Complex Deterrence: Strategy in the Global Age 9780226650043

Citation preview

Complex Deterrence

Complex Deterrence Strategy in the Global Age E D I T E D B Y T. V. PA U L , PATRICK M. MORGAN & JAMES J. WIRTZ

The University of Chicago Press chicago and london

t. v. paul is the James McGill Professor of International Relations in the Department of Political Science at McGill University. pat rick m . m organ is the Tierney Chair in Peace and Conflict at the Department of Political Science, University of California, Irvine. j am es j . w irtz is acting dean of the School of International Graduate Studies and a professor in the Department of National Security Affairs, Naval Postgraduate School, Monterey, California. The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London © 2009 by The University of Chicago All rights reserved. Published 2009 Printed in the United States of America 18 17 16 15 14 13 12 11

23 4 5

is b n -13: 978-0-226-65002-9 (cloth) is b n -13: 978-0-226-65003-6 (paper) is b n -10: 0-226-65002-2 (cloth) is b n -10: 0-226-65003-0 (paper) Library of Congress Cataloging-in-Publication Data Complex deterrence : strategy in the global age / edited by T. V. Paul, Patrick M. Morgan, and James J. Wirtz. p. cm. Papers presented at a conference at McGill University in Montreal in May 2007. Includes bibliographical references and index. is b n -13: 978-0-226-65002-9 (cloth : alk. paper) is b n -13: 978-0-226-65003-6 (pbk. : alk. paper) is b n -10: 0-226-65002-2 (cloth : alk. paper) i s bn-10: 0-226-65003-0 (pbk. : alk. paper) 1. Deterrence (Strategy)—Congresses. I. Paul, T. V. II. Morgan, Patrick M., 1940– III. Wirtz, James J., 1958– u162.6.c66 2009 355.02—dc22 2009011109 o The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, an si z39.48-1992.

Contents

List of Figures and Tables Acknowledgments i

i n t ro d u c t i o n

1

Complex Deterrence: An Introduction T. V. Paul

ii

deterrence and its challenges

2

Three Items in One: Deterrence as Concept, Research Program, and Political Issue Jeffrey W. Knopf Rational Deterrence against “Irrational” Adversaries? No Common Knowledge Janice Gross Stein

3

vii ix

1

31 58

iii d e t e r r e n c e a n d n o n s tat e a c t o r s 4 5

Complex Deterrence in the Asymmetric-Warfare Era Emanuel Adler Deterring Nuclear Terrorists S. Paul Kapur

iv

deterrence and smaller powers

6 7

Deterrence, Rogue States, and the U.S. Policy Robert Jervis Collective-Actor Deterrence Patrick M. Morgan

85 109

133 158

8

Complexity of Deterrence among New Nuclear States: The India-Pakistan Case Dinshaw Mistry 9 Unconventional Deterrence: How the Weak Deter the Strong Ivan Arreguín-Toft 10 Deterrence and Compellence in Iraq, 1991–2003: Lessons for a Complex Paradigm Frank P. Harvey and Patrick James v

183 204 222

deterrence and major powers

Deterrence among Great Powers in an Era of Globalization Patrick M. Morgan and T. V. Paul 12 The Endurance of Extended Deterrence: Continuity, Change, and Complexity in Theory and Policy Timothy W. Crawford 13 The Revolution in Military Affairs: Impact of Emerging Technologies on Deterrence Michel Fortmann and Stéfanie von Hlatky

304

Conclusions James J. Wirtz

321

List of Contributors Index

333 339

11

259 277

Figures and Tables

figure 9.1: Percentage of Victories in Asymmetric Conflict in Four Fifty-Year Periods table 10.1: Summary of Exchanges, Iraq, 1991–2003

206 246

Acknowledgments

t h i s volume reflects renewed academic interest in the theory and practice of deterrence that began to emerge in the aftermath of the September 11, 2001, attacks and the launching of the Iraq War in 2003. Although the actions and declaratory policies of the George W. Bush administration helped catalyze this debate about the efficacy of deterrence, challenges to deterrence theory and policy have come from multiple sources, thus creating an ongoing need for a vigorous debate on this crucial subject. With this aim, we organized a roundtable at the American Political Science Association convention in Washington, D.C., in September 2006 that helped define the parameters of this volume. A subsequent conference was held at McGill University in Montreal in May 2007, which allowed our contributors to present and discuss their chapters and to refine their arguments. Our contributors also introduced their ideas to the broader academic community at the 2008 San Francisco and 2009 New York International Studies Association (ISA) conventions. We wish to thank many individuals and organizations who helped us in developing this volume. The McGill University–University of Montreal Research Group in International Security hosted the May 2007 conference with financial assistance from the Security and Defense Forum, Ottawa, and Fonds de recherche sur la société et la culture, Quebec, through a grant for the Globalization and the National Security State project. Funding also came from the James McGill Chair and the Social Sciences and Humanities Research Council of Canada and the U.S. Navy Treaty Implementation Program. Our research was greatly facilitated by the able and dedicated work of

x

acknowledgments

Theodore McLauchlin and Stéfanie von Hlatky. The research assistance of Jessica Trisko, Mahesh Shankar, Imad Mansour, and William Hogg is greatly appreciated. The three anonymous reviewers at the University of Chicago Press and the chairs and discussants at the Montreal conference and the ISA panels, including Mark Brawley, Avner Cohen, Jeffrey Knopf, Peter Lavoy, John Mueller, Edward Rhodes, and Stephen Saideman, helped to us improve the arguments in the chapters. We also acknowledge the help of Gil Friedman, Aaron Hywarren, Emily Landau, Richard Schultz, and in particular David Pervin, the editor at the University of Chicago Press who showed much interest in the volume. Last but not least, we thank our families for their support and encouragement. Our hope is that this volume will stimulate further thinking and analysis on the multifaceted challenges facing deterrence as a theory and policy approach in the twenty-first century. T. V. Paul Patrick M. Morgan James J. Wirtz

R

part i Introduction

R

chapter one R

Complex Deterrence An Introduction t. v. pa u l

i n t e r n at i o n a l events since the end of the cold war have led many to question traditional understandings of international relations theory and military strategy. Deterrence, the leading theoretical and policy framework during the cold war, has come under criticism. Some believe that in the absence of a major great-power rivalry, deterrence has lost much of its value, and others contend that it has little utility in dealing with “rogue” states or cataclysmic terrorist groups.1 The most relevant expression of the uncertainty surrounding deterrence today is the treatment of nuclear weapons in U.S. policy. American policy makers in general seem to believe that deterrence has only partial relevance in dealing with new nuclear aspirants.2 However, U.S. policy has increasingly relied on nuclear deterrence and compellence as a way to prevent and preempt a chemical or biological attack on the United States, its military forces stationed abroad, and its allies. Along with the other “old” nuclear powers (France, China, Russia, and the United Kingdom), the United States continues to maintain a nuclear arsenal, although the number of warheads is expected to come down in most of these states in the coming decades as they plan limited force modernizations. Recent policy statements also highlight the important role that U.S. military forces play in dissuading potential competitors from enhancing their military capabilities or adopting policies that challenge U.S. interests. In other words, U.S. policy makers seem to be employing several assumptions when it comes to determining when rogue actors will be constrained by the logic of deterrence, when compellence is likely to work, and when dissuasion will turn potential foes toward less confrontational paths. 1

2

t. v. pa u l

The challenges and uncertainty surrounding deterrence call for a reexamination of the deterrence framework. This book responds to the need to reassess the relevance of deterrence as both theory and policy in light of an increasingly complex global political setting. This chapter lays out a conceptual agenda for deterrence theory and policy, especially as it applies to the nuclear realm. It begins with a discussion of the classic deterrence framework and the challenges that were levied against it during the cold war. It then turns to an assessment of how the assumptions of deterrence are met in the new security environment that has emerged since the demise of the cold war. It identifies five key types of deterrence relationships that prevail after the end of bipolarity as well as the current status of the tradition of nonuse of nuclear weapons. Finally, it sketches the significance of the analyses of deterrence presented in this volume and poses some theoretical questions and policy implications that should motivate a new research agenda. Ever since it became the basis of nuclear doctrines in the 1950s, deterrence theory has largely reflected the international context of the times. The theory arose as a response to the enormous impact of nuclear weapons and the awareness that they constituted a class apart, that is, the “absolute weapon,” in Bernard Brodie’s words, or an “uncontestable weapon,” given their unparalleled destructive power and the difficulties of defending against them on the battlefield when used as a weapon of war or when used against cities as a weapon of strategic coercion.3 The U.S.-Soviet arms competition, taking place within the intense East-West conflict, gave rise to the concept of mutual deterrence based on mutual assured destruction. Robert Jervis’s notion of a nuclear revolution followed Brodie’s lead by identifying the strategic consequences—the absence of war—that emerged when two enduring rivals possessed a secure nuclear second-strike capability. In this sense, the nuclear revolution had both technological and strategic dimensions.4 Although conventional deterrence has important theoretical and historical foundations, it is in the nuclear sphere that much of the research has taken place in terms of the theory and practice of deterrence.5 The classic conventional and nuclear deterrence theory is based on three core premises: (1) in order for deterrence to succeed, a deterrer should have sufficient capability, (2) its threat should be credible, and (3) it should be able to communicate the threat to its opponent. These elements should operate across all classifications of deterrence relationships, that is, in general, immediate, and extended deterrence and when both deterrence by denial and deterrence by punishment are attempted.6 Deterrence is achieved if and when a potential attacker, fearing unacceptable punishment or denial

complex deterrence

3

of victory, decides to forgo a planned offensive. The cost-benefit calculations inherent in instrumental rationality are assumed to be present in decision situations involving deterrence. Although deterrence can involve the threat of conventional military attack or the use of nuclear weapons, this chapter largely focuses on nuclear deterrence given that much of the cold war thinking about deterrence involved nuclear weapons. Other chapters in this volume discuss dimensions of deterrence, including conventional and asymmetric, although they treat the former in only a limited way given that complexity in deterrence relationships is mostly visible in nuclear and asymmetric areas. The discussion of collective-actor deterrence, however, expands this literature by including deterrence by international institutions through the threat of military intervention. Despite being elegant and parsimonious, deterrence theory has received considerable criticism for its weaknesses when examined in light of case studies of actual decision situations. During the 1970s and 1980s, the abstractness of deterrence theory was criticized by a number of scholars who saw substantial problems in the theory and its application in a policy framework, especially arising from psychological and domestic political constraints.7 Comparative case studies have shown that deterrence often does not work as theorists would predict, although convincing tests for its success or failure remain elusive. Political leaders under pressure can act quite contrary to deterrence axioms. Regime insecurity and psychological biases, such as wishful thinking, groupthink, and misperceptions, can upset deterrence calculations as well as outcomes. Highly motivated leaders can challenge deterrent threats even without expecting victory. Moreover, it is often difficult to discern whether a state abstains from military attack because of fear of retaliation or because of some other factor that has nothing to do with military considerations. In other words, deterrence theory has not produced a proper test to prove its veracity, and until it develops one, many will view deterrence theorists’ claims with suspicion. To supporters of deterrence, however, the pivotal case study for deterrence success is the cold war–era East-West relationship, in which no direct military conflict occurred between the two superpowers, the United States and the Soviet Union. Great powers in the prenuclear era fought intense wars in pursuit of their strategic goals, but they refrained from doing so during the four decades of the cold war era, despite engaging in an intense rivalry.8 However, with the end of the cold war, along with epochal changes in the international system, that is, the collapse of a superpower without war, many observers seemed willing to consign concerns about nuclear weapons

4

t. v. pa u l

and deterrence to the ash heap of history and move on to other theoretical and policy perspectives. Skepticism about the relevance of deterrence in the new security era arose partially as a result of concerns about whether the theory can account for the behavior of new nations that have acquired or are in the process of obtaining nuclear weapons. While nuclear proliferation is not a new phenomenon, its issue salience has increased in the post–cold war era because of its role in regional conflicts. Although many technologically capable states gave up their nuclear options immediately after the cold war, some vigorously pursued a weapons program.9 India and Pakistan, two states that actively began building nuclear weapons in the 1980s, conducted underground weapons tests in May 1998. Soon afterward, the two rivals entered into an intense phase of adversarial relationship characterized by a “stability-instability paradox.”10 They waged a shooting war in the Kargil region of Kashmir in 1999, thereby raising questions about deterrence in an intense conflict environment involving a nuclear dyad. Yet some observers argue that because the Kargil War did not escalate and the subsequent mobilization crisis in 2002–3 did not prompt a full-scale engagement across the entire region, largely because of the fear of nuclear escalation by the leaders of the two states, recent history demonstrates that deterrence can work even in a theater of active enduring rivalry such as South Asia.11 The fact that intense militarized crises emerged between two nuclear rivals, however, shows that nuclear possession did not prevent crisis initiation, especially by the relatively weaker side. In 1999, it also did not prevent the escalation of a crisis to a limited war. During the 1990s, Iraq, Iran, and North Korea also intensified their nuclear efforts. In the Iraqi case, sanctions and United Nations inspections destroyed any chance of an immediate acquisition of nuclear weapons, while coercive diplomacy and negotiations failed to work in the standoff with North Korea until February 2007, when the six-party negotiations led by the United States produced an agreement under which North Korea agreed to dismantle its nuclear facilities and weapons in return for normalization of relations, security guarantees, and economic assistance.12 The final outcome of these negotiations is difficult to predict given that the agreement does not address the nuclear weapons that North Korea may already possess. Iran has been pursuing a program aimed at nuclear weapons acquisition, apparently undeterred by the prospects of preventive strikes by the United States and Israel that threaten to sidetrack its nuclear ambitions. These nuclear hopefuls are often major challengers, both to their neighbors and to the great-power system. Some of them hold revisionist ideologies or policies meant to upset

complex deterrence

5

regional stability. The George W. Bush administration seems to believe that deterrence may not work with new nuclear states or nonstate actors and that the United States will be self-deterred or will be deterred by those actors if they possess nuclear weapons that can be used against the U.S. homeland, U.S. allies, or U.S. troops stationed abroad. From the regional powers’ vantage point, however, nuclear weapons will be the major deterrent against the United States, given its advocacy of regime change. In the post–September 11, 2001, era, deterrence has acquired further salience, given that some transnational networks such as al-Qaeda are seeking nuclear weapons and radiological materials. Al-Qaeda operatives may have already acquired components for dirty bombs through networks run by friendly rogue scientists, especially from Pakistan, or from the international black market. There are contrasting views among scholars and analysts about the comparative role of ideology and strategy in the behavior of terrorist organizations,13 but deterrence may have little application to terrorists motivated by a cataclysmic agenda who have much to gain from using nuclear weapons and who have little to lose, simply because these organizations lack responsible leadership or clear targets to hold at risk. A crucial issue for policy analysis is to determine whether deterrence axioms are at all relevant to the war on terror.

assumptions of the deterrence framework Deterrence is a theory, a tactic, a national security strategy, a larger defense policy approach, and a “critical component of security for the international system.”14 These various guises all share some common assumptions. Several of these assumptions, although logically valid, may occur infrequently or imperfectly in international politics. A theory cannot be challenged purely on the basis of its assumptions; however, deterrence theory is facing challenges to several of its most important precepts. If these prove empirically inadequate, it may be justified to seek a new theoretical framework. Moreover, deterrence is not only a theory, but a powerful policy instrument. In that respect, the reasonableness of assumptions does matter because policy options are determined by those assumptions. Bad theory, in this instance, is not just an academic matter: it can literally be a matter of life or death. The first assumption of deterrence theory is that states are rational actors and that they make cost-benefit calculations about whether and when to initiate a conflict. Logically, if the costs are higher than the benefits produced by some type of initiative, states will be deterred from taking some

6

t. v. pa u l

aggressive move against their opponents.15 This assumption is generally conceived to be universal in its application, that is, all states should ordinarily behave on this basis, although some believe the theory does not have to fit into actual deterrence situations.16 Can a theory survive if empirical reality is in dissonance with its key precepts? How does deterrence work if actors, limited by weak bureaucracies, internal strife, or ideological and religious zeal, are unable to execute policies in a rational fashion or calculate their interests in the way the theory postulates? Or do states in general behave as if they are rational actors in the face of certain nuclear destruction if they challenge the defender militarily? Historically, states faced with imminent defeat or that are subject to significant punishment from stronger rivals have not always given up, for various reasons. Some leaders might have committed too much of their reputation to retreat without major costs, for instance, or may have thought that their supposedly superior resolve, morale, strategy and tactics, or limited theater-specific advantages could offset the stronger power’s deterrent threats and the advantages of raw military capability.17 This risk-taking propensity of some leaders is not easy to factor into deterrence theory because the theory assumes that states rarely undertake highrisk confrontation. The rationality assumption used in the deterrence framework reflects “instrumental rationality,” that is, actors behave on the basis of cost-benefit calculations to advance their self-interest or obtain goods that maximize their utility. Under instrumental rationality, actors modify their goals, for example, when they realize that the costs of attaining objectives are exceptionally high. However, if actors are driven by “value rationality,” through which they pursue intangible goals with a high degree of commitment even when the costs are too high and success is not assured, deterrence may not work. These goals may include ideological and religious objectives or psychological goals, such as self-respect, dignity, and ethnic pride, for which individuals are willing to sacrifice everything including their lives, even though from an instrumental perspective it may make little sense to behave in this way. As Max Weber states, value-rational behavior is “determined by a conscious belief in the value, for its own sake of some ethical, aesthetic, religious, or other form of behavior, independently of its prospects of success.”18 A second crucial assumption of the deterrence framework is that deterrence operates chiefly among nation-states. This assumption arises from the fact that in the modern international system, the key actors and, therefore, the pivotal units of analysis have been nation-states. States are conceived of as rational bureaucratic entities with coercive power. They are legitimate

complex deterrence

7

entities that can start wars and conclude peace treaties. They are bestowed with bureaucratic-military organizations, and state leaders are expected to make rational decisions in the protection of their national interests. But what if the players are nonstate actors or failed states, which do not follow the rational-actor model of decision making? What if they lack the bureaucratic apparatus that generates and controls policy among normal states? Leaders of failed states may have very little, if any, national interest to speak of in their security objectives: they may be more concerned with internal security against their domestic opponents than defense of the state against foreign threats. Third, both conventional and nuclear deterrence theory assume intense rivalry among the parties. States in deterrent relationships consider war as a constant possibility; leaders believe that enemies would not hesitate to attack if an opportunity for success arose.19 What if nuclear states do not engage in such a rivalry? Threats are difficult to make and credibly execute among nonrivals. But does the possession of nuclear weapons act as insurance against future conflict or the reemergence of rivalry? Do nuclear weapons help to dissuade potential opponents from adopting future policies that can heighten tensions or even lead to war? Does bureaucratic rigidity influence security policy in the deterrence context? How many weapons does a state need to obtain deterrence in an environment of low levels of rivalry? One other assumption is that each class of weapons—conventional, chemical / biological, and nuclear—is at a different layer in the deterrence calculus of states. Threats and policies are relatively easy to discern for each category. What if the symmetry between threats and weapons is blurred as actors attempt to employ one type of weapon to deter the use of another? How credible are nuclear threats in the attempt to prevent a nonnuclear state from using chemical or biological weapons? Would a state be self-deterred from executing a nuclear deterrent threat in the face of a chemical weapons attack? If self-deterrence is a reality, would it undermine the credibility of nuclear deterrent threats against nonnuclear actors? These key issues in the post–cold war era have raised questions about the framework’s robustness and applicability in an increasingly multifaceted security environment. The deterrence framework may well have become much more complex in the new security environment along with the new international system that has emerged since the end of the cold war. Analysts have trouble describing the precise contours of the system, and hence they call it, variously, unipolar, multipolar, or informal empire. While not all elements of a complex system are applicable here, at least three elements of the current international sys-

8

t. v. pa u l

tem generate complexity for the deterrence framework: (1) small events can produce big changes, but it is difficult to predict when they will do so (nonlinearity); (2) the system is characterized by multiple and uncoordinated interactions among a number of different agents and components, and their effects may be complex and unstable; and (3) it is in the process of evolving and adapting and hence unfolding over time in unpredictable ways.20

five types of complex-deterrence relationships As the twenty-first century dawned, deterrence had become complex because of changes along several dimensions of the international system: an increase in the importance of multiple-state and nonstate actors; the distribution of power (the United States has a large military edge over all other states in the system, but the system cannot quite be called unipolar); power relationships (these are evolving, although great powers are in a state of relative peace); and goals, ideals, and issues (asymmetric challengers want to alter the regional or international status quo). Deterrence operates best when there is clarity on these elements, while ambiguity in them makes deterrent relationships complex, in the realms of both theory and policy. Structural indeterminacy is thus the root cause of complex deterrence. This indeterminacy manifests itself in the areas of signaling threats, attributing responsibility for hostile actions, and asymmetry of interests as diverse actors are driven by different calculations that are difficult to assess. “Complex deterrence” can thus be defined as an ambiguous deterrence relationship, which is caused by fluid structural elements of the international system to the extent that the nature and type of actors, their power relationships, and their motives become unclear, making it difficult to mount and signal credible deterrent threats in accordance with the established precepts of deterrence theory. What are the elements of complex-deterrence relationships? The deterrence framework during the cold war era was focused largely on the superpowers, especially as a strategic and psychological mechanism to manage their relations. Over time, leaders on both sides came to share a normative understanding of the operation of deterrence. The complex nature of the post–cold war international system has caused an expansion of deterrence relationships. During the cold war, the central international system was formed by the interactions of the two superpowers. In today’s near-unipolar international system, multiple layers of interactions and interconnectedness have emerged. Weak states no longer are all that weak, as asymmetric ca-

complex deterrence

9

pabilities and strategies are becoming effective in many situations in constraining the great powers, most prominently the United States. This has made deterrence outcomes more difficult to predict. Events occurring in regions such as the Middle East and South Asia, of limited consequence for the central system during the cold war, have become more consequential for the international system and the nature of power relations in it. Acquisition of weapons of mass destruction (WMD) capabilities by smaller powers has increased the complexity of deterrent relationships involving great powers and non–great powers. The new actors need not have the same learning curve or shared normative understanding of deterrence. Deterrence is an iterative relationship that requires regular communications between parties, which was achieved during the cold war through a series of arms-control negotiations and diplomatic engagements, interactions that are missing for the new nuclear states and nonstate actors.21 Dysfunctional learning can harm deterrent relationships as actors fail to observe its central precepts and try to bypass or design around it. In addition, increased economic globalization involving all principal states has made deterrence more complex. Making and executing threats between major trading partners is more difficult than among adversaries with limited economic, cultural and political interactions. There are at least five ideal types of deterrence relationships in the contemporary era that warrant our attention in order to speak of complex deterrence.22 They are 1. Deterrence among great powers, 2. Deterrence among new nuclear states, 3. Deterrence and extended deterrence involving nuclear great powers and regional powers armed with chemical, biological and nuclear weapons, 4. Deterrence between nuclear states and nonstate actors, 5. Deterrence by collective actors. Deterrence among Great Powers Among the great powers (the five permanent members of the United Nations Security Council), nuclear weapons are largely seen as a hedge against the emergence of great-power conflict in the future. The great-power relationships in the post–cold war era are characterized by “recessed general deterrence,” or dissuasion, in which states do not expect immediate militarized

10

t. v. pau l

conflict, but weapons are kept in the background as insurance given the inherent uncertainties of world politics. The end of the cold war witnessed substantial changes in the deterrence dynamics involving great powers, and, as a result, general deterrence and dissuasion became operational concepts. Although they do maintain large arsenals, neither the United States nor Russia is presumed to hold automatic launch-on-warning attack plans anymore, although some of the elements of the previous era are continuing.23 In addition, they have reduced the number of weapons they possess, although the numbers still exceed a minimum nuclear deterrence posture. The three other old nuclear powers—China, the United Kingdom, and France—also have been maintaining their smaller arsenals, but this might change as Chinese nuclear force modernization plans come to fruition in the coming decades. The logic behind the maintenance of nuclear capabilities is that the great powers want to be prepared in case their relations deteriorate in the future. Nuclear capability can also be construed as an assurance against the expansionist pathologies of great powers as described in perspectives such as offensive realism.24 Moreover, uncertainties in Russia and China give pause to Western nuclear powers, while, for Moscow, the fear of American influence in its former spheres in Eastern Europe and Central Asia is the cardinal source of anxiety. For the rising power, China, nuclear weapons offer a major insurance against direct assault on its strategic sphere, allowing it to rise peacefully. Nuclear weapons also offer a limited but crucial deterrent against potential conflict escalation between the United States and China involving Taiwan. The great-power deterrence calculations are thus based on “recessed general deterrence” as well as “existential deterrence”: no immediate expectations of war exist among them. However, as Patrick Morgan states, “if serious conflicts emerge again, then deterrence will be in vogue—if not, at least for a lengthy period, then deterrence will operate offstage, held in reserve, and will not be the cornerstone of security management for the system.”25 This does not mean that the relations in the U.S.-Russia and U.S.China dyads would remain the same in the long run. Power transition has invariably been turbulent in the international system, and herein lies the role that nuclear weapons may play in deterring a transition war. U.S.-Russia relations could deteriorate, and deterrence could become more relevant if tensions build up over the establishment of missile defenses in Eastern Europe and over Russian efforts to repudiate major arms-control agreements in its effort to regain its lost superpower status. As discussed in Morgan and Paul’s chapter in this volume, nuclear deterrence in this context has offered the major powers greater maneuverability.

complex deterrence

11

It has allowed the major power states to sustain their credentials as system managers and has prevented the emergence of active security dilemmas among them that can be caused by conventional arms races and technological breakthroughs. Absent the fear of existential wars, the potentially rival states have engaged in greater economic interactions. The increasing trade relations between the United States and China and China and India, an emerging power, suggest that general nuclear deterrence may offer economic spin-off benefits. To some extent, the stability in relations among the great powers, with no war in sight between them, points to the pacifying role that nuclear weapons may be playing, although other causes are present as well.26 In that sense, nuclear weapons may act as crucial factors in preventing a power-transition war akin to those that the world experienced in the nineteenth and twentieth centuries. For Russia, the superpower that declined, nuclear deterrence offers an opportunity not to be excessively alarmed by the expansion of the North Atlantic Treaty Organization. The continued possession of nuclear weapons, as well as the search for defensive capabilities including missile defenses and offensive capabilities such as antisatellite weapons involving the militarization of space, suggests that great-power deterrence may become more complex in the future. If a great power achieves a unilateral advantage in the revolution in military affairs (RMA), that power would be tempted to break out of the constraints imposed by the deterrence framework. Hence, as Michel Fortmann and Stéfanie von Hlatky rightly point out in their chapter in this volume, technological changes inherent in the RMA are crucial to understanding why great-power deterrent relationships could undergo major changes and assume higher levels of complexity in years to come. A further factor that will make deterrence more complex as the twenty-first century advances is the deepening of economic globalization involving major powers, especially China, Russia, and the United States and emerging powers such as India. Globalization would make military threats for anything but extreme challenges hard to signal or execute and would look incredible given the high costs of war among major trading partners. Deterrence among New Nuclear States The mutual-deterrence relationship among relatively new nuclear states has assumed prominence in the twenty-first century. The India-Pakistan deterrence relationship is the key one here, although it is conceivable that in the next decade or so such a relationship could emerge between Israel and Iran

12

t. v. pa u l

or between North Korea and Japan. India-Pakistan nuclear deterrence predates May 1998, when the two openly conducted nuclear tests. These South Asian neighbors had an opaque deterrence relationship since the late 1980s, when both were assumed to have acquired rudimentary nuclear capabilities. This deterrence relationship involves two active rivals that have fought four wars and engaged in several crises. Three wars predate the nuclear phase of the relationship, occurring in 1947–48, 1965, and 1971. But the Kargil conflict of 1999 occurred after nuclear acquisition in the late 1980s and indeed after their 1998 tests. The Kargil conflict provides ambiguous evidence for and against deterrence theory. On the one hand, a nuclear state, Pakistan, initiated a limited war against another nuclear rival, India, expecting no major retaliation from the larger neighbor. India did not escalate the war, but it did wage an intense localized battle that challenged deterrence axioms.27 Subsequently, India also engaged in a major mobilization crisis with Pakistan, as a result of the presumed complicity of the latter in launching an attack on the Indian parliament in December 2001. In the wake of these crises, India adopted a new military strategy, called “cold start,” which called for forward attacks on several points along the Pakistani border, even in the face of a WMD response, if an incident similar to the parliament attack were to occur again.28 The deterrence relationships involving new nuclear states have been questioned on the basis of doubts about four elements of deterrence: capability, credibility, communication, and stability. Smaller nuclear forces are vulnerable because of uncertainties about the extent of their offensive capabilities and survivability. The possibility of preventive and preemptive attacks as well as short distances and decision times lower the credibility of nuclear threats. The absence of political dialogue, the likelihood of military miscalculations, cultural differences, short distances and decision times, and historical animosities might all impair communication in a crisis. These relationships do not exhibit the stability that is traditionally characteristic of nuclear-backed interactions, since the actors engage in recurrent crises behavior and uncontrolled arms races, and have domestic reasons for crisis initiation that may not have much to do with their national security interests.29 In South Asia, Pakistan’s continued domestic instability and the proclivity of insecure military and civilian leaders to externalize internal conflicts add to the problem of deterrence stability. Others, such as Dinshaw Mistry in his chapter in this volume, have argued that even with all these pathologies, India and Pakistan have begun a process of nuclear learning, eliciting greater caution in policy and attempts at resolving the underlying conflict.

complex deterrence

13

Deterrence supporters also have charged that some strategic analyses of new nuclear powers may reflect ethnocentrism in Western scholarship.30 The nonescalation of crises in South Asia in the nuclear era attests to the effectiveness of deterrence. The initiation of a peace process in South Asia since 2004 suggests that the instability of the initial phase of nuclear proliferation may be changing as parties learn the limits of coercive behavior among nuclear-armed antagonists. Deterrence between Nuclear States and Nonnuclear WMD States Deterrence between nuclear states and regional powers, such as Iran and Syria, which are armed with chemical, biological, or nuclear weapons, is an issue that did not raise many theoretical or practical concerns during the cold war. Today, at least five nuclear powers—the United States, United Kingdom, France, Russia, and India—have made explicit policy changes that suggest the possibility of nuclear retaliation against attacks by chemical, biological, or radiological weapons. Israel, the undeclared nuclear power, also seems to be willing to retaliate with nuclear weapons against chemical or biological attacks by regional enemies. Israel’s deterrence relationship with its Arab neighbors is of a special variety. Israel is a single nuclear state surrounded by a host of Arab states. Some of Israel’s neighbors have developed an off-and-on rapprochement with the Jewish state, while the others are continuously hostile. Israel has declared several redlines that, if crossed, would result in a nuclear response. Nuclear deterrence in this instance is directed against conventional, chemical, or biological attacks by larger adversarial states. Israel may be asymmetrically deterred from using its nuclear arsenal, however, because its adversaries possess chemical and biological weapons, which are sometimes viewed as the “poor man’s atomic weapons.” The destruction of Iraq’s Osirak nuclear facility in 1981 and a Syrian facility in September 2007 suggests that Israel would be inclined to take a hard line against efforts by its Arab neighbors to expand past chemical and biological arms and develop nuclear weapons. While Israel’s nuclear possession may have been a factor for the Arab states’ not engaging in a state-destroying war, the 1973, 1982, 1991, and 2006 conflicts suggest that Israel’s nuclear arsenal has not prevented its neighbors from challenging it militarily.31 It is unclear how much nuclear deterrence operates in this context, other than as an Israeli deterrent against an existential threat. Because Israel does not publicly state its weapons capability and deterrence doctrine and because the Arab calculations often seem confusing,

14

t. v. pa u l

the Arab-Israeli conflict has emerged as a complex-deterrence relationship, adding another layer of complexity to the basic theory of deterrence. The deterrence relationships involving nuclear states and nonnuclear states in possession of other WMD capabilities are fraught with uncertainties. How credible is the threat of nuclear retaliation in response to a chemical or biological attack given the potentially disproportionate nature of a nuclear response? U.S. policy since the early 1990s has been to offer “calculated ambiguity” when facing a regional opponent armed with chemical and biological weapons. It is believed that Saddam Hussein did not use such weapons in the first gulf war because of a fear of nuclear retaliation. Although postwar statements by Iraqi leaders support this interpretation of events during the first gulf war, it is quite possible that Hussein had little capability to inflict chemical and biological damage using missiles, and he was probably deterred by the fear of regime removal, not just the fear of nuclear retaliation.32 He was not deterred from burning Kuwaiti oil fields, a redline for possible nuclear retaliation by the United States. It is also unclear why he was not deterred from attacking the heartland of a nuclear-armed Israel with missiles, even after Israel made ambiguous threats of retaliation. What if the Iraqi missiles had killed large numbers of Israelis? Moreover, Saddam Hussein had dispersed chemical and biological weapons and predelegated to local commanders the authority to retaliate with chemical weapons in the event of a U.S. invasion of Baghdad.33 There is also some evidence that the George H. W. Bush administration’s decision not to escalate the first gulf war in 1991 was partly owing to a concern that Saddam Hussein would have used his chemical and biological capability in the face of certain removal from power and possibly death.34 The question here is, if an actor is highly prone to risk taking, will it challenge the deterrent threats of stronger powers under certain circumstances? The case studies on immediate deterrence and compellence involving the United States and Iraq in the chapter by Frank Harvey and Patrick James offer substantial evidence for the complex nature of this form of deterrence relationship. The deterrence relationship in these contexts need not be one-sided. Deterrence could be both asymmetrical and mutual because the weaker side’s chemical and biological capability could dissuade a stronger power from escalating or intervening in a conflict. North Korea also has proven that asymmetrical deterrence is possible. Pyongyang has communicated this deterrent threat through the possession of conventional and limited nuclear capabilities that could cause unacceptable destruction to South Korea and Japan. It is also possible that this asymmetric capability has acted as a con-

complex deterrence

15

straint on the United States against launching a military response to North Korean provocations, even though such an option was considered by the Bill Clinton administration in 1994. As Ivan Arreguín-Toft argues in his chapter, unconventional or asymmetric deterrence is more prominent than standard international theories would have us believe. And as Timothy Crawford’s chapter demonstrates, alliance structures and extended deterrence, carrying over from the cold war, permeate likely U.S. responses to new regional antagonists, but these instruments may not be credible to a weaker challenger using asymmetric means. The expectations of regional states attempting to acquire WMD are clear: once they obtain such capabilities, they will not be easily susceptible to coercive intervention, especially attacks intended to remove the existing regime. In this sense, WMD could act as “great equalizers” in the calculations of weaker actors, even though they may not deter the great powers under all circumstances.35 For the so-called rogue regime, nuclear possession offers existential deterrence, given that the leaders of these regimes believe that they are targeted for removal by the United States. Deterrence is a two-way street, as shown in the chapter by Robert Jervis. The weaker side also can gain strategically as a result of current U.S. policies that simultaneously denigrate deterrence and exaggerate what the so-called rogues can accomplish through nuclear possession. However, a weaker state could prompt military intervention during the period in which it attempts WMD acquisition, as happened in the case of Iraq and potentially in the case of Iran. Asymmetric deterrence against military intervention could also be based on the ability of the regional power to mount challenges elsewhere. For example, Iran could unleash terrorism as a means to deter America’s military intervention toward it. Deterrence now looks complex, since even with asymmetric capabilities, the stronger side needs to pay attention to the weaker state’s capabilities and intentions. This form of deterrence can be constraining to dominant actors, who are often unwilling to accept defiance from a regional challenger. Deterrence between Nuclear States and Nonstate Actors The deterrence relationship between nuclear powers and terrorist groups, if it exists at all, is another issue that has entered the discussion on deterrence theory, adding one more layer of complexity to the deterrence framework. It is generally assumed that cataclysmic terrorist groups such as al-Qaeda may not be deterrable using a threat of retaliation that may be unacceptable to state actors. Since these groups do not often operate within a nation-state,

16

t. v. pa u l

the threat of retaliation of a state that offers base facilities may or may not be possible. Alternatively, deterring potential providers of nuclear arms offers, in theory, an opportunity to take action against nonstate actors through state sponsors. As Emanuel Adler rightly points out in his chapter, “damned if you do, damned if you don’t” could be the dilemma facing nation-states that confront such challenges, as a retaliatory attack would provoke unnecessary international criticism and it may not achieve any military goals, while no response implies the weakness of the credibility of the state. The opportunities for illicit acquisition of arms from nonstate actors, for example, from the A. Q. Khan network, may not make this any more tractable as a policy. Moreover, these groups operate in weak or failed states, and the capacity of central leaders to control them can be extremely limited. The complexity of this form of deterrence relationship, in both bilateral and extended domains, is likely to increase as the twenty-first century advances if nonstate actors acquire rudimentary nuclear capability or radiological weapons. In many respects, it may be characterized as microdeterrence, as the precise nature of the opponent’s capabilities and intentions are not known because of the absence of accurate intelligence assessments.36 The difficulty of nuclear attribution emerges as a major problem in the aftermath of an incident involving a nuclear weapon. In fact, deterrence theory faces its greatest challenge in this type of relationship. The differing construction of threats in different cultural settings and variations in the heuristics that dominate strategic thinking among boundedly rational human beings could, as Janice Stein contends, challenge deterrence in these circumstances. It appears, however, that few types of terrorist groups would engage in a nuclear attack. Groups that are seeking independent states such as the Chechens in Russia, Tamil separatists in Sri Lanka, or Palestinian factions such as Hamas are unlikely to use nuclear or chemical weapons because this type of attack could cause them to lose their legitimacy and potential international recognition.37 As Paul Kapur contends in his chapter, groups that are seeking more elusive objectives, such as the Japanese group Aum Shinrikyo, may be more prone to use nuclear weapons or dirty bombs because they are driven by goals that are not often earthbound.38 Collective-Actor Deterrence This form of deterrence involves international institutions and states. Relevant states are deterred from undertaking certain types of aggressive actions,

complex deterrence

17

whether they are internal or external, because of the fear of international opprobrium and collective punishment through interventions and sanctions under the aegis of multilateral institutions. This type of deterrence is not new, but it has become increasingly salient in the post–cold war era. The United Nations’ collective security provisions under Chapter 7 have included a deterrent component. The expectation was that as the threat of collective security action looms, states would not undertake certain aggressive actions that threaten international peace and security.39 Collective-actor deterrence could operate at two levels: the individual and the state. If deterrence succeeds, leaders should refrain from engaging in unacceptable behaviors because they may be afraid of losing power, being imprisoned or killed, or being brought to an international court after intervention by the collective actor. Interest in regime survival is important here. The other level is national—a state could lose its autonomy or become an outcast in the international system. Fear of this outcome could deter state officials from committing an aggressive act that the collective actors wish to prevent. There are limits to the viability of collective-actor deterrence, both in the range of states that may be targeted by it and in the logic of its operation. Collective deterrence can be constrained in terms of all three aspects of its operation: capability, credibility, and commitment—on a much wider scale here. Collective-actor deterrence is generally meant to restrain the behavior of small, weaker actors because it is unlikely to have sufficient capability to work against great powers and dominant regional actors. Failed or failing states, however, may not have the necessary capacity or institutions to comply with the dictates of the collective actor. In the logic of its operation, there are significant problems with taking collective action to deter aggression. Casualty-sensitive states may not want to intervene for fear of getting bogged down in a domestically unpopular war in which clear-cut success is unlikely. These are not new problems in terms of undertaking collective action to preserve the peace, but they are important because the global community no longer has a bipolar alliance system to control the behavior of smaller or weak states. They are further constrained in the area of asymmetric warfare because collective actors also could get bogged down waging unwinnable counterinsurgency warfare. Further, peacekeeping and peace enforcement are efforts for winning the minds of the local population, and excessive use of force can be counterproductive. Considering the costs of intervention, serious problems of collective action emerge, such as free riding and buck-passing. Targets have good reason to be skeptical of the capability

18

t. v. pau l

of collective actors and of the credibility of the deterrent threat. Here the complex aspect of deterrence in a complex international environment is all the more relevant.40 Apart from the five ideal types of deterrence, there are several other subcategories of deterrence, for example, involving conventional capabilities. In fact, many of the ongoing deterrence relationships involve conventional deterrence as states maintain and update conventional armed forces with the aim of obtaining effective defensive or deterrent and, in some cases, offensive capabilities. States in enduring rivalries are the most likely candidates to rely on immediate and general conventional deterrence. In the post–cold war era, conventional deterrence has assumed complexity partly owing to the arrival of deepened economic globalization.41 States in economic interdependence would find it difficult to threaten each other with military force as they attempt to retain or deepen their economic relationships. Globalization also constrains efforts to extend deterrence to allies who are threatened by their rivals. Furthermore, conventional deterrence could become more complex in deterrence relationships in which there is an intersection between conventional and chemical, biological, or nuclear weapons.

self-deterrence Beyond these challenges, deterrence has became complex as a result of the operation of a phenomenon called “self-deterrence.” A nuclear state attempting to deter asymmetric attacks by a nonnuclear challenger or a terrorist state may confront several problems. These challenges may be strategic, tactical, reputational, or moral and legal, all causing great difficulties for a nuclear state in mounting a credible threat of retaliation and executing that threat if the challenger goes ahead with an attack. This phenomenon can be aptly called self-deterrence. Self-deterrence could also be accentuated by the operation of the “tradition of nonuse” or a “taboo,” as some would call it, against the use of nuclear weapons.42 During the cold war, a tradition of nonuse emerged: nuclear states would not use their weapons against nonnuclear states owing to normative and reputational concerns. This tradition of nuclear nonuse created concerns about credibility, one of the core components of deterrence. According to Steven Lee, “The credibility problem concerns the strength of the belief on the part of the threatened party that the threatener would be willing to carry out its threats.”43 Kenneth Boulding described the issue involving the credibility of threats in a slightly different way: “if threats are not carried

complex deterrence

19

out their credibility gradually declines. Credibility, as it were, is a commodity which depreciates with the mere passage of time.”44 The tradition of nonuse raises questions about the political commitment and credibility necessary for successful deterrence. The difficulty with a nuclear deterrent is in translating capability into credible deterrent threats and in the constraints in committing one’s capability for anything other than the supreme interests of a state such as national survival. But such existential threats are nearly absent today. The credibility of declared as well as tacit commitment is questioned when a nonnuclear state or nonstate actor initiates a war or limited probing action against a nuclear-armed adversary. Nuclear states attempting to use nuclear threats to confront such limited challenges are likely to face considerable credibility problems. In the nuclear context, both operational and normative constraints can act as sources of self-deterrence; a nuclear state might be unable to execute a retaliatory threat owing not to a lack of capability but to other factors. The nuclear state is deterred either by a logic of consequences, which is produced by concerns about its reputation, prestige, power position, and general international standing, or by a logic of appropriateness, which is generated by normative considerations.45 Fear of other-imposed military costs is not the primary reason for the actor’s refraining from executing a threat of retaliation; instead, it results from a self-imposed rational or normative restraint.46 Self-deterrence could be strong in collective-actor cases or in deterrence involving nonstate actors because of the very elusive nature of the targets, especially their mode of operation and their physical presence. More importantly, the casualty-sensitive and reputation-conscious militaries of today, especially in the West, are more reluctant to use force than they used to be.47 War is now not only military operations but also competition for the hearts and minds of people, as the world media could show any excesses committed in the target state to a large audience instantaneously. This has added to the complexity of deterrence, as countries are reluctant to make credible threats that increasingly tend to be ignored by the potential targets of deterrence. The difficulty arising from possible self-deterrence is a reason why some in the United States want to build “more useable” mini nuclear weapons. With the end of the cold war and as the new international order has unfolded in the twenty-first century, the complexity of deterrence relationships has increased manifold. Nuclear states are frantically searching for a role for their weapons in the new security matrix but with rather questionable effects. Preventive and preemptive nuclear attack plans toward regional

20

t. v. pa u l

challengers seem to be spurring a new round of nuclear proliferation by regimes that are worried about survival. The trend toward additional states acquiring nuclear weapons or a virtual nuclear weapon capability is likely to accelerate as the twenty-first century advances, creating multiple layers of deterrence relationships involving states of different capability, resolve, and deterrence postures. All these developments suggest the need for rethinking deterrence by deviating from its previous straitjacket conceptions that were developed during the cold war era. This volume is an effort to understand the multidimensional problems deterrence faces in today’s international system.

significance of this volume Deterrence theory and policy attracted considerable attention during the cold war, but there have only been a handful of book-length studies on the subject undertaken by international relations scholars who explored the theory and practice of deterrence beyond a bipolar setting.48 Policy makers, especially those who do not fully subscribe to deterrence theory, have been at the vanguard of questioning its validity as both a theory and a basis for strategy. Officials in the George W. Bush administration have led the efforts to challenge deterrence, as they mix the logics of deterrence, compellence, and dissuasion as well as preemption and prevention in an effort to create a strategy to meet emerging security challenges. This questioning often is prompted by the United States’ frustration in its dealings with minor powers as well as nonstate actors who seem to be challenging U.S. dominance at a time when the disparity between U.S. military capability and the nearest great-power rival is overwhelming. The unipolar moment may be a fleeting one, as some scholars had feared.49 Those who question deterrence in the regional context are also often the main advocates of U.S. preponderance and global hegemony, a policy approach that does not necessarily mesh with deterrence theory. The critics seem to expect deterrence theory to answer some of the pressing security challenges that the United States faces, especially in its pursuit of global dominance. But they are reluctant to let regional states have their own ability to deter American intervention in their affairs. Deterrence is a double-edged sword, a point that deterrence detractors often ignore. A reassessment by international relations scholars of contemporary theoretical and policy issues related to deterrence is overdue. We have had two decades of experience since the end of the cold war to discuss the theoretical

complex deterrence

21

validity and applicability of the deterrence framework today. Deterrence is nothing if not complex; so are compellence and coercive diplomacy. Preventive and preemptive wars are even harder to wage successfully. In the post– cold war era, the elegance and parsimony of the theory are questionable, while as a policy instrument, it may be cumbersome at best to apply and operationalize. Despite its imperfections, deterrence is here to stay, because alternatives are few when dealing with states that have acquired WMD capabilities, especially nuclear weapons. The contributions in this book have major policy significance. It is important to evaluate the role of deterrence in the policy problems that confront decision makers. This is all the more pressing because, at present, policy is moving down a dangerous road as deterrence is rejected without a persuasive alternative. U.S. policy makers’ concerns about deterrence may have some validity, but the solutions that they have proposed or engaged in are worse than those under the deterrence framework. In the post–cold war era, U.S. policy toward regional challengers has been moving toward compellence in the sense of forcing states to abandon nuclear weapons programs. In fact, compellence has more weaknesses as a policy instrument than deterrence. Prevention and preemption produce turbulence (as evident in Iraq) and can lead to violent outcomes for regional orders, especially when nonstate actors can take advantage of state weakness to operate and potentially threaten their own campaigns of mass destruction. That actors sometimes behave irrationally is not a sufficient reason to jettison the deterrence framework. This would be akin to abandoning the market system because of irrational behavior by some consumers and producers. Actor rationality requires learning, and the behavior of revisionist states may be susceptible to modifications under conditions of mutual deterrence. Critics of deterrence theory have not yet developed a powerful alternative theory that can explain behavior under threat of retaliation. Describing behavior that is difficult to explain or pointing to the lacunae in one crisis or another is not sufficient to discredit the deterrence framework.50 Because deterrence is a military-psychological instrument of statecraft, politics matters too in conflict prevention. Seeking military solutions to essentially political problems could lead to further limitations of deterrence. Deterrence might be a better strategy to deal with a regional proliferator already on its way to acquiring a nuclear capability than preventive attack, which may simply embolden the state to accelerate its nuclear capability. Deterrence is one of several policy instruments, along with reassurance, diplomacy, dissuasion, economic rewards and punishments, regional

22

t. v. pau l

integration, and other conflict resolution methods. Socializing the new nuclear state on the need to avoid highly risky behavior is essential for deterrence to succeed. Regime change as a strategy has serious difficulties; a regime under existential threat could refuse to abide by the axioms of the deterrence framework. Research Questions Even if deterrence as a framework cannot remain as prominent in the new security environment as it did in the cold war, consideration of the questions raised in this and other chapters of this volume should allow theorists and policy makers to grapple with new and pressing problems. There are clear theoretical and policy imperatives for rethinking deterrence. Collectively, our contributors seek to address a series of questions raised by the new conditions facing deterrence as a theory and as a policy. These questions would also be useful for future research agendas on deterrence theory and policy. • Do the axioms and assumptions of the deterrence framework still hold true, or do they need to be recast to address the complex challenges in the current and emerging international or regional orders? In other words, is deterrence still feasible in the changed international context?51 • What is the role of rationality in deterrence? How much is required? What kind of rationality, for example, instrumental or value based? • What are the roles of culture, norms, and other shared understandings that are necessary for deterrence to work? Conversely, to what extent are differences in culture, and so on, likely to undermine deterrence? • What is the impact of domestic politics on deterrence? To what extent, and in what ways, do internal factors make deterrence either more or less likely to work? • To what extent are multiple actors likely to be involved in a deterrence situation, and do interactions among multiple actors alter the nature of deterrence relative to traditional dyadic models of deterrence? • To what extent are nonstate actors so different from states as to be outside the deterrence framework? • Why are policy makers, especially in the United States, moving away from the deterrence framework? • If deterrence is a failed or constrained framework, what can replace it? • Is deterrence a weak policy approach in confronting transnational terrorism?

complex deterrence

23

• What is the future of the deterrence framework? Will asymmetric warfare unseat it as a dominant approach to the management of security policy? • Will RMAs, especially the weaponization of space, displace the deterrence framework? • How will the deterrence framework operate when rising powers such as China and India gain increased power in the international system? t he chapters in this volume offer answers to many of these and other related questions and challenges deterrence faces in the twenty-first century. Our contributors in general point to the constraints in applying deterrence in the contemporary complex international system. Deterrence may still be relevant, but it now must take its place alongside other tools of statecraft. Strategies are context and time dependent, and they are not immutable. Leaders who forget this axiom can create unnecessary expectations for deterrence. Whereas during the cold war deterrence occupied the central role, in today’s world it is multidimensional and is part of the larger mix of strategies for war prevention that should include both coercive and noncoercive approaches. Even though deterrence as we knew it during the cold war era may be passé, in its new forms it still reigns as the most significant coercive strategy for war prevention.

notes 1. For example, Keith Payne, Deterrence in the Second Nuclear Age (Lexington, KY: University Press of Kentucky, 1996); and Payne, The Fallacies of Cold War Deterrence and a New Direction (Lexington, KY: University Press of Kentucky, 2001). On the complexities of deterrence in the twenty-first century, see George Quester, Nuclear First Strike (Baltimore: Johns Hopkins University Press, 2006), 1–52; Scott D. Sagan, “The Commitment Trap,” International Security 24 (Spring 2000): 85–115; Derek D. Smith, “Deterrence and Counterproliferation in an Age of Weapons of Mass Destruction,” Security Studies 12 (Summer 2003): 152–97. 2. President of the United States, “The National Security Strategy of the United States of America,” September 2002; White House, “National Strategy to Combat Weapons of Mass Destruction,” December 2002. 3. On the absolute character of the weapon, see Bernard Brodie, ed., The Absolute Weapon: Atomic Power and World Order (New York: Harcourt, Brace, 1946); T. V. Paul, Richard J. Harknett, and James J. Wirtz, eds., The Absolute Weapon Revisited: Nuclear Arms and the Emerging International Order (Ann Arbor: University of Michigan Press, 1998). 4. Robert Jervis, “Deterrence Theory Revisited,” World Politics 31 ( January 1979): 289–324; and The Meaning of the Nuclear Revolution (Ithaca, NY: Cornell University Press, 1990). 5. Exceptions include George Quester, Deterrence before Hiroshima (New York: John Wiley & Sons, 1966); John J. Mearsheimer, Conventional Deterrence (Ithaca, NY: Cornell University Press, 1985).

24

t. v. pa u l

6. In general deterrence, opponents keep military forces for future contingencies with the implicit understanding of their use (as they expect no imminent challenge), whereas in immediate deterrence, a crisis situation has arisen, the prospects for war are high, and a retaliatory threat is what prevents war initiation. Extended deterrence involves the protection of an ally by a stronger state using retaliatory threat. Deterrence by denial is obtained by convincing an opponent that it would be denied politicomilitary victory, while deterrence by punishment is obtained when an enemy forgoes an attack fearing unacceptable punishment. For definitions of these different types of deterrent relationships, see Patrick Morgan, Deterrence: A Conceptual Analysis (Beverly Hills: Sage, 1983), 30; Glenn H. Snyder, Deterrence and Defense (Princeton: Princeton University Press, 1961). For more nuanced conceptions of deterrence involving third-party efforts at preventing war between two states that are friendly to the dominant state, see Timothy Crawford, Pivotal Deterrence (Ithaca, NY: Cornell University Press, 2003). 7. Alexander L. George and Richard Smoke, Deterrence in American Foreign Policy (New York: Columbia University Press, 1974); Robert Jervis, Richard Ned Lebow, and Janice Gross Stein, eds., Psychology and Deterrence (Baltimore: Johns Hopkins University Press, 1985). 8. John Lewis Gaddis, The Cold War: A New History (New York: Penguin, 2005). For a counterargument, see John Mueller, Retreat from Doomsday: The Obsolescence of Major War (New York: Basic Books, 1989). On testing deterrence success and failure, see Frank P. Harvey, “Practicing Coercion: Revisiting Successes and Failures Using Boolean Logic and Comparative Methods,” Journal of Conflict Resolution 43 (1999): 840–71. 9. T. V. Paul, Power Versus Prudence: Why Nations Forgo Nuclear Weapons (Montreal and Kingston: McGill-Queen’s University Press, 2000); Mitchell Reiss, Bridled Ambition: Why Countries Constrain Their Nuclear Capabilities (Washington, DC: Woodrow Wilson Center Press, 1995). 10. Glenn H. Snyder, “The Balance of Power and the Balance of Terror,” in The Balance of Power, edited by Paul Seabury (San Francisco: Chandler, 1965), 185–201. See also Michael Krepon, “The Stability-Instability Paradox, Misperception, and Escalation Control in South Asia,” (Washington, DC: The Henry L. Stimson Center, May 2003). 11. Sumit Ganguly and Devin T. Hagerty, Fearful Symmetry: India-Pakistan Crises in the Shadow of Nuclear Weapons (Seattle: University of Washington Press, 2006); Rajesh Rajagopalan, Second Strike (New Delhi, Penguin, 2005). On the Kargil conflict, see Peter Lavoy, ed., Asymmetric War in South Asia (Cambridge: Cambridge University Press, forthcoming). 12. Department of State, “North Korea—Denuclearization Action Plan,” February 13, 2007, http://www.state.gov/ r/ pa / prs / ps / 2007 / february/ 80479.htm. 13. See Mark Juergensmeyer, Terror in the Mind of God: The Global Rise of Religious Violence (Berkeley and Los Angeles: University of California Press, 2003); Robert A. Pape, “The Strategic Logic of Suicide Terrorism,” American Political Science Review 97 (August 2003): 343–61. 14. Patrick M. Morgan, Deterrence Now (Cambridge: Cambridge University Press, 1993), 4. 15. It is true that in the U.S.-Soviet deterrence context, there was some recognition of the possibility for irrationality and “the ultimate solution adopted to the credibility problem in mutual nuclear deterrence was the inherent capacity for the irrationality of the antagonists so as a threat to use nuclear weapons had some inherent credibility” (Patrick Morgan, personal e-mail communication, May 22, 2007). However, the solution adopted for the credibility problem included hair-trigger alerts, launch on warning, and massive arms buildup, conditions that are not present in much of the current nuclear deterrent contexts. 16. For instance, Christopher H. Achen and Duncan Snidal, “Rational Deterrence Theory and Comparative Case Studies,” World Politics 41 ( January 1989): 143–69. 17. On the calculations of weaker states facing superior powers, see T. V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge: Cambridge University Press, 1994); Derek D. Smith, Deterring America: Rogue States and the Proliferation of Weapons of Mass Destruction (Cambridge: Cambridge University Press, 2006), 26–27; Ivan Arreguín-Toft, How

complex deterrence

25

the Weak Win Wars: A Theory of Asymmetric Conflict (Cambridge: Cambridge University Press, 2005). 18. Max Weber, Economy and Society: An Outline of Interpretative Sociology, volume 1, edited by Claus Wittich and Guenther Roth (Berkeley: University of California Press, 1978), 24–25. See also Ashutosh Varshney, “Nationalism, Ethnic Conflict and Rationality,” Perspectives on Politics 1 (March 2003): 85–99. 19. Morgan, Deterrence Now, 9. It is also true that a broadened deterrence policy would prevent crisis and crisis escalation even among nonrivals. However, the most salient deterrent relationships would involve rivals. 20. In recent years, complexity theory has assumed major importance in the study of systems, especially biological systems. On how complexity is working in different scientific fields, see the special issue of Science, 284, no. 5411 (April 1999). See also, Roger Lewin, Complexity: Life at the Edge of Chaos (New York: Macmillan, 1992); For social science applications, see Paul Cilliers, Complexity and Postmodernism: Understanding Complex Systems (London: Routledge, 1998); Ilya Prigogine, The End of Certainty (New York: Free Press, 1997); and Thomas Homer-Dixon, The Upside of Down: Catastrophe, Creativity, and the Renewal of Civilization (Toronto: Alfred A. Knopf Canada, 2006). For an application to international relations, see Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton, NJ: Princeton University Press, 1997). Components of the complex systems are “never quite lock in place, and yet never quite dissolve into turbulence, either . . . new ideas and innovative genotypes are forever nibbling away at the edges of the status quo,” and even the best-entrenched system can be overthrown over a period of time (M. Mitchell Waldrop, Complexity: The Emerging Science at the Edge of Order and Chaos [New York: Simon and Schuster, 1992], 12). The pivotal example of such a change in internationals relations was the collapse of the Soviet Union and the end of the cold war without much forewarning. 21. James Wirtz, personal e-mail communication, May 22, 2007. 22. There are other possible types: deterrence between nuclear powers and states seeking weapons of mass destruction (WMD) capabilities, conventional deterrence within nonnuclear dyads such as Armenia and Azerbaijan, conventional deterrence of external nonstate actors by states, or internal deterrence involving insurgent groups and governments. Some of these types of deterrence relationships are discussed in various chapters but not extensively because of space considerations. 23. T. V. Paul, “The Risk of Nuclear War Does Not Belong to History,” in The Waning of Major War, edited by Raimo Väyrynen (London: Routledge, 2006): 113–32. 24. John J. Mearsheimer, The Tragedy of Great Power Politics (New York: W. W. Norton, 2001). 25. Morgan, Deterrence Now, 242. 26. For various reasons for the absence of war including liberal causes, economic causes, and ideational changes, see various chapters in Väyrynen, The Waning of Major War. 27. This conflict challenged many axioms of the cold war deterrence framework: nuclear-armed states will refrain from war, expecting no military victory; nor will they use military coercion to change territorial status quo; and conventional capability is not significant in political outcomes. Peter Lavoy, “Introduction,” in Lavoy, Asymmetric War in South Asia. It was coercive diplomacy and compellence that forced the attacker in this instance to withdraw. 28. On this, see Subhash Kapila, “India’s New ‘Cold Start’ War Doctrine Strategically Reviewed,” (Paper No. 991, South Asia Analysis Group, May 4, 2004). Such a strategy, although it has some similarities to NATO’s war doctrine toward the Soviet bloc during the cold war, is more of a compellence approach, even though it is aimed at bolstering India’s deterrent / defensive posture. 29. S. Paul Kapur, Dangerous Deterrent: Nuclear Weapons Proliferation and Conflict in South Asia (Stanford: Stanford University Press, 2007). See also Saira Khan, “Nuclear Weapons and the Prolongation of the India-Pakistan Rivalry,” in The India-Pakistan Conflict, edited by T. V. Paul

26

t. v. pau l

(Cambridge: Cambridge University Press, 2005), 156–77; Timothy D. Hoyt, “Kargil: The Nuclear Dimension,” in Lavoy, Asymmetric War in South Asia. 30. On this debate, see Scott D. Sagan and Kenneth N. Waltz, The Spread of Nuclear Weapons: A Debate (New York: W. W. Norton, 1995), 1–45; Jeffrey W. Knopf, “Recasting the Proliferation Optimism-Pessimism Debate,” Security Studies 12 (Autumn 2002): 41–96. 31. Zeev Maoz, “The Mixed Blessings of Israel’s Nuclear Policy,” International Security 28 (Fall 2003): 207–32. 32. In August 1995, quoting Iraqi officials, United Nations Inspector Rolf Ekeus revealed that in December 1990, on the eve of the war, “they loaded three types of biological agents into roughly 200 missile warheads and aircraft bombs that were then distributed to airbases and a missile site.” This decision was taken after the United Nations Security Council had voted to liberate Kuwait “using all necessary means.” Ekeus states that Iraq decided not to use these weapons following the ultimatum by secretary of state James Baker at the Geneva conference. R. Jeffrey Smith, “U.N. Says Iraqis Prepared Germ Weapons in Gulf War; Baghdad Balked, Fearing U.S. Nuclear Retaliation,” Washington Post, August 26, 1995, A1. Other calculations of Saddam Hussein could have been his inability to overcome U.S. military power and effectively apply chemical weapons, especially after the crippling of the Iraqi command, control, and intelligence networks in coalition attacks. William J. Perry, “Desert Storm and Deterrence in the Future,” in After the Storm: Lessons from the Gulf War, edited by Joseph S. Nye and Roger K. Smith (Colorado Springs: Aspen Institute, 2000), 261. 33. Charles A. Duelfer, “Testimony before the Sub-Committee on Emerging Threats and Capabilities,” Armed Services Committee of the United States Senate, February 2002, http://csis.org. 34. On these possibilities, see Smith, Deterring America, 44. 35. On this, see T. V. Paul, “Great Equalizers or Agents of Chaos? Weapons of Mass Destruction and the Emerging International Order,” in International Order and the Future of World Politics, edited by T. V. Paul and John A. Hall (Cambridge: Cambridge University Press, 1999): 373–92. 36. On microdeterrence, see R. R. Subramanian, “Indo-U.S. Civil Nuclear Cooperation and Its Implications for the Non-Proliferation Regime” (unpublished paper, presented at McGill University, Montreal, March 16, 2007). 37. Pape finds that this type of goal is the most prevalent among suicide terror organizations. See Pape, “Strategic Logic.” 38. On the complexities of nuclear terrorism and deterrence, see Graham Allison, Nuclear Terrorism (New York: Times Books, 2004), 140–202; Rajesh M. Basrur, Minimum Deterrence and India’s Nuclear Strategy (Stanford: Stanford University Press, 2006), 122–47; and Jason J. Castillo, “Nuclear Terrorism: Why Deterrence Still Matters,” Current History 102 (December 2003): 426–31. 39. These forms of deterrence may sometimes be mixed with other forms of strategies such as coercive diplomacy and compellence. 40. Two illustrations of problems in collective-actor deterrence are the current experiences in Afghanistan and Sudan. In Afghanistan, an important Canadian complaint has been that other NATO members do not want to send troops to fight the Taliban. Even when they contribute, they send advisers to safe havens or noncombat troops. The Taliban is not deterred from attacks because they know the lack of commitment on the part of NATO. The same goes for Darfur in Sudan: the aggressors are not deterred because despite many pronouncements on the subject, no state wants to commit a sufficient number of troops for an effective operation. 41. On this, see Norrin Ripsman and T. V. Paul, Globalization and the National Security State (New York: Oxford University Press), forthcoming. 42. On nuclear nonuse, see T. V. Paul, The Tradition of Non-Use of Nuclear Weapons (Stanford: Stanford University Press, 2009); for a more stringent, taboo-like prohibition, see Nina Tannenwald, The Nuclear Taboo (Cambridge: Cambridge University Press, 2008).

complex deterrence

27

43. Steven P. Lee, Morality, Prudence, and Nuclear Weapons (Cambridge: Cambridge University Press, 1996), 124. 44. Kenneth Boulding, The Meaning of the Twentieth Century (New York: Harper and Row, 1964), 81. 45. James G. March and Johan P. Olsen, “The Institutional Dynamics of International Political Orders,” International Organization 52 (Autumn 1998): 943–69. 46. Avner Cohen and Steven Lee, “The Nuclear Predicament,” in Nuclear Weapons and the Future of Humanity, edited by Avner Cohen and Steven Lee (Totowa, NJ: Rowman and Littlefield, 1986). 47. Morgan, personal e-mail communication. 48. For example, see Morgan, Deterrence Now; Lawrence Freedman, Deterrence (Polity Press, 2004); Payne, Deterrence in the Second Nuclear Age and Fallacies of Cold War Deterrence; Frank C. Zagare and D. Marc Kilgour, Perfect Deterrence (Cambridge: Cambridge University Press, 2000); James H. Lebovic, Deterring International Terrorism and Rogue States: U.S. National Security Policy after 9 / 11 (London: Routledge, 2007); Smith, Deterring America; Muthiah Alagappa, ed., The Long Shadow: Nuclear Weapons and Security in 21st Century Asia (Stanford: Stanford University Press, 2008); and Stephen J. Cimbala, Nuclear Deterrence and Arms Control: A New Century (Westport, CT: Praeger, 2002). 49. Christopher Layne, “The Unipolar Illusion Revisited: The Coming End of the United States’ Unipolar Moment,” International Security 31, no. 2 (2006): 7–41. 50. Morgan states that deterrence critics have “tried to develop accurate descriptions of deterrence situations, what preferences, perceptions and judgments are like, how actors define their choices and evaluate them, and why outcomes are too often contrary to what deterrence was to achieve.” Morgan, Deterrence Now, 167. 51. Jeffrey Knopf helped to compose some of these questions.

chapter two R

Three Items in One Deterrence as Concept, Research Program, and Political Issue j e f f r e y w. k n o p f 1

t h e September 11, 2001, attacks on the Pentagon and the World Trade Center ushered in a dual crisis for deterrence. First, September 11 led to new doubts about whether the strategy of deterrence still works. In particular, a threat to retaliate seems unlikely to be effective against people who already plan to kill themselves in a suicide attack. Second, the September 11 tragedy helped bring about an apparent shift away from deterrence in U.S. policy. The George W. Bush administration began emphasizing a new doctrine of preemption, which it applied to rogue states as well as to terrorists. Many observers concluded that the Bush Doctrine meant that deterrence had been cast aside by U.S. leaders.2 Both inferences are incorrect. Although there are limits to deterrence, it is still possible to achieve some deterrence against both rogue states and terrorists. And the Bush administration did not abandon deterrence; it continued to be a component of U.S. national security strategy. The twin challenges posed by suicide terrorism and the U.S. embrace of preemption did, however, prompt many scholars to reexamine deterrence, leading to an outpouring of studies seeking to assess that strategy’s potential against the “asymmetric” threats that have dominated U.S. concerns in the early years of the twenty-first century.3 Despite rhetoric suggesting that “September 11 changed everything,” it is not the case that all pre–September 11 experience was rendered irrelevant. This chapter steps back from the more recent discussions of deterrence to identify certain themes in deterrence theory and practice before the al-Qaeda attacks that could prove helpful in ascertaining potential uses of deterrence 31

32

j e f f r e y w. k n o p f

in today’s security environment. The chapter begins with a conceptual analysis because the basic question raised by both the September 11 attacks and the Bush administration announcement of the preemption doctrine is whether deterrence still works. The answer one gives to this question, however, depends on how one defines the concept of deterrence. The chapter calls for a broader concept than specialists in security studies have traditionally employed. Developing a broader concept of deterrence might help people think more creatively about alternative ways to practice deterrence and hence enhance the chances for identifying effective deterrent strategies. The conceptual analysis is followed by a historical discussion of the role of deterrence as a tool of statecraft, a subject of policy debate, and the focus of an academic research program. This review will show that reservations about deterrence are not new. Changes in the status of deterrence derive partly from changes in the international environment. Yet even when the strategic incentives for using deterrence have been strong, the strategy has faced much criticism in both scholarly research and actual policy debates. Despite this, deterrence has typically remained an element of policy. This history suggests it should not be hastily dismissed in light of current challenges. Politically, deterrence tends to reflect a compromise around people’s second choices. When support grows for either more soft-line or hard-line policy alternatives, those who are skeptical about those alternatives tend to rally behind deterrence. This tendency to evaluate deterrence relative to alternatives shows the continuing relevance of the normative model of rational policy analysis. This chapter closes with some suggestions for how social science research could assist rational policy analysis with respect to deterrence.

assessing the role of deterrence in contemporary security policy There are three important reasons to assess the role that deterrence can play against contemporary threats. First, if a preventive war doctrine leads U.S. leaders to dismiss deterrence more completely than they need to, this could lead the United States into unnecessary wars. If the United States attacks or invades a country that could have been contained and deterred, the United States will pay potentially high and avoidable costs in blood, treasure, and diplomatic friction. The better a state’s ability to recognize when deterrence is likely to work, the better its chances of saving preventive action for only those cases where it is most likely to be advisable.

three items in one

33

Second, even if one has grave doubts about deterrence and strongly supports preventive war, it is not possible to undertake preventive military action against all potential threats simultaneously. While some threats are being eliminated, others will have to be contained and deterred until they can be dealt with. Hence, even supporters of preventive war should not dismiss deterrence and should instead be interested in identifying where it has the best chances of working in order to set priorities for more aggressive action. Third, because deterrence remains an element of policy, it is important to consider how to make it work best. Efforts to identify new ways of achieving deterrence, if they exist, would thus be valuable. Deterrence is not always appropriate, and employing it in such situations can be counterproductive. For example, when dealing with an actor that is motivated by insecurity, deterrent threats might only heighten its sense of insecurity and provoke it to escalate or attack in response.4 Given current threats, however, there are situations in which deterrence could be crucial, provided it can be made reasonably effective in today’s complex international system.

the concept of deterrence: reasons for a broad approach It is possible to distinguish the concept of deterrence from the variety of strategies for practicing deterrence.5 Adopting a broader deterrence concept would have policy benefits. There are still security threats against which deterrence might be necessary and less costly than preventive war, provided deterrence can be made effective. There are considerable doubts, however, about whether traditional deterrence will work against contemporary threats, especially those posed by nonstate actors. This makes it worthwhile to consider whether there are new or different ways to practice deterrence. Broadening the concept of deterrence may help us imagine new ways to practice deterrence that have better odds of success or lower costs. Broadening the concept of deterrence is also intellectually defensible for three reasons. First, deterrence already has been long employed in a much wider range of circumstances than is generally recognized in the international relations (IR) literature. Second, a broader definition is compatible with certain uses of the term in ordinary language. Third, recent developments in U.S. strategy are already moving toward a broader conception of deterrence.

34

j e f f r e y w. k n o p f

The Diversity of Deterrence Applications The first reason to embrace a broad understanding of deterrence is that this reflects reality. Deterrence has been employed since long before the cold war and in situations far beyond world politics. As an element of statecraft, deterrence has been around at least since the Roman general Vegetius declared, “If you want peace, prepare for war.” Moreover, while IR scholars think of deterrence in the context of trying to prevent military attacks, this is not the only context in which it is utilized. Deterrence is also important in criminology, where law enforcement measures are evaluated in part in terms of their ability to deter crime. Lawrence Freedman, as well as Robert Jervis in his chapter for this volume, both point out that attempts to deter even extend beyond the human race to other species. Freedman cites various insects that use deceptive markings and other techniques to persuade potential predators to keep their distance.6 Because terrorism occupies a gray zone between crime and interstate war, ideas about deterrence that have been developed in other contexts might be helpful in dealing with current asymmetric threats, and we should not limit our thinking about deterrence to the stock of ideas developed in strategic studies. Given the variety of contexts in which deterrence has arisen over time, it would be appropriate to have a concept of deterrence broad enough to capture the different ways in which deterrence has been employed. Ordinary-Language Analysis A second reason for thinking broadly comes from stepping outside IR research to consider ordinary language. There is some irony here because, in contrast to many terms used by IR scholars, we might expect the meaning of deterrence to be widely understood. Ask the average person on the street what an “epistemic community” is and odds are he or she will look at you blankly if not hasten to get away. Yet nearly every layperson could offer a definition of deterrence, and their definitions would likely all be about the same: threatening to do something that will make the costs of an action outweigh the benefits so as to convince others not to do it. Our very familiarity with the word “deterrence” threatens to unduly narrow our understanding of it. To see this, consider that the same root can be expressed as a noun or a verb. The terms “deterrence” and “deterrent” are usually nouns, or sometimes adjectives (for example, a deterrence strategy). The concept, however, also comes in a verb form: “to deter.” Logically, these

three items in one

35

should all have congruent meanings. Utilizing the passive voice, imagine an actor who is deterred. We could inquire about what caused this actor to be deterred. Anything we then identify that actually served to deter an actor might be something that could be utilized deliberately as a form of deterrence. The usages that exist in ordinary language make it clear that the category of things that can be described as “a deterrent” should be equivalent to the list of things that can act “to deter.” If the policy goal is to be able to deter, it is worth thinking broadly about the types of measures that might produce this effect rather than limiting ourselves to a single, narrow definition of deterrence. This cuts against the grain of much theoretical writing on deterrence. As with any concept that becomes important in academic study, there can be debate over whether or not to restrict usage to a narrow definition. In strategic studies, the preference is usually to maintain a narrow definition. The most thorough and careful discussion of the concept of deterrence remains the work of Patrick Morgan. Morgan argues deterrence should be defined as “the threat of military retaliation to forestall a military attack.” If we include other means, we risk conflating deterrence with other types of preventive measures, while if we include other ends we may make deterrence equivalent to foreign policy as a whole.7 This narrow definition made some sense in the context of cold war debates. During the cold war, deterrence came to be associated with the effort to prevent nuclear strikes by threatening massive nuclear retaliation. In effect, the connotations associated with massive retaliation and the subsequent situation of mutual assured destruction (MAD) displaced all other denotations of the word “deterrence.” To limit the definition to this usage, however, is to reduce the concept of deterrence to just one form of deterrent strategy. Morgan is right that not every effort at prevention should be considered deterrence. For example, there are efforts underway to delegitimize terrorism so that many who might support it today become convinced that it is morally wrong in any circumstance; these are a form of suasion not deterrence. Nevertheless, there are other steps that are being taken or that could be taken whose aim is to deter, even though they do not rely on threatening massive punishment in retaliation for a military attack. For instance, efforts to harden potential terrorist targets are intended in part to increase the chances that a terror operation will fail if it is attempted; such efforts also are intended to convince terrorists that such a prospective operation actually will fail, in the hope that the prospect of failure might deter them from launching an operation it in the first place.

