hymans admittedly does not want "to abandon the NPT regime," although he does "second-guess the continual urge to 'strengthen' it with ever-heavier supply-side controls." 11 Yet Hymans is by no means alone among the political scientists who have studied nuclear proliferation in criticizing continuing efforts to control nuclear technology through the NPT. Dong-Joon Jo and Erik Gartzke also find a strong positive correlation between a state getting "latent nuclear weapons production capabilities" and the initiation of a nuclear weapons program, and they maintain that "the inhibiting effect of the NPT is overcome by the stronger technological diffusion effect," concluding that "enthusiasm for the NPT among proliferation opponents thus appears to be misplaced." 12 Harsh Pant has been even more dismissive of the NPT, claiming that the treaty "was never sustainable and has had little, if any, effect on the pace of nuclear proliferation." 13 Such conclusions regarding the NPT are not warranted based on historical evidence. The underlying assumption, that more and more states have

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

85

become "nuclear weapons capable," depends crucially on how one defines and measures "nuclear latency." An important weakness, however, in the ways in which political scientists have measured nuclear latency becomes clear only after one delves deeply into the methodology used in their studies. What does it mean for a state to be "capable" of producing nuclear weapons? Hymans's assessment of what he calls "latent nuclear capabilities" is based on the methodology and data set used in Stephen Meyer's pioneering 1984 book The Dynamics of Nuclear Proliferation and Richard Stoll's update of the Meyer data through 1992. Meyer's study carefully measured a set of ten technical and economic indicators-national mining activity, indigenous uranium deposits, metallurgists, steel production, construction work force, chemical engineers, nitric acid production, electrical production capacity, nuclear engineers, physicists, chemists, and explosive and electronics specialists-to produce what he called "a list of nations with latent capabilities to manufacture nuclear weapons." 14 Not being able to measure directly whether the quantity or quality of a state's nuclear engineers and its explosive and electronic specialists were sufficient to build a nuclear weapon, Meyer used two proxy indicators: whether the state had been operating a research reactor for three years (the proxy for nuclear engineering expertise) and whether the state manufactured automobiles or assembled automobiles and manufactured radios and television sets (the proxy for explosive and electronics specialists). Based on this particular set of indicators, Meyer found that thirty-four states held the latent capability to build nuclear weapons in 1982. 15 Stoll updated the Meyer data set, but with a hidden yet significant change in coding rules, in the mid-1990s: While Meyer measured indigenous uranium sources, Stoll assumed that all states had access to nuclear materials, arguing that they were now freely available in the global marketplace. Stoll thus simply assumed away the crucial issue of whether a state had access to uranium that, once enriched, could be used in a nuclear weapons program. Based on the resulting data set, Stoll argued that forty-eight states had a latent nuclear weapons capability in 1992, noting that: A country is said to have a latent capacity when it has sufficient technical, industrial, material, and financial resources to support a wholly indigenous weapons program. Even though a state may have a latent capacity, it must still make an explicit decision to develop the particular facilities necessary to create weapons. However, once a state has a latent capacity, it is very difficult-perhaps impossible-to deny it nuclear weapons, since it is in essence self-sufficient. 16

86

SCOTT D. SAGAN

50

..----

45 40 35 Ill

--... $

1'0 I ll

30

0

25

4)

.c

E ::I

z

20

)

15 10 5

..-

/

~

.....

.../

I I I I

/

.....

0 1~1~1~1~1~1~1~1~

Year .....,. Nuclear-capable states Figure 5.1.

-

1~

2000

States with nuclear weapons

Stoll's nuclear latency estimate.

SouRCE: Data from Richard Stoll, "Latency Capacity Proliferation Model"; available at http://es.rice.edu/projects/ Poli378/Nuclear/Proliferation/

Hymans accepted the logic of that argument and claims that the Stoll data demonstrate that there was a "yawning gap between technical potential and military reality" in terms of the number of states that have capability to produce nuclear weapons compared to those that have actually done so. 17 (I depict the Stoll data in Figure 5.1.) A state can not make a nuclear weapon, however, unless it has HEU or plutonium from a large reactor, and Stoll's hidden assumption that any state could acquire uranium on the open market, coupled with his use of research reactor experience as the measure of required nuclear engineering expertise, essentially assumed away those two technical constraints. Moreover, looking at his data set reveals that even though Stoll argued that each of his ten criteria were "necessary conditions for the production of nuclear weapons," North Korea, which the IAEA discovered had taken spent fuel rods containing

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

87

plutonium from the Yongbyon reactor in 1989, 1990, and 1991,18 was not considered capable ofbuilding a nuclear weapon according to Stoll's model in 1992. North Korea lacked the necessary nitric acid production capability, chemical engineers, and electronic/explosives specialists (as measured by domestic automobile and radio/television industry). This observation obviously raises questions about whether these particular measures of nuclear latency really are "necessary conditions" for a state to develop nuclear weapons. Jo and Gartzke's 2007 Journal of Conflict Resolution study "The Determinants of Nuclear Weapons Proliferation" improves on the Stoll coding scheme by dropping three of the Stoll and Meyer indicators (construction workforce, steel production, and previous national mining activity) on the grounds that they are "too easily available to be thresholds" and modifying the coding for necessary "uranium deposits" (which, as we have seen, were assumed to be Table 5.1.

Comparison of Meyer/Stoll and Gartzke and Jo nuclear capability indicators. Definition of indicator Indicator

Meyer/Sto/1

Gartzke and Jo

Mining activity

Some fraction oflabor force in mining activity

Uranium deposits

Known uranium deposits (Meyer) Assumed market access (Stoll)

Known uranium deposits or produced uranium already

Metallurgists

Production of crude steel

Production of crude steel or aluminum

Steel

Production of crude steel

Construction work force

Production of steel and cement

Chemical engineers

Production of nitric acid or sulfuric acid

Production of nitric acid or sulfuric acid

Nitric acid production capacity

Nitric acid production or sulfuric acid production and nonorganic nitrogenous fertilizer production

Nitric acid production or sulfuric acid production and nonorganic nitrogenous fertilizer production

Electricity production capacity

Installed electrical capacity of200MWe

Installed capacity of 200MWe or produces equivalent of 50,000 metric tons of oil

Nuclear engineers/ physicists/chemists

Three research- reactor years

Three research-reactor years

Electronics/explosives specialists

Manufacture or assembly of motor vehicles and manufacture of radios orTVs

Manufacture or assembly of motor vehicles and manufacture of radios orTVs

88 SCOTT D. SAGAN

available for all states by Stoll) to include either a state with uranium deposits on its territory or one that has acquired "produced uranium" for a research or power reactor. 19 (See Table 5.1 for a comparison of the various coding schemes.) Yet, although Jo and Gartzke correctly note that "states that lack the basic material capabilities will be excluded from the group of potential proliferators," their model actually does not treat nuclear materials (enriched uranium or plutonium) as a necessary, but not sufficient, condition for building nuclear weapons. 20 Instead, their model implicitly assumes that "where there is a will, there is a way" and that a latent nuclear weapons state, or even a state that has already made nuclear weapons, may not actually have the necessary materials. That the Jo and Gartzke model does not therefore adequately capture the necessary conditions for nuclear weapons development can be best seen in the representation of their data set for 2001 presented as a map in Figure 5.2. 21 The Jo and Gartzke data set continues to show that North Korea did not have a full latent capability to develop nuclear weapons in 2001 (it still lacked sufficient chemical engineers, nitric acid production capability, and explosives specialists), even though the North Koreans were a major exporter of longrange missiles at the time and were known to have separated plutonium from the fuel rods of the Yongbyon reactor. 22 We now also know that North Korea tested its first nuclear weapon in 2006. The Jo and Gartzke coding rules also lead to the odd conclusion that South Africa, which built six nuclear weapons in the 1980s, dismantled the weapons in the 1990s but still maintained from 450 to 6oo kg of HEU under IAEA safeguards inspections in 2002, nevertheless lacked the full capability to build nuclear weapons. 23 (South Africa lacked sufficient chemical engineers and nitric acid production capacity, according to their data.) By not focusing attention on enriched uranium and plutonium, the weak proxy measures of nuclear latency used in this work clearly lead to bizarre results. Trinidad and Tobago (which "only" lacks uranium deposits, "produced uranium," and any research reactor experience) is assessed to have a higher degree of nuclear weapons latency in 2001 than is North Korea, only five years away from detonating its first nuclear weapon. In Africa, Egypt, which had only two research reactors in 2001, is assessed to have a higher degree of nuclear weapons latency than is South Africa. In addition, Jo and Gartzke usefully try to assess the nuclear diffusion effect in an attempt to understand whether "knowledge of how to construct nuclear weapons has spread with the passage of time" and how much this has

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

89

Key •

7 (Full latency)

-~ 03-4 D o.-2

Nuclear latency according to Jo and Gartzke. SouRcE: Data from Dong-Joon Jo and Erik Gartzke, "Dataset for 'Determinants of Nuclear Figure 5.2.

Proliferation: A Quantitative Model,"' journal of Conflict Resolution 51, no. 1 (2007), available at: http://jcr.sagepub.com/cgi/content/full/sl/ l /I67/DCI

influenced proliferation. 24 Their measure of "nuclear diffusion," however, is devoid of substantive content. Because Jo and Gartzke had no direct measure of the spread of nuclear knowledge over time, they simply assumed that "diffusion equals the log transformation of the number of years since 1938." 25 This variable may measure some temporal factor that affects changes in proliferation behavior over time but does not, contrary to their claim, "suggest that the NPT may actually contribute to the quickening pace of nuclear diffusion."26 In conclusion, these problems point to serious weaknesses in the political science literature analyzing nuclear latency and the effects of the NPT. All too often this literature measures "proxy variables" that are easily available, rather than collecting the data that reflect the variables of real interest. All too often there are hidden, but crucial, assumptions that have a strong impact on the findings in ways that are not acknowledged. Most importantly, the political science literature, like the misinterpretations of ElBaradei's statements cited earlier, often conflates two analytically different phenomena under the same labels of "nuclear weapons capability" or "nuclear latency"- first, what should properly be called "nuclear self-sufficiency" (a measure of how independent a potential long-term nuclear weapons program could be); and,

90

SCOTT D. SAGAN

second, "nuclear latency" (a measure of how quickly a state could develop a nuclear weapon if it chose to do so from its current state of technological development)Y The former may have value in estimating whether a state could develop nuclear weapons eventually, over an extended period of time, even if it received no technical assistance or nuclear materials from other states. This is a phenomenon, however, that has never happened in world history, as all nuclear weapons states have received some degree of assistance from others, as Itty Abraham has compellingly demonstrated: Even the first nuclear weapons states, the United States and the Soviet Union, were not self-sufficient, having received assistance from German scientists, scientists from closely allied nations, and spies from other states' programs. 28 In short, by focusing our attention away from the acquisition of the fissile materials needed to make a nuclear weapon, the political science literature has led to an exaggerated estimate of how many states currently have the technical capability to build nuclear weapons. The criticisms of the NPT regime cited above-that it has been ineffective or even may have led to the widespread diffusion of latent nuclear weapons capability-are therefore unwarranted. The NPT regime clearly has flaws, and we know of many cases of states that have violated their commitments, but the NPT has not (at least not yet) led to a world in which there are dozens of non-nuclear weapons states that could easily become nuclear weapons states in a short period of time. TECHNOLOGY, THE FUEL CYCLE, AND NUCLEAR WEAPONS

The biggest problem in these political science studies was their failure to focus on the fuel cycle technology necessary to produce the fissile materials needed for a weapon. Without a large nuclear reactor to produce plutonium (and a reprocessing facility) or the capability to produce enriched uranium, no state could build its own nuclear weapon. Does a state have nuclear power reactors or other reactors that produce plutonium? How large are its research reactors, and do they run on low enriched uranium (LEU) or HEU? Does it have the technological capability to produce, separate, or reprocess plutonium? Does the state have an enrichment facility to produce HEU? How long would it take to build such facilities and then to build a nuclear bomb with the materials? Such questions should be at the core of any assessment of a state's latent nuclear weapons capability. Fortunately, there is a strong technical literature on nuclear power and proliferation that addresses such questions. Yet even the most thorough of

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

91

these studies ignore important political factors that strongly influence proliferation decisions and their implementation. I will review three important studies of nuclear technology to illustrate the point. The pioneering study on this subject was the 1977 book Swords from Plowshares, produced by an interdisciplinary team of scholars led by Albert Wohlstetter.29 Wohlstetter and his colleagues divided non-nuclear weapons states into three categories-states with advanced infrastructure and fissile material, states with a research or power reactor, and states with no nuclear experience at all. In the first category were the nine states that were estimated to have "full access to the fissile material required to make a weapon" in 1977: Japan, West Germany, South Africa, Belgium, Taiwan, Italy, the Netherlands, Canada, and Sweden. The study argued that each of these states could take the following four remaining steps to produce a nuclear weapon within one year, assuming the steps were undertaken in parallel: converting the fissile material in their possession into metallic form, designing a weapon, fabricating the weapon and its components, and preparing for and conducting a nuclear test. 30 In the second category, Swords from Plowshares listed fifteen states that had a reactor in 1977 but no access to fissile material outside of the spent fuel rods: Israel, Argentina, Switzerland, Egypt, Spain, South Korea, Indonesia, East Germany, Czechoslovakia, Australia, Pakistan, Iran, Norway, Brazil, and Mexico. The study assumed that the most practical pathway for such states to get a nuclear weapon would be to construct a reprocessing plant, which the study estimated would take four years. It assumed that the four final weaponization steps listed above could be done in parallel to the reprocessing efforts (though the authors never explained how a state could produce plutonium metal before the reprocessing plant was completed and in operation) and thus estimated that any of these states could make a nuclear weapon in four years. 31 Finally, states in the third category were estimated to need at least six years to build a nuclear reactor and simple reprocessing facility "from scratch." (See Table 5.2 for the countries in each category.) Wohlstetter and his team acknowledged that these timelines were "engineering estimates based on American experience" and should therefore only be considered "rough approximations." 32 Swords from Plowshares was an influential and prescient study. Wohlstetter and his colleagues identified the danger of what they called "nearing the bomb without breaking promises not to make it" under the NPT. 33 They discussed the need for a more extensive system of IAEA inspections of non-nuclear weapons states' facilities and produced one of the first analyses of the pros and cons of building multinational nuclear fuel facilities. With the passage

92

SCOTT D. SAGAN

Table 5.2.

Wohlstetter et al.'s nuclear capability timeframes. Capability level (NNWS only)

Advanced nuclear infrastructure plus "full access to fissile material needed for bomb"

Research or power reactor; no fissile material except in reactors and spent fuel rods

No nuclear infrastructure

Countries qualifying (as of 1977)

Japan, West Germany, South Africa, Belgium, Taiwan, Italy, Netherlands, Canada, Sweden (9)

Israel, Argentina, Switzerland, Egypt, Spain, Republic of Korea, Indonesia, East Germany, Czechoslavakia, Australia, Pakistan, Iran, Norway, Brazil, Mexico (15)

All other states

(Assumed) obstacles remaining

(a) Convert fissile material to metal; (b) design weapon; (c) fabricate weapon; (d) test

(a) through (d) plus concurrent construction of reprocessing plant

Build reactor and reprocessing plant

Estimated time to first bomb

1 year

4 years

6+ years

SoURCE: Data from Albert Wohlstetter, "Thomas A. Brown, Gregory jones, David C. McGarvey, Henry Rowen, Vince Taylor, and Roberta Wohlstetter, Swords from Plowshares: The Military Potential of Civilian Nuclear Energy (Chicago: University of Chicago Press, 1977).

of time, however, one can also see some obvious weaknesses in this pioneering work. First, by basing the estimates for nuclear latency time lines on the U.S. experience, the study implicitly assumes that all proliferators are likely or able to follow the same technological pathways, with the same degree of bureaucratic and organizational competence and the same degree of wartime and Cold War urgency. Wohlstetter and his colleagues also assumed that once a government made a decision to develop nuclear weapons, there would be a high degree of political consensus behind it and the weapons program would progress without significant internal or external constraints. The estimate on the "starting from scratch" scenario seems particularly low and appears to contain a hidden assumption that the world market would encourage such growth in reprocessing facilities and that the United States and the global nonproliferation regime would not add additional political hurdles to constrain that growth. To be fair, the alarm bell sounded by the Wohlstetter study was a kind of self-denying prophecy because it was a major contributor to the development offuture nonproliferation innovations, at least in the United States. 34 Swords

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

93

from Plowshares also contained a useful discussion of the need for more rigorous agreements among governments that exported nuclear technology on what could be sold and what could not be sold, even to NPT member states. Still, this important 1977 technical study could have been supplemented at the time (and certainly needs to be supplemented now) with both analyses of alternative weapons development pathways and historical data on actual political decision making and bureaucratic implementation in different states' programs that could provide more accurate evidence to support estimates of future latency timelines. The second study focused on technology contributing to nuclear latency is the 2006 Science and Global Security article by Robert Harney and his colleagues entitled "Anatomy of a Project to Produce a First Nuclear Weapon." 35 The authors identify and provide a complex timeline of the 196 necessary tasks required to produce a uranium-based nuclear weapon by a state that has produced or acquired 120 metric tons of yellowcake. The tasks include production of enrichment-plant feed material (UF6), uranium enrichment (with timelines for different enrichment methods), production of HEU metal, and finally the design and construction of actual nuclear weapons. 36 Using measures of time, labor, energy, and necessary money required to complete these tasks under both normal and expedited ("crash") conditions, the authors provide an estimate of the earliest possible completion time: under normal conditions, approximately 338 weeks (six and a half years) would be needed to produce six weapons, and under crash conditions 260 weeks, or just under five years, would be required. 37 Harney and his coauthors also provide an estimate of an Iranian withdrawal from the NPT scenario: According to their model, it would still take 216 weeks (about four years) for a state that has a prototypic uranium enrichment plant already in place to produce its first nuclear weapon. 38 The Harney and colleagues study usefully highlights the complexity and time needed to produce uranium metal from highly enriched uranium and for subsequent steps in the weaponization process. By developing estimates of both crash programs and noncrash programs, the study brings one political variable into its analysis. But there are nonetheless limits to its utility as a guide to understanding nuclear latency in future proliferation scenarios. First, the authors, like those contributing to the Wohlstetter study, base their estimates on the U.S. experience with nuclear materials and weapons production. 39 This could bias the results in both directions, though it is difficult to

94

SCOTT D. SAGAN

know which bias would be stronger. On the one hand, the study assumes a high degree of organizational competence on the part of future proliferators, which would likely make the Harney estimates lower than is likely to occur in most new proliferant states in developing world. On the other hand, the study also assumes that future proliferators would copy the later U.S. Cold War penchant for careful testing, stringent safety, and high reliability, tasks that take time and would make the estimates derived from the model much too long. (For example, the study estimates it would take twenty-four weeks to verify gun velocity, twenty-four weeks to build a delivery vehicle compatibility mock-up, and forty weeks to finalize the weapons design. 40 ) From a U.S. weapons designer's view, such assumptions may reflect normal peacetime procedures; from a proliferation prevention perspective, they may constitute a wishful thinking, best-case analysis. For example, if a new proliferant government decides to use a simple uranium gun-type device or has access to the more advanced nuclear bomb designs that were peddled by the A. Q. Khan network, these time-consuming activities could be greatly reduced. The Harney study also assumes that a proliferant government has not developed covert facilities to jumpstart the weaponization process, having most essential tasks completed before the HEU is produced. The third, and most historically grounded, technical study on nuclear latency is a detailed 2005 report from the Pacific Northwest National Laboratory (PNNL): Nuclear Proliferation Technology Trends Analysis. 41 The PNNL study does not base its estimates on the U.S. experience with different nuclearrelated technologies but rather gathers and presents the available data on the experience and time lags seen in the history of many different states' uranium enrichment and plutonium production and reprocessing programs. The study is the most detailed analysis, at least in the unclassified literature, of the history of successful efforts either to enrich uranium or reprocess plutonium. The PNNL authors wisely do not "sample on the dependent variable" and therefore also analyze failed or abandoned programs to enrich uranium or reprocess plutonium. I reproduce the concluding estimates about nuclear materials production time lines from the PNNL study in Table 5-3· This PNNL study is valuable, but its claims to be able to use history to provide accurate estimates about technical time constraints that continue to apply today are problematic. The study correctly notes that "the time required for success varies widely and is strongly dependent on either help from nations that have already developed the technology or the nuclear and industrial maturity of the nation." 42 But the authors do not analyze the evolution of

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

ible5.3.

95

PNNL nuclear capability timeframes.

Technology

Number of countries interested in technology

Number of countries with successful production programs 1

Average time to pilot plant'

Average time to productioi

Gaseous diffusion enrichment

6

5

Centrifuge enrichment

18

7

Electromagnetic isotope separation

11

2 years

3 years

Chemical isotope separation

3

6 years

11 yean

Aerodynamic isotope separation

3

7 years

18 years

Laser enrichment

14

6 years 8 years

14 yean

Graphite-moderated production reactors

6

6

1 year

2-11 year

Heavy-water-moderated reactors

12

5

1 year

2-6 year

Research reactors

14

3

Reprocessing

19

13

4-5 year

6 years

10 years

SouRCE: M.D. Zentner, G. L. Coles, and R. ). Talbott, Nuclear Proliferation Technology Trends Analysis, Pacific .rorthwest National Laboratory, September 2005; available at www.pnl.gov/main/publications/external/technical _reports/PNNL-14480.pdf 'More than gram quantities of material produced [note in original]. 'Technological capability demonstrated [note in original]. 3 Significant quantities of material produced [note in original]. 'Note that "average" times given in this cell and the two cells immediately below it are in fact ranges, not average.

international export controls on nuclear technology, which have strengthened over time and thus influence late developers more than early developers of some nuclear materials production technologies. The authors maintain that "based on an evaluation of historical trends in nuclear technology development, conclusions can be reached concerning: 1) the length of time it takes to acquire a technology; 2) the length of time it takes for production of special nuclear material to begin; and 3) the type of approaches taken for acquiring the technology."43 History surely can provide an answer to point three, but the other two questions are historically contingent on the characteristics of the states involved and can change depending on the spread of other related technologies, organizational learning, illicit networks of suppliers, and shifts in foreign government assistance. The authors' calculations of"average" times to pilot plants and "average" times to production success are interesting but have a peculiar ahistorical character to them; and, even if the full ranges of historical time lines were presented, using the lowest number as a "worst-case estimate" would be problematic because these time lines include advanced industrial states as well as less-developed nations. In short, future scholars

96

SCOTT D. SAGAN

should be more cautious than were the PNNL authors when they concluded that "the timeframes identified above can be considered representative of current development efforts." 44 TOWARD A POLITICAL THEORY OF NUCLEAR LATENCY AND PROLIFERATION

These considerations lead me to conclude that any general measure of"nuclear latency" is likely to be a chimera. Nuclear latency is not like human pregnancy, in which all women have virtually the same nine month gestation period. Different non-nuclear weapons states, even those starting from the same technological threshold, are likely to take different lengths of time to move to possession of a single nuclear weapon or a usable arsenal. Where should the political science and technical community go from here to improve our understanding of nuclear latency and the risks of nuclear proliferation in the future? And how would such research help us understand how the NPT works to constrain proliferation? This conclusion outlines a multidisciplinary research agenda and suggests some principles that should guide future research to avoid some of the weaknesses of past efforts in this area. I then conclude with some policy-relevant observations about how new agreements regarding the fuel cycle facilities and produced fuels could strengthen the NPT regime. First, we need more research on the domestic characteristics of the regimes that have been successful and of those that have been less so in their attempts to develop uranium enrichment and plutonium reprocessing capabilities. Such analysis would have to take into account the degree to which the "failures" or longer time lines for successful states were due to internal characteristics or external constraints. The simple division between democracies and nondemocracies may be less helpful in this regard than a focus on the relationship among the political leadership and the scientific community and military. Jacques Hymans, for example, usefully theorizes that "neopatrimonial" or "sultanistic" regimes-governments characterized by extreme personalized rule, use of state resources to buy off clients, and an absence of checks and balances-will take longer to develop advanced levels of nuclear technology and will fail more often in attempts to move from one technological threshold to the nextY Hymans argues that North Korea fits this model and compares it to Romania, where an unsuccessful program was run by Elena Ceau~escu, the president's wife, who hired scientists based primarily on whether they would

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

97

promote her candidacy for the Nobel Prize in chemistry; and Libya, which was described by the 2005 WMD Commission as "an inept bungler, the court jester among the band of nations seeking biological and nuclear capabilities."46 This is a promising approach, yet it is important to note that North Korea, unlike Romania and Libya, was able to produce fissile materials despite international sanctions. North Korea tested a nuclear device in 2006, though this device was not as effective as was apparently expected, and another in 2009, widely considered to have been successful. 47 In addition, it is worth noting that Saddam Hussein's regime in Iraq was characterized by massive corruption, a culture of fear that led to exaggeration of progress by laboratory officials and military commanders alike, and a decision-making style that, according to his senior colleagues, "verged on the mystical."48 Despite such pathological decision making and leadership, however, the Iraqi covert program was discovered after the 1991 Gulf War to be much closer to producing a nuclear weapon than the CIA had estimated. 49 Another potential political constraint on nuclear weapons programs can be the rivalries for power between different leaders in potential proliferators. In Egypt in the 196os, for example, Gamal Abdel Nasser started a nuclear weapons program but did not give it high priority or a large budget, in part because the head of the nuclear program was a strong ally of Nasser's chief rival, Abdel Hakim Amer. As one former military officer later explained: "We didn't want to create heroes in the system that a nuclear bomb would create." 50 In short, more research on how regime characteristics influence the ability to develop both fissile materials and nuclear weapons should be pursued. We have a strong literature on how regime type can influence decisions about whether to seek nuclear weapons. But we lack broader studies of how regime characteristics influence the ability to implement decisions to acquire sensitive nuclear technology or use such technology to move closer toward developing a nuclear arsenal. Such research should avoid, however, the common assumption that governments that seek nuclear weapons options have already decided to get the bomb. As Itty Abraham argues, the proliferation/nonproliferation lens through which scholars commonly study nuclear history can blind us to the diversity of motivations of different bureaucratic actors within states and even the mixed motives of individual leaders. The nonproliferation literature commonly refers to "nuclear ambiguity": the lack of knowledge about whether a foreign government is pursuing nuclear weapons or only nuclear energy. In some cases, however, a government would be better described as experiencing

98

SCOTT D. SAGAN

"nuclear ambivalence" because its leadership is undecided or deeply conflicted about different options for future nuclear development. 51 A second line of research could focus on the time period between when states acquired weapons-usable fissile materials and when they tested a weapon or had a suspected nuclear weapons arsenal. Such research would be difficult, of course, because of the lack of firm information on the dates of nuclear program initiation and successful acquisition of a nuclear weapon in many historical cases. 52 Even if accurate dates were available, however, scholars should be careful not to assume that each of these different governments was seeking nuclear weapons with the same degree of urgency or unity. Even among the states that did eventually develop nuclear weapons, one finds some governments initiating nuclear weapons programs on a crash basis in wartime or crisis conditions, others slowly developing a hedge for an uncertain future in peacetime, and others initiating a nuclear power program or peaceful nuclear explosive program with only minimal interest in nuclear weapons applications. This line of research could also usefully focus on the political and technical factors that influence the size and characteristics of the nuclear arsenals sought by new nuclear proliferants. Any analysis of nuclear latency timelines should therefore be mindful of the assumptions used not only about the starting point (the capacity of an individual enrichment facility or reprocessing facility) but also about the end point sought by the state (a single weapon, an arsenal, a simple nuclear device, a miniaturized warhead?). The characteristics of the arsenal sought by a government have a major impact on the financial costs, the technical hurdles, and the time involved in a nuclear program. For example, although outside observers often claim that it would take Japan six months to a year to develop nuclear weapons if it chose to do so, 53 an internal Japanese government study after the North Korean test reportedly calculated that it would take Japan three to five years to construct a breeder reactor, expand its reprocessing facility, and build a prototype of a miniaturized warhead capable of fitting onto a missile to counter the DPRK threat. 54 Unfortunately, we know little about what would-be proliferators think about what kinds and levels of nuclear weapons might be needed to meet their security requirements. Third, it would be useful to have more technical and political science research on the effectiveness of past efforts to control the spread of uranium enrichment and plutonium reprocessing. Recent studies by Matthew Fuhrmann and Matthew Kroenig have contributed significantly to the field by examining

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

99

the political and economic causes and the security consequences of international exports of sensitive nuclear facilities, nuclear cooperation agreements, and sales of nuclear reactors. 55 This line of research could be extended to examine the effectiveness over time, or lack thereof, of strengthened export controls, the creation of the Nuclear Suppliers Group, and rise and fall of the A. Q. Khan network. Finally, there should be more integrated studies of supply and demand for nuclear weapons. Instead of thinking about these two "sides" of nuclear proliferation as separate issues, we need to recognize, and therefore study, the potential for complex connections between supply and demand. Three such interactions are obvious. First, how hard a government works on a nuclear weapons program-the resources it commits to the program and whether it is engaged in a crash effort or normal construction effort-is likely to be affected by the severity of its demand for a weapon. Second, a high degree of nuclear capability or latency could influence demand by enabling actors favoring a nuclear weapon to argue that acquiring a weapon is easier than would otherwise be the case. Third, a high degree of latency could make it easier for a pronuclear weapons party or individual leaders to implement a decision to acquire nuclear weapons if they are in power for only a brief period of time. NUCLEAR POWER WITHOUT NUCLEAR PROLIFERATION?