36

j e f f r e y w. k n o p f

There is a tension here between the logic of academic social science and the logic of policy. In social science, scholars are admonished to select on the independent and not the dependent variable, to define terms precisely, and to beware of conceptual stretching. All of this aids in the tasks of hypothesis testing and generalization. If deterrence is defined narrowly as using the threat of military retaliation to prevent military attack, it is possible to identify like cases for comparison or large-N data analysis. From these sorts of studies, we can learn a great deal about military threats, the overall likelihood they will work, other variables correlated with their success or failure, and the features of military threats that are likely to make them more effective. Such knowledge is policy relevant but does not provide a complete policy analysis. The threat of military retaliation is just one option that policy makers can select, and it has to be compared against other options. Some of these other options also might have the potential to deter an unwanted action; in some cases, they might work better than threats of highly punishing retaliation. If there are other ways to deter, they should also be considered deterrent options. In a full policy analysis, the various deterrent options would be evaluated against nondeterrent options. If the goal is to deter some unwanted action, then any measure that might contribute to achieving the goal should be considered a possible form of deterrence. Studying the effects of military threats is useful, but it is not equivalent to studying deterrence. In addressing today’s complex-deterrence situations, policy concerns make it advisable to embrace a broader definition of deterrence than normal social science would recommend. If there are alternatives to the use of traditional military threats that are morally acceptable and that could help deter terrorism or the use of weapons of mass destruction (WMD), it is important to identify them. Traditional definitions of deterrence facilitate the social science objective of cumulating knowledge within an existing research program. In the policy realm, however, embracing a narrow concept could limit our ability to imagine alternative approaches and unnecessarily restrict the range of options we analyze. If there is a nontraditional approach that would be an effective deterrent, it could provide security while making it possible to avoid preventive war. This makes it imperative not to reduce the concept of deterrence to a single, narrow strategy. New, nontraditional approaches would not replace traditional deterrence—threats of military retaliation are probably still an effective way to deter rogue states from launching WMD attacks. For some other activities, however, threats of military retaliation

three items in one

37

might not be credible, and it would help to have a broad framework that might assist in identifying other deterrence options. Deterrence Defined The suggestion to adopt a broader notion of deterrence has a parallel in recent debates over narrow versus broad definitions of security. Here as well it has proven useful to separate the underlying concept from its varied applications. David Baldwin has shown that there is a core concept of security that is recognizable in different applications of the concept but is still delimited in a way that distinguishes security from other values that might serve as policy goals.8 This means that adjectives become important as a way to distinguish different forms of security. National security, international security, and human security all involve a concept of security, but the adjectives specify different referent objects (the state, the system, and individuals, respectively) and imply different threats (military attacks by external actors, instability, and actions that harm people within a state, respectively). This suggests that scholars should start with an abstract, generic concept of deterrence and then use adjectives or other modifiers to indicate how deterrence will be applied in a particular context. With this approach, it is still possible to talk about and examine deterrence in the way scholars in security studies think about it, without conflating it with other possible approaches to deterrence. One could label it “traditional deterrence” or “classical deterrence” in those situations in which the addition of an adjective would help clarify the type of deterrence to which one is referring. Conceptually, deterrence is a form of preventive influence that rests primarily on negative incentives. When used as a conscious strategy, deterrence aims to prevent an action that another party might undertake. This distinguishes deterrence from the alternative coercive strategy of compellence, which aims to compel the target to take a new action or to stop or undo an action already underway. (However, it is not always easy in practice to tell whether the goal is deterrence or compellence, and policy may involve both simultaneously.) Deterrence is a form of influence because it operates by affecting an actor’s decision making. The way it works is more psychological than physical. In a classic discussion, Thomas Schelling distinguished between brute force and coercion; Lawrence Freedman has updated this description with a distinction between control strategies and coercive strategies.9 Control

38

j e f f r e y w. k n o p f

strategies seek to take away any choice for the other side by eliminating its ability to act or establishing physical domination. Deterrence, in contrast, leaves the other side with the ability to make a choice but conveys certain threats as a way to influence that choice. This distinguishes deterrence from some other forms of prevention. For example, military preemption is a control strategy that aims to take away the other side’s ability to launch an attack. A pure defense posture also does not aim to influence the other’s decision; it accepts that an attack may occur and aims only to ensure the attack will not succeed. Lastly, deterrence emphasizes negative inducements. Preventive influence might be pursued through the promise of rewards or offer of reassurances, but these represent positive inducements. Deterrence, in contrast, arises mainly because an actor expects a course of action to lead to a negative outcome. Deterrence success often requires pairing a threat with certain positive messages or inducements, so deterrence strategy usually does not rest exclusively on negative incentives, but the essence of deterrence is the target’s anticipation that taking a certain action will lead to a negative result. Since Glenn Snyder’s seminal work, most specialists have recognized at least two distinct paths to deterrence: punishment and denial.10 The threat to impose costs in retaliation for a transgression, which is deterrence by punishment, obviously involves a negative incentive. But deterrence by denial also rests upon a form of negative incentive. The ability to resist and ultimately frustrate another actor’s efforts can deny it any benefits while still leaving it with the costs of its efforts, again leading to a net negative outcome for the other side. An expectation of such a negative result is the key mechanism involved in deterrence. The relevant negative incentives must involve the potential aggressor’s own interests, values, or objectives. Concern about the well-being of the defender is usually not part of the deterrence equation. Hence, if a person says to a friend or loved one, “If you do this, you will hurt my feelings,” it might persuade the other person not to act, but we would not normally describe the friend or loved one as having been deterred. Deterrence involves influencing calculations of self-interest (although the self in question can involve a collective identity, such as one’s family, community, nation, or the like); changing behavior by appealing solely to other-regarding concerns is a different influence mechanism, separate from deterrence. In deterrence, it is fear of negative consequences for themselves or their goals in response to their actions that persuades other people not to act. Thus, deterrence involves anything that prevents (or attempts to prevent)

three items in one

39

an actor from taking an action by influencing its decision making through its anticipation that the action will lead to a negative result for its own interests or objectives. This generic concept or situation of deterrence can be distinguished from the strategy of deterrence. There might be situations in which an actor is deterred without anyone having tried to send a deterrent message or signal. Even where there is no conscious attempt to employ a strategy of deterrence, a condition of deterrence can still result. Actor B might decide for itself that a path of aggression is likely to end badly, and actor A might never have the slightest inkling that actor B ever contemplated hostile action. Deterrence arises when a possible course of action is prevented because the anticipation of negative consequences influences an actor’s decision making, even if no other actor took steps meant to bring about such a result. In any situation in which an actor was deterred (the verb form), it should logically be valid to say there was deterrence (the noun form). Some existing distinctions in the literature involve a similar differentiation between active efforts at deterrence and deterrence as more of a background condition. Morgan’s influential notion of “general deterrence” and Freedman’s more recently coined term “internalized deterrence” both point to situations in which deterrence is maintained without much explicit issuing of deterrent threats.11 Although deterrence might be unintended on the defender’s part, the questions of interest for this chapter mainly concern consciously chosen, deliberate efforts to deter. Where the intention to deter exists, we have a strategy of deterrence. Broadening Impulses in U.S. Strategy The preceding discussion of deterrence as a concept says little about either the means used or the ends sought—it does not require that either be military in nature. Recent U.S. military doctrine has utilized an “ends, ways, and means” framework that can be helpful here. In fact, this framework has already been applied to deterrence in a guidance document produced by the U.S. Strategic Command, the “Deterrence Operations Joint Operating Concept” (DO-JOC).12 As a generic concept, deterrence implies that the end sought is to prevent something, but it is open as to what that something might be. It could be to prevent an armed attack, but it could also be to avert states from giving assistance to terrorists or helping other states acquire WMD, two further objectives that are clearly relevant to national security. Beyond being preventive, however, the ends sought can be wide-ranging; hence, deterrence is most

40

j e f f r e y w. k n o p f

distinctly associated with the “ways” part of the “ends, ways, and means” formulation. It is a way of preventing an unwanted action by creating the ability to respond or threatening to respond in a manner that will likely produce negative results if some other actor takes a specified action. From this perspective, punishment and denial can be considered to be two different ways of practicing deterrence.13 There are other ways of trying to prevent something (such as moral suasion or pure defense), and there are other ends that can be sought besides prevention (such as compelling an action or eliciting cooperation), so this formulation still bounds the concept of deterrence without tying it specifically to the cold war understanding of the term. One also can utilize different means in seeking deterrence. Strategic deterrence has historically involved military means, but there is no logical reason why economic or informational or law enforcement measures cannot be used to achieve deterrent effects. In fact, U.S. doctrine already seeks to use a variety of instruments to produce deterrent effects. The DO-JOC, for example, states that sometimes nonmilitary means will be needed to achieve deterrence.14 Efforts to improve consequence management in the event of an attack involving chemical, biological, or nuclear weapons, for instance, could bolster deterrence. U.S. strategy documents since September 11, 2001, have regularly listed improving the ability to minimize the effects of WMD as a part of efforts to deter the use of these weapons.15 This provides a third argument for adopting a broad concept of deterrence: the fact that U.S. policy has already moved, albeit hesitantly and incompletely, to contemplate deterrent efforts that are not based on the threat of military retaliation. Because real-world security strategies are already moving in the direction of alternative approaches to deterrence, it makes no sense for scholars who study security to limit themselves to a definition that is narrower than the actual policies that they are supposed to be studying. Although nonmilitary means could produce deterrent effects, there is still something distinctive about the threat or use of military force.16 Threats of force will in most cases have a greater, more disturbing psychological impact than the use of nonlethal means, and if the threat is implemented this will usually result in greater destruction and more far-reaching political consequences. Hence, it is important not to group together military and nonmilitary approaches to deterrence in a way that would elide any distinction between them. Where relevant, appropriate adjectives or other descriptors should be used to clarify the means being considered. Hence, deterrence as conventionally understood in IR could be labeled “military de-

three items in one

41

terrence.” Other approaches to deterrence could be labeled “nonmilitary” or else described more specifically, using phrases such as “deterrence through improved physical protection” or “deterrence by threat of sanctions.” The distinctiveness of strategies that involve the threat or use of force actually points to a possible advantage of adopting a broad concept of deterrence. Because this approach encourages the use of adjectives to clarify the type of deterrence in question, it might make use of the term “military deterrence” more common in situations in which people would usually employ the word “deterrence” on its own. This would highlight the fact that a threat of force is involved and make it more explicit when military means, with their distinctive implications, are being used as the basis of deterrence. As a concept, deterrence can involve anything that influences an actor not to do something based on the actor’s expectation it will get a negative result (or any attempt to exert such influence). To turn deterrence from an abstract concept to a specific strategy or tactic, one has to become more specific about the ends, ways, and means. What ends does one seek to prevent, what means will be employed to do so, and in what way or ways will one seek to influence decision making? For most of the twentieth century, deterrence as an element of national security strategies was military in both ends and means: it used military means to seek the end of preventing military attacks by states. Historically, the ability to defend made possible a degree of deterrence by denial. With the advent of nuclear weapons and long-range missiles, however, states could not find a way to guarantee effective defenses, so the “ways” part of the equation shifted to deterrence by punishment. When a state cannot count on stopping a nuclear-armed missile from reaching its target, it tries to prevent such missiles from ever being launched by threatening retaliation in kind, creating the nuclear deterrence strategy that dominated the cold war. Nothing, however, inherently limits deterrence to retaliation in kind. The end of the cold war and the rise of the terrorist threat have led to a deemphasis on this one form of deterrence but not to the death of the concept. Instead, deterrence has been in transition, as the ends, ways, and means have all been in flux. Indeed, the list of potential national security ends for deterrence has been expanding. In addition to preventing military attacks launched by states, the United States and many other international actors are now concerned about preventing terrorist attacks mounted by nonstate actors. They also are concerned about preventing both states and private actors from assisting terror networks. And they want to prevent rogue states and terrorist organizations from acquiring or transferring WMD not just from

42

j e f f r e y w. k n o p f

using them. Far from being obsolete, deterrence is potentially being asked to do more than it has ever done. This makes it all the more important to understand why there is so much skepticism about deterrence. One major reason is the psychological impact of the September 11, 2001, terror attacks, which called into question the efficacy of relying on deterrence.17 Placing recent debates in a broader historical context, however, will show that there have often been doubts about this strategy.

deterrence as policy choice: lessons from its diplomatic, political, and intellectual histories Some critics of the Bush administration expressed shock that the administration had, in their view, jettisoned decades of U.S. commitment to strategies of containment and deterrence.18 There are good reasons to believe that the administration was too skeptical about the prospects for deterring rogue states and overstated the necessity for preemption.19 Many critics, however, ignored the larger historical picture. There is no basis for expecting that deterrence will remain forever ensconced as the lead element in a state’s security strategy. Strategy, if rationally chosen, is a response to the international security environment. It makes sense that the importance of deterrence will vary as circumstances change; for purely strategic reasons, we should not expect the role of deterrence to be a constant. Moreover, even in periods during which deterrence is important strategically, such as during the cold war, support for deterrence is often shaky. The U.S. commitment to containment and deterrence has never been unwavering. Reservations about deterrence have been expressed not just in the political arena but in the scholarly arena as well. For all these reasons, the status of deterrence can be both uncertain and variable. A particular deterrence strategy is unlikely ever to be set in stone. Conversely, the same factors that explain why deterrence does not usually go unchallenged as the lead element in national security strategy also make it clear that deterrence is unlikely ever to go away completely. Deterrence is a choice that emerges out of a strategic, political, and intellectual environment. Whether or not it is a good choice depends on what actors want to achieve and what other options are available. Sometimes deterrence will be the best alternative; sometimes it will not be the ideal way to deal with emerging threats. Ex cathedra pronouncements either that we need to hold fast to an existing strategy of deterrence or that deterrence no longer applies and should be abandoned are worse than useless—they are harmful.

three items in one

43

It is more realistic to expect that states will sometimes find deterrence an appropriate choice and sometimes not. Scholars typically treat deterrence as an independent variable. They inquire about whether or not a deterrence strategy is likely to work and the conditions under which it is more or less effective. One can, however, also make deterrence the dependent variable and ask when states will choose to make deterrence a major component of national strategy. As with many foreign policy choices, the commitment to deterrence can vary because of strategic factors as emphasized by realists, domestic political factors as emphasized by second-image theorists, and ideational factors as emphasized by constructivists and first-image theorists. All three sets of factors have contributed, at various points in time, to discouraging a reliance on deterrence. Variations in the Strategic Context Strategic incentives to adopt deterrent strategies have never been entirely absent, but they have varied in intensity over time. Historically, the centrality of deterrence in cold war strategy is an anomaly. Deterrence has rarely been the overriding strategic goal for major powers. For most of history, the leading powers were expansionist. They sought to expand their borders and influence and to acquire empires. These goals required offensive strategies and made deterrence a distinctly secondary concern, although fending off others’ offensive attacks did create some concern with defensive strategies. Thus, the classic nineteenth-century study of strategy by Clausewitz discusses both offense and defense—but not deterrence.20 This does not mean deterrence was irrelevant. The classic European balance-of-power system can be considered a mechanism of deterrence: to the extent that alliance commitments remained flexible, a potential aggressor would be prudent to consider whether expansionism on its part would trigger a superior counterbalancing alliance. But European diplomats of this period did not focus primarily on the potential deterrent effects of alliances. They perceived the balance of power as a means to defend themselves and contain potential hegemons, and there is little writing on strategy in this period that delves into the inner workings of deterrence. Eventually, popular attitudes began to shift from acceptance or even passionate embrace of imperialism to disapproval of war and aggression. At the same time, some of the leading states became status quo powers. These shifts made defense more important but still did not require states to think much about deterrence per se. Strong, effective defenses would naturally exert a

44

j e f f r e y w. k n o p f

deterrent effect, reflecting the logic of deterrence by denial, but deterrence was a side benefit of the commitment to defense, not a primary goal of strategy in its own right. Hence, even though the concept of deterrence is ancient, explicit efforts to employ deterrence strategies and develop theories of deterrence only became prominent during the twentieth century. This process began with the advent of air power. Beliefs that “the bomber will always get through” gave rise to a view that security might have to be based on deterrence.21 It was the development of nuclear weapons, however, that spurred systematic theorizing about deterrence as a way to avoid nuclear war. Although the atom bomb was sufficient to trigger Bernard Brodie’s famous observation that henceforward the goal of military forces would no longer be to win wars but to avert them, it took the subsequent development of the hydrogen bomb and the long-range ballistic missile to move deterrence fully to center stage.22 In a situation in which there were no effective defenses against ballistic missiles and good reasons to doubt that either disarmament or a first-strike knockout blow would be successful, most leaders came to the conclusion that deterrence was the only feasible strategic option. The unimaginable destructive power of the H-bomb made the recourse to deterrence not just desirable but imperative. If strategic considerations promoted a lead role for deterrence during the cold war, it is natural that the end of that confrontation might cause deterrence to be downgraded in U.S. national security strategy. The implications of strategic considerations in the post–cold war world, however, are not necessarily so clear-cut. The proliferation of chemical, biological, and nuclear weapons to so-called rogue states and the September 11 terrorist attacks have created new security challenges. Given new threats to security, one might predict that the role of deterrence would be restored. To the contrary, however, there has been no unambiguous return of deterrence to center stage. Many question whether deterrence can be effective against the new threats, especially religiously inspired suicide terrorism. Because the relevance of deterrence as a response to contemporary security threats has become the subject of considerable scholarly and public debate, probably the most accurate conclusion is that strategic considerations have become indeterminate when it comes to predicting the importance of deterrence in national defense policy. The shift from a bipolar to a unipolar system adds further ambiguity.23 As the world’s only remaining superpower, the United States has sought to maintain its freedom of action. Accordingly, the United States has not

three items in one

45

wanted smaller states to acquire the ability to deter U.S. action. This creates a systemic incentive to downplay the relevance of deterrence in the current international system, although in practice the United States has given indications that it can be deterred by regional states that acquire nuclear weapons or the ability to direct the activities of terrorist organizations. Two conclusions follow from the analysis in this section. First, the role of deterrence is historically contingent because strategic incentives in the international system can vary. The range of contingency, however, is bounded. On the one hand, use of deterrence in international relations is unlikely to disappear altogether as long as states remain important actors and the system stays anarchic. There will always be potential security threats for which deterrence might be an appropriate response. On the other hand, deterrence is unlikely ever to be embraced wholeheartedly and unreservedly as the only element of national security strategy. As an influence strategy and not a control strategy, deterrence leaves the final strategic choice up to the other side—even in the face of what should be a credible deterrent threat, it can still choose to attack and deterrence can therefore fail. It is always uncomfortable to live with a risk of deterrence failure, and there will always be a temptation to seek to eliminate the risk. For states with sufficient power, desire to assert control and not merely influence can lead them to attempt preemption or even preventive war. This leads to this section’s second conclusion: one cannot predict purely from an analysis of the current security environment whether or not the United States or other major powers will make deterrence a major goal of their strategies. There are systemic pressures to continue to employ deterrence, but there are also contrary pressures to dismiss the relevance of deterrence and to embrace more controlling strategies. The net result of strategic incentives in the complex post–September 11 international system is hence ambiguous. As a consequence, the status of deterrence is also being determined by other factors, including political and ideational factors. Consideration of the political and intellectual history of deterrence will show that conditions already existed that would have inclined many Americans to be skeptical about deterrence even before September 11. Policy Debates about Deterrence Doubts about deterrence long predate the September 11 attacks. Indeed, before the nineteen hijackers carried out their deadly mission, faith in deterrence had been considerably weakened by criticisms that arose in both

46

j e f f r e y w. k n o p f

public policy debates and academic research. Because of the legacy of past debates, many Americans would already have been inclined to seek alternatives to deterrence when the shock of September 11 gave them extra motivation to do so. Resistance to a reliance on deterrence was present in the political arena from the start of the cold war. Deterrence became prominent in part because of its importance in avoiding nuclear war, but it also received a boost because it fit well with the larger U.S. grand strategy of containment. If the U.S. political goal was to contain Soviet expansion, it made sense for the military strategy to emphasize deterrence of possible Soviet military aggression—not only at the nuclear but also at the conventional level. Yet, from the beginning, deterrence and containment had critics on both the left and the right. Those on the political right saw containment as a sell-out and argued for a strategy of rollback to “liberate the captive nations” in Eastern Europe. The most hawkish military strategists also urged preventive strikes against the Soviet Union before it could achieve its own nuclear strike capability.24 Those on the left, meanwhile, called for being more accommodating of Soviet concerns in an effort to preserve the World War II alliance. They also favored nuclear disarmament over deterrence as the way to avoid nuclear war. These basic splits reemerged periodically throughout the cold war. Social protests in the late 1950s and early 1960s sought to ban nuclear testing or even “ban the bomb.” Even more massive protests against the nuclear arms race emerged in the late 1970s and early 1980s, and some of the movement’s rhetoric criticized the expectation that nuclear deterrence would work indefinitely. The origins of the acronym MAD (for “mutual assured destruction”) also are instructive. The term was coined in the 1960s by Donald Brennan, who meant it as a pejorative label for what he viewed as the crazy logic of deterrence theory; like many conservatives, Brennan argued for building missile defenses instead of relying on deterrence.25 This perspective found a champion in President Ronald Reagan, who launched the Strategic Defense Initiative based explicitly on the belief that the United States could not count on deterrence to work—he called for replacing mutual assured destruction with mutual assured survival.26 Despite their criticisms, however, neither doves nor hawks could identify a convincing replacement for deterrence. Deterrence thus continued to feature prominently in U.S. strategy as a sort of centrist compromise, but it had few advocates who promoted it publicly with enthusiasm. After the end of the cold war and collapse of the Soviet Union, questioning of U.S. reliance on deterrence only increased. The sharpest criticism came

three items in one

47

from missile defense advocates. Although the Soviet threat had disappeared, new security concerns arose regarding countries that came to be known as rogue states. Given the continuing proliferation of WMD and ballistic missiles to these states, supporters of building missile defenses emphasized the potential undeterrability of rogue states, especially those led by fanatical dictators like Saddam Hussein or Kim Jong Il. Such leaders, they claimed, might be so crazy or risk acceptant, and so thoroughly insulated from reality, that they might not be deterred by the U.S. ability to annihilate their countries through nuclear retaliation (or, it might be added, by U.S. conventional superiority).27 The 1990s also witnessed renewed calls for nuclear disarmament, including some by prominent former political and military leaders.28 Their arguments for abolition generally included expressions of doubt that deterrence could prevent the use of nuclear weapons over the long term. Hence, the decade before the al-Qaeda attacks on New York and Washington involved some of the most vigorous challenges yet to deterrence in the political arena, from both the left and the right; it is hard to think of any equally energetic political advocacy for deterrence. The Deterrence Research Program Public debates in the political arena constituted one source of challenges to reliance on deterrence. Scholarly writings gave rise to a second source of doubt. What Jervis has labeled the first two waves of deterrence theory, which together ran from the end of World War II to the 1960s, drew on methodological tools like game theory to develop what became the standard model of deterrence.29 The initial waves of deterrence theorizing assumed both the necessity and the feasibility of deterrence, but eventually criticisms of deterrence arose in the academic community. A 1974 book by Alexander George and Richard Smoke represented the major turning point, inaugurating what Jervis called the third wave of deterrence theorizing. George and Smoke reviewed historical case studies of conventional deterrence failures and argued that deterrence can fail for several reasons, including the fact that states do not always act with the level of rationality assumed by existing deterrence theory.30 Subsequent research by Jervis, Ned Lebow, Janice Stein, and Patrick Morgan, among others, followed in this vein. Collectively, they developed a potent critique of what has come to be known as rational deterrence theory (RDT). Lebow in particular stressed that deterrence can fail even when deterring states do everything right in terms of establishing a credible deterrent: make

48

j e f f r e y w. k n o p f

a clear commitment (i.e., define one’s “red lines”), communicate one’s commitment, have the capabilities to act effectively on the commitment, and demonstrate the resolve to act. Researchers in the third wave claimed the case study evidence showed that even meeting all four conditions is not sufficient to ensure success. In particular, if potential challengers are highly motivated due to the fear that their international position is becoming more unfavorable or due to domestic political considerations, a variety of psychological biases can cause them to dismiss what should be credible signals of a deterrent commitment.31 The third wave is generally associated with the use of a comparative case study methodology and an emphasis on psychological variables. In the evolution of the deterrence research program, however, there is another important change from the first two waves to the third wave that has usually been overlooked. Because the initial waves of deterrence research focused on credibility, they had to emphasize actions taken by the deterrer to explain deterrence success or failure. RDT in effect blamed the victim: if deterrence failed, it would be because the defender did not take the necessary steps to make its deterrent threat credible. Although not given much attention in Jervis’s typology, the first empirical evaluations of RDT used statistical methods. From its beginnings in the 1960s, the statistical literature has largely emphasized structural variables, such as the military balance or, in the case of extended deterrence, the extent of political and economic ties between a protector and the state it seeks to protect.32 Here deterrence success or failure is explained largely in terms of underlying factors that are not under the short-term control of either the defender or the challenger. One further innovation of the third wave is that it put the focus more squarely on the deterree than did the previous research. Because deterrence is an influence strategy, not a control strategy, the proximate cause of deterrence success or failure is the decision of the challenger about whether or not to be deterred. One merit of the third wave is that it drew attention to the wide range of factors that could lead a challenger to initiate an action that the defender had hoped to deter. In light of the limitations of deterrence they identified, most of the scholars in the third wave advocated making greater use of positive incentives, strategies of reassurance, and diplomatic initiatives to address the underlying sources of tensions. In U.S. policy debates, this tended to align them with political liberals. Some strategic thinkers on the conservative side, however, also largely embraced the third-wave critique of RDT. Colin Gray and Keith Payne have been especially important figures in the hawkish case against deterrence.33

three items in one

49

They have cited the same reasons as the third-wave theorists for why deterrence can fail even when RDT would predict success. In addition, however, they have added a strong culturalist strand to the debate. Gray and Payne have argued that other actors will perceive and evaluate deterrent threats in terms of their own cultural values and frames of reference. Such cultural differences can undermine deterrence, they argue, by making it hard for the deterrer to know what values of the other side must be held at risk in order to make deterrent threats effective.34 In cold war debates over nuclear strategy in the 1970s and 1980s, Payne and Gray joined others, such as the historian Richard Pipes, who argued that Soviet leaders might be sufficiently willing to sacrifice the lives of their people that they would not be deterred by the threat of massive retaliation against the Soviet population.35 These critics called for seeking nuclear war–fighting capabilities that could target Soviet leaders directly and ensure Soviet defeat if a nuclear war occurred. In principle, this was not so much a dismissal of deterrence per se as it was an effort to shift strategy from deterrence by punishment to deterrence by denial. In practice, however, the critiques by hawks added to the body of literature that called into question U.S. reliance on deterrence. In the academic literature, others have risen to defend RDT, making two major responses to the critics. These defenses, however, have not fully addressed the sources of unease about relying on deterrence in the policy realm. First, champions of RDT have pointed out that much of the case study literature on deterrence suffers from selection bias: it has looked only at cases of deterrence failure but not deterrence successes.36 This is a problem if one wants to estimate the likelihood that deterrence will work. Looking only at deterrence failures makes it appear that deterrence often fails, but in the full universe of cases for which deterrence is potentially relevant it could instead be true statistically that deterrence works a high percentage of the time. Against an argument that deterrence is usually ineffective, this would be an important counterargument, but many deterrence critics do not make such a sweeping claim. Instead, they emphasize simply that “deterrence can fail,”37 even when states implement a deterrence strategy exactly as RDT prescribes. The historical cases they cite are sufficient to show that one cannot expect a 100 percent guarantee from deterrence. The second main defense of RDT actually reinforces this concern. Supporters of rational-choice analysis argue that many cases of deterrence failure can actually be explained in rational terms.38 They claim that in many of the cases cited in the case study literature the challenging state could rationally have believed that it could get around the deterrent threat and avoid

50

j e f f r e y w. k n o p f

retaliation. In other cases, states could rationally have believed the utility for attacking would be higher than for not attacking, even factoring in the costs of the defender’s retaliation, because of what the attacking state perceived as the even greater costs of not acting. According to Frank Harvey, the four conditions identified in the theory for making deterrence credible should only be considered necessary conditions, not sufficient ones.39 Meeting the conditions still improves the odds of success, Harvey claims, but because they are not jointly sufficient conditions, RDT predicts there will be some deterrence failures even when all four conditions are met. In essence, defenders of RDT say that failures of deterrence as policy do not automatically equate to failures of the theory. This is true, but it completely misses the point for those concerned about policy. For those seeking to make or influence policy, what matters is whether deterrence will succeed or fail. If deterrence does fail, it will be cold comfort to learn that the attack one has just suffered can still be explained in rational terms. If what one seeks from a deterrence strategy is certainty that it will work, the acknowledgment that a perfectly executed strategy can fail even against a rational actor means doubts will remain even if RDT proponents “win” their academic debates over theory and method.40

t h e c a s e f o r p o l i c y a n a lys i s The preceding sections show that deterrence was already on shaky ground when terrorists struck the World Trade Center and the Pentagon in 2001. Deterrence had been questioned by political leaders and social activists for half a century and by an important body of academic research for a quarter century. Debates about the dangers of nuclear proliferation and renewed calls for building ballistic missile defenses or seeking nuclear abolition kept arguments about the limitations of deterrence simmering through the 1980s and 1990s. By 2001, many people had already concluded that deterrence could be unreliable. The fact that the success of deterrence cannot be guaranteed in every situation had an impact in the policy realm. But should the absence of certainty rule out the use of deterrence? In the run-up to the second Iraq war, the Bush administration and its supporters implied the answer was yes. They argued that the risk that deterrence and containment would fail required the United States to undertake a preventive war against Saddam Hussein. But the limitations of one option do not automatically make another option preferable because none of the alternatives to deterrence offers an ironclad

three items in one

51

guarantee of success either.41 The possibility that deterrence might fail does not necessarily mean it should be abandoned. In particular, the potential limitations of a deterrence strategy do not by themselves make the case for preventive war. The likely results of preventive war must also be evaluated in their own right. As with deterrence, preventive war could fail, and the costs and risks of even successful preventive war must be considered. Nor are these the only two options; deterrence and preventive war must be evaluated not only against each other but against other possible alternatives. In short, traditional policy analysis remains necessary. Deterrence should not be analyzed in isolation. The compelling logic of an abstract deterrence model cannot by itself make the case for adopting a deterrence policy, while the potential flaws of deterrence revealed by empirical research do not by themselves make the case for ruling out a deterrence strategy. Because judgment rests on how one perceives deterrence relative to the alternatives, people can sometimes appear to switch sides. Prior to the Iraq war, the conservative columnist Charles Krauthammer highlighted what he saw as an inconsistency in the views of many of the opponents of going to war. He noted that many of those who argued that Saddam could be contained and deterred had been critics of the same strategies during the cold war, and he excoriated “the hypocrisy of the antiwar movement’s current newfound affection for deterrence.”42 If it is hypocrisy to support deterrence in some situations but not in others, however, then the same charge could be leveled at many hawks and conservatives. There are many who stressed the role of U.S. nuclear weapons in deterring the Soviet Union who later questioned the viability of deterring Saddam. One of the most interesting examples of this volte-face is George Shultz. In November 1983, the ABC Network broadcast a made-for-TV movie called The Day After, which presented a grim depiction of the aftermath of a U.S.Soviet nuclear war. After the movie, ABC granted airtime to then secretary of state Shultz, speaking on behalf of the Reagan administration. Shultz made the classic argument for deterrence, reassuring viewers that the United States maintained nuclear weapons only to prevent their use by others.43 In September 2002, however, the former secretary of state published an oped in the Washington Post stressing the urgency of preventive war in Iraq; Shultz’s analysis of the threat posed by Saddam did not even mention deterrence.44 Making the portrait even more complex, in January 2007 Shultz joined three other former high-ranking government officials to endorse “a world free of nuclear weapons.”45 Perhaps these positions are all inconsistent, and Shultz is guilty of serial

52

j e f f r e y w. k n o p f

hypocrisy. More likely, however, Shultz adopted positions based on the specifics of each situation. After the cold war ended, Shultz could contemplate the case for nuclear disarmament, even though he had perceived a continuing need for nuclear deterrence while the Soviet Union still existed. And comparing Saddam Hussein to the Soviet leaders he had dealt with, Shultz apparently decided the Iraqi leader was less susceptible to being deterred; because Iraq was much weaker than the United States, it was also possible to think that preemption would be feasible, while this was clearly not a viable option against the Soviet Union. These different policy positions were not necessarily all correct. But the Shultz case does provide an example of why we should not expect the status of deterrence to remain constant. Strategic circumstances can change, making deterrence less (or more) relevant than before. The actors and actions one seeks to deter can also be different, with deterrence success more likely in some situations than in others. Finally, the feasible alternatives can vary. Sometimes, there will be a better alternative to deterrence. The fact that people can support deterrence in some situations but not in others helps explain the politics of deterrence. When momentum starts to build for strongly hard-line policies, such as preventive war, propeace forces tend to align with moderates in support of deterrence as the preferable alternative. When doves gain momentum, as in the peak years of antinuclear weapons protest, hawks swallow some of their reservations about deterrence and join centrists in defending it.46 As a result, in foreign policy debates, the case for deterrence often echoes Churchill’s famous comment about democracy: in some circumstances, deterrence is the worst strategy there is, except for all the others. Deterrence is not always the best option, even when it does emerge politically as a compromise around most people’s second choices. Nor does majority support for a different option, such as preemption, necessarily mean deterrence was actually a worse choice. The fact that political debates tend to be framed around competing alternatives, however, does suggest that democratic political systems are capable of engaging in a comparison of alternative policy options. Such political debate would be aided by a more explicit commitment to the normative model of rational policy analysis. If political officials, military leaders, and nongovernmental experts on strategy seek to identify the full range of potential options and to evaluate carefully their pros and cons, they are more likely to come to sound policy conclusions and to be able to marshal stable public support for those policies. Deterrence will not and should not always be the choice that emerges from this pro-

three items in one

53

cess; sometimes, there are better options. Conversely, however, deterrence is not going away; sometimes, it will be an appropriate strategy. Either way, the status of deterrence will not be a constant. Even when the political and intellectual climates are receptive to deterrence, the strategic environment can vary in ways that induce shifts in the use of deterrence.

conclusion The September 11 attacks seemed to sound a death knell for deterrence. In response, many researchers sought to identify ways in which deterrence could still prove viable. The history of deterrence, however, shows that recent doubts about the efficacy of deterrence are hardly unique. In both the political arena and academic research, deterrence has always had critics but has still often emerged as a policy compromise around most people’s second choices. Systemic factors point in the same direction. In a changing strategic environment, it was unlikely that deterrence would remain as central as it was during the cold war. At the same time, the continued existence of security threats made it unlikely deterrence would go away completely. The complex nature of the contemporary security environment makes it desirable to broaden our understanding of deterrence to see if there are new ways to achieve deterrence that might prove more effective than traditional deterrent strategies. Social science research can help inform policy debates concerning deterrence. Scholars, however, should try to avoid framing research or presenting conclusions in a way that could have negative consequences if that work enters the public arena. Blanket statements that “deterrence is the best strategy” or “deterrence does not work” should be avoided. In most situations in which there is serious debate, deterrence has some probability of success and some probability of failure. Whether or not deterrence is the best choice will depend on what other options are available and on what alternative deterrent options might be feasible. In sum, this chapter reaches three primary conclusions. First, doubts about deterrence after the September 11 terrorist attacks were not unique. Just as past critiques did not lead to deterrence being abandoned before, they are unlikely to do so now. Second, new, complex challenges call for creative thinking about possible new approaches to deterrence. The broad concept of deterrence proposed in this chapter, which removes any presumption that deterrence must be military in either means or ends, is meant to help in the identification of new approaches to deterrence. Third, deterrence will not

54

j e f f r e y w. k n o p f

always be the best option; comparative analysis of policy options remains as important as ever. This means that some of the most important research on deterrence will not be on deterrence at all. It will focus on potential alternatives to deterrence so as to improve the basis for comparing deterrence against other options. Given recent U.S. policy debates, systematic research on preventive war and preemption is particularly important.47 At the same time, however, the debate on preventive war versus deterrence is remarkably narrow. There might be a range of other alternatives that could contribute positively to peace and security. The classic critique of deterrence theory by George and Smoke concluded that deterrence has to be evaluated in the context of a broader theory of influence that incorporates alternative strategies.48 I can do no better than to make this my conclusion as well.

notes 1. I thank T. V. Paul, Pat Morgan, and Mark Brawley for helpful comments and suggestions on an earlier draft of this essay. 2. Ivo H. Daalder and James M. Lindsay, America Unbound: The Bush Revolution in Foreign Policy (Washington, DC: Brookings Institution Press, 2003), 125; Colin S. Gray, “Maintaining Effective Deterrence” (Strategic Studies Institute Monograph, U.S. Army War College, Carlisle, PA, August 2003), 7. 3. Paul K. Davis and Brian Michael Jenkins, Deterrence and Influence in Counterterrorism: A Component in the War on al Qaeda (Santa Monica: RAND, 2002); Ian R. Kenyon and John Simpson, eds., “Deterrence and the Changing Security Environment,” Contemporary Security Policy 25 (April 2004); Robert F. Trager and Dessislava P. Zagorcheva, “Deterring Terrorism: It Can Be Done,” International Security 30 (Winter 2005 / 06): 87–123; Derek D. Smith, Deterring America: Rogue States and the Proliferation of Weapons of Mass Destruction (Cambridge: Cambridge University Press, 2006); David P. Auerswald, “Deterring Nonstate WMD Attacks,” Political Science Quarterly 121 (Winter 2006 / 07): 543–68; James H. Lebovic, Deterring International Terrorism and Rogue States: U.S. National Security Policy after 9 / 11 (London: Routledge, 2007). 4. Robert Jervis, Perception and Misperception in International Politics (Princeton: Princeton University Press, 1976), chapter 3; Janice Gross Stein, “Deterrence and Reassurance,” in Behavior, Society and Nuclear War, vol. 2, edited by Philip E. Tetlock et al. (New York: Oxford University Press, 1991), 8–72. 5. Alexander L. George, “The Need for Influence Theory and Actor-Specific Behavioral Models of Adversaries,” Comparative Strategy 22 (December 2003), 480. 6. Lawrence Freedman, Deterrence (Cambridge, UK: Polity, 2004), 6; Jervis, this volume. 7. Patrick M. Morgan, “Deterrence in Foreign Policy,” Armed Forces & Society 3 (Spring 1977): 495–502; the quotation is from Morgan, Deterrence: A Conceptual Analysis, 2d ed. (Beverly Hills: Sage, 1983), 29, emphasis in original. 8. David A. Baldwin, “The Concept of Security,” Review of International Studies 23 ( January 1997): 5–26. 9. Thomas C. Schelling, Arms and Influence (New Haven: Yale University Press, 1966), 2; Freedman, Deterrence, 26, 86.

three items in one

55

10. Glenn H. Snyder, “Deterrence by Denial and Punishment” (Research Monograph No. 1, Center of International Studies, Princeton University, Princeton, 1959); Snyder, Deterrence and Defense: Toward a Theory of National Security (Princeton: Princeton University Press, 1961). 11. General deterrence exists when a potential challenger becomes convinced it is not worth initiating any active challenge that would produce the type of crisis in which immediate deterrence becomes necessary; see Morgan, Deterrence: A Conceptual Analysis, and Morgan, Deterrence Now (Cambridge: Cambridge University Press, 2003), chapter 3. Internalized deterrence arises when an actor internalizes certain norms such that it no longer believes it would be legitimate to take certain actions; see Freedman, Deterrence. 12. United States Strategic Command, “Deterrence Operations Joint Operating Concept,” version 2.0, December 2006. 13. The “Deterrence Operations Joint Operating Concept” (DO-JOC) adds a third way of seeking deterrence, which it labels “deterrence by inducing adversary restraint.” In this approach, one tries to raise the benefits and / or lower the costs for the other side of choosing not to act. This third way reflects a well-known point derived from expected utility analysis: even a credible threat of retaliation can fail to deter if the other side perceives the costs of not acting as even greater than those of acting. To describe efforts to make the status quo more palatable to the other side as a form of deterrence requires broadening the concept even further than I have called for here. By my definition, encouraging restraint is not deterrence. What the DO-JOC describes instead encompasses some other terms in the literature. Partly, it reflects Schelling’s famous insight about the need to incorporate assurances in a deterrence strategy— but assurance is not a separate way to deter, it is a necessary component of the first two ways. If the strategy goes beyond assurance to try to assuage deeper fears about the future on the other side (i.e., to lower its expected costs of continuing the status quo), it becomes a strategy of reassurance. If it offers new benefits for restraint, it becomes a strategy of positive incentives. In practice, it is often a good idea to combine these with deterrence, but they are not best understood as a form of deterrence. In short, while I advocate a broader concept of deterrence than is typically found in the IR literature, I do also see limits on how far the term should be broadened. 14. United States Strategic Command, “Deterrence Operations Joint Operating Concept,” 16, 28. 15. Consequence management (CM) is described as a deterrent measure as early as the famous 2002 National Security Strategy, which is often viewed as having dismissed deterrence in favor of preemption (President of the United States, “The National Security Strategy of the United States of America,” September 2002, 14). The second iteration of the National Strategy for Combating Terrorism released by the Bush administration explicitly lists deterring terrorist WMD use as an objective, and also mentions CM as one measure that can contribute to that goal (United States Government, “National Strategy for Combating Terrorism,” September 2006, 14). 16. I thank Pat Morgan for drawing my attention to the importance of this point. 17. For an analysis of how “lessons of 9 / 11” contributed to support for the Iraq war, see Jeffrey W. Knopf, “Misapplied Lessons? 9 / 11 and the Iraq Debate,” Nonproliferation Review 9 (Fall /Winter 2002): 47–66. 18. Daalder and Lindsay, America Unbound, 125. 19. Jeffrey W. Knopf, “Deterrence or Preemption?” Current History (November 2006): 395–99. 20. Carl von Clausewitz, On War, translated by Michael Howard and Peter Paret (Princeton: Princeton University Press, 1976). 21. George H. Quester, Deterrence before Hiroshima: The Airpower Background of Modern Strategy (New York: John Wiley & Sons, 1966). The statement “the bomber will always get through” was made by British Prime Minister Stanley Baldwin before the House of Commons in 1932 (quoted in Quester, Deterrence before Hiroshima, 67). Interestingly, Baldwin lacked confidence in

56

j e f f r e y w. k n o p f

the prospects for mutual deterrence and instead advocated an essentially preemptive strategy of massive offensive operations against the other side at the outset of hostilities. 22. Bernard Brodie, “Implications for Military Policy,” in The Absolute Weapon: Atomic Power and World Order, edited by Bernard Brodie (New York: Harcourt, Brace, 1946), 76. 23. I thank T. V. Paul for this observation. 24. Marc Trachtenberg, “A ‘Wasting Asset’: American Strategy and the Shifting Nuclear Balance, 1949–1954,” in Nuclear Diplomacy and Crisis Management, edited by Sean M. Lynn-Jones, Steven E. Miller, and Stephen Van Evera (Cambridge, MA: MIT Press, 1990), 69–113. 25. Robert Jervis, “Mutual Assured Destruction,” Foreign Policy no. 133 (November/ December 2002): 40. 26. See Associated Press, “Reagan in Europe: Speech in Germany, Arrival in Spain,” New York Times, May 7, 1985, A8. 27. See, for example, Jesse Helms, “Opening Statement of Senator Jesse Helms (R-NC),” Ballistic Missiles: Threat and Response, Hearings before the Committee on Foreign Relations, United States Senate, April 15, 1999 (Washington, DC: Government Printing Office, 2000), 2; Baker Spring and James H. Anderson, “Making the Case for Missile Defense” (Backgrounder no. 1225, Heritage Foundation, Washington, DC, October 5, 1998),http://www.heritage.org / Research / MissileDefense / BG1225.cfm. 28. See, for example, Canberra Commission on the Elimination of Nuclear Weapons, “Executive Summary,” August 1996, http://www.nuclearfiles.org / menu / key-issues / nuclear-weapons / issues /civil-society/canberra-commission-executive-summary_1996–08–00.htm. Convened by the Australian government, the Canberra Commission was an international group of experts that included former high-level government and military leaders. 29. Robert Jervis, “Deterrence Theory Revisited,” World Politics 31 ( January 1979): 289–324. 30. Alexander L. George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974). 31. Richard Ned Lebow, Between Peace and War: The Nature of International Crisis (Baltimore: Johns Hopkins University Press, 1981); Robert Jervis, Richard Ned Lebow, and Janice Gross Stein, eds., Psychology and Deterrence (Baltimore: Johns Hopkins University Press, 1985); Stein, “Deterrence and Reassurance.” 32. For a good example, see Paul K. Huth and Bruce Russett, “What Makes Deterrence Work? Cases from 1900 to 1980,” World Politics 36 ( July 1984): 496–526. 33. Both authors have published extensively. For good recent summations of their thinking, see Keith B. Payne, The Fallacies of Cold War Deterrence and a New Direction (Lexington, KY: University Press of Kentucky, 2001); and Colin S. Gray, “The Reformation of Deterrence: Moving On,” Comparative Strategy 22 (December 2003): 429–61. 34. In addition to the writings of Payne and Gray, see also Adam Garfinkle, “Culture and Deterrence,” (E-Note, Foreign Policy Research Institute, Philadelphia, August 25, 2006), http://www.fpri.org / enotes / 20060825.americawar.garfinkle.culturedeterrence.html. 35. Richard Pipes, “Why the Soviet Union Thinks It Could Fight and Win a Nuclear War,” Commentary 64 ( July 1977). 36. Christopher H. Achen and Duncan Snidal, “Rational Deterrence Theory and Comparative Case Studies,” World Politics 41 ( January 1989). James D. Fearon argues that selection effects also affect the main large-N, quantitative tests of deterrence. James D. Fearon, “Selection Effects and Deterrence,” International Interactions 28 ( January–March 2002). 37. Keith B. Payne and Dale C. Walton, “Deterrence in the Post–Cold War World,” in Strategy in the Contemporary World: An Introduction to Strategic Studies, edited by John Baylis et al. (New York: Oxford University Press, 2002), 170ff. 38. Achen and Snidal, “Rational Deterrence Theory.” For an argument that most of Lebow’s cases specifically involve failures to implement deterrence correctly rather than failures of RDT,

three items in one

57

see John Orme, “Deterrence Failures: A Second Look,” International Security 11 (Spring 1987): 96–124. 39. Frank P. Harvey, “Rigor Mortis, or Rigor, More Tests: Necessity, Sufficiency, and Deterrence Logic,” International Studies Quarterly 42 (December 1998): 675–707. In part for reasons argued above, I believe the four conditions are only probabilistic factors, not necessary ones. There can be cases in which an actor is deterred even though no explicit deterrent threat was issued. Communicating a deterrent message when a genuine threat exists is likely to improve the odds of success, but it might not in all cases be necessary. 40. In addition to work on deterrence using rational-choice and case study methods, there are many studies that use statistics. The statistical research generally finds modest support for the propositions of RDT. Because the data sets contain cases of failure, however, the large-N literature also indirectly confirms that deterrence can either succeed or fail. There are factors that increase the probability of success, but there are no infallible deterrent strategies. For a good review, see Paul K. Huth, “Deterrence and International Conflict: Empirical Findings and Theoretical Debates,” Annual Review of Political Science 2 (1999): 25–48. Morgan, Deterrence Now, chapter 4, reviews and compares the findings from the formal modeling, statistical, and case study research. 41. Even Colin Gray, who generally supports hard-line positions, acknowledges this point and uses it to argue for not unduly dismissing deterrence. See his “Maintaining Effective Deterrence,” vi, 9. 42. Charles Krauthammer, “The Obsolescence of Deterrence,” Weekly Standard, December 9, 2002, http://www.theweeklystandard.com /Content / Public /Articles / 000 / 000 / 001 / 964dzkuf .asp. 43. Jon Niccum, “Fallout from ‘The Day After,’” Lawrence.com, November 19, 2003, http://www.lawrence.com / news / 2003 / nov/ 19 / fallout_from / . 44. George P. Shultz, “Act Now,” Washington Post, September 6, 2002, A25. 45. George P. Shultz et al., “A World Free of Nuclear Weapons,” Wall Street Journal, January 4, 2007, A15. 46. A partial exception is the early 1980s, when the Reagan administration responded to the nuclear freeze movement by holding out the dream that nuclear weapons could be rendered obsolete through its star wars proposal instead. 47. Some good recent analyses of these options include Dan Reiter, “Preventive War and Its Alternatives: The Lessons of History” (Strategic Studies Institute Monograph, U.S. Army War College, Carlisle, PA, April 2006); and Karl P. Mueller et al., Striking First: Preemptive and Preventive Attack in U.S. National Security Policy (Santa Monica, CA: RAND, 2006). 48. George and Smoke, Deterrence in American Foreign Policy.

chapter three R

Rational Deterrence against “Irrational” Adversaries? No Common Knowledge janice gross stein introduction: those who seek to deter seek the “rational” th e contemporary debate about deterrence as theory and as strategy unfolds along three paths. One of the principal threads of the debate—the rationality of adversaries—is long-standing but has been given new life in the contemporary international system. If adversaries are not rational, then deterrence is impossible.1 Indeed, if adversaries are not rational, then no strategy is possible and leaders are left with no choice but to resort to force when they are threatened. President George W. Bush, for example, seeking to deter Iran from developing nuclear weapons, looked to the rational people inside the Iranian government as partners.2 Two very large questions are embedded in this discussion by political leaders about deterrence and negotiation. The first is the meaning of “rationality,” which is far from obvious. What kind of rationality is required to enable a strategy of deterrence? What level of rationality? President Bush intuitively understood rationality as “stopping to think about the consequences of what is going on,” but the theory of deterrence requires more. If one or more of the parties to the conflict challenges deterrence in the face of clear military superiority and unquestioned resolve, are they necessarily irrational? Are there good reasons why strategies outside models of rational deterrence are nevertheless chosen by political leaders? These questions point to even bigger questions: How rational are political leaders when they make big and important decisions about the use of force? Are some kinds of leaders more likely to be rational than others? Are leaders more inclined to be rational in certain kinds of international systems than others? Does the structure of the global system matter? Finally, how well do concepts of 58

rational deterrence

59

rationality, as developed by philosophers or economists, travel into the world of politics? This chapter addresses these questions by exploring three issues that set the context for a reconsideration of deterrence as narrative in world politics. First, it explores the commonsensical understanding of rationality, an understanding that informs deterrence as strategy—rather than as a theory. Second, it identifies new research that challenges this commonsensical understanding of rationality and speaks directly to the limits of rationality in human decision making. It is this cumulating research that led Thomas Schelling recently to call game-theoretic models of deterrence “heuristics,” which at best set normative standards that can never be met.3 A second purpose is to describe the construction of “threat,” to identify the meaning threat is given in practice. It is often far from obvious what threat means, and it is the construction of threat that frames deterrence as theory and as strategy. To understand the way threats are constructed and interpreted, the chapter treats culture as an active matrix, which shapes narratives that can challenge as well as stabilize deterrence. Both these questions—the meaning of rationality and the construction of threat—are at the core of the theory of deterrence. They become especially challenging in a complex and changing international system. The structure of the international system is a third variable that affects deterrence in theory and practice. In a well-structured environment, in which the players are known and the rules are embedded, meaning gradually becomes established and responses become routinized. Response and meaning are learned over time. Under these conditions, the scope of deterrence theory becomes easier to define; what is “in” and what is “out” are easier to establish. The strategy of deterrence also becomes embedded in practice: its language becomes familiar and its meaning is established over time through repeated interaction among the parties. In a global system that is both structurally complex—with a large number of different kinds of actors—and uncertain—with changing power relationships that are difficult to interpret and unfamiliar challenges to the status quo—deterrence becomes far more problematic. In complex systems, in which units adapt and spontaneously organize and reorganize, change is generally nonlinear and harder to predict.4 It therefore becomes more difficult to define the scope conditions of the theory, and those who attempt deterrence increasingly find themselves in unfamiliar territory, with fewer codified practices and routines. What game theorists have called common knowledge, or communities of practice, is absent; shared understandings of

60

janice gross stein

the world and common interpretations of the meaning of action are much less readily available. The contemporary global system is both structurally complex and uncertain, and the absence of common knowledge compounds the problem of rationality. The chapter begins by first describing the theory of deterrence. It then examines the way deterrence theorists define and use the concept of rationality and how recent research in neuroscience and cognitive psychology calls into question even a circumscribed notion of rationality. It then explores how the strategic and diplomatic context of deterrence shapes its practice. Drawing on the research findings of cognitive and neuroscience, the chapter concludes by suggesting how the theory and practice of deterrence might be improved.

the theory of deterrence The theory of deterrence, unlike the practice, is elegant in its simplicity. Deterrence attempts to prevent someone from doing something he or she would otherwise like to do. Coercive diplomacy is a closely related strategy that seeks to get someone to do something he or she would otherwise not do. Theories of deterrence explain when the strategy is likely to succeed. The rational calculations of a would-be challenger to change the status quo could be altered by a credible threat, which raises the costs of that kind of action. To be credible, a threat must be reinforced by the capabilities to implement the threat of retaliation and a credible and firm resolve to do so.5 This abstract formulation of the theory of deterrence is deceptively simple. Its scope conditions are the presence of a rational adversary that can calculate the likely consequences of a challenge and then recalculate these consequences again in response to a credible threat of retaliation. The explanatory power of the theory is deeply grounded in microeconomic theory, which assumes that people consistently maximize their expected utility. The internal logic is tight. An adversary will proceed with a challenge if and only if the estimated benefits exceed the likely costs. The theory of deterrence holds even when the strategy of deterrence fails, if the challenger calculated that the likely benefits of a challenge would exceed the expected costs. Even when the calculation is in error—through a misreading of the intentions or calculations of a defender—the theory does not fail, although the strategy does. The theory fails under tightly specified and extraordinarily limited conditions: if and only if a challenger calculates that the likely costs will

rational deterrence

61

exceed the expected benefits and proceeds nevertheless to act in ways that the defender does not want. Even under these strict conditions, however, there is an escape: a challenger that does not choose to maximize expected utility cannot be considered rational. A challenger that is not rational is outside the scope of the theory. The theory of deterrence, in other words, is almost self-confirming because of the microeconomic model of rationality that structures its arguments. Almost all failures of deterrence are either the result of poor implementation of the strategy or they are beyond its explanatory scope. This formulation of deterrence theory leaves out almost all the interesting problems that theorists and practitioners face.

commonsensical understandings of rationality The impoverished concept of rationality at the core of deterrence theory makes it easy to confirm the theory. The commonsensical understanding of rationality is far more relaxed than the formal microeconomic model that informs the theory of deterrence. Rational decision making refers first to the processes people should use to choose among competing alternatives. In a rational decision-making process, people should be logical and orderly. Their preferences should be ranked, at least intuitively, in such a way that if they prefer A to B and B to C, then they also should prefer A to C. If policy makers prefer peace to all-out war and if they prefer all-out war to lowlevel insurgency, for instance, then they should prefer peace to insurgency. If policy makers violate this requirement of transitive preferences, that is, if they prefer C to A or insurgency to peace, then they are ruled out as rational decision makers. Decision makers whose preferences are not transitive are outside the scope of deterrence theory. Rational decision makers should also be good at attending to new information that comes along as they are making their choices. They need to update their estimates in response to new information that is reliable, valid, and diagnostic. The attentive reader may notice all sorts of caveats that complicate this conception of rationality. Reliable and valid information that comes from a trustworthy source and that can be independently confirmed is not always readily available. Policy makers also have to consider significant or diagnostic evidence that speaks to the likelihood of the consequences of different alternatives. When President George W. Bush was considering whether or not to go to war against Iraq, he was told that Saddam Hussein had sought to buy yellowcake uranium from Niger. That was new

62

janice gross stein

information to the president—he had not heard it before—and it was diagnostic; it signalled that Saddam was likely seeking to develop unconventional weapons. What the information was not, however, was reliable or valid, and therefore it should have been excluded from any kind of consideration. The reliability and validity of information is a threshold barrier that any piece of evidence should cross on its way into the decision-making process. Determining the trustworthiness of any piece of information, however, is often very difficult to do. Rational processes of information management are often swamped by the quick intuitive processes and deep cognitive biases that political leaders use to interpret evidence. So far this picture of a rational decision maker fits with a common-sense understanding of rationality. People who make important choices about the use of threats or force need to be logical, they need to be discriminating and open to evidence, and they need to be coherent and consistent in responding to logical arguments.6 Rational decision makers are open to arguments and evidence, free of serious biases as they weigh the evidence and think about the likely consequences of options. The minimal, commonsensical requirements of rationality expect that policy makers can learn from history, that they can draw some propositions from the past and apply these propositions in an appropriate way to the future as they think about the likely consequences of their policy options.7 The evidence is now overwhelming that even these relaxed requirements, which are far less demanding than the formal requirements of deterrence theory, are rarely met in practice. The most important evidence of the limits to rationality comes from well-established work in psychology. New, and still tentative, research results in neuroscience are also challenging the fundamental tenets of the rational model. Neuroscientists are reintroducing conflict into choice. What makes the work of both neuroscientists and cognitive psychologists even more important is that their findings tend to converge.

the cognitive revolution Forty years ago, psychologists started a cognitive revolution when they rejected simple behaviorist models and looked again at how the way people think shapes the choices they make. They brought the mind back into psychology. Although this was not its purpose, the cognitive revolution can be largely understood as a commentary on the limits to rationality. Research has now cumulated to show that people rarely conform to the expecta-

rational deterrence

63

tions of the rational model because of the difficulty of making inferences as models of rational choice expect and because of the limitations of the human mind.8 Cognitive psychology has demonstrated important differences between the expectations of rational decision models and the processes of attribution, estimation, and judgment that people frequently use. It explains these differences by the need for simple rules of information processing and judgment that are necessary to make sense of environments that are both uncertain and complex. People have a preference for simplicity, they are averse to ambiguity and dissonance, and they fundamentally misunderstand the essence of probability.9 We are not intuitively good at estimating probabilities. And we have risk profiles that depart from what models of rational choice would expect. For example, we are far more averse to loss than we are gain seeking. Together, these attributes compromise the capacity for rational choice. Political leaders considering a challenge need to make a very complex world somewhat simpler. To do so, they unconsciously strip nuance, context, and subtleties from the problems they face to build simple frames. To simplify a noisy, distracting, and complex environment, political leaders oversimplify, at times badly, to persuade themselves that a policy will solve a problem or that it will succeed or that it is justified. Especially when they think a challenge is justified, they become less sensitive to the likely negative consequences of their action than they ordinarily would. Cognitive psychologists have produced robust evidence that people strongly prefer consistency, that they are made uncomfortable by dissonant information, and that they consequently deny or discount inconsistent information to preserve their beliefs. This drive for consistency impairs processes of estimation and judgment. The well-established tendency to discount inconsistent information contributes significantly to the persistence of beliefs in the face of contradictory information. Indeed, exposure to contradictory information frequently results in the strengthening of beliefs.10 People, it seems, are hardwired to be conservative. The lengths policy makers will go to defend forecasts gone wrong are quite remarkable.11 Much of the work of cognitive psychology has been done in the laboratory, with students, and experts have questioned how well the results travel into international politics. That question has largely been put to rest by a remarkable study of the political forecasts made by foreign policy experts in different cultures.12 Experts on foreign policy generally continued to defend their forecasts, even though what they expected did not happen. Philip Tetlock identifies seven categories of belief system

64

janice gross stein

defenses: challenging whether the local conditions required by the forecast were satisfied; invoking the unexpected occurrence of a shock; invoking a close-call counterfactual (“I almost got it right”); using an “off-on-timing” argument (“I’m ahead of my time; history will prove me right”); declaring that international politics is hopelessly indeterminate and consequently unpredictable; defiantly asserting that they made the “right mistake” and would do it again; or insisting that unlikely things sometimes happen.13 The same kind of self-serving bias in argumentation that cognitive psychologists have documented in the laboratory has been confirmed among political experts. Tetlock finds a relationship between the size of the mistakes and the activation of defenses. The more confident experts were in their original forecast, the more threatened they were when they were faced with disconfirming evidence and the more motivated they were to use one or more of the seven defenses to preserve their beliefs. “Defensive cognitions,” Tetlock argues, “are activated when forecasters most need them.”14 When political experts most needed to revise their judgments, they were least open to revision. If these patterns of thinking are characteristic among experts in international politics, they are as likely, if not more so, to be present among political leaders thinking about overturning the status quo. Deeply rooted cognitive processes systematically work against rational expectations of appropriate, diagnostic updating. When beliefs and arguments do change, they generally change in lumpy, bumpy ways that reflect somewhat arbitrary patterns in information and basic processes of attribution. Neither of these has very much to do with rational expectations. There is some evidence, however, that is more encouraging. This evidence comes from the close analysis of differences among foreign policy experts in terms of their willingness to entertain the possibility that they were wrong. Not all experts were resistant to change all the time. Drawing on a well-known distinction made by Isaiah Berlin, Tetlock classified foreign policy experts as “foxes” or “hedgehogs.” Hedgehogs know “one big thing” extremely well and extend what they know into other domains of foreign policy analysis. Foxes, on the other hand, know many small things, are generally skeptical of grand overarching schemes, stitch together explanations with different threads of knowledge, and are skeptical of prediction in world politics.15 The evidence shows that the foxes do much better at short-term forecasting within their broad domain of expertise than do hedgehogs. The worst performers were hedgehogs who made long-term predictions, usually with

rational deterrence

65

considerable confidence. Hedgehogs are generally people who have strong needs for structure and closure and who are most likely to discount and dismiss inconsistent evidence when it contradicts their preconceptions. The more knowledge hedgehogs have, the better equipped they are to defend against inconsistency. Foxes are skeptical of deductive approaches, more likely to qualify analogies by looking for disconfirming information, more open to competing arguments, more prone to synthesize arguments, more detached, and, not surprisingly, more likely to admit they were in error and move on. The hallmark of the foxes was their more balanced style of thinking about the world. Foxes had “a style of thought that elevates no thought above criticism.”16 This evidence suggests that deterrence is context specific and that the cognitive styles of would-be challengers matter. The axiomatic theory of deterrence does not capture these differences in cognitive styles, the variation across political leaders who may be more like hedgehogs than foxes and consequently respond quite differently to a deterrent threat. Cognitive styles of leaders matter; they are an important part of the context of deterrence. Yet this context is precisely what is absent from the theory of deterrence. People are not intuitive estimators of probability and risk. They depart systematically from what objective probability calculations would dictate in the estimates they make. “Human performance suffers,” argues Tetlock, “because we are, deep down, deterministic thinkers with an aversion to probabilistic strategies that accept the inevitability of error.”17 Foreign policy experts are no exception. Where we can compare their estimates to those that would be generated by objective calculations of probability, experts do surprisingly poorly. Highly educated specialists in foreign affairs approached only 20 percent of the ideal across all exercises.18 Their performance is so poor because they think causally rather than pay attention to the frequencies with which events occur. Experts tend to overestimate the likelihood of war, for example, because they can easily imagine the causal pathways to war, a highly salient occurrence that they have likely studied. They pay less attention to the frequency of wars over an extended period of time.19 To make matters worse, likely states of the world are very difficult to estimate. In world politics, there are no repeated trials with large numbers. Neither party to a deterrence relationship generally lives in a world of risk, where the probability distributions are known and it is relatively easy to estimate the likelihood of events. Analysts, even the best informed, do not know the probability of another attack by militants against civilian infrastructure in the United States or the United Kingdom. There have been too few such

66

janice gross stein

attacks to generate a reliable distribution of probabilities. Those seeking to deter another attack are forced to work in structurally uncertain environments, where they have no access to probability distributions. This world of structural uncertainty is one that is particularly uncomfortable psychologically, and it is under these conditions that leaders, just like experts, are likely to seek the certainty, the false certainty, of order and control. In this world of uncertainty, leaders search for the relevant reference classes to anchor their judgments. Cognitive psychology has identified a number of heuristics and biases that people use in environments of risk and uncertainty that can impair processes of judgment.20 Heuristics refer to the rules leaders use to test the propositions embedded in their schemas, using convenient shortcuts or rules of thumb.21 What leaders know and what they can access has a disproportionate impact on their forecasts. For years in Britain, policy makers were likely to forecast negative consequences from appeasement due to the salience of the meeting at Munich between Neville Chamberlain and Adolf Hitler in British history. They paid correspondingly less attention to the unwanted escalation that could come from threatening the use of force and were overconfident in the success of deterrence. Hedgehogs who are experts on what happens “when we knuckle under to dictators” are especially likely to discount any negative consequences from the threat of force. Much the same is true among Israel’s leaders, where defending their reputation for deterrence becomes a value independent of other consequences.22 Cognitive biases also lead to serious errors in attribution, which can confound policy making. People exaggerate the likelihood that others’ actions are the result of their own prior behavior and overestimate the extent to which they are the target of those actions; cognitive psychologists call this pattern of attribution the “egocentric bias.” One of the most pervasive biases is the fundamental attribution error, where people exaggerate the importance of dispositional over situational factors in explaining the disliked behavior of others but explain their own behavior by the situational constraints that they face.23 When explaining Iran’s continuing development of a nuclear program, leaders attribute it to the irrationality of Iran’s leaders or to their aggressive ambitions; these are dispositional attributions. Leaders in Washington rarely make reference to Iran’s encirclement by Western forces to the east and to the west or to the alarming precedent of regime change in Iraq; both of these are situational factors. Foreign policy decision makers, like people generally, are usually averse to loss. Prospect theorists posit an S-shaped value function with varying

rational deterrence

67

risks as a function of losses and gains. Loss is more painful than comparable gain is pleasant, and people prefer an immediate smaller gain rather than taking a chance on a larger longer term reward.24 People systematically overvalue losses relative to comparable gains. The impact of loss aversion on deterrence is considerable. Leaders tend to be risk averse when things are going well and relatively risk acceptant when things are going badly. When they face a crisis in which they are likely to lose something that matters to them, they are predisposed to take greater risks. Leaders are also likely to take greater risk to protect what they already have—the endowment effect—than to increase their gains. They also are likely to take greater risks to reverse losses or to recapture what they once held than they would to make new gains. And when decision makers suffer a significant loss, they are far slower to accommodate and accept these losses than they would be to incorporate gains. President Anwar Sadat of Egypt, for example, never “normalized” for the loss of the Sinai to Israel in 1967. Even though Israel had an obvious advantage in military capabilities, Sadat was undeterred and highly motivated to recapture the Sinai. He designed around Israel’s military strengths and launched a war in 1973.25 Under these kinds of conditions, theories of rational deterrence, which do not systematically build in aversion to loss, are likely to mislead foreign policy decision makers by generating undue confidence in the stability of deterrence and blinding leaders to likely challenges. Finally, leaders reverse their preferences and make different choices when problems are reframed as losses rather than gains. This framing effect is so strong that people reverse their preferences when the same problem is put as a question of loss as opposed to gain. Preference reversal of this kind violates even the commonsensical understanding of rationality. It appears that preferences are not consistently transitive and can be reversed by how problems are framed. In deterrence, the story that is told matters. The need for simplicity and consistency, impediments to probabilistic thinking, the predisposition to loss aversion, and framing effects are often treated as deviations from rational models of deterrence. Rational choice remains the default and these deviations are treated as limiting conditions. Yet these deviations are so pervasive and so systematic that it is a mistake to consider rational models of deterrence as empirically valid. These patterns of choice need to be built into theories of deterrence as core assumptions and systematically tested in different contexts. The robustness of psychological models is now supported by a generation of research in neuroscience.

68

janice gross stein

neuroscience and deterrence We are at the cutting edge of an exciting revolution in our understanding of the human brain and how it works. New imaging technology is allowing scientists for the first time to “watch” the brain as it thinks, feels, remembers, and chooses, and the pictures scientists are seeing are revolutionizing our understanding of thought and decision. Even though the experimental results are still far from definitive, two results stand out. First, many decisions seem not to be the result of a deliberative thought process but preconscious neurological processes. The brain can absorb about 11 million pieces of information a second but can only process forty consciously. The unconscious brain manages the rest. Second, many decisions seem to be the product of strong emotional responses. In a by now famous set of experiments performed in the 1980s, Benjamin Libet, a physiologist, found that brain signals associated with movement occurred half a second before the person was conscious of deciding to move. The sequence was dramatically different from what models of rational choice expect; first came perception of motion and then conscious awareness of the decision to move. These results have been replicated thousands of times over the last two decades, well before the latest brain-imaging technology became available.26 As a recent reviewer of this research explained, “A bevy of experiments . . . suggest that the conscious mind is like a monkey riding a tiger of subconscious decisions and actions in progress, frantically making up stories about being in control. . . . The conscious brain is only playing catch up to what the unconscious brain is already doing.”27 If this image of neural activity is sustained as research progresses, it has profound consequences not only for our understanding of deterrence but also for all choice. The picture of reflective, deliberative, rational decision making that underlies so much of the analysis of deterrence theory and strategy fits poorly with this cumulating body of evidence of how we as human beings choose. New research suggests that emotion precedes choice, emotions trigger choices, and emotion follows in the wake of choice to shape what people learn from their choices. Emotions are the “inner states that individuals describe to others as feelings, and these feelings may be associated with biological, cognitive, and behavioral states and changes.”28 In an often-cited study, Antonio Damasio worked with a patient, “Elliot,” who had had a tumor removed from the part of his brain that controlled emotion. The flatness of his emotional responses impaired his practical reasoning so badly that Elliot made a succession of mistakes.29 Damasio rejected the widely

rational deterrence

69

accepted view that separate neural mechanisms underlie emotion and reason and argued, on the contrary, that emotion was necessary to reason and important to choice. How do neuroscientists think about emotion and cognition? Some think of two operating systems in the brain, emotion and reason.30 “Emotions influence our decisions,” argues Jonathan Cohen. “They do so in just about every walk of our lives, whether we are aware or unaware of it, and whether we acknowledge it or not.”31 Emotions are automatic processes associated with strong positive or negative response. The brain has different kinds of mechanisms; one, which includes emotional responses, can respond automatically, quickly, and definitively but is relatively inflexible, while the other is less rapid and has limited capacity but is more flexible. There is a trade-off between the speed and specialization of emotions and the generality of reflection. In the circumstances of modern life, these systems may prescribe different responses, and the outcome of this competition determines choice.32 This branch of what is called “neuroeconomics” has explicitly brought conflict back into decision making. Choice is a conflict between emotion and computation. Daniel Kahneman calls the first system of decision making “intuitive” and “associative” and the second “reasoned” and “rule governed.”33 The first system is preconscious, automatic, fast, effortless, associative, unreflective, and slow to change. The second system is conscious, slow, effortful, reflective, rule governed, and flexible. The vast majority of decisions are made through the first system, which draws heavily on emotions, and in a competition between the two, it always trumps the rule-governed, reasoned system. It is extraordinarily difficult, Kahneman concludes, for the second system to educate the first. The well-known ultimatum game highlights the computational, cognitive, and emotional elements at play in decision making. The game comes out of economics but has direct relevance to deterrence. One partner has access to a resource—ten dollars, vast oil wealth, highly sophisticated military technology—and can propose how the resource should be split. If the other party accepts the proposal, then the resource is divided as they have agreed. If the split is rejected, the party that controls the resource keeps it all and the game is over or war breaks out. Rationally, the second party should accept anything that is offered because anything is clearly better than nothing. And again, the first party, knowing that for the other party anything is better than nothing, should rationally offer as little as possible. Contrary to what rational models would expect, offers

70

janice gross stein

of less than 20 percent of the total are generally rejected out of hand. Why? Perhaps those who rejected the offer were worried about their bargaining reputation for the next round, as rational deterrence theory says they should. But people responded the same way even when reputational effects were removed from consideration, when they were told that they would play the game only once. When asked why they rejected an offer that would give them something, people responded that the offer was humiliating, insulting, patently unfair. They responded quickly and intuitively to reject an offer that gave them something but humiliated them in the process.34 Their rejection was driven by a strong negative emotional response. Scholars in international relations are now beginning to look carefully at the neuroscience of emotion and how it affects the stability of deterrence and the solution of collective action problems.35 People seem to have a hypersensitive capacity to detect and react to betrayal. The stability of deterrence is likely much more affected by a “hot” emotional reaction to betrayal, disappointment, shame, or humiliation than it is by a “colder” calculus of capability and resolve. Sustained research on the impact of emotion can help to unlock the paradoxes of irrationality that shape much of the logic of deterrence.

the systemic context: deterrence in practice Different styles of decision making—some more rational than others—are associated with different cognitive styles and with different emotional states. These styles and states are not wholly independent of the systemic environment in which leaders operate, of the values within cultures that are pushed to the fore in a relationship of deterrence, or of the motivation to challenge infused by a sense of grievance. Deterrence also is influenced by history, by the stories that challengers and defenders come to share over time. The rational theory of deterrence became prominent at a particular time in history. Although an age-old strategy, deterrence was theorized formally when the global system was tightly structured in a bipolar nuclear order. Both the context and the structure mattered. Theorists were heavily influenced by the experience prior to World War II, when Britain tried and failed to appease Hitler’s Germany. The failure of appeasement had a profound impact on postwar scholarship: scholars and practitioners alike used “appeasement” as a pejorative term and looked to strategies of deterrence as the principal strategy to avoid nuclear war. Theories of rational deterrence seemed particularly apt during the cold war, when the world was tightly

rational deterrence

71

structured into two blocs, each headed by a nuclear superpower, and a large group of uncommitted states in which the superpowers played out their rivalry. The simplified assumptions of theories of deterrence seemed intuitively plausible in a highly structured world in which the use of force by either superpower against the other would be catastrophic. The patently obvious consequences of nuclear war simplified the calculations of a challenge between the two superpowers, and direct deterrence between nuclear powers seemed relatively straightforward. There was no hard evidence, of course, that deterrence between the nuclear powers succeeded, but the absence of nuclear war was treated as self-evident confirmation of general nuclear deterrence. Even in that relationship, however, deterrence as strategy had its dangerous ambiguities that played out in the decades-long debate about extended deterrence. From the absence of major direct challenges escalating to nuclear war between the United States and the Soviet Union, theorists assumed deterrence success and then applied the theory—almost as ideology—to the deterrence of limited challenges against smaller allies. The taboos were less clear in extended deterrence, the thresholds of escalation were not as obviously demarcated, and the incentives to challenge were both more diverse and more complex. The risk of standing firm for too long was that the crisis would go over the edge into war, but the risks of backing down were an invitation to challenge again, elsewhere. In extended deterrence, unlike in general nuclear deterrence, a deterrent reputation became a value in and of itself. In an international system tightly structured at the top by two nuclear powers, the calculus of deterrence, although never as elegant and logical as proponents of deterrence theory suggested, was simple. It became more complicated as it was extended to allies who had their own motives and incentives to escalate, and both superpowers focused as much of their attention during the long cold war on controlling their allies as they did on deterring challenges. Superpowers found themselves entangled in situations of moral hazard.36 Even these kinds of challenges appear relatively simple, however, in comparison with the challenges in the loosely structured, complex, and multilayered diverse and evolving international system that followed the cold war. In this kind of global system, the challenges to the theory and the practice of deterrence have changed dramatically. Conventional war between states has declined, but ferocious civil wars rage within states, and civilians are their primary victims. Nonstate actors have also emerged as important players in international politics. All three

72

janice gross stein

were present during the cold war, but its long shadow obscured the conflicts that raged beneath the surface. In the aftermath of the cold war, it became apparent that the “sole remaining superpower” faced new kinds of challenges. These challenges were even less easily accommodated by theories of deterrence that assumed unitary actors and rational choice. In this structurally complex and uncertain system, the meaning of threat is contested, and consequently what constitutes a challenge is far less clearly demarcated. Deterrence now operates at the general level over time among the great powers, all of whom have nuclear weapons, and at an immediate level, when a challenge becomes imminent. Deterrence continues to be extended, especially by the United States to its allies. But deterrence is moving into uncharted territory. The ontology and metaphysics of threat are at the core of the debate, and the meaning that political leaders give to threat is constitutive of our understanding of contemporary global politics. After the attacks on the United States on September 11, 2001, President Bush quickly framed the challenge as a “war on terror.” That was not the only possible frame for these attacks. The challenge could have been framed as isolating a small group of hardcore militants who were prepared to use force against civilians, as a problem of policing and criminal justice, or as a challenge that the United States could meet with its existing policy instruments. By defining the challenge as requiring a war on terror that would require the preemptive use of force against irrational adversaries, Bush reframed the problem and drew a sharp dividing line between the past and the present. The fundamental challenge came from the breakdown of the established symbolic frameworks that legitimized the preexisting order. It was this rupture that led to the reframing of deterrence itself, a reframing that put attacks by irrational militants beyond the scope of both the theory and the practice of deterrence. Again, rationality was at the core of the decision to abandon deterrence, and the challenge to rationality came this time not only indirectly from scholars but directly from political leaders. Changes in the meaning, the explanation, and the understanding of threat have been informed by the increasing complexity and diversity of the international environment. Rather than linear models of causation, anchored by unitary actors making rational choices, contemporary theories are informed by analyses of complexity and epidemiology that focus on the “pathogens” that live within a system and explode once they cumulate and reach a critical mass. These kinds of threats are more difficult to see, more challenging to manage, and less amenable to the primitive notions of control that so

rational deterrence

73

informed the theories of deterrence—and coercive diplomacy—that were anchored in the cold war experience. A second factor speaks to the challenges to deterrence in theory and practice. It is the rise to prominence of new kinds of actors, with organizational structures that differ sharply from those of states. In the contemporary international system, networks of nonstate actors have joined states as important players in global politics. Today, almost all development assistance is delivered by nongovernmental organizations, security in war-torn societies is increasingly managed by private security firms, and militias, not connected directly to states, engage in violence across borders. In this far more crowded environment, states are the most important actors; but they are no longer the exclusive players in global politics. Networks are the form of social organization that increasingly defines the global system in the twenty-first century. Networks are flatter, less hierarchical than states, and more resilient insofar as they build in redundancy and duplication. The loss of one node in a network does not disable the network as a whole since information and decision making moves around the disabled node. Networks are often more nimble and more flexible in their response than hierarchies in which information travels up through channels to centralized decision makers and down again for implementation. In a “just in time” global system, state structures are slow and lumbering in their responses when compared with the lighter and more flexible responses of networks. Al-Qaeda, for example, resembles a distributed network, with a capacity for flexible response. It is resilient precisely because it is distributed; it operates through multiple nodes so that the destruction of any one is not fatal. Deterrence becomes far more difficult when a network has no fixed address, can move easily and hide, and can route its messaging around endangered nodes. After the attacks against New York and Washington on September 11, 2001, leaders in the United States struggled to craft a response that would be effective against this kind of distributed threat. They sought both to destroy the network and to deter further attacks, but the instruments at their disposal were cumbersome, ill suited to respond to a flexible network that operated below the radar. Very soon, frustration at the invisibility of their opponents led President Bush and his advisers to shift the understanding of the threat created by al-Qaeda. They moved from the unfamiliar to the familiar and lashed out against an adversary that they had long seen: Saddam Hussein in Iraq. That Saddam was unconnected to al-Qaeda and to the attacks was irrelevant. The president turned the clock back, abandoned

74

janice gross stein

deterrence and coercive diplomacy even against a state adversary, and launched a conventional, industrial-model war in a twenty-first-century global system running increasingly on networks. It is little surprise that in this kind of system, Saddam’s allies were able to organize quickly to fight a lowintensity insurgency. The United States faced simultaneously an insurgency, a civil war, and militants that found safe space in the chaos that was Iraq. In this kind of multilayered, low-intensity, and violent conflict, deterrence became virtually irrelevant in preventing or containing violence. Deterrence is more difficult in a global environment that is increasingly diverse in its ecology. Challenges come from a broader range of sources, some clearly identifiable, others not, some familiar, others new. Deterrence has consequently become more complex and poses strategic and coordination problems of a different order of magnitude. Even hypervigilant leaders face daunting challenges in scanning and reading complex environments in which chains of connection are often unclear until their effects are large enough to capture attention in a noisy and crowded world. The barriers to recognizing that a threat is looming are formidable in this kind of global system that is both tightly interconnected and evolving. Large organizations develop standard routines to monitor what they are doing. These routines do not equip them particularly well to notice what they do not expect to see. Similarly, intelligence agencies, working within existing frameworks, are less sensitive to the network effects that they are not monitoring at the time. It is, of course, a nearly impossible task to monitor the multiple chains that are weaving below the surface. Only when they amplify do they become visible, and by then the threat has erupted. In a world that is tightly interconnected across multiple levels, in a grid-like structure that works very differently from the more familiar hierarchies, those who are monitoring for threshold effects have many more places to look and much more to watch. It should not be surprising that they miss a gathering storm. The complexity of organizations, their overloaded channels of communication, and pervasive cognitive biases in the interpretation of data are not new obstacles to the identification of a growing threat, but these challenges are greater by orders of magnitude in the more tightly coupled contemporary global system. And they refract at home as well as abroad. The routines of deterrence work best when the environment is relatively stable and the challenges are familiar, but the practice of deterrence is far more difficult in a changing global system that throws up the unknowable and the unimaginable. Some crises are recurrent and familiar: the breakup of empire, the proliferation of nuclear weapons, and civil wars that erupt

rational deterrence

75

across borders. But many are not: the operations of clandestine networks bent on attacking civilians, the detonation of a dirty bomb or a biological weapon, the destruction of critical infrastructure, and the chain reactions of high-risk technologies that escalate through tightly coupled systems. The most challenging are those that policy makers have not imagined, that are unexpected and counterintuitive and that explode existing frames of reference. These are more likely in a global system that is evolving, that has no clear, well-defined structure. The challenges to deterrence in practice are broader and deeper than they were during the cold war. The construction of threat is in dispute, the meaning it is given is contested, as are explanations of its origins and life cycle. Linear explanations no longer capture the complexities of linked chains that cross a qualitative threshold to explode. The iconic image of deterrence as a strategy insulated in tightly controlled hierarchical structures is gradually fading, an artifact of a particular period of history.37

asymmetric warfare: deterrence in theory Asymmetric warfare and suicide bombing have received special attention as particularly difficult challenges to the theory of deterrence. Some argue that deterrence has become at best irrelevant in the face of these two related, but analytically distinct, challenges. In asymmetric warfare, deterrence becomes a trap in practice. The weaker party uses low-intensity warfare in a grinding strategy of attrition to provoke escalation by the stronger party in the conflict. The purpose of low-intensity warfare is to force the enemy to escalate so that the weaker “wins” a political victory. The genius of asymmetric warfare, when it is well done, is that it eliminates anything but poor choices for the stronger party. It is difficult to deter when those who challenge hope for a coercive response, so that conflict escalates and people turn against the perpetrators of violence. Soldiers kicking open doors of homes to find militants or firing on civilians to kill insurgents antagonize the civilian population whose support they badly need. Yet failing to respond is not an option when bombs are exploding in the streets, in schools, and in the markets. Damned if you do respond, damned if you don’t, as Emanuel Adler argues so cogently in this volume. The “deterrer’s dilemma” is compounded by the unwillingness of developed democratic societies both to sustain casualties and to kill others.38 Since asymmetric warfare requires far fewer resources than conventional warfare, it can be sustained over long periods of time. It relies on easily

76

janice gross stein

available technology—the construction of bombs and rockets is not demanding—which is widely available and simple to use. And it inflicts serious damage on far more expensive military assets and on civilians and domestic infrastructure, which in turn provokes widespread public anger and frustration. When leaders resort to escalation because they cannot be seen by their publics to do nothing and because they fear that their deterrent reputation will be weakened, insurgents waging asymmetric warfare achieve their political objectives and are emboldened. Deterrence quickly becomes a value rather than a strategy for “defenders” who continue to escalate to preserve their deterrent reputations. Deterrence then becomes part of the problem rather than part of the solution to conflict. The problem lies not only with deterrence as strategy but also with deterrence as a theory. The use of suicide bombing as one technique of asymmetric warfare poses a particular challenge to the core logic of deterrence as theory. Suicide bombing is not new. It was used in the modern era by the Tamil Tigers and was first used against U.S. troops in the Middle East in the 1980s. It is the highly visible use of suicide bombing against targets in New York and Washington, however, that has evoked arguments about “differences in culture” defeating deterrence. More accurately, proponents of deterrence theory would argue that suicide bombers are outside the scope of deterrence theory since they are irrational. In essence, if challengers are willing to commit suicide, to lose their lives, individually or collectively, then the strategic calculus of deterrence is beside the point. There is no credible threat that can deter those who are willing to sacrifice their lives. Deterrence is not simply fragile, it becomes impossible in a strategic culture that glorifies martyrdom.39 This argument is at its core cultural. Deterrence is impossible against those informed by a culture that considers death a lesser cost than dishonor. In this kind of cultural analysis, deterrence becomes no more than an imagined truth in a (Western) culturally idiosyncratic way.40 Choices are shaped not only by a capacity for rationality, not only by emotion, but also by cultures that frame strategic reasoning in ways that are outside the scope of deterrence. If this argument is interpreted literally, deterrence does become impossible. Schelling struggled with the problem of the credibility of the threat of nuclear retaliation during the cold war in the context of extended deterrence. Clearly, a threat to retaliate with nuclear weapons in response to an attack against an ally was not believable if that meant, in turn, that one’s homeland would be obliterated in response. To escape this paradox, he rec-

rational deterrence

77

ommended a “rational strategy of irrationality,” where leaders deliberately threaten irrational consequences and then, in order to be credible, give up their capacity to prevent their own irrational action.41 But throwing away the steering wheel of the car to communicate unequivocal commitment in a game of chicken on a one-lane road would not help to persuade an adversary who was willing to die in the car crash. Logically, it would not be effective against an adversary who values honor above life. It is difficult to conceive of a deterrence strategy that would work when people are willing to die, or even embrace death, to achieve their objectives. The death of deterrence, put forward as a cultural argument, is exaggerated. Almost all cultures instrumentalize their soldiers so that they become willing to sacrifice their lives for a cause that they believe in. Most soldiers face a possible rather than a certain death, and that is a significant difference from a culture that embraces certain death in the defense of honor. Nevertheless, armies throughout history have glorified death as a sacrifice for the homeland—or for freedom—and soldiers from a variety of cultures have willingly embraced missions that were viewed at the time as virtually suicidal. It is undeniable that within the narrow frame of deterrence, threats of harm cannot deter a would-be suicide bomber who embraces death. Not all who organize and send bombers to their death, however, refrain from a strategic calculus. Militant groups pursue positive goals on behalf of their constituencies; the individual may sacrifice life but does so in the service of a greater cause.42 Waves of suicide bombings have waxed and waned, for example, as Palestinian leaders responded to the changing imperatives of Israel’s negotiating strategies and to their domestic politics. What is important is a deeper understanding of the culture that shapes strategic choices. The strategic culture of those who engage in asymmetric warfare is certainly culturally predicated. For most, suicide bombing is a tactic of warfare within a larger cultural context. Death to avenge dishonor or humiliation, to expel the occupier, or to reclaim lost glory is a postulate that springs from a different cultural tradition. Just as the success of deterrence is contingent in part on the cognitive styles of leaders, so it is contingent in part on culture. This cultural interpretation of rationality forces a reexamination of deterrence theory, which focuses on avoiding death as the ultimate cost. It suggests that greater attention be paid to the avoidance of humiliation and shame, especially in those cultures in which honor and status matter a great deal. Strategy needs to be informed by the realization that deterrent threats

78

janice gross stein

and demonstrations of resolve can provoke rather than restrain violence if adversaries are humiliated. Options need to be framed to avoid deliberate or inadvertent humiliation and dishonor and to provide recognition and status. This is not a new concern but is central to effective strategy, as scholars argued decades ago.43

beyond deterrence as theory and as strategy I have identified three broad sets of conditions that constrain deterrence as a theory and as a strategy. The first is robust evidence, evidence that now converges from experimental psychology, from field studies of foreign policy experts, and from early research in neuroscience, that it is a category error to model rational choice as the default position and treat departures from rationality as deviant. The default position is “hot” emotional decision making that precedes reflection and analysis. There is, however, variation both in cognitive styles and in culture that creates different value frames, and these are important when thinking about the utility of deterrence. Context matters. Would-be deterrers need to pay attention to the cognitive style of their adversaries: are they more like hedgehogs or foxes? The second broad limiting condition is the difference in strategic culture that flows from deeper differences in political cultures. Leaders who come from cultures of honor and have a strong sense of grievance are especially likely to escalate in response to deterrent threats. Here the use of deterrence alone, without accompanying strategies that signal respect and recognition and provide positive incentives to address historic grievance, is likely to fail. When hedgehog leaders who come from honor cultures are the objects of deterrence, the risks of deterrence multiply. Cognitive styles and culture then become mutually reinforcing. These two streams of evidence do not wholly preclude reflection and deliberation, particularly after initial choices are made. Learning is possible, especially over time. Protracted conflicts end without the destruction of either party when the protagonists learn over time, when they develop some common knowledge and shared practices. At an individual level, the conscious brain can veto preconscious choices. There is, in other words, an opportunity to prime neural responses, to educate our brains for the next time. The brain has been “vulcanized;” just as rubber is treated with a substance to improve its strength, resiliency, and usefulness, so neurological systems leave room for reason over time.44 It is this capacity to reason and reflect after the fact that protects people against impulsivity the next time. This kind of reflec-

rational deterrence

79

tion is visible among leaders who choose, when the political conditions are appropriate, when the language of threat is complemented by respect and promises, to refrain from escalation or halt military action and enter into a process of negotiation. The sequencing of threats and promises, this argument suggests, matters. Traditionally, scholars have argued that a demonstration of resolve at the beginning is important to establish the political terrain for subsequent negotiation.45 The evidence suggests, however, that deterrence should, subject to variation in leader types and cultures, be accompanied with positive inducements at the outset to reduce the hot emotional sting produced by deterrent threats, especially in uncertain and complex environments. Deterrence works best when emotions are moderated, in stable environments. Lowering the political temperature becomes even more important when uncertainty and complexity reduce the stocks of common knowledge. A third set of limiting conditions on theory and practice comes from the structurally complex and uncertain contemporary global system. In this system, it is problematic to frame all adversaries as unitary states. Networks like al-Qaeda and militias that operate independently—or even against— their own governments are far more difficult to deter because they deliberately seek to provoke and escalate violence. As these networks proliferate, the challenge to threat-based strategies grows and deepens so that the practice of deterrence becomes an escalation trap. Here too leaders need to go beyond the logic of deterrence to think more broadly about policy, about promises as well as threats, about positive as well as negative inducements. Strategy must reach beyond the militants and the militias to create political alternatives that deepen the legitimacy of those the militants challenge. Alexander George, writing with Richard Smoke more than forty years ago, argued that deterrence was not a substitute for policy. Deterrence was not always necessary and at times not sufficient to achieve political objectives or prevent conflict. At best, they continued, deterrence could only buy time: Deterrence is a policy which, if it succeeds, can only frustrate an opponent who aspires to changing the international status quo in his favour. The consequences of continued frustration, however, are not easily predictable and are not necessarily benign. The most reliable benefit successful deterrence can offer is more time—time in which some of the conflict-generating or conflict-exacerbating elements in a historical situation can abate, so that deterrence will no longer be necessary or, at any rate, so critical for the maintenance of peace.46

80

janice gross stein

George’s sensibility was keenly political and acutely attuned to context, a sensibility that speaks to the more complex and uncertain global politics of today even more strongly than it spoke to the dilemmas of the cold war forty years ago.

notes 1. According to John Bolton, “When you have a regime [like Iran] that would be happier in the afterlife than in this life, this is not a regime that is subject to classic theories of deterrence. Retaliation for them, which would obliterate their society, doesn’t have the same negative connotations for their leadership” (“Questions for John Bolton,” New York Times Magazine, November 4, 2007, http:// www.nytimes.com / 2007 / 11 / 04 / magazine / 04wwln-Q4-t .html?adxnnl=1&adxnnlx=1). 2. George W. Bush, “Press Conference with President Nicholas Sarkozy” (Washington, DC, November 7, 2007). 3. Thomas C. Schelling. “The Limits to Rationality” (Rationality Lecture Series, Rationality: A Critical Perspective, Munk Centre for International Studies, University of Toronto, February 15, 2007). 4. Neil J. Harrison, Complexity in World Politics: Concepts and Methods of a New Paradigm (Albany: State University of New York Press, 2006); Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton, NJ: Princeton University Press, 1997). 5. Alexander George and Richard Smoke, Deterrence in American Foreign Policy (New York: Columbia University Press, 1974); Paul Huth and Bruce Russett, “Testing Deterrence Theory: Rigor Makes a Difference,” World Politics 42 ( July 1990): 466–501; Robert Jervis, “Deterrence Theory Revisited” World Politics 31 ( January 1979): 314–71; Richard Ned Lebow and Janice Gross Stein, “Beyond Deterrence,” Journal of Social Issues 43 (Winter 1987): 5–71; Lebow and Stein, “Rational Deterrence Theory: I Think, Therefore I Deter,” World Politics 41 ( January 1989): 208–34; Lebow and Stein, “Deterrence: The Elusive Dependent Variable,” World Politics 42 (April 1990): 336–69; Jack Levy, “Quantitative Studies of Deterrence Success and Failure,” in Perspectives on Deterrence, edited by Paul C. Stern et al. (New York: Oxford University Press, 1989), 98–133; Thomas C. Schelling, The Strategy of Conflict (Cambridge, MA: Harvard University Press, 1960); Thomas C. Schelling, Arms and Influence (New Haven: Yale University Press, 1966); Glenn H. Snyder, Deterrence and Defense (Princeton: Princeton University Press, 1961). 6. I deliberately exclude “reasonable” from these minimal requirements of rationality because it has much broader and deeper requirements. What “reasonable” means is both culturally and situationally specific and a contested concept. 7. Robert Jervis, Perception and Misperception in International Politics (Princeton: Princeton University Press, 1976); Jack Levy, “Learning and Foreign Policy: Sweeping a Conceptual Minefield,” International Organization 48 (Spring 1994): 279–312; Philip E. Tetlock, “Social Psychology and World Politics,” in Handbook of Social Psychology, 4th ed., edited by Daniel T. Gilbert, Susan T. Fiske, and Gardner Lindzey (New York: McGraw Hill, 1998), 870–912; Philip E. Tetlock and George Breslauer, eds., Learning in U.S. and Soviet Foreign Policy (Boulder: Westview, 1991); Yaacov Vertzberger, The World in Their Minds (Stanford: Stanford University Press, 1990). 8. Robyn M. Dawes, “Behavioral Decision Making and Judgment,” in Gilbert, Fiske, and Lindzey, eds., Handbook of Social Psychology, 497–548; William Goldstein and Robin Hogarth, eds., Judgment and Decision Making: An Interdisciplinary Reader (Cambridge: Cambridge University Press, 1996); Daniel Kahneman, Paul Slovic, and Amos Tversky, Judgment under Uncertainty: Heuristics and Biases (New York: Cambridge University Press, 1982).

rational deterrence

81

9. Philip E. Tetlock, Expert Political Judgment: How Good Is It? How Can We Know? (Princeton: Princeton University Press, 2005); Dawes, “Behavioral Decision Making.” 10. Craig A. Anderson, Mark R. Lepper, and Lee Ross, “Perseverance of Social Theories: The Role of Explanation in the Persistence of Discredited Information,” Journal of Personality and Social Psychology 39 (December 1980): 1037–49; Craig A. Anderson, “Abstract and Concrete Data in the Perseverance of Social Theories: When Weak Data Lead to Unshakable Beliefs,” Journal of Experimental and Social Psychology 19 (March 1983): 93–108; Edward R. Hirt and Steven J. Sherman, “The Role of Prior Knowledge in Explaining Hypothetical Events,” Journal of Experimental and Social Psychology 21 (November 1985): 519–43. 11. Tetlock, “Social Psychology.” 12. Tetlock, Expert Political Judgment. 13. Idid., 129. 14. Ibid., 137. 15. Isaiah Berlin, “The Hedgehog and the Fox,” in The Proper Study of Mankind (New York: Farrar, Strauss and Giroux, 1997), 436–98; Tetlock, Expert Political Judgment, 73–75; Arie W. Kruglanski and Donna M. Webster, “Motivated Closing of the Mind: ‘Seizing’ and ‘Freezing,’” Psychological Review 103 (April 1996): 263–68. 16. Tetlock, Expert Political Judgment, 88, 118. 17. Ibid., 40. 18. Ibid., 77. 19. Jonathan J. Koehler, “The Base-Rate Fallacy Reconsidered: Descriptive, Normative, and Methodological Challenges,” Behavioral and Brain Sciences 19, no. 1 (1996): 1–53; Amos Tversky and Daniel Kahneman, “Extensional versus Intuitive Reasoning: The Conjunction Fallacy in Probability Judgment,” Psychological Review 90 (October 1983): 293–315. People pay inadequate attention to the base rates of occurrence in estimating probabilities. They often attach high probabilities to low-probability events. 20. Kahneman, Slovic, and Tversky, Judgment under Uncertainty; Richard Nisbett and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment (Englewood Cliffs, NJ: Prentice Hall, 1980); Detlof von Winterfeldt and Ward Edwards, Decision Analysis and Behavioral Research (Cambridge: Cambridge University Press, 1986). 21. Susan T. Fiske and Shelley E. Taylor, Social Cognition (Reading, MA: Addison-Wesley, 1984); Robert Jervis, “Representativeness in Foreign Policy Judgments,” Political Psychology 7, (September 1986): 483–505; Amos Tversky and Daniel Kahneman, “Availability: A Heuristic for Judging Frequency and Probability,” Cognitive Psychology 5 (1973): 207–32. 22. Adler, this volume. 23. Fiske and Taylor, Social Cognition, 72–99. 24. Daniel Kahneman and Amos Tversky, “Prospect Theory: An Analysis of Decision under Risk,” Econometrica 47 (1979): 263–91; Kahneman and Tversky, Choices, Values, and Frames (Cambridge: Cambridge University Press, 2000); Amos Tversky and Daniel Kahneman, “Advances in Prospect Theory: Cumulative Representation of Uncertainty,” Journal of Risk and Uncertainty 5 (October 1992): 297–323. 25. Janice Gross Stein, “Calculation, Miscalculation, and Conventional Deterrence. I. The View from Cairo,” in Psychology and Deterrence, in Robert Jervis, Richard Ned Lebow, and Janice Gross Stein (Baltimore: Johns Hopkins University Press, 1985), 34–59; Stein, “Political Learning by Doing: Gorbachev as Uncommitted Thinker and Motivated Learner,” International Organization 48 (Spring 1994): 155–83. 26. It is not at all clear how informative data from functional magnetic resonance imaging (MRI) machines are. Neural activity occurs in milliseconds, on a scale of perhaps 0.1 millimeters. A typical MRI machine only measures neural firing indirectly by tracking blood flow and takes a picture every couple of seconds. It cannot detect anything less than three millimeters

82

janice gross stein

long. Functional MRIs also deal with regions with hundreds of thousands of neurons so they are imprecise spatially as well as temporally. Some of the neuroscientific work on choice by monkeys finds that an approximation of a mathematically derived formula appears to be hardwired into the primate brain. Paul W. Glimcher, “Decisions, Decisions, Decisions: Choosing a Biological Science of Choice,” Neuron 36 (2002): 323–32; Brian Lau and Paul W. Glimcher, “Dynamic Response-by-Response Models of Matching Behavior in Rhesus Monkeys,” Journal of the Experimental Analysis of Behavior 84 (November 2005): 555–79. 27. Dennis Overbye, “Free Will: Now You Have It, Now You Don’t,” The New York Times, January 2, 2007, D1. 28. Neta Crawford, “The Passion of World Politics: Propositions on Emotion and Emotional Relationships,” International Security 24 (Spring 2000): 125. 29. Anthony R. Damasio, Descartes’ Error: Emotion, Reason, and the Human Brain (New York: Putnam, 2004): xi. 30. This view of two operating systems is challenged by Glimcher, who argues that there is no evidence that there are two fully independent systems, one emotional, the other computational, operating inside the brain. He finds within the brain of rhesus monkeys a remarkably economic view of the primate brain in which the final stages of decision making seem to mirror a utility calculation. Glimcher, “Decisions, Decisions, Decisions.” 31. Jonathan D. Cohen, “The Vulcanization of the Human Brain: A Neural Perspective on Interactions between Cognition and Emotion,” Journal of Economic Perspectives 19 (Fall 2005): 3. 32. Colin Camerer, George Loewenstein, and Drazen Prelec, “Neuroeconomics: How Neuroscience Can Inform Economics,” Journal of Economic Literature 43 (March 2005): 9–64. 33. This summary draws on an informal presentation given by Daniel Kahneman at the Munk Centre at the University of Toronto on February 16, 2007. Daniel Kahneman, “A Psychological Perspective” (Rationality Lecture Series, Rationality: A Critical Perspective, Munk Centre for International Studies, University of Toronto, February 17, 2007). 34. Responses were monitored in functional MRI during the experiment. Activity was strongest in the interior ansula, the part of the brain that is consistently associated with negative emotional responses. Cohen, “Vulcanization of the Human Brain,” 14. 35. Jonathan Mercer, “Rationality and Psychology in International Politics,” International Organization 59 (Winter 2005): 77–106; James K. Rilling et al., “A Neural Basis for Social Cooperation,” Neuron 35 (2002): 395–405. 36. Crawford, this volume. 37. Knopf, this volume. 38. Ivan Arreguín-Toft, How the Weak Win Wars: A Theory of Asymmetric Conflict (Cambridge: Cambridge University Press, 2005); Arreguín-Toft, this volume. 39. Arreguín-Toft, this volume. 40. Adam Garfinkle, Culture and Deterrence (Washington, DC: Foreign Policy Research Institute, 2006). 41. Schelling, Arms and Influence. 42. Kapur, this volume; Monica Duffy Toft, “Getting Religion?: The Puzzling Case of Islam and Civil War,” International Security 31 (Spring 2007): 97–131. 43. George and Smoke, Deterrence in American Foreign Policy; Lebow and Stein, We All Lost the Cold War. 44. Cohen, “Vulcanization of the Human Brain,” 19. 45. George and Smoke, Deterrence in American Foreign Policy. 46. Ibid., 5.

chapter four R

Complex Deterrence in the Asymmetric-Warfare Era emanuel adler

t h i s chapter places deterrence theory and practice in context.1 Specifically, it applies the theory and practice of deterrence to a context that lies outside of the time and circumstances in which deterrence theory and practice originally developed. I call this new historical context the “asymmetric warfare era.” There is already a burgeoning literature on whether deterrence can work in asymmetric conflict situations. While some argue that deterrence ceases to work under these conditions,2 others argue that it has not lost the strategic value it possessed during the cold war.3 Whether deterrence works, fails, or even becomes detrimental is a function of social structural conditions, in particular the level of complexity of structural conditions, partly defined by the symmetry or asymmetry of strategic interactions. In asymmetric conflict situations, which pit nation-states against terrorist4 networks and other nonstate actors, such as insurgent groups and the radical revisionist states5 that support them, deterrence may not only not prevent violence but may actually help foment it. This is partly because the more states—including the most powerful states—believe in and practice deterrence, the more “weaker,” mostly nonstate, adversaries may build on deterrence beliefs and practice to elicit a violent response from which they stand to gain.6 When this happens, using force against asymmetrically weaker adversaries or exhibiting self-restraint will achieve the same result: the materially stronger state stands to lose. For lack of a better term, this dilemma, which turns deterrence into a self- defeating prophecy, can be labeled the “deterrence trap.” The deterrence trap is a structural situation that leaves social actors with 85

86

emanuel adler

the dreadful choice of reacting with force to a provocation that elicits a response or of practicing appeasement.7 Not responding is no better than responding. Thus, the deterrence trap is a lose-lose situation, in which states, including major powers, are provoked—typically by terrorist acts—into retaliating by violent means against the provocateur.8 Not responding has been socially constructed to mean loss of face, reputation, and deterrent capacity; in the end, it may spell defeat.9 Responding, however, is exactly what the weaker actor expects and hopes to achieve. The use of force against the weaker side enhances its social power and the credibility of its performance in front of domestic and foreign audiences, thus allowing it to win the war of the narratives, gain and maintain the support of a majority of the targeted population in question,10 delegitimize its enemies, and, ultimately, weaken its victims until they are beaten. The rational response to such provocation is self-restraint. But not responding is equally irrational, because it is taken as a show of weakness and undermines the capacity for deterrence, thereby inviting renewed extortion and actions that harm one’s interests and values. Thus, in the end, self-restraint is paradoxically rational and irrational at the same time. When one considers whether deterrence helps stabilize and govern the security relations of adversary states is a function of the complexity of the social structure that constitutes those relations, one moves away from realist systemic arguments, which link deterrence to the structure of the international system, defined mostly by international anarchy and the material distribution of power.11 Taking the constructivist view that strategic relations are constituted by the collective meanings that make relations possible and that, therefore, structures exist and constitute practice at the community, network, and social group level, this chapter suggests that the complexity of contextually situated deterrence social structures may help explain whether deterrence works. In other words, the complexity of deterrence social structures partly determines whether deterrence practice becomes mutually reinforcing and a self-fulfilling prophecy or whether it causes the violent behavior it is supposed to prevent, thus becoming a self-defeating prophecy. A deterrence social structure is material, but only to the extent that collective knowledge is represented in material things; it is contextual, changes or evolves in time and space, and may be characterized by higher or lower levels of complexity. While complex systems exhibit a variety of particular characteristics, such as spontaneous organization, adaptation, and nonlinear change of a relatively large number of interacting units or actors, all of which are deterministic,

complex deterrence

87

but epistemologically unpredictable,12 the notion of a complex-deterrence social structure is simplified here by referring mainly to the number of actors involved, the extent of their linear or nonlinear relations, and, most importantly, whether the main components, which constitute such structures, are characterized by symmetry or asymmetry. While this argument is a structural one, it is not deterministic: social structures do not determine things to happen, but rather they create propensities for them to happen; propensities are then actualized (or not) by human agency. The argument describes deterrence social structure to be constituted by three layers: (1) deterrence collective knowledge (deterrence culture and deterrence common knowledge), (2) the number of actors and their structural differentiation (or lack thereof ), and (3) the stratification of power, defined not only in material but also in social terms. When defining power in social terms, special emphasis will be given to what Jeffrey Alexander has called “performative power”—the social power to project a credible “performance” on the world stage, producing the desired material effects.13 The chapter argues that, on the one hand, the more symmetric the distribution of knowledge and power, the more linear the relationships and that, second, the lesser the number of actors, all of which exhibit similar structural characteristics, the higher the possibility that deterrence will become a self-fulfilling prophecy. On the other hand, the more the various deterrence social structure components are characterized by asymmetry, and the larger the number of players involved, the stronger the likelihood that deterrence will become a self-defeating prophecy and a security trap. The chapter’s next section briefly defines deterrence strategy and describes the critique that has been raised regarding its feasibility and effectiveness. In a further section, the chapter then defines deterrence social structure and refers to complex deterrence as a structural context affecting the relations of actors in a strategic situation. The third section describes the deterrence social structure that was prevalent during the cold war, which was characterized by a relatively low level of complexity. While the cold war deterrence social structure referred mainly to nuclear deterrence between the superpowers, a new and much more complex set of structural conditions, which represent the asymmetric-warfare era, were already beginning to develop then, for example, with the Vietnam War and the Soviet Union’s intervention in Afghanistan. Thus, this section also describes how deterrence social structure changed and became more complex. The fourth section explains the consequences of increased social structural deterrence complexity in asymmetric conflicts, in particular the

88

emanuel adler

notions that deterrence (1) is often used by states to prevent something that already has happened, (2) creates opportunities for nonstate actors to make states walk into a deterrence trap, (3) becomes the cause for the use of force, (4) strengthens actors who are interested in changing the status quo, and (5) eventually lowers the value of deterrence practice. It also shows that a compellence strategy, such as the one the United States adopted in the past decade, is no panacea because it requires a higher threshold of pain for an opponent and buries those who use it deeper in the social trap. At the same time, self-restraint, which most European states have exercised since the 1990s, may last only as long as Europe will be able to contain the domestic political and emotional factors that call for an adequate response. Where deterrence continues to be highly rated, like for example in Israel,14 selfrestraint against violent provocations, such as Hezbollah’s kidnapping of Israeli soldiers in the summer of 2006, can last only as long as the belief that state deterrence is not jeopardized by the provocations. Finally, the concluding section suggests that adapting deterrence to complex structural conditions requires the development of common knowledge through diplomatic engagement. Seen this way, deterrence and diplomacy are not only compatible practices but are in fact two sides of the same coin.

deterrence strategy and its critics Deterrence strategy is a coercive social logic aimed at dissuading the use of violence. It rests on premises that actors (1) are rational; (2) have common knowledge (of the rules and social logic of the game);15 (3) engage in tacit and explicit communication (understood as the exchange of information rather than efforts to arrive at collective understandings); (4) accurately assess the risks, costs, and gains of strategic games; (5) can control their emotions; and (6) hold normative assumptions about the appropriateness and proportionality of military actions. “Essentially,” as Glenn Snyder explained long ago, “deterrence means discouraging the enemy from taking military action by posing for him a cost and risk outweighing his prospective gain.”16 Deterring an opponent thus depends not only on a threat to increase the cost of violent action for the target; it also involves a promise that abstaining from violence will remove the threat.17 This definition covers both deterrence by denial—not allowing the adversary to have the capacity to commit successful aggression—and deterrence by punishment—dissuading the adversary from using force by making it more costly for the target to engage

complex deterrence

89

in unwanted behavior.18 Deterrence may be conventional or nuclear; the former encompasses all military means except nuclear weapons. The social logic of deterrence is as old as the organized use of violence, but only in the twentieth century was this logic encoded in conventional strategic doctrines and employed in political practice.19 Since the dawn of the nuclear age, what began as an acknowledgment of the prohibitively destructive potential of nuclear war—which led Bernard Brodie to note that “thus far the chief purpose of our military establishment has been to win wars. From now on its chief purpose must be to avert them”20—has been reflexively institutionalized as a strategic practice for discouraging nuclear war, both by denial and by threats of punishment; deterrence has become part of actors’ identities and how they make sense of their world.21 Ever since deterrence theory became an institutionalized practice for managing superpower relations and preventing nuclear conflict during the cold war and after, both theory and practice have been the targets of strong criticism. Using arguments from social and cognitive psychology, critics faulted its substantive rationality assumptions.22 Others focused on domestic and bureaucratic politics;23 still others rejected deterrence theory as historically specific and thus unscientific.24 Another argument was that the real problem was not the theory of deterrence but its strategic practice, which made the situation more volatile and in fact prolonged the cold war.25 With the advent of interpretive critical and constructivist international relations theories, deterrence theory was criticized for its instrumental conceptions of rationality, for its commitment to methodological individualism, which neglected the role of intersubjective understandings and the constructed nature of social reality, for its contractual account of international politics, which neglected the effect of identity on interests, and for its lack of attention to purposes and norms.26 These faults, however, did not reduce the belief in what experts, even during the cold war,27 considered to be general deterrence,28 namely, the internalized understanding of the existential dangers of nuclear conflict. After the cold war, experts maintained that general deterrence could be strengthened by agreed-upon norms of appropriate behavior,29 by a sense of community between deterring and deterred actors,30 and by deterrence communities that are constituted by the internalization of a deterrence norm.31 For example, the greater the internalization of international norms, the more difficult it would be for terrorist organizations and the radical revisionist states that assist them to act with impunity and thus the greater the deterrent effects. Deterrence, from this point of view, may still offer a practical and

90

emanuel adler

normatively superior alternative to preemption and prevention strategies, which are based on the use of force.32 Not necessarily, counter critics of post–cold war deterrence.33 First of all, threats against nonstate and nonterritorial actors, such as terrorist organizations, whose members can hide in their victims’ own cities or use parts of or entire states to launch their attacks from behind the cover of civilian populations, are likely to be empty, while promises of cooperation may simply invite more terrorism.34 Second, terrorist groups that are highly motivated and have no territorial base are, along with their state sponsors, unlikely to be deterred by fears of punishment or death.35 Finally, the critics say, the lack of common knowledge shared by Western states and terrorist organizations and the radical revisionist states that assist them, the absence of common norms, and the possibility that terrorists and radical revisionist states may be irrational, “and thus unresponsive to cost-benefit calculations required for deterrence,” disqualify deterrence as a viable strategy for dealing with post–cold war security threats.36

deterrence social structure Whether a deterrence strategy is effective and valuable or whether it may fail or even become detrimental to security depends on the structural complexity of the context in which the strategy or doctrine is situated. Thus, attention should be given not only to the conflicting actors themselves, their religious beliefs and their mode of operation, but also to factors beyond the individual and collective actors that impinge directly on deterrence’s strategic complexity. Deterrence social structure refers to a three-tiered set of material and epistemic background conditions and contexts, the material ones of which, according to Alexander Wendt, “only acquire meaning for human action through the structure of shared knowledge in which they are embedded.”37 The top tier, which stands in a constitutive relationship with the other two tiers, is deterrence collective or background knowledge, which consists of a macrodimension, deterrence culture, and a microdimension, common knowledge. Not unlike Wendt’s macrostructural “cultures of anarchy,”38 deterrence culture refers to intersubjective knowledge, which is not only in individuals’ heads but is also embedded in institutions and practices, in particular, cause-effect and normative understandings about the human quality and ability to deter and be deterred and collective understandings of rationality. Common knowledge means that we each know that the other

complex deterrence

91

knows and that the other knows that the other knows and so forth.39 Thus, while common knowledge is still intersubjective, confronting “actors as an objective social fact that cannot be individually wished away,”40 it nevertheless is also constituted by the strategic relations of individual actors and thus takes place at the individual level of intentionality. As such, the creation of common knowledge still influences the ordering of favored institutional outcomes and has an independent impact on actors’ bargaining power.41 Deterrence structure’s second tier consists of the number and differentiation of actors involved in international and transnational strategic conflict situations. The number of actors refers to the players in a deterrence game, that is, actors who enter into the calculation of other actors when engaging in deterrence practice, either as deterring actor or as deterred actor. The differentiation of actors not only reflects whether actors are morphologically and institutionally similar or different—in other words, whether they are either all state actors or all nonstate actors or whether they are a mix of state and nonstate actors—but also whether they are constituted, and separated from each other, by the same or different principles, for example, sovereignty in the case of nation-states and degrees of separation in the case of terrorist nonstate actors.42 This kind of differentiation (or lack thereof ) creates structural asymmetry or symmetry at the actor level. Structurally symmetric actors will most likely engage in symmetric warfare and deterrence relations; asymmetric structural conditions among actors will produce asymmetric warfare. Deterrence structure’s third tier consists of the distribution of power and knowledge among relevant actors. The distribution of power not only reflects the stratification of material capabilities but also the distribution of productive knowledge, which, to paraphrase Michel Foucault, defines the order of things.43 Power thus defined is social because it produces effects that shape the abilities of actors to determine their circumstances and fate through social relations.44 Social power depends on individual and collective capacities to construct social reality. A prime example of social power was the U.S. capacity during the cold war to define what the practice of deterrence and arms control meant in its strategic interactions. Social power is not only the power of making things happen as one wishes but also of constructing social reality or actors’ identities and practices.45 Social power is not only intrinsically related to actors’ authority and legitimacy46 but also to the ways in which meanings and discourse become socially authoritative and legitimate. Social power is also bound up with practices: only in and through practice can material resources be directed toward a particular purpose or target. For

92

emanuel adler

example, only through practice may states jointly know the deterrence value of specific material capabilities and technologies, whether these capabilities and technologies help control other actors’ circumstances and fate, and whether they may become counterproductive. Cold War Stable Nuclear Deterrence: Low-Deterrence Social Structure Complexity The cold war regime of stable nuclear deterrence and the social structure that constituted it were determined neither solely by military capabilities and technologies nor by the strategic interaction of rational actors who attempted to calculate the cold war into extinction. Rather, stable deterrence also resulted from a learning process, fraught with crises and opportunities, that consisted in the institutionalization of cause-effect and normative understandings of nuclear war in and through practice. So deep was this historical and cultural process that deterrence practices have outlasted the social structural conditions that constituted the cold war and that, to some extent, inform this book’s theory and empirics. The cold war’s nuclear deterrence practices were based on a deterrence culture founded on Western civilization’s secular beliefs and the Enlightenment’s emphasis on human reason and modernity. The latter stressed rational and instrumental means and efficient institutions, the notion that passions can be tamed by reason, diplomacy, raison d’état, and international institutions as well as a concept of social order that enshrined the inviolability of human life, the legal sanctity of private property, and pacta sunt servanda, the understanding that promises must be kept.47 The Enlightenment tradition also subscribed to the view that political power relations are primarily driven by dispassionate calculators who make political decisions on the basis of cost-benefit analysis considerations.48 Thus, what made American and Soviet political and military institutions and bureaucracies during the cold war increasingly converge toward similar strategic rationales was not only the common need to prevent nuclear war but also a modern, socially constructed, notion of rationality. This shared understanding of rationality helped social actors adopt practices based on cost-benefit analysis calculations, even if they also were guided by social psychological and emotional factors. Rational choice and instrumental strategic knowledge, albeit partially, played a role in helping constitute theories and practices of stable deterrence and arms control, which to some extent helped keep the cold war cold.

complex deterrence

93

The deterrence culture of nuclear deterrence (and arms control) during the cold war sustained (and was sustained by) common knowledge, which helped build stability into the deterrence relationship. Deterrence behavior, of course, has been around since time immemorial, perhaps as a result of human instinct, maybe also as the emotional response to danger and threats.49 The specific theory and practice of nuclear deterrence during the cold war, however, was made possible by common knowledge, consisting of common meanings, social technologies and discourses of deterrence, which were diffused and institutionalized by epistemic communities.50 For example, there was widespread agreement related to understandings of crisis stability, deterrence by denial, deterrence by retaliation, extended deterrence, first and second strike, mutual assured destruction (MAD), nuclear arms control, and more. These were first and foremost conceptual understandings, yet they became practices via political and institutional processes. To threaten nuclear punishment, but to promise, if the adversary was dissuaded, not to escalate and draw red lines and to limit defensive, rather than offensive, weapons— these are, in essence, what deterrence practice was all about. Not all American and Soviet practitioners and experts, however, accepted deterrence as the only, or the more efficacious, means of projecting military power to achieve political goals.51 Common knowledge also was not monolithic when it came to deterrence’s modalities, scope, and reach. Within the United States and the Soviet Union, the convergence around deterrence practice was often constrained by domestic politics, ideological disagreements, and civilian-military disputes about the role of nuclear weapons. Still, deterrence strategy and practice allowed cold war adversaries to share expectations of proper action and rationally to weigh policy options according to common knowledge of the situation.52 The deterrence social structure’s second tier (number and differentiation of actors) of mutual nuclear deterrence during the cold war consisted mainly of two actors, the United States and the Soviet Union, which were structurally undifferentiated. In other words, both actors were nation-states, which were constituted by sovereignty. This symmetric structural condition meant a low level of social complexity, thus turning the development of mutual deterrence into a self-fulfilling prophecy and a shared practice. Deterrence social structure’s third distribution of power and knowledge tier was no less important in setting structural conditions of low complexity, which promoted the practice of deterrence between the superpowers. The practice of nuclear deterrence during the cold war was thus not only enabled by a balance of material military capabilities and technologies, which

94

emanuel adler

produced so-called parity between the superpowers, but also by the symmetric cognitive recognition of parity and MAD, which affected actors’ expectations and enabled the control of nuclear events. The power inherent in knowledge distribution does not refer to what Joseph Nye called “soft power,” the purposeful diffusion of one’s culture.53 Rather, it reflects the ability to deter and be deterred, control events, and stabilize the situation. The symmetric distribution of knowledge strongly affected the capacity of both superpowers to practice mutual deterrence and sustain a deterrence culture as long as the cold war lasted. Thus, regardless of whether cold war nuclear deterrence directly helped prevent nuclear war or whether it indirectly created a stable environment that allowed diplomatic relations to do the work and for the cold war to end cold, the relatively symmetric (material and social) power relations between the superpowers was conducive to shared beliefs about the “rules of the game,” which helped constitute low complexity and thus turned mutual deterrence into a self-fulfilling prophecy. To sum up, in situations of low structural complexity—characterized by (1) symmetric deterrence collective knowledge; (2) the presence of mainly two actors, both of which are morphologically similar and constituted by the same set of principles; and (3) a more or less symmetric distribution of material and social power, the best example of which is nuclear deterrence during the cold war—deterrence is likely to become a self-fulfilling prophecy. Deterrence in the Asymmetric-Warfare Era: Complex-Deterrence Social Structure With the end of the cold war, the breakup of the Soviet empire, and the move of the center of strategic relations from superpower relations to regional security and asymmetric warfare, deterrence social structure became more complex than before. A change from a symmetric deterrence social structure to an asymmetric deterrence social structure can thus partly explain not only why “stronger” actors are recently having such difficulty in deterring “weaker” actors but also why deterrence as a theoretical concept and political practice has become part of the reason why this is so. To begin with, deterrence collective knowledge has changed drastically. Deterrence culture is now split into two opposing cultures. One culture is the extension of the cold war deterrence culture that is prevalent mainly, but not exclusively, in the West. The other deterrence culture can be understood as either building on a response to Western modernity and rationalism or as representing a genuine alternative path to modernity, whose conception of

complex deterrence

95

rationality may differ from that of the alternative deterrence culture.54 This deterrence culture conditions terrorists to take deterrence less as a game to be played for mutual gains and thus for preventing the use of violence than as a tool to win the war of the narratives. To actors constituted by this culture, deterrence is but yet another Western doctrine and practice, such as international law and human rights regimes, to be used instrumentally against those who believe in it. Because this deterrence culture is driven in part by terrorist actors’ strong ethnic-nationalist beliefs, deep religious sentiments, or worse, a deadly combination of the two, the structural possibilities of developing not only collective notions of legitimacy and fair game but also common knowledge are limited. It is in this context that we ought to understand Ivan Arreguín-Toft’s argument in this volume about the epistemic abyss separating Western states and the terrorist networks’ understanding of costs. The “fear of death,” which European elites cultivated for centuries and which as part of deterrence culture led to the mutual power to “take life” and to deter, stops at the terrorist actors’ gates.55 When this happens, a deeply troubling complex asymmetry sets in, not only because of the epistemic divide between the parties, but also because both terrorist networks and their target states become reflexively aware of the power implications of this asymmetry. Terrorist networks try to use the asymmetry to set the deterrence trap, and their victim states try to escape it. Equally important for deterrence culture in the asymmetric-warfare era is the notion that stable nuclear deterrence during the cold war was based on the collective assumption that when nuclear weapons and their deterrence came into play, emotions rested or were at least controlled by a combination of politics and technical means. As Janice Stein shows in this volume, however, recent advances in neuroscience put into question the substantive model of rationality by concluding that “decisions seem not to be the result of deliberative thought process,” but that “emotion precedes choice,”56 and that, as Daniel Kahneman has begun to show, decisions result from a double-edged system, the dominant side of which is preconscious, automatic, fast, effortless, associative, and unreflective, while the other side is conscious and reasoned.57 Thus, stability of deterrence is likely much more affected by a “hot” emotional reaction to betrayal, disappointment, shame, or humiliation than it is by a “colder” calculus of capability and resolve.58 This means that substantive rationality may have been less important during the cold war than we originally thought. What distinguishes the cold war from the asymmetric-warfare era, however, is that during the cold war, not only

96

emanuel adler

the lethal nature of nuclear weapons, but also the common knowledge that developed between the United States and the Soviet Union helped channel passions in the direction of preventing nuclear war. In a split deterrence culture, however, which characterizes our times and according to which Western actors still believe that substantive rationality and common knowledge can tame passions, while non-Western terrorist networks, although often acting strategically, let their passions drive their instrumental actions, the most likely scenarios are that Western actors are either going to walk into the deterrence trap or are going to adopt compellence or preventive practices, in the erroneous belief that the latter will help them escape the trap. The U.S. quagmire in Iraq shows that compellence and preventive practices elicit an even greater and more violent response, which makes the trap direr and more difficult to escape. Also, because terrorist networks use Western human rights norms to gain support for their cause and to win the war of the narratives with their local audiences, normative structures also become part of the current complex split deterrence structure. Terrorists may hide their weapons inside civilian houses, schools, and places of worship with the goal of reducing vulnerability and denying the adversary useful targets. When, however, civilians are targeted, because this is where the terrorists hide, or when because of the terrorists’ actions, civilians become the victims of collateral damage, terrorist networks skillfully appeal to human rights regimes via the media and accuse opponents of violating basic human rights principles—a charge that is often correct. This terrorist strategy of inviting and then lamenting human rights violations makes a mockery of basic human rights norms. Leaving aside the ethical dilemma in this situation, this behavior is structurally driven by an asymmetric deterrence culture. In situations of complex deterrence, common knowledge becomes more difficult to obtain. First, under asymmetric-warfare conditions, communication becomes more difficult and sometimes does not exist at all. While even at the height of the Cuban Missile Crisis, American and Soviet leaders exchanged messages and met in open and closed forums, hardly any channels of communication exist today between American leaders and al-Qaeda leaders or Israeli leaders and Hezbollah leaders. Second, because of the asymmetry in deterrence cultures, it is hard for common social technologies, institutions, and languages to develop. Under these conditions, the main concepts that gave deterrence its meaning, both in theory and in practice, become contextually unfit. Worse, they may be used instrumentally to tend the deterrence trap, as in the use of human rights norms by terrorists. Thus, for

complex deterrence

97

example, to sustain the argument that the United States or Israel will be able to deter a nuclear Iran requires a social structural assumption that common knowledge between them is possible. Changes in deterrence social structure’s second tier (the number and differentiation of players) also can render deterrence more asymmetric and thus more complex. When, for example, deterrence enters into the consideration of more than two actors, which are structurally differentiated— in other words, when in terms of their morphology and institutions and the principles that make them the way they are, actors are asymmetrically constituted—deterrence social structure becomes more complex. The growing asymmetric confrontation between nation-states and nonstate terrorist networks, such as al-Qaeda and Hezbollah, produces a major increase in complexity, uncertainty, and lack of mutual understanding. During the cold war, because nuclear deterrence took place between territorially separated entities, deterrence signaling in nuclear bargaining situations, such as in the 1962 Cuban Missile Crisis, was less difficult to understand by the adversary. When, however, actors confront each other less across territorial borders than in the territory and cities they mutually share—when, for example, alQaeda terrorists may be U.S. citizens who lie dormant in an American city ready to attack their fellow citizens—the complexity of, and difficulties involved in, deterrence signaling and social communication grows markedly. Under asymmetric conditions, thus, deterring adversaries becomes much more difficult, if not impossible. Deterrence social structure’s third tier (the distribution of power and knowledge) only adds to this complex situation. With the end of the cold war, and as the problem of asymmetric warfare became one of the most important features of the era that replaced it, material and social power, including collective knowledge, became asymmetrically distributed. From a material perspective, for example, the United States and Israel are vastly more powerful than al-Qaeda and Hezbollah. But al-Qaeda and Hezbollah have partly rendered the United States’ and Israel’s power impotent; neither state can actually use its overwhelming military superiority to overcome their weaker adversaries. Explaining this puzzle first requires returning to the asymmetric deterrence culture mentioned above.59 In asymmetric-warfare situations, when modern states, which owe their constitutive principles to Western civilization, have, as was mentioned before, internalized the fear of death and acquired the power to deter but their adversaries, nonstate terrorist networks, have not done so, the result is a deep lack of shared common knowledge about costs, which prevents states from deterring nonstate

98

emanuel adler

terrorist actors.60 A split deterrence culture thus helps to create the social power of terrorist networks and to make states structurally inferior. Modern states, in fact, lose their power to take life and deter, while terrorist actors reflexively acquire and instrumentally use the power to beat their enemies by virtue of their lack of fear of death and their insensitivity to cost. Under these conditions, asymmetric power and knowledge distribution becomes a distinctive feature of complex deterrence. Second, the explanation of the puzzle identified above lies neither in the seeming incapacity of states to deter determined terrorist networks, let alone suicide bombers, who are moved by deep religious beliefs, nor solely in the nature of terrorist networks and insurgents’ capacity to provoke states to use violence against them. Rather, the ability of the weaker side to overcome the stronger side lies in the former’s “performative power,”61 that is, the social power to project a dramatic and credible performance on the world stage, on the assumption that, to some extent, public audiences determine who are the winners and losers of international violent contests, thus bestowing terrorist networks’ narratives with legitimacy and helping the recruitment of new supporters to the cause. The best example of performative power is the performance Hezbollah “produced” for local Lebanese, Arab, Muslim, and global public opinion audiences in 2006. This performance, which involved not only media and image manipulation but also the use of Western human rights norms to score points in front of these audiences, enabled Hezbollah to claim victory over the Israelis in the last Lebanon war. Victory thus became a function not only of Israelis being unable to stop Hezbollah’s barrage of missiles against Israeli cities but also of Hezbollah’s performative power. Hezbollah’s performative power was effective because neither friend nor foe thought about power in these terms. If they did, Hezbollah’s social trap would have become evident and the performance would have lost its effect. The asymmetric distribution of material power and social power, including a capacity of weak actors to produce powerful and successful performances on the world stage, helps create a highly complex deterrence social structure that renders deterrence self-defeating. In highly complex structural situations, characterized by asymmetric (split) cultures, a lack of common knowledge, an asymmetric differentiation of actors and distribution of power and knowledge, there is a high propensity for deterrence practice to become self-defeating. Nonstate terrorist networks use states’ deterrence beliefs and practices to force states to react violently to provocations, thus augmenting the terrorists’ power and follow-

complex deterrence

99

ing. States, on the other hand, fall into the deterrence trap and use violence for deterrence’s sake. An increase in complex deterrence, thus, may not only challenge many of the epistemological assumptions on which deterrence theory and practice were based, but it may also change the ontological state of national and international security, where the practice of deterrence now becomes a cause for the use of force and the reason why weaker actors consistently tend to beat stronger actors in asymmetric structural situations.

the complex-deterrence context: the deterrence trap Complex deterrence can become a self-defeating prophecy. First, during the cold war, the main strategic problem or dilemma was surprise attack and preventing nuclear war, which meant threatening retaliation, but also promising self-restraint if the threat was successful.62 In the asymmetric-warfare era, which is marked by deterrence complexity, the main strategic problem is that retaliation and self-restraint are equally damned strategies. If, for example, the United States threatens to use force against Iran in order to prevent it from developing nuclear weapons, Iran may actually build on the threat to strengthen its determination to rally Iranians around the nuclear-weapons program and recruit supporters in the Muslim world. Allowing Iran to develop nuclear weapons without impediment, however, and then relying on deterrence to dissuade Iran from using or threatening to use these weapons would be equally problematic. This is because of the complex-deterrence structure, which includes the lack of common knowledge between the West and Iran as well as Iran’s power to control events via Hezbollah, Hamas, and even large sections of Muslim (including Sunni Arab) masses, for whom facing up to the United States and Israel may be more emotionally rewarding than beating the Shia, their traditional rival in the Muslim world. Second, in complex-deterrence situations, the belief in deterrence strategy may in fact be so deeply internalized that even when, owing to political and military circumstances, states may decide to adopt forceful compellence or preventive strategies, as the United States did after the September 11 attacks, the preventive use of force in the present is likely to aim partly at having a deterring effect in the future. Unless one believes that a very amorphous enemy, such as al-Qaeda, could be totally eliminated by military preventive actions, one cannot escape the conclusion that American post–September 11 preventive military actions, particularly in Iraq, were designed not only to prevent the next attack but also to dissuade the enemy from carrying out

100

emanuel adler

future terrorist attacks against U.S. territory and citizens. Thus, armed with inadequate knowledge of a situation of complex deterrence—which is characterized by high uncertainty about cause-effect relationships, the adequacy of goals, and the effectiveness of means—the United States went to war in Iraq partly to deter Islamic terrorism, expecting that terrorists in the Arab world would internalize the deterrent demonstration effect. Complex deterrence, however, frustrated American intentions, and the United States is now deeply buried in a long-term deterrence trap. The effects of complex deterrence are even more insidious. When states, such as the United States after September 11, use force for the sake of longterm deterrence against terrorist networks, the latter may be structurally positioned to place those states in jeopardy. Trying to deter terrorists in the long run may not only fail to achieve the desired goal, but it may in fact strengthen terrorist networks and their case in front of public opinion. And when states use force against terrorists due to the latter’s provocations, according to Lawrence Freedman, a “vigorous response [which supporters of preemption favor] is actually what the terrorists want, in that it will radicalize those who happen to get in the way and become victims of ‘collateral damage.’”63 Not responding, however, is perceived as weakness and invites further aggression; this too is out of the question. The deterrence trap, thus, means that using force may not be more effective than not using it. Escaping the deterrence trap requires neither inaction nor the use of more force, but—following Nathaniel Fick, who vividly declared, “Welcome to the paradoxical world of counterinsurgency warfare—the kind of war you win by not shooting”64—it requires taking actions that delegitimize the terrorist networks in the eyes of their current and potential supporters. Third, Israel’s use of force against Hezbollah in 2006 is a good example of how a strongly internalized and institutionalized belief that the security of the country depends on deterring the enemy triggers the deterrence trap. When Israel retaliated against Hezbollah, it did not abandon its deterrence doctrine and practice. To the contrary, force was used for deterrence’s sake, because Israeli political and security leaders firmly believe in deterrence as a social fact and therefore claim that Israel’s existence depends on its reputation to deter its enemies. Thus, rather than watching their deterrence power go by the wayside if, when repeatedly provoked, they exhibit self-restraint, Israeli leaders opted for using force for future deterrence’s sake. As a result, not only did they fall for the terrorists’ social traps, but they also turned deterrence theory and practice into a reason for, or cause of, their violent response.65 The deterrence logic helped determine the military response.

complex deterrence

101

One possible test of this argument would be to take a look at states that have abandoned deterrence doctrine and practice, as key European states have, and observe whether they react to provocations with the threat of military force. If the argument advanced here is correct, European states will not react to provocations militarily; instead, they will respond with cooperative security strategies as laid out in the European Security Strategy.66 While it would be possible to argue that the kind of self-restraint shown by Spain and Britain after the terrorist attacks in Madrid and London was the result of weakness, the alternative explanation is that a military response did not take place because the deterrence “chip” was not there or had been deactivated.67 While Europe’s overall military and political situation is obviously different from that of Israel, the comparison highlights the point about the relationship between deterrence’s complex logic and the use of force. Fourth, when, as in the asymmetric-warfare era, there are more than two actors involved in strategic relations, and when some of these actors are not states but terrorist networks, such as Hezbollah or al-Qaeda, the logical structure that deterrence strategy and practice used to depend on during the cold war is replaced by a different logic. This is not only because of the increased difficulty in the asymmetric-warfare era of signaling and communicating, of understanding the meaning of concepts and actions, and of threatening effectively, but also because nonstate actors do not respond to the same notions of rights and duties as states do. In the asymmetric-warfare era, therefore, deterrence logic is based on emotions, fears, and a desire for vengeance to a greater extent than it probably ever was during the cold war. Take, for example, the slowly brewing Iranian nuclear crisis. The main players are Iran, the United States, and Israel. But there are also Russia, China, the European Union, Syria, Saudi Arabia, Egypt, other Sunni Arab countries, Iraq, Hezbollah, the insurgents fighting in Iraq, and maybe also al-Qaeda. Strategically, these actors enter into each other’s calculations when determining how to act, against whom, and why. For instance, while all political actors in the United States are increasingly concerned with the possibility that Iran will develop nuclear weapons, some pundits advocate attacking Iran before it crosses the nuclear threshold,68 while others believe that Iran “can be deterred.”69 In the meantime, compelled by a long history of persecution and by Holocaust memories, which Iran’s president Mahmoud Ahmadinejad has helped to stir up, Israel must decide whether to launch a preventive attack against Iran if the United States does not—at which time the United States in turn would have to decide whether an Israeli attack is in

102

emanuel adler

its own interest. Successfully destroying Iran’s nuclear facilities may depend to a large extent on preventing Hezbollah from threatening Israel with thousands of missiles. Hezbollah’s action, meanwhile, may not be driven only by an Iranian rationale but also by domestic Lebanese politics. Thus, Hezbollah may decide once again to provoke Israel to respond violently, in which case Syria may become involved this time. And al-Qaeda, which can profit from the whole situation, may choose to attack U.S. and European targets. In the context of complex asymmetric deterrence, it will be more difficult to deploy a deterrence doctrine not only among a plethora of actors but also among a set of actors that includes both states and networks. Fifth, it is highly questionable in complex-deterrence situations whether deterrent threats work with radical revisionist states, such as Iran, which are in the process of creating power-projection capabilities to facilitate their regional hegemonic designs. Although the United States and Israel, for example, may be able to deter Iran from attacking Israel with nuclear weapons, they probably will not be able to deter a nuclear Iran from using Hezbollah or similar groups against Israel and the United States with impunity. At the same time, Iran may be able to deter the United States and Israel because in complex-deterrence conditions, being deterred from using force against radical provocateurs is slightly better than not being deterred and using force against them. Thus, the radical revisionist actor may deter the status quo actor but not the other way around. Finally, the concept of social power can add nuance to the mechanism through which the deterrence trap works. Terrorist organizations and their state sponsors enhance their social power by their ability to bear very large costs or by acting in ways that do not appear to coincide with the traditional notion of rational cost-benefit analysis. Whether their behavior is intentionally “irrational” (in order to confound their adversaries) or reflects a socially constructed set of rational rules (which happen to differ, historically and culturally, from Western rational rules), the result is the same: their social power is augmented.70 The lack of common norms between terrorist groups and states they target are not only about a Western versus non-Western or liberal versus nonliberal divide but also about a territorial state versus nonstate actor divide. Because the territorial state carries legitimacy and responsibility norms that do not apply to nonstate actors such as al-Qaeda and Hezbollah, nonstate actors—and terrorist groups in particular—may not be able to feel responsible for their dead in the same way as states do and therefore may be harder to deter. Moreover, although terrorist groups may be able to learn,71 their dependence on leadership authority, rather than on rational-legal or

complex deterrence

103

bureaucratic authority, can make their lessons more ephemeral, thus making it harder for them to be held accountable.72

conclusions Whether deterrence works, fails, or becomes detrimental is a function of social structural conditions, in particular the level of complexity of structural conditions, defined by the symmetry or asymmetry of strategic interactions. In asymmetric conflict situations, deterrence may not only be unable to prevent violence but may also help foment it. Using force against asymmetrically weaker adversaries or exhibiting self-restraint becomes a deterrence trap, which is likely to bring about defeat. Thus, curbing organized violence in asymmetric complex conditions requires not only coming to grips with the actors’ interests, beliefs, and modes of operation, but also with deterrence’s own structural complexity. I have argued elsewhere that defusing may help dismantle the deterrence trap. Defusing, both as an alternative and as a complementary strategy to deterrence, aims at thwarting terrorist networks and their supporters’ attempts to drag their adversaries into using force against them. By means of dramatic political changes and the creation of new agendas and rules of the game, defusing seeks to change the structural conditions that constitute the no-win choice of retaliating or appeasing terrorist provocateurs.73 This chapter explores a different strategy: common knowledge development as a means for adapting deterrence to complex structural conditions, thus making deterrence and diplomacy more compatible. Using the example of Iran’s nuclear-weapons challenge, the conclusion argues that diplomacy may be a means not only for exploring the conditions that may lead Iran to forego nuclear-weapons development but also, if Iran nonetheless acquires nuclear weapons, for creating shared meanings, rules, social technologies, and discourses of deterrence, which can make deterrence strategy and practice more effective for preventing nuclear war. From this perspective, diplomacy does not have to damage deterrence credibility; on the contrary, it may actually help create the epistemic and social conditions for its realization. Developing common knowledge in the asymmetric-warfare era requires, first, that strategic experts and political and military leaders identify diplomacy as a common knowledge-making instrument. Second, great care should be taken to advance the creation of common knowledge through bilateral and multilateral diplomacy—perhaps also with the help of “epistemic communities”74—not as the unilateral transmission of nuclear deterrence

104

emanuel adler

knowledge from the West to Iran, but rather as mutual learning. While Iranians may already be acquainted with various aspects of Western deterrence strategies—which they use to tend the deterrence trap—they nonetheless may be unwilling to buy into them completely and explicitly because they come from the West. Third, building common knowledge through diplomacy may require learning about each other’s identities and worldviews, for example, how religion enters rationality, not only in Iran, but also in Western countries; the contenders’ attitudes toward nationalism, victory, and death; or how honor and pride influence instrumental choices.75 Formal and informal diplomatic forums, for example, may help explore humiliation and shaming as part of the strategy of conflict. Denying the opponent the sense of victimization that often accompanies strategies that humiliate and cause shame can be important here; and a creative strategic dialogue may not only help produce common knowledge about such things but may also create a minimum measure of trust that may be needed to move forward. Fourth, it is likely that arriving at common knowledge may require contenders to learn different lessons. Western countries may need to learn about Iran’s tacit and explicit rationality rules and practices, the background knowledge on the basis of which Iranians make instrumental decisions, how Iranian leaders compare means to ends and assess risks, and whether their strategic thinking is affected by deterrence and, if it is not, which rules and standards they use. Iranians, on the other hand, need to get better acquainted with the nuances of Western rationality not only at the theoretical but also at the practical level, for example, in what ways rational choice and costbenefit analysis enter into the calculations of Western countries. Via common knowledge–creating diplomacy, Iranians may also come to appreciate better the practical knowledge that the United States and Russia acquired during the cold war, such as practical understandings and technologies about preventing unintended nuclear war. While Westerners may find it in their interest to transfer to the Iranians technologies that help prevent accidental nuclear war, Iranians may find it in their interest to learn that deterrence moves, which apparently signal malevolent intentions, may actually be intended to prevent mutual suicide.76 Finally, constructing common knowledge via diplomacy aims not only at placing the contending parties inside a common cognitive—individual and collective—bubble, but also at acquainting them with each other’s emotions, particularly each other’s fears,77 such as the fear of “the other,” of compromising deep values and identities, and of conceding too much. If

complex deterrence

105

only because fear may be as much a requirement for deterrence as it may be harmful for the ability to deter and be deterred, deterrence in complex structural conditions may seldom work if contenders do not know enough about each other’s fears and related emotions. To sum up, complex deterrence depends on the construction of common knowledge; diplomacy, formal or informal, was and still remains the best, if not the only, way to attain it.

notes 1. For their very useful comments and suggestions I would like to thank Michael Barnett, Timothy Crawford, Patrick James, Jeff Knopf, Oded Lowenheim, Amir Lupovici, Jennifer Mitzen, Patrick Morgan, T. V. Paul, Vincent Pouliot, Janice Stein, James Wirtz, and especially my outstanding research assistant Patricia Greve. 2. Robert A. Pape, Dying to Win: The Strategic Logic of Suicide Terrorism (New York: Random House, 2005). 3. James H. Lebovic, Deterring International Terrorism and Rogue States: U.S. National Security Policy After 9 / 11 (London: Routledge, 2007); Robert F. Trager and Dessislava P. Zagorcheva, “Deterring Terrorism: It Can Be Done,” International Security 30 (Winter 2005 / 6): 87–123. 4. Throughout this article I will refer to “terrorist” organizations because of lack of a better term and the frequency with which the groups in question engage in terrorist acts. I believe, however, that the concept of terrorism is losing its value; it describes everything, thus nothing. Even the goal of instrumentally socially constructing enemy militant organizations as terrorists has backfired. In the case of Hezbollah, for example, it has led Israel to believe its own construction and to be militarily unprepared for conflict with a powerfully armed militia. 5. Radical revisionist states express dissatisfaction with the status quo and thus would like to change it, if necessary by disregarding the international system’s regulative and constitutive norms. See Arnold Wolfers, Discord and Collaboration: Essays on International Politics (Baltimore: Johns Hopkins University Press, 1962). 6. Emanuel Adler, “Damned If You Do, Damned If You Don’t: ‘Performative Power,’ and the Strategy of Conventional and Nuclear Defusing” (unpublished manuscript, Munk Centre for International Studies, University of Toronto, January 2007); Amir Lupovici, “Security, Identity, and Israeli Deterrence during the Second Lebanese War” (unpublished manuscript, Department of International Relations, Hebrew University, Jerusalem, July 2007 [in Hebrew]); Barak Mendelsohn, “Israel’s Self-Defeating Deterrence in the 1991 Gulf War,” Journal of Strategic Studies 26 (December 2003): 83–107. 7. Adler, “Damned If You Do, Damned If You Don’t.” 8. See David A. Lake, “Rational Extremism: Understanding Terrorism in the Twenty-First Century,” Dialogue IO 1 (Spring 2002): 15–29. 9. For the classic argument that reputations for resolve are worth fighting over, see Thomas Schelling, Arms and Influence (New Haven: Yale University Press, 1966). For a critical view, see Jonathan Mercer, Reputation and International Politics (Ithaca: Cornell University Press, 1996). 10. For a related argument, see Ivan Arreguín-Toft, How the Weak Win Wars: A Theory of Asymmetric Conflict (Cambridge and New York: Cambridge University Press, 2005), 224–25. On the calculations of weaker powers see also T. V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge: Cambridge University Press, 2004). 11. Kenneth N. Waltz, Theory of International Politics (New York: McGraw-Hill, 1979).

106

emanuel adler

T. V. Paul captures the complexity of the international system in his introductory chapter to this volume. 12. Neil E. Harrison, Complexity in World Politics: Concepts and Methods of a New Paradigm (Albany: State University of New York Press, 2006); Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton: Princeton University Press, 1997). 13. Jeffrey C. Alexander, “Performance and Power” (working paper, Center for Cultural Sociology, Yale University, New Haven, 2005), http:// research.yale.edu /ccs / documents / public / alex_perfrmPower.pdf; Jeffrey C. Alexander, Bernhard Giesen, and Jason L. Mast, eds., Social Performance: Symbolic Action, Cultural Pragmatics, and Ritual (Cambridge: Cambridge University Press, 2006); Mark Juergensmeyer, “Understanding the New Terrorism,” Current History 99 (April 2000): 158–63. 14. Zeev Maoz, “Evaluating Israel’s Strategy of Low-Intensity Warfare, 1949–2006,” Security Studies 16 ( July-September 2007): 319–49. 15. Thomas C. Schelling, The Strategy of Conflict (Cambridge, MA: Harvard University Press, 1960). 16. Glenn H. Snyder, Deterrence and Defense: Toward a Theory of National Security (Princeton: Princeton University Press, 1961), 3. 17. Schelling, The Strategy of Conflict, 7. 18. Snyder, Deterrence and Defense, 14–16. 19. Lawrence Freedman, Deterrence (Cambridge: Polity Press, 2004), 6–25. 20. Bernard Brodie, “Implications for Military Policy,” in The Absolute Weapon: Atomic Power and World Order, edited by Bernard Brodie (New York: Harcourt, Brace, 1946), 76. 21. Jennifer Mitzen, personal e-mail communication, October 29, 2006. 22. Robert Jervis, “Rational Deterrence: Theory and Evidence,” World Politics 41 (1989): 183–207; Richard Ned Lebow and Janice Gross Stein, “Rational Deterrence Theory: I Think, Therefore I Deter,” World Politics 41 (1989): 208–24. 23. Graham T. Allison, Essence of Decision: Explaining the Cuban Missile Crisis (Boston: Little, Brown, 1971). 24. Bruce M. Russett, The Prisoners of Insecurity: Nuclear Deterrence, the Arms Race, and Arms Control (New York: W. H. Freeman, 1983). 25. Richard Ned Lebow and Janice Gross Stein, “Deterrence: The Elusive Dependent Variable,” World Politics 42 (1990): 336–69. 26. Keith Krause and Michael C. Williams, eds., Critical Security Studies: Concepts and Cases (Minneapolis: University of Minnesota Press, 1997). 27. McGeorge Bundy, Danger and Survival: Choices about the Bomb in the First Fifty Years, 1st ed. (New York: Random House, 1988). 28. Freedman, Deterrence, 41; Patrick M. Morgan, Deterrence Now (Cambridge: Cambridge University Press, 2003), 80–115. 29. Freedman, Deterrence, 4–5. 30. Morgan, this volume. 31. Amir Lupovici, “Let’s Talk! Deterrence Norm as a Case Study for the Synthesis of Rational and Normative Approaches” (paper presented at the Annual Meeting of the International Studies Association, San Diego, California, March 22, 2006). 32. Trager and Zagorcheva, “Deterring Terrorism.” 33. Keith Krause and Michael C. Williams, “From Strategy to Security: Foundations of Critical Security Studies,” in Krause and Williams eds., Critical Security Studies, 33–69. On the neoconservative critique, see Stefan A. Halper and Jonathan Clarke, America Alone: The Neo-Conservatives and the Global Order (Cambridge: Cambridge University Press, 2004). 34. Richard K. Betts, “How Superpowers Become Impotent,” Los Angeles Times, August 14, 2006, B11. 35. Pape, Dying to Win: The Strategic Logic of Suicide Terrorism.

complex deterrence

107

36. Cf. critics cited in Trager and Zagorcheva, “Deterring Terrorism,” 87. 37. Alexander Wendt, “Constructing International Politics,” International Security 20, no. 1 (1995): 73. 38. Alexander Wendt, Social Theory of International Politics (Cambridge: Cambridge University Press, 1999). 39. Pepper D. Culpepper, “The Politics of Common Knowledge: Ideas and Institutional Change in Wage Bargaining,” International Organization 62 (2008): 1–33; Schelling, Strategy of Conflict, 54–58. 40. Wendt, Social Theory, 160. 41. Culpepper, “The Politics of Common Knowledge,” 5. 42. “Degrees of separation” refers to the extent of connectedness within, thus also the differentiation between, networks. See Stanley Milgram, “Small World Problem,” Psychology Today (May 1967): 60–67. 43. Michel Foucault, The Order of Things: An Archaeology of the Human Sciences (New York: Pantheon Books, 1971). 44. Michael N. Barnett and Raymond Duvall, eds., Power in Global Governance (Cambridge: Cambridge University Press, 2005), 39. 45. Stefano Guzzini, “The Concept of Power: A Constructivist Analysis,” Millennium: Journal of International Studies 33 (2005): 495–521. 46. Ian Hurd, “Legitimacy and Authority in International Politics,” International Organization 53 (1999): 379–408; Max Weber, From Max Weber: Essays in Sociology, translated by Hans Heinrich Gerth and C. Wright Mills (New York: Oxford University Press, 1958). 47. Hedley Bull, The Anarchical Society (London: Macmillan, 1977). 48. Weber, From Max Weber. 49. Freedman, Deterrence, 7. 50. Emanuel Adler, “The Emergence of Cooperation: National Epistemic Communities and the International Evolution of the Idea of Nuclear Arms Control” International Organization 46 (1992): 101–45. 51. I thank an anonymous reviewer for helping me clarify this point. 52. Adler, “Emergence of Cooperation.” 53. Joseph S. Nye, Soft Power: The Means to Success in World Politics, 1st ed. (New York: Public Affairs, 2004). 54. Shmuel Eisenstadt, “Multiple Modernities,” Dædalus 129 (Winter 2000): 1–29. 55. Arreguín-Toft, this volume. 56. Stein, this volume. 57. Daniel Kahneman, “A Psychological Perspective” (Rationality Lecture Series, Rationality: A Critical Perspective, Munk Centre for International Studies, University of Toronto, February 17, 2007). 58. Richard Ned Lebow, “Thucydides and Deterrence,” Security Studies 16, no. 2 (2007): 163–88. 59. For a different explanation of this puzzle, based on the notion of war aims, see Patricia L. Sullivan, “War Aims and War Outcomes: Why Powerful States Lose Limited Wars,” Journal of Conflict Resolution 51 (2007): 496–524. 60. Arreguín-Toft, this volume. 61. Jeffrey C. Alexander, “From the Depths of Despair: Performance, Counterperformance, and ‘September 11’,” Sociological Theory 22, (2004): 88–105; Alexander, “Performance and Power”; Alexander, Giesen, and Mast, Social Performance. 62. This, however, does not mean that deterrence always worked. See, for example, Richard Ned Lebow and Janice Gross Stein, We All Lost the Cold War (Princeton: Princeton University Press, 1994).