Concerns about climate change and growing energy demand have sparked a global resurgence of interest in civilian nuclear power generation. The potential increase in nuclear power reactors in experienced states and the spread of nuclear power to new states will inevitably boost the demand for enriched uranium to fuel such reactors and increase interest in reprocessing to recycle plutonium from the back end of the fuel cycle. Will it be possible to have the expansion of nuclear power without enhancing the risk of global nuclear proliferation? A final research task would be to understand more about how the growth of civilian nuclear power bureaucracies in different states influenced their decisions to seek or renounce nuclear weapons. Indeed, how best to ensure that civilian nuclear power bureaucracies maintain a strong interest in opposing nuclear weapons proliferation may be the proverbial "$64,000 question" for the future of nuclear nonproliferation. (It will be closer to a $64 billion question in reality.) On the one hand, leaders of successful nuclear power enterprises would likely want to maintain strong ties to the global financial

100 SCOTT D. SAGAN

markets, nuclear power industry, and regulatory agencies and hence seek cooperation with the nuclear nonproliferation regime. On the other hand, leaders of less successful or struggling nuclear power enterprises might be more likely to support nuclear weapons development programs as tools to justify their existence and budgets within their state. Etel Solingen notes how crucial the former factor has been in promoting nuclear weapons restraint in East Asia; Itty Abraham has, in contrast, demonstrated that the weak record in producing nuclear power encouraged India's "strategic enclave" in the nuclear laboratories to lobby New Delhi to acquire nuclear weapons. 56 The degree to which the expansion and spread of nuclear power will increase individual countries' latent nuclear weapons capability will be largely determined, however, by who manages and controls uranium enrichment facilities and plutonium separation and reprocessing facilities. Fortunately, there are many potential reforms of the international regime, including the NPT, that could reduce states' incentives and capabilities to acquire nuclear weapons even in a world of expanded nuclear power. First, as demonstrated in E!Baradei's statements discussed earlier, there is significant interest in the creation of international nuclear fuel banks, multinational fuel production facilities, and spent fuel take-back arrangements that would permit expansion of nuclear power without expansion of national fuel production facilities. Second, the Nuclear Suppliers Group or even a future NPT Review Conference could discourage states from exercising their Article X right to withdrawal from the NPT by making future sales of nuclear fuels and sensitive technology subject to a "Return to Sender" agreement. U.N. Security Council Resolution 1887 explicitly encourages such a "Return to Sender" policy by requesting that nuclear supplier states: require as a condition of nuclear exports that the recipient State agree that, in the event that it should terminate, withdraw from, or be found by the IAEA Board of Governors to be in non-compliance with its IAEA safeguards agreement, the supplier state would have a right to require the return of nuclear material and equipment [ ... ] as well as any special nuclear material produced through the use of such material or equipment. 57 Unfortunately, many non-nuclear weapons states fear that efforts to enhance international controls over the nuclear fuel cycle cut against their "inalienable right" as specified in Article IV to enjoy the benefits of nuclear energy. These concerns should, however, be balanced by the hope that both nuclear

NUCLEAR LATENCY AND NUCLEAR PROLIFERATION

101

disarmament and nonproliferation will be enhanced by strong international control of the fuel cycle. George Perkovich and James Acton have compellingly argued that weapons states are less likely to agree to complete nuclear disarmament or go to extremely low numbers in the future if many other states have their own uranium enrichment or reprocessing facilities and could therefore develop nuclear weapons openly and more rapidly in an NPT withdrawal scenario or covertly in peacetime. 58 There is, therefore, an important logical connection between future controls over the nuclear fuel cycle and nuclear disarmament. The governments of non-nuclear-weapons states should recognize that entering into negotiations about international control of the nuclear fuel cycle is part of their Article VI commitment "to pursue negotiations in good faith on effective measures relating to cessation of the nuclear arms race." 59 If nuclear disarmament is ever to occur, it is obvious that all the nuclear weapons states will need to negotiate to reduce and eliminate their arsenals in a mutual and verifiable manner. It is less obvious, but no less true, that the prospects for eventual nuclear disarmament will be linked to the success or failure of the global effort to negotiate serious constraints on the global spread of enrichment and reprocessing facilities.

6

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"? An Exercise in Measurement Validation

Jacques E. C. Hymans

become a "nuclear weapons state"? Traditionally, the standard demarcation line has been a state's first nuclear test: Until the test, states are considered to be "nonnuclear"; after it, they are considered to have "gone nuclear." But in recent years, the perceived salience of the test as the definitive marker of a state's nuclear status has been declining. Many scholars and analysts have pointed out that, from a technical point of view, an explosive test may not be strictly necessary for the construction of weapons of fearsome destructive power. Therefore, understandably wishing to avoid strategic surprise, they have increasingly shifted their attention to a prior rung on the nuclear proliferation "ladder": a state's accumulation of enough fissile material for a bomb, also known as a "significant quantity" (SQ). Indeed, no less an authority than former IAEA Director-General Mohamed ElBaradei routinely labeled states that have accumulated an SQ as "virtual nuclear weapons states." 1 ElBaradei defines "virtual nuclear weapons states" as "countries that are able to develop nuclear weapons overnight." 2 Moreover, there is a strong tendency today, especially in the United States, to equate the "virtual" with the "real," given the supposed ease of conversion into weapons and the nearimpossibility for IAEA inspectors to know in timely fashion if an abrupt conversion has taken place. 3 There has never been an authoritative decision to shift the threshold for defining nuclear weapons stateness down from the test/no-test to the SQ/no-SQ indicator. Rather, the shift to SQ/no-SQ has been occurring gradually and even not entirely consciously, driven above all by the pragmatic concerns of policy makers. Not surprisingly, therefore, the test/no-test indicator as well as

WHEN DOES A STATE

102

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

103

the SQ/no-SQ indicator are both still often used. For instance, the well-known nonproliferation advocate Joseph Cirincione uses the test/no-test indicator to hail the NPT as "one of the most successful security pacts in history," but then he uses the SQ/no-SQ indicator to question the NPT's sustainability when even a treaty-respecting country like Japan can become, at least in some eyes, an "associate member of the nuclear club." 4 Although this sort of definitional fuzziness is likely to persist for years, the overall trend over the past two decades has unmistakably been toward the entrenchment of the SQ/no-SQ indicator. Nuclear testing is increasingly seen as a luxury with little real bearing on a state's ability to produce nuclear bombs, at least of the "simple" fission variety. The general trend toward SQ/no-SQ can be seen in the evolution of intelligence assessments of the nuclear status of North Korea. For instance, in 1983 the CIA could still write, "We have no basis for believing that the North Koreans have either the facilities or materials necessary to develop and test nuclear weapons." 5 By contrast, as detailed by Leon Sigal, "A National Intelligence Estimate [NIE] in late November 1993 concluded that North Korea already had one or two nuclear weapons," despite the fact that no test had taken place. 6 Trying to "protect our policy from the flak the estimate generated," top Clinton administration officials did attempt to highlight the NIE's precise language of a "better than even" chance, rather than a certainty, of actually existing North Korean nuclear weapons? But after George W. Bush became president, even such faint attempts to maintain some nuance died away. The CIA under Bush boldly stated in an unclassified 2002 briefing to Congress that the United States "has assessed since the early 1990s that the North has one and possibly two weapons using plutonium it produced prior to 1992." 8 The key point to focus on here is not the progress of North Korea's nuclear program but rather the evolving metric used by the intelligence community to assess that progress. In the 1980s, the key question had been whether North Korea would "develop and test" nuclear weapons; but, by the early 1990s, the intelligence community, having become very uncertain about the test/no-test indicator, was able to assert the existence of actual nuclear weapons simply on the grounds that Pyongyang had allegedly accumulated an SQ with proliferation intent. And this shift in metrics was hardly limited to analysts toiling away in the bowels of Langley. The notion of an actually existing North Korean nuclear weapons stockpile was taken so seriously in Washington that prominent liberal and neoconservative analysts agreed that it helped explain why the

104 JACQUES E. C. HYMANS

United States attacked Iraq but not North Korea in the spring of 2003. 9 Meanwhile even the IAEA's ElBaradei, when pushed in 2004 on whether Pyongyang had actually used its plutonium to produce weapons or merely weapons-ready fuel, replied, "What's the difference?" 10 But when North Korea actually did conduct its first nuclear test in 2006, it became clear that the Bush administration and ElBaradei had been wrong: There was in fact a big difference between having an SQ and having the bomb. North Korea's 2oo6 test device had a pitiful explosive yield of probably less than one kiloton, almost surely not what Kim Jong Il intended.U The failure of the test revealed (unless one gave Pyongyang a huge benefit of the doubt) that as of 2006 North Korea did not have nuclear weapons. Therefore, for over a decade U.S. policy toward North Korea was based on a gross overestimation of Pyongyang's capabilities. The SQ/no-SQ indicator had seriously failed. The case of North Korea shows clearly that the question of what indicators we should use to demarcate between nuclear and non-nuclear weapons stateness is one that carries very high analytic and political stakes. This chapter aims to assess the pluses and minuses of the two most widely used indicators of non-nuclear versus nuclear weapons stateness-test/no-test and SQ/no-SQ-from the perspective of measurement validity. In other words, the chapter asks, which of the two metrics for evaluating a state's nuclear weapons status is best if our goal is to maximize the accuracy of our descriptive inferences? This question, which at first may seem rather pedantic, is in fact a crucial foundation for proliferation assessment, analysis, and policy. My basic argument is that although the test/no-test indicator clearly has its problems, to replace it with the SQ/no-SQ indicator would be to throw the baby out with the bathwater. I then propose a theoretically driven "neotraditional" approach to this knotty question of measurement, which relies on the basic test/no-test indicator as our starting point but then supplements it with a theory-driven analysis of different states' incentives and disincentives to induct nuclear weapons without prior testing. The procedure for measurement validation followed in this chapter is one that was recently spelled out by the political scientists Robert Adcock and David Collier. 12 In their work, Adcock and Collier highlight the importance of carefully considering the validity of a proposed measurement strategy at three key levels: conceptualization, or the formulation of a systematized understanding of the thing to be measured in light of the theoretical meanings and understandings we associate with it; operationalization, or the development of

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

105

specific metrics, or indicators, that clearly reflect the basic concept; and scoring, or the application of those indicators to the universe of relevant historical cases. Although this procedure would appear on the surface to be a quite deductive one, in fact Adcock and Collier stress that information from the more "applied" steps of the process can (and should) stimulate rethinking at the more "fundamental" levels. The road map for the rest of the chapter is as follows. The next section deals with the conceptualization and traditional operationalization of nuclear weapons stateness (that is, the test/no-test indicator); the following one critiques the traditional operationalization; the section after that critiques the proposed alternative operationalization; and the final section suggests an improved "neotraditional" approach to the matter. CONCEPTUAL BACKGROUND AND THE TRADITIONAL OPERATIONALIZATION

It is important first to lay out the conceptual foundations of nuclear weapons

stateness. In other words, does making a clear distinction between "nuclear weapons states" and "non-nuclear weapons states" really help us to better describe reality? Or, for instance, should we instead speak of degrees of nuclear weapons possession in terms of many shades of gray, as we do for various levels of economic advancement? In fact, although perhaps the overall proliferation process can indeed be conceptualized as a "ladder," it has long been recognized that one of the rungs of that ladder must be marked in red. After all, the acquisition of even a small nuclear weapons arsenal is a revolutionary act in international politics. As such, it engenders uncertain but potentially vast consequences for the acquirer's international position and indeed for overall international stability.B Recognition of the revolutionary implications of the birth of new nuclear weapons states has led the international community to devote enormous diplomatic, financial, human, and other resources to prevent it from happening. Among academics, too, the question of why states "go nuclear" receives intense focus, whereas why states "go aircraft carrier," for instance, does not. In short, conceptually there is an important qualitative distinction between the two types of states, nonnuclear and nuclear. Our operationalization of the terms "nuclear weapons state" and "non-nuclear weapons state" needs to respect this conceptual background. The traditional focus on the nuclear test as marking the "birth" of a new nuclear weapons state fits the underlying concept already described quite well.

106 JACQUES E. C. HYMANS

A successful nuclear explosive test is typically a clear, focused, and public demonstration of power. 14 The explosive nature of the test closely mirrors our understanding of"going nuclear" as a revolutionary act in international politics. This is an important plus for the test/no-test indicator from the perspective of measurement validity. A second advantage of using the test/no-test indicator is that it greatly reduces the room for the politics and "spin" that unfortunately bedevil many measures of key concepts in international relations and even hamper supposedly "technical" assessments of the workability of a weapon's design or the quality of its manufacture. It is hard to argue with several kilotons of suddenly released energy. This is one important reason why the U.S. weapons labs, for instance, have been so reluctant to abandon "hot" testing in favor of computer simulationsY Of course, small or failed nuclear tests (for example, North Korea's two efforts to date) do pose problems for interpretation. There is no clear minimum explosive yield that a nuclear test must attain in order to be counted a technical success, but historically a yield of less than about six kilotons has tended to give rise to technical doubts. Therefore, for instance, after their 1974 test had a disappointingly small yield-perhaps in the two- to four-kiloton range-Indian nuclear scientists considered a new test prior to induction of weapons a "technical necessity." 16 Still, even a small nuclear test provides a clear signal that a state has at least arrived at the nuclear threshold. A third advantage of using the test/no-test indicator is quite simply the fact that this is the operationalization of nuclear weapons stateness that has been institutionalized in the foundational text of the nonproliferation regime, the NPT. Article IX of the NPT reads, "For the purposes of this Treaty, a nuclearweapon State is one which has manufactured and exploded a nuclear weapon or other nuclear explosive device prior to 1 January, 1967." To redefine nuclear weapons states as something other than states that have "manufactured and exploded" a nuclear device would mean overturning years of settled usage, inevitably leading to dangerous confusion about different states' rights and responsibilities under the treaty. Indeed, given the fundamental position of the non-nuclear versus nuclear weapons state distinction in the construction of the nonproliferation regime, a significant change in our understanding of that distinction implies the regime's complete overhaul. By way of comparison, Article IX also includes the 1967 cutoff date for "legal" nuclear weapons state status. Recent attempts to remove India's nuclear pariah status for testing later than 1967-even though it never signed the treaty-have been

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

107

strongly attacked by nonproliferation advocates as extremely destabilizing to the regimeY The nonproliferation community's concerns on this point are quite understandable; but it should therefore be even more concerned about the destabilizing consequences of abandoning the NPT's much more central "manufacture and explode" criteria for distinguishing between nuclear and non-nuclear weapons states. The foregoing has mentioned some considerable advantages to the test/ no-test metric. However, it must be pointed out that this metric is no more than an indicator of a state's potential nuclear weapons stateness, not the underlying quantity of interest itself. In an ideal world, nuclear weapons states would be identified by simply pointing to their actual operational nuclear weapons, as for instance birds are identified by their feathers. But the combination of the complexity of nuclear weapons systems and state secrecy often makes such a straightforward identification exercise impossible. Pointing to the test is the next best option; but it should not be forgotten that a nuclear test does not in itself equal a nuclear arsenal. It is actually merely a signal of intent and capacity, not the thing itself. For instance, while India's May 1998 nuclear tests did finally resolve the country's long debate over whether or not to "go nuclear," from an operational standpoint the country still had much work-about a year's worth-to do before it could credibly claim to own even what Ashley Tellis terms a nuclear "force-in-being"-"a deterrent whose effectiveness derives from its ability to be constituted into a viable retaliatory instrument under conditions of supreme emergency." 18 In other words, if we liken nuclear weapons to a fountain pen, India's 1998 tests proved that it had an adequate ink reservoir, but it still needed some time to add the nib that is essential for writing. In short, using the test as the measure of nuclear weapons stateness actually jumps the gun in most circumstances. But the evident imperfection of this indicator should not immediately lead us to condemn it as invalid. All operationalizations of key concepts in human affairs are imperfect. The proper standard for indicators in the social sciences is not perfection but adequacy. Furthermore, on a more practical level, given the destabilizing consequences of strategic surprise it can be argued that it makes sense to choose an indicator that alerts the members of international community to the reality of a new nuclear weapons state slightly in advance, while they still have some time for cool deliberation on how best to adjust their strategic postures in light of the emergent reality.

108

JACQUES E. C. HYMANS

A successful nuclear test clearly indicates nuclear weapons stateness because it mirrors our underlying concept, sends a relatively unambiguous signal of the state's nuclear attainment, and has for a long time underlain diplomatic praxis. But is the absence of a nuclear test a valid indicator of non-nuclear weapons stateness? If it is not, then as much as we might like to maintain the test/no-test indicator for the abovementioned reasons, in fact it is an invalid metric and needs to be replaced. Many have pointed out that there is an evident problem with using the notest criterion as the indicator of non-nuclear weapons stateness. The problem is that from a purely technical point of view a nuclear weapons test is not always strictly necessary for nuclear bombs to be built. For example, the "Little Boy" gun-type uranium bomb design had not been tested before the United States dropped it on Hiroshima. Even so, during the first decades of the nuclear age, few doubted that the test was the inevitable culmination of a nascent nuclear weapons program. All of the first five nuclear weapons states carried out explosive tests before inducting nuclear weapons into their military arsenals. Even in the U.S. case, its first nuclear blast was not over Hiroshima but instead near Alamogordo, New Mexico. And, despite the different design of the Alamogordo and Hiroshima devices, without the Alamogordo test U.S. scientists and policy makers would have been unconvinced that the bomb was indeed in their hands and that they could therefore order its use against Japan. 19 The consensus assumption that the birth of new nuclear weapons arsenals required prior nuclear testing was what led NPT Article IX to be written as it was. This consensus emerged not only because of the behavior of the first five nuclear weapons states but also because that behavior conformed to the tenets of deterrence theory as it had developed by the mid-196os. 20 As was memorably explained to the Soviet ambassador in the 1964 movie Dr. Strangelove, deterrence theory's basic insight was that, for the apocalypse to be avoided, its potential arrival must be heavily advertised. 21 The nuclear test is perfectly suited for that task. Indeed, it is the only sure means oflending an air of reality to the otherwise "fabulously textual" world of nuclear strategyY In short, from the perspective of mainstream deterrence theory, if you don't test, then you don't have the bomb-even if you do have it. So therefore you must test. During the Cold War, when it came to defining nuclear weapons states this strategic argument that testing was strictly necessary was generally seen to trump the technical argument that it was not strictly necessary. But many today are making a different calculation.

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

109

CRITIQUES OF THE TEST/NO-TEST INDICATOR

In recent years, the utility of using the first nuclear test as the key indicator of a state's nuclear weapons status has come under increasing attack. Although critics of the traditional thinking accept that a nuclear test reflects a state's entry into the weapons state club, they argue that the lack of a test does not mean that the state has not joined that club. In other words, in the phrase made famous by American neoconservatives, absence of evidence is not evidence of absence. Therefore, there is increasing sentiment that the real red line between nuclear and non-nuclear weapons states should actually be drawn at the acquisition of an SQ of fissile material. Why are new-school nonproliferation advocates so skeptical about the test/ no-test indicator's ability to capture the reality of proliferation? First of all, their point of view reflects the remarkable post-Cold War decline in the influence of mainstream deterrence theory. 23 As noted above, the United States knows from its own experience that some bomb designs may not require testing prior to induction; however, it started worrying that many other states would actually choose to go without testing only after it lost much of its faith in deterrence theory. New-school nonproliferation advocates are not at all convinced by the traditional notion that, because deterrence theory requires testing, testing will happen. Instead they pose a question: Wouldn't you rather be safe than sorry? Behind this deceptively simple question there stands an entire alternative strategic theory, which already some decades ago was elaborated by the neoconservatives' godfather, Albert Wohlstetter. On the definition of nuclear weapons stateness, Wohlstetter (and his coauthors) had this to say: If, in fact, technological transfers can bring a "nonnuclear weapon state" within weeks, days or even hours of the ability to use a nuclear explosive, in the operational sense that "nonnuclear weapon state" will have nuclear weapons. The point is even more fundamental than the fact that effective safeguards mean timely warning. A necessary condition for timely warning is that there be a substantial elapsed time. But if there is no substantial elapsed time before a government may use nuclear weapons, in effect it has them. 24 Wohlstetter and his coauthors continued: "Consider the case of a government which is not at war, but is capable of quickly assembling a nuclear device to use or threaten to use against another government without such a capability. Once again, there is no practical difference between the coercion it could use or the threat it could execute from what a nuclear power might manage." 25

110

JACQUES E. C. HYMANS

In short, whereas the traditional deterrence theory view was that if you might not have the bomb then you don't have it, the Wohlstetter view was that if you might have it then you do have it. Thus, he saw the necessity of shifting our indicator of nuclear weapons status from test/no-test to SQ/no-SQ. Wohlstetter's perspective has been updated and applied to the question of the reform of the NPT regime in a recent volume edited by the well-known nonproliferation advocate Henry Sokolski. Sokolski and his colleagues explicitly call for fundamentally reinterpreting the NPT not merely as a ban on non-nuclear weapons state acquisition of a nuclear weapons arsenal but in fact as a ban on their acquisition of an SQ or indeed even of facilities (such as uranium enrichment or fuel reprocessing plants) that can readily produce an SQ in a short period of time. 26 The argument runs as follows. Safeguards or no safeguards, timely warning of conversion of an SQ into operational weapons is impossible; therefore, outsiders cannot help but treat a state that has an SQ as if it already had an operational weapons arsenal. Furthermore, because under the NPT non-nuclear weapons states have taken on the obligation not to acquire nuclear weapons, and because in terms of its strategic implications an SQ is actually equivalent to a nuclear weapon, therefore non-nuclear weapons states are also obligated by the treaty not to acquire an SQ. In short, the decision to seek to acquire an SQ is tantamount to a decision to "go nuclear." New-school nonproliferation advocates also base their case on an empirical perception that some, and perhaps even many, states actually do prefer to forego testing before inducting nuclear weapons into their arsenals. For instance, Israel has been armed with operational nuclear weapons probably since around 1973, and some say since as early as 1967. 27 Indeed, most expert observers believe that Israel today has not merely "the bomb" but indeed a large and technically sophisticated nuclear arsenal mounted on multiple delivery systems. Yet in contrast to all of the other nuclear weapons states, Israel has apparently never carried out an explosive nuclear test. 28 Moreover, it refuses even to officially acknowledge the existence of its arsenal, although there have been many "slips of the tongue" over the years. 29 The Israeli case therefore appears to represent a major exception to the ideal-typical nuclear path that was followed by the first five nuclear weapons states. 30 Note that the problem posed by the Israeli case for traditional understandings of the natural course of proliferation is not that it developed the bomb in secret but rather that, having built the bomb, it maintained the secret. The mere fact of the Israeli exception is not in itself damning to the test/ no-test indicator. After all, no indicator is 100 percent error free. Therefore, if

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

111

one had good reason to believe that Israel's no-test path to the bomb was sui generis, one of a kind and unlikely to be repeated, then the Israeli exception would be just that, an exception to the rule. And indeed, for many years proliferation experts quite happily retained the test/no-test indicator in general, while simply appending an asterisk to the Israeli case. However, if the measurement error introduced by an indicator is not merely minor and random but large and systematic-in other words, if the indicator is seriously biased-then the basic validity of the indicator does come into question. 31 And, especially since the early 1990s, scholars and policy makers have developed the argument that the Israeli case is indeed not sui generis but is rather at the vanguard of a worldwide trend away from testing as the key step across the threshold to nuclear weapons acquisition. In their seminal work, Avner Cohen and Benjamin Frankel characterized the Israeli case of "opaque proliferation" as ideal-typical of a burgeoning second generation of new nuclear weapons states, which was being sorely underestimated by the traditional test/no-test indicator. 32 Cohen and Frankel also pointed to Pakistan and South Africa as states that had built "opaque" nuclear arsenals. Over the past two decades the Cohen-Frankel "opaque proliferation" model has become increasingly accepted in the policy world (though less so among academics) as a true picture of contemporary reality. Thus we are often informed that we are now living in a "second nuclear age" of rampant, but hidden nuclear proliferation. 33 And the response by policy makers to this perceived problem, as noted in the introduction to this chapter, has been to try to move back the red line from the first test to the production of an SQ of fissile material-the last chance, they believe, to catch the proliferant state red handed before it becomes capable of a strategic surprise. The next section evaluates this SQ/no-SQ alternative to the traditional test/no-test indicator of nuclear weapons stateness. CRITIQUES OF THE SQ/NO-SQ INDICATOR

As noted above, the emerging operationalization of the concept of nuclear weapons stateness focuses on the acquisition of a "significant quantity" (SQ) of fissile material, which some see as a more valid indicator of de facto nuclear weapons state status than the traditional test/no-test indicator. Does this argument compute? Should we choose definitively to abandon the test/no-test indicator in favor of the SQ/no-SQ alternative? In fact, there are several serious drawbacks to this idea. First, it is questionable to assert that the Israeli posture of nuclear opacity is appealing to the new generation of nuclear weapons states.