108

emanuel adler

63. Freedman, Deterrence, 122. 64. Nathaniel Fick, “Fight Less, Win More,” Washington Post, August 12, 2007, B1. 65. Zeev Schiff, “Analysis: For Israel, the Conflict in Lebanon Is a Must-Win Situation,” Ha’aretz, July 27, 2006. Zeev Maoz has recently shown that Israeli low-intensity warfare strategies responded to deterrence and compellence considerations. See Maoz, “Evaluating Israel’s Strategy”; Oded Löwenheim and Gadi Heimann, “Revenge in International Politics,” Security Studies 17 (2008): 685–724. 66. European Union, “European Security Strategy: A Secure Europe in a Better World,” 2003, http://www.consilium.europa.eu / uedocs /cmsUpload / 78367.pdf. 67. Adler, “Damned If You Do, Damned If You Don’t.” 68. Edward N. Luttwak, “Three Reasons Why Not to Bomb Iran—Yet,” Commentary 121 (May 2006): 21–28. 69. Kenneth Pollack and Ray Takeyh, “Taking on Tehran,” Foreign Affairs 84 (March /April 2005): 20–35. 70. Note that in Schelling’s celebrated conceptualization, the goal of acting “irrationally” was to enhance deterrence, not social power. Cf. Schelling, Strategy of Conflict, 16–18. 71. I thank Jennifer Mitzen and Patricia Greve for helping me see this point more clearly. 72. Adler, “Damned If You Do, Damned If You Don’t.” 73. Ibid. 74. Adler, “Emergence of Cooperation.” 75. Barry O’Neill, Honor, Symbols, and War (Ann Arbor: University of Michigan Press, 1999). 76. I thank Michael Barnett for helping me clarify this point. 77. I thank Neta Crawford for helping me clarify this point.

chapter five R

Deterring Nuclear Terrorists s. paul kapur

a terro ri st attack using a nuclear weapon would be extremely destructive, potentially killing hundreds of thousands of people and destroying miles of infrastructure.1 The resulting political, economic, and military effects would dwarf those of any past terror attacks. Many analysts believe that such an event is likely to occur. Graham Allison, for example, maintains “a betting person would have to” conclude that nuclear terrorism is “inevitable.”2 According to Anna M. Pluta and Peter D. Zimmerman, the threat of nuclear terrorism is so serious that “we should plan for its reality.”3 And Robert L. Gallucci argues that “more likely than not” terrorists “will detonate a nuclear weapon in a U.S. city within the next five to ten years.”4 Other scholars disagree with these alarming assessments. For example, Robin Frost argues that “the risk of nuclear terrorism . . . is overstated” and that “popular wisdom on the topic is significantly flawed.”5 And John Parachini maintains that “the historical record cautions against axiomatically suggesting” that terrorists “will inevitably successfully use [chemical, biological, radiological, or nuclear] weapons in a catastrophic attack.”6 Although some scholars may exaggerate its likelihood, nuclear terrorism is nonetheless a serious danger.7 Because a nuclear attack would be so destructive, the expected cost (damage × probability of occurrence) of nuclear terrorism is high, even if it is not likely to occur. States must therefore devise strategies to protect themselves against terrorists who somehow manage to arms themselves with nuclear weapons. Ideally, states would be able to deter nuclear terrorism. Deterrence would prevent terrorists from even trying to use nuclear weapons, thereby enabling 109

110

s. paul kapur

states to avoid the difficult and uncertain task of thwarting attacks in progress. This chapter explores the question of whether states can devise such deterrent strategies. It suggests that two broad opportunities for deterrence exist: one on nuclear terrorism’s supply side and one on its demand side. On the supply side, states can pursue deterrence by denial. Deterrence by denial would make the acquisition and use of nuclear weapons so difficult that terrorists would not even attempt it. On the demand side, states can pursue deterrence by punishment. Deterrence by punishment would make nuclear attacks so costly for terrorists that they would decide not to acquire and use nuclear weapons, even if they were physically capable of doing so.8 The chapter first explains the supply- and demand-side requirements for terrorists to use nuclear weapons. It then discusses possible target-state strategies for deterring nuclear terrorists. Next the chapter offers a framework that identifies four types of terrorist groups and ranks their susceptibility to these deterrent strategies. The chapter then illustrates this argument by briefly contrasting the potential effectiveness of deterrent strategies against two terror groups: al-Qaeda and the now largely defunct Aum Shinrikyo. The chapter concludes with a discussion of policy implications.

requirements for terrorists to use nuclear weapons Before a terrorist group can attempt to use nuclear weapons, it must meet two requirements. First, terrorists must be able to acquire and use a nuclear device. This is nuclear terrorism’s supply-side requirement. Much recent analysis of nuclear terrorism focuses on supply-side factors and emphasizes the need to keep nuclear material out of terrorists’ reach.9 Although nuclear terrorism’s supply-side requirement is important, it is not the only prerequisite for terrorists to use nuclear weapons. Potential nuclear terrorists must meet a second requirement: They must want to acquire and use a nuclear device. Many analysts assume that terrorists automatically meet this demand-side requirement, based on terrorists’ religious motivations, past statements, or previous attempts to acquire nuclear materials.10 Yet nuclear terrorism’s demand side may be more complicated than these analyses would suggest. The fact that some terror groups seek to use nuclear weapons does not mean that all terrorists wish to do so. And terrorists who under some circumstances want to use nuclear weapons might in other scenarios be persuaded to forgo them. Thus, a careful analysis of both demand- and

deterring nuclear terrorists

111

supply-side requirements is essential to the formulation of strategies to deter nuclear terrorism. Supply-Side Requirements To launch a nuclear attack, terrorists must acquire a nuclear device. How might they go about doing so? Terrorists could either secure a nuclear device intact, or they could construct one. Terrorists could procure an intact weapon in two different ways. First, a nuclear state could voluntarily transfer a weapon to terrorists for use against a designated target. This could enable the state to inflict massive damage on an enemy while maintaining deniability and potentially avoiding retaliation. Despite the possible benefits of such an approach, its occurrence appears unlikely. In this “transfer” scenario, the nuclear state would lose control of the weapon in question, forcing the state to place enormous trust in the terrorists’ loyalty and judgment. It is doubtful that a nuclear state’s leaders would be willing to trust terrorists to this degree.11 Terrorists also could acquire an intact weapon by stealing it from a nuclear state. This would be an extremely difficult feat, however, even for sophisticated terrorist groups. Nuclear weapons are protected by the most robust security regimes that nation-states can devise: personnel reliability programs; extensive physical barriers, including location in heavily guarded, often isolated military bases; electronic systems to prevent unauthorized weapons use; and storage of fissile cores separate from other weapons components. Even in the event of insider assistance or major political unrest within a nuclear state, terrorist theft of an intact nuclear device would remain difficult and unlikely.12 Given the challenges of acquiring a weapon intact, most analysts believe that terrorists are more likely to attempt to construct their own nuclear device.13 In order to do so, terrorists would need fissile material, either highly enriched uranium (HEU) or plutonium. Enriching uranium or reprocessing plutonium is sufficiently complex to require national-level capabilities and would probably prove impossible for nonstate actors. Terrorists would therefore most likely attempt to acquire fissile material that a state has already produced.14 Since it is easier to construct a nuclear device using HEU than using plutonium, terrorists would probably opt for HEU. They would need approximately fifty to sixty kilograms of very highly enriched uranium to make a relatively simple nuclear device similar to the one that the United States dropped on Hiroshima.15

112

s. paul kapur

Procuring fissile material would not be easy, but the challenge would be much less formidable than that of acquiring an intact weapon. This is the case for three main reasons. First, a great deal of fissile material—approximately 450 metric tons of plutonium and 1,700 tons of HEU—exists throughout the world. Second, because fissile materials are stored in bulk, they cannot be as easily inventoried as nuclear weapons and are susceptible to small-scale thefts over time. Finally, the measures in place to ensure the security of fissile material are far less robust than those currently protecting states’ nuclear-weapons arsenals.16 The former Soviet Union (FSU) offers a troubling example of fissile material security problems. Security systems in the FSU were originally designed during the Soviet era, when national borders were sealed and scientists and other nuclear workers were both well compensated and closely monitored by state security services. The current situation in the FSU, with porous borders and underpaid and poorly supervised personnel, is very different, and the existing security regime is ill equipped to handle it.17 As a result, two major sets of problems have emerged. The first is lax external security at nuclear material storage sites. Credible reports describe an environment in which guards lack ammunition, security doors are left open, fencing is damaged or missing, and trespassing repeatedly occurs. In addition to these external security woes, the FSU also suffers from a serious insider security problem. Low pay and an uncertain professional future increase the incentives for nuclear facility personnel to steal fissile materials or to cooperate with thieves who seek to do so. A number of recent thefts and attempted thefts of nuclear materials have occurred with insider cooperation.18 Although nuclear security problems are particularly acute in the FSU, they are not limited to those countries. For example, the A. Q. Khan proliferation network demonstrated serious security deficiencies in Pakistan. Khan, the director of Pakistan’s Khan Research Laboratories, apparently sold nuclear-related materials and technologies, including blueprints and components for gas centrifuges to enrich uranium, to several countries, including North Korea, Iran, and Libya. Khan and his associates may also have offered similar deals to Iraq and Syria.19 The Pakistani government claims to have had no knowledge of Khan’s activities.20 Many observers find such claims to be dubious.21 The Khan ring’s implications for nuclear security in Pakistan are deeply problematic. If the Pakistani government knew of and colluded in Khan’s activities, it was guilty of a serious breach of nuclear security. And if the Pakistani government was unaware of Khan’s actions, it exhibited colossal negligence. Further, the international community does

deterring nuclear terrorists

113

not know for certain the scope of Khan’s dealings. His activities could have been more extensive than are currently being reported.22 North Korea offers still another example of the potential insecurity of nuclear materials. The North Koreans are believed to have traded plans for their Nodong ballistic missiles for Pakistani uranium enrichment technology.23 Although there is no evidence that North Korea has sold or traded actual fissile material, it could do so at some point in the future. This possibility is especially worrisome given North Korea’s perilous economic condition. A cash-strapped DPRK could eventually decide to sell fissile materials to the highest bidder, which could include terrorists.24 Thus, security problems create a significant danger of unauthorized fissile material transfers, which in turn could enable terrorists to acquire HEU and construct a nuclear device. What level of resources would terrorists need to accomplish these tasks? Peter D. Zimmerman and Jeffrey G. Lewis offer the most comprehensive available cost analysis for such a project. Zimmerman and Lewis argue that a terror group could develop a gun-type nuclear device similar to the Hiroshima bomb for less than $5.5 million. According to their calculations, terrorists would purchase black market HEU for roughly $4 million and spend $1.4 million on construction and transportation of the weapon.25 Given this low price, nuclear attacks would, in Zimmerman and Lewis’ words, be a “bargain” for terrorists.26 Although a terror group wishing to commit mass murder could well find nuclear weapons attractive, they might not be quite the bargain that Zimmerman and Lewis suggest. As Zimmerman and Lewis note, there exists only one known instance of criminals obtaining a significant quantity of HEU. In that 1994 case, a team of Czech and Slovak smugglers attempted to sell HEU to undercover police in Prague for $1,600 per gram. At that price, the roughly sixty kilograms of HEU needed to fashion a Hiroshima-type nuclear device would have cost approximately $96 million. The smugglers were acting as middlemen for the uranium’s Russian suppliers, who were willing to sell the material for only $800 per gram.27 However, even if the terrorists had been able to acquire the uranium at the suppliers’ price, sixty kilograms would have cost $48 million, far above Zimmerman and Lewis’ cost estimate. Although Zimmerman and Lewis base their analysis on the requirements of a Hiroshima-type weapon,28 terrorists could seek to construct a smaller nuclear device; a detonation much less powerful than the Hiroshima bomb’s twelve to fifteen kilotons29 would still inflict considerable physical and psychological damage on its target.30 Even at the Russian uranium suppliers’

114

s. paul kapur

price of $800 per gram, however, a device using only half the HEU of the Hiroshima bomb would cost approximately $24 million. Such a weapon would significantly exceed Zimmerman and Lewis’s cost estimate, despite being approximately half the size of the device that they envision.31 Terrorists seeking to develop even a small nuclear weapon, then, will probably need to be able to spend tens of millions of dollars on fissile material alone. Of course, it is possible that terrorists could obtain HEU more cheaply; one cannot make definitive claims regarding black market fissile material prices based on a single incident. In the absence of evidence that the Czech incident is an outlier, however, one should expect other cases roughly to conform to it.32 Nuclear terrorism’s supply-side requirements suggest several points: First, lax security standards currently render nuclear materials vulnerable to unauthorized transfer. Improved fissile material safeguards, however, could make nuclear materials far more difficult to procure. Second, even in today’s security environment, only a small number of very wealthy terror groups stand a significant chance of acquiring nuclear weapons; groups that cannot afford a price tag of tens of millions of dollars are unlikely to be nuclear contenders.33 Zimmerman and Lewis believe that al-Qaeda is the only existing terror group that poses a real nuclear threat, assuming a price of just $5 million for the “fuel” needed to make a small nuclear weapon.34 Demand-Side Requirement Many analysts and policy makers assume that terrorist groups will wish to use nuclear weapons.35 It is not clear, however, that this is the case. Although analysts often characterize terrorism as an irrational activity,36 evidence indicates that terrorists in fact behave rationally, adopting strategies designed to achieve particular ends.37 Whether or not terrorists will want to use nuclear weapons therefore will depend on whether they believe that doing so is likely to further their goals. Scholars have identified a number of specific scenarios in which nuclear use could undermine terrorists’ objectives. For example, the massive destruction resulting from a nuclear detonation could subject the terrorists to harsh reprisals, alienate their supporters, and lead to dissension within their ranks.38 However, scholars have not specified the organizational characteristics that are likely to make terror groups particularly susceptible to such outcomes.39 These and other demand-side disincentives to nuclear use are associated

deterring nuclear terrorists

115

with “positive” organizational goals. By positive goals I mean that the terrorist group seeks to advance the welfare of an existing population or territory. Positive goals can take a number of forms, ranging from the promotion of a religious or economic system, to the protection of an ethnic group, to the preservation of a rainforest. Regardless of their differences, all groups with positive goals share a common element: they do not view violence either as an end in itself or as a means of wholly eliminating the current physical and social order. Rather, through their violent methods, the terrorists actually hope to promote the welfare of existing people and places. This means that even if the terrorists are willing to sacrifice their own lives, they nonetheless care about the well-being of others. A terrorist group with negative goals, by contrast, does not wish to promote the welfare of any existing population or territory. Instead, such a group seeks maximal violence and destruction, either as an end in itself or as a means of wholly eliminating the existing physical and social order. Terrorists with positive goals could find nuclear weapons to be unattractive. This is the case because by launching nuclear attacks, the terrorists could invite target-state retaliation against the things that are important to them. Even terrorists willing to die for their cause might find the punishment of the populations and territories about which they care to be unacceptable.

s u p p ly - a n d d e m a n d - s i d e d e t e r r e n c e Opportunities for deterrence exist on both the supply and the demand sides of the nuclear terrorism equation. On the supply side, states could make the use of nuclear weapons so difficult that terrorists would decide not even to attempt it. And on the demand side, states could exploit terrorists’ organizational goals to make nuclear use prohibitively costly. This could lead terrorists to avoid using nuclear weapons even if they were physically able to do so. Supply-Side Deterrence A supply-side deterrent strategy would seek to make nuclear use prohibitively difficult, thereby achieving deterrence by denial. States would do so by preventing terrorists from acquiring the materials that they need to develop a nuclear capacity. This would require the implementation of a rigorous safeguards regime to protect fissile material stockpiles. Some aspects of such a regime would include enhanced screening and monitoring of personnel

116

s. paul kapur

at nuclear facilities and fissile material storage sites; repair of crumbling security infrastructure such as fences, gates, walls, and doors; the installation of electronic surveillance devices at storage locations; implementation of statistical and computer-based inventory techniques; improved material control procedures; and independent audits of nuclear security measures.40 These policies would be challenging to implement but not excessively so. The international community knows the location of fissile material stockpiles around the world. Thus, cutting off terrorist access to fissile materials does not require highly accurate intelligence or sophisticated technologies. Rather it requires the political will to devise and implement rigorous physical protection regimes for fissile material stockpiles. The creation of a rigorous safeguards regime could prove to be a highly effective policy. Indeed, Graham Allison argues that doing so could render fissile material as secure as the gold stored in Fort Knox, making the unauthorized transfer of fissile material virtually impossible.41 Even if this level of perfection proved unattainable, however, improved safeguards could sharply reduce the potential supply of fissile material and significantly increase its price. This could render fissile material extremely difficult for terrorists to acquire, thereby deterring them from even attempting to pursue a nuclear capability. In principle, states also could achieve deterrence by denial by developing the ability to thwart a nuclear attack in progress. Such a tactical defense capability could convince terrorists that any nuclear attack was likely to fail and thus was not worth the cost of the attempt. A tactical defense would require either the apprehension of terrorists as they smuggled nuclear materials or the components of a preexisting nuclear weapon into the target state or the discovery of a nuclear device inside the target state before it could be used. Success would be difficult to achieve because it would require the effective policing of long borders and coastlines, monitoring of the enormous volume of goods coming into the country through ports of entry, and highly accurate real-time intelligence.42 Terrorists also could attack the target state using a nuclear-armed ship or aircraft. In order to foil such an attack, the state would have to intercept the weapon before it reached its target. Again, this would be a difficult task, requiring the effective protection of long borders and coastlines. Given these hurdles, tactical defense is probably of relatively limited deterrent value; it will be considerably easier for states to make fissile material prohibitively expensive by protecting a finite number of nuclear facilities than to make nuclear attack prohibitively difficult by securing myriad ports and lengthy borders.

deterring nuclear terrorists

117

Demand-Side Deterrence On the demand side, terrorists could be deterred from nuclear use by the possibility of punishment. The potential victim of a nuclear attack could threaten to retaliate against targets that the terrorists value, making nuclear use prohibitively costly. Conventional wisdom holds that punishment strategies are unlikely to deter terrorism. In this view, punishment is hobbled by two factors. First, terrorists are difficult for target states to locate and attack. Second, even lethal target-state retaliation might not convince suicidal terrorists that the costs of violence outweigh its benefits.43 Despite these challenges, target states may be able to deter potential nuclear terrorists through punishment strategies. Whether they will be able to do so will depend on the nature of the terrorists’ goals. If the terrorists have negative goals, punishment is likely to be ineffective, and denial strategies will probably be the target states’ sole option. This is the case because terrorists who seek to maximize destruction do not value the welfare of particular populations or territories. They therefore will be impervious to threats to harm such targets. By contrast, terrorists who have positive goals care about particular populations and territories, even if they are willing to sacrifice their own lives. In these cases, target states are no longer limited to denial-based strategies; states also can employ punishment-based strategies against the terrorists, threatening to harm the people and places that the terrorists value in the event of a nuclear attack. Such a third-party punishment approach would avoid the problems typically associated with deterring elusive, suicidal terrorists. Terrorists who personally accept death may nonetheless fear the demise of their compatriots and coreligionists or damage to their homelands and holy sites. These are the people and places to which the terrorist organization is dedicated and for which the terrorists are willing to sacrifice their lives. And while the terrorists themselves may be elusive, the things they care about should be easy to locate. Thirdparty punishment thus adds a potentially potent weapon to states’ defensive arsenals.44 Target states also can employ punishment strategies to deter terrorists’ sponsors, regardless of the terrorists’ goals. Third parties, especially nation-states, may provide terrorists with the money, training, and safe haven that they need to acquire a nuclear capacity. Despite sympathy for the terrorists, however, the leaders of such states may be unwilling to face death or ouster to support them. In such scenarios, target countries can threaten supporting states with punishment in the event that one of their client

118

s. paul kapur

groups launches a nuclear attack. This could deter states from providing crucial assistance to prospective nuclear terrorists.45

categorizing potential nuclear terrorists Scholars have ranked various types of terror groups’ propensities for nuclear use elsewhere in the literature, but they have not focused specifically on nuclear terrorists’ vulnerability to deterrent strategies. And most analyses turn on either terrorist goals or resources rather than on the interplay between the two sets of factors.46 The category of terror group least susceptible to deterrence will possess two key characteristics: On the supply side, it will have a high level of independent financial resources. In other words, the group is able to expend at least several tens of millions of dollars on the purchase of fissile material and it is not financially dependent upon a nation-state sponsor. On the demand side, the group will have primarily negative goals. This type of group is especially immune to deterrence for two reasons. First, the group probably possesses the minimum financial resources needed to acquire a nuclear weapon and may continue to do so even if fissile materials become somewhat more scarce and expensive. Because these resources are independent, target states cannot reduce them by threatening the group’s supporters. Second, the group is impervious to target-state punishment, since it does not value the welfare of any population or territory. This group’s susceptibility to deterrence is low. The category of terror group with the second lowest susceptibility to deterrence will possess either of two sets of characteristics: high independent resources and positive organizational goals or low independent resources and negative organizational goals. First consider a terror group with high resources and positive goals. High independent resources would facilitate the group’s acquisition of nuclear weapons, making supply-side deterrence difficult. The group’s positive goals, however, make it potentially vulnerable to punishment and increase the likelihood that the group could be deterred on the demand side. Thus, for this group, supply- and demand-side factors pull in opposite directions, yielding what I rank as a medium susceptibility to deterrence. Next consider a terror group with low resources and negative goals. Low independent resources would impede the group’s ability to acquire nuclear weapons, making supply-side deterrence more likely. Negative goals, however, lower the group’s vulnerability to punishment, reducing the likeli-

deterring nuclear terrorists

119

hood that the group would be susceptible to demand-side deterrence. Thus, supply- and demand-side factors again pull in opposite directions, yielding a medium level of propensity for nuclear terrorism. The category of terror group most vulnerable to deterrence will possess the following characteristics: low independent resources and positive organizational goals. Low resources would impede the group’s efforts to acquire and assemble nuclear materials, raising the likelihood of supply-side deterrence. And positive goals increase the group’s vulnerability to punishment, thereby making demand-side deterrence more likely. I rank such a group’s susceptibility to deterrence as high. This framework shows that states possess options for deterring most types of potential nuclear terrorists. Some groups may be deterred from nuclear use on the supply side, others on the demand side. Still other terrorist groups can be deterred on both the supply and demand sides of the nuclear terrorism equation. Only terror groups that combine high independent resources with negative organizational goals are likely to be undeterrable.

contrasting al-qaeda and aum shinrikyo The deterrence framework can be illustrated by exploring al-Qaeda’s and Aum Shinrikyo’s susceptibility to deterrence. Given their considerable wealth, neither group is likely to be vulnerable to supply-side, denial-based deterrent strategies. Al-Qaeda’s positive aims, however, render the group potentially susceptible to demand-side, punishment-based deterrence. By contrast, Aum Shinrikyo’s negative goals meant that it probably could not have been deterred from nuclear terrorism by the threat of punishment. Given its significant financial resources, Aum Shinrikyo was thus largely immune to deterrent threats. Although it is now mostly defunct, Aum Shinrikyo probably posed a greater nuclear danger when it was active than al-Qaeda does today. Al-Qaeda Al-Qaeda seeks to overthrow the “apostate” pro-Western regimes currently ruling the Muslim world and to replace them with an international Islamic caliphate.47 In order to do so, the group hopes to end the United States’ global hegemony, which it believes will result in the collapse of the pro-Western regimes.48 Al-Qaeda has carried out a number of deadly, high-profile operations, including the simultaneous 1998 bombings of U.S. embassies in

120

s. paul kapur

Kenya and Tanzania, the attack on the USS Cole in 2000, and the attacks of September 11, 2001. Al-Qaeda was based in Afghanistan until coalition forces ousted the Taliban following the September 11 attacks. The organization’s several thousand members are now splintered into small groups in South and Southeast Asia and the Middle East. The group secures funds from an international network of front businesses, charitable organizations, and individual supporters.49 On the supply side, al-Qaeda’s financial resources probably exceed the minimum necessary to acquire fissile material and develop a nuclear weapon. The group has a large number of members and affiliates, a sophisticated organizational network, and an independent financial base that could reach into the hundreds of millions or possibly billions of dollars.50 Thus even if improved safeguards made fissile material considerably scarcer and more expensive, al-Qaeda would probably not be deterred from trying to procure it.51 On the demand side, al-Qaeda leaders have announced their desire to acquire nuclear weapons. According to their statements, al-Qaeda seeks nuclear weapons for two reasons: to defend Muslim lands and communities against “Jews and Crusaders” and to inflict maximum harm on Americans.52 Given these factors, al-Qaeda is often cited as posing the world’s leading nuclear terrorist threat.53 Despite al-Qaeda’s apparent nuclear propensities, the group’s positive organizational goal could create disincentives for nuclear use. As Michael Doran explains, “the worldview of al-Qaeda differs from that of a doomsday cult . . . According to al-Qaeda, violence will not spark the apocalypse but instead will avert disaster and usher in a new dawn . . . Al-Qaeda’s goal is to take control of the state and to pull humanity back from the abyss by upholding Islamic values throughout society.”54 These beliefs and desires force the group to care about conditions in the Middle East. As a result, al-Qaeda could be susceptible to deterrence by punishment. If the group launched nuclear attacks to harm the United States or to force the withdrawal of outside powers from the Middle East, the victims could invade and occupy important sites in the region.55 Such large-scale foreign intervention could impede the creation of an Islamic state and thus directly undercut al-Qaeda’s central purpose. Nuclear use would therefore entail significant risks for al-Qaeda. Of course, such dangers cannot guarantee that al-Qaeda will never use a nuclear weapon. The group’s leaders could become sufficiently motivated to damage or coerce their enemies that they deem the potential benefits of launching nuclear attacks to outweigh even substantial retaliatory risks.

deterring nuclear terrorists

121

Also, as noted above, al-Qaeda is not a unitary actor. Although this discussion has focused on the likely behavior of the group’s core leadership, other al-Qaeda leaders could calculate the costs and benefits of nuclear use differently. Nonetheless, the potential costs of launching nuclear attacks could discourage al-Qaeda from engaging in nuclear terrorism, even if it were physically able to acquire a nuclear device. Given the danger of target-state retaliation, al-Qaeda could decide to limit itself to nonnuclear attacks, which would still inflict significant damage on its enemies without subjecting the group to the possible costs of nuclear use. Aum Shinrikyo In contrast to al-Qaeda, Aum Shinrikyo combined high resources with negative organizational goals. Shoko Asahara, a blind yoga instructor, founded the group in 1987. Its idiosyncratic mixture of Buddhist, Shinto, Hindu, Biblical, and New Age teachings stressed the decadence and inevitable destruction of the existing world and predicted the subsequent emergence of a new global order. The Aum made a brief detour into mainstream politics in 1990, standing 25 candidates for election to the Japanese Diet. After their resounding defeat, the group increasingly stressed the imminence of Armageddon, which it believed would be triggered by a nuclear war between the United States and Japan. Aum Shinrikyo was responsible for a series of chemical accidents and failed biological attacks in Japan, including a 1995 sarin gas attack on the Tokyo subway system. At the time of the Tokyo attack, the Aum possessed independent assets in the range of $300 million to $1 billion and claimed a worldwide membership of 40,000.56 Aum Shinrikyo is largely defunct today. In the wake of the Aum’s Tokyo subway attack, the Japanese government seized the group’s assets and incarcerated Asahara. Asahara was sentenced to death in 2004. Current estimates of the group’s size range from fewer than 1,000 to approximately 1,500 members. In 2000, under the leadership of Fumihiro Joyu, Aum Shinrikyo adopted the name Aleph and claimed to have renounced its violent goals.57 Despite its demise, Aum Shinrikyo’s past behavior helps to illustrate the nature of the nuclear dangers that high-resource, negatively oriented terror groups can potentially pose. On the supply side, Aum Shinrikyo resembled al-Qaeda. The group was extremely wealthy, and it probably would have continued attempting to procure nuclear materials even if they became more scarce and expensive. The group therefore would likely not have been deterred from the pursuit of

122

s. paul kapur

nuclear weapons by supply-side denial measures. Aum Shinrikyo appeared to resemble al-Qaeda on the demand side as well. Although its leaders did not publicly call for the acquisition of nuclear weapons, the group clearly wished to develop a nuclear capacity. Indeed, Aum Shinrikyo made attempts to develop nuclear weapons that, while unsuccessful, were more extensive than any known al-Qaeda nuclear acquisition efforts. For example, Aum Shinrikyo tried to mine uranium on land that it had procured in Australia. And the group went to great lengths to acquire a nuclear weapon in Russia, offering large bribes to an extensive network of contacts that it developed in the Russian scientific, political, and security establishments.58 Despite these similarities, a key demand-side factor differentiated Aum Shinrikyo from al-Qaeda: Aum Shinrikyo had negative goals. Specifically, the group believed that the end of the world was imminent, and it sought to trigger Armageddon through acts of mass violence.59 The Aum thus did not value the well-being of any particular population or territory. As a result, unlike a nuclear al-Qaeda, a nuclear-armed Aum Shinrikyo would not have been susceptible to deterrence by punishment. Target states would have been unable to inflict costs on the group by retaliating against any territory or population.60 In fact, the Aum probably would have viewed a violent target-state reaction as desirable since it would have furthered the group’s goal of maximizing destruction. Target states thus would have had neither supply- nor demand-side options in seeking to deter Aum Shinrikyo from engaging in nuclear terrorism. This brief comparison of al-Qaeda and Aum Shinrikyo suggests that even wealthy, violent terror groups could be deterred from using nuclear weapons. Although supply-side strategies will probably be ineffective against them, such organizations may be susceptible to demand-side punishment strategies, provided that they have positive goals. However, if potential nuclear terrorists combine high independent resources and negative organizational goals, then they are unlikely to be susceptible to any form of deterrence. While groups such as al-Qaeda do pose a real nuclear danger, wealthy, negatively oriented organizations like Aum Shinrikyo probably constitute an even greater threat of nuclear terrorism.

conclusion States can employ both supply- and demand-based strategies to deter potential nuclear terrorists. On the supply side, states can pursue deterrence by denial, seeking to make fissile material so scarce and expensive that terror-

deterring nuclear terrorists

123

ists will not even attempt to acquire it. On the demand side, states can pursue deterrence by punishment, threatening to harm targets important to nuclear terrorists and thereby making nuclear use prohibitively costly. Although deterrence will not be effective against all potential nuclear terrorists, it could discourage nuclear use by a range of groups, including wealthy and violent organizations such as al-Qaeda. What are the policy implications of these findings? On the supply side of the nuclear terrorism equation, states must implement a rigorous fissile material safeguards regime. This would require such measures as enhanced personnel screening and monitoring, better infrastructure repair, and improved electronic surveillance, inventory, and material-control techniques. On the demand side, target states must accomplish several tasks. First, states must make clear that terrorists and their supporters could face costly retaliation for launching a nuclear attack. Target states should announce that they reserve the right to take large-scale military action, including invasion and occupation, against territories of central importance to (1) any nonstate actor that attacks them with nuclear weapons and (2) any entity that provides the attackers with substantial material or financial support. This would not obligate states to retaliate following a nuclear terrorist incident. For example, if they were attacked by negatively oriented groups, states would probably be unable to identify third-party targets that the terrorists valued. In this scenario, states could decide not to launch reprisals. Or, if the evidence of links between nuclear terrorists and supporting states or organizations were murky, states could refrain from punishing possible supporters of terrorist groups. Nevertheless, by openly retaining the option of punishment, states would warn terrorists and their backers that the cost of launching nuclear attacks could be extremely high and thereby increase the likelihood of deterring them. Second, states must ensure that their threat to retaliate against nuclear terrorists is credible. This means that states must be both able and willing to carry out their deterrent threats. To be able to retaliate, states must identify the source of a nuclear device in the wake of an attack. This process of nuclear attribution would use techniques ranging from satellite imagery to mass spectrometry to determine the sophistication of a detonated nuclear device, the isotopics and efficiency of its fuel, and potentially the materials used to construct it. These factors, in turn, could enable investigators to identify the origin of the nuclear material and the actors behind the attack.61 Despite its importance, nuclear attribution remains a difficult endeavor.

124

s. paul kapur

Former Assistant Secretary of State Robert Gallucci characterizes the U.S. ability to determine the source of fissile material used in a nuclear detonation as “uncertain.”62 According to former director of the Defense Threat Reduction Agency Jay Davis, the problem of nuclear attribution “is sufficiently daunting that no confident statement can be made at present about the likelihood of success.”63 Thus it is not clear that target states would be able to identify the source of the nuclear material used in a terrorist attack. This uncertainty could reduce the credibility of a state’s retaliatory threats. States must therefore make improving their nuclear forensic capabilities a top priority, devoting resources for necessary research and development. Additionally, states should publicize their improved attribution capabilities. States must communicate to terrorists and their supporters that in the event of a nuclear attack, there is a high likelihood that the culprit will be quickly identified. Even if states were able to identify the source of a nuclear terrorist attack, a second question would remain: Would states actually be willing to punish the populations and territories that the terrorists valued, given that these targets bore no direct responsibility for the nuclear attack? Are threats to do so even remotely plausible? From the U.S. perspective, such threats could actually be more believable than those that underlay America’s cold war deterrent posture. During the cold war, the United States threatened to launch a nuclear attack on the Soviet Union in retaliation for a conventional Warsaw Pact attack on Western Europe. Now, by contrast, the United States would threaten to punish targets important to terrorists after the terrorists had detonated a nuclear weapon in the American homeland. Thus, the United States would be retaliating for a nuclear attack on the United States rather than for a conventional attack against another country. In addition, the American punishment of terrorist targets would not rely on nuclear weapons and thus would avoid the moral problems associated with a massive nuclear attack against civilian targets. Finally, American nuclear strikes on the Soviet Union would have invited a devastating Soviet nuclear response against the U.S. homeland. Terrorists, by contrast, could not inflict anything close to this level of damage on the United States in response to an American punishment campaign. Thus, a strategy of punishing third-party targets in the wake of a nuclear terrorist attack would not necessarily be implausible. Indeed, on both strategic and ethical grounds, it could be more believable and hence more credible than the U.S. deterrent posture during the cold war. Nonetheless, one might argue that in the context of its current coun-

deterring nuclear terrorists

125

terterrorism efforts, a strategy of deterrence by punishment would actually be counterproductive for the United States; attacks on third parties could further provoke the terrorists and increase sympathy for their cause. Although these are real concerns, the United States or another target country would take such action only in the wake of a nuclear attack on its homeland. Punishment would not be employed on a regular basis, in response to conventional terrorist violence. Indeed, if successful, deterrence by punishment would never resort to force at all and would simply use threats to persuade terrorists not to launch nuclear attacks. But if deterrence failed and a target state actually had to launch a retaliatory punishment campaign, it would do so in an environment in which potentially hundreds of thousands of its people had been killed and billions of dollars of its infrastructure had been destroyed. In this situation, the target state would hardly need to fear provoking the terrorists or encouraging their supporters; the terrorists would already have done their worst.64 Thus, a strategy that seeks to deter nuclear terrorism through threats of punishment need not be either implausible or counterproductive. Of course, the United States or other countries might nonetheless decide against adopting such a strategy. However, this would be a political choice on the part of target states. It would not mean that, in principle, states could not create significant disincentives for nuclear terrorism by threatening to punish the people and places that the terrorists value.

notes 1. The views expressed herein are solely the author’s and do not necessarily reflect those of the U.S. Navy or any other government agency. 2. Graham Allison, Nuclear Terrorism: The Ultimate Preventable Catastrophe (New York: Owl Books, 2004), 14–15. 3. Anna M. Pluta and Peter D. Zimmerman, “Nuclear Terrorism: A Disheartening Dissent,” Survival 48 (Summer 2006): 66. 4. Robert L. Gallucci, “Averting Nuclear Catastrophe: Contemplating Extreme Responses to U.S. Vulnerability,” Harvard International Review 26 (Winter 2005): 84. 5. Robin M. Frost, “Nuclear Terrorism after 9 / 11” (Adelphi Paper no. 378, International Institute for Strategic Studies, London, 2005), 7. 6. John Parachini, “Putting WMD Terrorism into Perspective,” Washington Quarterly 26 (Autumn 2003): 37–38. 7. The United States Code defines terrorism as “premeditated, politically motivated violence perpetrated against noncombatant targets by subnational groups or clandestine agents.” Title 22 United States Code sec. 2656f(d). I define nuclear terrorism as the detonation of a nuclear device in this manner. 8. On denial versus punishment, see Robert Jervis, The Illogic of American Nuclear Strategy (Ithaca, NY: Cornell University Press, 1989), 8–13; Robert A. Pape, Bombing to Win: Air Power and

126

s. paul kapur

Coercion in War (Ithaca, NY: Cornell University Press, 1996), 13, 19; Thomas C. Schelling, Arms and Influence (New Haven: Yale University Press, 1966), 1–34. 9. See, e.g., Graham Allison, “The Ongoing Failure of Imagination,” Bulletin of the Atomic Scientists 62 (September/October 2006); Matthew Bunn and Anthony Wier, “Securing the Bomb 2006” (Project on Managing the Atom, Harvard University, Cambridge, MA, 2006), v, 1; Pluta and Zimmerman, “Nuclear Terrorism,” 55, 66; Alexander Glaser and Frank N. Von Hippel, “Thwarting Nuclear Terrorism,” Scientific American 294 (February 2006): 56–63; Siegfried S. Hecker, “Toward a Comprehensive Safeguards System: Keeping Fissile Materials out of Terrorists’ Hands,” Annals of the American Academy of Political and Social Science 607 (September 2006): 121–32. 10. See, e.g., Allison, Nuclear Terrorism; Bunn and Weir, “Securing the Bomb 2006”; Pluta and Zimmerman, Nuclear Terrorism; Steven Simon, “The New Terrorism: Securing the Nation against a Messianic Foe,” Brookings Review 21 (Winter 2003): 18–24; Gallucci, “Averting Nuclear Catastrophe.” 11. Charles L. Glaser and Steve Fetter, “National Missile Defense and the Future of U.S. Nuclear Weapons Policy,” International Security 26 (Summer 2001): 55–56; Frost, Nuclear Terrorism After 9 / 11, 64–65; Charles D. Ferguson and William C. Potter, The Four Faces of Nuclear Terrorism (New York: Routledge, 2005), 55–57. 12. Ferguson and Potter, Four Faces of Nuclear Terrorism, 57–65, 119; Matthew Bunn, John P. Holdren, and Anthony Wier, “Securing Nuclear Weapons and Materials” (Project on Managing the Atom, Harvard University, Cambridge, MA, May 2002), 5. 13. Hecker, “Comprehensive Safeguards System,” 121. 14. Ferguson and Potter, Four Faces of Nuclear Terrorism, 119–20; Michael May, Jay Davis, and Raymond Jeanloz, “Preparing for the Worst,” Nature 443 (October 2006): 907; Hecker, “Toward a Comprehensive Safeguards System,” 121–22. 15. Bunn, Holdren, and Wier, “Securing Nuclear Weapons and Materials,” 5; Allison, Nuclear Terrorism, 95–98. 16. Ferguson and Potter, Four Faces of Nuclear Terrorism, 106–9. See also Hecker, “Toward a Comprehensive Safeguards System,” 122–25. 17. Ferguson and Potter, Four Faces of Nuclear Terrorism, 6–7. 18. Matthew Bunn and Anthony Weir, “Securing the Bomb 2005: The New Global Imperatives,” (Project on Managing the Atom, Harvard University, Cambridge, MA, and Nuclear Threat Initiative, Washington, DC, May 2005), 13–14, http://www.nti.org; Bunn, Holdren, and Wier, “Securing Nuclear Weapons and Materials,” 6–7, 10–11. See also National Intelligence Council, “Annual Report to Congress on the Safety and Security of Russian Nuclear Facilities and Military Forces,” December 2004. 19. Chaim Braun and Christopher F. Chyba, “Proliferation Rings: New Challenges to the Nuclear Nonproliferation Regime,” International Security 29 (Fall 2004): 15–18. 20. See Edward Luce, “Pakistan in Nuclear ‘Cover-up’ Row,” Financial Times, February 14, 2004, 1. 21. See William J. Broad and David E. Sanger, “As Nuclear Secrets Emerge in Khan Inquiry, More Are Suspected,” New York Times, December 26, 2004, A1; Sandro Contenta, “Pakistan’s Dirty Nuclear Secret,” Toronto Star, April 11, 2004, A6. 22. Siegfried Hecker identifies Pakistan as today’s greatest threat for fissile material theft or diversion. Hecker, “Comprehensive Safeguards System,” 130. 23. Braun and Chyba, “Proliferation Rings,” 11–13. 24. Michael O’Hanlon and Mike Mochizuki, Crisis on the Korean Peninsula: How to Deal with a Nuclear North Korea (New York: McGraw-Hill, 2003), 26–27. 25. Peter D. Zimmerman and Jeffrey G. Lewis, “The Bomb in the Backyard,” Foreign Policy no. 157 (November/ December 2006): 36. Zimmerman and Lewis do not estimate the terrorist

deterring nuclear terrorists

127

resources necessary for methods of HEU acquisition other than black market purchase, such as theft by the terror group itself. Terrorist HEU theft could vary from fairly cheap and easy to impractically expensive and difficult, depending on such factors as the target material’s location, the possessor’s internal and external security measures, and terrorists’ level of insider access. In most cases, a terror group’s surest means of acquiring fissile material would probably be to purchase it. This approach would afford terrorists access to a global fissile material market and would not require them to undertake risky and difficult theft operations potentially outside their area of expertise. I therefore judge terror groups’ ability to acquire fissile material according to their ability to purchase HEU. 26. Zimmerman and Lewis, “The Bomb in the Backyard,” 34, 39. 27. Ibid., 38. 28. Allison, Nuclear Terrorism, 95; Bunn, Holdren, and Wier, “Securing Nuclear Weapons and Materials,” 5. 29. Zimmerman and Lewis, “The Bomb in the Backyard,” 36; Nuclear Threat Initiative, “A Tutorial on Nuclear Weapons and Nuclear-Explosive Materials: Nuclear Weapons Design and Materials” (Nuclear Threat Initiative, Washington, DC, November 25, 2002), 6. 30. Pluta and Zimmerman, “Nuclear Terrorism,” 61. 31. Zimmerman and Lewis discount the importance of the Czech case, arguing that a 1993 al-Qaeda attempt to procure HEU offers a “more realistic estimate of the cost of acquiring fissile material.” In this incident, al-Qaeda spent $1.5 million to purchase a cylinder that supposedly contained HEU but in fact contained low-grade reactor fuel. The authors argue that two or three such purchases, at an average price of $4 million, would provide terrorists with sufficient material to construct a nuclear device. However, given that the material in the al-Qaeda cylinder was not HEU, it cannot reliably indicate the cost of HEU. Also, the amount of material supposedly in the cylinder is unknown. Therefore, even if the material had been genuine HEU, it is not possible to predict how many purchases would have been necessary to amass the requisite quantity to construct a nuclear device. Thus the Czech case, in which the market price per unit of genuine HEU is known, probably offers a more reliable indicator of the cost of fissile material. See Zimmerman and Lewis, “The Bomb in the Backyard,” 38; Sarah Daly, John Parachini, and William Rosenau, Aum Shinrikyo, Al Qaeda, and the Kinshasa Reactor: Implications of Three Case Studies for Combating Nuclear Terrorism, (Santa Monica: RAND, 2005), 31. 32. Note that recent evidence suggests that black market HEU is likely to cost at least as much as it did in the Czech case. See, e.g., Lawrence Scott Sheets and William J. Broad, “Smuggler’s Plot Highlights Fear over Uranium,” New York Times, 25 January 2007, A1. 33. Of course, a terrorist group that can afford to spend tens of millions of dollars will not necessarily be able to acquire nuclear materials or weapons. Such a group will simply possess the minimum financial resources necessary to do so. 34. Zimmerman and Lewis, “The Bomb in the Backyard,” 39. 35. Ashton B. Carter, “How to Counter WMD,” Foreign Affairs 83 (September/October 2004): 72–85; George W. Bush, “President Speaks to United Nations,” November 10, 2001, http://www .whitehouse.gov/ news / releases / 2001 / 11 / 20011110–3.html. 36. See, e.g., Walter Laqueur, The New Terrorism: Fanaticism and the Arms of Mass Destruction (New York: Oxford University Press, 1999), 4–5. 37. See, e.g., Robert A. Pape, “The Strategic Logic of Suicide Terrorism,” American Political Science Review 97 (August 2003): 344; Martha Crenshaw, “Relating Terrorism to Historical Contexts,” in Terrorism in Context, edited by Martha Crenshaw (University Park, PA: Penn State Press, 1995), 4. 38. Richard A. Falkenrath, Robert D. Newman, and Bradley A. Thayer, America’s Achilles’ Heel: Nuclear, Biological, and Chemical Terrorism and Covert Attack (Cambridge, MA: MIT Press, 1998),

128

s. paul kapur

181; Frost, Nuclear Terrorism, 68, 71; Richard A. Falkenrath, “Confronting Biological and Chemical Terrorism,” Survival 40 (Autumn 1998): 52. 39. For a partial exception see Ferguson and Potter, Four Faces of Nuclear Terrorism, 32–33. 40. See Allison, Nuclear Terrorism, 150; Hecker, “Comprehensive Safeguards System,” 126–127. 41. Allison, Nuclear Terrorism, 140. 42. It is worth noting that four years after suffering the September 11 attacks, launching the war on terror, and establishing a Department of Homeland Security, the United States managed to inspect only 5.6 percent of the containers entering the country through its ports. And those inspections that did occur were of questionable effectiveness; gamma-ray scanning devices often cannot see through dense materials, and radiation monitors usually cannot detect HEU, especially if it is shielded. See Hassan M. Fattah and Eric Lipton, “Gaps in Security Stretch from Model Port in Dubai to US,” New York Times, February 25, 2006, A26. See also Bunn and Wier, “Securing the Bomb 2006,” 4; Allison, Nuclear Terrorism, 119–20. 43. President of the United States, “National Security Strategy of the United States of America,” September 2002, 15. See also Allison, Nuclear Terrorism, 131–32; Gallucci, “Averting Nuclear Catastrophe,” 83–84; Simon, “New Terrorism,” 21. 44. Other scholars have also argued that punishment strategies can deter terrorists. See Robert F. Trager and Dessislava P. Zagorcheva, “Deterring Terrorism: It Can Be Done,” International Security 30 (Winter 2005–6): 87–123. 45. This is why the George W. Bush administration argues that, in principle, states supporting terrorists are just as guilty, and liable for punishment, as the terrorists themselves. See “National Security Strategy of the United States of America,” 5. See also Gallucci, “Averting Nuclear Catastrophe,” 83; James Kitfield, “Revoking al-Qaeda’s Checkbook,” National Journal 38 ( June 2006): 55; and Trager and Zagorcheva, “Deterring Terrorism,” 97. Nonstate actors could also support terrorists. However, their assistance is likely to be less extensive, and they will be harder for victims to target, than nation-states. 46. See, e.g., Frost, Nuclear Terrorism After 9 / 11, 70–73; Pluta and Zimmerman, Nuclear Terrorism, 64–66. Some scholars do note that nuclear terrorists will require both the desire to use nuclear weapons and the capability of doing so. However, they do not specify how different combinations of goals and resources are likely to result in varying terrorist susceptibility to target-state deterrence. See, e.g., Falkenrath, Newman, and Thayer, America’s Achilles’ Heel, 167–215. 47. Al-Qaeda is not a single, unitary organization. Rather, it is best described as a loose network of affiliated groups that, prior to September 11, 2001, “was headquartered in Afghanistan under the chairmanship of bin Laden.” Peter L. Bergen, Holy War, Inc.: Inside the Secret World of Osama Bin Laden (New York: Touchstone, 2002), 32. See also David J. Kulcullen, “Countering Global Insurgency,” Journal of Strategic Studies 28 (August 2005): 600–2; Hoffman, “Al-Qaeda and the War on Terrorism,” 424. I refer here to al-Qaeda’s goals as enunciated by its core leaders, including Osama bin Laden and Ayman al-Zawahiri. Philippe Errera calls this group al-Qaeda’s “first circle” of leadership. See Philippe Errera, “Three Circles of Threat,” Survival 47 (Spring 2005): 71–88. 48. Daniel Byman, “Al-Qaeda as an Adversary: Do We Understand Our Enemy?” World Politics 56 (October 2003): 146; Michael Doran, “The Pragmatic Fanaticism of al-Qaeda: An Anatomy of Extremism in Middle Eastern Politics,” Political Science Quarterly 117 (Summer 2002): 180; Osama bin Laden, “Declaration of War against the Americans Occupying the Land of the Two Holy Places,” 1996; Osama bin Laden, “Fatwa urging Jihad Against Americans,” 1998; “Encyclopedia of Afghan Jihad,” Introduction; “Declaration of Jihad Against the Country’s Tyrants, Military Series,” http://www.usdoj.gov/ ag / manualpart1_1.pdf (this document is also known as the “Al-Qaeda Training Manual”). 49. United States Department of State, “Patterns of Global Terrorism 2004”; Rohan

deterring nuclear terrorists

129

Gunaratna, Inside Al Qaeda: Global Network of Terror (New York: Berkley Books, 2003), 80–92; Kitfield, “Revoking Al-Qaeda’s Checkbook.” 50. Daly, Parachini, and Rosenau, Aum Shinrikyo, 28; Bruce Hoffman, “Al Qaeda and the War on Terrorism: An Update,” Current History (December 2004): 424; Gunaratna, Inside Al Qaeda, 81. 51. Al-Qaeda’s disaggregated status, however, could impede efforts to amass the resources necessary to purchase fissile material. This, in turn, could increase the group’s susceptibility to deterrence by denial. 52. Daly, Parachini, and Rosenau, Aum Shinrikyo, 25; Bergen, Holy War, Inc., 88; Gunaratna, Inside Al Qaeda, 65; Bruce Lawrence, ed., Messages to the World: The Statements of Osama Bin Laden (London: Verso, 2005), 72. Al-Qaeda claims the “right to kill four million Americans—two million of them children—and to . . . wound and cripple hundreds of thousands.” Suleiman Abu Ghaith, “In the Shadow of the Lances” (Special Dispatch Series No. 338, Middle East Media Research Institute, Washington, DC, June 12, 2002). Al-Qaeda offers a detailed justification of large-scale killing of civilians in “A Statement from Qaidat al-Jihad Regarding the Mandates of the Heroes and the Legality of the Operations in New York and Washington,” April 24, 2002, http://www.mepc.org / journal_vol10 / alqaeda.html. For an in-depth discussion of this document see Quintan Wiktorowicz and John Kaltner, “Killing in the Name of Islam: Al-Qaeda’s Justification for September 11,” Middle East Policy 10 (Summer 2003): 76–92. 53. See, e.g., Simon, “New Terrorism”; Gallucci, “Averting Nuclear Catastrophe.” 54. Doran, “Pragmatic Fanaticism of al Qaeda,” 181–82. 55. Alternatively, states could simply destroy high-value regional targets. However, invasion and occupation would probably be preferable. This would enable states to hold regional hostages, thereby maintaining coercive leverage over the terrorists in the event that they possessed multiple nuclear weapons. It would also enable states to avoid the moral problems associated with the destruction of historically or religiously significant sites. 56. United States Department of State, “Patterns of Global Terrorism 2004”; Memorial Institute for the Prevention of Terrorism, “Group Profile: Aum Shinrikyo /Aleph,” 2006, available at http:// www.tkb.org / Group.jsp?groupID=3956; William Rosenau, “Aum Shinrikyo’s Biological Weapons Program: Why Did It Fail?” Studies in Conflict and Terrorism 24 (2001): 291; Gavin Cameron, “Multi-track Microproliferation: Lessons from Aum Shinrikyo and Al Qaida,” Studies in Conflict and Terrorism 22 (1999): 279–80; Daly, Parachini, and Rosenau, Aum Shinrikyo, 6. 57. See Memorial Institute for the Prevention of Terrorism, “Aum Shinrikyo /Aleph.” 58. See Daly, Parachini, and Rosenau, Aum Shinrikyo, 10–20. 59. Some scholars believe that it is unclear whether Aum Shinrikyo sought to trigger Armageddon or simply wanted to defend itself through mass-casualty attacks when the inevitable cataclysm erupted. See Daniel A. Metraux, Aum Shinrikyo’s Impact on Japanese Society (Lewiston, NY: Edwin Mellen Press, 2000), 20, 52. Nevertheless, the weight of the evidence indicates that the Aum hoped to play an active role in spurring Armageddon. See Robert Jay Lifton, Destroying the World to Save It (New York: Henry Holt and Company, 1999), 6, 8; Ian Reader, A Poisonous Cocktail? Aum Shinrikyo’s Path to Violence (Copenhagen: Nordic Institute of Asian Studies, 1996), 93; Ian Reader, Religious Violence in Contemporary Japan: The Case of Aum Shinrikyo (Richmond, UK: Curzon Press, 2000), 27, 127, 160, 200, 201. 60. Aum Shinrikyo members were not suicidal. Indeed, the group went to great lengths to protect itself against the coming cataclysm. The Aum believed that these measures were so effective as to enable them to survive even a nuclear war. Thus, states probably could not have deterred them through threats of personal harm; the Aum would not have believed that such attacks could actually kill them. On the Aum’s defensive practices, see Metraux, Aum Shinrikyo’s Impact, 49; Reader, A Poisonous Cocktail?, 14, 27, 55; D. W. Brackett, Holy Terror: Armageddon in Tokyo (New York: Weatherhill, 1996), 96, 102; Lifton, Destroying the World to Save It, 44.

130

s. paul kapur

61. May, Davis, and Jeanloz, “Preparing for the Worst”; Michael A. Levi, “Deterring Nuclear Terrorism,” Issues in Science and Technology 20 (Spring 2004). 62. John Fox, “Khan Network Could Erode U.S. Deterrence, Expert Says,” Global Security Newswire, June 23, 2006. 63. Jay Davis, “The Attribution of WMD Events,” Journal of Homeland Security, April 2003. 64. Even if a punishment campaign risked further energizing the terrorists or their supporters, and thereby undermining the state’s counterterrorism efforts, the threat of punishment could still be useful as a deterrent. Deterrent threats may well harm the deterring state if they are actually carried out. Such threats can nonetheless be effective, provided that they promise to visit sufficient harm on the adversary to make the proscribed action prohibitively costly. See Schelling, Arms and Influence, 36–37.

chapter six R

Deterrence, Rogue States, and the U.S. Policy robert jervis

at the end of the cold war, deterrence seemed consigned to the dustbin of history. Hero to some for preventing World War III and eventually helping bring the Soviet Union down, it was a villain to others, some of whom saw it as more the generator than the moderator of conflict and others of whom saw it as substituting for a more vigorous policy that would have saved millions from Soviet oppression. But with the disappearance of the Soviet Union, all seemed to agree that we could turn our attentions to other problems and other policy instruments. Academic research on deterrence trailed off, public attention diminished, and policy makers talked about other things, such as the enlargement of the zone of democracy. This was not to be, however, as new problems and policies have made deterrence (and debates about its role) important once again. It is perhaps not surprising that deterrence has come back—it has always been with us in one form or another. Although many of the concepts used by academics and some of the tactics used by states are creatures of the mid-twentieth century, the basic phenomenon—or rather phenomena— are reflected in the history of the interaction of states and, more broadly, of human beings (and, in some forms, of animals).1 But seen in another way, the current concern with deterrence is odd. The United States no longer faces a peer competitor. Although Russia and the People’s Republic of China (PRC) could do horrendous damage to the United States, they would be utterly destroyed in any all-out war. More importantly, it is hard to conjure any conflicts of interests that would merit a major military clash between the United States and the PRC or Russia. Indeed, American leaders are less 133

134

robert jervis

concerned with these “near peers” than with much smaller states, often dubbed “rogues.” In the case of these small powers, the American military advantage is overwhelming on all levels of violence as well as in the economic sphere, and it would remain so even if these countries gained small nuclear arsenals. It is far from clear that the worry about rogue states shown by a series of U.S. governments and the George W. Bush administration in particular is merited. But it is clear that this worry has reinvigorated the notion of deterrence and disputes about it. This chapter concentrates on the Bush administration’s approach to these problems. It begins, however, by sketching the conceptual and historical background that has influenced American deterrent policies, and in concluding, it will return to the some of the broader issues surrounding the role of deterrence today.

basic ideas of deterrence The relevant categories and distinctions made about deterrence, although not unambiguous and uncontroversial, are well-known.2 Most broadly, one can distinguish between brute force and coercion. Brute force is the ability to take, hold, and protect physically whatever is in dispute (this notion works well with territory but has problems if what is in dispute is something else like honor, glory, or security, which are the goals and means to higher ends). Two armies fighting, with the victor taking the disputed ground, is the obvious ideal type of brute force. Brute force involves the clash of capabilities, and it is—or can be—zero sum since one side will win and the other will lose, which in turn means that bargaining is secondary, if it occurs at all, and the credibility of threats rests largely on judgments of capabilities. Offense and defense are the two subcategories of brute force. Coercion operates not by directly preventing the target from reaching its objectives but by inducing the target to comply with another’s desires by manipulating the incentives that the target faces. Essentially, the target is told that it will suffer pain or punishment if it does not comply (it is given “an offer it can’t refuse”). Situations of coercion, by definition, cannot be zero sum. Bargaining is at their core because some outcomes are worse for one side, and usually worse for both sides, than varying degrees of compliance and cooperation. Coercion applied to inhibiting the target from taking certain actions is deterrence; coercion in the service of getting the target to change its behavior is compellence, although unfortunately for the analyst many cases are ambiguous in this regard. Because the term “deterrence” is more frequently used than “coercion,” it is used here as a general term for

deterrence, rogue states, and the u.s. policy

135

threats involving punishment except when it is important to distinguish between deterrence and compellence. Scholars also distinguish between general deterrence, which refers to discouraging the target from making a challenge, and immediate deterrence, which occurs if general deterrence fails or is not exercised and the target issues a challenge. Another distinction deals with what is being deterred: direct or homeland deterrence refers to discouraging the other from a direct attack on the state; extended deterrence seeks to put a protective umbrella over other territories or values. Analysts were struck by the importance of deterrence as soon as they learned about atomic bombs. Indeed, in November 1945, Time magazine listed twelve key points about the impact of nuclear weapons on war and politics, and soon thereafter Bernard Brodie and William Borden wrote books that outlined many of the arguments that would frame later policies and debates.3 Brodie’s words summarized much about the new centrality of deterrence brought about by nuclear weapons when he declared that “thus far the chief purpose of our military establishment has been to win wars. From now on its chief purpose must be to avert them. It can have almost no other useful purpose.”4 The reason for this strategic revolution was obvious: defeating the other side’s army would not mean much when the level of destruction exacted on both sides would be enormous. Military superiority no longer could physically protect the state, which came as a psychological shock, especially to U.S. policy makers who always had this ability. In a real sense, each side’s fate was in the other’s hands, despite the fact that, or because, they were bitter adversaries.

deterrence and the cold war The basic ideas about deterrence and nuclear weapons emerged when the United States had a monopoly on the atomic bomb and before the rivalry with the Soviet Union took hold. These ideas might have developed differently in the absence of a cold war. One could argue that the American possession of the atomic bomb was itself a major cause of the cold war, which would make this counterfactual untenable. Nevertheless, it does not violate the laws of history or international politics to imagine several possible worlds with the United States, and perhaps others, having nuclear weapons in the absence of bipolarity. In the event, the cold war did develop, and it shaped our thinking about deterrence. For one thing, the Soviet-American rivalry put nuclear

136

robert jervis

deterrence front and center, and it highlighted the twin objectives at the heart of American policy, objectives that embodied a fundamental tension. On the one hand, the United States had to see that World War III did not occur. On the other hand, it had to protect itself and its allies and make sure that the Soviet Union did not gain important advantages. In an era in which significant power lay in Europe, this meant that, contrary to American tradition and initial planning, the United States had to be deeply involved on the Continent. With the perceived Soviet advantages in ground forces, atomic bombs were the “great equalizer.” Indeed, their existence and the temporary American monopoly on the atomic bomb facilitated an American security commitment to West Europe. At first, however, American decision makers doubted that a prolonged cold war was possible, thinking that it would lead either to a shooting war or to the gradual destruction of the democratic, social, and economic values at the heart of the American way of life.5 Nuclear weapons, combined with intelligence analysis indicating that Joseph Stalin did not want war and could be deterred, were crucial to the acceptance of the policy of long-run containment. As Europe stabilized, attention and conflict shifted to the third world. This was more complicated from the standpoint of deterrence. American interests were less deeply involved on “the periphery,” and as the Soviet nuclear arsenal grew, American deterrent threats became problematic. It was in this context that the importance of credibility and reputation loomed large. They were linked through the device of commitment, which meant that the United States staked its reputation on living up to its word and created reputational interests to supplement the weaknesses of its inherent interests in various parts of the world. Backing down in a crisis was then seen as not only exacting the immediate cost of whatever was at stake, but as endangering much else because it could lead to challenges throughout the world, leading others to view the United States as lacking resolve or a willingness to run risks. The context of mutual deterrence also led to what Schelling called the “rationality of irrationality”—if one side could lead the other to believe that it was not rational or in full control, the other would be inhibited from creating or taking advantage of crises. A strange and dangerous tactic, it was in fact employed (unsuccessfully) by Richard Nixon in Vietnam.6 These tactics have a belligerent tone, but deterrence theorists and sensible leaders understood that if they were to be effective, threats to punish the other side had to be coupled with promises to refrain from such action if the other behaved appropriately. If punishment was expected no matter what the other did, then the target of the deterrent threat would have no

deterrence, rogue states, and the u.s. policy

137

incentive to comply. The same basic model also led to the argument that intensity of preferences mattered a great deal. That is, it was easier to deter a state from doing something it did not care much about than it was to deter it from trying to reach or protect interests it considered to be vital. This means that deterrence was easiest where it was least needed and would be most difficult when and where the target was most motivated. Indeed, this observation dovetailed with the finding—and prescription—of standard realism: states should respect others’ interests. But this prescription was in some tension with the need to demonstrate credibility by engaging in conflicts even when the state’s own interests were not deeply implicated. Still, the admonition to pay careful attention to what the other side had at stake and not to expect from coercion more than it could reasonably provide was central to deterrence ideas. Partly because of the importance of the balance of interests, compellence usually was more difficult than deterrence. Changing the status quo is hard because actors have a major interest in holding on to what they have. Both logic and psychology are involved here. States devote most effort to what they care most about, and so the status quo usually reflects the participants’ most important interests. Retreats also raise the question of what the actor will fight for. Furthermore, possessing something or gaining a position increases the value that actors place on it, and people feel losses more keenly than they do gains.7 Although the United States did employ compellence in some cases during the cold war and the line between compellence and deterrence was often hard to draw because the status quo changed during many interactions and the two sides often had a different sense of what the status quo was, each side realized that using coercion to force the other to give up what it possessed was difficult and dangerous.

alternative positions on deterrence Little about deterrence is uncontroversial, in part because deterrence theory and its offspring, arms control, are a middle position between two other policy options that found different aspects of deterrence unconvincing as a theory and unwise as a policy. William Borden and his followers, most notably Albert Wohlstetter, the mentor of today’s neoconservatives, argued that deterrence was too passive and placed too much faith in threats that were both immoral (because they would involve killing millions of innocent civilians) and incredible (because carrying them out would lead to the state’s own destruction). The standoff of mutual stability was particularly

138

robert jervis

dangerous because it left the United States unable to deter convincingly anything other than an all-out attack on its homeland. Extended deterrence was undermined by what Glenn Snyder called the “stability-instability paradox.”8 Stability at the nuclear level could lead to instability at lower levels of violence because a state with military options could exercise them secure in the knowledge that the other side could not resort to massive retaliation because this would trigger an all-out nuclear war. The United States then needed to be able to match or surpass the Soviet adversary at any possible level of violence. Politically, standard deterrence was immoral, foolish, and ultimately self-defeating because it accepted the indefinite continuation of the Soviet regime. It was immoral because even if it worked, it condemned millions of people to living under tyranny; it was foolish and self-defeating because Soviet aggressiveness stemmed not from the structure of international politics or the need to deter the United States but from the nature of the Soviet regime, both its general characteristics as a tyranny and the values and outlook that stemmed from its communist ideology. As long as the Soviet Union existed, it would be a menace. Critics of deterrence from the opposite end of the spectrum agreed that the standard approach was flawed in both theory and practice, but for quite different reasons. It was too dangerous, relied excessively on threats rather than on political instruments, and heightened the conflict it was meant to deal with. Deterrence rested on fear, yet making the adversary fearful also was expected to lead it to be sensible if not fully rational9 and to be able to adjust its behavior to the incentives the state was providing. Threats and concern with credibility were difficult to keep under control, even though deterrence rested on a non-zero-sum view of the world. Heightened competition was likely to induce a mind-set and political pressures that made moderation and cooperation extremely difficult. Critics also suggested that even if deterrence might be needed in some instances, an excessive reliance on it in the cold war was a mistake because the Soviet Union, although oppressive internally, was not highly motivated to expand. Its behavior was explained less by seeking out weakness in the West or by searching for opportunities to expand its sphere than by its need for security. Although there were significant conflicts of values and interests between the two sides, they did not merit the fierce competition and tactics that characterized the cold war. While, in principle, deterrence was compatible with compromises, its practice generated fears and deep mistrust that blinded each side to the sincerity of overtures on those occasions when the adversary was willing to make them. It undermined diplomacy by inducing a rigid mind-set that de-

deterrence, rogue states, and the u.s. policy

139

valued the search for the common ground and triggered cognitive and emotional biases that crippled the state’s ability to accurately judge others and their intentions. It corrupted the political process by facilitating the ridicule of those who sought compromises with the Soviet Union and who wanted to explore and expand areas of common interest. It distorted appropriate state-society relations because the stress on resolve and reputation meant that American presidents were quick to identify dissent with disloyalty if not treason and to believe that protests against prevailing policy would have the effect if not the intent of emboldening the Soviet Union and reducing the credibility of the threats that served as the shield of the republic. Richard Nixon was extreme in this regard, but other presidents shared this attitude, which, while not endorsed by deterrence theorists, could be grounded in deterrence theory.

the impact of the past on the future When the apparent need for coercion became salient after September 11, 2001, one might think that the history of its employment during the cold war would have settled the debate among the three competing schools of thought about deterrence. But it did not. Testing deterrence theory is extremely difficult, and while many individuals have reached conclusions about the nature and effectiveness of deterrence theory, consensus eludes us.10 The reasons are many: the theory is underspecified, the historical events used to support various theories are hard to code, and the available information is far from conclusive. Just as the debate among historians between the orthodox and the revisionist school of thought about the outbreak of the cold war is not settled, so international politics scholars disagree as to the effects of American deterrence on the Soviet Union and the course of the cold war. Indeed, there is a resemblance between deterrence theorists and the orthodox school of thought on the origins of the cold war and between the critics of deterrence from the left and revisionism. Although the association is not close enough to collapse the two debates into one, the failure to settle one of the arguments does make the continuing disagreement on the other less unexpected. Both debates involve, among other things, difficult judgments on the simple and even childish question, Who started it? I do not mean to demean the severity and importance of major disputes in international politics by comparing them to playground squabbles, but both embody central questions about which (if either) party had reasonable objectives and if any party was pursing their objectives with due respect for others.11

140

robert jervis

With the evidence ambiguous, there is much room after the fact for the earlier disagreements to continue. While a few debates on specific issues have been resolved by the opening of documents (it is now clear, for example, that Stalin approved Kim Il Sung’s initiation of the Korean War) and the further evidence of Stalin’s brutality and Nikita Khrushchev’s restless ideology have moved many scholars a bit to the right, none of the three schools of thought is about to drive the others from the field. Here, as in so many areas, we see the world through the lenses of our general views about how the world works, if not through our formal ideologies. So despite the end of the cold war and what we have learned about it, attitudes toward the role of deterrence today not only resemble the three schools of thought, but most participants whose careers span both periods have maintained their positions, despite the changes in circumstances and adversaries. The scope for disagreement is more limited on one issue that, as we will see, is highly relevant to current concerns. Long after he left office, Robert McNamara declared that the “sole purpose [of nuclear weapons] . . . is to deter the other side’s first use of its strategic forces.”12 Even scholars like myself, who think that this understates the role of these weapons and who believe that they were able to provide a significant degree of extended deterrence, agree that there are severe limits on what nuclear weapons can accomplish. Unless the state’s adversary is entirely risk averse, these weapons cannot be readily used to change the status quo nor can they deter the adversary from all uses of force. The cold war also showed the limits of abstract models in capturing international interactions and in providing appropriate guidance for state action. Critics of the theory and practice of deterrence from both the left and the right pointed out the twofold difficulty. First, there is a great deal of variability from one actor to another in terms of goals, perceptions, and beliefs. Second, international communication—and coercion is a form of communication—is difficult because the two sides often live in different conceptual and perceptual worlds (and of course it is more complicated when there are more than two sides). Many discussions of international-relations theory imply a world that resembles a game of poker, but a better starting point is the Japanese story and movie Rashomon, in which each of the characters views the same scene very differently. Even though deterrence theory does a quite good job of explaining the behavior of the United States and the Soviet Union during the Cuban Missile Crisis, for instance, Khrushchev called this episode the Caribbean Crisis, and he believed its origins predated the Soviet placement of missiles on the island, which in his eyes was partly justified

deterrence, rogue states, and the u.s. policy

141

by the American attempts to overthrow Castro. Since the Americans saw things quite differently, they were slow to expect this move or to recognize that providing a noninvasion pledge could be an important tool in resolving the crisis.

deterrence and rogues Although the current situation is unlike the cold war in many ways, most obviously in the fact that the United States is worried not about a peer competitor but about much smaller actors13 and has to act in a more complex environment, the structure of attitudes toward deterrence displays a great deal of continuity. The terrorist attacks of September 11, 2001, focused the Bush administration’s attention on these issues, leading it to take low-probability events very seriously, thereby lowering its tolerance for risk, or at least for the risk of armed attack.14 Linked to and exacerbating this syndrome is the perception of a nexus among terrorists, weapons of mass destruction (WMD), and rogues: “imagine 9 / 11 with WMD” is the common phrase. While terrorists could perhaps make their own chemical or biological weapons and might buy radiological or nuclear weapons on the black market, the possibility that they could get them from rogue states obviously magnified the sense of danger from the latter. For the Bush administration, rogues were a greater danger than they were before September 11, and stiffer tactics were called for, both to combat them and to see that they do not gain nuclear weapons. For both psychological and real reasons, the end of the cold war elevated the danger from rogues. The removal of the threat of the Soviet Union and the concomitant danger of World War III means that our minds naturally turn to lesser threats, ones that would have been put aside before. It is an exaggeration to say that individuals and collective actors have a quota of worry that remains constant; however, when our primary sources of worry are removed, others previously lower in the queue will loom larger. There also is more reason to worry about rogues now because the disintegration of the Soviet Union removed an important source of both protection and constraint for at least some rogues, increasing their need for nuclear weapons and reducing the extent to which a patron could insist on moderation. Post– cold war globalization, even if exaggerated, also plays a role in increasing the danger from rogues. With increasingly dense interconnections throughout the world, regional disturbances can have larger consequences for the American economy and world order. Furthermore, one aspect of globalization is the increased role of networks and private actors whose activities can be a

142

robert jervis

grave menace. The obvious example is the A. Q. Khan network that spread nuclear technology in a way that few had imagined.15 It is hard to define rogues with any precision, and in any event the category is not so much socially as politically constructed. One can almost say that perceived danger leads the United States to designate a regime as a rogue as much as the label is applied to states with objective characteristics. Nevertheless, it seems that the crucial dimensions involved are unpredictability, the violation of international law and norms, expressed hostility to American interests, and domestic oppression. The analytical problem is that these four dimensions do not always coincide. Repressive regimes may be quiescent internationally, as is the case for Mugabe’s Zimbabwe. Even states that are not only repressive but also revolutionary may be conservative in their international behavior. Although there is room to disagree, Mao’s China generally fit into this category. Conversely, Iran, which tops Bush’s list of rogues, is at least semidemocratic, more so than our ally Pakistan was until early 2008. Syria sponsors extensive terrorism, yet is not generally labeled a rogue; North Korea has not engaged in terrorism since the mid-1980s and yet has been deemed a rogue threat. Libya removed itself from the list by being willing to give up its WMD, but this occurred without any changes in its domestic rule and led the Bush administration to overlook its recent attempt to assassinate a leading member of the Saudi royal family. Of course, policy makers are not academics, who presumably seek clear definitions. Problems occur only if the categories being used do not map well on to their preferred policies

the bush administration and rogues The Bush administration comes to the problem of rogues with the combination of the changes induced by September 11 and the hostility toward deterrence its members developed during the cold war, which were compounded by their well-known preference for doing ABC (Anything But Clinton). Although Bush himself had not been a participant in the cold war deterrence debates, many of his leading advisors were critics from the right, and many of the important second-level officials had studied with Albert Wohlstetter. Excessive attention has been given to their pro-Israel (actually pro-Likud) attitudes and the supposed influence of Leo Strauss, but more important was the influence of Wohlstetter and his beliefs about the fragility of deterrence. This was obvious even before September 11 in the strong drive to escape from the Anti-Ballistic Missile Treaty and to construct missile defenses

deterrence, rogue states, and the u.s. policy

143

aimed at China if not Russia, as well as at rogues. Of course, there are good if not convincing arguments for striving for self-protection and buying insurance against even unlikely contingencies, but the call for defense also may represent a manifestation of unilateralist impulses in that it seeks to remove the dependence of the United States on others’ behavior. The idea that American fate rests in the hands of adversaries is not a comfortable one, and even many proponents of deterrent stability came to it only because it was forced on them in the sense that there was no realistic way to escape from the situation of mutual assured destruction that existed during the second half of the cold war. No defense can be perfect, and an antiballistic missile system would be of no avail against weapons delivered unconventionally, perhaps by terrorists who had acquired them from rogues. In response, the Bush administration has stressed the need for what it calls preemption, but what is in reality is a policy of preventive war. “We cannot let the smoking gun be a mushroom cloud” was the phrase that became something of a mantra in the run-up to the Iraq war. There are some roots in cold war thinking here, and historians have discovered how seriously the Truman and Eisenhower administrations thought about a preventive war against the Soviet Union (Kennedy—and later the Soviet Union—also contemplated preventive strikes against the PRC).16 In retrospect, some have hinted that the decision not to launch preventive strikes was a missed opportunity. The fact that prevention received so little public attention before Bush’s renouncements should not blind us to the fact that it is a major strand in traditional international politics. Many wars have at least an element of prevention in them, sometimes a very major element, and even if John Gaddis exaggerates, prevention has played a significant role in American history.17 If prevention repudiates deterrence, another strand of the Bush policy for dealing with rogues and WMD relies heavily on it. Even before the September 11 terrorist attacks, the Bush administration said that in addition to the more traditional roles of deterring and defeating an adversary, our military posture should contribute strongly to what it called dissuasion. By this it meant convincing other countries not to build up their military forces. This was one of the main points of the proposed “Defense Planning Guidance” drafted by Zalmay Khalilzad for Paul Wolfowitz and his boss, Secretary of Defense Cheney, in 1992.18 But it is one thing to seek a military lead that is so great that others will see that it is pointless to try to match the American forces and another to hope to dissuade rogue states from acquiring nuclear weapons. Rogues are not seeking capabilities comparable to America’s; they

144

robert jervis

are seeking only a fairly small force. Although the administration has not publicly said a great deal about dissuasion, its hope that implicit and explicit threats can lead countries to forsake their nuclear programs asks more of coercion than many deterrence theorists would expect that it can achieve.

four crucial issues Four issues are at the heart of debates over how to treat rogues, how the United States should respond to the issue, and what U.S. policy makers can hope to accomplish. We should note that consistencies in the various positions on these issues are explained more by psychology than by logic: no one wants to believe that a task is both necessary and beyond our reach. The first issue concerns what the United States needs to achieve. What are U.S. objectives with respect to rogues? Should the United States embrace ambitious policies? Second, what is the source of rogues’ behavior? Why are they acting in undesirable ways? Third, what are the rogues’ goals and what are they seeking? Finally, how can the United States and other states best influence them? From the standpoint of coercion, the fundamental point is that if policy makers believe that the United States must change the rogues’ policies in ways that are of fundamental importance to them, then coercion will be extremely difficult. As Patrick Morgan says, “the challengers’ objectives are ultimately responsible for success or failure.”19 This observation could be expanded, because it is the combination of both sides’ objectives that renders coercion particularly easy or hard. The Bush administration’s objectives are very ambitious, and if its analysis of the rogues is correct, the rogues have ambitious goals as well. This means that coercive policies face great obstacles. More specifically, the Bush administration often seeks regime change. The president and his colleagues argue that U.S. security rests on the spread of democracy abroad because democracies behave fundamentally differently from tyrannies. States that oppress their own peoples are very likely to turn on others as well. A heterogeneous system composed of democracies and tyrannies is unstable, and although the United States is the leading power in the world, it cannot be the upholder of the status quo but rather must seek a system filled with democracies if it is to be safe.20 The Bush administration’s objectives are linked to its analysis of the sources of the rogues’ behavior, that is, they are driven by internal impulses not by their external environment. Evil does as evil is. As President Bush said in his 2002 State of the Union address, the rogues’ threat derives from “their true nature.” President Bush and

deterrence, rogue states, and the u.s. policy

145

many of his colleagues are what Kenneth Waltz would call “second-image” thinkers in believing that the nature of the state’s domestic system is the most important determinant of its foreign policy.21 This is one of the main reasons why realists recoil from the Bush administration’s stance. This second-image analysis tells us what the rogues’ goals are—or at least what they are not. They are not reacting to a difficult and hostile environment; they are not seeking security. In March 2007, Iran told the International Atomic Energy Agency that “its fear of attack from the U.S. and Israel prompted its decision to withhold information from the agency.”22 Although realists might not accept this explanation, they would find it plausible and would expect any regime in this situation to have real security fears. For the Bush administration, however, the claim is clearly false and the program stems from the tyrannical character of Iran’s government. Finally, the second-image view implies that conciliating rogues would constitute rewarding them for bad behavior, which would amount to appeasement, that there is little space for bargaining, and that only relentless pressure can succeed in modifying their behavior. Indeed, taken to its logical conclusion, this is a zero-sum situation because the United States needs to change the rogues’ regimes. Nothing else will permanently change their behavior; there is nothing to talk about and no common ground to be sought. Consistent with their views of the cold war and how it ended, however, Bush administration officials saw a way out. Coercion can at least temporarily restrain a rogue while internal dissent wears away at it. The United States can help end these regimes by covert support for local dissidents, as it did during the cold war. Tyrannies cannot last long in the modern world, and isolating them and keeping the pressure on can shorten their life spans. Critics of the Bush policy, both those who believe in the efficacy of deterrence and those who criticize it from the left, argue that the Bush administration’s goals are both more ambitious than is required and cannot be reached in practice. Rogues are not as dangerous as Bush says; the problem lies with their behavior, not the nature of their regimes. In most instances, they seek internal and external security and can be influenced by carrots as well as by sticks. Consistent with deterrence arguments, the American task is to affect the incentives the rogues face rather than to overthrow them. Furthermore, in many cases what the rogue values most is its security, which means both that its acquisition of nuclear weapons will not be a fundamental menace to American interests and that reassurances and guarantees could play a large role in getting them out of the nuclear business.

146

robert jervis

Of course, goals differ from one state to another as all rogues are not alike. But, all else being isolated, they do have legitimate security concerns. This is especially true of North Korea, which has been the target of American nuclear threats. Furthermore, it does not seem that the Democratic People’s Republic of Korea can any longer aspire to dominate, let alone conquer, South Korea and so cannot be seeking nuclear weapons for this purpose. Indeed, the North Korean foreign minister was quoted as saying, “If the regime’s security is guaranteed, there is no reason to possess a single nuclear weapon.”23 Here then, although the negotiations will not be easy, reassurances and inducements may be effective in reducing and perhaps eliminating North Korean nuclear weapons. This may not be the case with Iran, a country that has not only security fears but also regional ambitions and great pride. The United States may have to live with an Iran that has or could produce nuclear weapons, although this eventuality is years in the future. Just as Bush and his critics differ on the sources of rogue conduct, they differ on the likely effects of rewards and threats. For the Bush administration, inducements would reinforce rogues in their norm-defying behavior and strengthen them internally, while pressure will exacerbate the internal divisions, hearten the dissidents, and hasten the regime’s demise. Critics disagree, believing that external pressure feeds nationalism, which may be the only thing that keeps these regimes in power,24 and arguing that the way to speed the downfall of rogues is to bring them into greater contact with international society. While at least some members of the Bush administration appear to believe that bombing the Iranian nuclear facilities would not only set back the nuclear program but would trigger an uprising, critics expect that such an action would solidify the regime and think it could be undermined by more engagement. Of course tyrannical regimes try to limit such contact, and North Korea has been particularly successful in this regard, but in today’s world this degree of isolation brings with it economic ruin, as North Korea has discovered. Debates about the impact of isolation on the one hand and engagement on the other go back to the cold war, and here too are patterns that can only be explained by psychology and ideology, not by logic and empirical evidence. During the cold war, liberals called for isolating South Africa but expanding contact with the countries of Eastern Europe and, to a lesser extent, the Soviet Union. Conservatives advocated the opposite, arguing that the only way to bring down the apartheid regime was to reduce its fear of the rest of the world and to open it to outside contact, while saying that boycotts

deterrence, rogue states, and the u.s. policy

147

and pressures would be effective against Communist regimes. Although each camp could come up with convoluted explanations for their strained positions, the shunning response probably stemmed from extreme ideological distaste for the regime, and engagement was favored when the other country, although not viewed favorably, was seen as less abhorrent.

rejecting and embracing deterrence A more central tension in the Bush policy is that while it seems to reject deterrence, it is actually built on a strong and unconvincing version of it. Members of the Bush administration believe that if rogues get nuclear weapons, the United States and others will not be able to deter them from taking a range of undesired actions. These actions vary from one rogue to another. North Korea might expand its criminal activities, extort aid from South Korea, and perhaps resume its terrorist activities. Iran would seek to become a regional hegemon and use its nuclear weapons as a shield behind which it could intimidate neighbors, sponsor more terrorism, and engage in covert actions with the assistance of local Shia populations in the region. There are several problems here. To start with, the American stance may magnify the rogues’ bargaining positions. By claiming that rogues are so dangerous because they do not fear retaliation, the United States makes it easier for them to adopt Schelling’s “rationality of irrationality” strategy. Even if they do not, the Bush administration’s views imply that the United States will not be able to act on the assumption that they are rational. This perhaps increases the credibility of the threat to use force against them to prevent their getting nuclear weapons, but it undermines the credibility of the American threats if and when they do get the atomic bomb. The supposed drastic consequences of nuclear weapons in the hands of rogues may then be at least partly a self-fulfilling prophecy. The administration’s claims imply that American extended deterrence would falter in the face of a rogue’s nuclear weapons, a position that mirrors that held by the hawks during the cold war. Here too there could be a self-fulfilling prophecy: the United States will be deterred if it is unwilling to use force and believes that it cannot successfully contain the rogues. But the Bush administration position is odd in that while it invokes a version of the stability-instability paradox, this is done in a context in which the adversary has only a few nuclear weapons and the United States retains enormous superiority at all military levels. In the cold war, the Soviet Union not only had nuclear parity after the mid-1960s but had conventional superiority

148

robert jervis

that meant that the United States could thwart an invasion of West Europe only by resorting to nuclear weapons. This made the U.S. fears, even if not correct, at least plausible. The situation is very different now because the United States has both conventional and nuclear superiority. This military superiority does not guarantee that the United States can deter the rogues from taking any and all undesired activities, in part because the balance of interest might be in the rogues’ favor because they might care more about undertaking these activities than the United States cares about blocking them. But American interests should not be underestimated, and America’s obligations throughout the world may work in its favor because allowing a rogue to prevail in one area could have untoward effects on U.S. interests in other parts of the world. The United States would never allow itself to be at the mercy of a nuclear-armed rogue. The cold war experience indicates that while nuclear weapons discourage adversaries from even contemplating an all-out attack on the state, they are not a silver bullet that solves all foreign policy problems. They do not readily permit coercive changes of the status quo, nor are they likely to deter an American conventional response to a challenger’s adventures. Ironically then, the Bush administration simultaneously denigrates American deterrence and exaggerates what rogues will gain from nuclear weapons. The odd claim is that the enormous American nuclear arsenal and conventional military establishment cannot produce the coercion that the Bush administration thinks is needed but that the rogues can exert great influence if they have even a few nuclear weapons.25 Put this way, one can see that while the Bush admininstration denies deterrence in one way, it affirms a strong version of it in another. The denial of the American ability to deter the rogues if they get nuclear weapons affirms their ability to deter the United States from taking effective counteractions in the face of their provocations. Nuclear weapons in the hands of rogues have great deterrent power because the United States is so averse to casualties and risk that it would be paralyzed by rogues, who would then gain the freedom to take a wide range of undesirable actions. It is not surprising that the administration has not explicated the self-image implicit here, but it does fit with one strand of conservative thinking that is preoccupied with the specter of American decadence and lack of will.26 Whatever the cause, the belief that rogues are hard to deter and the United States is easy to deter is both underdeveloped theoretically and highly consequential, having underpinned the decision to overthrow Saddam Hussein.27 The claim that the United States could not deter the rogues if they gained

deterrence, rogue states, and the u.s. policy

149

nuclear weapons also implies that deterrence today is quite effective—nuclear weapons would enable rogues to do many things that are now beyond their reach. The raises several questions. What behavior can we now deter the rogues from that we could not if they had nuclear weapons? What makes our threats so effective today? In commenting on the Iranian seizure of British sailors in March 2007, a Wall Street Journal editorial declared, The world should keep in mind that Iran has undertaken this latest military aggression while it is still a conventional military power. That means that Britain and the U.S. can still respond today with the confidence that they maintain military superiority. That confidence will vanish the minute Iran achieves its goal of becoming a nuclear power. Who knows what the revolutionaries in Tehran will then be capable of.28

But if the sailors had been seized only after Iran had a atomic bomb, the editorial probably would have argued that it was nuclear weapons that enabled this bold move. In fact, it is by no means clear that nuclear weapons would give Iran much more freedom of action. They would probably give it direct deterrence—that is, nuclear weapons would deter an all-out American attack—but there is little reason to expect that a small nuclear arsenal would do much to compel others or provide a shield behind which it could engage in greater provocations than it is now undertaking. The experience of North Korea may be relevant here. North Korea is of course a rogue with a small nuclear arsenal, and if the claims of the Bush administration are correct, it should be behaving much more boldly, increasing its demands against its neighbors and the United States and flaunting its new power. In fact, it is hard to detect any change in its foreign policy after it acquired nuclear weapons. Of course, North Korea’s stockpile is small and it might be biding its time before behaving outrageously, but at least at this point North Korea’s behavior disconfirms the gloomy predictions underlying the Bush policy. Finally, coercing rogues is complicated by the fact that understanding other countries is extremely difficult, especially when contacts are limited and hostile. The cold war was filled with bizarre and dangerous episodes stemming from the Rashomon effect,29 and one reason why the United States was unable to coerce Saddam was that he was more worried about Iran, his generals, and a Shiite uprising than he was about the United States, which he believed would at most invade the southern part of his country.30 Deterring someone who is not paying attention to one’s threats is difficult, and the failure to understand that the United States will carry out its threats is a form

150

robert jervis

of rationality of irrationality. The lack of understanding often is mutual, and the United States and others often fail to understand the goals, beliefs, and perceptions of the rogues. One of the ironies of the war in Iraq is that if the United States had been able to foresee the resistance and civil war that followed its successful invasion, even Bush almost surely would have been deterred from undertaking a policy of regime change. Indeed, the post–cold war world has seen five major instances in which the United States has had to employ force because its threats were not efficacious, despite its military might and apparent incentives to carry out its threats if need be.31 This does not bode well for the future, especially since decision makers seem not to have learned this lesson and underestimate the difficulties of understanding others and conveying views, threats, and promises to them.

conclusion The Bush administration’s stance toward the rogues rested on twin beliefs: they will be very hard to deter (and the United States will be easy to deter) if they gain nuclear weapons, and their foreign policies stem from their tyrannical domestic rule. While there is nothing logically inconsistent about these two beliefs, they do point in somewhat different directions. The former implies that the United States should be willing to pay a high price to prevent rogues from getting nuclear weapons. While the price might be in the form of a war, it could also involve concessions, especially concessions in the form of promises not to overthrow the rogue’s regime and even to support it by establishing normal political and economic relations. Concessions of this type, while quite reasonable and indeed cheap to realists, who believe that the rogue is largely driven by the desire for security, contradict the Bush second-image claim. If the rogue is inherently dangerous because of the nature of its regime, then such bargains must be resisted. At best, they gain time in which the regime may collapse, but it is more likely that security and recognition will bolster the regime, which will redouble its search for ways to exert its malign influence. Insisting on regime change means that agreement is impossible, and, conversely, almost any conceivable denuclearization agreement will require the United States to provide some sort of guarantees that it will not attack or seek to overthrow the regime. As Robert Litwak puts it in his excellent book, the U.S. officials have to decide whether they aim for regime change or are willing to settle for a change of behavior.32 It was the choice of the latter that permitted the agreement with Libya, when President Bush ignored the advocates of regime change and agreed to give

deterrence, rogue states, and the u.s. policy

151

priority to the ending of Libya’s WMD programs.33 Although it is too early to tell how the negotiations with North Korea will turn out, the agreed statement of September 2005 and the agreement of February 13, 2007, were made possible by American promises to respect North Korea’s sovereignty, and the subsequent progress toward denuclearization has been correctly denounced by the neoconservatives as coming at the price of abandoning the administration’s initial assumptions and goals. Ironically, the drastic overestimation of the damage that would be caused by rogues gaining nuclear weapons may have led the Bush administration to a more appropriate policy. If the administration had adopted a more realistic view of what nuclear weapons could and could not do, it might not have been willing to sacrifice the objective of regime change to get rogues to agree to abandon their WMD programs. Although realist proponents of deterrence have some confidence that rogues could be contained even if they have nuclear stockpiles, they see this as a sensible bargain. Indeed, they do not see the willingness to forgo regime change as a major sacrifice. But for President Bush, abandoning regime change was a setback. Even if advocates of deterrence think that his worries were misplaced, they should recognize that this error in the end made the policy more sensible. Because George W. Bush is no longer president, many may see this assessment of U.S. policy to be mere history. The issues raised by the proliferation of WMD, however, have not gone away, and many of the impulses that drove President Bush’s stance toward rogues continue, especially as unsavory regimes and nonstate actors remain interested in acquiring nuclear weapons. Prevention is a motive that runs throughout international politics and that has contributed to many if not most wars, and while some of George W. Bush’s critics see him as strangely preoccupied by rogues, we should remember that it was Bill Clinton who was ready to attack the North Korean nuclear facilities if an agreement had not been reached at the last minute in 1994 and who endorsed the overthrow of Saddam Hussein. The world in which the new administration will have to operate will be similar to that confronting the Bush administration. Even if it is true that his misguided policy was responsible for a significant portion of the problems his successors face, this damage cannot be undone. Furthermore, there is no reason to expect that the broad camps established during the cold war and reinforced more recently will dissolve. Even if some of the ideas of the neoconservatives have been discredited, their skepticism of deterrence is widely held, and it is likely that if the post-Bush administration does not share it, a later one will.

152

robert jervis

Critics of the Bush administration—and I am one of them—must acknowledge that even if they are right in believing that it greatly underestimated the efficacy of American deterrence in inhibiting unacceptable behavior, deterrence and, even more, compellence are not entirely easy. At first glance, they should be, because the United States has such overwhelming nuclear, conventional, and economic superiority. The situation is nothing like it was in the cold war. Material superiority, however, does not automatically translate into influence. Even if it is true, as most security specialists believe, that American military might is more than sufficient to deter direct attacks by rogues, broader or extended deterrence is less easy. There are five reasons for this. First, the difficulty of successful coercion increases roughly in proportion to the ambition of the state’s goals. As I noted, the Bush administration’s belief that tyrannies are uniquely dangerous led it to seek regime change, although not consistently so. Coercion is almost impossible here. But even if the United States does not seek to overthrow the other side, it is likely to be acting on issues and in geographic areas in which its interests are less deeply involved than those of the adversary are. This makes successful coercion more difficult. Second, convincing other countries not to obtain nuclear weapons is difficult as long as the main instruments wielded are threats and coercion. These magnify the targeted state’s sense of being embattled and, if the target of coercion is driven largely by fear, will increase its incentive to secure such weapons. In this situation, a decrease in the threat and indeed security assurances will be more effective. Giving such “rewards” to tyrannical regimes, however, will meet with domestic opposition. The problem here is related to the widely shared premise that it is easier to deter a country from getting nuclear weapons than from behaving in unacceptable ways after it acquires a nuclear arsenal. In one way, this is common sense: these weapons do increase a nation’s power, and it is hard to see why a country would want them if its leaders did not think that they would help advance their interests. But this argument may be wrong. Countries may be less rather than more prone to disturb the status quo if they have nuclear weapons if for no other reason than that they know that they have become prime targets for preemption and, failing that, massive retaliation. Norms and legitimacy also focus on behavior rather than capabilities. Despite the increasing acceptance of preventive war, at least in the United States,34 attacking a country because it may be building powerful weapons is generally less accepted in the world than attacking it for unacceptable behavior. The third problem is that coercing rogues combines deterrence and compellence, or at least blurs the distinction between them. The United States

deterrence, rogue states, and the u.s. policy

153

not only wants to see that these countries continue to refrain from taking unacceptable actions; it also demands significant changes in their behavior. From Iran it seeks a halt in ostensibly civilian nuclear programs that could be harnessed to produce a nuclear arsenal; from North Korea, it seeks nuclear disarmament. From both countries it seeks moderate behavior and, in the case of Iran, ending support for groups U.S. officials label as terrorist. These are major modifications, which means that they will not come easily. Furthermore, effective nonmilitary coercion will require the cooperation of numerous countries. Their reluctance to adopt strong sanctions is not likely to change with a new American administration. Indeed, the irony is that their willingness to be as severe as they have been is partly the result of their fear that if they did not do so Bush would use military force. If they believe that the new administration is more reasonable, they might be tempted to do even less. Of course military action remains possible, but a fourth reason why massive superiority does not make coercion easy is that in many circumstances, the United States will be “self-deterred.” During the cold war, this phrase was used quite broadly, often including the obvious fact that the United States could be deterred by a Soviet military response to any of its moves. But this is an unfortunate expansion of the concept, which should be limited to the cases in which a country is deterred by factors other than retaliation by others. Ironically, the disparity in military power between the United States and its rogue adversaries increases self-deterrence, especially because the United States may want to befriend or rebuild those it is currently coercing. Although the cry “pick on someone your own size” carries much less weight in international affairs than in the school yard, and indeed good realists understand that the best adversary to fight is one that is weak,35 there remain some constraints imposed by advantage. During the war in Kosovo, the North Atlantic Treaty Organization tried to avoid killing not only civilians but Serbian soldiers as well. World and domestic opinion can be offended by using a sledgehammer against a fly. This is particularly true for the use of nuclear weapons. The public perception that nuclear retaliation constitutes excessive force would inhibit the first use of nuclear weapons, even against adversary WMD stockpiles, and might undercut the American threat to wipe a country off the face of the Earth if it was responsible for even a small nuclear explosion in the United States. The fear of self-deterrence lies behind the American desire to develop lower yield nuclear weapons, which would produce less collateral damage and therefore be more usable and provide the basis for more credible threats.

154

robert jervis

The continuing impact of Iraq creates two forms of self-deterrence, although these do verge on the broader form of the concept. First, the American presence in that country gives hostages to others, especially Iran, by creating vulnerabilities that they could use to harm or exert pressure on the United States. It is now harder to threaten the use of force against Iran because the American presence in Iraq has given Iran a number of powerful instruments it could use by increasing its support for the Shia and even Sunni insurgents. Second, Iraq has driven home the point made so well by Winston Churchill: “let us learn our lessons [from the Boer War]: never, never, never believe any war will be smooth and easy or that anyone who embarks on the voyage can measure the tides and hurricanes he will encounter.”36 The fifth reason why coercion is hard stems from its being a form of cooperation and communication. It is cooperation because it seeks to be “an offer the other side cannot refuse,” which means that the other side does indeed have to decide whether to accept it or not. While this decision is subject to incentives that the state can affect, Vietnam is a useful reminder that asymmetries of interest can sometimes offset asymmetries of military power. Coercion also requires communication. It can succeed in the face of significant misunderstandings—and is sometimes made possible by them—but in most cases it works best and may work only if the adversary understands the state’s threats and promises. Although this was not easy in the cold war, it was rendered less difficult by the concerted attention each side gave the other and the years of experience that they accumulated in dealing with one another. With rogues, communication is much more difficult. The post–cold war era has seen five instances in which coercion failed and the other side did not understand what the United States would do. Part of the reason lies in the complexity of deterrence under current circumstances, as T. V. Paul has discussed in the introductory chapter. Part also comes from the inherent difficulty in melding diplomacy and force.37 Part also stems from the difficulties dictatorships have in understanding their environments, perhaps compounded by the plethora of information the United States puts out that can make it more rather than less difficult for others to estimate what it will do.38 In many of our academic exercises, we assume that communication proceeds quite clearly. Policies that are based on this assumption are likely to fail. It is not only that the rogues have trouble understanding the United States, but American policy makers also have trouble understanding rogues. For all the debates about whether Iraq was or was not a menace, it is clear that we failed to understand Saddam’s regime. The extent of corruption, the degree to which Saddam was cut off from information about the world,

deterrence, rogue states, and the u.s. policy

155

and the fact the United States was of secondary importance to him escaped almost all observers.39 Perhaps we understand other rogues better, but this statement is little more than a hope. Life might be easier if all rogues were alike, but this is not the case. It is not a matter of understanding rogues as a category but of understanding each particular troublesome state. The United States then needs what Alexander George called actor-specific information, a kind of knowledge that is very hard-won. 40 It is unlikely that a onesize-fits-all approach will work, but a variegated diagnosis and policy are difficult. The felt need to coerce rogues is then not likely to go away. But even done sensibly, it presents us with significant problems. Critics of the Bush administration sometimes imply that once the United States abandons Bush’s misguided diplomacy, allies will flock to the American side and rogues will conform to what the United States government sees as normal standards. Some coercion may be needed, but a better combination of promises and threats should do the trick. Perhaps it will, but the history of the cold war, the interregnum before September 11, 2001, and the subsequent Bush years indicate that massive power, even used more skillfully than it has been recently, faces great obstacles in producing a better and a more peaceful world.

notes 1. George Quester, Deterrence before Hiroshima (New York: Wiley, 1966); Richard Overy, “Air Power and the Origins of Deterrence Theory before 1939,” Journal of Strategic Studies 15 (March 1992): 73–101; William Mason and Sally Mendoza, eds., Primate Social Conflict (Albany: State University of New York Press, 1993); Lee Dugatkin, Cooperation among Animals: An Evolutionary Perspective (New York: Oxford University Press, 1997). 2. See Patrick M. Morgan, Deterrence Now (New York: Cambridge University Press, 2003). 3. “Twelve Points,” Time, November 12, 1945, 28 (I am grateful to McGeorge Bundy for pointing this out to me); Bernard Brodie, ed., The Absolute Weapon (New York: Harcourt, Brace, 1946); William Borden, There Will Be No Time (New York: McMillan, 1946). 4. Bernard Brodie, “Implications for Military Policy,” in Bernard Brodie, ed. Absolute Weapon, 76. 5. Gregory Mitrovich, Undermining the Kremlin: America’s Strategy to Subvert the Soviet Bloc, 1947–1956 (Ithaca: Cornell University Press, 2000). 6. H. Robert Haldeman, The Ends of Power (New York: New York Times Books, 1994), 82–83, 98; Jeffrey Kimball, The Vietnam War Files: Uncovering the Secret History of the Nixon-Kissinger Era (Lawrence, KS: University Press of Kansas, 2004). 7. This is the basic argument of prospect theory. Robert Jervis, “Prospect Theory and Human Nature,” Political Psychology 25 (April 2004): 163–76. 8. Glenn Snyder, “The Balance of Power and the Balance of Terror,” in The Balance of Power, edited by Paul Seabury (San Francisco: Chandler, 1965), 184–201. 9. Patrick Morgan, Deterrence: A Conceptual Analysis (Beverly Hills, CA: Sage, 1977). 10. Morgan, Deterrence Now, chapter 4. 11. The touchstone of reasonableness is central to the analysis of the origins of the cold war

156

robert jervis

in Melvyn Leffler, A Preponderance of Power: National Security, the Truman Administration, and the Cold War (Stanford: Stanford University Press, 1992). 12. Robert McNamara, “The Military Role of Nuclear Weapons: Perceptions and Misperceptions,” Foreign Affairs 62 (Fall 1983): 68; also see 79 and his interview in Michael Charleton, From Deterrence to Defense (Cambridge, MA: Harvard University Press, 1987), 18. 13. Ivan Arreguín-Toft, How the Weak Win Wars: A Theory of Asymmetric Conflict (New York: Cambridge University Press, 2005); T. V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (New York: Cambridge University Press, 1994); Andrew Mack, “Why Big Nations Lose Small Wars: The Politics of Asymmetric Conflict,” World Politics 27 ( January 1975): 175–200. 14. Ron Suskind, The One Percent Doctrine: Deep inside America’s Pursuit of Its Enemies Since 9 / 11 (New York: Simon & Schuster, 2006). It did not adopt this stance toward other dangers, most obviously that posed by global climate change. 15. Gordon Corera, Shopping for Bombs: Nuclear Proliferation, Global Insecurity, and the Rise and Fall of the A.Q. Khan Network (New York: Oxford University Press, 2006). 16. Marc Trachtenberg, History and Strategy (Princeton: Princeton University Press, 1991), chapter 3; Willam Burr and Jeffrey Richelson, “Whether to ‘Strangle the Baby in the Cradle’: The United States and the Chinese Nuclear Program, 1960–64,” International Security 25 (Winter 2000 / 01): 54–99. 17. Dale Copeland, The Origins of Major War (Ithaca: Cornell University Press, 2000); John Lewis Gaddis, Surprise, Security, and the American Experience (Cambridge, MA: Harvard University Press, 2004); Scott Silverstone, Preventive War and American Democracy (New York: Routledge, 2007). 18. The document is summarized and excerpted in stories in the New York Times, March 8 and May 24, 1992. See Patrick E. Tyler, “U.S. Strategy Plan Calls for Insuring No Rivals Develop,” New York Times, March 8, 1992, 1; “Excerpts from Pentagon’s Plan: ‘Prevent the Re-emergence of a New Rival’,” New York Times, March 8, 1992, 14; Patrick E. Tyler, “Pentagon Drops Goal of Blocking New Superpowers,” New York Times, May 24, 1992, 1. Also see Zalmay Khalilzad, From Containment to Global Leadership? America and the World after the Cold War (Santa Monica: RAND, 1995), and Robert Kagan and William Kristol, eds., Present Dangers: Crisis and Opportunity in American Foreign and Defense Policy (San Francisco: Encounter Books, 2000). 19. Morgan, Deterrence Now, 113. 20. Raymond Aron, Peace and War, translated by Richard Howard and Annette Baker Fox (Garden City, NY: Doubleday, 1966), 99–124, 373–403; and Stanley Hoffmann, “International Systems and International Law,” in The International System, edited by Klaus Knorr and Sidney Verba (Princeton: Princeton University Press, 1961), 207–9; for the United States as a revolutionary power see Robert Jervis, “The Remaking of a Unipolar World,” Washington Quarterly 29 (Summer 2006): 7–19. 21. Kenneth Waltz, Man, the State, and War (New York: Columbia University Press, 1959); for further discussion of George W. Bush in these terms, see Robert Jervis, American Foreign Policy in a New Era (New York: Routledge, 2005), chapter 4. 22. George Jahn, “Iran: Attack Fears Spurred Nuclear Block,” Associated Press, March 30, 2007, http://www.truthout.org / docs_2006 / 033007R.shtml. 23. Norumitsu Onishi, “North Korea’s Leader Says He’s Ready to Resume Talks to End Nuclear Standoff,” New York Times, June 18, 2005, 5. 24. Borzou Daragahi and Ramin Mostaghim, “US Strategy on Iran May Have Backfired,” Los Angeles Times, April 3, 2007. 25. For the record of American attempts at coercion after the Cold War, see Robert Art and Patrick Cronin, eds., The United States and Coercive Diplomacy (Washington, DC: United States Institute of Peace Press, 2003). 26. Michael Williams, “What Is the National Interest? The Neoconservative Challenge in IR Theory,” European Journal of International Relations 11 (September 2005): 307–37.

deterrence, rogue states, and the u.s. policy

157

27. Jervis, American Foreign Policy in a New Era, chapter 3. 28. “Tehran’s Hostages,” Wall Street Journal, March 26, 2007. Note the bizarre claim that a few nuclear weapons would make Tehran the military equal of the United States. Also see Dan Senor, “The Long Arm of Iran,” Wall Street Journal, September 29, 2007. 29. See, for example, Benjamin Fischer, “The Soviet-American War Scare in the 1980s,” International Journal of Intelligence and CounterIntelligence 19 (Fall 2006): 480–518. 30. Kevin Woods et al., “Iraqi Perspectives Project: A View of Operation Iraqi Freedom from Saddam’s Senior Leadership” ( Joint Center for Operational Analysis, U.S. Joint Forces Command, Norfolk, VA, 2006). 31. Jervis, American Foreign Policy in a New Era, 125–30. 32. Robert Litwak, Regime Change: U.S. Strategy through the Prism of 9 / 11 (Baltimore: Johns Hopkins University Press, 2007). 33. For neoconservative criticisms of the agreement, see Claudia Rosett, “Deal with the Devil: Gadhafi Is Still a Tyrant, Not a Statesman,” Wall Street Journal, December 31, 2003, and Thomas Donnelly and Vance Serchuk, “Beware the ‘Libyan Model’,” National Security Outlook, American Enterprise Institute for Public Policy (March 2004), 1–8, http://www.aei.org / publications / pubID.20016 / pub_detail.asp. 34. Silverstone, Preventive War. 35. Indeed, democracies seem particularly adept at picking fights with much weaker adversaries: Dan Reiter and Allan Stam, Democracies at War (Princeton: Princeton University Press, 2002). 36. Winston Churchill, My Early Life (London: Cooper, 1989), 246. 37. See, for example, Wallace Theis, When Governments Collide: Coercion and Democracy in the Vietnam Conflict, 1964–1968 (Berkeley and Los Angeles: University of California Press, 1980). 38. Ralph White, “Why Aggressors Lose,” Political Psychology 11 ( June 1990): 227–42; Bernard Finel and Kristen Lord, eds., Power and Conflict in the Age of Transparency (New York: Palgrave, 2000); for the argument that the stance of the opposition in democracies makes it harder for them to bluff and easier for them to make credible threats, see Kenneth Schultz, Democracy and Coercive Diplomacy (New York: Cambridge University Press, 2001). 39. Woods et al., “Iraqi Perspectives Project.” 40. Alexander George, Bridging the Gap: Theory and Practice in Foreign Policy (Washington, DC: United States Institute of Peace Press, 1993).

chapter seven R

Collective-Actor Deterrence patrick m. morgan

th e theme of this volume is the complexity of deterrence today. Here we explore the complexity inherent in collective-actor deterrence, deterrence as employed by actors like the United Nations (UN) Security Council, the North Atlantic Treaty Organization (NATO), or the Organization of American States (OAS). As characterized in theory or as performed in practice, deterrence involving collective actors—as deterrers or targets of deterrence efforts—is a more elaborate endeavor than deterrence in a bilateral relationship. The fact that it is now common makes it necessary to try to better understand its complexities and their implications. Partly they are associated with the altered international context that has been evolving since the cold war, which is only briefly discussed here for lack of space. Partly it is related to the very different nature of a collective actor compared with the actors deterrence theory has envisioned. Application of the theory must be adjusted accordingly. Deterrence among states is a natural recourse for actors imbued with the use of force; it is an unnatural recourse for collective actors established to minimize the role of force and threats in international politics. The study of deterrence is traditionally associated with national security policy in conflicts between states. However, there has recently been a sharp decline in interstate wars, and many analysts link this to a decline in the intense interstate political conflicts typically associated with the practice of deterrence. Various developments tend to confirm this. For instance, a standard action taken with potential wars in mind is creating or joining an alliance focused on specific actual or potential threats; alliances are of less salience today, suggesting that threat perceptions that generate alliances 158

collective-actor deterrence

159

are less widespread. A central feature of the cold war was substantial military forces, maintained in anticipation of occasions for attack or defense, in various dyads and groups of states. Many of these force concentrations have disappeared or sharply declined. The cold war’s enormous nuclear arsenals, with many weapons on high alert and aimed at preselected targets, have shrunk markedly, with most weapons no longer on alert and specifically targeted. Most wars now occur within states. Major occasions for practicing deterrence feature actors like NATO, the UN Security Council, the African Union, and the OAS or venues like the Six-Party Talks on North Korea’s nuclear-weapons program. These are “collective actors” not just in that there is a group involved but because, unlike a traditional alliance, the group is meant to uphold the common good of the global or a regional international system, especially with respect to security. While there is disagreement about how to characterize what has been happening, a good working assumption is that there is an emerging international community or, more accurately, an emerging global community and smaller international communities or systems centered in geographical regions at varying levels of development. Conflict among the major states at the global level is quite low by historical standards and seems likely to remain so for some time, perhaps indefinitely.1 In Europe, community among states is at an historic high, and in East Asia prospects that interstate conflicts will lead to war are low, with one possible exception (the China–United States– Taiwan tangle). They remain similarly low in Latin America and Southeast Asia. The most dangerous interstate conflicts today are in South Asia, the Middle East, Central Asia, and East and Northeast Africa. It appears that the global community and several regional communities are developing significant norms to which members are expected to adhere and that are more honored than not. Factors that may be responsible include the enhanced relative power and global role of the Western world and therefore of Western values, the rise of globalization and increased interdependence, the relaxed security situation in many places, and the increasingly stable nature of many states.2 A natural outgrowth of all this is rising international systems management through intergovernmental organizations (IGOs), transnational organizations, and nongovernmental organizations (NGOs)—globally and in a number of regional systems. However, there will continue to be states and societies that are reluctant to accept and adhere to the existing and emerging norms and that will find it difficult to participate acceptably in system

160

patrick m. morgan

management. They will be regarded by the dominant members of the relevant international communities as security threats to those communities and their members. Various nonstate, internationally operating actors will behave and be treated in the same way. There is also a rising willingness to treat internal conditions in certain states and societies as, in the UN’s terms, “threats to international peace and security,” with strong pressures to curb or eliminate those threats. Under these conditions, international security management, like its domestic counterpart, needs deterrence, compellence, and enforcement capabilities. Without them, over the long run governance is absurd. However, in comparison with the range and scale of the security problems, and judging by the record in specific cases, capabilities that involve deterrence or compellence are often inadequate. They have emerged on an ad hoc, modestly evolutionary basis in response to the pressure of developments, in a context of governments’ reluctance to generate detailed rules, institutions, charters, and legal frameworks that would solidify and elaborate any more or less permanent security management. Instead, members of international communities have been feeling their way toward what is acceptable. This suggests there is a need to explore potential refinements in collective-actor applications of deterrence and compellence. Here it is useful to recall the evolution of cold war deterrence. Security concerns dominated, and deterrence and compellence (including arms control) dominated the pursuit of security. Deterrence was initially perceived as needed primarily to keep conflicts from escalating,3 ironically by confronting opponents with blatant threats of harm through escalation. Alexander George and Richard Smoke noted that by itself this did nothing to end a conflict. It was meant to frustrate the other side’s desire to resolve a conflict via an attack, and they suggested that frustrating an opponent was unlikely to make him or her go away.4 And deterrence, nuclear deterrence in particular, was clearly dangerous. Over time, however, it was adjusted with such concerns in mind. Parties in a deterrence relationship were depicted as locked into not only a conflict but an interdependent insecurity, which required better coordination of their decision making to avoid a joint disaster. This led to steps to contain the cold war’s intensity, restrain destabilizing weapons and deployments, curb destabilizing behavior of allies or others, and prohibit weapons of mass destruction (WMD) proliferation. Various bridges were built across the cold war divide for this purpose, with negotiations that helped shrink misperceptions, detect areas of compromise and cooperation, and

collective-actor deterrence

161

enhance joint understanding of the nature and implications of “common security.” Since the cold war, less emphasis has been placed on deterrence. Deterrence itself now seems less threatening. The George W. Bush administration’s expressed preference has been to make deterrence less central in American defense policy because the opponents of most concern today are unpromising targets for deterrent and compellent threats. There is also increased security management via multilateral institutions that primarily promote negotiation, mediation, conflict resolution, peacekeeping, and so on, while deemphasizing threats and force. Nevertheless, these institutions often find it necessary to practice deterrence or compellence in seeking to halt conflict escalation before it generates fighting or to halt fighting, prevent WMD proliferation, or end major human rights abuses. Some enforcement has been necessary as well, leading to numerous military interventions to pose threats or as repression.5 Little attention is paid to deterrence and compellence, however, as necessary components of international governance by collective actors. While there has been a revival in interest in compellence,6 and significant debate about the contemporary utility of deterrence,7 little of this focuses on collective-actor security management.8 And the burgeoning literature on that management,9 particularly through forceful interventions, makes little reference to deterrence and compellence.

some preliminary matters While the concept of security has recently been expanded to encompass economic, environmental, health, and other threats, this chapter is focused on security from deliberate threats or actual physical harm by domestic or international actors. While an important component of this involves interstate relations, nonstate actors have become increasingly important, so that they now can deliberately, both domestically and internationally, inflict serious harm on states and societies. Although deterrence and compellence differ conceptually in important ways, in practice they are often used simultaneously or in close association, and they often overlap considerably in the perceptions of conflicting parties. As a result, it is best to treat them as linked components of a broader coercive diplomacy, attending to their differences when this facilitates understanding. While reference is made constantly in the paper to the international “community,” global or regional, the other point of departure is appreciation of

162

patrick m. morgan

the continued importance autonomy and sovereignty have for governments and for some nongovernmental actors. The desire to maintain and increase autonomy is taken for granted, assumed to be constantly in play and always highly relevant to states, leaders, and national populations, as well as to groups seeking greater autonomy or their own states. Sovereignty is essentially a legal and political status. Autonomy is a more complex and, in various ways, more powerful factor. Basically it is freedom to decide and act independently in pursuit of one’s goals. Its varied manifestations are obvious in domestic affairs and on many matters in international relations. However, in the most important areas of international relations, an actor’s decisions and actions and the outcomes of interactions are significantly constrained by what others say and do. Others’ behavior may thwart the purposes of an actor’s decision and action or so escalate their burdens and costs as to cancel their utility or make life so difficult in other ways and on other matters that the decision and action are altered, abandoned, or regretted if pursued. Therefore, the exercise of autonomy in international matters inevitably involves compromising it.10 Thus autonomy for states may be mainly a matter of political utility and psychic comfort. Being able to decide for oneself means living with constraints but also having the capacity to disrupt, constrain, or burden others’ decisions and actions, and that constitutes politically useful bargaining power. Furthermore, autonomy is so appealing that it makes an important psychic contribution to sustaining any community, particularly a national community. Upholding autonomy is therefore politically useful, a compelling factor. Autonomy is so attractive that security management for an international community faces huge problems and complications from the tension between the two goals.11

i n i t i a l a n a lys i s Deterrence and compellence involve threats, backed by possible enforcement, to produce desired behavior. In promising or inflicting harm, they are complex, difficult, and controversial. The theory and application of these practices have been guided by images of sharp confrontations between parties that find more moderate ways of conducting their relationship that is unworkable on some matter. While deterrence and compellence and their enforcement are never fully successful, one sign of a working community is their presence in upholding the community with reasonable success. What is “reasonable?” First, members are satisfied that the capabilities exist and

collective-actor deterrence

163

are normally effective. Second, members therefore forego heavy reliance on their own capabilities to perform major security functions, ceding this responsibility to the community. Third, there is a high level of norm compliance even among those opposed to the community’s norms; no security management can cope with more than a modest but insistent noncompliance, particularly if that extends to violent resistance. Collective-actor deterrence has some unique features. In its classic conception and modeling, deterrence deals with severe conflict where there is little overarching community to restrain it and little or no sense of community between the antagonists. Collective-actor deterrence, by contrast, expresses the existence of a community, however limited, in trying to uphold and further inculcate community norms. This purpose and mission for deterrence was only modestly developed in the cold war, for example. While cold war superpower and bloc deterrence evolved into a limited system management, collective-actor deterrence is an established part of community governance. Collective-actor deterrence can therefore relate to autonomy as well, in significantly different ways. Unlike in a dyadic conflict, this community deterrence can itself be a security threat to the parties by shrinking some of their autonomy. It is hegemonic in character. On the other hand, it raises fewer prospects and fears of being employed for political blackmail. It can be undertaken to uphold the autonomy, in principle, of all community members (as in the ouster of Iraq from Kuwait), including the target’s. It may be a less selfish, less self-aggrandizing intrusion on autonomy than a state-to-state deterrence interaction. There are five organizational options for collective-actor security management:12 the international-hegemony model, the coalition-of-the-willing model, the concert model, the militia model, and the state model. In the international-hegemony model, a dominant state or coalition provides the main coercive capabilities. Because those capabilities are highly concentrated, decision-making responsibilities are also held tightly by one government or a few dominant members of the coalition. In the coalition-of-the-willing model, a subset of the community generates the coercive capabilities temporarily, acting on their own or with community authorization. Its operations might resemble a posse, which comes into being to respond to some immediate need for prompt action. In the concert model, a small number of dominant states constitute a coalition intended to manage international security, with membership based on their unusually large coercive capabilities. The members make decisions by consensus. In the militia model, all

164

patrick m. morgan

members are committed to contribute to the coercive resources when necessary, although only a few would typically take much action. A decisionmaking body operates to identify when contributions are needed from the members of the coalition. In the state model, a nonstate actor is permanently or temporarily created that resembles a state in the sense that it operates a central coercive capability. Alternatively, a state is empowered to provide this capability temporarily. Thus it stands somewhat apart from the members in its organization and operation and is utilized for the general welfare. With some variations, these are the only models available in international politics for collective-actor deterrence that reflect a liberal orientation toward the general welfare. The last one is largely hypothetical. Two others, international hegemony and the coalition of the willing, may not always actually exist. A hegemon is unlikely to allow a collective actor to play a significant role in responding to some dispute. The coalition of the willing also could be a phony collective actor—forming around a dominant actor mainly as an act of bandwagoning. The models are ideal types, but elements of each are present in contemporary international politics. The hegemon model was evident in the U.S. intervention in Haiti during the Clinton administration, while a coalition of the willing appeared in the second Iraq war and subsequent insurgency. It has been a struggle, but the concert model has taken hold in the Six-Party Talks on North Korean nuclear weapons. The militia model is best represented in today’s NATO interventions, in which decisions require consensus and members participate in many different ways depending on their resources. While the state model does not currently exist, it has often been suggested that the UN should have its own military units for small military interventions, and the burdens of repeatedly mobilizing military resources as needed may eventually make governments more amenable to the idea.13 Functional equivalents of it include the way the World Trade Organization can declare states in violation of international trade agreements and legitimize punitive responses by other governments and the way the UN commands for the Korean War and the gulf war actually operated. The ultimate goal of collective-actor deterrence is to make such effective responses to violations of community norms that eventually few violations occur or infringements quickly cease when community scrutiny and pressure is brought to bear. Even more attractive would be inculcation of norms, in part via deterrence, so violations become rare and coercive capabilities become mostly a recessed general deterrence as enforcement is rarely needed.14

collective-actor deterrence

165

Ideally, instilling adherence to the norms and the minimal use of threats and enforcement are both achieved.

ideal versus reality With the prospect of numerous security threats, some involving governments and societies of considerable power and size, the ideal for purposes of deterrence is a substantial coercive capability—ranging from nonmilitary pressure to major military force—under careful control of the community members. Some states, however, view almost any coercion on behalf of the international community as a dangerous precedent that threatens sovereignty and autonomy.15 Finding arrangements to suit all or most members of various collective actors is probably impossible. Collective actors also require arrangements that keep their capability from being placed mainly in the service of particular members. There is always a possibility that collective-actor coercion can be used mainly for a government’s or a group’s own selfish purposes. This is a standard complaint of the targets in recent UN or NATO deterrence and enforcement efforts— those efforts are said to be really serving the power and interests of the United States or its allies. A collective actor’s capabilities should also be sufficiently potent that members are comfortable with letting their own coercive capabilities shrink, except for contributing to a collective deterrence or enforcement effort. States should opt to maintain some forces as a hedge against weak or failed collective-actor security management, but the collective actor should be so successful that members feel little need to retain sole responsibility for their security. The present international communities have substantial collective-actor deterrence and enforcement capabilities when they are tapped. What actually results from these efforts, however, is far from ideal. First, for certain states, security management involves traditional nuclear deterrence, and for some others, security is derived in part from U.S. extended nuclear deterrence. Nuclear deterrence has a significant collective-actor component: U.S. leadership of NATO and the British, French, and U.S. nuclear guarantee that it provides. In terms of impending threats, this is not very important now, but it is not irrelevant to European security concerns. More pertinent is the contribution that extended nuclear and conventional deterrence makes to containing nuclear proliferation, plus what it also contributes (debate on this continues) to intensifying pressures for proliferation to occur.

166

patrick m. morgan

Second, this substantial community capability is concentrated heavily in the United States. Where only rather limited coercion is needed this does not matter, but if substantial forces are necessary, the United States supplies a large share and is vital in enabling its coalition partners to operate via its communications, logistics, and intelligence resources. This U.S. role as chief coalition architect is likely to continue into the future. The United States now has accumulated more combat experience than virtually any other state, spends as much as the rest of the world combined on military research and development, and continues spending as much on its armed forces as the other major states combined. Third, availability of appropriate military capability is skewed because most national forces are typically designed for local defense. They are not power-projection, often not really war-fighting, forces. As a result, the United States has been uncomfortable joining in multilateral operations that will result in significant combat, particularly non-NATO operations. It is simply not worth the effort to operate with forces that do not meet NATO standards. Only in follow-on operations after a war, and various war-suppression activities, are the necessary capabilities widely distributed, partly because for decades U.S. forces deliberately avoided those operations, but mostly because peace operations are the preferred military contributions to community security for many governments. It has been suggested that the United States should create a separate service for these operations, but thus far the Pentagon has settled for increased counterinsurgency, peacekeeping, and related training for its regular forces.16 The former policy of avoiding, not even preparing for, smaller military operations is now, at least temporarily, suspended. In general, this partial division of labor is beneficial: the Americans do the heavy combat, and the Europeans and others are better at the peacekeeping, peace enforcement, and postwar security efforts. Problems arise when the United States is unwilling to do serious fighting or, as in Iraq, opposition to a war incites unwillingness to help deal with the aftermath. It is also unclear whether the U.S. public will continue supporting combat operations for international security purposes without serious participation by others. Over the past decade, the United States has displayed discomfort with putting its forces at the disposal of collective actors to be used only with the collective actor’s approval. It has preferred either the hegemonic or the coalition-of-the-willing models and has been ready to evade or abandon collective-actor deterrence in selected cases. At times it has claimed it will act unilaterally if necessary on behalf of its national security and the best

collective-actor deterrence

167

interests of others without approval by a relevant international community. Thus, in the Iraq war, its coalition of the willing was an extension of American power and policy, not a true collective actor—the community coercion capabilities were not under community direction. There is a gap between real and ideal when the United States shows little confidence in collective-actor security management and seems comfortable in Europe only with a collective-actor arrangement it can dominate. In general, it is not moving toward greater reliance on collective-actor deterrence and enforcement for security either for itself or its friends. This is not necessarily permanent. The United States gradually shifted into supporting an ad hoc collective-actor effort to halt and reverse the North Korean nuclear-weapons program that used negotiations, sanctions, and threats. Similarly, the United States has gradually moved more in line with its European allies’ emphasis on negotiations and incentives in dealing with Iran, even though it remains more hawkish than they are. Also, American public opinion strongly supports security management via multilateral endeavors. Thus, it is possible that hegemony-oriented, unilateralist U.S. behavior will fade because it is unsustainable in view of the difficulties in Iraq and Afghanistan, the overstretch of American forces, and eroded public support. If the United States had followed its quick initial victories in Iraq and Afghanistan with successful efforts to impose peace and nation building, however, there might now be much international bandwagoning behind its leadership on global and regional security problems. Others might have sought to contribute in order to obtain a greater say in operations and would have been able to do so with strong domestic support for those and other future interventions. Another problem is the tension between community security management and autonomy. The use of deterrence or compellence by states is typically meant to protect autonomy. In contrast, collective-actor deterrence somewhat compromises national autonomy. The international coercion of a state erodes its autonomy—limiting its freedom of action or, as with Iraq in the gulf war, depriving it of the fruits of a prior autonomous action. It strikes at a state’s autonomy when the community forcibly intervenes, or threatens to, in a domestic situation considered a threat to international peace and security. If the intervention produces regime change or extensive efforts at social recovery and nation building, the affront to the state’s autonomy is enormous. This is also true when force is threatened or used to keep states from enlarging their ability to deter future intrusions, as in efforts to halt nuclear proliferation.

168

patrick m. morgan

Such encroachments on autonomy should be least problematic in a militia model, in which the necessary forces are temporarily supplied by members, who participate in, and perhaps even have a veto over, decisions on how this temporary force is used, the length and scale of their contributions, or the specific ways their forces are employed. This confines the intrusion on autonomy to the target state and its society, and members strongly committed to autonomy will resist carrying that encroachment very far. A coalition of the willing might have a similarly restrained impact, not readily setting disturbing precedents on autonomy. The hegemony, concert, and state models have far more serious implications for national autonomy. Since autonomy is central to national security, trying to enhance international system security by intrusive international coercion generates a security dilemma—the pursuit of security in this fashion at the system and state levels can enhance insecurity at each level. There are various suggestions on how to suppress this security dilemma. One is to reconceptualize sovereignty as ultimately residing in citizens not in their state and requiring that it be exercised by leaders so as to cherish citizen well-being. Sovereignty is a right and a responsibility.17 Hence, intervention against an inept, corrupt, or vicious regime is actually restoring sovereignty. The same could be said of autonomy: an intervention for regime change or peace building can greatly enhance autonomy for citizens by implanting democracy, respect for civil rights, gender and ethnic equality, or a higher standard of living. Nonetheless, international intervention is radical surgery on traditional international politics. Another solution to this security dilemma is to assert that since security is increasingly interdependent, threats to the community are thus perceived as threats to national security and that trends in this direction will likely accelerate. As in economic and other matters, national efforts to achieve security are inadequate, the pursuit of common security being more efficient and effective. As the experience of the European Union indicates, however, in security matters autonomy remains cherished, with powerful implications for the design and operation of system-level deterrence, compellence, and enforcement.

current security management In response to many significant security problems, there are numerous signs of emerging international communities. One sign is that outright defections from certain important norms are rare. In nonproliferation, the number of

collective-actor deterrence

169

nuclear-armed states is far below those with the resources needed to build weapons, while the total number of nuclear weapons and the number deployed have been sharply reduced. On the norm against interstate warfare, such fighting has become rare. Likewise, state sponsorship of terrorism appears to have dropped significantly. While not consistently leading to action, the norm on opposing and acting to halt humanitarian disasters now leads to interventions more often. Collective-actor interventions for the general welfare, or authorized for implementation by others, are at a high level in number of cases, number of troops involved, and costs being borne.18 Equally impressive is the scale of the ambitions. In various cases, elaborate attempts are under way to recast political systems and societies, promote nation building, and create competent states in an effort to strike at the roots of political violence. Nevertheless, there are extensive problems and deficiencies in current collective-actor deterrence for security management. Effective deterrence, for example, requires that the deterrer clearly define the unacceptable behavior, communicate a commitment to punish violations, have the means to deny the challenger his or her objectives or punish him or her beyond their value, and demonstrate resolve to do so. These requirements are often not met by collective actors. There also is an underlying fundamental cleavage among states about using international community action to deal with security problems, particularly in violating national autonomy.19 There is disagreement on the importance of sovereignty and on how seriously to take violations of human rights, on how damaging civil wars really are, and on how to deal with a government’s abuse of its citizens. There are clashes in deciding on interventions and over whether community coercion is being misused, including complaints that such actions really just reflect the international power structure and are therefore illegitimate. Numerous conflicting conceptions of which national interests are at stake inhibit necessary cooperation. Actors disagree on what features of international politics should be preserved (by deterrence) or changed (by compellence) as well as over the degree and nature of any changes in terms of their own interests. They do not agree on whether, in a particular case, deterrence or compellence belongs high on the list of resources to be used, should be a fallback when other steps fail, or should be a very last resort. There is also an insufficient commitment of some members to the continued development of the relevant international community itself; they are willing to see the community placed in jeopardy by a failure to act rather than support or

170

patrick m. morgan

acquiesce in the use of threats or coercion on a security matter. This is serious. If disagreements on specific cases are not to disrupt the international community and its security management, members must at least see upholding the community as in their interest. Another issue is that in refined versions of deterrence theory, as well as in negotiation theory, the starting point for the deterrer is identifying the specific target—who it is that must be persuaded. This is not just a detection problem; a collective actor can have difficulty in agreeing on whom to target and over shaping the threats accordingly. The difficulty is compounded in dealing with nonstate actors inside a society, which are frequently amorphous in terms of command and decision structures. And international terrorist groups make quite shadowy targets for threats. To enhance the chances of success, deterrence should be pursued early in a developing conflict, before the target’s position has hardened and it has considerable momentum behind its intentions. Collective-actor deterrence, however, is rarely timely.20 As Paul Diehl notes, the relevant organizations are “notoriously crisis-driven.”21 Sometimes even acknowledging the existence of the problem lacks a consensus; it is difficult placing a problem on the international agenda if some governments prefer not to look into it (this seemed to be the case with Rwanda). There is normally delay in deciding what to do because members disagree about the nature of the problem, its urgency, and what should be done first. There is a significant literature on the need for preventive diplomacy, starting with preventive diplomatic action and perhaps eventually a forceful intervention before serious damage occurs, the conflict has deepened, and perhaps becomes intractable,22 but it has regularly been ignored. Additionally, collective-actor deterrence often falls short on clearly conveying a commitment to punish violations and the resolve to uphold that commitment. Member disagreements will send conflicting messages, tend to push messages away from clear threats toward ambiguous statements of “concern,” and invite suspicions in the targets about the resolve behind any threats. A collective actor normally does not treat deterrence as its first priority (NATO is a possible exception). Coercion is typically considered a last resort, to be employed in a gingerly way. Even after a decision to act, members are apt to emphasize noncoercive steps like fact-finding, negotiations, and mediation. In contrast, individual governments often pursue deterrence or compellence much more readily because they consider other measures unlikely to work or they dislike what those measures require— patience, compromises, offering incentives, being conciliatory.

collective-actor deterrence

171

Hence, collective-actor deterrence often looks weak and comes too late. When it fails and force is necessary, this frequently comes only after situations have significantly deteriorated. The culprits include a reluctance to interfere, to bear the costs, and to use force, plus the difficulty in creating sufficient political support for taking action in IGOs, concerts, governments, ruling circles, and publics. As a result, challengers are more apt to take confrontations to the brink, such as by ignoring various “red lines” or sets of sanctions, which makes chances of a complete failure of deterrence unacceptably high. Deterrence may have to be carried to the point of enforcing threats, both to get the challenger to finally give way and in hopes of sustaining the collective actor’s credibility for the future. Also at stake at this point is future support of members who were strongly interested in taking action early on. A case in point is the NATO operations against the Serbs in Bosnia and then Kosovo when it began to seem clear that not acting or failing to sustain military action might well unravel the alliance. Alas, when enforcement is undertaken, it is often difficult to sustain. Governments are slow to send military contributions but quick to pull them out to curb costs and other burdens and to avoid casualties. Considerable free riding, particularly over time, is typical, in part because collective-actor deterrence is a form of alliance and alliances have intrinsic free-riding problems. This takes many forms: underparticipation, delayed participation, bounded participation, and nonparticipation. Extensive peace operations are therefore daunting to maintain. One result is little consistency across similar cases, with some serious situations badly neglected. Jeff Knopf in his contribution to this volume notes that deterrence is often the compromise choice between hard- and soft-line policies; for collective actors this means choosing between verbal complaints and threat enforcement. Many governments find it difficult to link a collective military security effort to their specific priorities. This has long been true about nuclear proliferation; it is regularly true about human rights depredations. Again, it is difficult to generate consistency in the international response to security problems and thus difficult to establish a strong reputation for action on which to rest threat credibility. Therefore, collective actors experience greater problems with credibility than individual states. Collective-actor deterrence, to most participants most of the time, is an extreme version of extended deterrence since the party to be protected and the target may be of little direct strategic, ethnic, or political interest to the collective actor. Extended deterrence is generally more difficult than deterrence of a direct attack because it is inherently less credible.

172

patrick m. morgan

Those outside the collective actor often believe that its commitments are apt to lack strong domestic approval and are therefore less likely to be upheld. As Timothy Crawford in his chapter in this volume points out, collectiveactor extended deterrence is also complex because it frequently involves a degree of “pivotal deterrence.” This is when the collective actor seeks to deter simultaneously two or more conflicting parties from beginning or continuing to fight. From the collective actor’s point of view, all members or all actors are entitled to its services; it is not fundamentally intended to take sides, so pivotal deterrence is a common pursuit. It also is hard for the collective actor to create the proper conditions for an effective use of force, which requires clear rules of engagement, a clear chain of command, avoiding the members’ interference and micromanagement, and properly trained and equipped forces available for an extended period. Additionally, a collective actor generally requires either cooperation or at least a pluralistic security community among a few of the most powerful states. Since they provide the ultimate community enforcement resources, the backbone of the collective actor’s threats, system management cannot confront them without threatening the community’s future. Thus, the states that make up the collective actor must responsibly avoid harsh and violent conflicts among themselves. For security and system management, a war among some or all of them is the equivalent of a state’s collapse for its society. These problems are particularly disturbing because collective actors are especially in need of highly effective general deterrence. Designed to operate along liberal lines, collective actors necessarily view coercion as compromising their basic mission of sustaining and promoting security without violence. Ideally, collective actors help generate a community in which norms of proper behavior make security challenges quite uncommon. They do this in many ways, but one of the most important is to visibly embody, and sometimes apply, deterrence and compellence effectively. Collective actors therefore have a strong stake in their deterrence and its credibility—they really do not want it challenged since their responses might be inadequate. Nevertheless, their inadequacies help encourage challenges. These kinds of credibility issues are reinforced by the way a collective actor’s security management rests on its perceived ability to provide an alternative to states’ relying on themselves in traditional ways for security. It must appear as a legitimate, highly effective purveyor of peaceful security management, offering effective general deterrence for this purpose. Being slow to react to security problems, being inadequate at tackling them early,

collective-actor deterrence

173

or seeming reluctant to threaten a serious military response to noncooperation are not suitable behavior patterns for solidifying and sustaining deterrence. Even if a collective actor overcomes all the barriers to mounting vigorous verbal threats and upholding them, success may still be hard to obtain. For instance, compellence is harder than deterrence, and many deterrence situations contain major elements of compellence or are perceived by the targets as compellence situations. When collective-actor coercion is late in dealing with a conflict, its threats are likely to be seen by the target as compellence: trying to force it to drop or reverse actions underway for some time or to abruptly initiate new ones. That can readily be taken as a demand to cede something, so the proposed outcome is framed as a loss. According to prospect theory, people and states resist losses more than other kinds of outcomes.23 The political and psychological elements that make compellence harder than deterrence may apply more often in collective-actor coercion than in comparable disputes among states. Studies of coercion find that applying increasing pressure via threats and then gradually escalating the use of force is typically less effective than expected. People, societies, and governments in such situations often ready themselves for the coming harm, making concrete and psychological preparations.24 When the pain finally arrives it is therefore more readily borne and often seems lower than expected. Being targeted also leads people to pull together. The problem for a collective actor is that its coercive efforts typically start by raising pressure and then move to sanctions and then to a limited use of force. This is a pattern unlikely to be successful; because of that, it may be difficult to sustain. The resolve of the target may get stronger while that of the collective actor is subject to increasing strain. If a collective actor has members that are unenthusiastic about a coercive endeavor, leading to extensive debates over even limited attempts at it, this sends signals that the motivation for the coercion is not high or will readily decline in the face of significant costs and burdens, particularly over time. Having this sort of image is always worrisome to a deterrer. The outcome of deterrence and compellence situations is partly shaped by perceptions of the strength of each side’s motivation. Barry Posen notes that in outside interventions on behalf of people inside a country, It is quite likely that the will of the local party, the assailant, is stronger than that of the outside rescuers, because the stakes for the local party are so much greater. Indeed outsiders are unlikely to have many classical “vital” interests

174

patrick m. morgan

at stake in these conflicts. Assailants often begin their depredations because of some deeply held beliefs about the necessity of their actions. . . . It is extremely difficult for the rescuer to convince the assailant that it cares more, particularly when the source of the concern seems to be an erratic and capricious humanitarian impulse, which varies with the extent of the international media coverage: the availability of other dramas in the global village, the skin color, culture or religion of the victims.25

Success would be more likely if the collective actor’s signals suggested it was passionate about the issue, seized by intense desire to rectify the situation, ready to steamroller any opposition. This would make for less delay in clarifying the perceived interests of each side, the scale of what each has at stake, the probable balance of motivation, and each side’s intent. The history of U.S. coercive diplomacy suggests that offering incentives will more likely work after significant harm has been delivered rather than earlier on.26 But a collective actor will normally be under heavy pressures from members to reverse this pattern: it will be encouraged to offer concessions to get negotiations started, inducements to keep them from collapsing, and further concessions to keep from resorting to force. That is, the incentives may be offered at the wrong time.27 Recent U.S. experience with collective actors tends to confirm the difficulty of offering incentives at the right moment in negotiations. Analysis of deterrence situations also suggests that movement toward shared perceptions about the parties’ relative intensity of motivation and commitment, the real issues at stake, and the nature of the parties’ interests is important to the deterrence process. Such learning is more likely early on; as the conflict builds and tension escalates, emotional and other factors can make positions rigid, compromise more difficult, and misperceptions more deeply rooted. This puts a premium on initiating deterrence or compellence early, but a collective actor will find it hard to behave accordingly. The best threat a collective actor could offer once a crisis has fully developed would be a highly credible promise that if negotiations fail overwhelming force will be brought to bear quickly in a determined effort to achieve a decisive result. A low-credibility threat, or one that is slow to emerge and carry out and that promises military steps that are likely to drag on perhaps indefinitely, is less likely to succeed. The goal of avoiding the use of force would be damaged by a lack of threat credibility. And the longer it takes to mount the threat the more serious and costly to implement it is likely to be, possibly eroding member support. This provides too

collective-actor deterrence

175

much incentive for the target to delay compliance or refuse to concede in any fighting. Threats to apply a good deal of force decisively are therefore better. But collective actors often get a late start in using force—too late to prevent a serious crime, major steps toward WMD proliferation, or the outbreak of war. The initial threats will be limited; they may then be followed by sanctions and only eventually get to the limited use of force, all the while lacking credibility. Thus collective actors often offer less than ideal threats. It is difficult to get coercive endeavors in general to work.28 It is therefore wise not to move in that direction politically and verbally without a real determination to use force if necessary29; collective actors should operate from the start as if force is likely. That type of position will bolster credibility, avoid unwarranted optimism, and make the eventual use of force easier—the decision to do so has already been made; it is only a matter of reconfirming it. Collective actors, however, never operate this way. Typically, many members want to defer any idea of using force until all other options have been exhausted; often they want no public statements that force might be used. Their concern is precisely that an early decision in principle to use force will make it too easy to slide into doing so later. Collective actors possess certain advantages over states in mounting deterrence or compellence, although these advantages can be hard to exploit. One advantage that must not be squandered is its valid claim to be pursuing the common interest. This confers a legitimacy so potent that simply dismissing its concerns and requests can undermine the target’s position. This helps sustain member cohesion and reinforce target isolation. A collective actor should therefore minimize vulnerability to the charge that it is dominated by a particular member or acting in the interests of a small coalition. Even better is seizing on opportunities to secure support or at least acquiescence from nonmembers, since this adds to the image of pursuing the general welfare. Unfortunately, particularly in coercive efforts, collective actors are usually heavily influenced by their leading members.30 Another advantage is that a collective actor’s threats are less likely to set off the classic stability problem in deterrence. The collective actor is less likely to appear as highly aggressive in intent and thereby incite the target to initiate fighting. Collective actors typically have the opposite problem—their threats are not taken seriously. Also, it is hard to imagine a collective actor attacking with no warning. A surprise attack might be used tactically, but only after carefully signaling an attack was imminent. A collective actor also poses a tough, indistinct target to attack, particularly for counterdeterrence.

176

patrick m. morgan

Is it best to focus on the weak links in the coalition or organization? Or the dominant member? Or the members driving the specific dispute? Or maybe those in politically sensitive positions, such as those holding swing votes? This may reduce the opponent to issuing vague general threats, which are harder to make credible, or to issuing threats just to fight hard when more pointed deterrence threats would be preferable. Even the disadvantages of collective actors can sometimes be of benefit. For instance, if forging a consensus of the members is difficult, a collective actor is likely to avoid asking too much, setting demands the target will never accept. It is unlikely to be so committed to military action as to neglect a last chance for a settlement. And a target can be more comfortable compromising under pressure from a collective actor than from another government. For instance, its fear that a concession will only mean further demands later is apt to be muted. Finally, analysts stress the importance of adding inducements for making deterrence and compellence more effective. With members reluctant to initiate warfare, a collective actor is likely to make an honest effort at using inducements, not just threats. And it may be easier for the hawkish members to agree to this than if they were dealing with the target on their own. In the same way, the collective actor can more credibly deliver necessary assurances that must accompany a deterrent or a compellent threat—assurances that compliance will not be followed by more threats or by application of the punishment anyway. These advantages are crucial to success of the collective actor. A good example of what may happen when they do not apply is the 1950 Chinese attack on UN forces in Korea. Beijing was primed to see the UN forces nearing the Yalu as a grave threat and was not reassured about UN intentions and particularly those of the Truman administration, despite strong U.S. and UN assurances about their limited intentions. This perception of threat was also certainly driven by the fact that Beijing believed that the UN was dominated by the United States. Believing that it might readily be attacked in the future, and with UN forces concentrated to provide a clear target, Chinese leaders launched their own attack. Because it is a somewhat diffuse target, a collective actor also benefits from its members’ different strengths in coercing for the common good. Some members may have unusual contacts with or special access to the target for negotiations, enhancing the collective actor’s ability to communicate credibly both deterrence and reassurance. Contemporary collective actors also have members with specialties in coercive endeavors, offering possibilities for more tailored threats and their implementation.

collective-actor deterrence

177

suggestions One difficulty in efforts to deal with gross deficiencies inside a state, particularly in a civil war, is the limited impact they may seem to have on the security and welfare of the international community. Such situations may not seem important. This has led to proposals that the inhibition on collective action be eased by assigning primary responsibility for such cases to states and societies in the neighborhood.31 In economic terms, ending the problems mainly supplies a regional public good so the greatest incentives to act are located in those regions.32 This is particularly suitable where there is intrinsic concern about interventions by Western countries, which raise memories of colonialism. The threats of a neighborhood collective actor would probably be more credible as well. Neighborhood collective actors are unlikely to emerge, however, in areas in which the neighbors are badly equipped to mount effective interventions. Considerable support is needed from outside the region, and the United States was right to launch an initiative to help African governments develop a regional intervention capability. Even then, an infusion of military forces from states outside the region will be important in cases like Darfur. In the long run, NATO-like organizations for regional communities might be best for neighborhood action that attracts member participation. It has often been said, “One NATO is not enough.”33 The UN Security Council might then subcontract some security management operations to these actors, as could regional political and economic collective actors that want to avoid having direct security responsibilities. An alternative is for the United States to go beyond its traditional alliance partners for interventions with other major regional powers such as China, India, and Brazil.34 It might also be possible to make major payments to members willing to participate in interventions when they have only a limited stake in doing so, with the recipients having to meet high standards with regard to the level of training, professionalism, and discipline of the forces they provide; the speed with which they can be ready; and their willingness to tackle missions ranging from classic peacekeeping to significant fighting. A variation of this plan could take the form of payments to private security forces to have them supplement member contributions. The credibility of collective-actor deterrence might be enhanced by developing a working substitute for having a constantly available force at the collective actor’s disposal, since common proposals for such a force remain unlikely to be realized any time soon.35 Credibility also can be enhanced

178

patrick m. morgan

if steps like this increase the likelihood a situation gets an appropriate response, since research suggests that deterrence and compellence effectiveness is influenced less by the threatened harm than the high likelihood of it. Greater availability of effective responses also would encourage less variation in the treatment of similar cases, meeting the standard complaints that collective-actor intervention varies so much that it cannot effectively instill important norms and principles. One practice worth investigating is the allocation of the major responsibility of designing, organizing, and applying a coercive campaign to a major coalition member on a more or less rotating basis. This leadership arrangement would resemble the rotating presidency of the European Union. More states would then periodically play a dominant role in system security management, helping reinforce the claim that it serves the general good. Rotating leadership would provide a focal point for management but would retain the overall weight of the collective actor. The collective actor could set the broad guidelines and then let a particular member take charge, with eligibility based on having the requisite capacities for conducting negotiations, deciding when being tough is necessary, and organizing and carrying out a coercive operation. Because the chief advantage of a collective actor is its scale, member cohesion is vital. Given member disagreements, this suggests that selective deterrence and enforcement are important—trying to do too much will adversely affect cohesion, effectiveness, and credibility. But the uncertainty of response to provocations can erode effectiveness. Several possible solutions to the problems created by provocation and selective response suggest themselves. First, thought should be given to developing principles for involvement that facilitate prioritizing interventions. The international community has been avoiding developing general rules on handling threats to international peace and security, preferring to meander in an evolutionary way through various responses. It does seem that, based on the experience to date, there could be general agreement that an outright aggressive interstate attack must get a sharp response as should clear state sponsorship of terrorism, that humanitarian disasters merit potent interventions, that regime change due to extreme mismanagement and civil rights concerns should be undertaken much less often, and that nonproliferation interventions lie somewhere in between in importance. Second, there should be a recognition that cases selected for action must offer a high chance of success and that governments should let noninvolvements embarrass the international community in hopes of building pres-

collective-actor deterrence

179

sure for a gradual evolution toward increased readiness to act. According to Martha Finnemore, there is more than strictly self-interest at work found “ . . . in surveying the global pattern of military intervention since the end of the Cold War. First, most of it is multilateral. . . . Second, in many cases the geostrategic or economic interests of the intervening states in the target state are negligible.”36 If other factors are important, their impact is likely to grow over time. Third, collective-actor deterrence should not be attempted unless member commitment to sustaining security management is profound in the case at hand. But who is to judge this? Individual members can do their own assessments, but someone must help inject and try to reflect the general interests of the community, its purposes, and its security responsibilities. Normally, that should be someone who heads, embodies, and stands for the collective actor, but those in charge of such institutions are in a rather weak position to do this. The responsibility typically falls on an extremely potent or greatly respected member or to some group equipped to act on behalf of other members (and thus to lead in shaping the decision). For the sake of collective-actor deterrence and security governance, we should better understand how and why leaders and governments emerge to fill this role. Institutionalizing it, as in the UN Security Council, offers insufficient leadership in many instances, so it must be gained in a more creative and flexible fashion. Fourth, collective actors might make it a practice to offer lead responsibilities for securing success via noncoercive practices in a particular case to governments who oppose direct military involvement and threats of it, which is how the Six-Party Talks on North Korea were run. Make the decision that the collective actor is seized of the issue and thinks it important; then give those who want to avoid coercion the responsibility of resolving it another way. If they succeed, that is all to the good. Their failure could facilitate coercion efforts likely to get the necessary support to work. This is another way of exploiting one of the most powerful forces for cohesion in collective actors. As in politics at other levels, one of the strong motivations here is the desire for a seat at the table, to be part of important discussions and decisions, to have a say. This was a potent factor in shifts in security policies in Europe and Japan after the gulf war and in fostering resentment over the George W. Bush administration’s approach to global security issues after September 11, 2001. It is somewhat underappreciated by analyses that stress traditional conceptions of national interest in anticipating states’ behavior, including behavior as members of collective actors.

180

patrick m. morgan

conclusion This is a preliminary attempt to merge the thinking and experience of collective actors in coercive diplomacy and outright military action with analyses and experience on deterrence and compellence. Stimulating discussion seems most useful at this point. Nevertheless, moving beyond introductions should be an important objective in the next few years, given the security workload of collective actors today.

notes 1. Robert Jervis, American Foreign Policy in a New Era (Milton Park, UK: Routledge, 2005). 2. A good recent discussion on this is Jean-Marc Coicaud, Beyond the National Interest: The Future of UN Peacekeeping and Multilateralism in an Era of U.S. Primacy (Washington, DC: United States Institute of Peace, 2007), 75–109. Linking this development to international coercive efforts is the major theme of Andrea Kathryn Talentino, Military Intervention after the Cold War: The Evolution of Theory and Practice (Athens: Ohio University Press, 2005). 3. Patrick M. Morgan, “Deterrence, Escalation, and Negotiation,” in Escalation and Negotiation in International Conflicts, edited by I. William Zartman and Guy Olivier Faure (Cambridge: Cambridge University Press, 2005), 53–79. 4. Alexander George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974). 5. An excellent overview of peace operations by collective actors, including coercive efforts, is Paul F. Diehl, Peace Operations (Cambridge: Polity Press, 2008). Recent information on violent conflicts worldwide can be found in Human Security Report Project, Human Security Brief 2007 (Human Security Report Project, Simon Fraser University, Vancouver, May 2008). 6. A good example is Lawrence Freedman, ed., Strategic Coercion: Concepts and Cases (Oxford: Oxford University Press, 1998). 7. Examples include Patrick M. Morgan, Deterrence Now (Cambridge: Cambridge University Press, 2003); Timothy W. Crawford, Pivotal Deterrence: Third Party Diplomacy and the Pursuit of Peace (Ithaca: Cornell University Press, 2003); Derek D. Smith, Deterring America: Rogue States and the Pursuit of Weapons of Mass Destruction (Cambridge: Cambridge University Press, 2006); James H. Lebovic, Deterring International Terrorism: US National Security Policy after 9 / 11 (Milton Park, UK: Routledge, 2007); James W. Davis, Jr., Threats and Promises: The Pursuit of International Influence (Baltimore: Johns Hopkins University Press, 2000); Keith B. Payne, Deterrence in the Second Nuclear Age (Lexington, KY: University Press of Kentucky, 1996); Payne, The Fallacies of Cold War Deterrence and a New Direction (Lexington, KY: University Press of Kentucky, 2001); Daryl G. Press, Calculating Credibility: How Leaders Assess Military Threats (Ithaca: Cornell University Press, 2005); and Anne E. Sartori, Deterrence by Diplomacy (Princeton: Princeton University Press, 2005). 8. Crawford’s illuminating analysis of pivotal deterrence, for instance, can readily be adapted to explore certain deterrence /compellence efforts by collective actors (Crawford, this volume). 9. Examples include Joseph Lepgold and Thomas G. Weiss, Collective Conflict Management and Changing World Politics (Albany: State University of New York Press, 1998); Talentino, Military Intervention. 10. Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton: Princeton University Press, 1997).

collective-actor deterrence

181

11. To quote Talentino, “states are reluctant to sanction violations of sovereignty and only acquiesce when the problem is absent . . . or when time has allowed an extreme justification to develop in spite of sovereignty, as in Kosovo.” Talentino, Military Intervention, 281. 12. These models also apply to collective-actor nonmilitary coercion, like sanctions. The hegemon model involves primary sanctions by a dominant state with others making limited contributions at the hegemon’s urging, like the decades of sanctions on Cuba. A coalition of the willing has group sanctions, as in the current Proliferation Security Initiative created by the United States. In a concert, dominant states adopt the sanctions and orchestrate others’ participation, as in sanctions on North Korea. The militia model fits the recent UN sanctions against Sudan. The state model is best represented by how the World Trade Organization can approve states’ applying sanctions against a violator of its rules. 13. See Daniel H. Deudney, Bounding Power: Republican Security Theory from the Polis to the Global Village (Princeton: Princeton University Press, 2007). 14. Lawrence Freedman, Deterrence (Cambridge, UK: Polity Press, 2004). 15. China’s position on North Korea, for instance, has been that it be strenuously urged to not develop nuclear weapons but that, being sovereign, it is entitled to develop them and using force to prevent that is unacceptable. 16. See Thomas P. M. Barnett, The Pentagon’s New Map: War and Peace in the Twenty-First Century (New York: G. P. Putnam’s Sons, 2004); and Robert M. Perito, Where Is the Lone Ranger When We Need Him (Washington, DC: United States Institute of Peace Press, 2004). 17. See Francis Deng et al., Sovereignty as Responsibility (Washington, DC: Brookings Institution, 1996). 18. Coicaud, Beyond the National Interest, 16–18. 19. Coicaud, Beyond the National Interest, 16–18. 20. Macedonia in the 1990s is one example to the contrary. The response to the renewed fighting in East Timor in 2006–7 is another. 21. Diehl, Peace Operations, 19. 22. See I. William Zartman, Cowardly Lions: Missed Opportunities to Prevent Deadly Conflict and State Collapse (Boulder: Lynne Rienner, 2005); David Hamburg, No More Killing Fields: Preventing Deadly Conflict (Lanham, MD: Rowman and Littlefield, 2002); Hamburg, ed., Preventive Diplomacy, Preventive Defense, and Conflict Resolution (New York: Carnegie Commission for the Prevention of Deadly Conflict, 1999); and Barry Steiner, Collective Preventive Diplomacy: A Study in International Collective Management (Albany: State University of New York Press, 2004). 23. On prospect theory, see Barbara Farnham, ed., Avoiding Losses, Taking Risks: Prospect Theory and International Relations (Ann Arbor: University of Michigan Press, 1995); Jack Levy, “Prospect Theory, Rational Choice, and International Relations,” International Studies Quarterly 41, no. 1 (1997): 87–112; and Jonathan Mercer, “Prospect Theory and Political Science,” Annual Review of Political Science 8 (2005): 1–21. 24. See Roger Fisher et al., Coping with International Conflict: A Systematic Approach to Influence in International Negotiation (Upper Saddle River, NJ: Prentice Hall, 1997), 194–208. 25. Barry Posen, “Military Responses to Refugee Disasters,” International Security 21 (Summer 1996): 21. 26. Robert J. Art, “Coercive Diplomacy: What Do We Know?” in The United States and Coercive Diplomacy, edited by Robert J. Art and Patrick M. Cronin (Washington, DC: United States Institute of Peace Press, 2003), 359–420. 27. The argument for this is that threats arouse more resentment and resistance, poison the atmosphere, stimulate greater cohesion in the target, and so on. 28. Paul F. Diehl, International Peacekeeping (Baltimore: Johns Hopkins University Press, 1993). 29. On sanctions, for example see Chantal de Jonge Oudraat, “UN Sanction Regimes and

182

patrick m. morgan

Violent Conflict,” in Turbulent Peace: The Challenges of Managing International Conflict, edited by Chester A Crocker, Fen Osler Hampson, and Pamela Aall (Washington, DC: United States Institute of Peace Press, 2001), 323–51. For this point applied to coercive diplomacy in general see Art, “Coercive Diplomacy,” 402–10. 30. On how this applies to the UN, see Donald J. Puchala, “World Hegemony and the United Nations,” International Studies Review 7 (December 2005): 571–84. 31. Zartman, Cowardly Lions. 32. See Todd Sandler, Global Collective Action (Cambridge: Cambridge University Press, 2004) 33. Joseph R. Nunez, “One NATO Is Not Enough,” International Herald Tribune, January 29, 2007. 34. See Barnett, The Pentagon’s New Map. 35. See Diehl, Peace Operations, 92–8, for a review of such proposals. 36. Martha Finnemore, “Military Intervention,” in Collective Conflict Management and Changing World Politics, edited by Joseph Lepgold and Thomas G. Weiss (Albany: State University of New York Press, 1998), 6.

chapter eight R

Complexity of Deterrence among New Nuclear States The India-Pakistan Case dinshaw mistry

pa k i s ta n and India are among the world’s newest nuclear powers. They developed small and recessed nuclear-weapons capabilities in the 1990s and conducted a series of nuclear tests in May 1998. Their motives for demonstrating their nuclear capabilities varied. New Delhi mainly sought to retain a nuclear “option” versus China and Pakistan and declared that it had “exercised” its option with its nuclear tests. India’s policy makers apparently had no formal deterrence doctrine that integrated nuclear weapons into their defense planning. Pakistan’s policy makers sought nuclear weapons to counter India’s superior conventional forces and its nuclear forces. New Delhi and Islamabad pursued nuclear-weapons programs with only the simplest understanding of nuclear deterrence—that it would help them counter their nuclear-armed neighbors. These simple notions of deterrence gave way to significant complexity in the practice of deterrence in South Asia. This chapter reviews these deterrent relationships and describes how the introduction of nuclear weapons affected the India-Pakistan conflict in multiple ways. The chapter also describes how nuclear weapons produced four effects from 1999 to 2003. First, nuclear deterrence, combined with factors such as a stability-instability paradox, encouraged Pakistan’s military to undertake limited military operations and to support militants against India, and this led to serious military crises in the region. Second, nuclear weapons and conventional military superiority influenced India to move beyond deterrence to a strategy of compellence. Third, by enabling both sides to undertake a variety of military operations that generated serious crises, nuclear weapons helped sustain the 183

184

dinshaw mistry

broader India-Pakistan rivalry. Fourth, nuclear weapons attracted external actors into South Asia’s military crises. This last development had both positive effects (external actors eased the crises) and negative effects (regional powers behaved provocatively to prompt external intervention in their disputes). Overall, a mixed deterrence outcome is seen in Indo-Pakistani relations from 1999 to 2003. Nuclear weapons did not deter Pakistan, a revisionist state dissatisfied with the status quo in Kashmir, from challenging this status quo and did not deter India’s political and military leaders from planning a strong military response in the face of provocations launched by Pakistan. Nuclear weapons did induce some caution in decision making, and dissuaded the parties from quickly escalating their crises to large-scale war. Yet, while they deterred quick escalation, nuclear weapons may not have ultimately deterred the parties from escalating military hostilities. Had these crises not been eased by third-party diplomacy, one or both sides could have undertaken significant military escalation, possibly leading to a larger war. Since 2003, evidence has emerged that nuclear learning has been occurring among policy makers in the subcontinent. Indian and Pakistani leaders learned lessons from their prior crises and adopted three policies. First, both sides took steps to defuse further military crises and undertook a dialogue seeking to resolve the Kashmir issue. Second, the two sides adopted stronger command-and-control mechanisms and developed new nuclear doctrines that mitigated but did not completely eliminate the potential for nuclear use during war. Third, India sought new technologies to overcome some of the limitations of deterrence—technologies that are not purely defensive but that could facilitate offensive operations. These new military technologies and doctrines add to the complexity of deterrence in South Asia. There is both change and continuity in the practice of deterrence among old actors (the superpowers) and new actors (Pakistan and India). Like the old actors, the new actors have undertaken and developed military forces and doctrines for a variety of military operations under their respective nuclear umbrellas. Deterrence relationships between new nuclear powers and their rivals also contain novel elements, especially when these new powers are deterring challenges from nonstate actors who act as proxies for their rivals. Deterrence of such proxy wars can be extremely difficult, leading states to seek deterrence, compellence, and defensive and offensive military options versus their revisionist rivals. This chapter elaborates upon these observations about nuclear developments in South Asia over the last decade. It begins by reviewing the military

complexity of deterrence

185

crises that occurred in South Asia in 1999 and 2001–2. It discusses how factors such as a stability-instability paradox and asymmetric military capabilities, combined with nuclear deterrence, contributed to crises in South Asia. It next reviews nuclear learning in South Asia. It also describes how new doctrines and new technologies add to the complexity of deterrence in the region. It concludes by comparing the India-Pakistan nuclear relationship with the superpower nuclear relationship and by discussing whether the behavior of other new nuclear powers—North Korea, Israel, and a possibly nuclear Iran—resembles the behavior of Pakistan and India.

the 1999 and 2001 – 2 military crises In 1999, India and Pakistan fought an intense localized conflict in the Kargil region of Kashmir.1 The conflict began in May after approximately one thousand Pakistani military soldiers, paramilitary troops, and militants occupied mountain peaks and positions across the line of control in Indian-held Kashmir. The primary Pakistani objective was to draw India into negotiations on Kashmir before ending the intrusion and to generate international involvement in negotiations on Kashmir.2 Its strategy was similar to the asymmetric war strategy adopted by weaker powers against stronger ones, such as the strategy adopted by Egypt when it attacked Israel in 1973.3 The Kargil crisis ended in July 1999. By then, India had recovered much of the territory occupied by Pakistani forces (but after incurring significant casualties, losing over four hundred troops). U.S. diplomatic efforts also began to have an impact on the crisis. In a July 4 meeting between President Bill Clinton and Prime Minister Nawaz Sharif, Pakistan agreed to withdraw its troops from across the line of control. If Washington had not provided an exit strategy and Pakistan’s military had continued to fight in Kargil, India would probably have escalated the crisis, especially if it was taking high casualties. The Indian army had mobilized its main strike and reserve formations and deployed them along the western border with Pakistan. It was prepared for an attack against Pakistan.4 India-Pakistan relations remained tense after Kargil, and militant violence in Kashmir intensified (the number of Indian security personnel killed by militants rose from approximately 340 in 1998 to about 600 each year in 1999, 2000, and 2001). This violence reinforced thinking among India’s civilian security policy makers that Pakistan was raising the level of conflict and that Pakistan’s decision makers believed that their nuclear weapons neutralized India’s conventional military superiority. India’s security

186

dinshaw mistry

planners sought to redress the situation through limited-war doctrines, which called for fighting small wars to retaliate against Pakistani aggression. The 2001–2 crisis was triggered by a December 13, 2001, militant attack on India’s parliament.5 India’s government—under intense media and political pressure to strike Pakistan—allowed the army to mobilize fully but sought to explore nonmilitary options before resorting to force. New Delhi demanded that Islamabad crack down on the militant groups linked to the parliamentary attack. India’s military mobilization was also a way to force the United States to push Pakistan toward accepting Indian objectives. Further, it was a preparation for quick military action if Pakistan failed to meet India’s demands. India came close to launching military operations on two occasions—in January 2002 and in May–June 2002. U.S. diplomacy eventually eased the 2001–2 crisis. Nevertheless, India had mobilized over 600,000 troops and could have undertaken significant military operations against Pakistan if the crisis had not been eased. India’s military plan was to seize Pakistani territory in the Punjab and Sindh-Rajasthan sectors, to draw Pakistani forces into combat and attrite them, and to trade Pakistani territory for concessions in ensuing negotiations.6 India’s military would have clashed with a fully mobilized Pakistani army, resulting in a major conflagration in the subcontinent. Three general observations can be made about these South Asian military crises. First, the crises did not escalate to a major war. Nuclear deterrence induced caution among security planners on both sides and was one factor that checked them from quickly escalating to large-scale military operations, although conventional deterrence and international diplomacy also contributed to this military restraint. Some scholars became more confident about South Asian crisis stability in the aftermath of the confrontation. They noted that purposeful war was unlikely in South Asia, even in the face of extremely provocative terrorist attacks or cross-border incursions, and that an Indian or Pakistani leader would not order a movement of ground forces across the international border in the face of possible nuclear retaliation.7 Second, contrary to the above perspective, if each crisis had not eased at key moments, both sides could have eventually escalated the encounter by ordering more than a strike corps across the international border (In 2002, India had mobilized some 600,000 troops and had all its three strike corps in place for a military invasion of Pakistan.) In 1999 and 2002, India’s political and military leaders were not going to be ultimately deterred from an attack on Pakistan.8 Rather, an Indian attack was averted primarily because

complexity of deterrence

187

some of India’s objectives were attained and the crises ended without the need for India to expand military operations. Third, during any escalation, India’s military planning called for keeping an attack limited by not crossing Pakistan’s “red lines,” which could cause Pakistan to escalate the conflict. According to Paul Kapur, “an Indian attack on Pakistan would have been highly circumscribed so as to avoid threatening Pakistan with catastrophic defeat.”9 Despite India’s intentions of keeping any attack highly circumscribed, Pakistan was likely to misperceive India’s intentions and respond forcefully to any Indian military action, causing India to widen hostilities beyond a localized encounter. In sum, from 1999 to 2003 Pakistan and India undertook and were prepared to embark on a variety of military operations under the nuclear umbrella. Their crises resulted in a mixed deterrence outcome.10 Some military actions, in this case quick escalation to large-scale war, were deterred, but other military operations, limited aggression and preparations for larger escalation, were not deterred.

nuclear deterrence, military operations, and conflict in south asia Nuclear deterrence had a number of effects on conflict in South Asia. First, nuclear deterrence failed to prevent low-intensity conflict in South Asia. Instead, it created a stability-instability paradox that facilitated conflict in the region. Second, nuclear weapons negatively impacted the broader IndiaPakistan rivalry. Third, nuclear capabilities influenced India to move beyond deterrence to a strategy of compellence. Fourth, nuclear weapons were instrumental in drawing external actors into South Asia’s military crises, and outside intervention shaped the conflict in positive and negative ways. Military Crises Influenced by Territorial Conflict and Asymmetry Nuclear weapons, combined with a stability-instability paradox, military asymmetry, the territorial nature of the conflict in South Asia, and an insurgency in Kashmir, contributed to the 1999 and 2002 military crises in South Asia. In general, a stability-instability paradox is manifest in South Asia because nuclear weapons deter large-scale war, but this tends to set the stage for low-level conflict.11 One side is emboldened to undertake military operations with the assumption that its nuclear weapons will deter retaliation by the other side. Under these circumstances, conflict can expand to a