112

JACQUES E. C. HYMANS

Second, because the accumulation of an SQ is much less visible than a nuclear test, using that metric to assess nuclear weapons status allows a much wider range of potential interpretations and even opens the door to politically biased misinterpretation. Third, the SQ/no-SQ indicator considerably underestimates the obstacles, both technical and political, that stand between the accumulation of fissile material and the production of operational bombs. And last but not least, shifting to the SQ/no-SQ indicator may have a significant impact on real-world behavior and quite possibly could have the effect of encouraging more nuclear proliferation. The chapter fleshes out these points one at a time. Israel as a Sui Generis Proliferator

First, as noted above, Israel has been showcased as the ideal-typical case of "opaque proliferation"; but is the Israeli path really ideal-typical of a larger class of cases, or is it just an anomalous exception? One way of beginning to answer this question is to ask whether Israel's stance of nuclear opacity has proven to be a smart choice for it, one that others might therefore be tempted to follow. It is certainly possible to argue that the undeclared Israeli bomb has proven key to the country's security. 34 But Israel's consignment to the realm of nuclear opacity has also been a difficult experience in many ways. Opacity has led to unclear nuclear signaling externally, a stunted nuclear doctrinal debate internally, regular embarrassment at international diplomatic forums, and even the cancerous growth of a secret state that undermines Israeli democracy.35 Indeed Israel's leaders often appear, at least at some level, to want to put an end to the ambiguity-as evidenced by their numerous "slips of the tongue" about the fact of the Israeli bomb over the years. 36 It is hard to imagine that many other states would be eager to join Israel in

this dubious club. 37 And indeed most of the other states that were once viewed as part of it-notably India, North Korea, Pakistan, and South Africa-clearly are not in it today. 38 Even Avner Cohen, the coauthor of the seminal work on "opaque proliferation," now himself argues that Israel is a "sui generis proliferator," whose choice to rely for decades on an untested nuclear arsenal is historically unique. 39 SQ or No SQ? The Difficulty of Assessment

A second problem with the SQ indicator lies in the question of how we can know that an SQ has actually been achieved. For instance, as was noted at the outset of this article, the U.S. intelligence community believed that North Korea had at least one plutonium-based bomb by the early 1990s. Moreover,

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

113

in 2002 the United States publicly alleged that Pyongyang had also developed a uranium enrichment capability with the help of the Pakistani scientist A. Q. Khan. But it now appears that the country did not have a secret SQ of plutonium, let alone nuclear weapons, in the 1990s. As part of the Six-Party Talks process, in May 2008 North Korea turned over about 18,ooo pages of technical documents to the United States and other negotiating partners. According to reports, these documents-which would be very difficult to fabricatereveal that, contrary to U.S. estimates, North Korea had removed no plutonium from its Yongbyon reactor prior to 1992. In other words, the experts now believe that the assumption that North Korea even had an SQ in the early 1990s was simply wrong. 40 Moreover, the equally widespread notion that, thanks to A. Q. Khan, North Korea may have been on the brink of acquiring a considerable stockpile of HEU at the start of the 2ooos also now appears quite suspect. The U.S. intelligence community's uranium enrichment assessment turns out to have been based largely on North Korea's importation of the same type of aluminum tubes that led to wildly inaccurate estimates of the progress of Iraq's nuclear program in the early 2ooos. 41 Of late, the United States has been pretty quiet on the uranium enrichment issue. The retreat of the analysts from their prior strong statements about the advancement of the North Korean nuclear program is not just a proper correction in light of new evidence. Rather, it shows a deeper flaw in the reliance on the SQ/no-SQ indicator as the measure of nuclear weapons stateness. The validity of any indicator depends heavily on the degree of measurement error that it introduces. Yet the application of the SQ/no-SQ indicator requires a great deal of guesswork, even if the state's nuclear facilities are known and easily observable via satellite. If there are suspicions of hidden facilities, the ratio of guesswork to known facts becomes much more lopsided-and this ratio increases yet further if we entertain the possibility that a state might have acquired fissile material from smuggling rings. Therefore, the SQ/no-SQ indicator seems to open the door widely to all manner of interpretations of the nuclear progress of particular cases. It therefore also opens the door to political abuse of the sort that characterized the reports on Iraq's alleged WMD programs in the run-up to the 2003 Iraq war. 42 From an SQ to the Bomb: A Difficult Road

An additional serious drawback to the SQ/no-SQ indicator is the assumption of easy conversion to bombs on which it rests. For instance, to quote the U.S. official Laura Holgate, "Once you have the fissile material, it's a matter of basic

114

JACQUES E. C. HYMANS

chemistry, basic machinery, and a truck."43 This statement grossly underestimates the degree of technical difficulty of putting all the pieces of the puzzle together, and perhaps even more importantly it also ignores the political and organizational difficulties that can hamper and delay the shift from bomb program to weapons arsenal. 44 For these reasons, even if we are sure that an SQ has been achieved and that the country's leadership intends to go all the way, the SQ/no-SQ indicator is still prone to exaggerate the number of bona fide nuclear weapons states. The mistake of assuming that the road from SQ to a bomb is a short and easy one is again highlighted by the misbegotten U.S. experience with North Korea. Former Assistant Secretary of State for East Asian and Pacific Affairs James Kelly, who had earlier happily buried the U.S.-DPRK Agreed Framework, told the press shortly after leaving office in 2005 that because it was clear that North Korea already had the bomb, "I personally have never been worried about a test. A test would just tell us what we already know." 45 Yet, as noted above, when the first North Korean nuclear test actually happened in October 2006 it was quite underwhelming. The blast measured somewhere between 0.2 and 1 kiloton in yield-making it by far the least powerful first nuclear test in history. Thus, far from confirming the DPRK as a nuclear weapons state, the test actually showed the opposite to be true. As Jungmin Kang and Peter Hayes of the Nautilus Institute put it after reviewing the technical evidence, "We conclude that the DPRK test did not enable it to pole-vault into the ranks of nuclear weapon states .... The DPRK inhabits a peculiar and ambiguous status between having declared itself to be nuclear-armed, and having demonstrated by its test that it is not capable of such armament." 46 North Korea's second nuclear test, in 2009, appears to have been somewhat more successful, but doubts about it persist as well. The most the U.S. intelligence community could offer on the 2009 episode was that North Korea had "probably" conducted a nuclear explosionY Moreover, even if one judges that the second North Korean test did meet the bare minimum standard for "success," there remains a very open question as to whether Pyongyang has the ability to take the next step and produce reliable, deliverable nuclear warheads. Real-World Effects of the Shift to SQ/No-SQ

Finally, we must consider not merely the validity of shifting to the SQ/no-SQ indicator on conceptual and empirical grounds but also what consequences such a shift might have in the real world. After all, unlike the natural sciences,

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

115

in the social sciences we must be mindful that our theories, concepts, and measurement strategies may not merely describe but also influence the behavior of the subjects under study. 48 Indeed, one of the main motivations for those who are attempting to install the SQ/no-SQ indicator is precisely the hope that the widespread use of such a metric would change state behavior. They reason that, if states can be convinced that the choice to make fissile material actually itself represents the choice to "go nuclear," then many fewer states will in fact end up going even that far down the nuclear path; meanwhile, others' true intent will be unambiguously identified sooner. 49 Such a scenario would of course be a boon for international security. But, although this is a plausible story, other stories are plausible, too. For instance, the shift to SQ/no-SQ represents a significant "dumbing down" of the barriers to entry into the nuclear "club." Why should this not encourage more states to try to get in? Indeed, why should it not lead even frankly incompetent states like Zimbabwe or Myanmar to imagine that with just a little nuclear contraband they, too, would be able to claim nuclear weapons state status? Moreover, might not states that already have an SQ, such as Japan, simply accept the world's new judgment of their nuclear weapons stateness, despite never having made that decision for themselves? After all, the most important factor encouraging nonproliferation up to the present day has been state leaders' perception that to "go nuclear" is a revolutionary choice, with potentially vast and unknowable consequences for their states' security and power positions. 50 But if the international community pushes the definition of nuclear weapons stateness backward to SQ/no-SQ, it will essentially have relieved some states' leaders from the burden of facing that revolutionary choice. Furthermore, even the mere effort to replace test/no-test with SQ/no-SQ could be harmful to overall international systemic stability. For although the lines we draw between two qualitatively different states of being may be social constructions, it does not logically follow that these lines can be drawn arbitrarily or unilaterally. 51 The test/no-test indicator became the red line between non-nuclear and nuclear weapons state status because, in light of our background understandings of the nuclear issue, it was the obvious focal pointand it still is. 52 In politics, and particularly in international politics where the dangers of miscommunication are so rampant, the wise course is usually to keep our concepts and indicators as closely as possible in accordance with ordinary usage. Attempts to change by fiat the common web of understandings

116 JACQUES E. C. HYMANS

that lie at the very foundations of international order are likely to produce confusion or even chaos. 53 In short, the SQ/no-SQ indicator for nuclear weapons stateness is not only analytically flawed; it is also politically dangerous. None of the foregoing should be taken to imply that efforts to enhance the tracking or to control the growth of fissile material stockpiles are meaningless or self-defeating. Rather, such efforts may well be an important means of promoting nonproliferation. 54 But still, fissile material is not "the bomb," and to equate the two is a fundamental error in terms of both social science and public policy. FROM WHETHER TEST OR NO TEST TO WHY TEST OR NO TEST?

The previous section demonstrated the many problems from the perspective of measurement validity of moving wholesale from test/no-test to SQ/no-SQ as our preferred indicator of nuclear weapons stateness. In light of these problems, it makes much more sense to retain the traditional test/no-test indicator, which has actually not served us all that badly up to now. However, the fact that the simple test/no-test indicator is superior to the usual proposed alternative does not mean that we should be completely happy with it. As previously noted, new-school nonproliferation advocates are certainly right to note that the simple test/no-test indicator is not infallible, and they could possibly be right that the indicator will prove more fallible in the future. So, can we do better? In their important work on measurement validity, Adcock and Collier propose a method of"nomological" or "construct" validation, which they also catchily term the "AHEM" method of validation-standing for "Assume the Hypothesis, Evaluate the Measure." 55 In other words, Adcock and Collier suggest that when we are confronted with the task of measuring something that we cannot see directly, one possible response is to look for the presence or absence of the established causes of that thing. Where those causes are seen to be present, we can make at least a cautious claim that their typical effect should also be present. An AHEM approach to measurement validation in this context leads us to ask not just whether a state has arrived at an SQ or has conducted a test but also why states test, or not, at the outset of their careers as nuclear weapons states. In other words, although the test/no-test indicator has historically done pretty well as a measure of nuclear weapons stateness, the truth is that we do not know why. As noted previously, the traditional confidence in the indicator

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

117

was based on the assumption that states should instinctively understand and comply with the requirements of deterrence theory. But in light of even U.S. and Soviet nuclear behavior during the Cold War, it is hard to claim that states really act as traditional deterrence theory imagines they do. 56 The test/no-test indicator has thus actually fared better empirically than it should have, considering its heavy debt to this rather dubious theory of state behavior. This lack of proper theoretical grounding was what opened the door for advocates of the SQ/no-SQ alternative in the first place. A more theoretically grounded AHEM approach to operationalizing the concept of nuclear weapons stateness would also improve on the current test/ no-test indicator by allowing us to report our level of confidence in the validity of that indicator, both globally and on a case-by-case basis. After all, valid measurement means not merely making careful descriptive inferences but also carefully estimating the degree of uncertainty attached to those inferences. 57 This is always a difficult task, but it is much more difficult if we have not first identified the main factors that drive countries to choose the route of testing or not testing. Paradoxically, by delineating the key potential problem cases for the test/no-test indicator, the AHEM approach should actually strengthen our faith in the overall value of that indicator. A full elaboration of the AHEM approach would require a book-length effort. But to provide a flavor of the approach, I propose the following hypothesis as a possible starting point: Civilian-run nuclear weapons programs are highly unlikely to induct nuclear weapons without a test, but military-run nuclear weapons programs may sometimes choose to do so. To understand the thinking behind this hypothesis, first note that the decision to induct nuclear weapons without prior testing most likely reflects a highly offensive strategic orientation. This is because, by not testing, the state is maximizing its potential for strategic surprise. Strategic surprise can be had even if the nuclear weapons program is not a complete secret because without a test the adversary will be unaware of the true advancement of the arsenal, and it may even be lulled into complacency by official denials of nuclear weapons possession. In short, induction without prior testing is not a sign of restraint but rather quite the opposite. The state may be readying its unannounced nuclear arsenal for a "splendid first strike," or it may have other plans, such as to use a stunning test of operational, battlefield nuclear weapons to put the enemy on warning, as South Africa was ready to do during the 1980s. 58 But even the latter would clearly be an extremely aggressive maneuver.

118 JACQUES E. C. HYMANS

If inducting without testing reflects a highly offensive orientation, then who in the state is likely to support such a policy? It is widely claimed in the international relations literature that professional military organizations, in contrast to most civilian politicians and agencies, have a decided bias in favor of offensive strategies. Various reasons have been adduced for this preference: Military organizations perform better when they can follow their own script; militaries embrace offensive strategies as a means of gaining larger budgets and operational autonomy; military professionals tend to be fixated on, and therefore to overestimate, the likelihood of war; military culture tends to glorify offensive daring; militaries are accustomed to maintaining the secrecy required for launching first strikes; and so forth. 59 In short, many scholars concur that the military preference for the offensive is a strong, enduring, global tendency. (Beyond professional militaries, the most likely candidates might include paramilitary or secret police organizations.) As evidence for this tendency in the nuclear area, despite numerous claims by analysts that nuclear weapons are the defensive weapon par excellence, various states' militaries have demonstrated a strong bias for exploiting nuclear weapons in strategies of compellence as well as deterrence, and many have also shown a clear inclination toward nuclear war-fighting doctrines. 60 Because militaries tend to prefer offensive strategies, and because induction without testing likely reflects a highly offensive strategy, the natural conclusion is that militaries should be more likely than civilians to do without testing. However, "more likely" here should not be taken to mean "likely." Even offensive-minded militaries may still prefer first to garner the presumed deterrence benefits of the bomb by carrying out a successful test and only then turn their minds to thinking about possible offensive uses of the weapon. 61 The military's general preference for offense can hardly be seen as automatically indicating a desire to bypass testing in order to create an option for strategic nuclear surprise. Moreover, and more prosaically, militaries also have a strong desire to know that the weapons in their arsenal will actually work. This need is especially acute if nuclear weapons are indeed slated to be employed offensively. After all, if the plan is to carry out a splendid first strike, that first strike had better be truly splendid. This desire for confidence in their nuclear arsenal also militates in favor of testing. One way for the military to satisfy its desire for reliability while maintaining the option of strategic surprise could be to carry out a secret test. This

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

119

was the option pursued by the United States and apparently also by the Soviet Union, although the Soviet blast was quickly discovered by outside intelligence.62 Such an option was still barely possible in 1979, when South Africa may have carried out a nuclear test in the South Atlantic. 63 But due to the increase in quantity and quality of remote-sensing equipment of various kinds (especially the largely deployed International Monitoring System for verifying compliance with the Comprehensive Test Ban Treaty), a secret test is clearly not an option anymore. So if the military wants to go for nuclear opacity today, it has to do without a real test, which considerably raises the costs of pursuing such an offensive nuclear strategy. I have argued that, unlike most civilian politicians and agencies, militaries may be at least tempted to induct nuclear weapons without prior testing. The question now becomes, Under what circumstances are those military preferences more likely to be realized? This will of course depend largely upon the specific institutional balance of power inside the state. However, in general it may be surmised that the bureaucracy that controls the nuclear weapons program-that is, either directly runs the program or is directly and uniquely reported to by those who do run the program-is likely to have a disproportionate say over the question of whether or not to test prior to induction. This is because that bureaucracy would be in a position to monopolize the information flow to the other parts of the state, and thus to twist the information toward or away from the testing option as it saw fit. Thus, we arrive at the hypothesis stated above, that while civilian-run nuclear weapons programs are highly unlikely to induct nuclear weapons without a test, military-run nuclear weapons programs may sometimes choose to do so. A brief study of the historical record appears to provide tentative confirmation of the power of this hypothesis. Before turning to that historical record, however, it must be emphasized once again that most nuclear weapons programs have never gotten to the stage where they were faced with the decision to test or to induct without testing. For instance, North Korea's failure to test nuclear weapons in the early 1990s was not a sign of a strategic choice for nuclear opacity but rather simply a reflection of the program's lack of sufficient technical advancement. Since 1945, ten states are generally believed to have crossed the line and either tested a nuclear fission device or inducted untested but operational nuclear weapons into their arsenals. Table 6.1 shows that the hypothesized distinction between civilian- and military-run programs does seem to matter

120 JACQUES E. C. HYMANS

Table 6.1.

Nuclear testing choices of military- versus civilian-run nuclear weapons programs. Civilian-run

Military-run

Test, then induction

China, North Korea

Secret test, then induction

United States, possibly South Africa

Soviet Union

No test prior to induction

Israel, Pakistan

n/a

United Kingdom, France, India

NoTE: See the appendix to this chapter for explanation of Table 6.1 codings.

in determining how those ten states decided to cross the line: with a test, secret test, or no test. As Table 6.1 clearly shows, most states have chosen to test in advance of acquiring operational nuclear weapons arsenals. Moreover, all programs run by civilian agencies have chosen to test before inducting nuclear weapons, although, notably, the Soviet nuclear program, run by the KGB secret police organization, chose not to immediately announce the test's success. Militaryrun programs have shown a somewhat greater willingness to do without testing, but even in that category at least three out of six tested prior to induction, four out of six if South Africa indeed conducted a test in September 1979. 64 (Note that although South Africa's program was initially civilian run, by September 1979 control of the program had passed to the military parastatal Armaments Corporation of South Africa. 65 ) A more thorough empirical study might suggest different codings for this table-for example, the Soviet Union may have been planning to announce its test itself if the United States failed to do so; Israel may have participated to some degree in South Africa's supposed test; Pakistan may have carried out an earlier test in China; or North Korea's program may still be in civilian hands-but most of these possible coding changes would not substantially change the basic conclusion. 66 What Table 6.1 does not show is that, of the two states that certainly did not test prior to induction, only Israel definitely committed itself to developing a sizeable arsenal without testing. By contrast, whatever genuine operational capability, if any, Islamabad had during its brief period of nuclear opacity was almost certainly very limited. 67 Again, the Israeli case appears to be sui generis. It might also be noted that, in both the Israeli and Pakistani cases, U.S. intelligence actually learned about their induction of nuclear weapons very shortly after it had occurred. 68 Therefore the United States, at least, was not vulnerable to strategic surprise in either case.

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

121

The tentative takeaway from this empirical exercise is that it is best to retain the test/no-test indicator as our basic metric for nuclear weapons stateness unless there is direct and incontrovertible evidence that successful induction without testing has taken place in a given case. But it also makes sense to append an asterisk to the non-nuclear weapons state status ascribed to militaryrun nuclear weapons programs that have built up an SQ of fissile material. "Appending an asterisk" means retaining the test/no-test indicator but indicating less than complete confidence in this judgment. The precise level of confidence for a given case could then be fine-tuned through in-depth qualitative analysis of the strategic culture of the military running the program and in particular of the degree to which it has generally been prone to exploit opportunities for strategic surprise. Qualitative analysis would also be important to study the overall quality of management of the nuclear weapons program, in order to assess the state's ability to move from an SQ to an operational arsenal that, despite remaining untested, could be counted on to function as planned. On this point it bears repeating that the technical bar for inducting operational nuclear weapons is much higher than the bar for testing a simple nuclear fission device. In short, the test/no-test indicator is always preferable to the SQ/no-SQ indicator as a first-cut approximation of the state of play, even for militaryrun nuclear weapons programs; but a better-theorized test/no-test indicator is much more preferable. The key lesson here is less the specific hypothesis of civilian- versus militaryrun programs and more the overall approach to this knotty problem of measurement validity. The AHEM, or what one might term "neotraditional," approach to measuring nuclear weapons stateness provides us with a way forward beyond the test/no-test versus SQ/no-SQ impasse. The approach is "traditional" because it retains the focus on the testing stage as the key threshold point. But the approach is also "neo" because it admits that the traditional barefoot empiricist indicator of test/no-test is not enough and indeed that it may even contain a systematic bias toward false negatives in a small subset of cases. The neotraditional approach to the measurement of nuclear weapons stateness will not be sufficient in itself to produce a consensus coding of the nuclear weapons status of a country such as Iran. But it can help to set the parameters for a more analytically sound debate on Iran or other states' nuclear evolution, and also-even especially-it should reassure us about the continuing overall health of the nonproliferation norm.

122

JACQUES E. C. HYMANS

APPENDIX: EXPLANATION OF TABLE 6.1 CODINGS

United States: Program run by U.S. military with civilian input. First test in July 1945, prior to induction, although note that the Hiroshima attack employed a different, untested bomb design. Test carried out in total secrecy and results not announced until after Hiroshima. 69 Soviet Union: Program run by KGB, not military. First test in 1949, prior to induction. Test secret but above ground and easily detected by U.S. intelligence. 70 United Kingdom: Program run by civilian atomic energy authority with military input. First test in 1952, prior to induction. Test results announced. 71 France: Program run by civilian atomic energy authority with military input. First test in 1960, prior to induction. Test results announced. 72 China: Program run by the military. First test in 1964, prior to induction. Test results announced.7 3 Israel: Program run by people inside the General Staff of the Defense Forces, reporting to civilian leadership in the Ministry of Defense. No test prior to induction, although possibly participated in the supposed South African test. 74 India: Program run by civilian atomic energy authority with some military input. First test in 1974, prior to induction. Test results announced. Second test in 1998, and also probably prior to induction, although by that point the country may have had a very crude operational capacity. These test results also announced. 75 South Africa: Program run by civilian atomic energy authority until1979, when military parastatal corporation took over. Possible test in 1979, prior to induction. Secrecy maintained until weapons dismantled in early 1990S.76 Pakistan: Program run by the Pakistani military after 1976. According to U.S. intelligence, probable induction in 1990, prior to testing. First test in 1998, results announced. 77 North Korea: Program run by civilians reporting directly to the top politicalleadership at least until early 1990s?8 Since then, great uncertainty, but given Kim Jong Il's overall "military-first" policy it is unlikely that the military would have been excluded from the project, particularly after

WHEN DOES A STATE BECOME A "NUCLEAR WEAPONS STATE"?

2002. First test in 2006 with paltry results, though regime publicly announced it as a success. Second apparent test in 2009, with slightly better results, again announced. Induction almost certainly not before 2oo6, and indeed serious questions remain about whether induction could even take place at this point.

123

7

THE LITTLE-KNOWN STORY OF DEPROLIFERATION Why States Give Up Nuclear Weapons Activities

Harald Mi.iller and Andreas Schmidt

IN THIS CHAPTER, we

tackle one of the major puzzles of international politics in the last fifty years: the failure of nuclear proliferation to assume a pandemiclike dynamic. Realist theory and the discourse informed by it see proliferation as inevitable in an anarchic world where states have to fend for themselves and cannot take for granted alliances that may temporarily relieve their security problem, as the politics of the allies may change and umbrellas may thus be closed. The curve of proliferation appears to confirm this prediction, though at a much lower speed than one could have expected (Figure 7.1). A very different picture emerges if we do not simply look at the number of countries that built nuclear weapons but focus on the entire set of countries that have conducted nuclear weapons activities since 1945, including those that reversed course. Taking into account the number of such "reversals" is essential to understand the dynamics of proliferation and nonproliferation. We adopt, for our purposes, the definition of"reversal" proposed by Ariel Levite: ... the phenomenon in which states embark on a path leading to nuclear weapons acquisition but subsequently reverse course, though not necessarily abandoning altogether their nuclear ambitions .... At the core of this definition is the distinction between states that have launched (indigenously or with external assistance) a nuclear weapons program and then abandoned it and those that never had such a program in the first place. 1 As applied here, this definition does include cases in which "a governmental decision to acquire the bomb could not be ascertained." 2 This definition takes into account the crucial step over the first important policy threshold, to con124

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

en Q)

125

10

a;

Ui en c 8 0

c.

ro Q) !: .... ro 6 Q)

u:::J c

0.... 4 Q)

..c E :::J

z

2

o,_~r--r--~-,---r--~~--.---r-~--~--r--r~

1945

1955 1950

Figure 7.1.

1965 1960

1975 1970

1985 1980

1990

2005 1995 2015 2000 2010

The commonsense view of the proliferation process.

sider seriously the acquisition of nuclear weapons, and it enables one to review a larger number of cases than a more restricted definition would permit (Figure 7.2). By this definition, we have identified thirty-seven states with "nuclear weapons activities" since the beginning of the nuclear age (see the appendix at the end of this chapter). As Figure 7·3 shows, the number of states that started nuclear weapons activities but reversed course is more than double the number of those who still conduct them. At the end of our observation period, 2005, the latter reached the lowest score since the early 1950s. The process went as follows: Addition of new states with nuclear weapons activities stagnated in the mid-1970s and came to a virtual halt in the early 1990s, where the only "new ones" were innocent "inheritors"-Belarus, Kazakhstan, and Ukraine. Renunciation grew in the late 1960s, stagnated in the early 1980s, and stampeded starting in the mid-198os. We come back to the interpretation of these data later in the chapter. The fact that the ten states that we believe have nuclear weapons activities today3 represent less than 6 percent of the U.N. membership points to

126

MULLER AND SCHMIDT

~

z

£ 20 ·;: (/)

I 0

....

Q)

..c E 10 :::J c

~

~

04---~~--~--~--~--r-~---r--~--~~---;

1951-55 1961-65 1971-75 1981-85 1991-95 2001-2005 1945 1946-50 1956-60 1966-70 1976-80 1986-90 1996-2000 Period Figure 7.2.

States with current nuclear weapons activities (NWA), 1945 to 2005.

the unexpected low outcome of the proliferation process. States with nuclear weapons activities were always a minority, and today they are the smallest minority since 1945 (see Figures 7-4 and 7.5). This finding does not correspond to the frequently expressed view that nuclear weapons proliferation is getting progressively worse. Indeed, it is a situation that should be at the forefront of proliferation and nonproliferation research: Far more countries have "reversed" their nuclear pursuits than followed them through. Among the reversals, there are states whose economic and technological capabilities to build nuclear weapons were beyond doubt and those for whom the lack of resources and expertise created considerable, but not insuperable, hurdles on the way to "success." In this chapter, we try to provide an explanation of the overall phenomenon without claiming that the statistics we present obviate the need for indepth case studies. Indeed, in collecting the data we relied on country studies, and each case shows idiosyncrasies that necessitate a thorough process tracing

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

127

60

40

20

04-~~~~~~~--~~--~~--~~~

1951-55 1961-65 1971 - 75 1981 - 85 1991-95 2001-2005 1945 1946-50 1956-60 1966-70 1976-80 1986-90 1996-2000 Period

--Cum. terminations as % of cum. NWA

-Cum. no. of states that terminated NWA

-

cum. no. of states with NWA

Cumulative number of states with nuclear weapons activities relative to cumulative number of states that terminated such activities, 1945 to 2005.

Figure 7.3.

to uncover the particular causal path that led a given country to explore or pursue the nuclear weapons option and later to reverse course. We start by briefly summarizing the state of the art and then discuss in more detail two widely cited quantitative and four qualitative studies of proliferation/ nonproliferation that have helped us develop our own approach. We then test a series of well-established explanations for proliferation and reversals, using simple statistical methods. We conclude that theories of technological or economic pull are incorrect, that theories that claim alliance guarantees are the strongest cause for reversal need much more sophistication, and that economic liberalization is not a comprehensive explanation for the process of reversal. We find the most significant variables to be normative change-that is, the

128 MULLER AND SCHMIDT

200

100

0 ~~~-------.--.--.-.--.-~--~ 1945 1951-55 1961-65 1971-75 1981- 85 1991-95 2001-2005 1946-50 1956-60 1966-70 1976-80 1986-90 1996-2000 Period

-

Total no. of states with NWA

Figure 7.4.

-

No. of states at end of period

The proliferation process relative to the world of states, 1945 to 2005.

coming into force of the NPT-and democratization processes. The combination of the existence of the NPT and democracy/democratization is a sufficient cause for nuclear abstinence or reversal. However, we note a ratchet effect, as this combination does not reverse nuclear weapons programs that were well entrenched bureaucratically and politically before the NPT came into being. WHY STATES PURSUE, FOREGO, AND ABANDON NUCLEAR WEAPONS PROGRAMS: STATE OF THE ART

Most case studies exploring the motives for proliferation date back to the 1970s, 1980s, and early 1990s.4 This set is sufficient, however, as no new nuclear weapons programs have been launched since 1985 (with the possible exception of Syria). The findings of individual case studies and comparative research on the motives of aspiring nuclear weapons states during those years are still relevant

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

129

20

10

0;---,---.--,---.---.--.---.---.--.,--.---.--~

1945 1951-55 1961-65 1971-75 1981-85 1991-95 2001-2005 1946-50 1956-60 1966-70 1976-80 1986-90 1996-2000 Period

-

NWA as a percentage of all states

Figure 7.5.