188

dinshaw mistry

large-scale conventional war if the opponent responds more forcefully than anticipated.12 The stability-instability paradox influenced Pakistani decision makers to raise the level of conflict with India by launching the Kargil operation and by supporting militant attacks in Kashmir. Pakistani strategists assumed that nuclear weapons would deter an Indian military response. Pakistani security elites believed that their small and “recessed” nuclear force deterred Indian military attacks against Pakistan on three previous occasions—in the mid-1980s, 1987, and 1990—and this belief increased their confidence that they could deter future Indian military action.13 Without a nuclear deterrent, Pakistan’s decision makers would have been more cautious about raising the level of conflict with India, because India could have retaliated with its stronger conventional forces and defeated Pakistan’s military. Pakistan’s nuclear deterrent changed this situation, because its decision makers then assumed that their nuclear forces would deter such an Indian conventional attack, and this emboldened them to raise the conflict with India. The asymmetric capabilities of the two states and their position toward the status quo further influenced Indian and Pakistani military strategies. The weaker side, Pakistan, challenged India in only a limited way but still inflicted high costs on India. Kargil resulted in over four hundred Indian military fatalities, while militant attacks in Kashmir have taken thousands of lives. Yet Pakistan had a strong motivation to challenge India because it was highly dissatisfied with the territorial status quo in Kashmir, and despite the asymmetry in the two sides’ conventional capabilities, Pakistan possessed sufficient capability to challenge India.14 Insurgency in Kashmir also provided Pakistan with an opportunity to challenge India. Pakistan could claim that it had little involvement in attacks undertaken by insurgents. Pakistan thus challenged India over many years, tying down several hundred thousand Indian army troops in Kashmir, with the militancy resulting in tens of thousands of fatalities. India’s security planners and domestic polity grew increasingly weary of this sustained Pakistani proxy war against India. In this context, aggressive Pakistani behavior led to forceful Indian responses, which aggravated the conflict. The Indian response was influenced by military superiority. In both 1999 and 2001–2, India threatened a conventional war primarily because of its overall military superiority vis-à-vis Pakistan. Indian planners adopted conventional war strategies with the assumption that they could fight and win a conventional war without prompting escalation to the nuclear level. India’s military also was prepared to counter Pakistani conventional or nuclear escalation, and it used its superior forces

complexity of deterrence

189

to deter Pakistan from launching a full-scale conventional war or a nuclear strike. Thus, during Kargil, one motivation behind India’s conventional force mobilization was to deter Pakistan from widening hostilities outside the Kargil sector. During 2001–2, India issued nuclear warnings three times to establish dominance in any situation in which Pakistan might be tempted to escalate the conflict, thereby warning Pakistani strategists that any use of nuclear weapons would be matched by an Indian nuclear response. Finally, the combination of nuclear capability and conventional superiority not only allowed India to threaten conventional war against Pakistan but also gave it the capability to deter Pakistan from low-intensity conflict with India by threatening war if Pakistan continued to support militants.15 Thus, from 1999 to 2003, nuclear deterrence not only failed to deter military challenges to the status quo in South Asia but, when combined with a territorial dispute, the asymmetry of forces, and a stability-instability paradox, facilitated such challenges resulting in military crises and conflict.16 The India-Pakistan Rivalry From 1999 to 2003, nuclear weapons also exacerbated the broader IndiaPakistan rivalry. Nuclear weapons helped generate military crises. Pakistan’s nuclear deterrent created an atmosphere that prompted its leaders to undertake the Kargil operation and intensify the proxy war with India in 1999–2002. Events in Kargil set back prospects for termination of the India-Pakistan rivalry, which could have occurred under the February 1999 Lahore Accords.17 The duration of Pakistan’s proxy war was prolonged, and military crises recurred because of the perception by the two sides that they could trigger crises and threaten military retaliation without the risk of military escalation. Thus, as crises recurred and Pakistan’s proxy war against India unfolded, the level of hostility and rivalry between India and Pakistan became elevated. As this elevated rivalry became more entrenched, paths to termination of the rivalry remained unexplored. By enabling one side to militarily challenge the status quo that led to repeated military crises, nuclear deterrence prolonged the India-Pakistan rivalry. From Deterrence to Compellence One more effect of nuclear deterrence was that India’s military superiority enabled it to go beyond deterrence and to adopt a strategy of compellence.18 India’s military mobilization in 2001–2 was an exercise in compellence and

190

dinshaw mistry

coercive diplomacy. India demanded that Pakistan roll back its support for terrorists. India’s demands and compellence strategy were made in the context of its substantial military superiority, which was highlighted by its massive mobilization of conventional forces. External Actors The presence of external actors had two effects on the conflict in South Asia. First, nuclear weapons induced external actors—primarily the United States, acting in coordination with close allies such as the United Kingdom—to intervene, ultimately easing South Asia’s military crises. Second, regional powers used their nuclear capabilities to provoke or prolong the crises with the aim of drawing external powers into the conflicts.19 In 1999, one of Islamabad’s primary motivations was that the fear of nuclear war would attract U.S. and international intervention to Kashmir. According to Ashley Tellis and Christine Fair, “many senior Pakistani military officers associated with the planning for this operation believe that these same [nuclear] capabilities . . . served to catalyze U.S. diplomatic interest” in the crisis.20 In the next crisis, Indian decision makers were reluctant to stand down their military mobilization after the U.S. diplomatic effort in January 2002 because they sought to keep the United States engaged in the crisis. They wanted Washington to keep up the pressure on Pakistan at least until the spring. Because India’s military remained mobilized, the threat of war remained significant in May–June 2002 following a major militant attack in Kashmir. This history suggests that India and Pakistan, but also other regional nuclear powers, may be more inclined to resort to violence, and less inclined to back down in military crises, because they anticipate that the United States and other international powers will intervene, thereby bolstering their position vis-à-vis their enduring rivals.

n u c l e a r l e a r n i n g a f t e r s o u t h a s i a’s 2002 military crisis In the years following the 2002 crisis, India-Pakistan relations improved. In 2004, the two sides commenced a “composite dialogue” under which New Delhi agreed to discuss the subject of Kashmir—something that it was reluctant to undertake until Pakistan ended its support of militants. Islamabad agreed to the talks because it included the subject of Kashmir.21 In

complexity of deterrence

191

subsequent years, militant violence in Kashmir decreased.22 The composite dialogue still did not produce a solution to Kashmir, and it may simply have produced a mutually satisfactory absence of a solution. Pakistan may have thought that India was satisfied with the lower level of violence in Kashmir, which was down to a few hundred deaths from militant attacks in 2006 and 2007, compared to over a thousand fatalities annually in prior years, while India may have thought Pakistan was content with simply securing talks about Kashmir’s future.23 In addition, the two sides discussed and adopted limited nuclear confidence building measures. In October 2005, they signed an agreement to provide advance notice of ballistic missile tests. In February 2007, they concluded an agreement to inform each other of any nuclear accidents. Much earlier, at the time they initiated the composite dialogue, the two sides began thinking of their nuclear weapons as sources for stability in the region rather than as the source of threat.24 In the words of the Pakistani ambassador, both countries acknowledged that “their nuclear capability is a factor for stability in South Asia.”25 Since 2003, Pakistan and India have undertaken a composite dialogue and improved their bilateral relationts. The main question remaining is whether nuclear deterrence or other diplomatic or military factors provided the impetus for this improvement in India-Pakistan relations. Nuclear deterrence was probably one of several factors that contributed to the absence of military crises in South Asia since 2002. First, both sides found it less costly and more beneficial to avoid crises. Second, militant attempts on President Pervez Musharraf ’s life made Pakistani officials realize the dangers created by a policy that supported militants who could not be controlled as a finely tuned instrument of foreign policy. Third, a combination of changing power dynamics in the region and India’s increasing economic and military capabilities, the changing, pro-India, position of the United States, which made it less likely to support Pakistan in a future crisis, and less permissive international norms about state support for terrorism after September 11, 2001, all made it harder for Islamabad to continue its proxy war against India.26 By 2002, nuclear weapons complemented these considerations by providing apparent constraints. India and Pakistan remained under international scrutiny. The international community expected that both states would behave responsibly primarily because they were nuclear armed, and the leaders of both states claimed that they were responsible custodians of their nuclear arsenals. The composite dialogue was one way to demonstrate responsible

192

dinshaw mistry

behavior: their February 2007 agreement specifically noted that both sides “[r]ecogniz[e] that the nuclear dimension of the security environment of the two countries adds to their responsibility for avoidance of conflict between the two countries.” Thus, because of their nuclear weapons, the two sides recognized that they had a responsibility to avoid conflict, and the composite dialogue was a means to achieve that end.

nuclear command and control and nuclear doctrines Two factors that influence the stability of deterrence are command-andcontrol systems and nuclear doctrines. In the years after their 1998 nuclear tests, Pakistan and India strengthened their nuclear command-and-control mechanisms. They also clarified key aspects of their nuclear doctrines. Although India-Pakistan relations have improved since 2003, neither state has ruled out the possibility of undertaking military operations against their nuclear-armed opponent. Their nuclear doctrines retain the possibility of nuclear use, and new technologies could further undermine deterrence and facilitate offensive military operations in the subcontinent. Pakistan established a National Command Authority (NCA) to control its nuclear assets in March 1999 and it announced the creation of this command structure in February 2000.27 The NCA is chaired by Pakistan’s president, with the prime minister serving as vice-chairman. This authority comprised an Employment Control Committee (dealing with policies to employ nuclear weapons), a Development Control Committee (concerned with developing Pakistan’s nuclear arsenal and implementing the decisions of the Employment Committee through strategic organizations), and a Strategic Plans Division, which acts as a secretariat for both committees. This division is responsible for the daily management of Pakistan’s strategic assets, oversight of the budget and administration of nuclear laboratories and facilities, and overseeing a security division of 9,000–10,000 personnel who secure Pakistan’s strategic infrastructure. Overall, Pakistan’s military retains significant control of its nuclear assets. The main nuclear management body, the Strategic Plans Division, is under the military’s Joint Chiefs of Staff Committee and is headed by a three-star general. The election of a new government in 2008 somewhat altered Pakistan’s nuclear command-and-control system; a civilian leader then replaced Pervez Musharraf as the president of Pakistan after these elections. While the NCA was established in 1999–2000, key aspects of Pakistan’s

complexity of deterrence

193

nuclear security were strengthened only after revelations of Pakistani scientist A. Q. Khan’s nuclear technology exports in 2003. The NCA then monitored more closely the activities of Pakistan’s nuclear laboratories and scientists. It also developed a stronger security system in the areas of fissile material protection and accounting, sensitive facility perimeter security, transportation security, and a personnel reliability program. Pakistan’s nuclear weapons were also fitted with permissive action links (PALs) that create additional barriers against unauthorized use. PALs were originally not included on Pakistan’s nuclear devices. Pakistan still did not announce a formal nuclear doctrine, although it generally hinted at such a doctrine during the 2002 crisis, when it announced that India’s crossing of any one of four thresholds would trigger nuclear use by Pakistan.28 After the 2002 military standoff with India, Pakistan developed a more coherent nuclear doctrine, which has a number of elements.29 First, Pakistan emphasized that its main line of defense was conventional deterrence rather than nuclear deterrence, and it sought to build up effective conventional forces to respond to Indian conventional threats in any future crisis. Second, Pakistan sought a minimum deterrent that could survive an Indian attack and offer an assured second-strike capability. This implies that a force of some 50–150 deliverable nuclear weapons may be sufficient to provide Pakistan with a minimum deterrent, one that provides an assured second-strike capability against India. Still, Pakistan has not announced a no-first-use policy. In fact, it declared in 2002 that it would be willing to use nuclear weapons if India crosses the four thresholds in any future military crisis. Pakistan’s use of nuclear weapons, however, would presumably be tightly controlled: Islamabad has created a stronger command-and-control structure via the NCA to ensure tight negative control during peacetime and prompt operational readiness (positive control) during crisis and war. India has announced several elements of a nuclear doctrine and clarified aspects of its command-and-control apparatus.30 Within days of its May 1998 nuclear tests, India announced two principles of its nuclear doctrine: (1) that it only sought a minimum nuclear deterrent and (2) that it would adopt a no-first-use policy of nuclear weapons against nuclear-armed states and nonuse against nonnuclear states. In August 1999, India announced an unofficial draft nuclear doctrine, developed by a twenty-two-person National Security Advisory Board (NSAB). This was not an official document but simply represented the combined views of the members of the NSAB. It essentially reiterated existing government pronouncements, stating that India should adopt a no-first-use policy and that it should have a credible

194

dinshaw mistry

minimum nuclear deterrent. The fact that India required a credible and survivable nuclear deterrent led the NSAB to recommend that New Delhi should develop submarine-launched missiles, which constitute a highly survivable weapons delivery system. India first tested such a missile in 2008, but it will take a few more years to deploy this missile on submarines actually on patrol. The board also noted that India should retain civilian control of nuclear weapons with the prime minister or a designated successor retaining nuclear release authority.31 On January 4, 2003, India’s Cabinet Committee on Security officially clarified India’s nuclear doctrine. It confirmed that India sought a “credible minimum deterrent.”32 New Delhi has still not clarified the number of nuclear weapons that would be sufficient for a minimum deterrent. The January 2003 statement also reiterated but diluted India’s no-first-use policy, noting that, “in the event of a major attack against India, or Indian forces anywhere, by biological or chemical weapons, India will retain the option of retaliating with nuclear weapons.” It also mentioned that India would have a Nuclear Command Authority, comprising a Political Council and an Executive Council, and that “the Political Council is chaired by the Prime Minister. It is the sole body that can authorize the use of nuclear weapons.” A Strategic Force Command (SFC), with representation from all three branches of India’s armed forces, serves as the operational arm of the Nuclear Command Authority. This authority presumably controls India’s nuclear warheads and delivery systems. Two of India’s missile groups and some air force and naval assets were designated to the SFC, although each branch of India’s armed forces was still responsible for training and servicing its particular missile, aircraft, and naval assets. India also developed an offensive military strategy called “cold start” that is intended to mobilize quickly its armored divisions for military strikes against Pakistan. India apparently now has the capability to employ its military forces faster than international diplomacy could intervene in a future crisis.33 India’s and Pakistan’s nuclear doctrines and command-and-control arrangements have mixed implications for crisis stability. Both countries strengthened their command-and-control arrangements, which make the unauthorized or accidental use of nuclear weapons less likely during a crisis. India’s nuclear doctrine clarified that only the Political Council, chaired by the prime minister, could authorize the use of nuclear weapons. Yet India took a step back from its no-first-use policy by declaring that it could use nuclear weapons even if Indian territory is not subjected to a nuclear attack. Indian officials noted that they might use nuclear weapons if Indian

complexity of deterrence

195

forces are attacked with nuclear or even chemical or biological weapons in Pakistani territory or if India faces a biological or chemical weapons attack. And while Pakistan’s evolving nuclear strategy since 2002 has emphasized conventional forces as a first line of defense versus India, it could, if conventional defense fails, quickly resort to nuclear use. Islamabad does not have a no-first-use policy. It also has a more delegative command-and-control system for its nuclear weapons, whereby commanders in the field could be authorized to make decisions on nuclear use at critical points in a crisis, such as when Islamabad fears it is losing, or losing control over, its nuclear assets.

the impact of the revolution in military affairs and new technologies The nuclear capabilities sought by India to deter and compel Pakistan were not sufficient to address specific Indian security concerns. Therefore, India sought new capabilities, such as missile defenses.34 After September 11, 2001, Indian government officials feared that if Pakistan’s government collapsed and Islamabad lost control over its nuclear arsenal, India could be the likely target of a Pakistani nuclear attack. Under these circumstances, nuclear deterrence would be insufficient and active defenses would be necessary to counter an attack. Indian officials also observed that the A. Q. Khan network had created an international black market, diffusing nuclear materials and delivery systems to substate groups that could threaten India. Again, India’s nuclear weapons would be inadequate to deter this threat, and defenses would be necessary to counter nonstate actors armed with nuclear weapons. The 2001–2 crisis with Pakistan made clear to Indian policy makers that Pakistan could use nuclear threats to respond to India’s conventional and nuclear superiority in a future crisis. Indian policy makers sought new countervailing strategies against Pakistan, including missile defenses. They assumed that defenses would provide an additional layer of reassurance and protection and would enable India to counter what they termed “nuclear blackmail” by Pakistan, that is, waging terrorist and low-intensity operations against India under the cover of its nuclear umbrella. As a result of these concerns, India’s senior policy makers began embracing the idea of missile defenses. As of 2008, India had not actually acquired a capable missile defense system, although it had purchased missile detection radars from Israel and had tested two indigenously built missile defense interceptors. India’s eventual acquisition of capable missile defenses,

196

dinshaw mistry

combined with its space assets and “revolution in military affairs” technologies such as precision-guided munitions, could undermine Pakistan’s deterrent. Once New Delhi acquires military reconnaissance satellites that provide better coverage of Pakistan’s military installations, it could obtain a counterforce first-strike capability versus Pakistan, especially as long as Pakistan’s nuclear arsenal is small and its delivery systems are situated at a few known airfields and missile bases. India’s satellites could detect Pakistan’s air and missile delivery systems. Its aircraft with precision-guided munitions could target Pakistan’s mobile missiles, while its missiles with improved accuracy could target Pakistan’s airfields. Missile defenses also could provide protection against retaliatory strikes by the limited number of Pakistan’s nuclear delivery systems that survive an Indian preemptive strike. These potential Indian counterforce capabilities could undermine rather than enhance crisis stability. Deterrence rests on both perceptions and capabilities. Therefore, even if Indian officials do not intend to launch a first strike, New Delhi’s capability to undertake a conventional first strike against Islamabad’s nuclear assets weakens Pakistan’s deterrent or at least increases Islamabad’s perceptions that its deterrent is vulnerable. In turn, this potential Indian capability could make Islamabad more likely to deploy its mobile missiles into the field to increase its survivability during a crisis or consider a preemptive strike of its own. These factors could decrease the stability of deterrence in the subcontinent. Missile defenses also could undermine the political basis of crisis stability. Defenses make political decision makers more risk acceptant, causing them to escalate rather than retreat in a crisis,35 because defenses give them confidence that they can destroy some of their opponent’s attacking force. India may still lack the technical capabilities to conduct “splendid first strikes” on Pakistan’s nuclear assets, and Indian missile defenses may only protect a few cities rather than the entire Indian landmass against the “fearsome retaliation” that India would expect from Pakistan.36 India’s continued vulnerability may therefore restrain India’s decision makers from risky offensive operations. Proponents of missile defense therefore note that missile defenses are not destabilizing and instead offer New Delhi limited but important insurance against some catastrophic contingencies in a major crisis.

comparing the old and new nuclear powers There are some similarities between deterrence in South Asia and deterrence between the superpowers. Deterrence has been complex in both cases. For

complexity of deterrence

197

the superpowers, deterrence operated at three levels: deterrence of a nuclear attack, of a conventional war in Europe, and of limited conflicts.37 It has operated at similar levels for Pakistan and India. To deal with this complexity, the superpowers, as well as Pakistan and India, adopted a variety of nuclear and conventional military strategies aimed at deterrence, compellence, and countermilitary operations. Neither the superpowers nor Pakistan and India could rely primarily on their nuclear forces or on a single nuclear deterrence strategy in their defense planning. The fact that serious crises were followed by the absence of crises in India-Pakistan relations also follows a pattern set between the superpowers during the cold war. The superpowers were initially involved in major military crises, but, after the 1962 Cuban Missile Crisis, they began a period of détente. A related point concerns the possibility of nuclear learning. During the cold war, the superpowers learned rules of prudence, such as avoiding direct large-scale combat and avoiding direct challenges in their respective spheres of influence, and this helped avoid serious crises after 1962.38 In South Asia, there is some evidence that nuclear learning and the recognition of the nuclear danger facing both countries has influenced the two sides to undertake a composite dialogue since 2003. There also are important differences between the India-Pakistan enduring rivalry and the superpower standoff, differences that could easily emerge in the relationships between other new nuclear powers. First, Pakistan and India had very small nuclear forces and relatively small nuclear programs in the years after their nuclear tests and during their two major crises.39 Pakistan and India also were slow to develop medium-range missile delivery systems and thermonuclear weapons. Even today, many observers believe that India’s single thermonuclear test was at best a partial success. Pakistan has not tested a thermonuclear device. Further, their nuclear weapons are believed to be kept separate from rather than mated to their delivery systems. Second, external involvement was very significant in the India-Pakistan crises and in deterrence outcomes. Involvement by the major powers, principally the United States, was crucial to easing South Asia’s military crises, but such involvement also encouraged provocative behavior by the participants in the crisis. In contrast, there were no outside powers to intervene significantly in superpower confrontations. Third, in South Asia, the two sides clashed over contiguous territory, and some of this territory was closely tied to national identity for both countries. Thus they had strong motivations and opportunities to challenge and defend the status quo. They fought limited wars and were on the threshold of larger

198

dinshaw mistry

military conflagrations on each other’s home territories. In contrast, during the cold war, the superpowers sought to deter and defend against challenges in the more distant territory of their allies and clients. Fourth, in South Asia, challengers to the status quo were not entirely in control of the means with which they challenged their opponents. The challenger, Pakistan, used irregular forces not directly tied to a client state to challenge the status quo. At times the proxy militant groups behaved according to the intentions of the Pakistan army and intelligence agencies, but at times they worked on their own. This adds to the complexity problem of deterrence. If Pakistan weakens and loses control over elements of its nuclear program, deterrence will become much more complex. Both the Soviet Union and the United States had difficulty controlling their clients during the cold war, but the potential for inadvertent escalation or irrational behavior seems greater when nonstate actors are used as an instrument of war. If a terrorist group acquires nuclear material to build dirty bombs and detonates them within an Indian city, it is unclear how India would retaliate and whether retaliation would involve nuclear weapons. Further research could explore whether the above four aspects of deterrence between Pakistan and India may be manifest among other new nuclear powers. These aspects could be examined for the cases of North Korea since 2003, the behavior of a possibly nuclear Iran, and Israel’s deterrent versus its neighbors and regional rivals since it acquired nuclear-weapons capabilities in the late 1960s. Each of these cases raises important theoretical and policy-relevant questions such as whether the revisionist nuclear state challenges the status quo by means of proxy war and whether the status quo nuclear power can deter or compel its rival to cease such challenges. In the case of North Korea, its nuclear-weapons capabilities could, in theory, embolden it to challenge the status quo versus South Korea, using unconventional tactics such as infiltration and terrorism, which it has adopted in prior decades.40 Yet North Korea has not militarily provoked South Korea since 2003, largely because ties between the two states have gradually improved after Seoul adopted a “Sunshine Policy” of engagement with Pyongyang since 1998. Thus, the North Korean case suggests that if a nuclear state’s political ties with its rivals improve, its nuclear capability does not necessarily result in military provocations versus regional rivals. By contrast, Tehran could follow a policy against Israel or neighboring countries similar to the one employed by Pakistan against India. It could arm proxy groups to challenge the status quo. Tehran already employed such a policy in the 1990s and 2000s—when it did not have a nuclear arsenal—and

complexity of deterrence

199

it could bank on the stability-instability paradox to pave the way for even more effective proxy wars and harassment of its opponents. Israel’s ability to deter or compel Tehran to halt such proxy wars also warrants closer examination. In general, Israel’s nuclear arsenal has not deterred many challenges to the status quo.41 For instance, Israel’s nuclear capability did not deter Egypt from launching a conventional attack in 1973. It also did not deter Iraq from firing Scud missiles at Israel in 1991. And while Iraq did not arm its Scuds with chemical weapons, it remains unclear whether Israel’s nuclear deterrence or other factors were behind Iraq’s nonuse of chemical weapons.42 Further illustrating the limitations of deterrence, Israel’s nuclear capabilities have not deterred terrorism and suicide bombing against Israel; and they have not deterred Iran and Syria from arming proxy groups such as Hezbollah and Hamas that launch attacks against Israel. Thus, it is possible that Israel may follow in India’s footsteps by moving beyond deterrence to a strategy of compellence, and it could militarily retaliate against Tehran to compel it to halt its support for proxy militant groups. In response to Israeli actions, Tehran might escalate its asymmetric war against Tel Aviv. Tehran is a revolutionary state governed by a theocracy, with a risk-taking elite that is willing to push the envelope to achieve its goals.43 Further research could provide more detailed insights into how the strategies and behavior of a nuclear Israel and a nuclear or nuclearizing Iran may resemble or differ from the behavior of the new nuclear powers Pakistan and India.

conclusions Nuclear deterrence between Pakistan and India has been characterized by complexity. In the period from 1999 to 2003, India’s nuclear weapons did not deter Pakistan from military operations seeking to alter the status quo, and Pakistan’s nuclear weapons did not deter India’s plans for an aggressive military response to provocations by Pakistan and Pakistan’s proxies. Several factors contributed to the complexity of deterrence in the India-Pakistan relationship: the existence of a serious territorial dispute over Kashmir; Pakistan’s deep dissatisfaction with the status quo in Kashmir; insurgency in the disputed territory that provided opportunities to challenge the status quo; a stability-instability paradox; asymmetry in military capabilities between the two sides; and the involvement of external actors. Since 2003, however, relations between India and Pakistan have improved. Nuclear learning and the recognition of the nuclear danger facing both countries was one factor that

200

dinshaw mistry

influenced the two sides to undertake a dialogue to reduce the prospects of war and to make some modest progress toward resolving their differences. This nuclear behavior between India and Pakistan has some similarities to that between the superpowers during the cold war, but it also has important differences. Some similarities in the nuclear behavior between India and Pakistan and the cold war superpowers could emerge in the relations among other new nuclear powers. One theme among new nuclear states is the changing nature of deterrence relationships between them and their rivals, especially when they are confronting nonstate actors that act as proxies in their conflict. Revisionist states can mount significant challenges to the status quo, and the defending states cannot easily deter such challenges and may adopt strategies of compellence, defense, or preemption and preventive war against such challenges. Thus, asymmetric military challenges to the status quo, on the home territories of the contending parties, add to the complexity of deterrence among new nuclear powers. Such military challenges under the nuclear umbrella are a major theme in the case of Pakistan and India, and further research would provide better insights into the conditions under which similar military conflicts could take place among other new nuclear powers.

notes 1. Ashok Krishna and P. R. Chari, eds., Kargil: The Tables Turned (New Delhi: Manohar, 2001); Y. M. Bammi, Kargil 1999: The Impregnable Conquered (Noida, India: Gorkha, 2002); Peter Lavoy, ed., Asymmetric Warfare in South Asia (Cambridge: Cambridge University Press, forthcoming). 2. Lavoy, Asymmetric Warfare; Shireen Mazari, The Kargil Conflict 1999: Separating Fact from Fiction (Islamabad: Institute of Strategic Studies, 2003), 42–50; and Shaukat Qadir, “Analysis of the Kargil Conflict 1999,” RUSI Journal 147 (April 2002): 24–30. 3. Egypt could not expect to defeat a conventionally stronger Israel and therefore adopted a limited-aims strategy of intruding only some ten to twenty miles into the Sinai and then defending the intruded area and attracting international involvement and negotiations over Sinai. See T. V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge: Cambridge University Press, 1994), 126–45. 4. Ashley Tellis, C. Christine Fair, and Jamison Jo Medby, Limited Conflicts under the Nuclear Umbrella: Indian and Pakistani Lessons from the Kargil Crisis (Santa Monica: RAND, 2001), 56; V. R. Raghavan, “Limited War and Nuclear Escalation in South Asia,” Nonproliferation Review 8 (Autumn–Winter 2001): 6. 5. For this crisis, see V. K. Sood and Pravin Sawhney, Operation Parakram: The War Unfinished (New Delhi: Sage Publications, 2003); and P. R. Chari, Pervaiz Cheema, and Stephen Cohen, Four Crises and a Peace Process (Washington, DC: Brookings Institution, 2007). 6. S. Paul Kapur, “India and Pakistan’s Unstable Peace: Why Nuclear South Asia Is Not Like

complexity of deterrence

201

Cold War Europe,” International Security 30, no. 2 (2005): 138–39; and Sood and Sawhney, Operation Parakram, 80–83. 7. Sumit Ganguly and Devin Hagerty, Fearful Symmetry: India-Pakistan Crises in the Shadow of Nuclear Weapons (New Delhi: Oxford University Press, 2005), 199. 8. Dinshaw Mistry, “Tempering Optimism about Nuclear Deterrence in South Asia,” Security Studies, forthcoming. 9. Kapur, “India and Pakistan’s Unstable Peace,” 148, citing interview with V. K. Sood, former vice-chief of the Indian army. 10. Alexander George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974). 11. The original stability-instability paradox was that strategic stability at the nuclear level (the belief that conventional war will not escalate to the nuclear level because a nuclear war would be devastating) allows for instability at the conventional level, since it facilitates conventional aggression (by the conventionally stronger side, which is confident of prevailing in a conventional conflict). On this original paradox, see Glenn H. Snyder, “The Balance of Power and the Balance of Terror,” in The Balance of Power, edited by Paul Seabury (San Francisco: Chandler, 1965), 194–201. S. Paul Kapur notes that this original paradox is not manifest in South Asia because it is the weaker, rather than the stronger, side that has initiated limited conflicts. S. Paul Kapur, Dangerous Deterrent: Nuclear Weapons Proliferation and Conflict in South Asia (Stanford: Stanford University Press, 2007). Yet, as this chapter clarifies, there is still a paradox because stability at the nuclear level has produced instability at the subnuclear level. 12. For more on this issue in South Asia, see Tellis et al., Limited Conflicts; Michael Krepon and Chris Gagne, ed., The Stability-Instability Paradox: Nuclear Weapons and Brinkmanship in South Asia (Washington, DC: Henry L. Stimson Center, June 2001). 13. Agha Shahi, Zulfiqar Ali Khan, and Abdul Sattar, “Securing Nuclear Peace,” News International (Karachi), October 5, 1999. 14. Although Pakistan was the weaker side (India’s population is seven times larger than Pakistan’s), the asymmetry was truncated because of rough parity in the theater of conflict, Kashmir, as India and Pakistan had roughly the same number of army divisions in this region. Also, Pakistan was neither too small nor incapable of mounting a sustained challenge to India, and its elites had the motivation and wherewithal to sustain the conflict with India even if it involved economic costs. Further, Pakistan reduced the asymmetry through appropriate strategies and tactics (by supporting militants versus India), alliances, and arms acquisitions. See T. V. Paul, “Why Has the India-Pakistan Rivalry Been So Enduring? Power Asymmetry and an Intractable Conflict,” Security Studies 15 (October 2006): 600–30. 15. Rajesh Basrur, Minimum Deterrence and India’s Nuclear Security (Stanford: Stanford University Press, 2006), 87. 16. Kapur, Dangerous Deterrent. 17. Saira Khan, “Nuclear Weapons and the Prolongation of the India-Pakistan Rivalry,” in The India-Pakistan Conflict: An Enduring Rivalry, edited by T. V. Paul (Cambridge: Cambridge University Press, 2005), 156. 18. Basrur, Minimum Deterrence, 88 19. During the cold war, superpower clients were sometimes more provocative in conflicts because they expected the United States and Soviet Union to intervene and they did not seek to compromise. India and Pakistan were behaving similarly—even after getting nuclear weapons. 20. Tellis et al., Limited Conflicts, 49 21. Stuart Croft, “South Asia’s Arms Control Process: Cricket Diplomacy and the Composite Dialogue,” International Affairs 81 (2005): 1039–60. 22. The number of security personnel killed in militant attacks in Kashmir was approximately 120 (2007) and 150–60 (2006), compared with 190–220 (2005), 280–325 (2004), 310–60

202

dinshaw mistry

(2003), 450–70 (2002), and 540–90 (2001). Civilian deaths were approximately 160 (2007) and 350–90 (2006), compared with 520–60 (2005), 530–700 (2004), 660–800 (2003), 840–1000 (2002), and 900–1100 (2001). Data compiled by the author from South Asia Terrorism Portal, 2007, http://www.satp.org, and Government of India, Union Ministry of Home Affairs, Annual Report 2006–2007, 2007. 23. Croft, “South Asia’s Arms Control Process,” 1053. 24. V. R. Raghavan, “The Double-Edged Effect in South Asia,” Washington Quarterly 27 (Autumn 2004): 147–55. 25. Munir Akram, “Pakistan’s Perspective on Arms Control, Disarmament, WMD Proliferation, and Use by Non-State Actors” (paper presented at the International Peace Academy, Japan, October 4, 2004). 26. Paul, “Why Has the India-Pakistan Rivalry Been So Enduring?” 630. 27. Kenneth Luongo and Naeem Salik, “Building Confidence in Pakistan’s Nuclear Security,” Arms Control Today 27 (December 2007): 11–17. 28. These were an economic or sea blockade, a military threat in the plains, meddling in Pakistani politics, or cutting off Indus River waters. 29. These elements are mentioned in Peter Lavoy, Pakistan’s Nuclear Posture: Security and Survivability (Washington, DC: Nonproliferation Education Center, 2007). 30. Harsh Pant, “India’s Nuclear Doctrine and Command Structure: Implications for CivilMilitary Relations in India,” Armed Forces and Society 33 ( January 2007): 238–64. 31. P. R. Chari, “India’s Nuclear Doctrine: Confused Ambitions,” Nonproliferation Review (Autumn–Winter 2000): 123–35. 32. Government of India, “The Cabinet Committee on Security Reviews Operationalization of India’s Nuclear Doctrine,” January 4, 2003. 33. Rodney Jones, “Conventional Military Imbalance and Strategic Stability in South Asia” (South Asian Strategic Stability Unit Research Paper, University of Bradford, Bradford, UK, March 2005); Walter C. Ladwig, “A Cold Start for Hot Wars? The Indian Army’s New Limited War Doctrine,” International Security 32 (Winter 2007 / 08): 158–90. 34. Ashley Tellis, “The Evolution of U.S.-Indian Ties: Missile Defense in an Emerging Strategic Relationship,” International Security 30 (Spring 2006): 113–51. 35. Robert Powell, “Nuclear Deterrence Theory, Nuclear Proliferation, and National Missile Defense,” International Security 27 (Spring 2003): 86–118. 36. Tellis, “Evolution of U.S.-Indian Ties.” 37. George and Smoke, Deterrence in American Foreign Policy. 38. Joseph S. Nye, “Nuclear Learning and U.S.-Soviet Security Regimes,” International Organization 41 (Summer 1987): 371–402. 39. In the period 1999–2002, India had weapons-grade plutonium sufficient for perhaps one hundred nuclear weapons, while Pakistan had weapons-grade uranium for perhaps fifty nuclear weapons, but these countries are believed to have actually assembled no more than ten or a few tens of nuclear weapons. In contrast, the U.S. stockpile increased from approximately three hundred nuclear weapons in 1950 (within five years of testing its first atomic device) to 15,000 in 1959 and 27,000 in 1962 (the year of the Cuban Missile Crisis), while the Soviet stockpile increased from approximately 150 in 1954 (five years after its first nuclear test) to about one thousand in 1959 and three thousand in 1962. 40. Joseph Bermudez, The Armed Forces of North Korea (New York: I. B. Tauris, 2001); Victor Cha and David Kang, Nuclear North Korea: A Debate on Engagement Strategies (New York: Columbia University Press, 2003); Selig Harrison, Korean Endgame (Princeton: Princeton University Press, 2003); Leon Sigal, Disarming Strangers: Nuclear Diplomacy with North Korea (Princeton: Princeton University Press, 1998). 41. Avner Cohen, Israel and the Bomb (New York: Columbia University Press, 1998); Seymour

complexity of deterrence

203

Hersh, The Samson Option: Israel’s Nuclear Arsenal and American Foreign Policy (New York: Random House, 1991). 42. Zeev Maoz, “The Mixed Blessings of Israel’s Nuclear Policy,” International Security 28, no. 2 (2003): 44–77; Louis Rene Beres, “Israel’s Bomb in the Basement: A Revisiting of Deliberate Ambiguity vs. Disclosure,” in Between War and Peace: Dilemmas of Israeli Security, edited by Efraim Karsh (London: Frank Cass, 1996), 113–33. 43. Anthony Cordesman and Khalid Al-Rodhan, Iran’s Weapons of Mass Destruction (Washington, DC: Center for Strategic and International Studies, 2006). See also Ray Takeyh, Hidden Iran: Paradox and Power in the Islamic Republic (New York: Council on Foreign Relations, 2006); Shahram Chubin, Iran’s Nuclear Ambitions (Washington, DC: Carnegie Endowment for International Peace, 2006).

chapter nine R

Unconventional Deterrence How the Weak Deter the Strong ivan arreguín-toft

two developments have made it easier for weak actors to deter stronger actors from military intervention in conflicts of interest.1 First, strong actors have encountered a class of weak actors who have not only come to reject the idea that “others” may legitimately rule them, but these same weak actors are increasingly apt to use an indirect strategy—for example, guerilla warfare—to prevent others from coercing them. This combination of nationalism and indirect strategy made it possible for weak actors to defeat their materially stronger adversaries in a string of asymmetric conflicts following World War II. Strong actors are therefore less able to coerce weak actors with violence or the threat of violence. Second, since the end of the cold war (1991), strong actors have grown increasingly reluctant to use their advantages in killing power to harm weak actors, especially weak-actor noncombatants, who are more vulnerable when weak actors use indirect strategies than when they use direct strategies. Strong actors—in particular democratic strong actors—are therefore increasingly less willing to coerce weak actors with violence or the threat of violence.2 These trends create a serious challenge to the ability of strong actors to coerce weak actors. The costs of coercing weak actors have risen since World War II, yet the benefits have declined. Weak actors will therefore find it easier to deter strong actors from intervening in their affairs even when the disparity in material power is very large. This chapter explains how the weak deter the strong.3 It focuses on the United States, which, although currently flagging, remains the strongest military and economic actor in the international system. The chapter begins 204

unconventional deterrence

205

with an overview of a key trend: killing power has been less likely to cause victory in asymmetric conflicts, even when the strong actor’s killing power is ten times greater than that of its adversary.4 Policy makers in both materially weak and strong states are aware of this trend, but because explanations for the trend have varied widely, the policy implications that follow are not obvious. Prior to the string of unexpected losses by strong actors in asymmetric conflicts following World War II, no attempt by weak actors to deter by denial would have been credible. A core argument of the chapter is that for “Western” societies, fear of death and killing power evolved in tandem. Western societies then developed specialized political (the state) and military (disciplined, subordinate, combined-arms armed forces) institutions, which privilege a narrowly defined self-preservation above all other values. The result is a system that relies for the creation of coercive power on the existence of a target community that fears death above all else. As a sine qua non of rationality, “fear of death above all else” is spatially and historically bounded, not universal. The chapter concludes by observing how the George W. Bush administration has attempted to reverse both the trend of losses in unconventional asymmetric conflicts and the policy implications that follow from that trend. The likely defeats in Afghanistan and Iraq indicate that unconventional deterrence—understood here as what Lawrence Freedman refers to as “internalized deterrence”5—is likely to remain an important feature of contemporary interstate politics.

the diminishing utility of killing power In the roughly two hundred years of conflict following the conclusion of the Napoleonic Wars in Europe, when strong actors fought wars with weak actors, strong actors won a clear majority of such wars. The overall success rate of strong powers, defined as those states that outpower their competitors by a margin of at least ten to one, is 71.5 percent (see Figure 9.1).6 Yet when these two centuries are divided into four fifty-year subperiods, a startling pattern emerges: weak powers have been winning more over time. In the first of the four periods (1816–49), strong powers won nearly 90 percent of the time. In the most recent period (1950–2000), weak powers won a majority of such asymmetric wars. Power, as commonly understood, has not correlated with expected outcomes. By 1954, the West had experienced three losses against insurgencies and three wins. On the loss side stood China, Indochina, and Algeria; on the win

206

ivan arreguín-toft

figure 9.1: Percentage of Victories in Asymmetric Conflict in Four Fifty-Year Periods. s ou rc e : Arreguín-Toft, How the Weak Win Wars, 5.

side stood Greece, Malaya, and the Philippines. The conflict in Indochina continued into the 1970s. Vietnamese nationalism played a crucial role in the struggle for control of Indochina.7 By the time the United States entered the conflict, nationalism had made it easier for the Viet Cong and Democratic Republic of Vietnam to paint U.S. support of South Vietnam as a proxy for foreign domination of Vietnam than it was for the United States and the government of (South) Vietnam to commit to self-sufficient Vietnamese national independence. This was especially true after the John F. Kennedy administration reluctantly agreed in 1963 to allow the assassination of Ngo Dinh Diem, Vietnam’s only credible noncommunist nationalist leader. The subsequent increase in the number of foreign combat troops in South Vietnam only made matters worse. In the Malayan Emergency, which lasted twelve years, the British had been favored by a more patient public, so that when its military made mistakes, public opinion gave it the luxury of a chance to correct them.8 In Vietnam, U.S. public opinion in support of the war declined after three years, just as the U.S. military was beginning to see gains from its efforts to adapt to counterinsurgency warfare. In Malaya, the British were able to use nationalism to counter socialist ideology, and the former proved devastating to the latter. By credibly committing to Malayan independence, Britain deprived the insurgency of its most potent mobilization grievance. In the Philippines, the United States, with the help of Ramon Magsaysay—a genuinely reformist Filipino leader—accomplished the same thing against the Hukbalahap. Having forsworn an invasion of North Vietnam, the United States could never promise more than independence for South Vietnam, and because its approach to counterinsurgency was overwhelmingly militarized,9 its increasing presence in South Vietnam had several bad effects. First, it made the South Vietnamese government look dependent for its existence on the

unconventional deterrence

207

presence of foreign troops. Second, U.S. forces had been specialized to fight and win wars with massive firepower against identifiable adversaries. North Vietnam, where uniformed forces abounded, was effectively off-limits except to U.S. air power; and South Vietnam did not present U.S. forces with many identifiable adversaries. In that context, massive firepower meant massive collateral damage; and massive collateral damage became a crucial mobilization grievance for the Viet Cong. Whereas the British eventually switched from a failed strategy of “more force to counter more resistance” to a successful strategy of “more law enforcement to counter more resistance,” with a few notable exceptions the United States responded to increased resistance with increased firepower. By 1954, the civil war in Vietnam had already been won by the North. The best any intervening power could have therefore managed would have been to delay the unification of North and South under communist rule. Had promised elections been held in 1956, the North would certainly have won. Had the South not been supported by the United States with arms, advisors, and eventually combat forces, the North would certainly have won. Given that at the time the United States and its Western allies believed that Vietnamese unification under communist rule would necessarily be a disaster for the Vietnamese people and a blow to political stability throughout Asia, the fact that the United States was able to delay this outcome for twenty years should stand as a tribute to, rather than an indictment of, its military and political elites. Vietnam affected the United States in another crucial respect: news images and antiwar activism, along with some genuinely brutal and illegal conduct on the part of U.S. armed forces in Vietnam, combined to convince most Americans that the war in Vietnam could not be fought justly and was in fact brutalizing noncombatants as a matter of course. Pulitzer Prize–winning images—such as a little girl running from a napalm attack or of a South Vietnamese general summarily executing a suspected, but untried, Viet Cong insurgent—came to characterize the U.S. military’s engagement in that war for many Americans. As the argument that preserving an independent South Vietnam was vital for U.S. national security became less and less credible, the toll of casualties—low by historical standards—and the quantity and quality of the killing—unprecedented by historical standards—both became intolerable. In sum, after World War II, nationalism, guerrilla warfare strategy, and a loss of confidence in domino logic combined to turn the chief advantage of advanced industrial state militaries—killing power—into something of a liability. Rather than leading to victory in most conflicts as it had in the past,

208

ivan arreguín-toft

killing power turned out to be an advantage in some wars against some adversaries but a serious disadvantage in other wars against other adversaries. Deterrence and Rationality Deterrence means the creation of circumstances that privilege inaction along a specific issue dimension. Inaction is made desirable when the actor who intends to deter another communicates new information—that is, that costs will be high—in order to offset an anticipated benefit of action to the target actor. Deterrence often combines some form of capability increase (or transparency) with some form of specific intent message, for example, we will use our capacity to hurt you should you attempt to undertake a specified action. There are many types of deterrence, but common to nearly all is an assumption that the actor to be deterred is rational.10 Rationality is understood to mean that the target of a deterrent threat can undertake a net assessment: when costs of possible action or inaction are calculated to exceed benefits, that specific action will not be undertaken.11 The wider that margin, the less likely is action across that particular issue dimension. Rational actors thus invariably (1) calculate and (2) attempt to reduce the costs to themselves of proposed or expected interactions with other actors. The key mechanism of deterrence is simple: deterrence is a credible attempt by one actor to manipulate another actor’s appreciation of the costs of a given action. Deterrence is said to “hold” when the target actor acknowledges that it has chosen not to act because of the threat of increased costs by the actor attempting to deter its otherwise desired action. Deterrence is said to “fail” when the target actor undertakes an action even after a credible threat has been made that communicates that the costs of undertaking the specified action will be higher than any expected benefits. For example, if the Soviet Union both asserted a desire to use its numerical conventional military advantage in Eastern Europe to invade Western Europe, and then acknowledged it never followed through on its desire because of the possession of tactical nuclear weapons by North Atlantic Treaty Organization (NATO) forces, nuclear deterrence could have been cited without controversy as the best explanation for what historian John Lewis Gaddis has called the “Long Peace” in Europe. The actual motives and cost-benefit calculations behind the end of the cold war, however, were more mixed and muddied than this.12 So deterrence must remain a potential, but not the sole, explanation for the lack of a major conventional war in Europe from 1948 to 1990.

unconventional deterrence

209

Deterrence and Relevant Power Power is a crucial component in considerations of costs and benefits, but it is difficult to define in general terms.13 When in a bargaining situation, one actor will often have an idea about the other actor’s wants and fears, and it may be wrong about both or either. Generally, the closer one is to one’s target in terms of history, culture, social class, and language, the less likely one is to err in estimating how another will perceive one’s threats or bribes. As Freedman puts it: Deterrence works best when the targets are able to act rationally, and when the deterrer and the deterred are working within a sufficiently shared normative framework so that it is possible to inculcate a sense of appropriate behavior in defined situations that can be reinforced by a combination of social pressures and a sense of fair and effective punishment. Norms, therefore, do not develop and exist independently of assertions of power and interest.14

The corollary to this proposition should also prove true: the more “foreign” the target, the more likely one is to err in estimating how the target will react to a threat or bribe. Allan Bloom, in the introduction to his translation of Jean-Jacques Rousseau’s Emile, points out that European bourgeoisie (and by implication their descendants in North America, Australia, and elsewhere) have shared a common denominator that could be relied upon to provide coercive leverage: “Most simply, following Hegel’s formula, he is the man motivated by fear of violent death, the man whose primary concern is self-preservation or, according to Locke’s correction of Hobbes, comfortable self-preservation.”15 Since, in the Western canon, death stops all arguments, credibly threatening to destroy a target state becomes a useful and presumably universal way to coerce a target’s behavior. In the language of constructivism or critical theory, both fear of death and the threat of lethal violence have become naturalized, such that they take on the aspect of axioms as opposed to culturally or historically specific variables. What type of power is most useful for deterrence? So long as the target fears death above all else, then the relevant power must be the power to kill. This axiom is reflected in the course of European history. From the time of the fall of the Roman Empire to the twentieth century, the middle classes of Europe were indispensable in creating and refining the means to kill on a scale and in a span of time previously unimaginable. The triumph in all places of the middle classes over the aristocratic classes was in part due to

210

ivan arreguín-toft

the vital links between warfare and scientific developments in metallurgy, chemistry, physics, mathematics, manufacturing, steam combustion, and navigation, none of which were of much interest to Europe’s aristocrats or working classes. The foundation of the contemporary European states system is fear of death, and Europeans thus sought and mastered the technology of killing as the preeminent form of power—the power to coerce, deter, and destroy. Within the same cultural and historical space, the power to kill is like a shared language, and becomes synonymous with the broader concept of power itself. It follows that where power is generally understood to mean killing, coercive and deterrent strategies would revolve around calculations of which actor or coalition of actors possessed the greatest capacity to kill. Those with a lot of killing power will, ipso facto, find themselves able to coerce and deter those with significantly less killing power. A subsidiary but crucial factor must be added. The increased willingness of insurgents to court death is necessary but not sufficient to account for what I call unconventional deterrence. In addition to a nationalism-inspired or religiously inspired willingness to sacrifice life for other values, and the refinement of a strategy designed to take advantage of this idea against advanced industrial states, another crucial trend has emerged in interstate politics: a decreased willingness to kill or, in terms of the present analysis, a decreased willingness to use power to achieve foreign policy objectives where the target is a nonstate actor. This is also the most common form of organized violence today, interstate war having largely taken a hiatus in international affairs. Killing and Coercion over Time The last half of the twentieth century witnessed a steady erosion of the ability of killing power to effect coercion in asymmetric conflicts. Something is causing it to be less effective in political terms. The technologies and strategies employed by the stronger actors in such asymmetric conflicts resulted in more killing in such wars over time (U.S. intervention in Vietnam and Soviet intervention in Afghanistan are exemplary in this regard), but outcomes are not mapping as expected. It is no coincidence that within this same historical period, the consolidation of nationalism as a systemwide norm of legitimacy and as a force multiplier and an increasing use of indirect-defense strategies by nominally weak actors can be observed. When the Western or bourgeois

unconventional deterrence

211

conception of power is exported beyond its boundaries in the era of nationalism, it runs into difficulties as a coercive tool.16 Even credible threats to kill are proving less effective at coercion. By extension, weak actors may not be deterred in pursuit of actions to which strong actors object. The string of unexpected—indeed, largely unexplained—defeats of the strong by the weak after World War II led to the preposterous condition of unconventional deterrence. Having seen the trend, political elites have been less willing to engage indirect threats with armed forces, even knowing that in all such engagements their troops would enjoy a considerable technological or numerical advantage.17 As Freedman puts it, “The unacceptable practices of foreign governments are denounced but they are left untouched; ideological ambitions are shelved; inconveniences, disruptions, outrages are tolerated; punches are pulled.”18 In U.S. military history, the Vietnam War stands as a watershed conflict in this regard. During the war, a great deal of debate surrounded the question of how many casualties North Vietnam could sustain before giving in.19 U.S. defense secretary Robert McNamara went on record as speculating that there might be no number of deaths short of actual genocide that could coerce the North into accepting a divided Vietnam.20 If true, the level of killing would have shifted our characterization of the use of force in Vietnam from coercive to destructive (brute force) in the way suggested by Thomas Schelling.21 The coercive use of armed force, under limited circumstances, falls well within the purview of legitimacy for most Americans; but brute force does not. The toll of unexpected losses in the third world has caused a core policy constituency—the military—to internalize the idea that employing the state’s killing power against adversaries in unconventional (substate) settings is either of low utility or counterproductive.22 Operations against “unconventional” adversaries share three liabilities in this view. First, they invariably involve the killing of noncombatants because, as Michael Walzer notes, insurgents systematically hide behind noncombatants.23 Second, they promise the killing of others when those others cannot, by definition, levy a direct or mortal threat to the killers’ state (and by extension, their families and livelihoods). Like the noncombatant liability, this adds to the morale corrosion of unconventional campaigns due to an additional normative axiom: lethal force may only be justified in defense of a lethal threat. So long as political elites could sustain the case that nonlethal threats necessarily constituted a lethal threat in combination (domino logic), Western publics could go along with the use of lethal force—even against noncombatants. But once

212

ivan arreguín-toft

real-world events made that linkage implausible (dominoes did not fall far, and, as observed above, the shift to the victory of weaker powers was accelerated by the end of the cold war itself ), the legitimacy of the application of lethal force came into direct question, making it impossible to use armed forces in just wars. Third, the political imperative to improve battlefield performance often acts to force militaries to alter doctrine, training, and leadership. These new capabilities then become a potential threat to the conventional mission24 (generally defense against the regular armed forces of other advanced industrial states), which favors heavily equipped, firepower-intensive, and hierarchically disciplined units.

unconventional deterrence Unconventional deterrence has now come into play in a powerful way, but only in some conflict dyads and not in others. The key determinant of unconventional deterrence success turns out to be a function of the strategic interaction of adversaries.25 This is because strategies contain and mix cultural, historical, technological, and ideational elements (including victory conditions). Once adopted, therefore, strategies have the effect of shifting culture. This is partly what Samuel Huntington means when he says a challenge for developing states is “to modernize but not to Westernize.”26 The continuing integration of modern weapons and Western military strategies provided tremendous coercive power in the West. But the export of modern weapons to the developing world,27 which only makes sense in combination with Western military strategies, did not provide the same power to the states that imported them (or rather the power they may provide depends on the nature of the adversary). When a strong state approaches a weak adversary that uses an indirectdefense strategy—that is, is prepared to sacrifice proportionately many more lives and territory—the leadership of the stronger state is apt to find its killing power discounted to the point at which the costs of war exceed the benefits. When a strong state threatens a weak one within the same cultural and historical space (and the weak party deploys a direct-defense strategy), war is actually less likely. But when war does happen, the strong state is much more likely to win because its killing power is less apt to be discounted across the dyad. Belgian resistance to Germany in 1914, Argentina’s attempt to hold onto the Falkland Islands, and Iraq’s attempt to hold onto Kuwait are three examples of weak actors offering resistance against a more powerful adversary

unconventional deterrence

213

when both sides employ similar strategies. In the Belgian case, the fight was over in seventy-two hours. But the fact of Belgium’s resistance hurt Germany both by slowing down its plan to survive a two-front war (the Schlieffen Plan) and by making Germany look like a rude bully (a huge stimulus to British army recruitment). In the Falklands, Argentine armed forces attempted to defend the islands with a conventional defense and lost rather handily to the more powerful and technologically advanced British. Finally, in Iraq in 1991, Iraqi armed forces sought to maintain Iraq’s occupation and control of Kuwait using a conventional defense strategy against a U.S. military that had done nothing else since World War II but train and practice to defeat just such an adversary. The result, again, was a rapid defeat of the weaker party by the stronger adversary.28 But in Indochina, Algeria, Vietnam, and Afghanistan, the weaker party proved able to deny the stronger side its political objectives, and the stronger side was forced to give up after a costly effort. In order to win, the materially weaker side needs social support, sanctuary (either physical, such as tough terrain, or political, such as a weakly defended interstate border), an idea capable of making self-sacrifice seem both necessary and noble (e.g., nationalism, jihad), and a strategy capable of tying all three advantages into a single effort. With these four assets, weaker opponents can delay, denying their nominally stronger adversaries the ability to use their advantages in technology, materiel, and numbers to obtain expected political objectives, imposing sufficient costs to force strong opponents to reconsider the expected benefits of victory. Precisely because strategies mix in cultural, material, and historical elements, neither strong nor weak actors are entirely free to choose strategies. The U.S. military, for example, has always disdained special operations forces and the use of a guerrilla warfare strategy to defeat its adversaries (“sneaky” and “dishonorable” would be two adjectives commonly applied).29 This was the situation that faced Ethiopia’s emperor Haile Selassie as the numerically superior and highly mechanized Italian army was bearing down on his spear-wielding warriors in 1935. He begged his princes to fight a hitand-run style of war, but they refused because such tactics were beneath their dignity.30 Instead, they chose to engage the Italians (and their Eritrean allies) in brave but suicidal frontal assaults. The result was defeat. The problem for the United States and other advanced industrial states is that since World War II, the distribution of conflict types has shifted to favor conflicts of interest across rather than within the cultural and strategic divide; “unconventional” adversaries have been more common than “conventional” adversaries. Thus, unconventional deterrence can be expected to

214

ivan arreguín-toft

be robust whenever the weaker party can credibly threaten an unconventional defense. This explains why, since the end of the cold war, repressive regimes from Asia to Africa and Latin America have been able to thumb their noses at the superpowers, in many but not all cases effectively deterring these mighty military powers from setting foot in their territories or halting whatever abuses they chose to inflict on portions of their population. What began as a difficult and costly process of deterrence by denial evolved into internalized deterrence. Especially after the cold war, strong actors with adversaries whose location and history suggest a credible threat of indirect defense (e.g., guerrilla warfare) can rely on the precedent of successful insurgencies to deter would-be aggressors from attempting coercion via military intervention. As late as 1992, a year after the collapse of the Soviet Union, the United States and its allies were deterred from military intervention in Bosnia-Herzegovina,31 despite the similarity of the barbarism there to what Europe had witnessed under Nazi occupation. Serb nationalists, led by Slobodan Milosevic, claimed that the Balkans would become another Vietnam for any European power that attempted to intervene with ground forces.32 The widely perceived defeat of U.S. forces in the Battle of Mogadishu a year later was not the result of a failed attempt by Somali warlords to deter the United Nations (UN) from intervention, but the debacle did materially impact UN and U.S. calculations regarding the feasibility of military intervention to halt a planned mass murder of ethnic Hutus and political moderates in Rwanda the following year. In that case, the United States was deterred from intervention to halt a genocide,33 even though it was perpetrated by poorly armed and equipped paramilitaries. In 1994, North Korea failed to deter the United States from plans to destroy its nuclear facilities by air assault but was spared an assault by the quixotic intervention of former U.S. president Jimmy Carter, who brokered a last-minute deal that averted the planned strike. Perhaps the only case of unconventional deterrence failure (where deterrence was of the internalized type) was in the 2002 Operation Enduring Freedom, in which the United States sought the destruction of the Taliban in Afghanistan, a place said to be the “graveyard” of modern militaries. But by the time of the United States–led assault on Afghanistan, the Taliban had so alienated its popular base there that its strategic options were limited to a conventional or direct defense, as opposed to the indirect defense that had given the Soviets in the 1980s and the British a century earlier so much trouble. On the cost side, there is little controversy about why the Taliban failed to deter the United States from military intervention: the

unconventional deterrence

215

connection between the Taliban and the recent assaults against the United States by al-Qaeda meant that the United States would have cost tolerances approaching those of World War II. This explains why the 2003 assault on Saddam Hussein’s Iraq was so much more controversial: nearly two years had passed since September 11, and the connection between Hussein and al-Qaeda could never be established. Yet, the attack on Iraq was not a failure of unconventional deterrence, because Iraqi armed forces were directed to defend Iraq with a conventional or direct-defense strategy and were even weaker vis-à-vis their adversaries than they had been a decade earlier. Why Unconventional Deterrence Works Advanced industrial states, such as the United States and Britain, must maintain the forces, doctrines, and technologies appropriate to deter and if necessary defeat peer adversaries.34 They also must be prepared to engage very real but less direct and intense threats beyond their immediate neighborhoods.35 Unfortunately, the forces, doctrines, and technologies appropriate to high-intensity conventional combat tend to be counterproductive when used in unconventional conflicts. Several trends are converging to produce this outcome. First, national self-determination as a master principle of legitimacy spread from Europe to Europe’s colonies before a strategy designed to harness its power came into play. In the 1920s, civil wars in Russia and China pitted increasingly skilled prosecutors of indirect or guerrilla warfare against one another, but these forces had been motivated by ideological considerations over nationalism. The Nazi occupation of the Soviet Union and the Balkans, China’s invasion of Korea, and Japan’s attempt to control Asia linked nationalism to guerrilla warfare as a method, and the stage was set for a series of major confrontations that the advanced industrial colonial powers would, against all expectations, lose. Whereas, prior to World War I, most insurgent resistance could be demobilized by brutal countermeasures, since World War II, brutal countermeasures have tended to stimulate resistance.36 What one might call “inadvertent brutality”—now “collateral damage”—had the same effect. Firepower-centric militaries that engaged guerrillas in Asia following World War II generally stimulated more resistance than they eliminated. Second, the collapse of cold war rivalry reduced the strategic value of the network of allies the United States had built to contain the Soviets. Even assuming that the costs of intervention in substate violence had remained constant, this benefit reduction had the net effect of raising the perceived

216

ivan arreguín-toft

costs of any military intervention. These costs continue to rise, so the argument for military intervention in the developing world is becoming increasingly difficult to sustain. Two key events following the collapse of the Soviet Union in 1991 support this claim. First, the rapid coalition victory over Iraq’s conventional forces that same year appeared to rehabilitate U.S. faith in the general utility of armed force. U.S. victories in Grenada (1983) and Panama (1989) had not had this effect. In the absence of a threat of escalation to world war, it appeared to Americans that the United States had the capacity to achieve victory at very low costs. The second event appeared to contradict this possibility. The United States had defeated the fourth largest army in the world in a matter of weeks, forcing it to leave Kuwait and chasing it all the way back to Baghdad. Yet, its well-intentioned humanitarian intervention in Somalia (1993), in which Task Force Ranger was opposed by a few thousand Habr Gidr clansmen in Mogadishu, went horribly wrong. Perhaps, as then chairman of the Joint Chiefs of Staff Colin Powell had warned,37 the costs of intervention vary with the nature of the opponent. What if the opponents of the future came to look less and less like Iraq in 1990 and more and more like Aideed’s clansmen in Mogadishu? The Somalia debacle was followed by genocidal mass murder in Rwanda (1994), mass rape and ethnic cleansing in Kosovo (1998–99), and now Darfur (2003–present). During the cold war, there may have been a rationale for intervention in Rwanda or Darfur, if only to support an anticommunist ally or to undermine a communist proxy. In the absence of either a vital strategic resource or a superpower rivalry, however, there is no strategic rationale for intervention, and by that logic we should have expected what actually transpired: no meaningful intervention. In Rwanda, we have a smoking gun in the form of a specific reference to “no more Somalias” in a message from UN headquarters, forbidding the UN force commander, Roméo Dallaire, from raiding known Hutu arms caches in Kigali.38 This was a striking example of what T. V. Paul has referred to elsewhere in this volume as self-deterrence. One “successful” military intervention, however, occurred following the cold war’s end. In 1999, NATO forced a halt to Serb-supported barbarism in Kosovo. NATO’s military operation actually failed, because Serb-supported paramilitaries succeeded in accomplishing their primary political objective, which was the mass expulsion of almost all of Kosovo’s Kosovar Albanian population. Serb-supported forces escalated their attacks against noncombatants after NATO’s bombing campaign got underway. NATO’s killing power proved unable to deter Serb-sponsored atrocities. The exact cause of Serbia’s

unconventional deterrence

217

decision to retire from Kosovo is still the subject of considerable debate. But as NATO considered a ground invasion, which promised the possibility that Serbia’s helpless victims might find allies who actually shot back, Serbia withdrew and the mass rapes, murders, and looting ceased. The reason NATO’s military intervention is today considered a success is due to the unprecedented rate of return of Kosovar Albanian refugees (over 90 percent), not because killing power effected deterrence or coercion. Because refugee repatriation historically averages no more than 20 percent, by the time Serb political elites agreed to retire from Kosovo, they had every reason to believe their armed and paramilitary forces had achieved their objectives. International response and Kosovar Albanian courage, anger, and determination after the war undid this Serb goal. Today, nationalist actors in tough terrain, using indirect-defense strategies, or those actors whose defeat implies no material benefit to the attacker, generally find it much easier to deter the United States and its peers from intervention, even in cases that shock or sicken, such as Rwanda and Darfur. The benefits of a positive outcome are low to nil, and the costs are likely to be high—perhaps a Vietnam-like quagmire. It is beside the point whether this is due to social squeamishness on the part of G7 countries39 or to simple public cost sensitivity (the requirements of long-term political success in such contingencies are well-known—generally ten to eleven specially trained troops per 1,000 inhabitants for from seven to twenty years40— and are suspect not only because of the unbearable costs in relation to expected benefits but because where successful they will look suspiciously like colonialism).

conclusions When conventional deterrence works, it works best when both sides define power in similar ways. After World War II, the distribution of violent conflicts shifted decisively from matches pitting advanced industrial states against each other to matches pitting advanced industrial states against a relatively less developed state or nonstate actor employing what Western militaries often referred to as unconventional strategy. After 1991, cold war rivalry could no longer justify the costs and risks of military intervention, which either stayed as high as they had been during the cold war or increased. With perceived benefits in decline and costs high or increasing, and with weak actors more apt to defend themselves by hiding combatants among noncombatants, the general utility of military intervention would

218

ivan arreguín-toft

seem to be in grave doubt. Not only has winning such wars become more difficult, but winning the peace following a victorious military outcome has become all but impossible. This leads to what Richard Falk aptly characterized as a core dilemma of contemporary world politics: “non-intervention is intolerable, but intervention remains impossible.”41 Unconventional deterrence works and has become more prominent over time because the nature of power itself—the power to destroy, the power to coerce and deter—turns out to be historically and culturally specific. In times and places in which peoples (and by extension, states) fear death above all else, the credible threat to kill effects coercive leverage. Where fear of death is not paramount, however, as in many places in the developing world after the establishment of national self-determination as a powerful systemwide norm of legitimacy, the same threat does not effect the same leverage and may even backfire, creating nationalist or religious martyrs who, although dead, dramatically increase the subsequent costs of using force to achieve political objectives. Moreover, the fact that strong states share the view that national self-determination is legitimate imposes a cost on them, which is that killing those fighting for national self-determination—as opposed to other goals—becomes a bad thing. Thus, in addition to being less able to coerce weaker parties over time, strong states have shown themselves to be progressively less willing to do so as well. Western states, burned so many times in military interventions during and after the cold war, have come to see the writing on the wall: they intuitively grasp the significance of the empirical trend reported in Figure 9.1 above. They have been deterred by nominally weak actors (those with small populations, few resources, no powerful allies, and small low-technology militaries), by the promise of high costs (counterintuitive), and by declining political or economic benefits. The George W. Bush administration appears to have made a go at reversing these trends. Having identified previous U.S. failures as a problem of insufficient will, his administration has moved forward with aggressive rhetoric punctuated by actual military interventions in Afghanistan (2001–present), and Iraq (2003–present). But the “will” the United States lacked in Vietnam was ultimately not the will to win but the will to win at the cost of perpetrating genocide against the Vietnamese people. Today, the complexities and accelerating costs of the Afghan and Iraqi interventions have taken their toll on the president’s popularity, to say nothing of U.S. forces themselves and the U.S. budget. U.S. foreign policy appears to be returning now to its status quo ante, as the

unconventional deterrence

219

president and his advisors seek to reestablish basic diplomatic contacts and credibility earlier discounted in the “force first” era. The only way to overcome unconventional deterrence and facilitate effective interventions to halt genocides, such as those in Rwanda and Darfur, would be for strong states to build and deploy armed forces (supplemented by economic and administrative resources) able to execute a well-designed counterinsurgency strategy, such as that followed by the British in Malaya in 1952, and for strong states to be able to sustain that strategy and those forces over time, which calls for steady leadership. Once the likelihood of denial by materially weak actors is diminished, self- or internalized deterrence will fade, leaving advanced industrial states with more coercive options in a crisis. The alternative is to simply follow General Powell’s advice: leave the military unchanged and forswear all interventions beyond the advanced industrial world. That is, unfortunately, not an option likely to make the world more stable, just, or prosperous.

notes 1. “Weak” and “strong” refer to relative material power across a conflict dyad only. Asymmetric conflicts are singled out for analysis because they by definition present the greatest puzzle: if deterrence is, inter alia, about the threat of harm, how is it that the weak ever deter the strong? 2. For an elaboration of this thesis, see Gil Merom, How Democracies Lose Small Wars: State, Society, and the Failures of France in Algeria, Israel in Lebanon, and the United States in Vietnam (Cambridge: Cambridge University Press, 2003). 3. For an excellent summary of deterrence logic and typology, see T. V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (New York: Cambridge, 1994): 7–10. For a broader introduction, see Lawrence Freedman, Deterrence (New York: Polity Press, 2004). On “internalized deterrence,” see Freedman, Deterrence, 29–32. 4. I define “power” more explicitly below. 5. Freedman, Deterrence, 29–32. 6. Arreguín-Toft, How the Weak Win Wars, A Theory of Asymmetric Conflict (Cambridge: Cambridge University Press, 2005), 3. 7. See, e.g., Walker Connor, “Ethnology and the Peace of South Asia,” World Politics 22 (October 1969): 51–86; and Chalmers Johnson, “The Third Generation of Guerrilla Warfare,” Asian Survey 12 ( June 1968): 435–47. 8. On the Malayan Emergency and British counterinsurgency strategy, see Robert Thompson, Defeating Communist Insurgency: Experiences from Malaya and Vietnam (London: Chatto & Windus, 1966); John Nagl, Counterinsurgency Lessons from Malaya and Vietnam: Learning to Eat Soup with a Knife (Westport, CT: Praeger, 2002); and Ivan Arreguín-Toft, “How to Lose a War on Terror: A Comparative Analysis of a Counterinsurgency Success and Failure,” in Jan Ångström and Isabelle Duyvesteyn, eds., Understanding Victory and Defeat in Contemporary War (London: Frank Cass, 2007), 142–67. 9. See Geoffrey D. T. Shaw, “Policemen versus Soldiers, the Debate Leading to MAAG Objections and Washington’s Rejection of the Core of the British Counter-Insurgency Advice,” Small Wars and Insurgencies 12 (Summer 2001): 51–78.