-

Total no. of states with NWA

States with nuclear weapons activities as a percentage of all states, 1945

to 2005 . today. Various authors hypothesized-and this has been largely repeated in more recent research-that security, status/prestige, domestic factors (such as bureaucratic politics), and "technological pull" are the dominant independent variables pushing states down the proliferation road. 5 While the temptation to generalize has always been considerable, warnings to pay attention to the specifics of an individual case are well taken. 6 Much less work has been done to understand why states have not pursued the nuclear path or why the majority of those who have eventually reversed gear. A shortcut to answering this question would be the proposition that nonproliferation is the opposite of proliferation and that the decision not to pursue nuclear weapons rests on the absence of the four groups of factors that motivate proliferation in the cases where it occurs. If security problems do not

130

MULLER AND SCHMIDT

exist for a state or can be solved by nonnuclear means (for example, an alliance), when status and prestige are irrelevant either because the state lacks the necessary ambition for high status or finds other ways to satisfy this desire (for example, moral high ground, economic success), when domestic interests in a nuclear program are absent and the technological basis is lacking, then the state would not be compelled to go nuclear. There are problems with this argument, however. In an anarchic world, it is hard to see (at first glance) why states should shun the option to develop the most powerful weapon of their time, just in case. Whether substitutes for nuclear weapons in terms of prestige and status are available depends, of course, on what is considered a status/prestige symbol not only by the state in question but also by its peers. These problems have motivated some researchers to study nuclear reversal in a more systematic way. Mitchell Reiss has been a pioneer in this respect? A considerable number of single case studies and comparative works have also enhanced our knowledge on this issue. 8 Risks of Quantitative Research

In the field of nonproliferation, quantitative research is largely lacking. However, two recent studies on the "correlates" or "determinants" of proliferation are relevant. 9 Singh and Way and Jo and Gartzke both employ multivariate statistical analysis with a view to identifying the independent variables that drive nuclear proliferation. 10 Both use year-by-year accounts of proliferation programs. Singh and Way adopt a multistage model of proliferation that ranges from "exploring option" to "nuclear weapons program" to "possession of nuclear weapons." Stages 1 and 2 are characterized by "additional further steps aimed at acquiring nuclear weapons, such as a political decision by cabinet-level officials, movement towards weaponization or development of single-use dedicated technology." 11 Jo and Gartzke use crisper definitions for the start and end of a nuclear weapons program: We regard the year in which the highest decision maker in a given state authorized a nuclear weapons program as the year in which the state first possesses a nuclear weapons program. Similarly, we assume that the year in which the highest decision maker ended an existing nuclear weapons program is effectively the final year of the program. 12 For countries with insufficient information available to use that definition, they substitute "the year in which a suspect state's nuclear activities increase

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

131

noticeably as the start of their weapons program," where the phrase nuclear activities means "nuclear reactor construction or purchase, uranium milling or enrichment plant construction, and plutonium reprocessing facility construction and exclude small nuclear research reactor construction or purchase and simple nuclear research." 13 We regard these design decisions to be inappropriate in light of data hazards.14 Jo and Gartzke recognize, in principle, the data deficiencies but underestimate the consequences tremendously. As Levite has pointed out, the problems involved are multifold. First, due to the nature of nuclear activities, secrecy is all encompassing. Second, nuclear weapons activities do not necessarily start at the top level of leadership but can be initiated at the level of military commanders or nuclear experts and bureaucrats. Third, top leadership behavior toward these activities is often more permissive than determining. Fourth, significant activities may be directed at the initiation of a nuclear weapons option rather than determined pursuit of nuclear weapons. The Argentine and Swedish cases fall in this category. Fifth, a nuclear option orientation frequently fits the political and psychological needs of leaders better than an unambiguous decision to go nuclear or to forgo a nuclear weapons program. In a hedging situation, the demands of both nuclear proponents and opponents can be equally satisfied: Opponents can be told that no decision in favor of nuclear weapons has been taken, while proponents can hold out hope that their preferred activities have not been foreclosed. It is more than two decades since Meyer wrote his pioneering study on proliferation dynamics and detailed a differentiated model between "zero" activities and a "nuclear weapon state." 15 Today we know that there is no textbook method on how to proliferate. States may try to buy the bomb, as did Australia under Prime Minister Robert Menzies. They may try to build the bomb together with an advanced partner, as did Germany and Italy in 19571958. They may try to jump-start the program by buying foreign technology, expertise, or material, as did Iran and Iraq. They may wish to get the bomb but lack any knowledge about how to do so, stumbling forward without a clear strategy, as was the case with Libya. They also may follow the same bumpy path but half-heartedly and with many diversions as appears to have been the case with Nigeria 16 or, at a higher level of sophistication, Egypt. In addition, military intent might be behind a nuclear program from the beginning but only become discernible to the outside world much later. Political leaders may pursue an "option" policy rather than a weapons policy, as the Shah of Iran did. Alternatively, they may pursue a strategy of permissiveness, letting

132 MULLER AND SCHMIDT

ambitious scientists develop an option without committing themselves to it to preserve deniability toward the outside world, oppositional domestic forces, or even themselves, as may have been the case with South African Prime Minister Vorster in 1971-1974 and Indian Prime Minister Nehru. Leaders may pursue a nuclear option without any clear strategy just in case the country's security situation deteriorates or a rival proceeds down the nuclear road, as was the case with Spain and Chile. As a consequence, it is hard to determine precisely where nuclear activities stand along the spectrum from "no interest" to "full-fledged weapons." The definitional problem is evident in Singh and Way's attempt to distinguish between "exploration" and "program." Some of the activities they note as indicators of a "program" are virtually impossible to detect unless the program is completely documented, as with the United States and the United Kingdom (through their own historians) or South Africa, Iraq, or Libya (through the efforts of the IAEA and the collaboration of the inspected state). For example, "steps toward weaponization" would include work on design, calculations of fission product and neutron multiplication, implosion experimentation, work on metallurgy and electronics, and the like. These activities take place in secret on computers, in small shops and labs, and in bunkers and underground, and they may not be revealed until long after the program has been terminated. For example, after Germany's aborted attempt to participate in a collaborative endeavor with France and Italy to produce the bomb, key German politicians such as Franz Josef Strauss did not give up. It appears that, during the 196os, Strauss shielded a large conglomerate of institutes of the "Fraunhofer Gesellschaft" so that they could conduct applied research on military projects, including those aimed at improving the German capability to build the bomb. By sheer accident, we have learned about flyer plate experiments that were conducted in pursuit of this research, although there is no available corroborating documentation of these activitiesY This can be regarded as among "steps toward weaponization," but the activity was almost unobservable. Sweden, classified by Singh and Way to have reached the first (exploration), but not the second (program), stage of nuclearization, chose a policy of ostensibly forfeiting active nuclear weapons research, including work targeted at developing designs, but pursued "defensive" research to explore the consequences of other states' using nuclear weapons against Sweden. For this purpose, the scientists had to understand how various types of weapons would work, which necessitated studies that looked very much like research on warhead design. According to Jan Prawitz, this work included

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

133

"neutron studies, shaped charges and explosion studies, electronics application, nuclear chemistry and plutonium metallurgy, fission product." 18 Jonter's 2001 study, based on declassified materials of the Swedish National Defense Research Institute (FOA), confirms that "various calculations concerning the construction and operation of nuclear weapons research had also been done as well as theoretical calculations of the critical size of nuclear explosive devices, and pressure, temperature and radiation conditions in nuclear weapons explosions."19 FOA also "looked closely into problems of initiation, which has to do with the actual triggering of nuclear explosive devices," "explosives technology," the calculation of "the smallest possible diameter for a grenade with a nuclear charge," and "testing and production of conventional explosive charges and the development of manufacturing methods in order to be able to analyse different types of charges." 20 When "design research" was officially closed in the mid-196os, studies on design continued under the guise of" defensive research" and with the consent of the leadership. 21 Only after the signature of the NPT in August 1968 was "FOA's more design-oriented nuclear weapons research phased out." 22 For the reasons outlined above, we regard the designs chosen by our four colleagues as method driven rather than adapted to the field and its data constraints. We have thus adopted a simpler research design that is more robust when it comes to dealing with data limitations. We strive to dichotomize time into "before" and "after" periods and, where we perform time series analysis, to deal with five-year periods. We pull together all types of behavior beyond "no nuclear activities" into a single category "nuclear weapons activities," which includes Singh and Way's "exploration," "program," and "weapons possession" phases. This decision is informed by the views that moving from "no interest" to "doing something" is a crucial step and that our knowledge about this step is in many cases better than for the movement between the other ones. In addition, this step requires authorities to take decisions-or a series of them-to return to the "no activity" status. This gives us three categories of states: those with "no activity," "continuing activity" (nuclear weapons possession or some active program) and "past activity" (once started, but reversed). Tables 7.1 and 7.2 demonstrate where we differ in coding decisions from Singh and Way and Jo and Gartzke. The absence of several real "cases" in the Singh and Way and Jo and Gartzke data sets is in part due to the problematic coding rules of the two articles. However, it also probably reflects a lack of knowledge about the field, thus the failure to list Canada, a country once

Table 7.:1..

Differences between Jo/Gartzke and our coding of cases.

States that conduct(ed) nuclear weapons activities, coded according to Jo/Gartzke

United States

States lacking in the fo/Gartzke list, according to Muller/Schmidt

Norway'

Soviet Union/Russia

Germany (Postwar)b

France

Japan (Postwar)'

China

Italy"

Israel

Canada'

India

Australia'

South Africa

Libyas

Pakistan

ChiJeh

Germany (Wartime)

Spain'

Japan (Wartime)

Switzerlandi

'Astrid Forland, "Norway's Nuclear Odyssey: From Optimistic Proponent to Nonproliferator," The Nonproliferation Review 4, no. 2 (Winter 1997): 1-16. b Miiller, "German National Identity and WMD Nonproliferation"; Paolo Cacace, L'Atomica Europea. I progetti dellaguerrafredda, il ruolo dell'Italia, le domande de/futuro (Rome: Fazi Editore, 2004). 'Kurt M. Campbell and Tsuyoshi Sunohara, "Japan: Thinking the Unthinkable," in Kurt M. Campbell et al., Tlze Nuclear Tipping Point, 218-253. 'Leopold Nuti, "Me Too, Please: Italy and the Politics of Nuclear Weapons, 1945-1975," Diplomacy & Statecraft 4, no. 1 (March 1993); Miiller, "German National Identity and WMD Nonproliferation"; Cacace, L'Atomica Europea. 'Potter, Nuclear Power and Nonproliferation: Duane Bratt, "CANDU or CANDON'T: Competing Values behind Canada's Nuclear Sales," The Nonproliferation Review 5, no. 3 (Spring-Summer 1998): 1-16; Duane Bratt, "Canada's Nuclear Schizophrenia," Bulletin of the Atomic Scientists s8, no. 2 (March/April2002): 44-SO. 'Jim Walsh, "Surprise Down Under: The Secret History of Australia's Nuclear Ambitions," The Nonproliferation Review 5, no. 1 (Fall1997): 1-20; Jim Walsh, "Bombs Unbuilt: Power, Ideas, and Institutions in International Politics" (PhD diss., Massachusetts Institute of Technology, 2ooo). 'International Atomic Energy Agency (IAEA), "Implementation of the NPT Safeguards Agreement of the Socialist People's Libyan Arab Jamahiriya," Report by the Director General, GOV/2004/12, Vienna, February 20, 2004; IAEA, "Implementation of the NPT Safeguards Agreement of the Socialist People's Libyan Arab jamahiriya," Report by the Director General, GOVhoo4/33, Vienna, May 28, 2004; IAEA, "Implementation of the NPT Safeguards Agreement of the Socialist People's Libyan Arab jamahiriya," Report by the Director General, GOV/2oo4/59, Vienna, August 30, 2004; Harald Muller, "The Exceptional End to the Extraordinary Libyan Nuclear Quest," in Morten Bremer Maerli and Sverre Lodgaard, eds., Nuclear Proliferation and International Security (London/New York: Routledge, 2007), 73-95. h On Chile, a good study is still lacking. We include it because of heavy circumstantial evidence. Run by a dictatorship, it had an enduring territorial conflict with Argentina that almost led to war in the late 1970s. Significantly, Chile had an extensive nuclear research program and acceded to the NPT and accepted full scope safeguards only when Argentina did. 'Angel Vinas, "Spain," in jozef Goldblat, ed., Non-Proliferation: The Why and the Wherefore (London: Taylor and Francis, 1985). ; Theodor Winkler, Kernenergie und Aufienpolitik: Die internationalen Bemuhungen urn eine Nichtweiterverbreitung der Kernwaffen und die friedliche Nutzung der Kernenergie in der Schweiz (Berlin: Berlin Verlag, 1981).

(Continued)

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

Table 7.1.

135

(Continued)

States that conduct(ed) nuclear weapons activities, coded according to ]o!Gartzke

States lacking in the fo/Gartzke list, according to Muller/Schmidt

Sweden

Egyptk

Yugoslavia

Indonesia1

Taiwan (70s only)

Nigeriam

South Korea (70 only)

Syria"

Iran Iraq Argentina Brazil Romania North Korea 'Walsh, "Bombs Unbuilt"; Robert). Einhorn, "Egypt: Frustrated but Still on a Non-Nuclear Course," in Kurt M. Campbell et al., The Nuclear Tipping Point, 43-82. 1 Kusnanto Anggoro, "The Domestic Roots ofProactivist Non-Nuclear Policy: The Case of Indonesia," Paper for the PRIF Project on Proactivist Non-Nuclear Policy, jakarta, Centre for Strategic and International Studies, 1997 (mimeo); Robert M. Cornejo, "When Sukarno Sought the Bomb: Indonesian Nuclear Aspirations in the Mid-1960s," The Nonproliferation Review 7, no. 2 (Summer 2000): 31-43. m Adibe, "0Jigeria: The Domestic Basis of a Proactive Non-Nuclear Policy." "Avner Cohen and Leonard S. Spector, "Israel's Airstrike at Syria's Nuclear Reactor: Implications for the Nonproliferation Regime," Arms Control Today 38, no. 6 (July/ August 2008).

involved in the Manhattan project, or Italy and Germany, which attempted to collaborate with France to acquire a national deterrent. Those and other missing cases, such as the Australian, Swiss, Japanese, and Egyptian efforts, have been well documented in English-language publications. The most egregious mistake is the judgment by Jo and Gartzke that Libya stopped nuclear activities in favor of chemical weapons in the late 1980s. Their only reference to Libya is a book published in 1990, 23 even though Libya made headlines in December 2003 by renouncing all weapons of mass destruction, including nuclear weapons, and admitting it had acquired, throughout the 1990s, material, equipment, and even weapons design from the A. Q. Khan network. Limited knowledge about nonproliferation is also evident in Jo and Gartzke's comments about the NPT. For example, they maintain that "NPT protocols requiring the dissemination of nuclear knowledge and materials suggest that the NPT may actually contribute to the quickening pace of nuclear diffusion." 24 The NPT, however, has no protocols. Article IV of the NPT confirms the "inalienable right" of all states to enjoy the peaceful uses of nuclear

136

MULLER AND SCHMIDT

Table 7.2.

Differences between Singh/Way and our coding of the cases involved.

States that conduct(ed) nuclear weapons activities, coded according to Singh/Way

States lacking in the Singh!Way list, according to Muller/Schmidt

Algeria

Canada

Argentina

Germany

Australia

Italy

Brazil

Egypt

China

Indonesia

France

Norway

Iran

Spain

Iraq

Chile

India

Nigeria

Israel

Japan

Libya

Syria

Pakistan Romania Soviet Union/Russia Sweden Switzerland South Africa South Korea Taiwan (twice) United Kingdom USA Yugoslavia

energy and contains an undertaking of all states parties to cooperate in this endeavor but in accordance with Articles I (nontransfer of nuclear weapons by nuclear weapons states) and II (nonacquisition of nuclear weapons by nonnuclear weapons states). The treaty thus contains no "requirement to disseminate." As for the sixty-seven countries that obtained nuclear research reactorsthe most significant first "dissemination" on the way to a nuclear weapons capability-almost two-thirds did so before the NPT entered into force. In fact, the limited nature of post-NPT assistance explains the eternal complaint of the nonaligned countries that Article IV was never fully implemented. The fundamentally false proposition about the "dissemination" dysfunction of the

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

:137

NPT plays a major role in the causal reasoning of Jo and Gartzke about the alleged insignificance of the NPT in stemming proliferation. 25 The omission of cases in the two articles is not distributed randomly but reflects a bias that heavily influences the results. They depict fewer cases of allied countries exploring a nuclear option or seeking nuclear weapons and omit countries that gave up activities after the NPT entered into force. These omissions are of great importance as the security guarantee as a decisive nonproliferation instrument has a major bearing on their findings, as does the insignificance of the NPT for Jo and Gartzke. We also find the thesis that nuclear threats are likely to stem proliferation to be both disturbing and unconvincing and believe it is most probably an artifact of definitional and coding mistakes. Looking at the most critical cases, the countries that possess nuclear weapons or still active related programs (see Table 7.3), in all but one case-Israel-the nuclear threat played a direct or indirect role. The United States launched the Manhattan project against (supposedly) proliferating Germany. The USSR had a sharp political conflict with a nucleararmed United States, and France was even directly threatened by the Soviet leader Nikolai Bulganin during the Suez War. China received a threat from the United States during the Korean War and the Taiwan Strait crisis, and India was threatened by China during the war against Pakistan in 1964. Pakistan was in conflict with proliferating India, while South Africa feared the Soviet presence in Africa. Iran was involved in a shooting naval war with the United Table 7.3.

Nuclear threats and proliferation. Explicit nuclear threat

Military conflict withNWS

Indirect nuclear threat

Conflict with proliferator

X

United States USSR/Russia

X

United Kingdom

X

France

X

China

X

Israel India

X

X

X

Pakistan

X

South Africa Iran

X

X

X

138

MULLER AND SCHMIDT

States in 1987 and learned that "all options are on the table" several times during the last several years. We readily accept the importance both studies attach to the external security situation. However, we doubt that the objective measure applied by either makes sense. Security is very much a state of mind: The external situation has to be assessed and evaluated, and strategies to cope with it have to be devised. As Hymans and Sagan have rightly pointed out, it is security perception by the actors, not security assessment by faraway scholars, that determines the policies of a state. 26 Scholars can argue forever that South Africa had no security problem, but: As long as they ignore the very specific "Wagenburg" mentality of the South African leaders at the time, they will not understand South African decision making. 27 The idea of the "total onslaught" orchestrated by a Soviet Union, which may-improbably but not impossibly-use or make available for use chemical or even nuclear weapons, appealed to this mentality and gave the experiment-happy parts of the nuclear bureaucracy the window of opportunity to frame the issue accordingly. 28 In that sense, the security situation must provide the resources for "nuclear myth makers" 29 to make their case, but it is not a variable from which one can deduce outcomes without taking into account perceptions, and it is far from certain what strategies political leaders in different countries with different leading coalitions and political cultures will choose in response to a threat assessment. Nonproliferation Research: Qualitative Studies

The most stimulating studies on why states have renounced nuclear weapons are by T. V. Paul, James Walsh, Ariel Levite, Etel Solingen, and Jacques Hymans. 30 While we share some of their conclusions and have greatly profited from all, we have some differences with each one. Paul and Levite side with the realist thesis, while Solingen and Hymans focus on internal factors. These theoretical predilections have had an impact on their assessment of cases: A slight paradigmatic bias tilts the interpretation of some cases in one direction, and this in turn influences their comparative conclusions. Paul has difficulty coping with the impact of moral and political factors that were significant in Sweden and also in Switzerland (where economics also played an important role). A realist bias leads him to focus on both countries' interests in avoiding the wrath of hostile nuclear powers. 31 Detailed country studies, however, demonstrate that security arguments were employed to save face, and thereby co-opt, the military brass in Sweden32 and honorable patriots

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

139

in Switzerland33 who favored the bomb but were outnumbered, over time, by bomb opponents. The studies by Levite and Paul conceptually are the most sophisticated in the "nuclear reversal" field. Levite pays some attention to norms but subscribes to the cynical view that F. W. de Klerk gave up South Africa's nuclear weapons to prevent them from falling into the hands of the black majority government. The most meticulous studies on South Africa, however, emphasize other aspects: de Klerk wanted to reenter the international community and was aware that being an (illegal) nuclear state would run counter to this objective. The National Party did not want to let its successors receive the laurels for denuclearization, and racist South Africans were keen to keep the nuclear option open, in case of a reversal. 34 The "race" thesis is at least contested and most probably not tenable. Levite also paints too generous a picture of the positive U.S. influence on the (non)proliferation game. We agree with him that Washington helped in a couple of cases, most obviously in East Asia, and to some degree, but possibly less than generally believed, in Europe. At the same time, U.S. policies drove other countries down the nuclear road, starting, to some degree, with the Soviet Union, followed by China, India, Israel (where Washington denied what it was willing to offer to Japan), North Korea, Libya, Iraq, and Iran. This is not to speak of the general symbolic example of "nuclear achievement" the United States might have given to status seekers among middle powers, as the country with the most sophisticated nuclear arsenal and the most articulate nuclear strategy. In addition, U.S. policies stimulated temporarily the nuclear ambitions of certain allies (Taiwan, South Korea, Japan) before Washington remedied this undesired effect by countermeasures, as Levite himself notes. In short, the record is more mixed than Levite's very positive balance suggests. Solingen and Hymans have done the most impressive work on domestic drivers of (non)proliferation. Solingen develops a theory focused on the politicalideological orientation of ruling coalitions. Nationalist, inward-looking, autarkic coalitions are liable to nuclear temptations; liberal, outward-looking coalitions are less interested in nuclear weapons. The change from one regime/ coalition to the other helps explain changes in nuclear policies. Solingen gives a convincing account of some cases, but problems remain. For example, Solingen denies that security concerns played a major role in Libya's decision to move from erratic attempts to "buy the bomb" to a systematic if not always well-devised procurement program to produce the bomb at home. However,

l.40

MULLER AND SCHMIDT

this shift happened exactly at the time when Libya was under great political, economic, and, above all, military pressure from the Reagan administration, which culminated in several air and naval battles close to Libyan territory. 35 That the Libyan leadership denied security problems in its explanations for giving up the program-the main evidence for Solingen's thesis-should be interpreted more as the leadership's attempt to justify the policy change to its supporters than as an authentic account of the motivations that drove its policy in the early and mid -1980s. A second problem with Solingen's approach is the relationship between democratization and liberalization. It might be that the latter is-in cases where liberalization occurred-an intervening rather than an independent variable. The liberalization movement could as well result from the change of mind of the leader, or a small group of leaders, in reaction to external events (which would make the "change of mind" an intervening variable as well) rather than through a change in ruling coalitions. We attempt to test below how much Solingen's approach captures the pattern of nuclear reversals. Hymans's work has the great merit of pointing us toward the belief system ofleaders as opposed to supposedly objective environmental/systemic determinants of behavior. His four-set ideal-type model identifies "oppositional nationalist" as the mind-set likely to desire and pursue nuclear weapons. The emergence of leaders with different mind-sets, then, is responsible for reversal. Hymans's approach does not appear incompatible with Solingen's, and it might complement it when a country is not led by coalitions but is very much dependent on the will of a single leader or a very small group of powerful individuals. We disagree, however, with some ofHymans's assessments of specific cases. His argument that Australia moved toward the bomb only under the nationalistic Prime Minister Gorton ignores the continuous attempts to acquire the bomb from the United Kingdom. In the 1950s and early 1960s, one could not rule out such a possibility, as the antiproliferation attitude of the great powers had not yet taken shape. Australia could have hoped to find British sympathy, as Britain used Australian territory to test its nuclear arsenal. 36 The second case where we view things differently is Argentina, which Hymans excludes from the list of countries that pursued nuclear weapons activities. 37 We respect the fact that he had access to new, official sources, but the fact that he found no indication that the option was considered is a typical "dog-that-didn't-bark" phenomenon. It seems implausible, given that India tested a device and Brazil

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

141

started its own program, that no one in Buenos Aires at least raised the issue_ This seems too good to be true, and the explanation of the puzzle might be found in Hymans's admission that the access to documents was selective_ It appears reasonable, on the basis of his findings, to assume that the Argentinean pursuit of the option was slow and lacked the degree of determination that we found in the United States, China, or even Iraq_ In terms of Hymans's model, it is hard to understand how a regime that massacred several thousands of its citizens because of a supposed danger of a leftist takeover lacked the deep hostility that Hymans uses as indicator of"oppositional" thinking and that a leadership that almost went to war with a neighbor, Chile, and invaded the Falkland Islands does not fall under an extreme nationalist category_ 38 We are concerned that the approaches by Solingen and Hymans pay too little attention to external conditions under which decisions were taken_ While the focus on internal variables is a healthy antidote to the realist fixation on the international system, it is advisable to look at the interplay ofboth levels. 39 We try to do so in testing the most popular explanations for nuclear reversal. THEORIES OF NUCLEAR ABSTENTION AND REVERSAL Economic and Technological Pull

One important hypothesis about proliferation is that it is capability driven. Capability may be defined in economic or technological terms. 40 To test the "economic pull" hypothesis, we first identified the economic threshold necessary to start nuclear weapons activities on the basis of the state with the lowest economic capability that initiated such activities. Next, we calculated for each period the number of states whose economic capability met or exceeded the required threshold. The economic capability of a country is usually estimated on the basis of its GDP. To attain the best possible comparability of states' GDPs over the investigated time frame (1960-2005), we assessed economic capability in constant "1990 international Geary-Khamis dollars." 41 We thus ascertained that the economic threshold to start nuclear weapons activities is 19.64 billion dollars ofGDP, which corresponds to Egypt in 1951. This threshold is the result of two exceptions we made in analyzing our economic data set. First, for a long time Israel's GDP fell far below the assumed limit. The lowest value is 4.7 billion dollars in 1951, by which time Israel had already invested significant resources in the development of nuclear weapons. This extremely low value supports our thesis that economic factors are

142 MULLER AND SCHMIDT

almost completely irrelevant for the initiation of nuclear weapons activities. Rather, it is a question of political willingness to devote a considerable share of state's resources to the military sector instead of, say, strengthening the public welfare program. Nevertheless, we treated Israel as an "outlier" and therefore regarded the second-lowest GDP in the group of states with nuclear weapons activity as the necessary threshold for the initiation of such activities. Secondly, the economic data concerning Libya appeared inaccurate, as the assessment of its GDP is consistently too low. This estimation is confirmed by such data sets as the CIA World Factbook and the "World Military Expenditures and Arms Transfers" (WMEAT) yearbook, where Libya's GDP always exceeds that of Egypt. Because Maddison's data set ends in 2003 we supplemented the missing data on the basis of the Total Economy Database, which also assesses countries' economic capability in "1990 international Geary-Khamis dollars."42 The

100

80

60

40

0+----.----r---.----.----.----.---.----.----4 1960

1965

1970

1975

-o- NWA as a percentage of NWA-capable states Figure 7.6.

-

1980

1985

No. of states with NWA

1990 -

1995

2000

2005

No. of NWA-capable economies

Economic capability and nuclear weapons activities, 1960 to 200 5 .

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

143

70 60 50 40

30 20 10

0~~~~~~--~~~--~~~--~~ 1945

1955 1950

1965 1960

1975 1970

-States with RR and NWA Figure 7.7.

1985 1980 -

1995 1990

2005 2000

Cu. no. of states with RR

Nuclear research reactors (RR) and nuclear weapons activities, 1945 to 2005.

absence of data for Afghanistan, North Korea, Libya, and Nepal was completed on the basis of the "World Development Indicators" (2007) data set of the World Bank. As we employ only an ordinal-scaled data set for economic capability, this measure is adequate for our purposes. Figure 7.6 depicts our findings: The statement that nuclear weapons activities are minority business, and a declining one at that, is confirmed. Economic capability is not a sufficient cause for states to have initiated such activities. On the contrary, the closer a country approaches the minimum level necessary to start nuclear weapons activities, the less it appears to be inclined to do so. To test the technological pull hypothesis, we have compared the evolution of research reactor possession with the evolution of nuclear weapons activities (see Figure 7.7). More precisely, we examined for each period how many states

144

MULLER AND SCHMIDT

whose research reactor(s) passed the criticality threshold pursued nuclear weapons activities at the same timeY While during the early period nuclear weapons activities and the possession of a research reactor were closely related, this is not the case any longer. An increasing number of states have such reactors, while a decreasing number are conducting nuclear weapons activities. The "technology pull" thesis is further refuted by the fact that more states started nuclear weapons activities before they acquired research reactors (twenty-one of thirty-seven), and more abandoned them after their acquisition (twenty-five of thirty-seven)Y There is no indication at all of a "technological pull." Security Needs and Guarantees

The most prominent security-centric thesis for nuclear abstention is the alleged role of alliance guarantees in dissuading non-nuclear weapons states from pursuing nuclear weapons. 45 To test this hypothesis we have compared U.S. allies' behavior in Europe, East and Southeast Asia, and the Pacific with that of nonaligned countries in the same regions (see Table 7-4). We have selected these cases because they belong, by and large, to the same conflict complex (Cold War) and can thus be assumed to have a comparable security environment. According to the "alliance hypothesis," the share of nonaligned countries with nuclear weapons activities should be significantly higher than that of U.S. allies. We excluded states that did not have the economic capability to initiate nuclear weapons activities on the basis of the economic threshold discussed previously. The hypothesis is not confirmed by the empirical data: The correlation is not significant at the level of p > 0.1. At most, we were able to identify a tendency in the opposite direction: Allies were slightly more likely to launch Table 7.4.

U.S. allies and nuclear weapons activities: Europe, East Asia, Southeast Asia, Pacific region (bi- and multilateral alliances), controlled for economic capability. Nuclear weapons activities

Status of alliance

Allied Nonaligned Total

Nuclear weapons activities 10

No nuclear weapons activities 9

Total 19

4

7

11

14

16

30

Chi-Square-Value: 0.74 (p > 0.1) I Phi-Coefficient: 0.16

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

145

nuclear weapons activities. But as the chi-square test and the phi coefficient show, this relation is not statistically significant. Another proposition could be that the readiness of allies to renounce nuclear weapons grew with the quality and credibility of the nuclear guarantee. We identified the following turning points in the European and Asian alliances:

Europe: From massive retaliation (unconditional nuclear guarantee against aggression based on U.S. superiority) to flexible response (conditional nuclear guarantee depending on the course of war, type of aggression, and U.S. decision making under conditions of uncertainty); weakening of guarantee after 1962 Asia/Pacific: Guam Doctrine of 1969: Asian allies responsible for their defense, reduction of U.S. role to "backup" guarantee; weakening of guarantee after 1969 There appears to be no correlation between the strength of the guarantee and nuclear weapons activities; contrary to the "strength of alliance" hypothesis, the majority of U.S. allies that started nuclear weapons activities did so before the weakening of the guarantee (nine out of ten). Moreover, while the guarantees tended to erode, nuclear weapons activities decreased as well: Eight out of ten allies renounced their nuclear weapons activities after a weakening of the guarantee, the exceptions being France and Great Britain. Our point is not that alliance plays no role in countries' nuclear renunciation decisions; the evidence from some individual case studies is too strong to completely refute this assumption. What is clear, however, is that the causal relationship is not as straightforward as often suggested. Alliance alone is not a sufficient condition, and additional reinforcing factors are needed, such as domestic opposition, external pressures, normative concerns, more benign security environment, and the like. The "High Posture" Hypothesis

There is another "realpolitik" hypothesis, which Enid Schoettle has labeled the "high posture" assumption: If the big powers are strong, others do not dare to proliferate. 46 If they disarm, and the marginal utility of small arsenals becomes higher, incentives for proliferation will rise. Arms build-ups are usually connected to tensions between the superpowers, while disarmament is associated with periods of detente. We use "political climate" between the superpowers as an indicator of the overall political climate, which creates expectations of future behavior and is thus likely to have an impact on longer-term strategic decisions, including the pursuit of nuclear weapons activities. 47

146

MULLER AND SCHMIDT

Table 7.5.

Detente periods versus Cold War periods. Trend in nuclear weapons activities

World political climate Cold War

NWA growth

NWA stagnation

2

Detente 6

Total 4

4

Between Cold War and detente

Total

NWA reduction

2

2

5

2

3

4

12

As Table 7.5 demonstrates, the relationship is, at best, spurious. During the Cold War period, a relatively large number of countries started nuclear weapons activities. In the early 196os-after the Cuban Missile Crisis and the conclusion of the Partial Test Ban Treaty-proliferation stagnated in relative terms. Stagnation held during the 1970s. The "New Cold War" in the early 198os saw a slight revision of the trend. With the new detente, terminations accelerated, while new activities did not start. If anything, the relationship of detente/disarmament and proliferation is the reverse of what is assumed by the "realpolitik" thesis. The "Norms" Thesis

Those who do not subscribe to realist assumptions frequently maintain that nuclear abstention or reversal is strongly influenced by the normative environment. According to this perspective, the NPT has had a decisive impact on the curve of proliferation and nonproliferation. 48 After 1960, the number of states with nuclear weapons activities declined relative to the total number of states. After 1970, the number of states with nuclear weapons activities remained almost constant for a decade in absolute terms. From 1985 on, there was a steep decline in the number of states with nuclear weapons activities. What could account for the beginning of the relative decline of nuclear weapons activities? In December 1961, the "Irish Resolution" was adopted unanimously by the U.N. General Assembly. This was the first significant manifestation of a universal norm against the further spread of nuclear weapons. Is it possible that this event had a serious impact? To look deeper into the issue, it is revealing to distinguish between "old states," who were independent before 1960 (that is, before the lobbying for the new norm started), and those who gained their independence afterward (see Table 7-6). The chi-square test points to a significant relationship between the time of the foundation of a state and its inclination to conduct nuclear weapons

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

Table 7.6.

147

"Old" and "new" states and their nuclear weapons activities. Number of states before 1960

Number of states after 1960

Total

States with NWA

33

States without NWA

73

86

159

106

87

193

Total

34

Chi-Square-Value: 29.60 (p < 0.001) I Phi-Coefficient: 0.39

activities at the level ofp < o.oo1. The phi coefficient is 0.39. Thus, we can assume a moderate correlation between the two variables. The picture does not change significantly if we control for economic capability, as Table 7-7 demonstrates. While quite a few of the new states did not dispose of the economic capabilities necessary to start nuclear weapons activities, a sufficient number did, and the share of those that embarked on a nuclear course is considerably lower than that of the "old states." For states that gained their independence after the Irish Resolution, being nonnuclear was seen as an appropriate status, the attribute of a "good citizen" of the world community of states. For the "old states," the new norm competed with the old understanding that a state was entitled to acquire armament according to the standard of the time. This is an indication, but not proof, that the debate over and codification of a new, if still weak, international norm influenced the way the new states viewed proper behavior and shaped their own understanding of security. For the old states, the impact was weaker, but the series of terminations of nuclear weapons activities started during that period. In 1968, a much stronger norm was created: The NPT was signed, and it entered into force in 1970. Our assumption about norm diffusion differs from the conception of Jo and Gartzke, who suppose that norm strength grows with the accession of countries to the NPT. A certain linearity is implied in that concept that misses the salience of the signature/entry into force event (1968-1970), which Table 7.7.

"Old" and "new" states and their nuclear weapons activities, controlled for economic capabilities. Number of states before 1960

Number of states after 1960

Total 33

States with NWA

32

States without NWA

48

22

70

Total

80

23

103

Chi-square-value: 10.43 (p < o.oo1) /phi-coefficient: 0.32

148

MULLER AND SCHMIDT

Table 7.8.

Nuclear weapons activities initiation and termination before and after the NPT as a percentage of economically capable states. BeforeNPT

After NPT

NWA start

41.38% (24 of 58)

15.66 % (13 of 83)

NWAstop

12.50 % (3 of 24)

67.65 % (23 of34)

changed the normative environment decisively. For that reason, a before/after differentiation is more telling. The figures in Table 7.8 give a very strong indication that the NPT had a significant impact. Before the treaty's entry into force, more than two-fifths of the states that possessed the required economic capability to start nuclear weapons activities did so. In addition, only a fraction of these states renounced their nuclear weapons programs in the absence of a strong nonproliferation norm. Yet the situation looked very different after the NPT altered the normative environment. Since 1970, less than 15 percent of the economically capable states have started nuclear weapons programs. Even more striking is that since the inception of the nonproliferation norm, almost 70 percent of all nuclear weapons activity states gave up those activities. To sum up, while the Irish Resolution appears to have contributed to the slower pace of proliferation in relative terms, following the conclusion of the NPT one can speak about deproliferation in absolute terms. The next question we wish to tackle, then, is the sudden steep rise of weapons activity termination in the second half of the 1980s. The change of Soviet policy, and the ensuing improvement of international relations, must have played a role. But what kind of role? Rather than looking for a direct relationship, it is instructive to ask about the most important indirect effect: the change in states' domestic political systems. To categorize regime types, we relied on the well-established Polity IV data set. The assumption is that states that go through a process of democratization are more inclined to subscribe to the international nonproliferation norm and terminate their nuclear weapons activities than autocratic regimes or states that are undergoing autocratization. Democratization or autocratization, respectively, is defined as a change of a state's polity value in either direction. Therefore, we examined the states' polity values at the start of their nuclear weapons activities and at the time of their termination. For states continuing nuclear weapons activities, we took the polity value as of 2004.

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

149

Definition of regime type by means of its polity value: -10

too:

1 to 6: 7 to 10:

autocracy semidemocracy democracy

The results presented in Table 7-9 are twofold: On the one hand, both democracies that had nuclear weapons activities but had not yet achieved sufficient progress and states in the process of democratization were sensitive to changes in the normative environment. Furthermore, countries that switched from authoritarian regimes to more democratic systems were likely to have submitted to the existing norm and to have renounced nuclear weapons activities that the ancien regime had begun. On the other hand, democracies that had started their nuclear weapons activities long before the spread of the nonproliferation norm showed no inclination to get rid of their "acquis," no matter what the norms-or majority opinion-prescribed. We assume that some kind of path dependency may account for this finding. Where nuclear weapons activities had already created powerful bureaucracies and decisionmaking momentum, if not nuclear weapons hardware, the reversal of such a process can be hard to achieve. Where full-fledged nuclear arsenals had become integrated into the national security doctrines of democracies or where being a nuclear weapons state had become part of the national identity, this effect was even stronger. 49 To test this assumption, we compared the duration of democracies' nuclear weapons activities before the NPT came into force. The thesis is that democracies that still continue their nuclear weapons programs pursued weapons activities over a considerably longer period than those democracies who renounced these activities with or soon after the advent of the nonproliferation norm. As Table 7.10 reveals, our thesis is confirmed: States whose nuclear weapons activities lasted less than fifteen years before the NPT came into force renounced these activities relatively quickly. The duration of their weapons activities averages 10.60 years. States whose nuclear weapons activities lasted sixteen years or more before the NPT became effective exercised the nuclear option. The duration of their programs averages 23.60 years. These results suggest that the probability of nuclear renunciation declines with the duration of nuclear weapons activities and the accompanied institutionalization and bureaucratization of such activities. Democracy thus appears not to be an absolute guarantee against nuclear temptations. The reason might well be due to the ambivalent relationship

Change of polity value and nuclear weapons activities (Belarus, Ukraine, and Kazakhstan excluded).

Table 7.9.

State

United States United Kingdom India Israel France Pakistan North Korea

Political system: start

Political system: Stop/continuing

Democracy Democracy

Democracy Democracy

Democracy Democracy

Democracy

Democracy Semidemocracy Autocracy

Democracy Autocratization Autocratization Autocratization

Iran China

Autocracy

USSR/Russia

Autocracy

Democratization

Syria Canada

Autocracy Democracy

Democratization

Norway

Democracy

Sweden Australia

Democracy Democracy

Democracy Democracy

Germany Italy

Democracy Democracy

Switzerland Japan Brazil Chile

Democracy Democracy Semidemocracy Sem idemocracy Semidemocracy

Democracy Democracy Democratization

Semidemocracy Autocracy Autocracy

Democratization Democratization Democratization Democratization

South Korea South Africa Argentina Yugoslavia Taiwan Egypt Nigeria

Autocracy

Autocracy Autocracy

Democratization Democratization

Democracy

Democracy Democracy Democracy

Democratization Democratization

Autocracy

Democratization Democratization

Autocracy

Democratization

Algeria

Autocracy Autocracy

Democratization Democratization

Indonesia

Autocracy

Autocratization

Iraq

Autocracy

Autocratization

Libya

Autocracy

Autocracy

Spain Romania

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

Table 7.10.

151

Democracies' start ofNWA before the NPT (1970)."

State

Start ofNWA (year)

Years ofNWA (max-NP1}

United States

1939

31 years (1970)

United Kingdom

1939

31 years (1970)

France

1954

16 years (1970)

India

1948

22 years (1970)

Israel

1952

18 years (1970)

Australia

1956

14 years (1970/1973 stop)

Germany

1957

13 years (1970)

Italy

1957

13 years (1970)

Switzerland

1958

12 years (1970/1977 stop)

)apan

1969

1 year (1970)

'We excluded Canada, Norway, and Sweden because they renounced their NWA before 1970.

between nuclear weapons and the basic liberal values that democracies embrace. On the one hand, the devastating effects of nuclear weapons use clearly run counter to the fundamental belief of democracies in human dignity and the right to life. On the other hand, the strong aversion to war among many democratic populations may encourage the possession of nuclear weapons as a deterrent against all war. In the context of French thinking about nuclear weapons, this has been labeled nuclear pacifism. 50 India is another case in point. It started its nuclear activities almost immediately after gaining independence. While the entrenched bureaucracy advocated going nuclear, the political leadership-influenced by the moral heritage of Gandhi-pursued disarmament for a comparatively long time, especially in light of the security situation vis-a-vis China and Pakistan. So why did India pass the threshold in 1998 and move from mere capacity to actual nuclear weapons? While realists such as Sumit Ganguly point to external security threats, 51 we assume there are more important reasons for the change in India's security policy. First, the nuclear weapons states showed no inclination to disarm. Instead, the NPT was extended indefinitely in 1995 and the conclusion of the Comprehensive Test Ban Treaty (CTBT) was looming on the horizon, closing the window of opportunity (for testing) and perpetuating the regime of "nuclear apartheid." 52 Second, the newly elected and power-savvy coalition led by the Bharatiya Janata Party (BJP) concluded that the possession of nuclear weapons was the ultimate precondition for recognition as a "global player" and inclusion as one of the "centers in a polycentric world." 53

152

MULLER AND SCHMIDT

Table 7.11.

Political system and start of nuclear weapons activities after the NPT.

State

Political system: Start

South Africa

Semidemocracy

Nigeria

Autocracy

Iran

Autocracy

Iraq

Autocracy

Spain

Autocracy

North Korea

Autocracy

Romania

Autocracy

Algeria

Autocracy

Libya

Autocracy

Syria

Autocracy

Under the right historical conditions in which there is no entrenched pronuclear bureaucracy or prevailing national nuclear identity or ideology, democracy may have helped to counter nuclear weapons ambitions by fostering the inhumane image of nuclear weapons and an awareness of the new international nonproliferation norm. In that context, it was important that the NPT was viewed as embracing the complete elimination of nuclear weapons both in its preamble and in Article VI. Once the norm was framed in a way that resonated widely with the public, the mechanisms of democracy influenced adherence to the norm by national leaders. 54 This phenomenon might explain why, after the NPT came into force, not a single democracy was inclined to start nuclear weapons activities, as Table 7.11 demonstrates. Except for Iran and North Korea, all of the states in Table 7.11 renounced their nuclear weapons activities sooner or later (although Syria's activities were stopped two years after the end of our observation period, in 2007). A vast majority of those states also passed through a process of democratization. These findings support the assumptions that democracies are more sensitive to the normative structure of the international environment and that increasing democratization contributes to norm conformity-or, in this case, the renunciation of nuclear weapons activities. Iran and North Korea fit well the criteria for "pariah status," which Richard Betts has identified as one of the driving forces for going nuclear: an isolation from the international community that makes an existential security threat even more ominous and, at the same time, reduces the effect of the norms that the international community propagates.

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

153

Isolated, autocratic regimes, in addition, are prone to paranoid attitudes that push them even more in the direction of nuclear weapons. 55 One could argue that it is not the polity change toward democracy that influences a state's decision to continue nuclear weapons activities or to terminate them. Possibly, it is the rising degree of economic liberalization that correlates with the renunciation of nuclear weapons activities. The "Liberalization" Thesis

To assess the power ofEtel Solingen's "liberalization" hypothesis, we compared the average degree of liberalization of states that stopped their nuclear weapons activities after the NPT with states that continued their nuclear programs (see Table 7.12). We used trade as a percentage of GDP as a proxy indicator to estimate a state's degree of liberalization and relied on the "World Development Indicators" (2007) data set of the World Bank. Admittedly, caution should be exercised in interpreting the results because of the lack of reliable data for a significant number of states. Despite data shortcomings, the following picture emerges: Taking into consideration the slight difference in the averages of trade as proportion of GDP and the standard deviation, it appears that the political economy hypotheses are not supported. As Table 7.13 shows, for all nuclear weapons states, the share of trade in GDP rose during the investigated time frame. Nonetheless, all states stuck to their nuclear weapons programs or even accelerated their activities during the 1990s, for example, India and Pakistan. None of the NWS party to the NPT abandoned their nuclear weapons programs, although they are bound by the provision of Article VI and are all liberalizing states. As with the alliance hypothesis, we do not see our results as an outright refutation of Solingen's hypothesis, but one has to examine individual cases Table 7.12.

Average degree of liberalization and nuclear weapons activities after the NPT.

Trade (as a percentage of GDP), states that:

N

Min

Max

Average

Standard deviation

Stopped their NWA after the NPT'

13

20

86

50.77

23.31

Continued their NWA after the NPTb

10

25

97

54.10

20.27

'No data available for Taiwan, Yugoslavia, Iraq, Romania, Belarus, Ukraine, and Kazakhstan. "No data available for Korth Korea.

154 MULLER AND SCHMIDT

Table 7.13.

Trade (as a percentage of GDP) trends of current NWS from 1970 to 2004.

State

Trade (as a percentage

Trade (as a percentage

ofGDP), 1970

ofGDP), 2004

United States

11%

25%

United Kingdom

44%

54%

France

31%

51%

China

5%

65%

Russia

no data untill989 (43 %)

57%

lsrael

79%

97 o/o

8%

40%

22 o/o

31 o/o

India Pakistan

very carefully. In some instances, liberalization might be the consequence of democratization. In others, the ratchet effect of far-reaching activities and bureaucratic build-up might pose an insuperable hurdle to reversal, or the perception of the security situation is too pessimistic to stop activities. In any case, the normative environment plays a major role in whether a state is inclined to pursue nuclear activities or to give them up: The ratchet effect works only if the normative environment was permissive before the ratchet had clicked in. CONCLUSIONS

Nuclear weapons proliferation is not natural, is not a historical law, is not irresistible, and is not irreversible. A considerable number of states have launched nuclear weapons activities. Most of them started during the peak of the Cold War. Most of them terminated their activities after the basic nonproliferation norm-embodied in the NPT-was codified. Most of them stopped when the nuclear-armed superpowers themselves began a process of arms reduction. We discover no economic or capability pull. Decisive capability for program development-research reactors-does not create an irresistible temptation to work on nuclear weapons, and weapons programs can be terminated after significant technical capabilities have been acquired. In that sense, technical skills can be "disinvented." 56 Being allied to a nuclear weapons state is not a necessary condition for nonnuclear weapons states to abstain from nuclear weapons activities, while being nonaligned is not a sufficient condition for the pursuit of nuclear weapons. Nuclear proliferation does not grow when nuclear weapons states move toward

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

155

disarmament The emerging norm and domestic political system in combination with the norm appear to exert the strongest influence on state behavior. Why is this so? Norms affect state behavior in several ways. 57 Under a utilitarian perspective, they lead to a new probability distribution of sanctions and rewards: The NPT has changed states' expectation of the utility and the costs of nuclear weapons. Second, they create a system of reference that changes the influence of the parties to a domestic debate on nuclear weapons pursuit, thereby affecting the intraelite and intragovernmental balance of power, as Walsh's study on Australia has highlighted. 58 Thirdly, and maybe most importantly, they change the assumptions about "what states do." From the 1950s until the mid-196os, it was widely assumed that more states would procure nuclear weapons. There was no opprobrium connected to that practice; Chancellor Adenauer and Prime Minister Sato assumed it was quite natural that, once the shadow of the dark past of their countries disappeared, Germany and Japan would become nuclear weapons states. The Australian military even opined that having nuclear weapons would be the condition to be taken seriously in the Alliance. 59 The entry into force of the NPT changed these assumptions about appropriate state behavior; from then on, going nuclear was decisively not "what states did." In a relatively short period of time, this notion became internalized as a political habit to the extent that when changes occurred in the security environment they did not necessarily lead to a reconsideration of one's nonnuclear weapons status. In the sense of Bachrach and Baratz's concept of "nondecisions," restarting nuclear activities ceased to be a policy option because it was simply kept off the agenda by a self-explanatory, cogent norm. 60 Democracies show a relatively higher probability to abide by the rule oflaw and to take efforts to be good international citizens. 61 Democratizing states feel a particular need to posture as norm abiding. Thus, young democracies frequently hasten to document norm obedience by ostentatious nuclear reversals. This tendency seems to be at odds with the finding that young democracies are among the most war-prone political systems. 62 However, this is not necessarily a contradiction. Joining the NPT is a form of good citizenship and is likely to attract the sympathy of established democracies, which is what young democratic states tend to desire for both normative and utilitarian reasons. At the same time, it does not prevent their hostility toward their nondemocratic regional environment and their perceived need to defend against potential aggressors. It would be worthwhile to conduct separate in-depth research on such cases (for example, Georgia).

156

MULLER AND SCHMIDT

Faithful norm observation is also frequent among many nondemocracies. This finding should not be surprising: Abiding by international law and its established norms is, by and large, normal behavior in the international society of states. However, some nondemocracies are more likely to end up in the minority that deviates from this normalcy. Totalitarian states with a power-seeking or paranoid leadership are more likely to breach their obligations openly or clandestinely. That only nondemocratic states started nuclear weapons activities after the NPT entered into force-some while still party to the treaty-underlines this assessment. The force of the nonproliferation norm proved ineffective in democracies in which nuclear weapons had become part and parcel of national identity and where there existed deeply entrenched pronuclear bureaucracies with a vested interest in retaining advanced weapons capabilities. Some of them gained (temporary) norm compliance through the acceptance of their (interim) nuclear weapons status in the NPT. This finding should encourage research that addresses the issue of which domestic constellations, in terms of both institutional checks on the power of nuclear bureaucracies and the framing of national discourses on security, work in favor of nuclear renunciation and disarmament or in the opposite direction. What we can learn from the research on the "antinomies of democratic peace" is the fundamental ambivalence of liberal-democratic ideology with regard to security policy in general and nuclear policies in particular. The direction in which democracies move in their security policies appears to depend greatly on idiosyncratic factors, notably such path-dependent ideational factors as identity, role perception and political culture, and particular constitutional and bureaucratic settings. 63 Politically, the important conclusion is the key role of the NPT. To underestimate its value 64 may lead to deeply flawed and potentially fatal policies. If the nonproliferation norm is eroded by too many instances of member states cheating or the failure of the nuclear weapons states to honor their obligations to disarm, its documented effectiveness may unravel. 65 Given the less than impressive overall impact on proliferation decisions by other variables, one would be inclined to make pessimistic predictions about rampant nuclear proliferation in the absence of a potent nonproliferation norm.

THE LITTLE-KNOWN STORY OF DEPROLIFERATION

157

APPENDIX: BEGINNING AND END OF NUCLEAR WEAPONS ACTIVITIES BY STATES, 1945 TO 2005

Start (cumulative)

Period

Stop (cumulative)

TotalNWA

1945

4

0

1946-1950

2 (6)

1 (1)

5

1951-1955

7 (13)

0 (1)

12

1956-1960

5 (18)

0 (1)

17

1961-1965

3 (21)

1 (2)

19

1966-1970

3 (24)

4 (6)

18

1971-1975

4 (28)

2 (8)

20

1976-1980

2 (30)

2 (10)

20

1981-1985

2 (32)

0 (10)

22

1986-1990

1 (33)

5 (15)

18

1991-1995

3 (36)

9 (24)

12

4

1996-2000

1 (37)

1 (25)

12

2001-2005

0 (37)

1 (26)

11

1945

US, UK, CA, SU

1946-1950

IN, SE (6)

1951-1955

AR, CN, IL, FR, YU, NO, EG (13)

4 CA (1)

5 12

1956-1960

AU, BR, DE, IT, CH (18)

1961-1965

CL, ID, PK (21)

NO (2)

1966-1970

KR, TW, JP (24)

DE, SE, ID, IT (6)

18

1971-1975

IQ, IR, ZA, ES (28)

AU, JP (8)

20

1976-1980

NG,KP(30)

EG, CH (10)

20

1981-1985

RO,LY (32)

1986-1990

DZ (33)

YU, TW, RO, ES, KR (15)

18

1991-1995

KZ, UA, BY (36)

DZ, AR, IQ, CL, KZ, UA, BY, NG, ZA (24)

12

1996-2000

SY (37)

BR(25)

12

LY (26)

11

2001-2005

17

22

Abbreviations:

Argentina Australia Brazil Belarus CA Canada

AR AU BR BY

19

CH Switzerland CL Chile CN China DE Germany

DZ Algeria

158 MULLER AND SCHMIDT

EG ES FR ID IL IN IQ IR IT

Egypt Spain France Indonesia Israel India Iraq Iran

Italy Japan JP KP North Korea KR South Korea KZ Kazakhstan LY Libya 2.

NG NO PK RO SE

su SY TW UA UK

us YU ZA

Nigeria Norway Pakistan Romania Sweden USSR/Russia Syria Taiwan Ukraine United Kingdom United States of America Yugoslavia South Africa

Definitional elements of "nuclear weapons activities": Either of the following activities counts:

Ambiguous activities (nuclear facilities not under safeguards, no clear commitment to renounce nuclear weapons) Serious consideration of weapons program (repeated statements by political or military leaders, studies on utility and feasibility of nuclear weapons acquisition) Weapons program (construction of facilities, studies on weaponization, tests) Nuclear weapons status (actual possession of nuclear explosive devices) Inheritance (nuclear weapons on territory inherited from former empire or occupation forces)

8

WHY DO STATES PROLIFERATE? Quantitative Analysis of the Exploration, Pursuit, and Acquisition of Nuclear Weapons

Philipp C. Bleek

After more than sixty years, during which time thirty countries explored, sixteen pursued, ten acquired, and one relinquished nuclear weapons, answers to this question remain contested. 1 Scholars have advanced a rich menu of explanations, but there remains substantial disagreement about which factors matter and especially about how much they matter relative to one another. The systematic study of the various causes of nuclear weapons proliferation remains at a very early stage. Quantitative approaches are particularly suited to assessing the importance of multiple variables that jointly and probabilistically affect outcomes, precisely the task confronting proliferation analysts. Hence this chapter seeks to complement the rich qualitative literature on the various factors that affect why states proliferate by employing quantitative tools to analyze whether and how much these factors matter relative to one another. Quantitative hazard modeling, a statistical technique for examining risk and its relationship to time, is employed to analyze various theoretically supported variables, using a new proliferation data set created by the author. The following section makes the case for quantitative approaches to studying the causes of proliferation. The subsequent section reviews the modest quantitative literature on this topic. Next, the analytic approach of this chapter is outlined. Results are presented and discussed. Finally, implications for forecasting proliferation are explored and conclusions offered. WHY DO STATES PROLIFERATE?

159

160

PHILIPP C. BLEEK

WHY USE QUANTITATIVE METHODS?

Scholars occasionally bemoan the supposedly rudimentary state of knowledge about the causes of nuclear weapons proliferation. For example, in her recent book on the proliferation impact of regimes' international economic orientations, Etel Solingen approvingly cites Brad Glosserman arguing that a central obstacle to countering nuclear proliferation is that: We still don't know why governments proliferate nuclear weapons. Several explanations have been offered-to provide security, to establish international status, or as a result of internal political and bureaucratic dynamics-but no single explanation convinces. Until we know why governments acquire nuclear weapons, it will be difficult to stop them from doing so. 2 As Glosserman suggests, the proliferation literature contains a variety of often competing, and sometimes complementary, explanations. But he implies that some silver bullet explanation remains to be discovered, some crucial variable that will explain all cases of proliferation. At this point, more than six decades after nuclear weapons appeared on the international scene and after decades of intense scholarly focus, it seems unlikely that so powerful a variable would have been overlooked. What seems more plausible is that the literature identifies most of the variables that explain why states do and do not proliferate but is still at a rather early stage in assessing how important these variables are relative to one another. 3 This is a task to which quantitative analysis is particularly well suited and for which qualitative analysis has some shortcomings. Qualitative analysis of one or a few cases-the predominant analytic mode in the proliferation literature-can shed light on whether a particular variable matters in at least some cases and, if it does, by what mechanisms it operates. Qualitative methods are consequently indispensable to the generation and refinement of theories that can identify key factors that influence proliferation behavior. Further, qualitative methods can be used to test theories that posit either necessary or sufficient conditions. If a dire security environment is posited to be necessary or even sufficient for states to be motivated to pursue nuclear weapons, then one case in which a state pursued nuclear weapons despite not facing a dire security environment, or in which a state faced a dire security environment and could have pursued nuclear weapons but did not, would falsify that theory. If theories posit neither necessary nor sufficient relationships but instead merely probabilistic ones-in other words, if a dire security environment

WHY DO STATES PROLIFERATE?