220

ivan arreguín-toft

10. See, e.g., Paul, Asymmetric Conflicts, 8. 11. See Lawrence Freedman’s qualification, however, that one can undertake net assessments from within different cost-benefit frameworks and arrive at results that are both entirely “rational” and entirely unexpected by others who calculate from within a different framework. Freedman, Deterrence, 49. I discuss this in more detail below. 12. Freedman, Deterrence, chapter 3. 13. For a good comprehensive treatment of “power,” and debates concerning its meaning and definition, see Steven Lukes, ed., Power (New York: New York University Press, 1986). For purposes of this chapter, I follow the tradition of previous work on asymmetric conflict in which a resource-based or material conception is favored. See, e.g., Paul, Asymmetric Conflicts, 20, 22; and Arreguín-Toft, How the Weak Win Wars, 2. 14. Freedman, Deterrence, 5. 15. Allan Bloom, “Introduction,” in Jean-Jacques Rousseau, Emile: or On Education, translated Allan Bloom (New York: Basic Books, 1979), 5. 16. See Albert O. Hirschman, The Passions and the Interests: Political Arguments for Capitalism Before Its Triumph (Princeton: Princeton University Press, 1977); Stephen Toulmin, Cosmopolis: The Hidden Agenda of Modernity (Chicago: University of Chicago Press, 1990). On nationalism as a threat to interstate order, see Barry R. Posen, “Nationalism, the Mass Army, and Military Power,” International Security 18 (Autumn 1993): 80–124. On the link between religion and rationality in the context of interstate stability (and the Thirty Years’ War), see Monica Duffy Toft, “Getting Religion?: The Puzzling Case of Islam and Civil War,” International Security 31 (Spring 2007): 97–131. 17. Again, internalized deterrence. See especially Caspar W. Weinberger, “The Uses of Military Power,” speech before the National Press Club, Washington, DC, November 28, 1984; and Colin L. Powell, “U.S. Forces: Challenges Ahead,” Foreign Affairs 71 (Fall /Winter 1992): 32–45. In separate but very similar arguments now known as the Weinberger and Powell doctrines, both men argued that the key to maintaining U.S. national security (and the morale and fighting power of U.S. armed forces) was avoiding the use of U.S. armed forces to engage indirect threats to U.S national security, particularly any contingency that might devolve into a Vietnam-like quagmire. 18. Freedman, Deterrence, 30. 19. John Mueller, “The Search for the ‘Breaking Point’ in Vietnam: The Statistics of a Deadly Quarrel,” International Studies Quarterly 24 (1980): 497–519; Frederick Z. Brown, “Comment on Mueller: American Misperceptions,” International Studies Quarterly 24 (1980): 525–29. 20. See Steven Rosen, “War Power and the Willingness to Suffer,” in Peace, War, and Numbers, edited by Bruce M. Russett (Beverly Hills: Sage, 1972), 167–68; Mueller, “Search for the ‘Breaking Point’,” 497–519; Brown, “Comment on Mueller,” 525–29; and Stanley Karnow, Vietnam: A History (New York: Viking Press, 1983), 454, 596. 21. Thomas C. Schelling, Arms and Influence (New Haven: Yale University Press, 1966). 22. The Powell and Weinberger Doctrines cited above are strong evidence of this, as is a widely reported shouting match between General Powell and U.S. ambassador to the United Nations Madeleine Albright in 1993, in which Albright reportedly reacted hotly to Powell’s reluctance to use U.S. troops in the Balkans. Said Albright, “What’s the point of having this superb military you’re always talking about if we can’t use it?” Powell reportedly replied, “I thought I would have an aneurysm . . . American GIs are not toy soldiers to be moved around on some global game board.” As quoted in the Observer, “Reluctant Warrior,” September 30, 2001. 23. See Michael Walzer, Just and Unjust Wars: A Moral Argument with Historical Illustrations (New York: Basic Books, 1992), 185. 24. Susan L. Marquis, Unconventional Warfare: Rebuilding US Special Operations Forces (Washington, DC: Brookings Institution Press, 1997). 25. Arreguín-Toft, How the Weak Win Wars, chapter 2.

unconventional deterrence

221

26. Samuel P. Huntington, “The Clash of Civilizations,” Foreign Affairs 72 (Summer 1993): 41. 27. Small arms, such as rifles, do not carry as much cultural baggage as heavy weapons and combat aircraft, which cannot be decoupled from such baggage. See Christopher S. Parker, “New Weapons for Old Problems: Conventional Proliferation and Military Effectiveness in Developing States,” International Security 23 (Spring 1999): 119–47. 28. Note that in this context, “defeat” must be limited—as it increasingly has been in fact—to defeat of the adversary’s armed forces, followed by political concessions, but never occupation. The George H. W. Bush administration understood quite clearly that defeat of Saddam Hussein’s armed forces in 1991 would not translate into a low-cost postwar occupation: “Would [occupying Baghdad to capture Saddam Hussein] have been worth the inevitable follow-up: major occupation forces in Iraq for years to come and a very expensive and complex American proconsulship in Baghdad? Fortunately for America, reasonable people at the time thought not. They still do.” Powell, “U.S. Forces: Challenges Ahead,” 38. 29. Marquis, Unconventional Warfare. 30. See Arreguín-Toft, How the Weak Win Wars, 121–22. 31. See the Observer, “Reluctant Warrior.” 32. See, e.g., Marlise Simons, “Militia Leader Arkan Is Indicted in War Crimes; Yugoslavs are Warned,” New York Times, April 1, 1999. 33. Here one might object that the key issue was not the costs of a potential intervention so much as the lack of benefits, so that deterrence did not come into play. But we may question whether acting in accord with U.S. national identity—deeply invested in a self-image as a defender of the weak and helpless and committed to intervening to halt genocide—might not have counted as an important benefit (albeit an intangible one) in its own right. 34. It has often been said that the apparently myopic focus on major conventional or nuclear war—especially by military elites—represents a kind of organizational and institutional woodenheadedness, or inappropriate connections between a defense industry and politics. Yet although such arguments have merit, in order for them to be decisive, proponents would have to be willing and able to show that the likelihood of a future major conventional or nuclear war was zero. Even a highly advanced and industrialized state will require some time to mobilize to meet a crisis, and given that historically the only kinds of crises capable of truly threatening a state’s survival come from other states, it may not be inappropriate to maintain considerable major conventional or nuclear war capabilities even at a time, such as now, when the likelihood of engaging in such a conflict seems very low. 35. On the debate over whether third world contingencies should engage the interests of advanced industrial states—and in particular, the United States—see Michael C. Desch, “The Keys That Lock Up the World: Identifying American Interests in the Periphery,” International Security 14, (Summer 1989): 86–121; and Steven R. David, “Why the Third World Still Matters,” International Security 17 (Winter 1992 / 93): 127–59. 36. See, e.g. Caleb Carr, The Lessons of Terror (New York: Random House, 2002); and Ivan Arreguín-Toft, “The [F]utility of Barbarism: Assessing the Impact of the Systematic Harm of Noncombatants in War” (paper presented at the Annual Convention of the American Political Science Association, Philadelphia, PA, August 2003). 37. Powell, “U.S. Forces: Challenges Ahead.” 38. Samantha Power, “Bystanders to Genocide: Why the United States Let the Rwanda Tragedy Happen,” Atlantic Monthly (September 2001): 90. 39. Merom, How Democracies Lose Small Wars. 40. See James T. Quinlivan, “Force Requirements in Stability Operations,” Parameters (Winter 1995): 59–69. 41. Richard Falk, “Hard Choices and Tragic Dilemmas,” Nation (December 1993): 757.

chapter ten R

Deterrence and Compellence in Iraq, 1991–2003 Lessons for a Complex Paradigm f r a n k p. h a r v e y a n d p a t r i c k j a m e s

this chapter tests rational deterrence and compellence theory based on evidence from six major exchanges that occurred from 1991 to 2003 involving the United States (with support from Britain and the United Nations) and Iraq. The study is based on the preinsurgency era, but it may have implications for present policy making. Dissecting the United States–Iraq rivalry reveals a series of separate (although interrelated) encounters that highlight the crucial importance of specifying the precise time frame and exact sequence within which appropriately designated deterrent or compellent threats, countercoercive strategies, and retaliatory responses unfold over time. By identifying discrete exchanges, counterproductive debates over coding successes and failures can be replaced by a more relevant level of analysis. Most wars or crises rarely encompass a single, dominant encounter—they typically include multiple exchanges that provide theoretically relevant information about the core prerequisites for testing deterrence success and failure. In contrast to conventional approaches to testing,1 the coding decisions that are applied in this chapter do not require compressing all relevant information from these exchanges into a single data point (e.g., Cuba, 1962; Six Day War, 1967; Yom Kippur, 1973). One implication of this reassessment of the United States–led coalition’s exchanges with Iraq is that complex, multidimensional encounters may represent the norm rather than the exception in the history of coercive diplomacy. Some crises encompass both deterrence and compellence encounters and both successes and failures. Complexity is the basic characteristic of the encounters explored in this chapter. Some conflicts are recurrent, and 222

deterrence and compellence in iraq

223

their complexity is reflected in the stages of both escalation and deescalation within a given crisis and the connections across the discrete events that are often given a single designation by historians and political scientists alike. T. V. Paul, in the introduction to this volume, identifies five “ideal relationships” vis-à-vis deterrence: between great powers; between nuclear states; between nuclear great powers and regional powers armed with weapons of mass destruction (WMD); between nuclear states and nonstate actors; and between collective actors and target states of concern. The events covered by the present study fall within the third category; their aftermath—a chaotic occupation and insurgency—belongs in the fourth. Dissecting a crisis or military encounter into its constituent events can also help to answer important questions about deterrence stability in the same case over time. Deterrent failures at one stage of a conflict, for instance, can provide relevant information for interpreting and explaining successes later on because credibility and resolve are typically reinforced after failures. Success may also breed complacency and discourage innovation on the part of the victor. Our approach to assessing the success of deterrent threats also helps to disentangle the relevance of deterrence prerequisites as they interact in different contexts throughout the same case. Instead of assuming that the presence of resolve and capabilities are always positive forces when combined, a protracted crisis framework helps to identify different ways in which the presence or absence of prerequisites matter. The chapter thus explores the relationship between context and causation, a subject that remains relatively unexplored in the literature on coercive diplomacy.2 By expanding the pool of evidence appropriate for testing, the protracted crisis approach adopted in this analysis will provide a more relevant, accurate, and complete body of information for evaluating deterrence and compellence theory. This chapter unfolds in four stages. It first outlines a framework of analysis for deterrence theory. Second, deterrence and compellence are assessed over six exchanges between the United States and Britain (and loosely associated allies) on one hand and Iraq on the other. Third, it offers a series of observations derived from the preceding exchanges. Fourth, it offers some conclusions about the viability of the deterrence paradigm.

f ra m e w o r k o f a n a lys i s The standard, four-condition defender-oriented model of rational deterrence theory stipulates that a retaliatory threat should succeed in preventing a

224

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

challenge if leaders of the deterring state (A) clearly define the behavior deemed to be unacceptable, (B) communicate to challengers a commitment to punish violations, (C) possess the means (capability) to defend the commitment by punishing adversaries who challenge it or by denying the challenger the specific objectives sought through its aggression, and (D) demonstrate their resolve to carry out the retaliation if the challenger fails to comply.3 If these four prerequisites are satisfied, the expected net costs of the threatened sanction (to the challenger) should be greater than the expected net gains from noncompliance because the punishment (if carried out) would prevent the challenger from achieving intended military, political, or economic goals.4

deterrence / compellence exchanges The six exchanges described below occurred after the end of major military hostilities in the first Gulf war in 1991 and just before the onset of major hostilities in 2003. Initiation and termination dates for all of these exchanges are assigned as precisely as possible. A brief description of the events leading up to each exchange is followed by a narrative that conveys the deterrence or compellence threats, probes, and responses that came to characterize the event. Discussion then turns to outcome in terms of the success or failure of the deterrent and compellent threats that occurred. Exchange 1: April 10–24, 1991 The Establishment of a Northern No-Fly Zone in Iraq: Operation Provide Comfort (Operation Northern Watch) The creation of a no-fly zone (NFZ) in Iraq resulted from the United Nations Security Council Resolution (UNSCR) 688, on April 5, 1991, which stated that the Security Council condemned the repression of the Iraqi civilian population in many parts of Iraq and demanded that Iraq put an end to its domestic repression and open a dialogue to ensure that the human and political rights of all Iraqi citizens were respected.5 The United States ordered Iraq to end all military activity in northern Iraq. U.S. officials established an NFZ to prevent Iraqi aircraft from attacking the Kurdish population. U.S., British, and French planes patrolled Iraqi air space north of the thirty-sixth parallel in a deployment the U.S. military dubbed “Operation Provide Comfort.”6 Operating from Incirlik in eastern Turkey, coalition aircraft had a

deterrence and compellence in iraq

225

relatively short transit to enforce the NFZ above the thirty-sixth parallel. The NFZ also provided air cover for ground forces and humanitarian assistance to the Kurds, who remained fearful about another Iraqi military incursion. From April to September 1991, the operation led to over 40,000 sorties, helping to protect the relocation of over 700,000 refugees and the restoration of nearly 80 percent of the villages destroyed by the Iraqis. Deterrence / compellence threats, probes, and responses. On April 10, 1991, the Iraqi government rejected a European plan to establish a Kurdish safe haven in the northern Iraq.7 On April 11, in response to this Iraqi action, the George H. W. Bush administration issued a warning to the Iraqi government not to use military force in northeastern Iraq—a message directly conveyed to Iraqi officials in Washington and at United Nations (UN) headquarters in New York.8 This warning was characterized by the New York Times as an explicit demand; the Bush administration made it very clear that “no land or air forces would be allowed to function in the area involved,” with an explicit threat to fire on helicopters or planes approaching Kurdish areas. But the credibility and resolve of U.S. and European threats to prevent Iraqi ground forces from attacking Kurdish areas was seriously weakened by subsequent statements by the President on April 11: I am not going to involve any American troops in a civil war in Iraq . . . we are going to do what is right by these refugees, and I think the American people expect that, and they want that. But I don’t think they want to see us bogged down in a civil war by sending in the 82nd Airborne or the 101st or the 7th Cavalry. And so, I want to get that matter cleared up.9

On April 12, following this weak threat, Iraqi forces launched attacks on Kurdish forces north of the thirty-sixth parallel, in direct violation of U.S. warnings.10 This Iraqi action was consistent with deterrence theory—a watered-down threat apparently failed to make much of an impression on the Ba’athist regime. Other Iraqi attacks reportedly occurred in Kurdish areas south of the parallel, including Sulaymaniyah.11 Bush administration spokesmen did not confirm the reports, adding that there had been no attempts by Iraq to impede the refugee relief operations for the Kurds in northern Iraq. On April 14, President Bush reaffirmed his reluctance to “interfere” in Iraq’s civil war.12 Consistent with findings from case studies of Bosnia and Kosovo,13 weak threats are often worse than no threat at all. Like Milosevic, Saddam often

226

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

exploited signs of weaknesses and used these occasions as an opportunity to demonstrate relevance and control to domestic and regional constituencies. As security concerns for Kurdish civilians grew, however, a stronger, more credible retaliatory threat was issued by the president on April 16, 1991, in an effort to deter further Iraqi incursions into Kurdish territory. The president announced a joint U.S.-European plan to send U.S. troops into northern Iraq to build refugee camps and guarantee protection for the Kurds.14 This commitment superseded the rhetoric about limiting U.S. involvement and dramatically upgraded the credibility and potency of the coalition threat to Saddam Hussein. Coalition solidarity bolstered U.S. deterrent threats, temporarily removing an important element of complexity that often is an obstacle to deterrence and compellence in other contexts. Outcome. As predicted by traditional deterrence theory, the revised threat issued by the United States on April 16 produced results. On April 18, 1991, Iraq and the UN signed a Memorandum of Understanding (MOU) providing for a UN humanitarian presence in Iraq, which included the stationing of five hundred UN security guards in northern Iraq to protect the relief operations. This agreement was to remain in force until December 31, 1991, and carried with it the possibility for renewal.15 While Saddam Hussein responded with statements that “the Western [U.S.] relief operation was ‘not necessary in practical terms’ because Baghdad and the United Nations had reached agreement in principle on easing the crisis,” nevertheless, Iraqi troops satisfied the condition stipulated in the MOU on April 21. Approximately two hundred Iraqi policemen entered the Kurdish town of Zakho as Iraqi soldiers withdrew, a direct result of the agreement between the allies and the Iraqis to establish a safe haven for the Kurds.16 Several senior U.S. military and political officials interpreted the stationing of Iraq police units as an initiative that undermined the spirit of the agreement. They viewed this as an all too common probing effort by the Iraqi regime to test the resolve of coalition forces.17 Additional probing efforts followed on April 22 as Iraqi troops, which had withdrawn from Zakho under the agreement with the U.S. military, returned to the outskirts of Zakho. This Iraqi troop deployment deterred many Kurds from returning to their homes. Iraqis in and around Zakho were generally viewed as friendly, with a preference to avoid confrontation with the Kurds in the area. Some Iraqi troops overlooking Zakho, however, did fire machine gun and mortar rounds in an effort to harass the Kurdish population.18 On April 24, and in direct response to additional U.S. warnings, Iraqi offi-

deterrence and compellence in iraq

227

cials agreed to move their troops several kilometers away from Zakho. Iraqi Kurdish leaders announced their agreement in principle with the Iraqi government on a formula to allow them autonomy in northern Iraq. In an obvious public relations move, Iraq welcomed several foreign journalists into Kurdistan to witness the conditions facing the local population.19 The Bush administration followed these events with an additional diplomatic and military warning, ordering Iraq to withdraw all police and military forces from the area near the Kurdish refugee camp in northern Iraq. A similar note was presented in the UN.20 Consistent with Paul’s observation in the introduction to this volume that deterrence can operate between both parties in a given incident, the circumscribed and largely symbolic actions from Saddam Hussein served notice regarding further interactions. Specifically, Iraq hoped to deter further encroachments upon its sovereignty, while simultaneously being deterred itself from directly challenging the terms of the distasteful agreement it had just made with the United States. Exchange 2: August 26, 1992, to January 19, 1993 Establishment of Southern No-Fly Zone: Operation Southern Watch On August 26, 1992, the United States, Britain, and France established an NFZ in southern Iraq along the thirty-second parallel. President George H. W. Bush stated that the NFZ was a direct response to “Saddam’s use of helicopters and . . . fixed-wing aircraft to bomb and strafe civilians and villages” in southern Iraq, a clear indication, he added, that Iraq was “failing to meet its obligations under United Nations Security Council Resolution 688 . . . demand[ing] [that] Saddam Hussein end repression of the Iraqi people.”21 Iraq was given twenty-four hours to comply with the NFZ. With the president’s announcement, U.S. Central Command activated Joint Task Force– Southwest Asia ( JTF-SWA), a command-and-control unit for coalition forces monitoring the NFZ consisting of American, British, French, and Saudi Arabian units. JTF-SWA’s mission, dubbed Operation Southern Watch, was intended to monitor the NFZ and to respond if Iraqi forces infiltrated the area. While Baghdad vowed to eliminate the NFZ through the use of force, the Pentagon reported that Iraqi forces had begun removing their helicopters and airplanes north of the thirty-second parallel by the evening of August 26, 1991.22

228

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

Deterrence / compellence threats, probes, and responses. JTF-SWA coalition forces flew their first Operation Southern Watch sortie within twenty-four hours of the NFZ activation and reported no Iraqi activity. In fact, for the next several months, Iraq complied completely with the no-fly restrictions and coalition forces controlled the airspace south of the thirty-second parallel. Tensions between Iraq and the coalition over the southern NFZ flared up in December 1992. On the morning of December 27, 1992, while flying a routine mission in the NFZ, U.S. Air Force F-15s encountered a pair of Iraqi MiG-25s flying south of the thirty-second parallel. They were warned by the U.S. patrol to return north of the NFZ. The MiGs did so, but a second set of Iraqi MiG-25s were intercepted by two U.S. Air Force F-16s later in the day. One of the Iraqi MiGs, after having locked its air-to-air radar on one of the F-16s, was shot down.23 American officials stated that Baghdad’s decision to breach the thirty-second parallel was a deliberate move by Iraq to “test the willingness of the United States” to uphold the NFZ, especially in light of American involvement in Somalia and Bosnia, the relocation of the USS Ranger—an American aircraft carrier central to enforcing the NFZ—from the Persian Gulf to the waters off Somalia, and the upcoming transfer of power to the newly elected president Bill Clinton.24 In an effort to counter a seemingly emboldened Iraqi policy regarding the southern NFZ, Washington sent two additional squadrons of navy warplanes to the region on December 28, 1992, to augment coalition forces already enforcing the NFZ.25 While Iraqi air incursions into the NFZ abated slightly in the days that followed, by the beginning of January 1993, Iraqi ground forces had begun deploying surface-to-air missiles (SAMs) south of the thirty-second parallel that could be used destroy coalition airplanes patrolling the zone. Thus, the United States–led deterrence effort that was based on the NFZ was met by an Iraqi military effort to compel the coalition to abandon its policy of protecting the Shia by denying Baghdad sovereignty over its airspace. On January 6, 1993, an ultimatum was presented to the Iraqi representative to the UN by American, British, French, and Russian representatives. The ultimatum read, “All SAM systems which have been moved into new positions south of the 32nd parallel should be returned to their original sites or configurations within 48 hours of the delivery of the demarche.” This ultimatum warned that if Iraq did not remove the missiles from the southern edge of the NFZ, coalition forces would respond “appropriately and decisively.”26 Initially, Iraqi officials rebuffed the coalition’s ultimatum and continued to voice their displeasure with the demands of the United

deterrence and compellence in iraq

229

States and its allies. Yet, despite Iraq’s defiant rhetoric, continued political pressure by the United States and the UN Security Council, along with a doubling of JTF-SWA flights over the NFZ on January 8, 1993, compelled Iraq to dismantle and remove some missile batteries in the NFZ before the deadline in the coalition’s ultimatum. By January 9, White House officials asserted that Iraq had indeed “backed down” in the face of allied demands and had removed several SAM sites north of the thirty-second parallel.27 While the immediate confrontation was defused, Iraq continued its provocative actions on January 11, 1993, by relocating several dismantled SAM sites to the northern NFZ and taking the unprecedented step of sending several hundred armed Iraqi soldiers across the border into Kuwaiti territory, apparently to retrieve six antiship missiles left behind when Iraqi forces were driven from Kuwait in 1991. Both the United States and the UN Security Council strongly condemned the incursion, warning of “strong consequences” if any further action was taken on the border.28 Despite this warning, a second Iraqi incursion into Kuwait took place on January 12, 1993. This second incursion, and Iraq’s continued ambivalence over removing the remaining SAM sites from the NFZ, prompted a strong and vocal response from the Bush administration. Pentagon officials reported that a plan to attack Iraq had been adopted “in principle,” adding that it was “just a matter of when to pull the trigger.” British and French officials further reiterated the American report that agreement had been reached to respond to these Iraqi provocations.29 On January 13, 1993, more than one hundred American, French, and British strike aircraft launched a thirty-minute bombing raid on Iraqi targets inside the NFZ. This raid was the first large-scale attack since the cessation of conflict in 1991. It destroyed various missile sites, radar units, and lowlevel command bunkers.30 In a move that was designed specifically to deter further Iraqi incursions, a battalion task force of 1,250 American troops also was dispatched to the Kuwait-Iraq border.31 Officials in Washington, Paris, and London all expressed their resolve to uphold the NFZ and warned that further military action against Iraq would be taken if Baghdad continued to violate UN-sanctioned restrictions by following what appeared to be a “cheat and retreat” strategy.32 In the days immediately following the bombing raid, further warnings of renewed hostilities were issued by allied officials. These warnings were backed up by action. Forty-five Tomahawk cruise missiles were launched against the Saafaraniay industrial complex near Baghdad on January 17, 1993, and a daytime air raid against SAM sites

230

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

in the NFZ, involving American and British fighter planes, took place on the following day.33 Outcome. Both raids were of limited military scope but represented a strong political message to Iraq that allied forces were prepared to continue enforcing the UN mandate. On January 19, 1993, in an apparent policy reversal, Iraq announced that it had ordered a cease fire against allied jets in the NFZ.34 Although further small-scale attacks, mostly against Iraqi fighter-jets and SAM sites, continued in the southern NFZ in June and July of 1993, the January Crisis ended with an allied victory. Iraq was compelled to accept the southern NFZ and was ultimately deterred from continuing its attacks against Shiites in southern Iraq. Iraq’s reaction to the January 1993 air strikes also initiated a pattern of cyclical response that was repeated in subsequent confrontations with allied forces. Iraqi actions began to follow a predictable a tit-for-tat pattern: (1) Iraqi officials initiated an interaction with rhetorical rejections of UN resolutions that resulted in eventual policy moves that challenged U.S. resolve; (2) Iraqi forces would then hunker down while under allied attack; (3) Iraqi officials would then announce a policy reversal that would acquiesce to Western demands. This cycle would then start anew with an Iraqi rejection of some other UN demand or condition. This pattern is consistent with the presence of escalation dominance—a standard concept in the lexicon of deterrence theory—in favor of the United States–led coalition. In other words, Iraqi officials would continuously test the coalition’s resolve, backing down when it became apparent that they were about to encounter the full force of allied military power. It also reinforces the continuing relevance of the rationality axiom even under conditions of multistage, multiactor complex deterrence. Iraq backed off from extreme confrontation in the face of clear and powerful threats; when the message began to fade, it started to probe again and test for new limits. Exchange 3: October 7–16, 1994 Iraqi Aggression against Kuwait: Operation Vigilant Warrior In October 1994, renewed Iraqi aggression against Kuwait resulted in a brief yet significant international crisis in the Persian Gulf. Operation Vigilant Warrior, the coalition operation launched in response to Iraq’s provocation, was a direct result of Baghdad’s apparent hostility toward Kuwait. On October 7, 1994, Iraq repositioned an estimated 10,000 troops, includ-

deterrence and compellence in iraq

231

ing a contingent of elite Republican Guard, to the southern town of Basra, thirty kilometers from the Kuwaiti border. Pentagon officials reported that the deployment brought the number of Iraqi troops on the border with Kuwait to approximately 50,000.35 Washington’s reaction was firm, with President Bill Clinton taking immediate “precautionary steps” in the Gulf region and further warning Iraq, in a White House press release, that “it would be a grave mistake for Saddam Hussein to believe for any reason that the United States would have weakened its resolve” on issues similar to those that had developed in 1990.36 Several U.S. officials conceded that the Iraqi deployment was probably saber rattling intended to underscore Baghdad’s demands for a lifting of UNapproved economic sanctions and elicit support from members of the Security Council—notably Turkey, France, and Russia, all of which had already voiced concerns about sanctions. Tariq Aziz, Iraq’s deputy prime minister, indicated as much in a speech that day to the General Assembly, where he called on the UN to lift sanctions that had caused “unbearable hardships” to Iraqis.37 Thus, accompanying the overarching coalition effort to deter Iraqi aggression against Kuwait is a limited Iraqi attempt at compellence in terms of generating pressure to lift UN sanctions against the Ba’athist regime in Baghdad. Iraqi saber rattling was intended to force an end to the sanctions that were damaging the Iraqi economy. Deterrence / compellence threats, probes, and responses. American resolve was reaffirmed in a speech given by Madeleine Albright, the chief American delegate to the UN, responding to Aziz’s assertions, further warning Iraq against pursuing aggressive action against Kuwait.38 In addition to these verbal warnings, President Clinton ordered the aircraft carrier George Washington to set sail from its position in the Adriatic to the Red Sea, instructed a marine amphibious group to leave the United Arab Emirates and steam toward the northern Gulf, ordered the deployment of maritime prepositioning ships toward the area from the American naval base at Diego Garcia in the Indian Ocean, and stepped up reconnaissance flights over the southern NFZ.39 Further American signals were given on October 8, 1994, when 4,000 U.S. Army troops were ordered to leave for Kuwait. President Clinton reiterated his earlier resolve, stating to the press that he wanted “to make it clear . . . [that] it would be a grave error for Iraq to repeat the mistakes of the past or to misjudge either American will or American power.”40 U.S. secretary of state Warren Christopher also was dispatched to the Middle East, stating to

232

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

reporters that Saddam Hussein “must not be allowed to intimidate either Iraq’s neighbors or the United Nations Security Council.”41 The president of the UN Security Council also expressed his disapproval of the Iraqi troop movements and reaffirmed the UN’s readiness to uphold the “sovereignty and territorial integrity of Kuwait.”42 Ignoring these initial statements and warnings, Iraq continued to reinforce its military presence on the border, sending in a third division of elite Republican Guards to the area, increasing the total number of Iraqi troops on the border with Kuwait to 80,000. Many American officials believed that this Iraqi force was capable of staging an invasion of Kuwait “in a matter of days.”43 Pentagon officials called the Iraqi buildup “irresponsible and dangerous” and a “serious challenge” to regional stability.44 As the gravity of the Iraqi buildup became clear, and as an indication of how seriously Washington regarded these developments, President Clinton dispatched 36,000 troops along with fifty-one additional combat planes to Kuwait on October 9, 1994.45 This third exchange between Iraq and the coalition was characterized by “credibility inflation.” Compared with early exchanges, the number of forces wielded to demonstrate interests and intentions was greatly increasing. Large bodies of troops and aircraft operating in close proximity to each other create a greater threat of inadvertent escalation and greater consequences if hostilities somehow manage to erupt in the absence of direct political authorization. In making these increasingly large deployments, each side conveyed its willingness to go closer to the brink of war and to endure greater risk than in past exchanges. Outcome. On October 10, Iraq’s ambassador to the UN, Nizar Hamdoon, told the General Assembly that Iraqi troops had been given the order to withdraw from the border with Kuwait.46 Because his statement could not be immediately confirmed, the remark was met with skepticism. In a television address from the Oval Office that evening, President Clinton reaffirmed that the United States was “committed to defending the integrity of [Kuwait] and to protecting the stability of the gulf region.” He also confirmed the deployment of another 350 Air Force aircraft to the Persian Gulf region.47 Sensing that the immediate crisis had been averted, a debate about how to “contain” Iraqi aggression in the future began in the White House. The idea of establishing a southern containment zone in Iraq apparently emerged from these deliberations.48 Confirmation that Iraqi troops had in fact begun moving north from their earlier positions near Basra came on October 11, and the confrontation be-

deterrence and compellence in iraq

233

tween Iraq and the United States eased somewhat.49 Nonetheless, U.S. deployment to the Gulf region continued—and a further 155,000 U.S. ground troops were put on alert—amid a flurry of discussion regarding a U.S. plan for the future containment of Iraq. Clinton administration officials commented that the United States was considering establishing a buffer zone in southern Iraq, the territorial equivalent of the southern NFZ, that would bar Iraq from moving its elite ground forces within striking range of Kuwait.50 To that end, Ambassador Albright held discussions, on October 12, with her counterparts in the Security Council about establishing an exclusion zone that would restrict Iraq’s Republican Guard from deployment into southern Iraq.51 During these Security Council talks, the first visible fissures between the allies became apparent, with France suggesting that American deployment to the Gulf was “not unconnected with domestic politics” and Russia announcing that Foreign Minister Andrei V. Kozyrev would fly to Baghdad on an official state visit.52 With support for the containment zone plan dwindling both within the United States53 and within the Security Council, American Defense Secretary William J. Perry announced on October 13 a revised plan to deter Iraq that involved increasing “the American military presence in the Persian Gulf region.” This plan included the positioning of American warplanes and a division’s worth of tanks and armor on the borders of Iraq— mainly in Saudi Arabia, the United Arab Emirates, and Kuwait—even after the immediate crisis passed.54 Perry also commented on an American proposal to the Security Council, to be officially tabled within a few days, that would condemn Iraq’s recent deployment and acts of aggression. American troop deployments to the Gulf region continued, and a small-scale military exercise, involving 125 U.S. Marines—accompanied by an “army of international photographers and reporters”—began in Kuwait.55 One way to interpret the French and Russian statements regarding the October 1994 crisis is in connection with the mounting costs of commitment and danger of escalation, especially given the provocative Iraqi behavior. French and Russian equivocation suggests at least temporarily successful deterrence by Iraq, which all parties believed to possess WMD and which seemed to be more willing than ever to go to the brink against the United States, its allies, and the UN. Even though they ultimately voted with the United States on the Security Council, France and Russia retreated into the background in terms of the diplomatic exchanges related to the October 1994 crisis. All of this follows the pattern described in the introduction to this book: multiple interactions among different agents producing a situation best described as complex deterrence. Exchange 3 is archetypal in that

234

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

sense, with deterrence and compellence initiatives being undertaken by both parties in the international interaction. On October 14, 1994, the Clinton administration, in conjunction with the British government, formally introduced a Security Council resolution and, ignoring Russian pleas for delay, pressed for a “quick vote” on it. The vote took place on October 15 and UNSCR 949 passed unanimously.56 The wording of UNSCR 949 clearly stated what was expected of Iraq. The Security Council condemned recent military deployment by Iraq toward Kuwait, demanded that Iraq continue to “withdraw its troops to their original position,” and demanded that Iraq “not redeploy to the south” or use its forces in any “hostile or provocative manner.”57 On October 16, government-controlled news media in Iraq indicated that UNSCR 949 had been accepted by Baghdad and that the government would comply with its demands to withdraw elite forces from southern Iraq. In an unusual move, all the details of the resolution were read aloud on Iraqi radio and television news bulletins, further signaling to the allies that Iraq would comply with their demands.58 This action effectively and intentionally substituted for the naturally credible signaling that would have taken place in a regime with open media. Iraq’s compliance with UNSCR 949 marks the last phase of October 1994 crisis. Exchange 4: August 31 to September 13, 1996 The Irbil Offensive: Operation Desert Strike On August 30, 1996, U.S. military officials in the Gulf region reported observing renewed mobilization and the movement of Iraqi troops toward Kurdish districts in northern Iraq.59 Clinton administration officials responded with warnings that “we will consider any aggression by Iraq to be a matter of very grave concern.”60 The administration’s objective was to send a strong signal to deter Iraq from crossing the so-called exclusion zone. Glyn Davis, a State Department spokesman, reiterated the threat, stating, “we would view any aggressive moves by Iraq with utmost seriousness. The Iraqis are in no doubt of our views on this.”61 Other media reports confirmed Pentagon plans to deploy additional warplanes to the region if Iraq violated the Security Council exclusion zone resolution and began, once again, to move forces in preparation for retaliation against any Iraqi incursions into northern Iraq.62 On August 31, a 30,000 to 40,000 strong Iraqi armored force invaded the

deterrence and compellence in iraq

235

Kurdish provisional capital, Irbil.63 In the view of most observers, Saddam Hussein was repeating previous patterns of probing to test U.S. and coalition resolve.64 In this case, however, the seizure of Irbil by the Patriotic Union of Kurdistan posed a more direct challenge to Iraqi sovereignty and status. It therefore would be more difficult for the United States to mount a sufficiently capable and costly threat to deter Iraqi forces from responding to this incursion. In addition, the credibility of the retaliatory threats was becoming less potent (and, therefore, less effective and more difficult to mount) as France and Russia became increasingly vocal about their opposition to the sanctions regime and to U.S.–United Kingdom (UK) military action. These French and Russian signals were continuously exploited by Saddam Hussein in ways that made it more difficult for deterrence to succeed. This Russian, French, and Iraqi interaction became increasingly more challenging (and detrimental) over time because it undermined the credibility of U.S. threats. This pattern helps to explain the last coercive diplomatic failure in the period preceding the 2003 Iraq war—a point that will be addressed in more detail in the overview of exchange 6. Deterrence / compellence threats, probes, and responses. On September 1, 1996, the Clinton administration issued several explicit military threats to punish the Iraqi regime for its blatant violation of multiple Security Council resolutions.65 These threats were followed by action. On September 3, U.S. forces launched twenty-seven cruise missiles at military targets in southern Iraq. Fourteen missiles were fired from U.S. warships in the Persian Gulf, and thirteen missiles were fired from two B-52 bombers flying from bases in Guam.66 President Clinton’s speech following the attacks confirmed the U.S. intention to pressure Iraqi forces to retreat: “I want to reaffirm to you what I said to all the American people early this morning: Our objectives there are limited, but our interests are clear . . . These steps are being taken to further all of these objectives and the policy of containing Iraq that I have pursued for four years now, and it was developed before me under President Bush.”67 President Clinton followed this speech with an official statement on September 3: “Earlier today I ordered American forces to strike Iraq. Our missiles sent the following message to Saddam Hussein: . . . to make Saddam pay a price for the latest act of brutality, reducing his ability to threaten his neighbors and America’s interests.”68 These efforts produced limited diplomatic and military support from key allies.69 Russian officials called the United States–led military operation “an inappropriate and unacceptable reaction to the latest events in northern

236

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

Iraq.”70 The reception at the UN was even colder.71 These fissures within the coalition led Saddam Hussein to dismiss the Security Council resolutions and to issue several countercoercive threats by promising to down any foreign aircraft flying over Iraq territory.72 Defense Secretary William J. Perry expressed Washington’s concerns about the newest challenge created by Baghdad: “the issue is not simply the Iraqi attack on Irbil . . . our concern is that if Saddam Hussein is emboldened by what he would see as a success in the north he might strike out in areas which are of greater strategic importance to him as well as to us in the south.”73 In response, on September 4, 1996, President Clinton announced his intention to widen the southern NFZ, extending it northward from the thirty-second to the thirty-third parallel. He also rejected calls for a limited oil sale by Iraq (under a recent UN resolution) until Iraq abandoned its current policy of using military force against the Kurds.74 Outcome. The coalition’s military actions and diplomatic threats succeeded. On September 4, 1996, President Clinton announced that “our mission has been achieved . . . There has been a withdrawal of the forces, a dispersal of the forces . . .” and that Iraq had “pulled back most of its troops from a contested Kurdish enclave [and] United States–led air patrols met only token resistance enforcing a newly expanded prohibition on Iraqi military flights.”75 On September 13, the Iraqi government announced that it would no longer attack coalition fighters enforcing the flight-exclusion zone in northern and southern Iraq.76 Pentagon officials also reported that Iraq had halted repairs to its air-defense sites in response to specific U.S. demands.77 Debate over the success of the operation, however, lingered. Saddam maintained a strong military presence just outside the exclusion zone. Russia, France, and China also were beginning to reject virtually all U.S. and UK diplomatic and military strategies toward the region, a development that improved Baghdad’s ability to resist coalition pressure. These divisions within the Security Council over how to respond to Iraqi challenges would continue and grow deeper in the coming years.78 Exchange 5: July to December 1998 Operation Desert Fox On February 23, 1998, the UN secretary-general obtained an undertaking from the Iraqi government to resume full cooperation with the UN Spe-

deterrence and compellence in iraq

237

cial Commission on Iraq (UNSCOM) and the International Atomic Energy Agency (IAEA).79 An MOU was signed in Baghdad by the secretary-general and Tariq Aziz, whereby the government of Iraq reconfirmed “its acceptance of all relevant resolutions of the Security Council” and undertook “to accord UNSCOM and the IAEA immediate, unconditional and unrestricted access.”80 Under the terms of the MOU, a special group of senior diplomats was assigned to accompany UNSCOM and IAEA members during inspection of the eight disputed presidential sites. The MOU also noted that the lifting of sanctions was of “paramount importance to the people and Government of Iraq” and that the secretary-general undertook to “bring this matter to the full attention of the . . . Security Council.”81 The UN Security Council endorsed the MOU in UNSCR 115482 on March 2, 1998, and warned that any violation of the agreement would have the “severest consequences for Iraq.”83 Washington and London had pushed for a more explicit threat but met with resistance from the other permanent members of the Security Council. There also was a dispute within the council over the precise meaning of the term “severest consequences,” with the United States maintaining that the resolution provided the authority to act in the event of further Iraqi noncompliance. China insisted, however, that the resolution was “not an automatic green light to use force.”84 The UN secretary-general, Kofi Annan, also stated that the United States should hold “some sort of consultations with the other members [of the Security Council]” before any military action was launched.85 Reports that the UN was “divided over the Iraq issue,” emerged in a variety of media sources.86 The pattern of diplomatic activity within the Security Council was virtually identical to events preceding the 2003 Iraq war. On March 5, 1998, UNSCOM inspectors returned to Iraq and carried out a number of site inspections, without hindrance.87 Questions remained, however, about access to the presidential sites, especially after the initial inspections found that all the rooms targeted by UNSCOM inspectors had been stripped bare of files, personnel, and even furniture.88 The Iraqi government claimed it had not agreed to any further inspections of the sites, whereas inspectors believed the initial visits only served to establish the right of access prior to future visits. In spite of the declarations of cooperation, relations between UNSCOM and Baghdad began to deteriorate during late June and July 1998, following allegations that Iraq had sought to conceal the extent of its program to develop and weaponize the nerve agent VX. Iraq claimed the allegations were fabricated by UNSCOM in an attempt to delay the lifting of sanctions,

238

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

and Iraqi officials insisted they had provided all the necessary evidence that UNSCOM had demanded about their weapons programs.89 Talks between Baghdad and UNSCOM on the next stage of the inspection process broke down in early August 1998, despite indications from UNSCOM that work was almost complete on both the missile and chemical weapons files. On August 4, Tariq Aziz demanded that Richard Butler report to the UN Security Council that the disarmament process was complete, but the UNSCOM chief executive refused, saying he did not have sufficient evidence to make such a declaration.90 On August 5, 1998, Iraq announced it was suspending all cooperation with UNSCOM and the IAEA inspection teams and would restrict international monitoring activities to existing sites.91 In contrast to the rhetoric of the February 1998 crisis, the international response to the Iraqi decision was relatively muted. The UN Security Council declared the move to be “totally unacceptable,”92 but London and Washington sought to minimize the crisis, believing that Iraq had created the dispute to provoke a split in the Security Council over how to respond.93 On August 26, 1998, UNSCOM inspector Scott Ritter resigned in protest over what he perceived to be a weakening of U.S. and UK policy toward Iraq. Ritter, a former intelligence officer and member of the U.S. Marine Corps, joined UNSCOM in the autumn of 1991 and played a key role in developing a more proactive approach to UN arms inspection in Iraq.94 In his resignation letter, Ritter accused the United States and Britain of putting pressure on UNSCOM to abandon planned intrusive inspections of controversial sites so as to avoid a fresh confrontation with Iraq. He declared that the Security Council was propagating “the illusion of arms control,” which was “more dangerous than no arms control at all.”95 He also accused Kofi Annan of acting as a “sounding board” for Iraqi complaints and claimed that the failure of the Security Council to punish Iraq for its decision of August 5, 1998, to suspend cooperation with UNSCOM constituted “a surrender to the Iraqi leadership.” Ritter wrote that “the issue of immediate, unrestricted access is, in my opinion, the cornerstone of any viable inspection regime, and as such is an issue worth fighting for. Unfortunately, others do not share this opinion, including the Security Council and the United States.”96 U.S. secretary of state Madeleine Albright dismissed the allegations and claimed that Ritter did not “have a clue” about the overall policy toward Iraq, arguing that the United States had in fact been the “strongest supporter of UNSCOM.”97 U.S. officials admitted in private that Washington and London had sought to control the pace of confrontation with Iraq in order to keep other members

deterrence and compellence in iraq

239

of the Security Council on board. According to one Clinton administration official, “We’re learning that this is a marathon race and we cannot threaten military action every time Saddam Hussein rattles his cage.”98 The perception that the coalition was constrained in terms of making deterrent threats and taking military and diplomatic action in response to Iraqi provocations also represented a nonlinear effect produced by the complexity of multiactor, iterated deterrence and compellence. U.S. and UK officials realized that, with the rest of the UN Security Council off the bandwagon, pursuit of forceful and prompt compellent or deterrent initiatives toward Iraq no longer could take place with certainty, reducing their escalation dominance when it came to threatening Baghdad. In a number of interviews following his resignation, Ritter made additional allegations, saying he believed Kofi Annan had secretly promised Baghdad that intrusive inspections of the disputed “presidential sites” would take place only once within a four-month time frame. If true, such a deal would have effectively created safe havens for the Iraqis to hide weapons and material.99 It also would have undermined any threat emanating from the Security Council. Deterrence / compellence threats, probes, and responses. On September 9, 1998, the Security Council adopted UNSCR 1194,100 condemning the August 5 decision by Iraq to suspend cooperation with UNSCOM and the IAEA. The Security Council also decided not to conduct the semiannual review of sanctions scheduled for October 1998 or any further reviews until Iraq rescinded its decision.101 As a potential incentive to Baghdad, the resolution offered the prospect of a comprehensive review of the state of Iraqi compliance if cooperation resumed. Iraq was initially keen on the idea of a review, believing it would accelerate the lifting of sanctions by demonstrating that most of the requirements had been met. It soon became clear, however, that the United States had refused to include in the terms of the review any reference to paragraph 22 of UNSCR 687, which stipulated that the oil embargo would be lifted once Iraq was declared free of WMD.102 Baghdad declared the terms of the review to be too restrictive and refused to cooperate. Tariq Aziz said on September 28, 1998, “The idea of a comprehensive review is not a reward to be given to Iraq and then Iraq has to make concessions before such a review takes place. When the review is conducted in an honest and professional manner, we hope that it will lead to the lifting of sanctions.”103 In October, the UN assistant secretary-general and chief UN relief

240

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

coordinator for Iraq, Denis Halliday, resigned in protest over the impact of sanctions on the Iraqi population.104 The lack of blame attached to Saddam Hussein for any of the side effects from sanctions, such as the suffering and death of children, is instructive regarding the overall UN position as it evolved across the exchanges surveyed in this chapter. Kofi Annan held direct talks with Iraqi officials during early October in an attempt to resolve the impasse over weapons inspections but failed to make progress.105 The Iraqi ambassador to the UN, Nizar Hamdoon, reiterated Baghdad’s refusal to hand over a document sought by UN weapons inspectors.106 And on October 31, the Ba’ath party and Revolutionary Command Council declared an end to all cooperation with UNSCOM and the restriction of the IAEA to monitoring activities only. The UN continued efforts to defuse the crisis with Iraq.107 On November 2, British prime minister Tony Blair repeated his warning to Saddam Hussein, who had halted cooperation with UN weapons inspectors. Blair stated, “‘if the use of force is necessary that is the course that will be taken.’ . . . Defense Secretary George Robertson [had] already warned President Hussein to stop obstructing UN weapons inspections or ‘face the consequences’.”108 These threats were followed on November 5 by the passage of UNSCR 1205,109 which condemned Iraq’s decision to halt cooperation with Security Council resolutions as a “flagrant violation of resolution 687 [the cease-fire resolution of 1991] and other relevant resolutions,” and demanded that Iraq provide “immediate, complete and unconditional co-operation” with UNSCOM and the IAEA.110 The possibility of a comprehensive review of sanctions was kept open to encourage Iraq to comply with the UN. The threat of military action was reinforced on November 11, when it was announced that all UNSCOM personnel had been withdrawn from Iraq on the recommendation of the United States.111 The secretary-general and Arab leaders put additional pressure on Saddam by urging Iraq to find a diplomatic solution to the crisis. This was followed by a statement by eight Arab states, including Egypt, Saudi Arabia, and Syria, warning President Saddam Hussein that he would be “solely responsible” for the consequences of his noncompliance.112 This introduction of new actors into the confrontation— including several with established anti-U.S. and anti-UK reputations—represented a qualitative change at both psychological and material levels for Saddam Hussein. The strong statement of the Arab states was a credible voice of international disapproval of Baghdad’s policies and warned of Iraq’s isolation in the event of escalation of the crisis. In a Veterans Day speech at Arlington National Cemetery, President Clinton made clear that U.S. military action was a possibility if the weapons inspectors

deterrence and compellence in iraq

241

were not allowed to carry out their work.113 The final component of the threat was issued on the afternoon of November 14: the American and British Governments authorized “substantial military action” against Iraq as a wave of strike aircraft was launched. Simultaneously, the special representative of the UN secretary-general in Baghdad, Prakash Shah, was involved in a last-ditch attempt to find a diplomatic solution by providing Iraq with a face-saving formula to deescalate the crisis. Although Kofi Annan could not personally attempt another mediation mission to Baghdad without Security Council authorization, he insisted on sending a personal letter to the regime in Baghdad through his special representative, stating his conviction that “Iraq [should] be allowed to join the community of nations free of sanctions.”114 Outcome. Just hours before the first strikes were about to land, the response from Baghdad arrived. The Ba’athists declared that they were willing to comply with UN demands. Upon receiving further clarification from Iraqi officials concerning their agreement to comply unconditionally with Security Council resolutions and to rescind their earlier decisions to halt cooperation, the United States and Britain called off the strikes but warned that their forces would remain ready to act. UNSCOM inspectors returned to Iraq on November 17, 1998. The UNSCOM executive chairman, Richard Butler, stated that full Iraqi cooperation would enable the inspectors to complete their work on chemical weapons inside two to three months, before long-term monitoring could be put in place. Unfortunately, this apparent coalition victory in the war of wills with Iraq was short-lived. On December 8, 1998, Butler declared Iraqi officials were still impeding inspections.115 In his official report submitted on December 15, 1998, the executive chairman of UNSCOM wrote to the UN secretary-general with an updated report on the state of Iraqi compliance since the resumption of cooperation in mid-November.116 This report, which came to be known as the “Butler report,” asserted that Iraq’s claims to have fulfilled its disarmament obligations could not be accepted without further verification. It stated that the Iraqi government had provided some clarifications sought by the commission, but that in general Iraq had “not provided the full co-operation it promised on 14 November 1998.” Moreover, the report concluded, “Iraq’s conduct ensured that no progress was able to be made in either the fields of disarmament or accounting for its prohibited weapons programs.”117 On December 16, American and British forces initiated military action against Iraq in response to the refusal of the Baghdad government to

242

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

comply with UN weapons inspectors. The operation ended on December 19, just prior to the onset of the Islamic holy month of Ramadan, which lasted from December 20, 1998, until January 19, 1999. In a televised address on December 16, President Clinton declared that the action, called Operation Desert Fox, was “designed to degrade Saddam’s capacity to develop and deliver weapons of mass destruction and to degrade his ability to threaten his neighbors.”118 Prime minister Tony Blair also issued a statement on December 16, saying, “There is no realistic alternative to military force. We are taking this military action with real regret but also with real determination. We have exhausted all other avenues. We act because we must.”119 The air strikes, however, provoked a strong response from some countries. Russia, for instance, recalled its ambassadors from the United States and Britain in protest. China also criticized the action. As the air strikes began on December 16, the UN Security Council met in emergency session to discuss developments.120 Russia and China stated their opposition to military action. The Russian ambassador to the UN “rejected outright” the justifications given by the United States and Britain and declared that no one country could act for the council or “assume the role of a world policeman.” Russian president Boris Yeltsin demanded an end to the U.S.-UK strikes. Beijing also condemned the attacks, and in the UN, the air attacks met severe criticism.121 On December 19, 1998, military operations in Iraq were called off by President Clinton, who suggested that coalition objectives had been achieved. Clinton said, “I am confident that we have achieved our mission. We have inflicted significant damage on Saddam’s weapons of mass destruction.”122 In the aftermath of the air strikes, Iraq sought to demonstrate that it was unbowed by the U.S. and British action. On December 21, 1998, the Iraqi vice president, Taha Yasin Ramadan, declared Iraq was no longer willing to cooperate with the UN inspectors, saying, “all that has to do with inspection, monitoring, and WMD is now behind us.”123 Iraq declared that UNSCOM would never be allowed to return.124 Exchange 6: November 8, 2002, to April 2003 UNSC 1441 and Operation Iraqi Freedom In December 1999, the Security Council replaced UNSCOM with the UN Monitoring, Verification and Inspection Commission (UNMOVIC). Iraqi officials rejected the move and continued to refuse any and all inspections.

deterrence and compellence in iraq

243

Periodic talks collapsed over the next two years, with the final straw for Washington coming on July 5, 2002, when Iraq-UN talks in Vienna broke down without any agreement. On September 12, 2002, President George W. Bush addressed the UN and issued the first formal threat of military action if Iraq failed to comply with seventeen UN resolutions. This was followed on September 16 with formal acceptance by Iraq of the “unconditional” return of UN inspectors.125 Formal negotiations between the UN and Iraqi officials resumed on September 30, but Iraq insisted on keeping eight presidential compounds off-limits to international inspectors. The deal was rejected by U.S. officials, who insisted on a final UN resolution before the return of inspectors.126 On November 8, 2002, the Security Council voted unanimously to endorse UNSCR 1441, declaring Iraq in “material breach” and reinstating inspectors after a four-year absence. Saddam Hussein issued a formal reply on November 13 to the UN secretary-general, accepting the conditions outlined in the resolution. Five days later UNMOVIC weapons inspectors arrived in Baghdad. Deterrence / compellence threats, probes, and responses in a multilateral context. The prerequisite for a return of UNMOVIC inspectors to Baghdad was UNSCR 1441. But to get unanimous consensus on UNSCR 1441, other permanent members of the UN Security Council (France, Russia, and China) needed a push—the threat of unilateral deployment by the United States and Britain of nearly 100,000 troops to Kuwait, Qatar, and Turkey. The prerequisite for such a substantial and sustained deployment of American troops to the region began with the “multilateral” domestic support Bush obtained in the 2002 midterm elections. These midterm elections were the first time since 1934 that a sitting president’s party picked up seats in the House of Representatives during their first term. Moreover, these first-term elections were the first time that the party controlling the White House gained enough seats in the Senate to become the majority party. This level of public support for Bush’s foreign and security policy was essential for building multilateral consensus within Bush’s cabinet, the Republican leadership, rank and file Republicans in the House and Senate, and even the Democratic opposition. This momentum behind the Bush administration’s policies accelerated when the Senate voted 77-23 on October 11, 2002, authorizing President Bush “to use all means that he determines to be appropriate, including force, in order to enforce the United Nations Security Council Resolutions (regarding Iraq), defend the national security interests of the United States

244

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

against the threat posed by Iraq, and restore international peace and security in the region.” The House added its support to Bush’s evolving “multilateral” coalition by approving an identical resolution by a vote of 296-133. The passage of these two resolutions was driven by a strong conviction held by a significant majority of politicians in both parties that the only way to get France, Russia, China, and nonpermanent members of the UN Security Council to accept UNSCR 1441 was to convey to these leaders the U.S. commitment to respond (alone if necessary) to what Americans perceived at the time as a growing threat to their security. In the aftermath of the September 11, 2001, terrorist attacks on the United States, the potential threat posed by the possibility that Saddam Hussein was reconstituting a nuclear or chemical weapons arsenal was extraordinarily salient in the minds of most Americans. The deployment by the United States and Britain of close to 100,000 troops to Kuwait and Qatar established the credibility of U.S.-UK resolve. Without this credible threat, and the fear in Paris, Moscow, and Beijing that their substantial economic interests in a post-Saddam Iraq were in jeopardy, the UN would not have come close to achieving a unanimous resolution. Saddam Hussein allowed UNMOVIC inspectors back into Iraq in 2002 (with a significantly more robust mandate than UNSCOM) as a direct consequence of the application by the United States and Britain of credible and resolute coercive diplomacy throughout this entire period. If the coercive threat of unilateral action had not been issued at each stage, none of this domestic or international momentum to act against Iraq would have materialized. The United States succeeded in producing the multilateral consensus in 2003 (and in 1999, 1998, 1995, and 1991, for that matter) by linking coercive military threats against the target with inducements of political or financial rewards (grants, loans, or military aid) to those whose interests and priorities might have been adversely affected by the target’s compliance with U.S. demands. Multilateral coercion requires a major push involving coercive threats of unilateral action by major powers.127 Outcome. The problem with multilateral approaches to deterrence is that it they are significantly more difficult to sustain over time and even more difficult when coercion begins to have a direct impact on the core interests of those in the original coalition. Saddam began to misinterpret French, Chinese, and Russian reluctance to support a second resolution as a clear indication that the Security Council would prevent war. As was evident in the previous exchanges surveyed in this chapter, suc-

deterrence and compellence in iraq

245

cessful coercive diplomacy requires effective communication (e.g., an explicit threat of “serious consequences” and a timeline for compliance), a clear commitment to the issue(s) at stake, the capability to impose sufficiently painful military costs on the opponent, and the resolve to carry through on one’s threats.128 But heavy reliance on multilateral consensus as the benchmark for a “legitimate” application of force occasionally undermines the credibility and effectiveness of coercive threats, because opponents may mistakenly conclude that “legitimacy” and “credibility” are linked, when in reality they are not. The mistake that ultimately led to the demise of Saddam Hussein (and Slobodan Milosevic after the election of 2000) was the assumption that the U.S.-UK threat lacked credibility because it did not satisfy the requirements for legitimacy. Because of this miscalculation, Saddam underestimated (1) the probability of an attack that did not first obtain UN endorsement and (2) the risks of noncompliance. Ironically, the obligation to demonstrate U.S.-UK resolve became even more pronounced in light of efforts by France and Russia to deny legitimacy to the threat of military action. From Washington’s point of view, this was the precedent-setting case for establishing the credibility required to manage and prevent threats from Iraq and other rogue regimes.

observations, or lessons for complex deterrence Deterrence and compellence succeeded and failed in each case for all the right reasons, consistent with expectations derived from the standard fourcondition model of deterrence theory. Table 10.1 provides a summary of exchanges in the Iraq case from 1991 to 2003. It is apparent that the four conditions for deterrence noted at the outset of this study (labeled A through D in the table) are connected to success versus failure across the six exchanges. Nuances within the respective exchanges are explained in notes to the table. What follows is a summary of events and outcomes for the six exchanges in narrative form to accompany the table’s summary of results. The first exchange, in which all conditions for deterrence were present, ultimately proved successful for the United States and Britain. The coalition compelled Iraq to accept the southern NFZ and deterred it from attacking Shiites in southern Iraq. Iraqi compliance offers impressive support for deterrence theory because there is some “white noise” along the way, namely, equivocation by the president regarding how long the United States would remain committed if troops had to be deployed. But strong follow-up statements, which showed high levels of credibility and intensity of response,

246

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

tab le 10.1 Summary of Exchanges, Iraq 1991–2003 1 April 5–24, 1991

2 Aug.26– Jan. 19, 1993

3 Oct. 7–16, 1994

4 Aug. 31– Sept. 13, 1996

5 July– Dec. 1998

6 Nov. 8, 2002– April 2003

A. Unacceptable behavior defined

Y

Y

Y

Y

Y

Y

B. Commitment to punish communicated

Y/ N

Y

Y

Y

Y/ Na

Y

C. Possess the means /capability to defend commitment

Y

Y

Y

Y

Y

Y

D. Demonstrate resolve

Y/ N

Y

Y

Y/ Nb

Y/ Na

Y/ Nc

Success?

Y/ Nd

Y

Y

Y

N

Y/ Ne

Condition

n o t e : N = no; Y = yes. a On December 16, 1998, military action against Iraq was called off without Iraq complying, which diminished the commitment to punish Iraq. b Fissures in the coalition make resolve dubious and lead to Iraqi counterthreats on September 1, 1996. c Resolve called into question by public opposition of the antiwar powers (Russia, China, France). d The successful outcome is ambiguous for two reasons. The problem arises because there appears to be a concerted effort by Iraq in the initial stages of this exchange to defy weak U.S. threats, which would be a failure. However, the final outcome is Iraqi compliance with the NFZ (success). The target state sometimes ascertains resolve through probing. The question is, when does probing occur in order to test resolve, versus as a response to a lack of resolve? The answer would seem to lie in the pervasiveness of the probing. In the initial stages of this exchange, the Iraqi seizure of Irbil would appear to be a concerted effort of defiance due to a lack of commitment and resolve from the United States. For this reason, the first exchange really consists of two exchanges: An initial exchange between April 5 and 15, which resulted in failure (as the model predicts), and a second exchange between April 16 and 24, which resulted in Iraqi compliance and success. Disaggregating this exchange (for the reasons stated above) eliminates the ambiguity in the outcome and adds to the predictive power of the model. e One could argue exchange 6 ended in war and should therefore be coded as N. Coercion must have failed if military force was used. However, Washington’s coercive efforts throughout 2002 (described in the main part of the paper) were designed in large measure to compel Saddam to let inspectors back into the country (after a four-year absence) with a robust inspections mandate. On these terms, the coercive threats (and all related actions designed to generate the consensus that produced UNSCR 1441) succeeded. But Saddam never quite understood his predicament, or Washington’s post-9 / 11 mind-set, and continued to probe for weaknesses by challenging inspectors over time. These probes reinforced the false (but entirely logical) impression that he was hiding WMD, which set the stage for war.

deterrence and compellence in iraq

247

proved sufficient to gain the attention of Saddam Hussein and his supporters. In the second exchange, there was an escalation from words to deeds. Bombing raids and cruise missile attacks became necessary and represent deterrence by punishment. As a result of these actions, Saddam Hussein accepted the NFZ and eschewed attacking Shias in southern Iraq. Success in the third exchange revealed further escalation in the commitment of capabilities, ultimately producing an explicit show of compliance by Iraq. The way this exchange was terminated serves as strong confirmation for deterrence theory: Iraq went out of its way to prevent miscommunication that could lead to a United States–led attack. Results in the next three exchanges began to change and combined to hint at ultimate failure by the United States and Britain to deter Saddam Hussein. Mixed results occur in the fourth exchange. Support from France, Russia, and China evaporated at this point in time as the confrontation between Iraq and the United States became even more militarized. Saddam gave up on his efforts toward compellence (i.e., no longer shooting at coalition fighters enforcing flight-exclusion zones), but maintained a military presence just outside the NFZ. This partial achievement of compellence hinted at the pattern of activity that would come to characterize future interactions between the Security Council and Iraq. Exchange 5 featured very strong U.S. actions that compelled Iraq to accept IAEA and UNSCOM inspectors, but an important subtext also emerged: there was an increasing amount of public commentary emanating from Beijing and other national capitals about the lack of a “green light” for the use of force in the future. Although the United States succeeded in the short term, the necessary conditions for long-term success became increasingly problematic. The United States and Britain engaged in air strikes in response to Iraqi noncompliance on inspections. Iraq, however, remained defiant after the strikes ended. This defiance was probably produced by the realization that air strikes produced limited damage while affording the regime in Baghdad the opportunity to enjoy major political gains from standing up to the United States and Britain. Exchange 6 began well and ended badly for the United States and Britain. By this point, both states were on their own in their efforts to deter Iraq. International inspectors returned to Iraq for a time, but miscommunication between Saddam Hussein on the one hand and the antiwar powers (Russia, China, and France) on the other distorted Iraq’s perception of the threat from the United States and Britain. Hussein’s overestimation of the power of Russia, China, and France to stop an attack on Iraq led to the failure of

248

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

compellence, full-scale war, and ultimately the downfall of his regime and his own demise. After this review of events, it is clear that Alexander George and Richard Smoke129 are correct: deterrence fails over time and in stages. Their assertions anticipate this volume’s focus on complex deterrence. A direct answer to the question posed at the outset of this chapter is now possible. Deterrence remains relevant even in a complex security environment, although at times failure is to be expected when conditions for its implementation are not satisfied. One aspect of complex deterrence that comes out in this case is the “game within the game.” While the United States and Britain attempted to deter or compel Iraq on various occasions, Iraq also probed its adversaries for weakness. Probes were a natural and almost inevitable feature of Iraq-U.S.-UKUN confrontation. Probes provide an opportunity to demonstrate resolve, credibility, and commitment, so their occurrence cannot be viewed exclusively as an illustration of deterrence failure (in terms of both theory and strategy). In the absence of retaliation following probes, the theory predicts failure as opponents exploit weakness to demonstrate their own success. States, no matter how powerful, rarely have the capacity to sustain credible threats over time without experiencing a few probes; the exceptions to this rule are nuclear rivalries between advanced nuclear powers. Probes, therefore, should be expected because opponents in conventional rivalries rarely know precisely what constitute “red line” conditions until the point of retaliation, when it is too late to compromise to avoid military action. If the unwanted behavior (e.g., an attack) continues despite the presence of credible and resolute commitments on the part of the actor intending to issue a deterrent threat, then such an outcome would constitute grounds for judging that both deterrence policy and theory had failed. In sum, the Iraq case reveals that no threat is better than weak threats because the latter tend to lead to escalation as opponents exploit weakness to demonstrate power, relevance, and control. Deterrence complexity is an important yet still largely unexplored subject—behavior throughout the pre-Iraq 2003 crisis was a direct consequence of interactions in the previous ten years. Comprehending the failure of 2003 is impossible without a clear understanding of successes and failures and associated interactions over this entire decade-long rivalry. Iraq updated beliefs about coalitional breakdown but went too far and engaged in wishful thinking about the implications regarding an attack by the United States and Britain.

deterrence and compellence in iraq

249

conclusions Deterrence remains relevant to today’s complex security environment. Three of the four assumptions of the deterrence paradigm have been confirmed by the historical survey offered in this chapter. First, the paradigm assumes that states function more or less as rational actors. When events usually considered to be a single case are divided into their constituent diplomatic or military parts, cost-benefit calculations become clearer. Even Iraqi officials, who are often stereotyped as irrational or ruled by a mad dictator, were able to respond in a coherent way to the evolving situation. In terms of instrumental rationality, it would be difficult to find a more prudent performance than the one put in by Saddam Hussein as he confronted a series of three American presidents from 1991 to 2003. During each of the six exchanges, he appeared to follow the advice given by the coach to the team’s quarterback at the start of every football game: Take what the defense gives you. A second axiom of the paradigm, which asserts that deterrence operates mainly among states, is confirmed. While the UN played an intermediary role, and the activities of Secretary-General Kofi Annan should not be overlooked, the decisive exchanges involved states, with the United States and Iraq in the foremost positions. Rivalry, the third assumption, is perhaps the most easily confirmed by this case. The United States and Iraq sparred with each other for almost fifteen years before regime change occurred. Each state waited eagerly for its next opportunity to gain an advantage over the other, despite the fact that they faced an ever-increasing risk of war. Fourth, the axiom regarding the differences among conventional, chemical, biological, and nuclear weapons forming the basis for different types of deterrence remains unclear. Much of the conflict between Iraq and the international community surrounded efforts on the part of the coalition to discover and to end Iraq’s effort to acquire WMD. But it is not clear that the quasi endowment of Iraq with these weapons, which may have existed mostly in the minds of its adversaries anyway, played a role in determining strategy and tactics on the part of the nuclear powers at various stages of confrontation. Instead, the latter seemed willing to run the risk of a chemical or biological weapons attack in their quest to force Iraq to back off from abusing its population and reveal more about its military infrastructure. In sum, this chapter supports the notion that deterrence encounters are complex and multidimensional. Complex deterrence is a paradigm that

250

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

recognizes that multiple actors and iterations are a reality in today’s security environment. Complexity is likely to remain as long as the power disparity among actors in the international system is large. Given its position of military dominance, the United States will try to compel and deter at the same time; success or failure may be based on the complexity of a particular situation, the importance of the issues involved, and the potential for issuing credible threats.

notes 1. Paul Huth and Bruce Russett, “What Makes Deterrence Work? Cases from 1900–1980,” World Politics 36 ( July 1984): 496–526; Huth and Russett, “Deterrence Failure and Crisis Escalation,” International Studies Quarterly 32 (March 1988): 29–45; Huth and Russett, “Testing Deterrence Theory: Rigor Makes a Difference,” World Politics 42 ( July 1990): 466–501; Richard N. Lebow and Janice Gross Stein, “Beyond Deterrence,” Journal of Social Issues 43 (Winter 1987): 5–71; Lebow and Stein, “Rational Deterrence Theory: I Think, Therefore I Deter,” World Politics 41 ( January 1989): 208–24; Lebow and Stein, “When Does Deterrence Succeed and How Do We Know?” (paper presented at the annual meeting of the International Studies Association, London, March 28–April 1, 1989); Lebow and Stein, “Deterrence: The Elusive Dependent Variable,” World Politics 42 (April 1990): 336–69. 2. Frank P. Harvey, “Practicing Coercion: Revisiting Successes and Failures Using Boolean Logic and Comparative Methods,” Journal of Conflict Resolution 43 (December 1999): 866. 3. Paul (this volume) collapses these conditions into three premises: “a deterrer should have sufficient capability; its threat should be credible; and it should be able to communicate the threat to its opponent.” These premises for success are nearly identical to those noted in the text, although condition A makes the requirements imposed in this chapter more restrictive. It recognizes an additional complication—the inability of the target to comprehend what is unacceptable behavior—that appears in a range of cases. 4. Frank P. Harvey, “Rigor Mortis or Rigor, More Tests: Necessity, Sufficiency, and Deterrence Logic,” International Studies Quarterly 42 (December 1998): 675–707; Harvey, “Practicing Coercion.” 5. United Nations Security Council, Resolution 688, adopted by the Security Council at its 2982nd Meeting, April 5, 1991, S/RES/0688, http://www.fas.org/news/un/iraq/sres/sres0688.htm. 6. GlobalSecurity.Org, “Operation Provide Comfort,” n.d., http://www.globalsecurity.org / military/ ops / provide_comfort.htm. 7. United Nations Security Council, Resolution 688; Paul Lewis, “Europeans Back off Plan to Help Kurds,” New York Times, April 10, 1991, A12. 8. Elaine Sciolino, “U.S. Warns Against Attack by Iraq on Kurdish Refugees,” New York Times, April 11, 1991, A10. 9. George H. W. Bush, “Exchange with Reporters on Aid to Iraqi Refugees,” April 11, 1991. http:// bushlibrary.tamu.edu / research / public_papers.php?id=2863&year=1991&month=4. 10. Clyde Haberman, “U.S. Military Takes over Relief for Kurdish Refugees in Iraq,” New York Times, April 13, 1991, A4. 11. Kenneth Katzman, “Iran and Iraq: U.S. National Security Problems Since the Gulf War—A Chronology” (Congressional Research Service, CRS 93–638F, Washington, DC, July 8, 1993). 12. Maureen Dowd, “Bush Stands Firm on Military Policy in Iraqi Civil War,” New York Times, April 14, 1991, A1.

deterrence and compellence in iraq

251

13. Harvey, “Rigor Mortis”; Harvey, “Practicing Coercion.” 14. George H. W. Bush, “Remarks on Assistance to Iraqi Refugees and a News Conference,” April 16, 1991, http:// bushlibrary.tamu.edu / research / public_papers.php?id=2882&year=1991 &month=4. 15. Katzman, “Iran and Iraq,” 20; see also U.S. State Department Dispatch, “US Expands Kurdish Relief Efforts,” April 22, 1991, http:// dosfan.lib.uic.edu / ERC / briefing / dispatch / 1991 / html / Dispatchv2no16.html. 16. Alan Cowell, “Iraq, Assailing Bush’s Plan, Says U.N. Will Open Camps,” New York Times, April 18, 1991, A16; Chuck Sudetic, “Iraqi Forces Begin Pullback in North as the Camps Rise,” New York Times, April 22, 1991, A1; Sudetic, “Kurds Reluctant to Return to Iraq,” New York Times, April 23, 1991, A1. 17. Sudetic, “Iraqi Forces Begin Pullback,” A8. 18. Katzman, “Iran and Iraq,” 20. 19. Ibid., 20; Chuck Sudetic, “U.S.-Iraqi Standoff Hinders Kurd Relief at Camp,” New York Times, April 24, 1991, A10. 20. The New York Times news analysis noted the political motivation behind this Iraqi concession, and expressed doubt regarding the intentions behind the about-face. Patrick E. Tyler, “New About-Face in Iraq,” New York Times, April 25, 1991, A1, A12. 21. President George H. W. Bush, “Excerpts from Bush’s Talk: Iraqi Air Zone is Off-Limits,” New York Times, August 27, 1992, A14. 22. William E. Schmidt, “Iraq Says It Is Ready to Fight Allies over Air Zone,” New York Times, August 27, 1992, A14. 23. Michael R. Gordon, “U.S. Shoots Down an Iraqi Warplane in No-Flight Zone,” New York Times, December 28, 1992, A1. 24. Ibid. 25. Michael, R. Gordon, “U.S. Sends Jets to Iraq Zone from a Carrier off Africa,” New York Times, December 29, 1992, A4. 26. Michael R. Gordon, “Iraq Given Friday Deadline on Missiles,” New York Times, January 6, 1993, A8. 27. Eric Schmitt, “U.S. Tries to Verify Compliance as Iraqi Missile Deadline Passes,” New York Times, January 9, 1993, A6; Michael R. Gordon, “U.S. Says Baghdad Removed Missiles,” New York Times, January 10, 1993, A10. 28. Youssef M. Ibrahim, “Iraqi Aide Defends Removal of Equipment in Border Zone,” New York Times, January 12, 1993, A2. 29. Michael R. Gordon, “Bush Said to Plan Air Strike on Iraq over Its Defiance,” New York Times, January 13, 1993, A1, A2. 30. R. W. Apple, Jr., “U.S. and Allied Planes Hit Iraq, Bombing Missile Sites in South in Reply to Hussein’s Defiance,” New York Times, January 14, 1993, A1. 31. Eric Schmitt, “Allied Strike: Swift and Unchallenged,” New York Times, January 14, 1993, A8. 32. Michael R. Gordon, “Hitting Hussein with a Stick, Keeping a Sledgehammer Ready,” New York Times, January 14, 1993, A8. 33. Thomas L. Friedman, “U.S. Leads Further Attacks on Iraqi Antiaircraft Sites; Admits Its Missile Hit Hotel,” New York Times, January 19, 1993, A1. 34. Michael R Gordon, “Iraq Says It Won’t Attack Planes and Agrees to U.N. Flight Terms,” New York Times, January 20, 1993, A1. 35. Michael R. Gordon, “U.S. Sends Force as Iraqi Soldiers Threaten Kuwait,” New York Times, October 8, 1994, A1, A6. 36. Ibid. 37. Barbara Crossette, “Iraqi Denounces Sanctions,” New York Times, October 8, 1994, A6. 38. Crossette, “Iraqi Denounces Sanctions,” A6.

252

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

39. Gordon, “U.S. Sends Force,” A6. 40. Michael R. Gordon, “Pentagon Moving a Force of 4,000 to Guard Kuwait,” New York Times, October 9, 1994, A1, A10. 41. Ibid., A10. 42. United Nations Security Council, Statement by the President of the Security Council, October 8, 1994, S / PRST/ 1994 / 58, http://www.fas.org / news / un / iraq / sprst / sprst1994–58.htm. 43. Robert S. Greenberger, “Clinton Orders More Soldiers to Persian Gulf,” Wall Street Journal, October 10, 1994, A14. 44. Ibid., A14. 45. Douglas Jehl, “Clinton’s Line in the Sand,” New York Times, October 10, 1994, A1. 46. Jeffrey H. Birnbaum, “U.S. Dispatches More Aircraft to Persian Gulf,” Wall Street Journal, October 11, 1994, A3. 47. President William J. Clinton, “Transcript of Clinton’s Address: No Evidence the Iraqis Have Pulled Back,” New York Times, October 11, 1994, A11. 48. Michael R. Gordon, “Keeping the Iraqis Back,” New York Times, October 11, 1994, A12. 49. Robert S. Greenberger, “U.S. Plan to Block Future Iraqi Moves on Kuwait Amid Signs of Withdrawal,” Wall Street Journal, October 12, 1994, A3. 50. Michael R. Gordon, “U.S. Sees Signs of Iraqi Retreat but Continues Buildup,” New York Times, October 12, 1994, A1. 51. Robert S. Greenberger, “U.S. Plan Would Bar Elite Iraqi Units from Moving Toward Kuwait Again,” Wall Street Journal, October 13, 1994, A18. 52. Elaine Sciolino, “U.S. Offers Plan to Avoid Threat from Iraq Again,” New York Times, October 13, 1994, A1, A16. 53. Some Pentagon officials called the containment plan speculative, because the United States did not want to encourage Iraq to test its resolve to uphold the zone by sending in one or two tanks at a time and because the United States feared creating a “military vacuum” in southern Iraq that could result in the dismemberment of the country. See Sciolino, “U.S. Offers Plan to Avoid Threat,” A16. 54. Michael R. Gordon, “U.S. Plans to Keep Planes and Tanks in the Gulf Area,” New York Times, October 14, 1994, A1. 55. Joseph B. Treaster, “Marines Stage Desert Exercise in Show of Force,” New York Times, October 14, 1994, A16. 56. Barbara Crossette, “U.S. Is Demanding a Quick U.N. Vote on Iraqi Pullback,” New York Times, October 15, 1994, A1; Crossette, “Security Council Condemns Iraq’s Threat to Kuwait,” New York Times, October 16, 1994, A12. 57. United Nations Security Council, Resolution 949, adopted by the Security Council at its 3438th Meeting, October 15, 1994, S / RES / 949, http:// www.fas.org / news / un / iraq / sres / sres0949.htm. 58. Youssef M. Ibrahim, “Iraq Signals Acceptance of U.N. Move,” New York Times, October 17, 1994, A10. 59. Steven Lee Myers, “Pentagon Sees a New Threat by Iraq Forces,” New York Times, August 31, 1996, A1, A2. 60. Ibid. 61. Ibid., A2. 62. Jamie McIntyre, “U.S. May Send More Warplanes to Gulf,” CNN World News, August 30, 1996, http:// edition.cnn.com /WORLD/ 9608 / 30 / us.troops.gulf / index.html. 63. Steven Lee Myers, “U.S. Calls Alert as Iraqis Strike a Kurd Enclave,” New York Times, September 1, 1996, A1. See also Kenneth Katzman, Alfred Prados, and Clyde Mark, “CRS Issue Brief: August 1996 Crisis,” December 5, 1996, http://www.globalsecurity.org /wmd / library/ report / crs / 94–049.htm.

deterrence and compellence in iraq

253

64. Chris Hedges, “Baghdad’s Move Puts the Future of Kurdish Safe Haven in Doubt,” New York Times, September 1, 1996, A8. 65. Steven Lee Myers, “U.N. Halts Deal for Iraq Oil Sales As U.S. Pledges Action on Attack,” New York Times, September 2, 1996, A1, A6; CNN World News, “Clinton Signs off on Military Action in Iraq,” September 2, 1996, http:// edition.cnn.com /WORLD/ 9609 / 02 / iraq.update / . 66. Katzman, Prados, and Mark, “CRS Issue Brief ”; Steven Lee Myers, “U.S. Attacks Military Targets in Iraq,” New York Times, September 3, 1996, A1, A6. The Department of Defense briefings from that day are U.S. Department of Defense Office of the Assistant Secretary of Defense (Public Affairs), “News Transcript,” September 3, 1996, http://www.defenselink.mil / transcripts / 1996 / t090396_t0903dst.html; U.S. Department of Defense Office of the Assistant Secretary of Defense (Public Affairs), “DoD Press Briefing with Mr. Kenneth Bacon,” September 3, 1996, http://www.defenselink.mil / transcripts / 1996 / t090396_t0903asd.html; see also Myers, “U.S. Attacks Military Targets in Iraq.” 67. The White House Office of the Press Secretary, “Remarks by the President to the National Guard Association Annual General Session,” September 3, 1996, http://www.clintonfoundation .org / legacy/ 090396-speech-by-president-to-national-guard-association.htm. 68. The White House Office of the Press Secretary, “Statement by the President,” September 3, 1996, http://www.clintonfoundation.org / legacy/ 090396-presidential-statement-on-action -against-iraq.htm. 69. Alison Mitchell, “U.S. Continuing Bid to Smash Air Defense,” New York Times, September 4, 1996, A8; Craig R. Whitney, “From Allies, U.S. Hears Mild Applause or Silence,” New York Times, September 4, 1996, A10. 70. Whitney, “From Allies,” A10. 71. Barbara Crossette, “Clinton Finds Little Support at the U.N. for Iraqi Strikes,” New York Times, September 5, 1996, A10. 72. Mitchell, “U.S. Continuing Bid,” A1, A8. 73. Eric Schmitt, “Targets Were Chosen to Punish and Weaken Hussein, U.S. Officials Say,” New York Times, September 4, 1996, A9. 74. Clinton’s remarks on Iraq from September 4 can be found at The White House Office of the Press Secretary, “Remarks By President on Iraq at Lunch Meeting,” September 4, 1996, http:// www.clintonfoundation.org / legacy / 090496-remarks-by-president-on-iraq-at-lunch -meeting.htm. See also Mitchell, “U.S. Continuing Bid,” A1, A8; Katzman, Prados, and Mark, “CRS Issue Brief.” 75. Eric Schmitt, “Clinton, Claiming Success, Asserts Most Iraqi Troops Have Left Kurds’ Enclave,” New York Times, September 5, 1996, A1. 76. Philip Shenon, “Iraq Orders Halt to Missile Strikes on American Jets,” New York Times, September 14, 1996, A1. 77. Shenon, “Iraq Orders Halt to Missile Strikes,” A1 As Shenon stated, “after several days of taunting the United States, the Iraqis appeared today to be backing down . . .” (Shenon, “Iraq Orders Halt to Missile Strikes,” A3). 78. CNN, “Strike in the Gulf Special Coverage: The Strike,” http:// edition.cnn.com /WORLD/ 9609 / gulf.strike / index.html#strike; CNN, “The Lead Up,” http://www.cnn.com /WORLD/ 9608 / 30 / us.troops.gulf / index.html; CNN, “The Attack,” http:// www.cnn.com /WORLD / 9608 / 31 / iraq / . 79. BBC News Online, “Annan Signs Deal with Iraq,” February 23, 1998, http:// news.bbc .co.uk/ hi / english / events /crisis_in_the_gulf / latest_news / newsid_59000 / 59232.stm. 80. “Memorandum of Understanding between the United Nations and the Republic of Iraq,” February 23, 1998, as quoted in Tim Youngs and Mark Oakes, “Iraq: ‘Desert Fox’ and Policy Developments” (Research Paper 99 / 13, International Affairs and Defence Section, UK House of Commons Library, London, February 10, 1999), 9–10.