161

merely makes states more likely to pursue nuclear weapons-then qualitative analysis gains less traction. Further, while qualitative analysis can shed light on whether and by what mechanisms different variables matter, it sheds far less light on how much they matter relative to one another. Perhaps states are far more likely to acquire nuclear weapons if they face a dire security environment and if they have substantial economic resources at their disposal, because developing nuclear weapons is quite costly. Relatively poor states facing dire security environments might be only modestly more likely to proliferate and might instead pursue one or more possible second-choice options, such as a security guarantee from a nuclear-armed ally or accommodation with the source of the security threat. Relatively rich states would still confront high political and technological barriers to proliferation but might make the decision to proliferate more readily than poorer ones. Qualitative analysis of one or a few cases could focus on both these variables, looking for evidence that policy makers were motivated to pursue nuclear weapons by their security environment and, simultaneously, whether and to what degree resource constraints appeared to be a factor. But such analysis would be hard pressed to evaluate how much these two factors mattered relative to one another. As a result, qualitative analysis is less well suited to answering the question of which would have a greater impact on the risk that a state might acquire nuclear weapons, a dramatic increase in its national wealth or a drastic worsening of its security environment. The problem is only compounded if more variables are posited to be relevant to proliferation behavior; in addition to wealth and threat environment, perhaps also membership in the NPT and regime type. Precisely where qualitative approaches gain less traction, quantitative approaches gain more. Quantitative approaches are suited for assessing multiple variables that are jointly and probabilistically related to the dynamic under study. Of course, quantitative approaches also have limitations. Some variables cannot readily be approximated quantitatively. For example, there appears to be a robust norm against the use of nuclear weapons, although whether it rises to the level of a taboo remains open to debate, and some suggest a similar norm restrains states from proliferating. 4 If such a norm affects proliferation behavior, its strength presumably fluctuates over time, and some states are presumably more susceptible to it than others. But quantifying precisely how strong the norm was in a given year or how susceptible a given state was to it poses enormous and perhaps insurmountable challenges.

162

PHILIPP C. BLEEK

Further, quantitative approaches only provide average results across states and time. This may mask variation in individual cases, including outliers that differ substantially. The numerical results quantitative approaches generate can imply a misleading degree of precision while masking crucial underlying assumptions. Finally, the devil is in the details; just as there are more and less sound qualitative methodologies, there are more and less sound quantitative ones. Hence both more quantitatively and more qualitatively oriented analysts have some grounds for the skepticism with which they sometimes approach one another's work. But a more productive approach begins with the recognition that the two approaches complement many of each other's strengths and weaknesses and that the greatest analytic leverage may derive from employing them in concert. Qualitative methods can be used to study individual cases to develop theories about why states proliferate. Both qualitative and quantitative approaches can be used to test those theories. This, in turn, provides fodder for further qualitative analysis to explore nuances and refine theories, and the cycle continues. The virtues of such a multimethod approach make the present divide between qualitatively and quantitatively oriented scholars, particularly in the proliferation realm but also in other political science domains, all the more unfortunate. Scholars need not master each other's approaches; in particular, more qualitatively oriented analysts need not master the tools required to conduct cutting-edge quantitative analysis. But better understanding these approaches, including both their virtues and shortcomings, would yield considerable dividends for the study of proliferation. For their part, quantitative analysts can be far more transparent about the assumptions and decisions underpinning their analysis, as this chapter attempts to do. QUANTITATIVE STUDIES OF THE CAUSES OF PROLIFERATION

Despite quantitative approaches' evident strengths for studying the causes of proliferation, relatively few studies have employed them for this purpose. Two early attempts used rather crude statistical tools by present-day standards and are useful primarily as aids to identifying potentially relevant explanatory variables. 5 More recently, four articles have employed sophisticated statistical techniques to shed light on why states proliferate. 6 Two of these focused on particular explanatory variables-civil and sensitive nuclear assistance from other states-albeit while controlling for other potentially confounding variables and are briefly discussed in the following pages? The other two assessed

WHY DO STATES PROLIFERATE?

163

the relative importance of a broad array of theoretically supported variables and are reviewed in the following discussion at greater length. These articles all represent significant steps forward, although each also has shortcomings related to some combination of dependent variable coding, selection of explanatory variables, and modeling approach. This chapter builds on these articles, adopting the best elements of each and improving on them where feasible. The first recent article assessing a broad set of explanatory variables, Sonali Singh and Christopher Way's "The Correlates of Nuclear Proliferation: A Quantitative Test," published in December 2004, employed quantitative hazard modeling, supplemented by multinomiallogit. 8 The authors code the dependent variable of proliferation into four categories: no activity, exploration, pursuit, and first explosion/assembly of weapons, coding the behavior of 154 states in the years 1945 through 2000. 9 They explain observed proliferation behavior with reference to three broad categories of explanatory variables: technological, external, and domestic. Technological variables include GDP per capita, GDP squared to capture a potential curvilinear relationship where at lower levels increases make proliferation more feasible but at higher levels they contribute little, and an industrial capacity index based on whether a country produces steel domestically and has installed electricity-generation capacity greater than 5,000 megawatts (MW). External variables include participation in one or more enduring rivalries, a five-year moving average of the number of militarized interstate disputes in which a country is involved, and whether the country has a security guarantee from a nuclear-armed ally. Domestic variables include the level of democracy, the degree to which the country is in the process of democratizing, economic interdependence as measured by the trade ratio (the proportion of GDP accounted for by imports and exports), and changes in the level of economic interdependence, analogous to the democratization variable. 10 Singh and Way find strong support for economic, industrial capacity, and threat environment variables. They do not find support for the widely held and intuitive claim that nuclear security guarantees reduce proliferation proclivity. Finally, while they do not find support for Solingen's argument that countries' level of economic integration is a predictor of proliferation, they do conclude that movement toward greater openness is negatively correlated with exploration, although positively correlated with pursuit.U The second recent article assessing a broad array of variables, Dong-Joon Jo and Erik Gartzke's "Determinants of Nuclear Weapons Proliferation," published in February 2007, relied on probit regression, both uncensored and

164 PHILIPP C. BLEEK

censoring the pool of states being analyzed for prior pursuitP They code proliferation behavior into three categories: no activity, active nuclear weapons development program, and nuclear weapons possession, coding states on an annual basis between 1939 and 1992Y (As discussed in the following section, Jo and Gartzke's more conventional regression approach does not allow them to distinguish between initial acquisition and subsequent possession, and because states that acquire nuclear weapons subsequently possess them for many more years, the latter swamps results for the former.) The authors seek to explain observed variation with reference to variables in two broad categories, opportunity and willingness. Instead of GDP per capita, which is not available for most countries back to 1939, they construct an index of national economic capacity based on energy consumption and iron and steel production. They also construct a measure oflatent nuclear weapons production capability based on the presence or absence of seven resources: uranium deposits, metallurgists, chemical engineers, nuclear engineers/physicists/ chemists, electronic/explosive specialists, nitric acid production, and electricity production. Their final opportunity variable is diffusion of knowledge about how to build nuclear weapons, operationalized as a function of time. 14 Their willingness variables include a conventional threat variable that compares the capabilities of all rivals to that of each state, a nuclear threat variable representing one or more rivals armed with nuclear weapons, a nuclear security guarantee variable, diplomatic isolation, domestic unrest, level of democracy, NPT ratification, the proportion of states that have ratified the NPT, and major and regional power status. Gartzke and Jo find, oddly, that latent capability is strongly associated with programs but not with weapons possession, while, more plausibly, economic development is associated with the possession of weapons but not with the initiation of programs. Not surprisingly, they find that conventional threats are strongly associated with both programs and possession. Contra Singh and Way, they find that nuclear security guarantees are negatively associated with weapons possession but have no effect on whether states initiate programs. Diplomatic isolation has no effect on either stage. Strikingly, they conclude that having a nuclear-armed rival makes states less likely both to pursue and to acquire nuclear weapons. Among the domestic variables, democracies are no more or less likely to pursue nuclear weapons but more likely to acquire them. Domestic unrest has no effect in either direction. As for norms, NPT ratification is negatively associated with having a nuclear weapons program,

WHY DO STATES PROLIFERATE?

165

not surprising because many states presumably join the NPT because they have decided not to pursue nuclear weapons-in other words, the causal relationship likely runs in both directions. The NPT systemic effect variable, which quantifies the percentage of states in the international system that have joined the treaty, is never significant; this is not necessarily a damning conclusion for the NPT, because it might very plausibly have an effect that was not tightly correlated with the proportion of states that had joined it at any given time. Both major power and regional power status make states more likely to pursue and to acquire nuclear weapons. Finally, Jo and Gartzke observe various interesting temporal effects. The longer a state does not pursue nuclear weapons, the less likely it is to do so. And the longer states have nuclear weapons programs, the less likely they are to see these through to acquisition (although presumably this only applies past some threshold because all states that acquired nuclear weapons first had programs for some number of years). At a systemic level, states' likelihood of proliferating has increased over time as nuclear technologies have presumably diffused through the international system. The two recent articles focused on civil and sensitive nuclear assistance merit mention. Matthew Kroenig's "Importing the Bomb: Sensitive Nuclear Assistance and Nuclear Proliferation," published in April 2009 in a special issue of the Journal of Conflict Resolution, uses both cursory qualitative analysis of a few cases and quantitative hazard modeling to assess whether sensitive nuclear assistance helps states proliferate. 15 Kroenig explicitly builds on Singh and Way's prior work. He concludes, plausibly, that sensitive nuclear assistance helps recipient states proliferate more readily and more broadly argues for not neglecting supply-side factors by focusing solely on state motivations, as the literature has tended to do in recent years. Matthew Fuhrmann, in "Spreading Temptation: Proliferation and Peaceful Nuclear Cooperation Agreements," published in summer 2009, uses qualitative analysis of a few cases and quantitative probit regression and rare events logit to examine whether receiving civilian nuclear cooperation helps states proliferate. 16 Fuhrmann's quantitative work, like Kroenig's, builds on Singh and Way's. He concludes that receiving civil nuclear cooperation does help states proliferate and therefore that "proliferation will occur as the nuclear renaissance unfolds," although he also notes that most states receiving civil cooperation do not proliferate, and the international community can do much to constrain potential proliferators. 17

166

PHILIPP C. BLEEK

RESEARCH DESIGN

This chapter explicitly builds on prior work, attempting to synthesize the best elements of earlier studies and to improve on them where feasible. The analysis adopts the hazard modeling approach taken by Singh and Way, regarding it as superior to the more conventional regression approach employed by Jo and Gartzke. The coding of the dependent variable of proliferation behavior by both Singh and Way and Jo and Gartzke leaves room for improvement, so this analysis is based on a new data set compiled by the author. Jo and Gartzke employ various theoretically supported explanatory variables that Singh and Way do not; this analysis integrates some of these. Each of these decisions is discussed in turn in the following pages. Hazard Modeling

Hazard models, also called event history, duration, or survival models, are suited to analyzing rare events whose incidence is probabilistically related to multiple and varying risk factors, precisely what is needed for studying proliferation.18 As do Singh and Way, I employ them to model proliferation risk over time, while controlling for, and exploring the importance of, various theoretically supported explanatory variables. Jo and Gartzke's more conventional regression approach is less suitable to the task at hand. Unlike the hazard approach, in Jo and Gartzke's models states do not exit the risk pool once they have engaged in a given level of proliferation activity. For example, once states have acquired nuclear weapons, every year they retain them is counted as though they had reacquired them that year. This has two effects. First, the results help explain why states possess nuclear weapons far more than why they acquire them. But the reasons that states retain nuclear weapons are presumably substantially different than those that motivate their acquisition; nuclear weapons are likely "sticky" or path dependent, as the fact that only one state that developed them subsequently gave them up suggests. Second, it means that states that possessed nuclear weapons for a long time exert a far greater effect on the conclusions reached than states that acquired them more recently. 19 On the other hand, Jo and Gartzke make a very useful contribution by censoring their data, so that only states with nuclear weapons programs are potential candidates for weapons acquisition. Censored analysis addresses the question of what differentiates states with nuclear weapons programs that acquire weapons from those with programs that are not seen through to acquisition. Uncensored analysis asks what makes states that acquire nuclear weapons

WHY DO STATES PROLIFERATE?

167

different from all those that do not. Censoring is useful but need not invalidate uncensored analysis, both because each method addresses somewhat different questions and because various stages of proliferation can be thought of as various operationalizations of the dynamic under study. Of the three proliferation stages analyzed here-exploration, pursuit, and acquisition-we can be most confident about the coding of acquisition, but the number of states that acquired nuclear weapons is modest, making it harder to discern correlations. At the other extreme, we can be least confident about the coding of exploration, but far more states engaged in this level of behavior, so correlations can be more readily discerned. And while the case universe is large enough to support censoring for Jo and Gartzke's more conventional regression approach, that is not the case for the more demanding hazard analytic approach employed here. 20 One brief methodological aside: Hazard analysis requires the analyst to decide whether to make assumptions about the hazard function, that is, about the character of risk over time. When analyzing time-varying covariatesfor example, GDP per capita, which fluctuates annually-two approaches are possible. One of these, semiparametric or Cox regression, does not require the analyst to make assumptions about the hazard function. The other, parametric, requires the hazard function to be specified in advance. The advantage of the latter is that, assuming the assumptions are correct, it yields more precise estimates. 21 The smaller the data set, the less precise Cox estimates will be relative to parametric estimates. 22 When employing parametric approaches, analysts compare the fit of models using various distributions to select the distribution that best fits the data. Following Singh and Way, I employ parametric models with a Weibull distributionY The Weibull is extremely flexible and can mimic the functional forms of a variety of other probability distributions. As a robustness check, I also tested the model employed here using the Cox-regression approach. The results I obtained, which I do not report, were broadly consistent with those derived using the Weibull distribution, although as expected they were also broadly less robust. New Data Set of Proliferation Behavior

Quantitative analysis depends on consistent, empirically supported coding of individual cases. Both Singh and Way's and Jo and Gartzke's data sets rely on too few sources or sources of suspect reliability and are not fully transparent about why they made the coding decisions they did, although Jo and Gartzke are considerably more transparent than Singh and Way. 24

168 PHILIPP C. BLEEK

The analysis in this article is based on a new data set of the proliferation behavior of states in the international system between 1945 and 2000. The data set builds on Singh and Way's and adopts their time frame. Singh and Way presumably began their analysis in 1945 because they surmised that proliferation dynamics were different in a world in which the possibility of nuclear weapons had been unequivocally demonstrated, as it was by the United States that year. There is a case to be made for extending the analysis back farther; several countries initiated nuclear weapons programs prior to 1945; and, at least among certain scientists, nuclear fission explosives were judged feasible well before their first demonstration. But extending would require data collection on and consistent coding of all the explanatory variables to be retained, and this analyst regards keeping the 1945 time frame as the lesser of two analytic evils. This chapter employs multiple, high-quality sources for each coding decision. The results are detailed in Table 8.1, and a lengthy coding appendix that discusses each decision is available on request from the author and will be published as part of a forthcoming paper. 25 Following Singh and Way, I code states in each year as engaging in one of four levels of nuclear weapons proliferation activity: no discernable activity, exploration, pursuit, and acquisition. For exploration, I follow Singh and Way's coding approach, although I code a number of cases differently. As do they, I look for evidence of either explicit political authorization to explore (but not pursue) a nuclear weapons option or formal linking of atomic research to defense agencies. Further, in some cases states appear to have engaged in proliferation-relevant activity but not with sufficient intensity to merit coding pursuit, and these too earn an exploration coding. Exploration is both the least clearly defined and hardest to ascertain category of proliferation behavior, and perhaps for this reason Jo and Gartzke chose not to include a similar category in their analysis. I regard this variable as useful because a considerable population of states appears to have been modestly interested in nuclear weapons and engaged in exploration but not overt pursuit. Because each stage of proliferation behavior is modeled separately the inclusion of the exploration variable does not affect results for either pursuit or acquisition. In other words, at worst the results for exploration can be discounted, and at best they shed some additional light on the proliferation puzzle. On pursuit, I judge Singh and Way's and Jo and Gartzke's definitions to be essentially analogous. As do both sets of authors, I look for evidence that heads of state authorized the pursuit of nuclear weapons with the goal of

WHY DO STATES PROLIFERATE?

Table 8.1.

Proliferation behavior over time.

Country

United States Russia United Kingdom France China Israel South Africa Pakistan India North Korea Yugoslavia Brazil South Korea Libya Iran Iraq Germany Germany, West Japan Sweden Switzerland Norway Egypt Italy Australia Indonesia Taiwan Argentina Romania Algeria

169

Explore

1939194219401945195619491969-1991 1972194819621949-1962, 1974-1987 1966-1990 1970-1975 1970-2003 1974-1979, 19841976-1991 1939-1945 1957-1958 1941-1945, 1967-1970 1945-1970 1945-1969 1947-1962 1955-1980 1955-1958 1956-1973 1964-1967 1967-1976, 1987-1988 1978-1990 1978-1989 1983-1991

Pursue

1942194319411954195619551974-1991 19721964-1966, 1972-1975, 19801980-

Acquire

1945194919521960196419671979-1991 198719872006-

1953-1962,1982-1987 1975-1990 1970-1975 1970-2003 19891976-1991

NOTE: A coding appendix (42 pp.) is available on request and forthcoming as part of an occasional paper to be published by the james Martin Center for Nonproliferation Studies, Monterey Institute of International Studies, with the working title, "When Did (and Didn't) States Proliferate? Coding the Spread of Nuclear Weapons throughout the Atomic Age." Coding is based on multiple, high-quality secondary sources for every case listed above, and a number of others were deemed as not meriting inclusion on the basis of those sources.

acquiring them; but, acknowledging that such information is not always available, I also consider circumstantial evidence. Finally, I judge acquisition to have occurred in the year in which states first had access to deliverable nuclear explosive devices. For most states, this corresponds to the year of their first nuclear explosive test, but one state tested without, at that time, acquiring

170 PHILIPP C. BLEEK

nuclear weapons (India), and several acquired without or prior to conducting tests (South Africa, Israel, Pakistan). 26 The secondary literature is sufficiently rich to allow high confidence in the majority of coding decisions. Where alternate codings are also plausible, these are noted in the coding appendix. It bears emphasizing that, while coding for quantitative analysis can capture much of the proliferation behavior of states, much is also necessarily beyond its purview. For example, states may engage in various degrees of hedging that is difficult to reliably code. 27 Many analysts regard Japan's civil nuclear power infrastructure, including the accumulation of substantial plutonium stocks from reprocessed power reactor fuel, as a form of nuclear weapons hedging. This analyst believes that it is nonetheless important to ask why some states overtly pursue or acquire nuclear weapons while others do not, even if some of the latter engage in some degree of hedging. The bottom line is that quantitative analysis can capture much but unequivocally not all of states' proliferation behavior and therefore can make useful contributions as long as analysts are transparent about its limitations. Explanatory Variables

Scholars have advanced a rich menu of explanations for proliferation, but there remains substantial disagreement about which factors matter and especially about how much they matter relative to one another. This analysis therefore attempts to test a broad array of theoretically supported variables. Most extant proliferation theories can be operationalized and tested quantitatively, even if such operationalizations are never perfect. But some arguments do not lend themselves to quantitative testing. 28 Scholars have categorized proliferation explanations in various ways. Scott Sagan's typology of security, status, and domestic politics is widely cited, but it focuses only on states' motivations for attempting to acquire nuclear weapons and ignores the capabilities they may or may not have to realize these. 29 Singh and Way categorize their variables into technological, external, and domestic determinants. 30 Jo and Gartzke categorize their variables into opportunity and willingness, and they break the latter down into international security, domestic politics, norms, and status. 31 Although both these broad categories are plausible and useful, there is significant overlap between variables in each of them. For example, Singh and Way categorize Solingen's arguments about the impact of ruling coalitions' international economic orientations as domestic, but the posited mechanism is at least as much international. And Jo and Gartzke put their NPT variables into the willingness category; while the NPT

WHY DO STATES PROLIFERATE?

171

almost certainly affects proliferation motivations, it also affects states' opportunities to realize their proliferation motivations, for example via its monitoring and inspection provisions. I categorize proliferation explanations into five categories: economic resources and technological capacity; security motivations; institutional constraints; status motivations; and domestic politics. 32 Each is briefly summarized in the following discussion, although this is not intended as a comprehensive literature review. Earlier in the nuclear age, there was substantial focus on states' capabilities, but in recent years scholarly attention has shifted primarily to motivational factors. 33 Analysts often treat financial and industrial capabilities as a threshold issue, and because several less affluent and developed countries successfully acquired nuclear weapons, assume the threshold is relatively low. This analysis assumes that there is instead a capability continuum, with poorer and less-developed states facing a higher bar for proliferation while still being capable of doing so with sufficient motivation, whereas richer and more highly developed states may more readily proliferate, given external motivation. In other words, Pakistan was willing to "eat grass," as Zulfikar Ali Bhutto famously observed, and was therefore able to acquire nuclear weapons, while Japan, should it decide it required nuclear weapons, would be able to do so much more readily. 34 Security motivations are the most widely cited proliferation motivation. 35 These include both conventional insecurity, as a function of the conventional military threats states face or the wars and lesser conflicts in which they are engaged, and nuclear insecurity, as a function of proliferation by or nuclear coercion at the hands of rival states. In recent years, some analysts have argued that the importance of security motivations is exaggerated, and it is occasionally even suggested that they offer little traction on the proliferation puzzle. 36 This argument is sometimes made by highlighting the fact that some specific state not deemed to have been facing dire security threats chose to proliferate or that some state facing dire security threats chose not to do so. But the relationship between security and proliferation is almost certainly probabilistic rather than deterministic-a dire security environment may make states more likely to proliferate.

Institutional factors are widely assumed to affect proliferation behavior, although the scholarly literature on these is more modest. 37 The NPT is the centerpiece of the institutional architecture relevant to proliferation. The treaty

172

PHILIPP C. BLEEK

codifies and reinforces the norm against proliferation. It also serves a signaling function, allowing states to more credibly signal to others that they do not intend to pursue nuclear weapons. And it plays a functional role, making it harder for states to covertly pursue nuclear weapons in the face of monitoring and verification and easier for other states to act in concert when possible proliferation is discovered. Status motivations are often cited as contributing to proliferation proclivity.38 States may pursue weapons because they desire the additional international status their possession entails or because they view possession as consistent with their perceived status. Conversely, states may not pursue nuclear weapons because of the stigma associated with doing so. Finally, domestic politics may affect proliferation. 39 There is a substantial international relations literature suggesting that democracies are less likely to engage in armed conflict against each other, and some have suggested they may also be less likely to acquire nuclear weapons. 40 Some additional literature suggests that states in the process of democratizing are more warlike, and it is plausible that they may be similarly more likely to proliferate. 41 Solingen has suggested that ruling coalitions' orientation toward the international economy affects proliferation proclivity because states that are highly integrated with the international economy have more to lose if they proliferate and also reap lesser benefits from doing so, while less integrated states have less at stake and more to gain from proliferating. 42 Also falling into this general category are Jacques Hymans's arguments about the psychological profiles of national leaders and Peter Lavoy's arguments about domestic "mythmakers."43 Singh and Way employ a relatively comprehensive set of explanatory variables, and this analysis imports many of them. One ofJo and Gartzke's innovations is to introduce a number of additional variables that are theoretically compelling, and several of these are also integrated. 44 Variables that capture Kroenig and Fuhrmann's arguments about sensitive and civil assistance are tested, though ultimately not reported, as explained in the following discussion. Each of these studies fails to capture some key dynamics, or does so in ways that leave room for improvement, so a few variables are also created for the purposes of this study. Variables described in the following discussion are drawn from Singh and Way unless otherwise noted. These variables are employed as indicators of both economic resources and industrial development potentially relevant to proliferation. The GDP-squared term is intended

Gross Domestic Product per Capita and GDP Squared

WHY DO STATES PROLIFERATE?

173

to capture a potential curvilinear relationship, where additional units of GDP at the lower end of the spectrum have a significant effect on proliferation; but that effect drops off and even reverses at the higher end of the spectrum. Industrial Capacity To examine industrial resources available for a nuclear program but not fully captured by the GDP variables, this dichotomous variable takes on a value of one if a country produces steel domestically and has installed electricity-generating capacity greater than s,ooo MW, and zero otherwise.45 Jo and Gartzke take a different approach, creating a latent nuclear weapons production capability indicator, based on the presence of seven factors listed earlier. 46 Although ingenious, this appears to be substantially endogenous to the decision to pursue nuclear weapons, and it overlaps directly with some of the authors' criteria for coding the dependent variable of a weapons program; I do not employ it here. Finally, I imported from Kroenig variables for both sensitive and civil nuclear assistance, to test Kroenig's argument about the former and Fuhrmann's arguments about the latter, both detailed in the preceding discussion. 47 Note that, while Fuhrmann includes all nuclear cooperation agreements in his analysis, many such agreements were never executed and therefore should have had no impact on proliferation, whereas the variable I import from Kroenig includes only technology transfers that actually took place. 48 Neither sensitive nor civil nuclear assistance is significant in any of my models (with one plausible exception detailed in the endnote), but they exert a confounding effect on the results for other variables, and therefore I do not report results for them in the following discussion. 49 Dispute Intensity To examine the effect of states' conventional security environments, I create a variable that is a moving average of the cumulative intensity of the militarized interstate disputes (MIDs) in which a state has been involved in the prior five years. MIDs are threats or outright uses of force between states and are coded on a five-point scale from mere threats to full-scale war. 5° Singh and Way create a similar variable but sum the number of MIDs rather than their intensities, treating low-level threats as equivalent to full-scale wars. Singh and Way also examine whether states are engaged in one or more enduring rivalries, but this variable is never significant in my models, and I do not report results for it. 51 Jo and Gartzke create a more complex threat variable in which they sum aggregate national capacity indicators for all of each state's enduring rivals and divide these by the state's own capacity; I do not obtain significant results from this variable and do not report results for it. 52

174 PHILIPP C. BLEEK

Rival Nuclear Weapons Program To examine possible "reactive proliferation" in response to proliferation by security rivals, I create a dichotomous variable equal to one if states have one or more enduring rivals pursuing (or having acquired) nuclear weapons and zero otherwise. Enduring rivalries are the most intense rivalries in the international system, and the majority of wars and militarized disputes occur between enduring rivals. 53 I choose pursuit because I judge this to be the stage of proliferation behavior most likely to trigger a response; exploration is too far removed from acquisition, while, if states have not responded to a serious weapons program, I judge them as less likely to do so if and when their rival finally acquires (although it bears emphasizing that focusing only on rival acquisition, rather than pursuit and acquisition, yields broadly similar results). 54 Nuclear Security Guarantee To explore the possible proliferation-ameliorating effect of a security guarantee from a nuclear-armed ally, I create a dummy variable equal to one if states have a formal defense pact with a nuclear-armed ally and zero otherwise. 55 Defense pact is the highest of three levels of alliance; the other two, which I deem insufficient to constitute a robust security guarantee, are ententes and neutrality treaties. 56 I count all nuclear-armed states as capable of extending security guarantees, although in practice the vast majority were extended by one of the two Cold War superpowers and none by non great powers. An alternate variable, focusing only on superpower security guarantees, yields almost identical results and is not reported here. Democracy and Democratization To test whether regime type influences proliferation, these variables measure the level of democracy in each state in each year by integrating rankings for both democracy and autocracy and movement toward (or away from) democracy over a five-year time span. Singh and Way also employ a variable based on the percentage of democracies in the international system, but I judge it to have little theoretical support and do not include it. Economic Interdependence and Liberalization Two variables attempt to get traction on Solingen's argument that domestic political coalitions seeking to integrate with the international economy will face higher costs and fewer benefits from proliferating, while inward-oriented coalitions will face lower costs and greater benefits. 57 Most straightforward is the trade ratio, imports and exports as a percentage of GDP. To capture the impact of change, analogous to

WHY DO STATES PROLIFERATE?

175

arguments about democratization, the change in the trade ratio over five-year spans is also included. Major Power and Regional Power To explore whether status has an impact on proliferation, I import two variables from Jo and Gartzke. The first is equal to one if a state is a major power, using the Correlates of War classifications, and zero otherwise. 58 To preempt one possible concern that is problematic for Jo and Gartzke's analysis but not for this chapter, there is not a significant selection problem in the fact that countries are categorized as major powers in part because they have acquired nuclear weapons because the hazard modeling approach employed here considers risk factors only up until states reach the proliferation threshold being examined, at which point they are dropped from the risk pool. The second variable is equal to one if a state is a regional power and zero otherwise. 59 Regional powers are all states that are not major powers but are at least half as powerful as the most powerful state in their regions, as measured by the aggregate capacity indicators discussed in the context of the dispute intensity variable above. I also attempt to examine the effects of diplomatic isolation, importing from Jo and Gartzke a variable that captures the number of states with which a state lacks diplomatic relations relative to the total number of neighboring states and major powers. 60 This variable never has a statistically significant relationship to any stage of proliferation in my models, but exerts a substantial and hard-to-explain effect on the results for a variety of other variables, and I do not report results for it. NPT Ratification and NPT Systemic Effects To examine whether individual membership in and the aggregate strength of the NPT constrain proliferators, I import two variables from Jo and Gartzke. The first is a dichotomous variable equal to one if a state has ratified the NPT and zero otherwise. Because the decision to ratify the treaty is presumably substantially endogenous to the decision not to proliferate-in other words, many states ratify the treaty because they do not intend to try to obtain nuclear weapons-this variable must be interpreted with care, but meaningful inferences are still possible, especially if the variable yields results other than the expected negative relationship. The second variable is equal to the proportion of states worldwide that have ratified the NPT. The intuition is that the more states that have ratified, the more robust the nonproliferation norm the treaty formalizes. This is at best an imperfect operationalization. The norm the NPT codifies may not be a direct function of the number of members but instead a threshold function

176

PHILIPP C. BLEEK

dependent on the behavior of some proportion of states or some subset of highly respected states. And the treaty's effects are not merely normative. It also operates instrumentally-for example, nonnuclear NPT members commit to reporting civil nuclear activities and submit to periodic inspectionsand these instrumental functions may have little to do with the proportion of states that have joined. The treaty may also operate as a signaling mechanism, perhaps allowing regional antagonists to more credibly signal that they do not intend to proliferate, and this function, too, may have little to do with the proportion of member states. All these caveats notwithstanding, the variable may still shed some light on at least some aspects of the treaty. RESULTS

This section outlines and explores results obtained from running hazard models for nuclear weapons exploration, pursuit, and acquisition, employing the new data set of proliferation behavior and explanatory variables detailed in the preceding section. Before delving into specific conclusions, a brief comment on correlation and causation is in order. The statistical models employed here highlight correlations, and these are not necessarily tantamount to causal relationships. For example, two variables may be correlated, but the causal arrow may be reversed. Alternately, two correlated variables may not be directly causally related but instead related to some third variable. There is no magic bullet solution for discerning causation when conducting quantitative analysis; the onus is on analysts to consider plausible alternate explanations for the patterns they find and to test these where feasible. To revisit a theme highlighted earlier in this chapter, this is yet another area where quantitative and qualitative analysis can complement one another. Qualitative process tracing, which examines the specific mechanisms by which policy outcomes manifest themselves, is particularly suited to assessing causation. 61 The bottom line is that the correlations highlighted below should be considered highly suggestive of potential causal relationships but are perhaps closer to starting points for further analysis than definitive conclusions about causation. A. What Factors Affect States' Likelihood of Proliferating?

Table 8.2 presents estimates for hazard models for three stages of proliferation: exploration of a nuclear weapons option, pursuit of weapons (that is, an active program to acquire them), and acquisition of deliverable nuclear weapons. 62 Note that, unlike with more conventional regression analysis, the results reported in Table 8.2 allow us to assess only the statistical significance

Table 8.2.

Why do (and don't) states proliferate? Quantitative hazard modeling results. Dependent variable (proliferation) Explore

Pursue

Acquire

Economic resources and technical capabilities .0007" 018 (.0003)

.002'-" 01 (.0006)

.014·007 (.005)

-5.08e-08·0' 7 (2.13e-08)

-2.03e-07 205,206,215,242n11,246n4s,268n3: and neoclassical realism, 61-69; nuclear ambivalence, 97-98; and nuclear latency, So, 81, 96-97, 99; and nuclear testing, 119; outward-looking regimes, 9, 40,

283

41-44,45-48, 50-57, 172, 174, 181, 197, 215, 243n26, 268n3; role in nuclear reversal, 129, 130, 138, 139-41, 148-54, 155-56, 200, 219; Solingen on, 6, 7, 9, 28-31,36,138,139-40,141,172,174, 181-82, 187-88,241n7,242n11,268n3. See also democracy/democratization; prestige motivations; state, the; status motivations domino thesis. See proliferation chains Dougherty, James E.: Contending Theories of International Relations, 250n34 Dr. Strange love, 108 dual-key arrangements, 74, 209-11, 223 dual-use research, 16 Dueck, Colin: Reluctant Crusaders, 249n27 Dunn, Lewis A.: Controlling the Bomb, 241n3, 263n4; on nonproliferation regime, 56; on proliferation chains, 3, 12; Trends in Nuclear Proliferation, 3, 12, 276n4 Du Preez, Jean, 265n34 Duval, Marcel: Histoire des forces nucleaires franfaises depuis 1945, 262n72 East Asia: financial crisis of 1997, 54; financial crisis of 2008-2009, 54; vs. Middle East, 41-42,43,45,52-54,55, 56;Solingen on, 100 East Germany, 91, 92 Eckstein, Harry, 245n38 economic liberalization, 189, 190, 191, 197, 273n62; of outward-looking regimes, 9, 40, 41-44, 45-48, 50-57> 172, 174> 181, 197, 215, 243n26, 268n3; role in nuclear reversal, 127, 140, 153-54, 242m5; Solingen on, 9, 153-54, 163, 170, 172, 174, 181-82, 187-88, 274n68; trade ratio, 163, 174-75, 17h 181,183,185, 243n15 economic resources, 161, 171, 193, 197, 275n7o; economic pull hypothesis, 127, 141-43; GDP, 141-42, 153, 154, 163, 164, 167, 172-73, 177, 178, 181, 182, 183, 185, 186, 189, 190, 191-92, 273n62; industrial capacity, 173, 177, 178, 181, 182, 183, 186, 189, 190, 191-92, 272nn45>46, 274n69. See also technological capacity Egypt, 38, 243n26; economic conditions in, 141, 142; nationalism in, so; nuclear latency of, 88, 91, 92; nuclear weapons program of, so, 97, 131, 135, 136, 141, 150, 157, 158, 169, 254n21, 271n24; political conditions in, 41, 43, 49-51, 53, 150;

284

INDEX

Egypt (Continued) relations with Iran, 50-51, 208; relations with Israel, 49, so, 51 Einhorn, Robert J., 49, 264n8 ElBaradei, Mohamed: on virtual nuclear weapons states, 82-83, 89, 100, 102, 104, 257n2 Elman, Colin, 247m2, 248n2o, 249n25 Elman, Miriam Fendius, 249n25 enrichment facilities, 45, 82-83, 91, 93, 94, 95.96,98-99.100-101,110,113,216, 219, 243n26, 25snn30,31; see also fissile material environmental movement, 27, 34, 82 Epstein, William, 8, 13-14; The Last Chance, 15-18; on NPT, 15-18 European Union (EU), 215; relations with Iran, 201, 202, 204, 207; relations with Russia, 207; relations with Turkey, 53,216 Evan, Gareth, 38

France: economic conditions in, 191; nuclear weapons program of, 15, 32, 120, 122, 134> 136, 150, 151, 157> 158, 169, 184, 189, 191, 198-99. 214, 215, 231ll5, 244n29, 263n3; political conditions in, 150; relations with Italy, 132, 135; relations with Soviet Union, 137; relations with United States, 145, 215; relations with West Germany, 132, 135, 276n76; Suez War, 137 Frankel, Benjamin, 2, 9, 14-15, 25, 249n29, 271n35; Opaque Nuclear Proliferation, 22-24; on opaque proliferation, 22-24, 28, 111; on prestige, 233n22; The Proliferation Puzzle, 263n4 Friedberg, Aaron L., 66, 68, 251n44; The Weary Titan, 248n23 Froscher, Torrey, 258n14, 262n62 Fuerth, Leon, 246n47 Fuhrmann, Matthew, 98-99, 243m6; on civil and sensitive assistance, 165, 172, 173

Fearon, James D., 244n31 Feaver, Peter, 233n21 Feffer, John, 83 Feiveson, Harold A.: Nuclear Proliferation, Motivations, Capabilities and Strategies for Control, 263n4 financial crisis of 2008-2009, 54, 55 Finnemore, Martha, 234n29 first strike, 73, 117, 118 Fischer, David: Nuclear Non-Proliferation and Global Order, 266n48 fissile material: accumulation of, 10, 19, B1,82,84,ss,ss,9o-96,98, 102-5,

Gaddis, John Lewis, 247nn2-4 Ganguly, Sumit, 151 Gartzke, Erik, 267nso, 269n6; on initiation of weapons programs, 254n21; on NPT, 135-37, 147-48, 170-71, 268n64; on nuclear diffusion, 88-89, 135-36; on nuclear latency, 84, 87-88, 89, 164, 173; quantitative analysis of nuclear proliferation, 21, 130-31, 133-38, 163-65, 166-6A 168-69,170,172,173,175,179, 235n4s, 270nn12-14,24, 271nn32,46, 273nss; on regime type, 272n40 Gay, Corey, 275n72 Gellman, Barton: Angler: The Cheney Vice Presidency, 26on42 geography and proliferation, 65, 67, 73, 74 George, Alexander L.: Case Studies and Theory Development in the Social Sciences, 245n41, 273n61 Georgia, 155, 207, 211-12 Germany. See East Germany; Nazi Germany; West Germany Gibler, Douglas M., 273ns6 Gilpatric Committee Report, 1 Gilpin, Robert, 249n29 globalization, 31, 40, 55 Glosserman, Brad, 160 Gold blat, Jozef: Nonproliferation, 263n4 Gorton, John, 140 Gourevitch, Peter, 234n27

109, 110, 111-16, 121, 170, 216, 221,

259n26; control regime, 205; enrichment facilities, 45, 82-83, 91, 93, 94, 95, 96, 98-99,100-101,110,113,216,219, 243n26, 255nn30,31; highly enriched uranium (HEU), 17, 81, 82, 86, 88, 90, 93, 94, 96, 113, 221, 255n31; nuclear fuel cycle, So, 81, 82, 83, 90-96, 99, 100-101; plutonium, 81, 82, 86, 88, 90, 94, 96, 170, 216, 255n31; reprocessing facilities, 82-83,91,92,94.95.96,98-99,100-101, 110, 170, 243n26, 255n3o; significant quantity (SQ) as indicator of nuclear weapons stateness, 102-5, 109, 110, 111-16, 117, 121, 257n2, 259n26 Flank, Steven, 25 Ford/MITRE study, 255n34

INDEX

Gowing, Margaret: Independence and Deterrence, 262n71 Gray, Colin, 25 Greenpeace, 82 Greenwood, Ted: Nuclear Proliferation,

Motivations, Capabilities and Strategies for Control, 263n4 Grieco, Joseph M., 249n29 Grillot, Suzette R., 2651127 Gulf War of 1991, 97 Gurr, Ted Robert, 64, 251n38 Gusterson, Hugh, 2581115 Harney, Robert: "Anatomy of a Project to Produce a First Nuclear Weapon," 93-94, 256n}9 Hayes, Peter, 114 Hersman, Rebecca K., 264n8 Hibbs, Mark, 265n34 Hintze, Otto, 252n72 Hiroshima, 108, 122, 168, 188, 238n39 Hohenemser, Christopher, 232n7 Holgate, Laura, 113-14 Holloway, David: Stalin and the Bomb, 262n70 Holsti, Ole R., 2491129 Horowitz, Michael, 267n50 House, Karen Elliott, 258n9 Hudson Institute Study: Trends in Nuclear Proliferation, 1975-1995, 195 Husbands, Jo. L., 242n11 Hussein, Saddam, 41, 97 Hymans, Jacqnes E. C., 258n11, 259n23, 2611144, 263n5, 266n37; criticisms of, 32-33; on domestic politics, 6, 7, 138, 139, 140-41, 172; on national identity conception, 31-33, 234n34, 239n74; on North Korea, 96-97; on nuclear abstinence, 31, 36, 86, 138; on nuclear latency, 84-85; on opaque proliferation, 26on32; on oppositional nationalist leaders, 7, 31-32, 33, 36, 140-41; The

Psychology of Nuclear Proliferation, 6, 7, 31-33.36,84-85,253n8,264n8,265n30, 268n3, 271111128,36; on regime type, 96-97; on Romania, 96-97; on security concerns, 138, 2711136; on uncertainty, }1, 36 hypotheses: economic pull hypothesis, 127, 141-43; "motivational" hypotheses, 18-22, 23; "technological imperative" hypotheses, 18-22, 23; technological

285

pull hypothesis, 127, 143-44, 154; testing o£7,8, 10,18-21,25,32,33,37,49-54, 116-21, 127-28, 140, 141-54> 160-61, 162, 170 IAEA. See International Atomic Energy Agency Ikenberry, G. John, 2481119, 249n3o, 2511148 India, 1-2, 19, 46; Abraham on, 33-35; Bharatiya Janata Party (BJP), 151; economic conditions in, 153, 154, 2441129; nationalism in, 31; and NPT, 106-7; nuclear test of 1974, 5, 15, 17, 24, 34, 35, 106, 122, 170, 244n29; nuclear test of 1998, 31, 34, 35, 107, 122, 151; nuclear weapons program of, 23, 26, 32, 100, 112, 120, 122, 132, 134, 1}6, 139, 140-41, 150, 151, 153, 154, 157, 158, 169, 189, 191, 203, 205, 244n29, 263n3; political conditions in, 31, 32, 33-34, 47, 150, 151; relations with China, 137, 151, 204; relations with Pakistan, 137, 151, 203, 205 Indonesia: nuclear latency of, 91, 92; nuclear weapons program of, 135, 136, 157, 158, 169, 224, 254n21, 271n24; political conditions in, 43, 54, 150 Intermediate and Shorter-Range Nuclear Forces (INF) Treaty, 212 Intermediate-Range Ballistic Missiles (IRBMs), 74 International Atomic Energy Agency (IAEA), 16, 24, 56, 58, 132, 213, 216, 259n26, 266n43; Additional Protocol, 45, 2431126; E!Baradei on virtual nuclear weapons states, 82-83, 89, 100, 102, 104, 257n2; misleading of, 41, 45, 207, 209; and NGOs, 27-28; and North Korea, 86-87, 88; and nuclear latency, 81, 82-83, 89, 102; and South Africa, 88, 221 International Commission on Nuclear Nonproliferation and Disarmament, 38 Iran: economic conditions in, 201, 244n27; Iranian Revolutionary Guard Corps (IRGC), 210; and Middle East, 38, 199, 206, 207, 208, 209, 212-1}, 221, 225; and NPT, 53-54, 56, 93, 2oo-2o1, 207, 226; nuclear latency of, 91, 92; nuclear weapons program of, 10, 43, 45, 47, 50-51, 53-54, 56, so, 121, 131, 135, 136, 139, 150, 152, 157. 158, 169, 179. 195, 199-202, 203-4> 205,207-8, 210, 211, 212-13, 216,218,225,226,242n11, 270n24;

286

INDEX

Iran (Continued) political conditions in, 43, 47, 53, 150, 152, 201-2; and proliferation chains, 38, 199, 206,20h208,209,212-13,221,225; relations with China, 200, 204; relations with Egypt, 50-51; relations with Iraq, 41; relations with Israel, 41, 200, 201, 203, 204, 205, 207, 210; relations with Russia, 200, 202, 204, 207; relations with Syria, 210; relations with Turkey, 53, 208, 216; relations with United Nations, 54, 201, 243n26; relations with United States, 137-38,139,199,200,201,202,203,204, 207-8 Iraq: economic conditions in, 244n27; Gulf War of 1991, 97; nuclear weapons program of, 17, 23, 24, 45> so, 56, 97. 113, 131, 132, 135, 136, 139, 141, 150, 152, 157, 158, 169, 196, 200, 209, 210; political conditions in, 150, 152; relations with Iran, 41; relations with Israel, 209, 210; relations with United States, 97, 104, 113, 139, 200, 207, 209; war of 2003, 113 Ireland: as nuclear abolitionist, 196 Israel: as case of opaque proliferation, 22, 23, 24, 110-11, 112, 120, 205, 259n32, 26onn36,37,39; economic conditions in, 141-42, 183, 191, 274n69; and NPT, 207; nuclear latency of, 19, 91, 92; nuclear weapons program of, 46, 49, so, 110-11, 120, 122, 134, 136, 137> 139, 141-42, 150, 151, 15h 158,169,170,183,189,190, 191, 199, 204, 205, 207, 259nn27-29, 260n36, 262n66, 263n3, 274n69; political conditions in, 22, 47, 112, 150; relations with Egypt, 49, so, 51; relations with Iran,41,200,201,203,204, 205,20h210; relations with Iraq, 209, 210; relations with Libya, 209, 210; relations with Palestinians, 203; relations with South Africa, 120, 122, 259n28, 271n26; relations with Syria, 190, 197, 200, 209, 210, 263n3; relations with United States, 22,120,139,217

Italy: nuclear latency of, 91, 92; nuclear weapons program of, 131, 134, 136, 150, 151, 157, 158, 169, 254n21, 271n24; political conditions in, 150; relations with West Germany, 131, 132, 135, 276n76 japan: civilian nuclear power in, 52, 170; economic conditions in, 189, 192; nuclear latency of, 1, 2-3, 52, 91, 92, 98,

103,115,155,171,185,188,189, 231n4, 243m8, 245n44, 254n15, 256n53; nuclear weapons activities of, 135, 136, 150, 151, 157, 158, 169, 270n24, 271n24, 275n7s: political conditions in, 43, 45, 52, 54, 150, 155, 188, 246n46; relations with Australia, 52; relations with Canada, 52; relations with China, 52, 54, 189; relations with France, 52; relations with North Korea, 206, 216-17; relations with South Korea, 54; relations with Soviet Union, 189; relations with United Kingdom, 52; relations with United States, 29, 52, 139, 180, 189-90, 192, 196, 214-15, 216-17, 274n65; during World War II, 134, 169, 254n15,270n24, 275n75 jervis, Robert, 67,2481124, 251n56, 252n6o, 2581113, 269n4 jo, Dong-Joon, 267nso; on initiation of weapons programs, 254n21; on NPT, 135-37,147-48, 170-71,268n64;on nuclear diffusion, 88-89, 135-36; on nuclear latency, 84, 87-88, 89, 164, 173; quantitative analysis of nuclear proliferation, 21, 130-31, 133-38, 163-65, 166-6h 168-69, 170, 172, 173, 175, 179, 235n45> 270nn12-14,24, 271nn32,46, 273n55; on regime type, 272n4o jones, Bradford S., 27onm8,22 Jones, Daniel M., 273n30 Jones, Gregory, 109 )onter, Thomas, 133 jordan, 189; economic conditions in, 190-91; political conditions in, 44, 45, 243n26; relations with Israel, 191 Kahler, Miles, 2491126 Kahn, Herman: Trends in Nuclear Proliferation, 3, 12, 276n4 Kaneko, Kumao, 52 Kang, Jungmin, 114 Kant, Immanuel: Perpetual Peace, 272n4o Kaplan, Fred, 258n9 Kapur, Ashok, 2711134; on proliferation chains, 1, 2 Kapur, S. Paul: Dangerous Deterrent, 262n61 Karpin, Michael I.: The Bomb in the Basement, 259n27, 262n66 Kase, Yuri, 275n75 Katzenstein, Peter, 234n3o, 251n48 Kazakhstan: nuclear weapons possessed by, 125, 157, 158, 196, 211, 220, 268n1 Kegley, Charles W., 269n5

INDEX

Kelly, James, 114 Kennedy, John, 1, 19, 203 Keohane, Robert 0., 2331126; Designing Social Inquiry, 26m57 Khan, A. Q., 83-84, 94, 99, 113, 135, 211 Khan, Feroz Hassan, 263n77 Khatami, Mohammad, 2431126 Khrushchev, Nikita, 203 Kier, Elizabeth, 261n51; Imagining War, 26lll59 Kim jong-Il, 31, 104, 122, 197,205 King, Gary: Designing Social Inquiry, 261n57 Korean War, 73, 137, 206 Koring, Paul, 82, 253n4 Kotter, Wolfgang: Nuclear Non-Proliferation and Global Order, 266n48 Krasner, Stephen, 66; Defending the National Interest, 251 n43 Krieger, Wolfgang: The Germans and the Nuclear Question, 276n76 Krishnamurthy Rao, S. V., 35 Kroenig, Matthew, 98-99; on civil and sensitive assistance, 165, 172, 173, 269n6 Kuwait, 2431126 Labs, Eric, 249n29 Lake, David A., 248n19, 2491130, 25m48 Lalman, David: War and Reason, 250n32 Lamborn, Alan C., 251n43 Lapp, Ralph E.: Arms Beyond Doubt, 2661140 Lavoy, Peter, 25, 172, 242n11, 263n77, 265n29 Lebanon,45 Lebow, Richard Ned: The Tragic Vision of Politics, 26ms3; We All Lost the Cold War, 26m56 Legro, jeffrey, 4, 2491132 Leventhal, Paul, 256n53 Levite, Ariel, 124, 131, 265n3o; on nuclear reversal, 138, 139, 253n8, 271n27 Levy, jackS., 61, 244n3o, 248m7, 252n70 Lewis, John: China Builds the Bomb, 262n73 Liberman, Peter, 255n23, 261n58, 262n65, 264n8,2651134 Libya: economic conditions in, 142, 143, 244n27; nuclear weapons program of, 21-22, 24, 44, 45, 47,49-50, 56, 97, 131, 132, 134> 135, 136, 139-40, 142, 150, 152, 157,158,169,196,209,210,254n21; political conditions in, 43, 47, 53, 139-40, 150, 152, 2431126; relations with Israel, 210; relations with United States, 139-40 Lieber, Keir A.: War and the Engineers, 261n6o

287

Lindley, Dan, 259n21 Lipson, Charles: Reliable Partners, 267n54, 268n61 Li, Xiaojun, 243m6 Lobell, Steven E.: Neoclassical Realism, the State and Foreign Policy, 2491127 Long, William)., 2341127, 2651127 Lynn-Jones, Sean M.: Offense, Defense, and War, 261n59 Mack, Andrew, 256n53 MacKenzie, Donald, 2581115; Inventing Accuracy, 267n56 Maddison, Angus, 142, 266n41 Maettig, Thomas, 265n34 Mansfield, Edward D.: Electing to Fight, 268n62,272n41 Mansourov, Alexandre: The North Korean Nuclear Program, 263n78 Maoz, Zeev, 260n35 Mardor, Munya, 262n66 Martin, Kevin, 83 Mastanduno, Michael, 248m9, 249n3o, 251n48 McDermott, Rose, 244n30 Mearsheimer, John J., 2; The Tragedy of Great Power Politics, 2491129 measurement validity: Adcock and Collier on, 104-5, n6-q, 259n31; AHEM approach to, n6-21 Menem, Carlos, 43 Menzies, Robert, 131 Mercer, Jonathan, 244n30, 2611151 Mexico: as nuclear abolitionist, 196, 224; nuclear latency of, 91, 92 Meyer, Stephen M.: on dependent variables, 18; on domestic politics, 38; The Dynamics of Nuclear Proliferation, 8, 14-15, 18-22, 24, 85, 131, 247m, 255n27, 263n4, 267ns5: vs. Hymans, 32; on nuclear decision making, 18-22, 30, 32, 38; on nuclear latency, 18, 19-20, 23, 85, 87, 254TII5, 255n27; on proliferation puzzle, 24; on "technological imperative" hypotheses vs. "motivational" hypotheses, 18-22 Mian, Zia, 83 Middle East, 195, 196; vs. East Asia, 41-42, 43, 45, 52-54, 55, 56; and Iran, 38, 199, 206,207,208,209,212-13,221,225 militarized interstate disputes (MIDs), 173, 179, 186,190,267n50,273nso Miller, Steven E.: Offense, Defense, and War, 261n59

288

INDEX

Minami, Kiyoe, 256n53 Moltz, James Clay: The North Korean Nuclear Program, 263n78 Mangin, Dominique: Histoire des forces nucleaires franraises depuis 1945, 262n72 Monterey Institute of International Studies: James Martin Center for Nonproliferation Studies, 7-8 Montgomery, Alexander H., 254n9, 256n52, 264ll14, 267n50 Moravcsik, Andrew, 4, 249n32 Morgenstern, Oskar, 232n7 Morgenthau, Hans J., 67; Politics among Nations, 232n11 Morocco, 243n26 Morrow, James D., 68, 252n61 Mi.iller, Harald, 254n21, 265nm7,35, 268nn61,63; A European NonProliferation Policy, 263n4; Nuclear NonProliferation and Global Order, 266n48 multicausality, 59, 79, 193 Myanmar, 43, 115 Nagasaki, 238n39 Nasser, Gamal Abdel, 43, 50, 97 nationalism: in China, 47, 52; in Egypt, so; Hymans on oppositional nationalists, 7, 31-32, 33, 36, 140-41; in North Korea, 30, 31, 41; Solingen on, 30-31 National Planning Association: 1970 without Arms Control, 194, 276n3 NATO: after Korean War, 73; relations with Turkey, 53, 214, 215, 216; and security guarantees, 196, 212, 213-14, 217, 218; and West Germany, 73, 74, 75 Nautilus Institute, 114 Nazi Germany, 254n15, 275n76 Nehru, Jawaharlal, 33-34, 132 neoliberal institutionalism, 7, 25, 234nn27,28; criticism of, 27-28, 29; and NPT, 26-28; and nuclear abstinence, 5, 29 Nepal, 143 Netherlands, 91, 92, 254n21 New Zealand: as nuclear abolitionist, 196, 224 Niebuhr, Reinhold: Moral Man & Immoral Society, 249n3o Nigeria: nuclear weapons activities of, 131, 135> 136, 150, 152, 157> 158, 264n16, 266n44; political conditions in, 150, 152 Nikitin, Mary Beth, 256n47 nonproliferation norms, 5, 34, 36, 44, 145, 170, 193, 203; role in nuclear abstinence,

6, 7, 11, 146-54, 175-76, 206-7; role in nuclear reversal, 127-28, 133, 146-54, 155-56; role in opaque proliferation, 22 nonproliferation regime, 12, 34; arms control treaties, 83; confidence in, 27-28, 42, 200,209,212-13,215-16,219,222-23, 224, 226; future of, 56, So, 82, 84, 96; and Iran, 53-54; nongovernmental organizations (NGOs), 27-28; nuclearweapons-free zones, 5, 42, 243n16. See also International Atomic Energy Agency (IAEA); nonproliferation norms; Treaty on the Non-Proliferation of Nuclear Weapons (NPT) North Korea: Agreed Framework, 24, 114; economic conditions in, 41, 55, 143, 182, 183; Korean War, 73, 137, 206; nationalism in, 30, 31, 41; and NPT, 56; as nuclear recidivist, 195, 197, 198, 205-6, 218, 220; nuclear test of 2006, 88, 97, 98, 104, 106, 114, 123, 216-17; nuclear test of 2009, 97, 106, 114, 123, 238n39; nuclear weapons program of, 13, 24, 38, 41,53, 56,86-8h88,96-97,98,1o3-4, 112-13,114,119,120,122-23, 135> 139, 150, 152, 157> 158, 169, 177, 182, 195> 197> 198,205-6,210,238n39,239n74,263n3; political conditions in, 41, 45, 46, 53, 55, 150, 152, 197, 205-6; relations with China, 205, 206; relations with Japan, 206, 216-17; relations with South Korea, 206; relations with United States, 24, 103-4,112-13,114, 139,205,206;SixParty Talks, 113, 205, 2o6; Yongbyon reactor, 87, 88, 113 Norway: nuclear latency of, 91, 92; nuclear weapons program of, 134, 136, 150, 151, 157, 158, 169, 254n21, 271n24; political conditions in, 48, 150 NPT. See Treaty on the Non-Proliferation of Nuclear Weapons "Nth country problem," 2, 232n7 nuclear abolitionists, 196, 219, 221, 224, 227 nuclear abstinence, 38, 211-14; and constructivism, 6; Hymans on, 31, 36, 86, 138; and neoliberal institutionalism, 5,29;Paulon,25-26,36,138-39; prevalence of, 13-14, 129-30, 142, 143; proliferation puzzle regarding, 13-14, 15, 24; and realism, 4, 14, 15, 20, 25-26, 29, 49, 6o, 79; role of nonproliferation norms in, 6, 7, 11, 146-54, 175-76, 206-7; role

INDEX 289

ofNPT, 11, lS-18, 22, 96, 128, 146-48, 164-65, 175-76, 206-7, 221, 27sn7o; role of security guarantees in, 20, 29, 144-4s, 1S4, 178,179-80,183,184,187,207-8, 27sn7o. See also nuclear reversal nuclear ambiguity, 97-98 nuclear disarmament, 16, So, 82, 101, 156, 193,196,221,222,226 nuclear fuel cycle, So, 81, 82, 83, 90-96, 99, 100-101 nuclear latency, 80-101, 255nn27,31, 2s7n2, 271n27; and IAEA, 81, 82-83, 89, 102; Jo and Gartzke on, 84, 87-88, 89, 164, 173; measurement of, 10, 81, 82-83, 8s-93, 2ssn3o, 272nn45,46; Meyer on, 18, 19-20, 23, 8s, 87, 2s4n1s, 2ssn27; vs. nuclear self-sufficiency, 89-90; and political science, 81, 83-90, 96-99; Stoll on, 8s-88; and supply-side control measures,81,83-84,98-99 nuclear pacifism, 1S1 nuclear power, civilian, s, 16, 17, 97; relationship to nuclear weapons capability, S2,80,99-101, 162,16S,170, 199,216, 272n49 nuclear research reactors, 143-44, 1S4, 266n43 nuclear reversal, 138-41, 1S7-s8, 200, 218-22, 242ms; defined, 124-2s; North Korea as nuclear recidivist, 19s, 197, 198, 205-6, 218, 220; prevalence of, 124-28, 129-30; role of alliances in, 127, 130, 144-4s, IS3; role of democracy/ democratization in, 46-48, 128, 140, 148-s3, IS4· ISS; role of domestic politics in, 129, 130, 138, 139-41, 148-s4, 155-56, 200, 219; role of economic liberalization in, 127, 140, 1s3-54, 242n15; role of nonproliferation norms in, 127-28, 133, 146-54, 155-56; role ofNPT in, 127-28, 133, 134, 137, 147-54, 155-56; role of security guarantees in, 127, 137; of South Africa, 24, 43, 88, 139, 196, 220-22, 268n1;ofSweden,138-39,151,218,219; of Switzerland, 138-39, 218, 219. See also nuclear abstinence nuclear suppliers/exports, 17-18, 36, 93, 95, 206; A. Q. Khan network, 83-84, 94, 99, 113, 135, 211; civilian nuclear assistance, 100-101,162,165,172, 173,272n48; Nuclear Suppliers Group, 58, 99, 100; and "Return to Sender" agreements, 100;

sensitive nuclear assistance, 41, 56, 90, 98-99,162,165,172,173,209-11,223, 269n6 nuclear terrorism, 13 nuclear testing, 49, 169-70, 204, 258ms; by China, 5, ss, 120, 122, 244n29; detection of, 106, 258n14; and deterrence, 107, 108, 109-10, 117, 118; by India, 5, 15, 17,24,31,34.35,106, 107,122,151, 170, 244n29; as indicator of nuclear weapons stateness, 10, 22-23, 102-12, 11s-23, 2S9ll28, 260n38; military-run vs. civilian-run programs regarding, 116-23, 262n66; by North Korea, 88, 97, 98,104,106,114, 123,216-lh2}8n39;by Pakistan, 24, 120, 122; in secret, 118-19, 120, 122; by Soviet Union, 119, 120, 122; by United States, 108, 109, 119, 120, 122 nuclear weapons: arsenal characteristics, 98; vs. conventional weapons, 77-79, 105, 180, 202; fourth generation/low-yield weapons, 202, 204; modernization of, 193, 202, 204; use of, 6, 108, 122, 161, 168, 188,223,224,226-2h238n39,269n4 nuclear-weapons-free zones, 5, 42, 243m6 nuclear weapons programs, 21-22; civilian control of, 10, 117-21; military control of, 10, 117-21 nuclear weapons states (NWS): definition of nuclear weapons stateness, 10-11, 102-23; dispute propensity among, 266nso; ElBaradei on virtual nuclear weapons states, 82-83, 89, 100, 102, 104, 2s7n2; Israeli exception, 110-11, 112, 120; and NPT, 16-18, 23, 24, 106, 136, 151, 1S3. 1S6, 193, 19S, 198, 202, 211, 221, 222; nuclear revivalism and end of abolition among, 199, 202-5, 211-12, 21},214,218-19,220-21,222-23,22S, 226; number of, 13-1s, 16, 22, 23, 24, 3S-J6, 119-20,124, 125-26,198,263n3; SQ/no-SQ indicator of nuclear weapons stateness, 102-5, 109, no, 111-16, 117, 121, 257n2, 2s9n26; test/no-test indicator of nuclear weapons stateness, 10, 22-23, 102-12,11S-23,259n28,26on38 Obeidi, Mahdi: The Bomb in My Garden, 2S6n49 Ogilvie-White, Tanya, 263n5 oil prices, 203, 208 Olmert, Ehud, 259n29

290

INDEX

Oneal, John R., 268n61 opaque proliferation, 22-24, 28, 35-36, 119; Israel as case of, 22, 23, 24, uo-11, 112, 120, 205, 259n32, 260nn36.37.39 Overholt, William: on proliferation chains, 3 Pacific Northwest National Laboratory (PNNL): Nuclear Proliferation Technology Trends Analysis, 94-96 Pakistan: economic conditions in, 153, 154, 178; as example of opaque proliferation, 111, 120; nuclear exporting to, 17; nuclear latency of, 91, 92; nuclear testing by, 24, 120, 122; nuclear weapons program of, 17, 23, 24, 111, u2, 120, 122, 134. 136, 150, 153. 154. 157. 158, 169, 170, 171,178,20},205,260n38,263n3; political conditions in, 150; relations with India, 137, 151, 203, 205; relations with United States, 120 Pant, Harsh, 84 Partial Test Ban Treaty (PTBT), 146 Paul, T. V., 8, 9; on balancing, 69, 252n68; on nuclear abstinence, 25-26, 36, 138-39; Power Versus Prudence, 25-26, 36, 265n3o; on status, 233n22 Payne, Keith B.: Deterrence in the Second Nuclear Age, 26on33 peace movement, 27 Peres, Shimon, 262n66 Perkovich, George, 101; India's Nuclear Bomb, 262n75, 267n52, 272n39 Peron, Juan, 43 Perry, William, 38 Peters, Robert, 264n8 Pfaltzgraff, Robert L., Jr., 252n76; Contending Theories of International Relations, 250n34 Pitzer, Kurt: The Bomb in My Garden, 256n49 Plassnik, Ursula, 253n3 Poland, 212, 214 political science, 162; and nuclear latency, 81, 83-90, 96-99 Polity IV data set, 148 Porter, Bruce D.: War and the Rise of the State, 252nn72,74 Posen, Barry, 248n24; The Sources of Military Doctrine, 252n69 Potter, William C., 236n6, 253n8, 258m7, 263n5; Nuclear Power and Nonproliferation, 263n4 Pratt, Manes, 262n66 Prawitz, Jan, 132-33, 264n8

prestige motivations, 3, 4, 41, 129, 130, 202, 207, 219, 226, 242n11, 269n10. See also status motivations prisoner's dilemma (PD), 27, 28 proliferation chains, 1-5, 6, 7, 9-10, 36, 39, 49, 54-55, 73, 77, 179, 195, 227; Dunn and Kahn on, 3, 12; and Iran, 38, 199, 206, 20h208,209,212-13,221,225 proliferation restraint. See nuclear abstinence prospecttheor»46,47-48 Purkitt, Helen E., 255n23; South Africa's Weapons of Mass Destruction, 259n28, 262nn63,64,76 quantitative research, 11-12, 159-92, 264n14, 269nn6,9, 270n19, 273nns1,56,62, 274n63; censored vs. uncensored analysis, 163-64, 166-67; correlation vs. causation in, 176; Cox-regression approach in, 167; and data set reliability, 21-22; hazard modeling, 11, 159, 163-64, 165, 166-67, 175, 176-92, 270n14, 274n67; parametric/ Weibull distribution approach, 167; regression analysis, 163-64, 165, 166, 167, 176, 178, 182, 27onn12,13; risks of, 130-38; vs. qualitative research, 8, 66, 68-69, 12h 159,160-62,170,176,187-88, 192, 271n28 Quester, George, 255n27; The Politics of Nuclear Proliferation, 263n4, 272n38 Raskin, Marcus, 83 Rauchhaus, Robert, 267n5o reactive proliferation, 3-4, 7, 8, 12, so, 174, 179, 186-87, 273n54. See also proliferation chains realism: balancing, 2, 3, 7, 9-10, 49, 55-56, 59-61,63,64,6S,66-77.78-79,247n12, 248n24, 251n52; classical realism, 2, 61-62,63,65,232n11,250n34;conceptof power in, 64-66, 68, 71; criticisms of, 4, 5-7, 20, 25-26, 29, 49-51, 55-56, 6o-61, 63.72-73.79. 141,248n24,249n32, 271n36; defensive perspective, 62, 71, 250n33; and domestic-level variables, 61-69,73. 74-77. 78-79. 248ll24, 249n32, 250n33; emulation of successful practices, 6o, 64, 69, 70; maximization of power, 2, 71, 241n6, 250n33, 251n52; neoclassical realism, 9-10,61-79, 248nn23,24, 249nn25,27, 250nn33-35. 251n53,54, 252n69; neorealism, 2, 3, s. 6-7. 9-10, 20, 22, 39. 49-51, 59-62,

INDEX

64, 66, 69, 70, 79, 232nn12,14, 241n6, 247n10,248n24,251n52,271n35;and nuclear abstinence, 4, 14, 15, 20, 25-26, 29, 49, 6o, 79: offensive perspective, 62, 71, 24m6, 250n33; role for perception in, 64, 66-69, 70, 71-73, 76-n, 78; security concernsin,2, 3,4,7,9-10,20,24,29, 49, 50, 51, 59-63, 67, 70-79, 124, 151, 24In6,247n8,2S0n33,251n52,252n76, 271nn35,36: self-help strategies, 2, 49, 6o, 70, 79, 124; structural realism, 63, 250n34 Redick, John R., 266n38 Reed, Thomas C.: The Nuclear Express, 270ll15 Reichart, John F., 253n2 Reiss, Mitchell, 25, 49, 130, 234n28, 264n8, 266n45, 268n64; Bridled Ambition, 253n8, 264n7, 265n34; Without the Bomb, 241n3,264n7 Rhodes, Richard: The Making of the Atomic Bomb, 262n69, 275n75 Richelson, Jeffrey: Spying on the Bomb, 240n88, 262nn67,68 Ripsman, Norrin M.: Neoclassical Realism, the State and Foreign Policy, 249n27 Robb, Charles, 240n2 Romania: nuclear weapons program of, 96-97, 1}5, 136, 150, 152, 157, 158, 169, 254n21, 271n24; political conditions in, 150, 152 Rose, Gideon: neoclassical realism of, 62, 63, 64, 65, 66, 250n36, 251nn37,43,53,56 Rublee, Maria Rost, 234n31 Ruggie, John Gerard, 234n27 Rumsfeld Commission report, 23 Russett, Bruce, 268n61 Russia: nuclear weapons program of, 46,134,136,169,198-99,204,263n3: political conditions in, 203, 204; relations with Georgia, 207, 211-12; relations with Iran, 200, 202, 204, 207; relations with Ukraine, 214; relations with United States, 200, 203-5, 207, 214, 215, 217, 226, 266n47 Rynning, Sten: Changing Military Doctrine, 249n27 Sadat, Anwar, 41, 43, so Sagan, Scott, 256n52, 257nns9,2, 264nn6,14, 267nso; on domestic politics, 24, 272n39; on models of proliferation, 24-25, 59, 170, 247nns,6,8,11, 253n8; on neorealism,

291

232n14; on security concerns, 24, 59, 6o, 138, 247nn9,11, 271ll35 Sarabhai, Vikram, 35 Sarkees, Meredith, 273n56 Sato, Eisaku, 15 5 Saudi Arabia, 38, 41, 189, 191 Scheel, Andrea, 45 Scheinman, Lawrence, 271n37 Schelling, Thomas: The Strategy of Conflict, 261n52 Schmidt, Andreas, 254n21 Schoettle, Enid: on high posture assumption, 145-46 Schultz, Kenneth A., 48, 245n36 Schwartz, David N.: NATO's Nuclear Dilemmas, 74, 253n79 Schweller, Randall: Deadly Imbalances, 248n23; neoclassical realism of, 64, 65, 249n25; Unanswered Threats, 64, 249nn27,31, 250n35, 251nn39,41 security concerns, u-12, 44, 98, 129-30, 132, 137-38, 139-40, 152, 155, 156, 160, 161,164,170,171,203-4,214,215-18, 222-23, 224, 226, 245n42; acute security dilemmas, 67, 72, 73, 76, 77, 78-79; dispute intensity, 173, 174, 177, 178-79, 183-84,187,190,191, 192;enduring rivalries, 21, 25, 26, 36, 163, 173, 174, 177,179,183,184,189,190, 191,273nsi; militarized interstate disputes (MIDs), 173,179,186, 190,267n50,273n50; nuclear neighbors, 29, 46, 49-50, 54-56, 195,196, 201,202,206-8,212-1},224, 244n34; in realist theories, 2, 3, 4, 7,9-10,20,24,29,49, so, 51,59-63, 66-69, 67, 70-79, 124, 151, 241n6, 247n8, 250n33, 251n52, 252n76, 27Inn35,36; rival nuclear weapons programs, 174, 177, 178-79, 184, 186-87; Sagan on, 59, 6o, 138, 247nn9,11, 271n35 security guarantees, 2, 17, 59, 161, 163, 164, 191; during Cold War, 5, 74, 144-45, 174; role in nuclear abstinence, 20, 29, 144-45. 154, 178, 179-80, 18}, 184, 187, 207-8, 275n70; role in nuclear reversal, 127, 137, 145, 153; by United States, 29, 74· 144-45,174,189-90,192,196,207-8, 209,213-18,226,274n65 Serbia, 211 Shah oflran, 131 Sherwin, Martin J., 258m9 Sigal, Leon, 103 Sikkink, Kathryn, 234n29

292

INDEX

Simons, Howard, 232n7 Singapore, 44 Singer, J. David, 273nn50,52 Singh, Jaswant, 267n52 Singh, Sonali: on economic liberalization, 42, 242n15; on enduring rivalries, 21, 173; on NPT, 272n44; quantitative analysis of nuclear proliferation, 21, 130, 132, 133, 136,137-38,163,164,165,166,167-69, 170, 172, 173. 174> 182, 235n45. 242ll15, 269nn7,10, 270nn13,14,24, 272n45. 273n55; on regime type, 272n4o Snyder, Jack, 248n24; Electing to Fight, 268n62, 272n41; Myths of Empire, 248n23 Snyder, Jed C.: Limiting Nuclear Proliferation, 263n4 Sokolski, Henry, 110, 259n26, 261n5o Solingen, Etel, 8, 25; criticisms of, 29-31; on domestic politics, 6, 7, 9, 28-31, 36, 138,139-40,141,172,174.181-82,187-88, 241n7, 242n11, 268n3; on East Asia, 100; on economic liberalization, 9, 153-54, 163,170,172,174,181-82,187-88, 274n68; on inward-looking regimes, 9, 139-40,172,174, 187-88,242n11,268n3; on nationalism, 30-31; on nuclearization vs. denuclearization, 29-31, 138, 139-40, 187-88, 241n9; Nuclear Logics, 6, 7, 9, 28-31,36,42,55. 160,241n3,253n8, 264n8,265n30,268n3,271n36;on outward-looking regimes, 9, 172, 174, 181, 268n3; on world time, 40, 241nn7,8 Soros, George, 261n48 South Africa: economic conditions in, 190; as example of opaque proliferation, 23, 111; as nuclear abolitionist, 221; nuclear latency of, 88, 91, 92; nuclear reversal of, 24, 43, 88, 139, 196, 220-22, 268m; nuclear weapons program of, 23, 24, 43, 44, 88, 111, 112, 117, 119, 120, 122, 132,

134. 136, 137. 138, 139. 150, 152, 157. 158, 169,170,189,190,196,220-22,244n28, 255n23, 262n64, 265n34, 268m; political conditions in, 43, 139, 150, 152, 185, 221; relations with Israel, 120, 122, 259n28, 271n26 South Korea, 26; nuclear latency of, 91, 92; nuclear reversal of, 139, 219; nuclear weapons program of, 17, 135, 150, 157, 158, 169, 218, 219; political conditions in, 43, 45, 54, 150, 219; relations with China, 54;

relations with Japan, 54; relations with North Korea, 205, 220; relations with United States, 139, 196, 214, 219 Soviet Union: former republics of, 3, 125, 15h158, 196,209,211-14,212,213,214, 217, 220, 222, 268n1; nonproliferation policies, 4-5, 16; nuclear testing by, 119, 120, 122; nuclear weapons program of, 90, 117, 119, 120, 122, 134, 136, 139, 150, 157, 158, 169, 198, 244n29; political conditions in, 150; relations with France, 137; relations with Japan, 189; relations with United States, 72-73, 74, 137, 139, 148, 244n29, 266n47; relations with West Germany, 74, 190; Sputnik, 74. See also Cold War Spain: nuclear latency of, 91, 92; nuclear weapons program of, 132, 134, 136, 150, 152, 157, 158; political conditions in, 43, 45. 150, 152 Spector, Leonard S.: Nuclear Ambitions, 2651123; Nuclear Proliferation Today, 263n4 Spinardi, Graham, 258n15 Sri Lanka, 196, 224 state, the: Composite Index of National Capability (CINC), 273n52; economic pull hypothesis regarding, 127, 141-43; origin in war, 70; regime type, 11, 96-97, 148-53, 155-56, 161, 163, 164, 174, 187, 190,191, 19h268n3,272n40,275n73; relationship to society, 64-67, 69, 70, 74-75, 76-79; secrecy regarding decision making, 6, 14, 27, 97, 107, 117, 118-19, 120,122,200,201,20h209,213-14, 232n11; strong vs. weak states, 66, 70, 75, 76-77, 78-79; technological pull hypothesis regarding, 127, 143-44, 154. See also democracy/democratization; domestic politics; economic resources; technological capacity status motivations, 3, 21, 24, 129, 130, 160, 164,170,171, 172,201,219,226,269n1o; great power aspirations, 20, 25, 26, 36, 233n22; major powers vs. regional powers, 165, 175, 177, 180-81, 183, 185, 186. See also prestige motivations Stein, janice Gross: We All Lost the Cold War, 261n56 Stillman, Danny B.: The Nuclear Express, 270n15 Stoll, Richard: on nuclear latency, 85-87

INDEX

Strauss, Franz Josef, I32, 275n76 Stuckey, John, 273n52 Subrahmanyam, K., 267n53 Sukarno, 43 Sweden: National Defense Research Institute, I33; as nuclear abolitionist, 196, 2I9, 224; nuclear latency of, I, 9I, 92, 255n30; nuclear reversal ot~ I38-39, ISI, 218, 2I9; nuclear weapons program of, I3I, I32-33> I35, I36, I38-39, ISO, ISI, IS7, ISS, I69, 2I8, 219; political conditions in, 48, I3S-39,I50, 2I9 Switzerland: and NPT, I7; nuclear latency of, 9I, 92; nuclear reversal of, I3S-39, 218, 219: nuclear weapons program of, 134, I3S, I36, I38-39, I 50, ISI, I57, ISS, I69, 2IS, 219; political conditions in, 48, I3S-39, ISO, 2I9 symbolic motivations, I39, 242111I, 2691110. See also status motivations Syria: and NPT, 56; nuclear weapons program of, 128, I35, I36, ISO, I 52, I 57, ISS, I89,I95, I96, I97,200,209,2IO, 254n2I, 263n3; political conditions in, 43, 45, 150, I 52; relations with Iran, 210; relations with Israel, I90, I97, 200, 209, 210, 263n3 Taiwan: economic conditions in, 185, I90; nuclear latency of, 9I, 92; nuclear reversal of, 139, 185; nuclear weapons program of, 17, I35, 136, I39, I50, I 57, I5S, I69, I89, I90, I96, 2I8, 2711124; political conditions in, 43, 45, ISO; relations with China, 52, I90, 220; relations with United States, I39, I85 Talbott, R. ).: Nuclear Proliferation Technology Trends Analysis, 256n4I Taliaferro, Jeffrey, 2481119, 249n29, 250n36, 25Inns6,5S, 252nn69,73,75; Balancing Risks, 249n27; neoclassical realism of, 64, 65, 67, 68-69, 70, 25onn32,34,35;

Neoclassical Realism, the State and Foreign Policy, 249n27 Tannenwald, Nina: The Nuclear Taboo, 269n4 Taylor, Theodore B.: Nuclear Proliferation,

Motivations, Capabilities and Strategies for Control, 263n4 technological capacity: nuclear research reactors, I43-44, I54, 266n43; relationship to proliferation, Io-n,

293

IS-22,23,73-74,I2~ I}O, I43-44> I54, I63, I64, I77,I78,225,275n70; "technological imperative" hypotheses, I8-22, 23; technological pull hypothesis, I27, I43-44, I 54· See also economic resources Tellis, Ashley, 107, 25Sm6 Tetlock, Philip, 57 Thayer, Bradley A., 263n5 Tilly, Charles: on the state, 70 tipping point. See proliferation chains. Total Economy Database, I42, 266n42 Trachtenberg, Marc, 259n2o trade ratio, I63, I74-75, 177, I81, 183, 1S5, 243nl5 Treaty on the Non-Proliferation of Nuclear Weapons (NPT), 42, 58, 18I-82, 189, 190, 192, 200, 209, 223, 272n44; Additional Protocol, 45, 243n26; Article I, 135, 2n; Article II, 135; Article IV, 100, I01, I35-36; Article IX, 106-7, 10S; Article VI, 152, 153, 220, 221, 222, 226; Article X withdrawal clause, So, IOO, IOI; and Brazil, 220, 222; civilian nuclear energy in, 16, 17, 135-36, 176; criticisms of, 81, 84-8s,89,90,10J,I10,I35-37:Daion, 26-2S; Epstein on, IS-IS; future of, So, S2, 84, 96; and Iran, 53-54, 56, 93, 2oo-2oi, 207, 226; monitoring and inspection provisions, 24, 27-2S, 41, 45, 88, 91, 102, I}2, I72, 176, 207, 209, 2I3; and nonnuclear weapons states (NNWS), 16-18, 23,24,29,41,43,45, 52,53, s6,90,9I-92, 100-IOI, llO, I}6, I6I, 164-65, 170, 171-72,175-76,1SO, IS3, 184,187, I91, 196,2o6,211,215-16,274n66,275n7o; and nuclear weapons states (NWS), I6-18,23,24, I06,I36, I51, 153, IS6, 193, 195, 198, 202, 2ll, 221, 222; preamble, 152; Review and Extension Conference of I995, 206-7; Review Conference of 1975, 16; Review Conference of 2005, 56; Review Conference of 2010, 56; role in nuclear abstinence, 5, ll, I5-18, 22, 96, I28, 146-4S, 164-65,175-76,206-7,221, 275n7o; role in nuclear reversal, I27-2S, I33, I34, 137, 147-54, 155-56; and West Germany, I6, 17. See also International Atomic Energy Agency (IAEA) Tunisia, 243n26 Turkey, 3S, I89; economic conditions in, 191; political conditions in, 52-53, 2431126;

294

INDEX

Turkey (Continued) relations with EU, 53, 216; relations with Iran, 53, 208, 216; relations with NATO, 53, 214, 215, 216; relations with United States, 214, 216 Ukraine, 3, 217; and NATO, 212; nuclear weapons possessed by, 125, 157, 158, 196, 209, 211, 213, 222, 268n1; political conditions in, 212; relations with Russia, 213, 214 United Arab Emirates, 243026 United Kingdom: nuclear testing by, 120, 122; nuclear weapons program of, 120, 122, 132, 136, 150, 151, 157> 158, 169, 184, 198-99,215,244n29,2S4n21,263n3; political conditions in, 150; relations with Australia, 140, 254n21; relations with United States, 145, 215 United Nations: Irish Resolution, 146-47, 148; relations with Iran, 54, 201, 243026; Security Council, 54, 100, 201, 211, 243026 United States: George W. Bush administration, 27, 103-4; CIA, 103; Clinton administra· tion, 103; dual-key arrangements with, 74, 210-11; Guam Doctrine, 145; Kennedy administration, 1, 19, 203; National Intelligence Estimates (NIEs), 1, 10, 103, 231nn4,5, 258n6; neoconservatives in, 109; nonproliferation policies, 4-5, 16, 17, 23,29,92, 139,185, 199,203,206,219; nuclear testing by, 108, 109, 119, 120, 122; nuclear weapons program of, 90, 92, 93-94,106,108,109,117,119,120,122, 132, 134, 136, 137-38, 141, 150, 151, 157, 158, 168,169,198-99,244n2g,263n3, 269n9, 275n71; Obama administration, 203-4; relations with China, 137, 139, 204-5, 215, 217, 226; relations with France, 145, 215; relations with India, 139; relations with Iran, 137-38, 139, 199,200,201,202,203,204,207-8: relations with Iraq, 97, 104, 113, 139, 200, 207, 209; relations with Israel, 22, 120, 139, 217; relations with Japan, 29, 52, 139, 180, 189-90, 192, 196, 214-15, 216-17; relations with Libya, 139-40; relations with North Korea, 24, 103-4, 112-13, 114, 139, 205, 206; relations with Pakistan, 120; relations with Russia, 200, 203-S,207,214,215,217,226,266n47:

relations with South Korea, 139, 196, 214, 219; relations with Soviet Union, 72-73, 74, 137, 139, 148, 244029, 266n47i relations with Taiwan, 139, 185; relations with Turkey, 214, 216; relations with United Kingdom, 145, 215; relations with West Germany, 16, 17, 74, 76; security guarantees by, 29, 74, 144-45, 174, 189-90,192,196,207-8,209,213-18, 226, 274n6s. See also Cold War Vanaik, Achin, 35 Van Evera, Stephen, 2, 248n24 Vargas, Getulio, 43 Verba, Sidney: Designing Social Inquiry, 261ll57 Vietnam, 44 Vorster, B. J., 132 Walker, Mark: German National Socialism and the Quest for Nuclear Power, 276n76 Walker, William, 268n65 Wallerstein, Immanuel, 24m8 Walsh, James (Jim), 134, 138, 155, 239n77, 265n36, 266nn39.48, 267057 Waltz, Kenneth N.: Adelphi Paper, 3; on balance of power, 59-61, 63, 64, 65, 89, 247012, 248n24, 251n52; on emulation, 70; neorealism of, 25, 59-61, 63, 64, 65, 66, 69, 70-71, 8g, 232n12, 247nn10,12, 248n24, 251n52; Theory ofInternational Politics, 232012, 247010, 250n32, 251n52 Way, Christopher: on economic liberaliza· tion, 42, 242n15; on enduring rivalries, 21, 173; on NPT, 272n44; quantitative analysis of nuclear proliferation, 21, 130, 132, 133, 136, 137-38, 163, 164, 165, 166, 167-69,170,172,173, 174,182,235n45, 269nn7,10, 270n013,14,24, 272n4s, 273n55; on regime type, 272n40 Wayman, Frank W., 250n32 Weber, Max, 31 Weeks, Jessica L., 47 Weinburg, Neil, zs6ns3 Wells, Samuel F.: Limiting Nuclear Proliferation, 263n4 West Germany: economic conditions in, 75-76, 190; and NATO, 73, 74, 75; and NPT, 16, 17; nuclear latency of, 2, 19, 91, 92, 155, 190; as nuclear supplier, 17; nuclear weapons program of, 131, 132, 134, 135, 136, 150, 151, 157, 158, 169,

INDEX 295

189,254n21,271n24,275n76,276n76; political conditions in, 150, 155; relations with Brazil, 17; relations with France, 132, 135, 276n76; relations with Italy, 131, 132, 135, 276n76; relations with Soviet Union, 74, 190; relations with United States, 16, 17, 74, 76; and tactical nuclear weapons, 74, 76 Winkler, Theodor: Kernenergie und Aujlenpolitik, 265n33 Wisotzki, Simone, 268n63; Die Nuklearwaffenpolitik Grojlbritanniens and Frankreichs, 266nn49,50 Wohlstetter, Albert, 263n4; on nuclear weapons staleness, 109-10; Swords from Plowshares, 91-93, 255nn27, 29-31 Wohlstetter, Roberta, 109 Wohlundforth, William, 68; The Elusive Balance, 248n23 Wolf, Albert B., 48, 245n36

Wolff, Jonas, 268n63 Woods, Kevin M.: The Iraqi Perspectives Report, 256n48 Xue Litai: China Builds the Bomb, 262n73 Yager, Joseph A.: Nonproliferation and U.S. Foreign Policy, 263n4 Yoshida, Shigeru, 52 Yugoslavia: nuclear weapons program of, 135, 136, 150, 157, 158, 169, 270n24; political conditions in, 150 Yusu~Moeed,2,4

Zakaria, Fareed, 65-66, 249n25; From Wealth to Power, 248n23, 250n32, 252nn66,69 Zentner, M. D.: Nuclear Proliferation Technology Trends Analysis, 256n41 Zimbabwe, 115, 189, 191-92