254

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

81. “Memorandum of Understanding between the United Nations and the Republic of Iraq,” 10. 82. United Nations Security Council, Resolution 1154, adopted by the Security Council at its 3858th Meeting, March 2, 1998, S / RES / 1154, http:// daccessdds.un.org / doc / UNDOC / GEN / N98 / 050 / 79 / PDF/ N9805079.pdf?OpenElement. 83. United Nations Security Council, Resolution 1154, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 10. 84. David Buchan and Laura Silber, “UN-Iraq Accord Faces Early Test,” Financial Times, March 4, 1998, 5, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 10; BBC News Online, “China says UN Iraq Resolution Doesn’t Allow Automatic Use of Force,” March 3, 1998, http:// news.bbc.co.uk/ hi / english /world / monitoring / newsid_61000 / 61626.stm. 85. Laura Silber, “US ‘Must Consult UN on Iraq Strikes’,” Financial Times, 9 March 1998, 6, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 10. 86. BBC News Online, “UN Divided over Action against Iraq,” March 3, 1998, http:// news.bbc .co.uk/ 1 / hi / events /crisis_in_the_gulf / 61576.stm. 87. BBC News Online, “New UN Weapons Teams Arrive in Baghdad,” March 5, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 62385.stm. 88. Daily Telegraph, April 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 10. 89. Youngs and Oakes, “Iraq: ‘Desert Fox’,” 11. 90. Ibid. 91. World Mediawatch in Association with BBC Monitoring, “Iraq’s Statement,” August 5, 1998, http:// news.bbc.co.uk/ 1 / hi /world / monitoring / 146208.stm. 92. Max Ruston, “UN / Iraq Inspections”, Voice of America, August 6, 1998, http://www.fas .org / news / iraq / 1998 / 08 / 980806-iraq1.htm. 93. BBC News Online, “Butler: No Crisis with Iraq,” August 5, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 145102.stm, and BBC News Online, “US Dismisses Iraqi Action,” August 5, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 146169.stm. 94. BBC News Online, “UN Weapons Inspector Resigns, Guns Blazing,” August 26, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 159276.stm. 95. Sunday Telegraph, September 27, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 11. William S. Ritter, Jr., “Resignation Letter,” August 26, 1998, http://cns.miis.edu / research / iraq / friday.htm. 96. Irish Times, August 28, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 12. 97. Associated Press, September 1, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 12. 98. Mark Tran, “Inspector Claims UN Caved in to Iraq,” Guardian, August 28, 1998, 14, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 12. 99. For more detail on Scott Ritter’s allegations, see Peter J. Boyer, “Scott Ritter’s Private War,” New Yorker, November 9, 1998. 100. United Nations Security Council, Resolution 1194, adopted by the Security Council at its 3924th Meeting, September 9, 1998, S / RES / 1194, http://www.un.org / Depts / unscom / Key resolutions / sres98–1194.htm, 101. BBC News Online, “Security Council Cancels Review of Iraq Sanctions,” September 9, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 168129.stm; BBC News Online, “UN Suspends Iraq Sanctions Reviews,” September 10, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle _east / 168134.stm. 102. Financial Times, “Iraq Rules out More Arms Inspections,” September 29, 1998, 6, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 15. 103. Ibid. 104. Middle East International, November 13, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 16.

deterrence and compellence in iraq

255

105. BBC News Online, “Iraq and the UN: The Standoff Continues,” October 7, 1998, http:// news.bbc.co.uk/ 2 / hi / middle_east / 188764.stm; BBC News Online, “The UN Continues Efforts to Defuse the Crisis with Iraq,” October 2, 1998, http:// news.bbc.co.uk/ 2 / hi / americas / 185507.stm. 106. BBC News Online, “Iraq Refuses UN Request on Weapons Document,” October 15, 1998, http:// news.bbc.co.uk/ 2 / hi / americas / 194536.stm. 107. World Mediawatch in Association with BBC Monitoring, “Iraq’s Statement in Full,” October 31, 1998, http:// news.bbc.co.uk/ 2 / hi /world / monitoring / 205206.stm. 108. BBC News Online, “UK Prepared to Use Force against Iraq,” November 2, 1998, http:// news.bbc.co.uk/ 2 / hi / uk_news / politics / 205356.stm. 109. United Nations Security Council, Resolution 1205, adopted by the Security Council at its 3494th Meeting, November 5, 1998, S / RES / 1205, http://www.un.org / Depts / unscom / Key resolutions / sres98–1205.htm. 110. United Nations Security Council, Resolution 1205. 111. BBC News Online, “UNSCOM Withdraws Monitors from Iraq,” November 11, 1998, http:// news.bbc.co.uk/ 2 / hi / middle_east / 212123.stm. 112. Financial Times, “Arab States Increase Pressure on Saddam,” November 13, 1998, 1, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 18. 113. BBC News Online, “Clinton Warns Iraq: Speech in Full,” November 13, 1998, http:// news .bbc.co.uk/ 2 / hi / middle_east / 212578.stm; BBC News Online, “Clinton Warning to Iraq,” November 11, 1998, http:// news.bbc.co.uk/ 2 / hi / middle_east / 212416.stm. 114. Financial Times, December 17, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 18. 115. BBC News Online, “Iraq’s Inspection Block Unacceptable,” December 9, 1998, http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 231562.stm. 116. “Letter from the Executive Chairman of UNSCOM to the Secretary-General of the United Nations,” December 15, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 24. The text of this letter is included in “Letter Dated 15 December 1998 from the Secretary-General Addressed to the President of the Security Council,” December 15, 1998, S / 1998 / 1172, http://www .un.org / Depts / unscom / s98–1172.htm (For further news reports, see GlobalSecurity.org, “Iraq Special Weapons News, 16 December 1998,” http://www.globalsecurity.org /wmd / library/ news / iraq / 1998 / index_12_16.html; “UNSCOM Chairman Butler’s Report to UN Secretary General,” 15 December 1998, http://www.fas.org / news / un / iraq / s / butla216.htm.) 117. See the December 15, 1998, documents (fn. 116). 118. Richard Wolffe and Stephen Fidler, “U.S. and Britain Launch Air Strikes On Iraq,” Financial Times, December 17, 1998, 1, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 25. 119. “Statement by the Prime Minister, Mr. Tony Blair, Downing Street, London,” December 16, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 25. 120. See GlobalSecurity.org, “Iraq Special Weapons News, 16 December 1998,” for UN Security Council minutes and the US announcement of the attacks. 121. BBC News Online, “From Beijing, Condemnation of Iraq Attack,” http:// news.bbc.co.uk/ 1 / hi /world / middle_east / 236871.stm, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 32; BBC News Online, “Yeltsin Demands End to Strikes: Statement,” December 17, 1998, http:// news .bbc.co.uk/ 1 / hi /world / monitoring / 237086.stm; BBC News Online, “UN Hears Calls for End to Raids,” December 17, 1998, http:// news.bbc.co.uk/ 1 / hi / events /crisis_in_the_gulf / latest_news / 236594.stm. 122. BBC News Online, “Iraq Bombing Completed,” December 19, 1998, http:// news.bbc .co.uk/ 1 / hi / events /crisis_in_the_gulf / latest_news / 238856.stm. U.S. government statements and U.S. and global news reports from this day are available at GlobalSecurity.org, “Iraq Special Weapons News, 19 December 1998,” http://www.globalsecurity.org /wmd / library/ news / iraq / 1998 /

256

f r a n k p. h a r v e y a n d p a t r i c k j a m e s

index_12_19.html; see also Nick Simeone, “Clinton / Iraq,” December 19, 1998, http://www .globalsecurity.org /wmd / library/ news / iraq / 1998 / 981219-end01.htm. 123. BBC Summary of World Broadcasts, “Vice-President Ramadan Says UNSCOM Work Is Over,” December 21, 1998, as quoted in Youngs and Oakes, “Iraq: ‘Desert Fox’,” 36. 124. The Office of the Historian, Bureau of Public Affairs, “History of the Department of State During the Clinton Presidency (1993–2001),” http://www.state.gov/ r/ pa / ho / pubs / 8528.htm. Excerpts from his speech can be found on online at BBC News Online, “Iraqi Deputy Premier’s Baghdad News Conference,” December 21, 1998, http://news.bbc.co.uk/1/ hi/world /monitoring/ 240169.stm. 125. Matthew Engel, Ewen MacAskill, and Nicholas Watt, “Saddam Caves in on Inspectors,” Guardian, September 17, 2002, http://www.guardian.co.uk/ Iraq / Story/ 0,2763,793642,00.html. 126. Ian Traynor and Julian Borger, “US Rejects Iraq Inspections Deal,” Guardian, October 2, 2002, http://www.guardian.co.uk/ Iraq / Story/ 0,2763,802828,00.html. 127. Frank P. Harvey, Smoke and Mirrors: Globalized Terrorism and the Illusion of Multilateral Security (Toronto: University of Toronto Press, 2004). 128. For a more complete discussion of the methodological and evidence-related requirements for evaluating success and failure, see Harvey, “Practicing Coercion.” 129. Alexander George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974).

chapter eleven R

Deterrence among Great Powers in an Era of Globalization pa t r i c k m . m o r g a n a n d t. v. pa u l

this chapter examines contemporary deterrence relationships among great powers, that is, the five permanent members of the United Nations Security Council. Our central argument is that deterrence has become significantly more subtle and complex among the great powers, a change initiated by the epochal transformation of the international system brought about by the end of the cold war and the onset of intensified economic and political globalization. A working premise is that deterrence is heavily influenced by the nature and context of political relations among leading state actors. If they avoid intense enduring politicomilitary rivalries, deterrence has little immediate relevance. Today, only general deterrence is at work in great-power relationships, and then only as one of several background conditions contributing to peace and stability.1 This type of deterrence among the great powers is rather recessed, not very prominent—a “recessed general deterrence.” A second working premise is that this deterrence, in practice, is partially shaped by elements familiar from the cold war. They are elements produced by certain central tendencies and tensions that exist in all deterrence relationships, somewhat independent of the specific politicomilitary rivalries that animate such situations. The chapter initially discusses the cold war deterrence relationships among the great powers, and this is followed by a description of their deterrence relations today. The chapter also looks briefly at how deterrence could help ensure that the ongoing transition of the international system remains peaceful. It also explains why certain self-sustaining elements of 259

260

pa t r i c k m . m o r g a n a n d t. v. pa u l

deterrence, particularly nuclear deterrence, make it difficult to completely set it aside. Thus, the chapter does not suggest that the great powers have rejected deterrence. Instead, it posits that they have made deterrence less salient in their relations. It continues to be relevant and strongly influenced by political conditions among those actors. As their political relations have become more complex, deterrence has done the same, even as severe great-power confrontations, in which deterrence would move to the forefront, now seem improbable in the near term.

cold war deterrence relationships among the great powers The cold war was an intense enduring rivalry between the two power blocs led by the superpowers. That it never resulted in a great-power war (the “long peace”) was due in part to deterrence, itself resting, to a significant degree, on nuclear weapons and culminating in a situation of mutual assured destruction. Although scholars differ on the reasons for this peace of over five decades, many believe that without deterrence the great powers’ intense ideological and strategic rivalry would have ultimately resulted in a major war.2 Nuclear weapons were first designated as primarily deterrent weapons early in the cold war.3 The “nuclear revolution” emerged out of the realization that the immense destructive potential of nuclear weapons generally exceeded the political stakes involved in great-power rivalries or in their wars with smaller states. Eventually, this significantly altered statecraft among the great powers to exclude using nuclear weapons and thus avoid conflicts that would make their use very likely.4 The uncontestable nature of nuclear weapons made actors circumspect about even nuclear threats, including when their deployment postures focused on possible first strikes.5 Several crises—over Korea, Berlin, and Cuba—occurred before Moscow and Washington grasped the significance of the nuclear revolution and the deep constraints it imposed on their competition.6 This did not prevent them from engaging in indirect or secondary conflicts, some of which, such as Vietnam for the United States and Afghanistan for the Soviet Union, had major repercussions. But they went to great lengths to avoid direct and visible military clashes with each other on even a small scale. Soviet “volunteers” participated in the Korean War but only secretly, and the Chinese “railroad” troops deployed to North Vietnam during the Vietnam War were similarly unadvertised. During the Cuban Missile Crisis, Washington worked hard

deterrence in an era of globalization

261

to not kill Russians, and Moscow actively avoided killing Americans out of fears that uncontrollable escalation could result. The adversaries’ restraint culminated in a settlement of the crisis with virtually no casualties. The cold war’s second-tier great powers also maintained nuclear weapons for both deterrence and, in their view, as contributions to the central strategic balance between the superpowers. These small arsenals also were intended to give them a greater say in world affairs. With a no-first-use policy and very small arsenal, China had what has since been termed a “limited deterrence posture,” while Britain and France built more substantial and varied nuclear forces. Each gained some modest strategic autonomy with their nuclear weapons but seemed unable to more fundamentally shape the course of the cold war or its conclusion. While they obtained a “seat at the table” of sorts, their small arsenals provided only limited deterrence against the superpowers.7 However, fears these governments might be reckless and destabilize international politics were unfounded.8 Nuclear weapons simplified deterrence by clarifying the nature of the risks great powers, and others, would run in provoking great-power warfare or even just crises. Eventually, they also highlighted the risks associated with minor conflicts among smaller states that might provoke great-power involvement. Under the circumstances, it was hard for nuclear powers to make threats to use nuclear weapons highly credible, but the awesome damage they could inflict was so clear that it helped promote caution in many governments caught up in serious conflicts. In addition, with nuclear weapons it was not necessary to calculate precisely what it would take to deter, and since any serious military action by nuclear-armed states below the nuclear level might readily escalate, even lesser threats of retaliation could be mounted effectively by a nuclear power. Nuclear weapons, however, did not simplify strategy or force design, particularly for the superpowers. The destructiveness of nuclear weapons and the speed with which they could be delivered raised serious threats to survival of the retaliatory forces on which deterrence rested. In response, the United States developed a nuclear triad: three different clusters of delivery systems (intercontinental ballistic missiles [ICBMs], submarine-launched ballistic missiles [SLBMs], and long-range bombers), each with enough nuclear weapons to destroy an attacking country in retaliation, an elaborately redundant retaliatory capability. Several justifications were offered for these deployments: such large and diverse forces ensured that no Soviet attack could completely eliminate all retaliatory forces; a disarming first strike would therefore seem virtually impossible to Soviet leaders; and no single

262

pa t r i c k m . m o r g a n a n d t. v. pa u l

Soviet technological breakthrough that endangered one leg of the triad would endanger the others. The Soviet Union followed suit with its version of a triad, while Britain and France developed smaller triads of land-, air-, and sea-based nuclear retaliatory forces. Thought also had to be given to the possibility of confining the use of nuclear weapons in a war to the battlefield and of using preparations for this possibility to deter attacks. The superpowers installed nuclear weapons to either reinforce their conventional forces or to escalate a war to the nuclear level if necessary. The thousands of tactical nuclear weapons added greatly to the complexities of battlefield operational planning. Thought had to be given to keeping the weapons secure in foreign countries, arranging with those governments about when and how they might be used, deciding when they would be turned over to some of those governments to use in a war, and implementing technical and procedural safeguards to keep the weapons from being used if they were lost or stolen. Fear of having to escalate a conventional war to avoid losing it, or of losing it on a strictly conventional level, led the superpowers to also conclude that deterrence stability required maintaining huge conventional forces in addition to their large nuclear arsenals. They also accumulated giant chemical and biological warfare arsenals for deterrence. In short, deterrence involved many kinds of forces on many levels, something other states could not match, although the other great powers maintained quite respectable arsenals of chemical and biological weapons and conventional forces. The superpowers also went well beyond what deterrence required in designing their triads. They pursued a high-damage limitation, even first-strike, capability—aiming to destroy the other side’s forces in an attack. They also looked into achieving victory in an initial attack by decapitation strikes or by attacking enemy command-and-control systems using the electromagnetic pulse produced by high-altitude nuclear detonations. Eventually, each had over 10,000 strategic nuclear weapons, leading to what many considered to be a senseless capacity for “overkill.” In reality, most of these strategic nuclear weapons were not aimed at civilian targets but at the other side’s nuclear forces. Gaining a first-strike capability would have reduced or eliminated the constraints of mutual deterrence by escaping mutual vulnerability, flatly contradicting deterrence theory on what best provided security in the nuclear age. In emphasizing mutual vulnerability, deterrence theory clashed with the logic of national defense and survival in case of a deterrence failure and an imminent attack. If an attack was imminent, it would obviously be best to have a significant damage-limitation capability if not the ability to

deterrence in an era of globalization

263

conduct a “splendid”9 first strike, and military planners were sensitive to this need to limit damage in the event of nuclear hostilities. While it often seemed that nuclear weapons and deterrence contributed greatly to sustaining the cold war by intensifying suspicion and fear, the fact that it ended peacefully with the great-power nuclear and conventional forces still intact strongly suggests that political conflicts were more important than the military balance when it came to explaining the extent and duration of the superpower standoff. And nuclear weapons also did not, as some analysts suggested, dictate the structure of the international system. Russia’s nuclear arsenal did not prevent its disappearance as a superpower and the consequent emergence of unipolarity. In fact, the Soviet and U.S. nuclear arsenals might have facilitated the end of the cold war by reassuring Moscow that Washington would not exploit the military weaknesses that would inevitably emerge as the Soviet empire collapsed. Concern among the great powers, and particularly the superpowers, about maintaining international system stability, especially averting great-power warfare, led to significant system management in areas like crisis management, crisis avoidance, arms control (particularly on nuclear proliferation), efforts at détente, and the isolation of irksome regimes. Apparently, statesmen on both sides came to believe that however effective deterrence was, it was best not to place it under too much stress.

great-power deterrence in the post–cold war era The end of the cold war significantly altered international politics. The strategic situation was transformed, and with that the reign of deterrence as the central security principle in great-power relations ended. Thus, the practice of keeping many nuclear weapons on high alert and already targeted has been considerably curtailed, although at times Russia has threatened to adopt a launch-on-warning posture in response to the decay of its conventional forces and weaknesses in its early-warning systems.10 The great powers have greatly reduced the total number of nuclear weapons through various arms agreements, most recently the May 2002 Strategic Offensive Reductions Treaty that requires the United States and Russia to cut deployed strategic nuclear weapons to 1,700–2,200 each by 2012.11 Similar independent cuts by Britain and France, smaller in the numbers and percentages of the weapons involved, are also underway.12 Modernization of nuclear weapons and delivery systems has been occurring on a much more limited

264

pa t r i c k m . m o r g a n a n d t. v. pa u l

basis than in the past. The United States has not developed new warheads or strategic missiles—its last nuclear-weapons test was in 1992, and it last manufactured a nuclear weapon in 1989—but has modernized many existing ones. Russia is now deploying some new strategic missiles. Britain and France have not devised new nuclear weapons but continue to upgrade their delivery systems. China has reportedly been slowly moving toward introducing new strategic missiles and warheads. With no great power displaying deep antagonism toward any of the others, their conventional forces also have been significantly reduced, drastically so in Russia, and none are now primarily deployed in anticipation of soon having to fight a great-power war. This is a sharp departure from the scale and deployments of conventional forces during the cold war. Yet the great powers continue to retain a substantial number of nuclear weapons in different forms. Many of those dismantled by the superpowers, or their central components (particularly U.S. weapons), have been stored so that thousands of the weapons can be readily reconstituted—the United States plans to have some 5,000–6,000 of these available in reserve by 2012 in addition to the 1,700–2,200 weapons it will have deployed.13 France has stabilized its nuclear force at 348 weapons. Britain is keeping, and modernizing, its four Trident missile submarines with plans to backfit more advanced missiles, developed with U.S. cooperation, into them. Russia’s new strategic missiles include advanced terminal guidance and missile defense avoidance capabilities. China has maintained roughly eighty strategic nuclear weapons, less than twenty of intercontinental range, but is widely expected to introduce new mobile, solid-fuel ICBMs of greater range, to begin deploying new SLBMs on new subs, and to increase the total size of its strategic nuclear forces within the next few years. Nuclear weapons are now maintained for several reasons: as a hedge against the possible reappearance of great-power threats and arms racing; as a symbol of great-power status; to provide extended deterrence that contributes to slowing nuclear proliferation and provide stability in key regions (only of major concern to the United States); as a deterrent against several non-great-power states; and in response to internal pressures inherent in nuclear deterrence postures. The first four reasons are relatively clear. The last concerns several matters. Wanting to be able to practice nuclear deterrence if necessary requires being able to recreate nuclear weapons if need be after they have been discarded. But that takes an elaborate nuclear establishment of complex facilities and talented experts, difficult to sustain if there is nothing for them to

deterrence in an era of globalization

265

do. Keeping some nuclear weapons does not fully solve this problem but makes it easier to solve. Next, as the numbers of nuclear weapons decline, the gap between the scale of U.S. and Russian forces on the one hand and the other nuclear-weapons powers shrinks, not a politically comfortable outcome for the United States and Russia. In addition, to completely eliminate nuclear weapons would require a vast verification system, out of fear that elimination would provide enormous incentives to cheat—getting just a few nuclear weapons would provide a vast military superiority—and thus other enormous incentives to cheat in order to hedge against others cheating. Finally, for several of the great powers, nuclear weapons substitute for large numbers of more expensive conventional forces in terms of what is considered adequate for national security; eliminating them would provoke awkward budgetary and other political problems. Even though they want to retain nuclear weapons, given the great powers’ relatively benign political relations,14 their nuclear weapons are instruments of “existential” and “recessed general” deterrence in their relations with each other. They are also kept in the background in dealing with other potential challengers, meant to discourage such states from even thinking about pursuing military adventures or provoking military crises. Sustaining conventional forces is also hedging on the great-power level, a symbol of great-power status, and a contribution to regional or global stability, but those forces also have more immediate functions. All the great powers know they are quite likely to be using those forces at some point, not against another great power but to deal with lesser but important contingencies—trying to deter them but militarily responding to their emergence anyway. These forces, however, add complications to political relations and deterrence among the great powers. Why is deterrence recessed among the great powers? General deterrence is particularly relevant for states in intense enduring rivalries, strategic competitions with numerous militarized disputes—repeated interactions involving threats or displays of force, and actual violence.15 Enduring rivalries last for two decades or more16 and usually arise from major events like wars and crises and over issues of territory, ideology, or identity, or they are based on major strategic concerns over relative power and status.17 Once up and running, they often last until a major political shock alters the participants’ long-standing attitudes.18 Deterrence is pursued when there is a distinct possibility of war, usually because the target finds the status quo unacceptable or sees an opportunity for a major strategic gain by force. Because of the rivalry, the deterrer sees a

266

pa t r i c k m . m o r g a n a n d t. v. pa u l

war as likely under these circumstances without effective deterrence.19 Enduring rivals are the most likely, therefore, to consistently practice general deterrence and periodically get into immediate deterrence situations.20 Fortunately none of today’s great-power dyads is an enduring rivalry— their political conflicts are too muted. Moreover, historically great powers eagerly sought territory in conflicts with each other to enhance their wealth, power, and prestige, but they largely ignore this today. With no significant, and particularly no enduring, rivalries, great powers are unlikely to see their core national interests gravely threatened, making it difficult for them to expect to use force against each other. Thus, threats of retaliation to prevent the use of force have lost salience, and emphasizing capabilities and threats of that sort seems unimportant, or even counterproductive if it arouses antagonism with other great powers. Political and diplomatic means seem far more appropriate for dealing with the limited disputes and conflicts that arise; deterrence seems largely irrelevant for keeping the peace. Much of the deterrence in great-power relations is inertial, rather than being the crux of any great power’s national security policy. Activities with respect to deterrence capabilities and how they relate to each other do not appear to be managing any major great-power relationship today. Officials spend far more time on deterrence of rogue states or actors in internal conflicts around the globe than on deterrence in relations with each other. Nevertheless, deterrence among the great powers is not completely outmoded. For existential deterrence, nuclear deterrence offers the great powers the ultimate barrier to direct military assaults on one another no matter how politically improbable they are now. They also preserve their status as dominant states in the system. Thus two of their fundamental goals are safe for the indefinite future. But elaborate capabilities and doctrines for existential deterrence are unnecessary.21 The great powers hardly need it, or balance-of-power confrontations and wars, to prevent territorial aggrandizement or blackmail among each other. In today’s international system, they are not driven by fears that U.S. hegemony or the rise of China will erode their sovereignty or territorial integrity. Hence, they increasingly focus on lesser security challenges.22 In addition to disinterest in direct territorial conquest, they also do not mount the hegemonic quests of the Hapsburgs, or a Napoleon or Hitler, that threatened the existence of many sovereign states (including other great powers) and made active military balancing and other deterrence efforts essential. As a result, great powers increasingly focus on soft balancing, hedging, buck-passing and other strategies in pursuit of their international security goals.23

deterrence in an era of globalization

267

But equanimity is not totally pervasive in great-power relations, so deterrence capabilities still play a role. Certainly these governments are not entirely satisfied with the nature of contemporary international politics, and they frequently disagree about current policies, issues, and trends. The United States has an unparalleled military position. The impact of the revolution in military affairs (“military transformation” in Pentagon terms), on top of the collapse of the Soviet bloc, unraveled the rough triadic balance of conventional land forces among the United States and its allies, the Soviet Union, and China that was one component of deterrence in great-power relations. The United States now has a unique superiority in conventional warfare and especially in power projection. American policy makers welcome this, and it is U.S. policy to maintain the superiority indefinitely, at such a high level and so vigorously sustained as to discourage the emergence of potential peer, or “near-peer,” competitors for the foreseeable future. This preoccupation with clear military superiority vis-à-vis the other great powers, abandoning—on the conventional level—any mutual deterrence based on mutual vulnerability, has created uneasiness and suspicion among several of them, particularly when combined with a U.S. interest in preemptive or preventive attacks as well. Linked to this is an elaborate, much broader and more complex, approach to deterrence whereby military, economic, diplomatic, political, and information-warfare resources are to be blended to “dissuade” possible challenges to the United States and its deterrence. The goal is “tailored deterrence,” designed and applied with a specific target’s motivations, risk acceptance, worldview, and capabilities in mind.24 Seeking to forestall challengers and perhaps preempt attacks also has led U.S. planners to seek more strategically valuable conventional forces, blurring cold war distinctions between deterrence roles, between nuclear and conventional forces, and between deterrence capabilities and actual combat resources. The goal is a broad spectrum of forces for destroying a wide range of targets with great precision and speed, a new Global Strike capability.25 The United States is installing conventional warheads on several SLBM submarines, and the U.S. Air Force wants to do the same with some ICBMs, for destruction of strategic targets with conventional weapons. It also has significantly upgraded the hard-target kill capability of its ballistic missiles through precision guidance of their warheads and has sought permission (from Congress) to develop new nuclear warheads that would be smaller as well, presumably to make their use more feasible in a war.26 It is looking to develop cruise missiles with strategic range and terminal guidance for great

268

pa t r i c k m . m o r g a n a n d t. v. pa u l

accuracy as another counterforce capability. In recent wars, it has displayed the much improved accuracy of conventional bombs and tactical missiles, which are more capable against heavily fortified targets. It has gained experience, in Iraq and Afghanistan, in the use of special operations forces to attack strategic targets. There are suggestions that the United States has been seeking, and may have attained, a first-strike capability against all great powers and other states.27 This is a particularly sensitive subject in view of gradual American progress in developing and deploying ballistic missile defenses. If it is true, then the United States clearly dislikes and wants to avoid a true mutual deterrence relationship again, particularly with another great power. Having good political relationships with the other great powers is fine, but the United States clearly wants a military way to try to avoid mutual deterrence if mutual conciliation and cooperation break down. These developments are nearly always said to be directed at rogue states, the weaknesses of deterring such states, the need to protect American forces intervening in places like the Middle East, the need to reassure allies and allied forces in joint interventions, and the goal of preventing nuclear proliferation. They are rarely said to be aimed at other great powers, but this provokes skepticism abroad and in critics at home. It is a triumph of the logic of planning for deterrence weaknesses and failures, noted earlier, over the logic of relying on and thus strengthening mutual deterrence, a struggle continuing from the cold war era. What encourages uneasiness about U.S. military predominance among certain great powers is the fear that the United States is a revolutionary state. As a hegemon, the United States seeks to strengthen the existing international system, a status quo objective. Nevertheless, it openly champions transforming the domestic systems of many other societies, including two great powers, claiming in effect that their existing systems are illegitimate. Such an objective linked to a hegemon is inevitably disturbing, offering a possible basis for reviving conflict among the great powers. It has certainly made the Chinese and Russians less sanguine about current international politics than the U.S. policy makers, and U.S. military developments are often interpreted in this light in Moscow and Beijing. Britain has retained nuclear weapons, but it has greatly simplified its nuclear deterrent. It now relies only on submarine-based missiles and has discarded other nuclear weapons and their delivery systems. It has expressed interest, however, in possibly placing conventionally armed missiles on the submarines, paralleling U.S. missile deployments. It also now maintains strong, flexible conventional forces for regional power projection and inter-

deterrence in an era of globalization

269

vention. Nothing in its foreign policy suggests that any of this is directed at other great powers, and it expresses little sense of threat from any of them, so its forces arouse uneasiness only in that they have been closely associated with U.S. military interventions that other great powers have opposed. Britain’s deterrence concerns appear primarily oriented toward the Middle East. France was for years quite concerned with the emergence of American hegemony but in a political, not military security, sense. But it has slowly shifted toward supporting American goals such as North Atlantic Treaty Organization (NATO) expansion and deterrence of Iran’s proliferation efforts. To a certain extent, France’s nuclear policies also resemble policy changes undertaken by Washington. It has adopted a more flexible nuclear posture and planning process, with regional powers as the threats in mind. This includes preparations for preventive demonstration strikes, counterforce and countereconomic targeting, and the possible use of strategic nuclear forces in support of a wider potential range of national interests, including ones beyond Europe—presumably in the Middle East. France has yet to devote the necessary resources, however, for reequipping its deterrence forces for this shift in its orientation and missions, that is, making French nuclear weapons smaller and more accurate.28 French conventional forces are still significant, although reduced from cold war levels, but the government seems to see no direct threats to France from other great powers that need serious deterring, so the forces are increasingly designed for interventions. China has been concerned about U.S. power and intentions since the Tienanmen Square incident, while having fair to good relations with the other great powers. This concern includes consistently opposing U.S. alliances in the Asia-Pacific region, U.S. threats against North Korea, the forward deployment of U.S. military forces in East Asia, and the U.S. commitment to Taiwan. At first glance, this would seem like a continuing great-power conflict, inviting a central role for deterrence. But this opposition to U.S. interests in Asia is not the sole driving force behind China’s behavior. China is placing great emphasis on modernizing and professionalizing its conventional forces, clearly in preparation for a possible war over the Taiwan issue, which could involve the United States. Beijing apparently hopes that this conventional force modernization will deter both a Taiwanese drive for independence and U.S. military intervention on behalf of Taiwan in the event of hostilities. The modernization includes naval forces and land-based missiles that could put carriers at risk, missiles that might paralyze the island so China could seize it and present the United States with a fait accompli,

270

pa t r i c k m . m o r g a n a n d t. v. pa u l

and testing of a possible system to attack U.S. command, control, and surveillance assets in space. Yet China has done very little to modernize its strategic nuclear forces, that is, to really strengthen its deterrent vis-à-vis the United States. For years it has been widely assumed that China will modernize and enlarge those forces, introducing new mobile missiles and SLBMs with multiple independently targetable reentry vehicle warheads, all with the ability to reach the United States,29 but this has yet to lead to new deployments. And its conventional forces have not yet been strongly reoriented toward power projection, a capability needed to enhance Chinese security in the event of a major conflict with the United States. This suggests that China is not anticipating a great crisis or real threat of an imminent U.S. attack and that it is carefully calibrating its behavior so as to avoid provoking any of those developments. That in turn indicates that deterrence is not now at the heart of China’s national security policy, that it is currently responding to its narrow preoccupation with a regional issue, one on which it has also now adopted a more engagement- and diplomacy-oriented approach. Russia is also embroiled in disagreements with the United States and the West, some fairly intense. Like China, it objects to Western pressure about democracy and human rights. It has strongly opposed NATO expansion. It opposes American missile defense arrangements for Europe and the creation of ballistic missile defense bases in Eastern Europe. It has disliked U.S. military interventions and is uncomfortable with U.S. hegemony. Thus, it is still in the business of deterrence. As its conventional forces virtually collapsed in the 1990s, Russia became more dependent on nuclear weapons for practicing deterrence. It had to rely more heavily on its landbased missiles because the submarine missile fleet rapidly deteriorated from lack of maintenance and regular operations and its aged strategic bombers became obsolete. One response was putting the land missiles on higher alert, closer to a launch-on-warning posture. Another was announcing that nuclear weapons might be used first in a wide range of military contingencies. Most recently, it introduced new ICBMs, one with advanced penetration aids to evade missile defenses. Nevertheless, it has not been using serious military threats over unacceptable international political developments. It has not rebuilt the conventional forces that could credibly pose such threats. While Russian military leaders make much of the threats from the West, Russian policy seems a long way from using deterrence to deal with them.

deterrence in an era of globalization

271

As for possible crises leading to fighting among the great powers, only United States–China relations over Taiwan are influenced directly by deterrence. During the 1990s, the Taiwan issue spawned a real crisis, with a potential for escalation as each side postured militarily in the area and put some forces on alert. In the early 2000s, Taiwanese officials threatened to declare independence and China engaged in some saber rattling for deterrence, partly with the United States in mind. China has been doing a good deal of military modernization and redeployment, with a war with the United States in the background. The United States has responded by shifting ballistic missile submarines from the Atlantic to the Pacific, reconfiguring its overseas conventional forces to take on missions anywhere, including East Asia, improving its bases on Guam for power projection, and developing more extensive defense cooperation with Japan partially with a Taiwan contingency in mind. This is the one great-power relationship in which the parties have been upgrading forces for a possible war with each other, as reflected in their arms development and procurements, military deployments, and military exercises. Still, barring a major miscalculation or a serious crisis driven by domestic politics, even this situation is unlikely to turn into an intense test of deterrence. China continues to pursue its peaceful-rise strategy, including its highly interdependent economic relationship with the United States. It now emphasizes, on the Taiwan problem, a promising engagement strategy. American policy is hedging against a hostile China and continues to offer it a major stake in system management and extensive additional opportunities for peaceful cooperation. Deterrence is not dominating the United States– China relationship today.

nuclear weapons and the peaceful transition of the international system American analysts regularly assert that American extended deterrence provides a stability and security that has allowed major system change in a nonprovocative fashion, such as with China’s rise to true great-power status. By extension, nuclear weapons and conventional deterrence, although recessed, may cushion other prospective transitions in the international system. China’s minimal nuclear deterrence and U.S. nuclear forces can not only facilitate China’s rise without destabilizing its relations with its neighbors, but they can also make it easier for Japan to continue becoming a normal and active great power. Extended deterrence and nuclear weapons have so far

272

pa t r i c k m . m o r g a n a n d t. v. pa u l

also prevented anything like a Sino-Indian crisis as India emerges as a great power. In the past, such power transitions have often been accompanied by war. Preventing hostilities during great-power transitions would allow the rising powers to continue focusing on economic growth, not on showdowns with other uneasy great powers. The economic globalization now helping China and India is facilitated by systemic stability that is provided partly by deterrence and American hegemony. Once again, however, this deterrence acts as a hedge, not a critical factor, because transitions today are being shaped and promoted by other factors. Deterrence is in the background.

great-power deterrence: future trajectories There is nothing on the horizon in great-power relations that would readily disrupt the recessed nature of contemporary deterrence among them. The chief sources of their political frictions are a fundamental disagreement about the sanctity of sovereignty in the face of behavior that the international community regards as unacceptable and the democracies’ objections to authoritarian rule in Russia and China. The current great-power deterrence situation, however, might not last indefinitely. New technologies, for instance, could unevenly allow great powers to try to escape from mutual deterrence. A major breakthrough in missile defense or in directed energy weapons and rail guns, the continuing genetic engineering revolution, or a breakthrough in space operations could undermine some great powers’ satisfaction with their deterrence situation, particularly if they were then caught up in a major international crisis. The low-level salience of deterrence also could be imperiled by domestic political changes that bring on highly aggressive regimes or the emergence of highly risk-acceptant leaders with ambitious international agendas. Extended-deterrence commitments could draw a great-power patron into engaging militarily with another great power on behalf of a client government—the Taiwan scenario. As Dinshaw Mistry notes in his chapter on the India-Pakistan rivalry, the regime in Islamabad hoped for such external intervention when staging a crisis with New Delhi in 1999. Much has been made of how globalization may facilitate the decline of intense rivalries. Where regime legitimacy and political survival are tied to delivering rising living standards, military conflicts would put the economic opportunities globalization offers at great risk. On the other hand, globalization will produce economic winners and losers, and losing will undermine the political, cultural, and other noneconomic roots of various regimes. The

deterrence in an era of globalization

273

result can be intrinsic or politically stimulated nationalism and resentment that drives destabilizing behavior in the international system. There has been more than a little of this in Russia, for example, since the end of the cold war. And economic winners—the United States in the 1990s—can embrace grandiose ambitions and overconfidence as a result. Power transitions among great powers have historically been very stressful, with rising tensions and often warfare between rising and declining states.30 Technological change and globalization are already creating developments that, if continued, will produce important power transitions in the foreseeable future. Under these circumstances, the great powers may feel compelled to give deterrence greater salience to sustain peace by constraining each other’s military options.

conclusion Deterrence has not disappeared in great-power relationships. It continues to play various important roles. This deterrence is, however, considerably more complex than in the cold war. It is more recessed, more implied. Being quite open about it can get a government charged with inciting suspicion and hostility, playing on insecurities, and accenting the negative. The Russians rail against NATO expansion and basing United States–sponsored ballistic missile defenses in Poland and the Czech Republic, but they have to live with the fact that if they actively try to deter these “threats” with direct military preparations, retargeting their missiles and deploying ground forces, such hostile behavior could totally upset Russian relations with the West, something no Russian government wants to contemplate. The U.S. preoccupation with missile defense must refer constantly to North Korea, Iran, and the like and never to China or Russia, so Washington bends over backward to stress that “national missile defense” will not erode Chinese or Russian security. It is no coincidence that actual U.S. missile defenses remain relatively modest. Deterrence remains, yet strategic offensive systems continue to decline in number and to deteriorate with age, while in China strategic modernization, long overdue, is approached with great caution. China and the United States merely test antisatellite systems without moving toward deployment. Russia threatens to pull out of the Treaty on Conventional Armed Forces in Europe but has been clearly reluctant to do so. The deterrence threats are more indirect, more subtle, more politically acceptable. In effect, the great powers are behaving much like a great-power concert

274

pa t r i c k m . m o r g a n a n d t. v. pa u l

should. Collectively, they want (1) to proclaim they are not enemies and often behave accordingly, (2) to cooperate on global / regional security management and other important matters, (3) to avoid posing obvious threats to each other, and (4) to keep their political and other conflicts within limits. At the same time, they also want to be able to take the various steps necessary to hedge—and thus be ready to resume active deterrence if necessary—and to maintain the necessary military capabilities to achieve their objectives. Considerable complexity is introduced into their deterrence calculations by their efforts to hedge while still attempting to achieve their four primary objectives. Of the great powers, the United States has the most active concern about deterrence but only, among those powers, with China, and that concern has not been expanding. Russia has perhaps the most immediate concerns about the deployment of ballistic missile defenses in Europe and NATO expansion. But Moscow is rather weakly employing military deterrence and is instead relying more on political and economic measures to achieve its objectives. China’s deterrence concern is in anticipation of a Taiwan conflict, and it is carefully preparing only for that deterrence effort and not frantically ramping up its overall deterrence posture. Britain and France have deterrence concerns but are only marginally worried about other great powers. There are hawks on deterrence in each country who want to give it more attention, but they do not seem to be directing great-power relations. The great powers will need to deal responsibly with threats to the benign configuration of their current relationships, some that arise out of their efforts to deal with threats and problems beyond great-power relations and the others that are linked to the great changes taking place in the international system. There are also all the problems arising from technological change and the pursuit of more precise uses of force and threats. Given the complexity being added to deterrence as a result, pursuing further understanding of deterrence in theory and practice is not likely to become obsolete.

notes 1. On contemporary general deterrence among great powers, see Patrick Morgan, Deterrence Now (Cambridge: Cambridge University Press, 2003). 2. For varying positions, see John Lewis Gaddis, We Now Know: Rethinking Cold War History (Oxford: Clarendon Press, 1997); John Mueller, “The Essential Irrelevance of Nuclear Weapons,” International Security 13 (Fall 1988): 55–79; John Vasquez, “The Deterrence Myth: Nuclear Weapons and the Prevention of Nuclear War,” in The Long Postwar Peace: Contending Explanations and Projections, edited by Charles A. Kegley, Jr. (New York: HarperCollins, 1991), 205–23.

deterrence in an era of globalization

275

3. On the context-dependent nature of deterrence, see Paul C. Stern et al., eds., Perspectives on Deterrence (New York: Oxford University Press, 1989), 9. 4. Robert Jervis, The Meaning of the Nuclear Revolution (Ithaca: Cornell University Press, 1989). 5. Bernard Brodie first called nuclear weapons “absolute” and useful only for deterrence, in The Absolute Weapon: Atomic Power and World Order (New York: Harcourt, Brace, 1946), chapters 1–2. On their being “uncontestable,” see Richard J. Harknett, “State Preference, Systemic Constraints, and the Absolute Weapon,” in The Absolute Weapon Revisited: Nuclear Arms and the Emerging International Order, edited by T. V. Paul, Richard J. Harknett, and James J. Wirtz (Ann Arbor: University of Michigan Press, 1998), 53. 6. Joseph S. Nye, Jr., “Nuclear Learning and U.S.-Soviet Security Regimes,” International Organization 41 (Summer 1987): 371–402. 7. China, for instance, was taken to the brink of war by the Soviet Union at one point, despite its nuclear weapons. 8. Avery Goldstein, Deterrence and Security in the 21st Century: China, Britain, France, and the Enduring Legacy of the Nuclear Revolution (Stanford: Stanford University Press, 2000); Lawrence Freedman, The Evolution of Nuclear Strategy (New York: St. Martin’s Press, 1981). 9. A term of art, of a sort, used to indicate an attack that thoroughly wipes out the opponent’s retaliatory forces and thus is meant to be vast in scale. 10. T. V. Paul, “The Risk of Nuclear War Does Not Belong to History,” in The Waning of Major War, edited by Raimo Väyrynen (London: Frank Cass, 2005), 113–32. 11. Treaty between the United States of American and the Russian Federation on Strategic Offensive Reductions, May 24, 2002, http://www.nuclearfiles.org / menu / library/ treaties / strategic -offensive-reduction / trty_strategic-offensive-reduction_2002–05–24.htm. 12. China has apparently not reduced its nuclear forces and may even have modestly increased them. 13. Arms Control Association, “What Are Nuclear Weapons For? Recommendations for Restructuring U.S. Strategic Nuclear Forces” (Washington, DC, November 8, 2007). Moscow has proposed renewing the 1991 Strategic Arms Reduction Treaty that ends in 2009, with more cuts and with restrictions on these recessed weapons, but the United States has not agreed. See Keir A. Lieber and Daryl G. Press, “Superiority Complex,” Atlantic Monthly ( July–August 2007): 86–92. 14. See Robert Jervis, American Foreign Policy for a New Era (New York: Routledge, 2005). 15. Charles S. Gochman and Zeev Maoz, “Militarized Interstate Disputes, 1916–1976,” Journal of Conflict Resolution 28 (December 1984): 587. 16. “Enduring rivalry” means at least twenty to thirty years long and six to eight conflicts, according to Paul F. Diehl and Gary Goertz, War and Peace in International Rivalry (Ann Arbor: University of Michigan Press, 2001), 44, 48. 17. Gary Goertz and Paul F. Diehl, “The Initiation and Termination of Enduring Rivalries: The Impact of Political Shocks,” American Journal of Political Science 39 (February 1995): 30–52. 18. For example, the Cold War; United States–China 1949–71; Soviet Union–China 1960–90; Arab-Israel 1948–; India-Pakistan 1947–. On the last see T. V. Paul, ed., The India-Pakistan Conflict: An Enduring Rivalry (Cambridge: Cambridge University Press, 2005). 19. Paul Huth and Bruce Russett, “General Deterrence between Enduring Rivals: Testing Three Competing Models,” American Political Science Review 87 (March 1993): 61–73. 20. In general deterrence, an actor believes a rival might readily attack given a clear opportunity but is not, for the time being, ready and eager to do so, and general deterrence is meant to discourage him or her from getting ready to do so. In an immediate deterrence situation, an actor believes an opponent is poised to attack and uses deterrence to prevent the attack from being carried out. See Patrick M. Morgan, Deterrence: A Conceptual Analysis, 2d ed. (Beverly Hills: Sage, 1983), 38, 42–3.

276

pa t r i c k m . m o r g a n a n d t. v. pa u l

21. On existential deterrence see Morgan, Deterrence Now, 23. 22. This is illustrated in Russian and Chinese military doctrines. See http://www.globalsecurity .org / military/world / russia / doctrine.htm; and Government of the People’s Republic of China, Information Office of the State Council, “White Paper on China’s National Defense in 2002,” December 9, 2002, chapter 1, http:// english.people.com.cn / features / ndpaper2002 / nd1.html. 23. See Richard Rosecrance, The Rise of the Trading State (New York: Basic Books, 1986), chapter 2; T. V. Paul, “Soft Balancing in the Age of U.S. Primacy,” International Security 30 (Summer 2005): 46–71. 24. T he basic draft plan is in Director, Plans and Policies, Deterrence Operations Joint Operating Concept, Version 2.0 (Washington, DC: United States Strategic Command, Department of Defense, December 2006). See also Hans M. Kristensen, “US Strategic War Planning After 9 / 11,” Nonproliferation Review 14 ( July 2007): 373–90. 25. Kristensen, “US Strategic War Planning after 9 / 11.” 26. Roger Speed and Michael May, “Assessing the United States’ Nuclear Posture,” in U.S. Nuclear Weapons Policy: Confronting Today’s Threats, edited by George Bunn and Christopher F. Chyba (Washington, DC: Brookings Institution Press, 2006). 27. Keir A. Lieber and Daryl G. Press, “The Rise of US Nuclear Primacy,” Foreign Affairs 85 (March–April 2006): 42–54. 28. Olivier Debouzy, “French Nuclear Deterrence Doctrine: An Aggiornamento,” European Affairs 7, nos. 1–2 (Spring–Summer 2006): 70–76. 29. A recent example: Office of the Secretary of Defense, “Annual Report to Congress: Military Power of the People’s Republic of China, 2007” (Washington, DC: Department of Defense, 2007). 30. On power-transition theory and the perils of power transitions, see Paul Kennedy, The Rise and Fall of the Great Powers (New York: Random House, 1987); A. F. K. Organski and Jacek Kugler, The War Ledger (Chicago: University of Chicago Press, 1980); Jacek Kugler and Douglas Lemke, eds., Parity and War (Ann Arbor: University of Michigan Press, 1996); Ronald Tammen, ed., Power Transitions, Strategies for the 21st Century, (New York: Chatham House, 2000); Charles F. Doran, The Politics of Assimilation: Hegemony and Its Aftermath (Baltimore: Johns Hopkins University Press, 1971); and Dale C. Copeland, The Origins of Major War (Ithaca, N.Y.: Cornell University Press, 2001).

chapter twelve R

The Endurance of Extended Deterrence Continuity, Change, and Complexity in Theory and Policy t i m o t h y w. c r aw f o r d

o n e often hears that the simple rules of deterrence that provided “comfort” and “enormous confidence” in the cold war are now obsolete.1 But there never was a golden age of uncomplicated superpower nuclear deterrence. Such claims gloss over the big problem of protecting allies in Europe that drove America’s heavy dependence on nuclear threats. That awkward reliance did not inspire comfort or confidence. It created “fundamental and vexing dilemmas” and “critical tensions” that fueled intense doubt and debate.2 The problem of “basic” deterrence was tied to extended deterrence, and one could not understand either without reference to alliance politics. Today the junctures between deterrence and alliance politics remain important focal points of inquiry and theoretical progress.3 If extended deterrence was key to the cold war deterrence paradigm, then that paradigm has not been cashiered. In October 2006, after North Korea’s nuclear test, U.S. officials leaped to make deterrent threats against Pyongyang and reassure regional allies. In Tokyo on October 18, U.S. secretary of state Condoleezza Rice declared in a joint news conference, “The United States has the will and capability to meet the full range—and I underscore full range—of its deterrent and security commitments to Japan.”4 That emphasis on “will and capability” came straight from the playbook of extended-deterrence theory. Three days later, U.S. defense secretary Donald Rumsfeld pledged “assurances of firm U.S. commitment and immediate support” and South Korea’s continued inclusion under a protective “nuclear umbrella.”5 Extended-deterrence priorities and principles have profoundly 277

278

t i m o t h y w. c r aw f o r d

shaped U.S. policy in Asia since the end of the cold war and will continue to in the foreseeable future.6 The logics of extended deterrence also feed into American involvement in the Middle East. Formal defense treaties between the United States and Iraq have been struck, and one recent commentator called for the United States to deploy “‘tripwire’ or deterrent forces athwart the most likely invasion routes from Syria and Iran and so signal to both that they would have to attack the United States if they attack Iraq.”7 Such external arrangements beg a more pressing question of extended deterrence: how U.S. officials should order their commitments vis-à-vis Iraq’s internal factions—or future rump– states—so as to create stability rather than incentives to escalate ongoing conflicts.8 Extended deterrence also is a potential response to Iran’s nuclear threat. A most startling demonstration of this occurred in May 2008 when Hillary Clinton, in the midst of the Democratic party’s presidential primaries, declared that as president she would “provide a deterrent backup” to Arab allies and Israel, promising “massive retaliation” against Iranian nuclear use and noting, just for good measure, that the United States “would be able to totally obliterate them.”9 Efforts to bolster U.S. extended-deterrence commitments against Iran are likely to persist for two reasons beyond the obvious one. First, doing so offers a plausible alternative to the certain costs, and almost certain failure, of the preventive war option. Second, it is hoped that U.S. security commitments can dissuade Middle East allies from pursuing their own nuclear options.10 On a broader scale, the long-term viability of the Nonproliferation Treaty regime may hinge on the credibility of nuclear powers’ “positive assurances” to non-nuclear-weapons states that they will be protected from nuclear aggression and coercion—on the credibility, in other words, of extended-deterrence threats and promises.11 Extended-deterrence thinking also has penetrated into other growth areas of international security, especially under the aegis of what Patrick Morgan has termed “collective-actor deterrence.”12 In the heyday of the 1990s “new peacekeeping,” the United Nations (UN) Security Council passed numerous resolutions concerning “safe areas” in the former Yugoslavia and elsewhere that explicitly tied the deployment of peacekeepers to the goal of deterring attacks on civilian enclaves including, unfortunately, Srebrenica. The frequently heard claim that the Rwandan genocide could have been averted by a timely deployment of just a few thousand “robust” UN peacekeepers rests on an implicit—and heroic—assumption about the effectiveness of extended deterrence.

the endurance of extended deterrence

279

Calls for third-party intervention and guarantees to end civil wars involve complex extended-deterrence problems of credibility and perverse incentives.13 How can parties removed from a potential conflict credibly commit to intervene, even when the ex post incentives to abstain are high? And if they make such commitments too strongly, will they encourage the local actors to run greater risks and escalate disputes?14 These problems have always been embedded in extended-deterrence theory, and the tension between them helps to define some of its main political features. To address these issues, the chapter first describes the strategic setting known as “extended deterrence.” It then explains why certain elements of the extended-deterrence problem have declined in importance, while others continue to command attention, and why still others have become more salient today.15 The result is a mixed picture of progress in thinking about extended deterrence, moving toward a more complex view of its political contours and participants. By way of conclusion, connections will be drawn between these richer conceptions of extended deterrence and current U.S. security policy challenges.

e x t e n d e d d e t e r r e n c e : p o l i t i c s , c o m p l e x i t y, and concatenated theory “Extended deterrence,” as one observer put it, “is essentially political, only more so.”16 Pure calculations of the military balance aside, political factors— and complex compounds of actor beliefs and motives—loom large in the dynamics and outcomes of extended deterrence. Extended-deterrence research thus revolves, as Alexander George and Richard Smoke suggested, around a “concatenated” theory: a merger of assumptions and propositions about different actors, from different vantage points, that describes and explains a pattern of strategic interaction that converges “on some central point.”17 It presents a picture of potential interaction among at least three strategic actors that “converges on the central point” of an attempt by one of them to deter another from attacking a third. Those three parties are conventionally designated the defender (the extender of deterrence), the challenger (the actor to be deterred), and the protégé (the target of the challenge and party to be defended). Beyond this basic cast of characters, extended-deterrence theory distinguishes between two types of deterrence situation—general and immediate.18 General extended deterrence occurs when there is a standing “covering” relationship between defender and protégé and the challenger does not

280

t i m o t h y w. c r aw f o r d

attempt to coerce the protégé because it anticipates and fears the defender’s response. An immediate extended-deterrence attempt begins after general deterrence has failed—that is, after a challenger has begun openly to coerce the protégé—and the defender has threatened overtly to intervene. If the challenger backs off because of the defender’s actions, then immediate extended deterrence succeeds. It fails if, defying the defender, the challenger attacks the protégé. A variety of motivations, intentions, and relative capabilities can be attributed to the actors in this triangle, without rendering nonsensical the “convergence on the central point.” In the cold war, a particular configuration of assumptions about such things became dominant. What drove the conventional assumptions was the strategic situation confronting the United States. There was always a connection between security policy problems and the development of extended-deterrence theory. The 1990s ushered in important shifts on both sides of the equation, but that connection continues. As the salient policy problems have changed, so too has the emphasis placed on elements of the concatenated theory. Nevertheless, some of the focal points of cold war theorizing remain compelling and controversial.

past problems in decline Uncredible First Use The cold war in Europe case of extended deterrence evinced extreme values on a number of critical variables. This was especially true in respect to what Lawrence Freedman called “the dominant issue in Western nuclear strategy,” the North Atlantic Treaty Organization (NATO) threat to go nuclear first in response to a conventional Soviet attack, “despite the evident risk of a Soviet nuclear counterattack.”19 This dominant issue was an especially ugly version of an ever-present predicament of extended deterrence: to protect others from attack, the party making the deterrent commitment may risk paying costs that exceed the inherent value of the party being protected. Today, this problem has receded, at least for the United States. The need to rely on the incredible nuclear threat stemmed from NATO’s conventional weakness. It does not arise if the challenger is inferior in conventional power or lacks a nuclear retaliatory capability. The United States is the only power extensively engaged in extended-deterrence efforts today, and none of the potential challengers are so strong conventionally as to require nuclear threats to bolster the credibility of U.S. positions. The closest approximation

the endurance of extended deterrence

281

is the U.S. position in the Taiwan Strait, and although mainland China does have a minimum nuclear deterrent, there is no need for the United States to adopt a first-use threat when it has, especially in conjunction with allies, the conventional upper hand. The United States still must deal with the risk that trying to deter attacks on others will end up costing more than they are worth. But now the whole equation does not sum up to an implausible promise to commit suicide. The Perverse Incentive to Free Ride In cold war Europe, the perverse incentives of a strong U.S. extendeddeterrence posture were conceived in a singular way—as a problem of “free riding.” To many observers—especially in the U.S. Congress—the allies’ frequent failure to meet NATO force goals was proof of shirking.20 The primary concern was that the allies were withholding from NATO the strength it needed to deter Soviet aggression. But it was intermingled with the suspicion that the allies shirked because the United States was “overcommitted.” As Henry Kissinger argued in 1957, the conviction that the United States would have to respond to a Soviet attack on Western Europe “with all-out war whatever might be the contribution of its allies . . . has been the central reason for NATO’s lagging military effort even before the missile age.”21 Although a burden-sharing debate continues in today’s transatlantic strategic discourse, the logic and consequences of the free-riding problem are no longer tied directly to the extended-deterrence equation but instead to out-of-area operations relating to the war on terrorism and other concerns. Shrinking European defense spending is due mostly to the facts that state-based military threats to Europe have declined, that the European members of NATO have become pacific toward each other, and that Europeans do not aspire to an activist role in policing global security. There also may be a shared presumption that the United States will inevitably uphold common out-of-area interests. But if so, that presumption does not grow out of U.S. extended-deterrence commitments per se. There is no longer a major threat to Europe to be deterred by such commitments or to be emboldened by buck-passing allies. The Surprise Attack Scenario The theory of progression from general to immediate deterrence helps to elucidate the logic of the surprise attack scenario. The presumed sequence

282

t i m o t h y w. c r aw f o r d

explicitly gives the defender a chance to attempt immediate deterrence after general deterrence has failed but before the challenger attacks. But in the surprise attack scenario that worried so many during the cold war, the Soviets would try to leapfrog the immediate deterrence phase—going from a static situation directly to an all-out attack on Europe. In addition to all of the military advantages, surprise would garner political leverage by eliminating any opportunity for the defender to issue explicit deterrent threats or otherwise bolster its commitment. Again, the concern was driven by fear that the Soviets were in a position of such conventional strength that they might be tempted to launch a “standing start” attack in Europe that would present Washington with a fait accompli. The element of surprise was critical because it would prevent the United States from being able to work up to its nuclear commitment incrementally. The Soviets could pose the United States and NATO with the stark choice between acquiescence and a suicidal nuclear response all at once because they would be able to make rapid gains on a conventional battlefield against much weaker NATO defenses.22 At the outset of Surprise Attack, a 1982 study of NATO scenarios, Richard Betts mused that a generation hence “the best possible verdict” on the book “would be that it had been superfluous or alarmist.”23 Although it was neither of those things, it is nevertheless true that the problem of surprise attack created by general extended deterrence in Europe has faded. The specter of a similar nightmare scenario is occasionally raised in relation to a North Korean attack on South Korea or of a mainland China attack on Taiwan. But the suspension of disbelief required to imagine a massive and unanticipated attack in these circumstances is much harder to achieve. The arguments for the pure strength of conventional deterrence and defense also are much stronger.24 So this surprise attack version of extended-deterrence failure has gone into semiretirement, to be replaced, perhaps unfortunately, with the surprise scenario of mass casualty terror attacks by nonstate actors, a marked departure from the extended-deterrence paradigm.

continuities and change Extended deterrence entails, by definition, a credibility problem. The discrepancy between self-interest and interest in others is what sets it off from “basic” deterrence. In basic deterrence, the driver of influence boils down to capability: does the defender have the ability, after it is attacked, to impose

the endurance of extended deterrence

283

costs greater than the challenger’s benefits of attacking? If it does, there is a strong chance the defender will impose them if it can—strong enough, at least, that a reasonable challenger aware of those capabilities will assume that the defender will hit back.25 The capabilities question of credibility pertains to all kinds of deterrence situations: does the defender have the power readily available to impose costs and does the challenger know it? The question of resolve involved in credibility (does the defender have the political will to impose costs?) becomes trickier when the issue shifts from basic to extended deterrence. If there is little doubt that a defender will pay the price of punishing an attack upon itself, there is rather more doubt that it will pay the price of punishing an attack on another actor. If the defender’s price for meting out punishment were zero, there would be little doubt that a response would be forthcoming regardless of the target of the initial attack. But the price of responding to an attack is never zero, and in the important scenarios in international politics, the potential price involved in honoring one’s deterrent threats is usually quite high. Commitment Measures Deterrence theorizing, especially in the wake of Thomas Schelling’s pioneering work, gravitated toward finding ways to generate resolve, or “commitment,” through political trickery and theater.26 There are general and immediate modes of commitment. The deployment of “trip-wire” forces, or a transfer of arms and military advisors, to a vulnerable locale, prior to a challenge, can bolster the credibility of a general extended-deterrence posture by “coupling capabilities to objectives.”27 Indeed, most of the sinews of “alignment” can be put into the category of general-deterrence commitment. By contrast, public declarations of intent or warning signals sent through reputable intermediaries or the dispatch of gunboats to an area after a challenge has been made are commitments to bolster the credibility of immediate extended deterrence. The general mode of commitment is central to prevailing U.S. grand strategy and thus a ubiquitous feature of today’s security environment. In recent years, from Eastern Europe to the Middle East, Africa, Central Asia, and South East Asia, the United States has recruited new formal and informal allies, “sinking costs” in these partners in the form of concrete basing infrastructure, military-to-military training and exercises, and preferential arms sales, investing prestige and political capital in these relationships. And the

284

t i m o t h y w. c r aw f o r d

conventional wisdom is that these efforts to shape the international environment through alliances and “forward presence” will tend to discourage regional aggression—in other words, that it will tend to promote general deterrence. As for immediate commitment, the focus has increasingly been on the role of “audience costs” in generating credibility for deterrent threats, especially for democratic states. Audience costs are incurred when leaders are punished politically at home for reneging on strong foreign policy commitments (especially those ratified by popular or institutional means).28 Leaders can thus use these expected audience costs to improve their bargaining positions, signaling to adversaries that they mean business by making public threats that are hard to renounce. Audience-cost logic provides the most traction in the realm of immediate extended deterrence, when challengers have already chosen to disregard the threats embodied in existing relations between the defender and protégé. Since some challengers will disregard those threats because they doubt the defender’s commitment, the defender’s costly public reaffirmation of commitment may be enough to alter challengers’ expectations and stop them from seeking further escalation of a crisis. Although compelling, the audience-cost argument does run into methodological and, more importantly, scope limitations.29 Most limiting is the disconnect between the abstract logic of firm and crisp public commitments and the reality that in the heat of crises coercive diplomacy often unfolds in secret or at least beneath a veneer of public statements carefully tailored to permit flexibility and avoid provocation.30 Moreover, there are other ways in which the importance of audience costs in making extended-deterrence threats may be diminished. If foreign policy in general—or the particular issue over which the leader makes a threat—is not a salient concern to the public, then leaders will have more freedom to bluff and back down. In politically complex situations, it also may be hard to designate just one of the parties as a clear-cut threat to the peace, which means it will be hard to convey to internal audiences a compelling conflict narrative that would inspire a strong conviction that commitments must be upheld. Finally, in conflicts involving ambiguous and contradictory interests and commitments, leaders will have more leeway to define and redefine how and whether they keep their word because, in these circumstances, “people are likely to take their cues” from their leaders.31 Such considerations point to a diminishing ability of U.S. leaders to generate leverage from audience-cost tactics.

the endurance of extended deterrence

285

Reputation Over the years, heavy emphasis, especially among practitioners, was placed on reputation as a way to bolster the credibility of extended-deterrence commitments. The main themes married the politics of deterrence and alliance relations. A reputation for standing firm and paying high costs, even for lesser stakes, would generally strengthen deterrent threats and reassure the allies who depended on them.32 This thesis became an important target of critics and for empirical testing in the later years of the cold war, and it continues to be so today.33 Jonathan Mercer, for example, mobilized psychology to argue that countries tend to believe that their adversaries back down because of external constraints rather than weakness of will, in which case, it makes no sense to stand firm in defense of weak material interests just to protect or build a reputation.34 Later, Daryl Press reinforced the skeptic’s case, arguing that challengers’ decisions were based heavily on estimates of their adversaries’ current military capabilities and interests at stake, not on inferences drawn from their previous actions in similar circumstances.35 In a seminal critique, Patrick Morgan argued that the impulse of U.S. leaders to stand firm in situations of limited interest (e.g., Vietnam), in order to build a reputation for resolve, said little about what they would be willing to do in a nuclear-tinged crisis in Europe. “Nuclear powers may uphold lesser commitments because they are the only ones they can certainly decide to fight for . . . This converts the concern for reputation from a rational extension of the art of commitment into a pretense used to hide a pervasive insecurity over what to do if one’s most important commitments are challenged.”36 With the cold war over, and the need to rely on the incredible nuclear threat for extended deterrence reduced, this critique seems less telling. But its core contention—that there was more behind the U.S. pursuit of reputation than rational technique—gained force in the 1990s. Even when the United States had overwhelming conventional superiority, and thus highly credible options for extending deterrence without going nuclear (and risking nuclear retaliation), its leaders would still routinely invoke the need to gain or save face for the sake of credibility in all manner of unilateral and multilateral security affairs.37 The idea that resolve and reputation are both important to credibility, and easily squandered, remains in currency today. It is often claimed that al-Qaeda was emboldened to strike on September 11, 2001, by the U.S. failures to stand firm in Lebanon and Somalia. And with even more conviction, the George W. Bush administration, along with other U.S. officials, claimed that the United States must persevere in Iraq lest failure in

286

t i m o t h y w. c r aw f o r d

the Middle East cause U.S. allies to lose faith in American commitments and embolden terrorists.38 Even if the possible connection between reputation and credibility is stronger than skeptics admit, there is reason to doubt that the connection today is very robust, especially when it comes to U.S. extended-deterrence commitments. If a reputation for resolve added to the credibility of U.S. commitments in the cold war, logic would suggest that it has lost some of that value today. The nature of the cold war contest supported a plausible argument that reputation did matter when it came to deterring the other superpower and its clients. After all, there were many important commitments continuously vulnerable to challenge by the same superpower competitor or its allies and proxies. The unipolar logic of American preponderance does not provide such a rationale for the importance of reputation. A military and economic superpower, without any serious near-term peers, has much more freedom of action than one that confronts a roughly equal, and determined, competitor. It can pick and choose battles, change its mind, and renege on promises with relative ease.39 Doing so will not increase the power and relative prestige of a peer competitor. Potential partners in one part of the world also are unlikely to spurn American protection from local foes because the United States treated a protégé in another part of the world shabbily. When you are the only major “defender” in the system with global reach, protégés come cheap. A reputation for resolve might matter now more than ever. There are, after all, many circumstances in which the United States might like to ride on a reputation for sticking to its guns, defending friends, allies, and other assorted protégés. But the demand for a reputation is not equivalent to its supply. A longer look at today’s circumstances suggests that the systemic costs that would discipline the United States into paying a price to avoid a bad reputation are much lower. Consequently, even if U.S. leaders are convinced of reputation’s importance, U.S. adversaries are unlikely to believe that it would really pay high costs to stand firm when the concrete strategic consequences of backing down are minimal. If attempts at extended deterrence by collective actors become more frequent—especially in civil war situations—then there are strong reasons to doubt that the reputation-credibility connection will play a part in such activities. Two basic premises of these arguments are that actors cannot easily decouple their prior behavior in one place from their behavior in another, and that reputations especially matter when there are strong contextual similarities between past and present situations. By contrast, collective-actor

the endurance of extended deterrence

287

deterrence activities diffuse responsibility and blur the connection between past and present multilateral actions and actors. From the perspective of member states, one of the great virtues of working through the UN is that they can shift blame to “it” when things go badly. And parties that do confront a UN collective actor are not likely to presume that it is the same UN actor, with the same resolve propensities, that another party, in another place, at another time, confronted. A Sudanese warlord would not be more likely to believe that Egyptian blue helmets would defend a safe area in Darfur because Dutch peacekeepers had fought to the last man in Srebrenica. Intrinsic Interests These changes in the strategic setting tend to elevate the importance of context-specific political factors—the “intrinsic” or “inherent” interests at stake.40 The idea of intrinsic interests thus tries to identify the elemental political value of the protégé to the defender, apart from any additional utility the defender may derive (i.e., in reputation or prestige) from standing up to those who threaten it. It presumes that adversaries judge the credibility of the threats they face in terms of the strength of the underlying interests motivating the parties posing the threat. Intrinsic interests are thus not identical to intense interests—they may be strong or weak or somewhere in between—but attention to intrinsic interests inevitably segues into drawing distinctions among them and thus to the matter of vital and secondary interests, values that are presumed to be independent of commitment tactics or prestige concerns. Thus, a “trip-wire” force deployed on the territory of a contiguous or otherwise critical ally will not generate the same credibility as a trip-wire force deployed elsewhere. One hundred American military observers sent to Israel in a crisis is a stronger signal than one hundred sent to Macedonia in a crisis—context matters. Those who subscribe to a grand concept of post–cold war U.S. leadership often note that America’s interests have expanded with its climb to system preponderance. More plausible and coherent is the opposite view. If some remote corners of the cold war world gained strategic value to the United States because controlling them, or denying them to Soviet control, would improve the ability of United States to protect vital interests against Soviet threats, they must inevitably matter less with the Soviets out of the equation. Since 1991, the United States has been in a position in which the absence of a major geopolitical rival has exposed its extended-deterrence attempts to much scrutiny. Paradoxically then, the credibility of American extended

288

t i m o t h y w. c r aw f o r d

deterrence vis-à-vis local challengers has been weakened by post–cold war preponderance. At the very least, the gains in credibility that come with its unmatched military power are offset by a weakening of strategic interests that comes with having no peer competitor. Credibility in specific situations must rest heavily on the local military balance, because where the United States seeks to be a defender, it usually confronts challengers with much higher stakes in the outcome of a potential conflict. The slogan that “a nation’s interests increase with its power” glosses over the question of the relative importance of those additional interests. If increasing one’s power increases one’s security—and it would be perverse to maintain that the most powerful are the least secure—then the interests accumulated “on the way up” should diminish in intrinsic value. To be the superpower in a unipolar world, therefore, is to be constantly tempted to intervene in matters of less importance to you than to the local protagonists. The logic of intrinsic interest suggests even more difficulty in achieving collective-actor extended deterrence. Collective actors (by any reasonable understanding of the term) do not have intrinsically vital interests at stake in a conflict; however, their opponents often are literally fighting over matters of life and death. Looking at how issues of commitment, reputation, and interest relate to the shape of current international politics suggests confounding dilemmas. From the perspective of the United States, military dominance has multiplied the political complexities of acting abroad, and while the reasons and invitations to extend deterrence proliferate, the political support needed to sustain such efforts has diminished. Extended deterrence is a promising option in fewer circumstances than may seem to warrant deterrent commitments. This also implies limits to the extent to which confidence in extended deterrence can be transposed to such collective-actor domains as conflict management in civil wars or the use of broad positive assurances to support nonproliferation. There will likely be new venues for unalloyed and narrowly tailored extended deterrence. Depending on how China’s rise unfolds, Vietnam, Singapore, or Indonesia could become beneficiaries of new and explicit American security guarantees. If they consolidate under their current governments, Afghanistan and Iraq will likely become bona fide American allies, with a U.S. extended-deterrence commitment to back up their political and territorial integrity. And if Iran crosses the nuclear-weapons threshold, more of its neighbors may be swept under a reinforced American protective umbrella. Beyond these fairly straightforward scenarios there are other, more complicated ones coming into view. That the political conditions may

the endurance of extended deterrence

289

not be ripe for extended-deterrence commitments will not alleviate the need to make them. Threats, (Re)Assurances, and Accommodation “The object of threat,” wrote Schelling, “is to give somebody a choice. To say, ‘one step more and I shoot’ can be a deterrent threat only if accompanied by the implicit assurance, ‘And if you stop I won’t.’”41 This theme was embellished in the deterrence theorists’ claim that threats are more likely to succeed when they are coupled with bargaining tactics that reduce or “limit the costs of retreating to the other side.”42 Attention to the role of assurances and promises in coercive diplomacy emerged in key works of the 1970s. By the 1980s, injunctions to deal with challengers through “deterrence and reassurance” and “firmness but flexibility” were staples of leading extended-deterrence studies.43 Yet some confusion about these matters remains. In particular, the term “reassurance” has been used to denote two opposite relations in an extended-deterrence situation: as a form of positive influence to be combined with threats to curtail an opponent driven by insecurity, needs, and weakness and as a form of positive influence toward allies, which calms them and inspires confidence.44 There are four types of influence that may concatenate in an extendeddeterrence attempt, thereby contributing to its success. Reassurance bolsters a protégé’s confidence in the defender’s commitment. Accommodation, by contrast, makes some sacrifice on behalf of the challenger’s interests for the sake of preserving the peace.45 Restraining threats try to control a protégé by raising its fear of abandonment. And straightforward deterrent threats try to stop challengers by threatening punishment or active defense. Two key points follow from this framework. In an immediate deterrence crisis, reassurance is not merely an alternative or supplement to extended deterrence but a part of it. When a defender attempts to derail a direct challenge against its protégé by threatening to intervene, it also reassures the protégé (unless the protégé views these threats as unnecessarily provocative). Likewise, when a defender restrains a protégé in a crisis, it also accommodates the challenger.46 The combination of such techniques may be what determines an extended-deterrence outcome. To be deterred, a challenger may need to be partly accommodated because it is motivated by vulnerability as well as by opportunity. The protégé, likewise, may need to be restrained if it is to be protected, because it is motivated by opportunity as well as vulnerability.

290

t i m o t h y w. c r aw f o r d

These mixed assumptions about players’ motivations generate a politically complex pattern of deterrence dynamics, and we will return to them.

change and complexity There is much disagreement about the role of ambiguity in deterrence. The dominant view is that ambiguity is detrimental insofar as it muffles signals, suggests weakness and evasiveness, and encourages miscalculation: if deterrence is the goal, it is best to make clear, credible, publicly communicated threats.47 Other scholars have noted how ambiguity may contribute to deterrence success.48 Vague threats may make it easier for targets to comply while saving face and may help to preserve a range of relationships with a target that would be sacrificed by explicit threats. Military considerations also may increase the importance of ambiguity. A highly explicit threat can reveal too clearly the threat maker’s capabilities, creating the opportunity for countermeasures that can weaken the deterrent threat. It is important to distinguish between the general motivation to remain flexible in one’s foreign policy commitments and the strategic rationale for pursuing deterrence through ambiguity. The conventional view is that ambiguity is useful insofar as it hedges the costs of closed-off options but provides little influence over the adversary’s behavior. By contrast, the strategic rationale for deterrence through ambiguity posits that flexibility can influence adversaries by shaping their choices and constraining their freedom of action. According to Schelling, the basic idea is that “one may deliberately choose to be unclear and to keep the enemy guessing . . . to enhance his anxiety.”49 Or, as Glenn Snyder and Paul Diesing put it, “the ‘keep-em-guessing’ formula attempts to give some coercive mileage by introducing uncertainty into the opponent’s calculations.”50 Several assumptions underpin the claim that strategic ambiguity is more dangerous than clarity because it invites war through miscalculation. The first is that only the challenger will be a miscalculating war causer. Starting with that assumption makes it harder to see how there could be a downside to maximizing the clarity of a deterrent threat. Granted, the challenger might be tempted to press matters right up to the red lines, but the implication then is that one should be careful about where one draws lines not that they should be blurred. If the challenger needs face-saving ambiguity, then sequencing is crucial: clear and firm threats must come first, then a little sweetener of ambiguity to make capitulation easier to swallow. The downside of rigid public commitments is only apparent if one considers the

the endurance of extended deterrence

291

protégé a potential participant in the steps to war. This does not happen in most of the literature on extended deterrence because the protégé is usually depicted as a passive security seeker. Such a protégé will only cause trouble if it is unsure about the defender’s commitment and overreacts to a challenge. So the prescription is still the same. Increasing the strength of the defender’s commitment removes sources of miscalculation on both sides. The reassured protégé is less jumpy, and the challenger is more certain that the defender will protect it. When it is made explicit that the challenger or the protégé can escalate a crisis and that they both can do so in the pursuit of gains, the war-causing effect of firm one-sided defender commitments becomes evident. Firm commitments made by the defender create an incentive for the protégé to engage in adventurism. Recent efforts to explain the utility of ambiguity in extended deterrence all build upon the idea of the protégé as a potentially ambitious player in the strategic interaction.51 Just as the best strategy for dealing with a challenger depends on its motivations, the best way to restrain a protégé depends on its motivations. If it is driven by vulnerability, then strong reassurances will encourage moderation; if it is driven by opportunism, then strong reassurances will encourage it to act provocatively. Introducing that possibility into the extended-deterrence equation makes the basis for connecting strategic ambiguity to deterrence clearer. It is one way—however imperfect—to apply deterrent pressure in both the direction of the challenger and the protégé. The Protégé and Moral Hazard The possibility that challengers may be motivated by fear implies the possibility that protégés may be motivated by greed. But there is a strong tendency in most extended-deterrence research simply to ignore the protégé as a strategic actor.52 A good illustration of this is Paul Huth’s significant body of work on extended deterrence. He hypothesized that four critical factors were likely to determine whether a challenger complied with or defied an immediate extended-deterrence threat: the military balance between the challenger and defender, the defender’s interests at stake in the protégé, the defender’s reputation for resolve, and the defender’s style of military escalation and diplomatic bargaining.53 Nowhere did Huth take into account the protégé’s political posture or bargaining behavior. These factors can have an important impact on the challenger’s decisions and the outcome of extended-deterrence crises.

292

t i m o t h y w. c r aw f o r d

The fact that the protégé’s incentives and behavior have received little attention from scholars reflects a number of concerns and biases. First, there is a methodological incentive to simplify the model of interaction to a twoplayer game to make it more tractable for formal analysis.54 Second, there are at least two generic realist postulates that could encourage such a simplification. One is the notion that alliances are built on some degree of identical interests, which could support the assumption that the defender and protégé must share status quo preferences. But to start the analysis presuming that the protégé will not want to make gains at the challenger’s expense because the defender does not want the challenger to gain at the protégé’s expense is to approach this interaction from a politically naïve standpoint. It also eliminates an important domain of extended-deterrence interactions from consideration. The other realist idea is that weak states, dependent on others for security, can only have defensive ambitions.55 This passive-defensive assumption implies that the protégé does not seek conflict gains, will not initiate anticipatory defensive measures, and will not exploit the cover provided by an extended-deterrence policy. Perhaps such protégés are not rare, but neither are they so typical that extended-deterrence theorists can afford to assume that they represent the general population. There is no reason to equate dependence and “limited strength” with “passivity and resignation.”56 Yet, in the cold war context, the passive-defensive assumption made some sense. The European beneficiaries of U.S. deterrence were in fact strategically passive if not paralyzed security seekers. Greco-Turkish conflicts aside, there was no great worry that protection would encourage NATO allies to act rashly.57 The worry was that protection would encourage shirking. Because the European protégés were safely assumed to be passive-defensive and the European context dominated research and debate, the problem of restraining protégés was downplayed in extended-deterrence theorizing. For all of the emphasis on credibility problems, little attention was paid to the moral hazard created by too credible commitments.58 Moral hazard and the restraint of allies were always of greater concern in studies of alliance politics.59 Well before formal economics became engrossed with moral hazard, David Hume, in his 1742 essay “Of the Balance of Power,” observed that the British were “so declared in our opposition to French power and so alert in defense of allies, that they [allies] always reckon upon our force as upon their own; and expecting to carry on war at our expense, refuse all reasonable terms of accommodation.”60 The moral hazard problem was first discerned in the deterrence literature by George

the endurance of extended deterrence

293

and Smoke who noted that a “major risk and limitation of deterrence policies” was “the opportunities it may give a weak ally to manipulate the nature of the commitment made to it.”61 A bit later, Snyder and Diesing would put it more explicitly: “the issuing of [deterrent] threats or the underlining of alliance commitments may make the ally more confident or dangerously intransigent.”62 The whole idea was most powerfully elucidated next in Snyder’s article on the security dilemma in alliance politics, which established the risk of “entrapment” as a fundamental concept of alliance liability.63 And this, in turn, was the touchstone for James Fearon’s 1997 argument that the “problem of moral hazard in alliances and extended deterrence” could help to explain why defenders “shy away from absolute commitment” when the costly signals logic of credibility would instruct them to do otherwise.64 Nevertheless, extended-deterrence scholarship has been slow to absorb the larger implications of moral hazard. Despite intensive scrutiny of the tie between reputation and credible commitments, it was only recently that a scholar pointed to the perverse incentive created by a strong reputation for alliance reliability—that a state with reputable allies will be emboldened to challenge its enemies.65 The link to intrinsic interests has also received little attention. If extended deterrence is more credible the stronger the interests at stake, then as the defender’s intrinsic interests grow, so too does the risk of moral hazard. Thus, moral hazard may explain why some extended-deterrence attempts fail when the defender’s vital interests are at stake. High levels of moral hazard can even be associated with defender-protégé partnerships that entirely lack a formal alliance “contract.” As students of alliance politics know, when the natural bases for security cooperation are very strong, such formalities may be unnecessary. Recent work by Vesna Danilovic on the regional dimensions of intrinsic interests suggests additional ramifications of the moral hazard problem. Her quantitative results showed that the strength of a defender’s regional interests is a good predictor of whether it will intervene on behalf of a protégé.66 Danilovic’s findings thus help to establish a scope and domain for moral hazard effects in relation to certain conflicts. These points all suggest a larger fusion between moral hazard and extended deterrence that demands greater scrutiny. A highly credible general-deterrence situation may create for the protégé an enhanced baseline of security that encourages it to take more foreign policy risks that can trigger immediate extended-deterrence crises.67 In other words, in a strong general-deterrence situation, an ambitious protégé will be incited to press its adversaries to the point at which they may decide to mount a direct

294

t i m o t h y w. c r aw f o r d

challenge to end the provocation. Thus, the source of the challenger’s willingness to run risks may be the protégé’s provocative behavior, which is itself a reflection of the strength of the defender’s general commitment to the protégé’s security. This backstory to the initiation of challenges suggests a generic causal pathway to general extended-deterrence failure. Moral hazard incentives can operate at the immediate deterrence stage too, if commitment tactics do increase credibility. If so, then high moral hazard also can be associated with situations of low intrinsic interest to the defender. When it deploys costly public signals to overcome an intrinsic interest deficit, it erects an incentive structure that can be gamed by an ambitious protégé. As the defender’s commitment mounts, the protégé will gain increasing liberty to press its advantage in bargaining with the challenger. Thus, the presence of moral hazard and the possibility of an ambitious protégé can have major ramifications for the causal dynamics of extended-deterrence crises—both for the move from general to immediate deterrence and for whether the crises end in war or peace. Additionally, the moral hazard dynamic highlights how a defender’s accommodation attempts can contribute to successful extended deterrence. For the defender in an immediate extended-deterrence crisis involving an ambitious protégé, credible deterrent threats against the challenger may need to be coupled with a credible effort to restrain the protégé. Looking at the problem this way gives more political substance to the firm but flexible approach and “conditional promises of limited compromise” that defenders often must couple with threats to make extended deterrence work.68 Finally, the moral hazard angle restores the protégé’s role as a potential player in the outcome of immediate extended-deterrence crises. It can refrain from escalating a crisis and thus encourage the challenger to comply with the defender’s threat. Or it can do things that make it harder for the challenger to back down. Attempts to explain extended-deterrence outcomes that fail to consider the protégé’s role may suffer from a colossal omitted-variable bias. Pivotal Deterrence In pivotal deterrence, the defender sits in a pivotal position between ambitious adversaries, and it seeks simultaneously to deter both of them. Pivotal deterrence is a kind of extended deterrence in which each of the other two actors is part protégé and part challenger, and all four of the forms of influence described above are likely to be brought into play in crisis interaction. This concept introduces a high level of strategic complexity into

the endurance of extended deterrence

295

the extended-deterrence equation, and it raises questions about the scope conditions of certain key principles of extended deterrence. For example, for pivotal deterrers it makes little sense to establish strong clear commitments early in a conflict, since doing so will create moral hazard, increase the possibility of escalation, and hamper the pivot’s ability to apply offsetting pressures among the potential belligerents. The concept of pivotal deterrence can help explain extended-deterrence situations that otherwise fit badly into an abstract defender-protégé-challenger mold that presumes stark lines of amity and enmity. A good example is a country trying to deter escalation between two of its allies, as United States tried to do a number of times during and after the cold war vis-à-vis Greece and Turkey and Pakistan and India. Likewise, when a country tries to deter war between parties that are more or less its adversaries—as the United States did when it tried to prevent a nuclear escalation of the 1969 Sino-Soviet border fighting—the pivotal-deterrence framework can explain more than the standard extended-deterrence model. Moreover, the pivotal-deterrence concept complements Morgan’s collective-actor concept by describing the deterrent goals and dynamics often associated with collective-actor interventions in budding or broken-out conflicts. NATO’s initial attempts to deter escalation of the civil war in Kosovo, for example, are better understood in terms of pivotal deterrence than of the standard extended-deterrence scenario.69

conclusion: conceptual complexity and current challenges Extended-deterrence research has been enriched and sustained by the integration of political concepts that help to identify the techniques, motivations, interests, identities, and even the composition of the principal actors. The recognition, for example, that the outcome of an extended-deterrence crisis could depend on whether the defender used a hard-line or firm but flexible strategy pointed to a more nuanced image of defenders and challengers, which led to the insight that challengers may be motivated by fear more than greed. With an understanding of entrapment gleaned from alliance politics came the recognition that protégés may be motivated by greed more than fear. Out of these problem shifts came a clearer grasp of the dangers of overcommitment, the utility of ambiguity, and the drawbacks of overly simplistic assumptions about the actors’ interests and allegiances. Complexities are ramified by introducing the concept of collective-actor defenders that must

296

t i m o t h y w. c r aw f o r d

coordinate and cooperate among themselves while attempting to deter other parties from attacking each other. A look at two of today’s pressure points in U.S. security policy shows how richer conceptions of extended-deterrence situations are valuable. U.S. deterrence in the Taiwan Strait awkwardly conforms to the storyline of defender and protégé against the challenger. The United States does maintain close military ties to Taiwan and works to improve joint defense against a mainland attack. But near-term stability requires that Taiwan refrain from challenges to the status quo.70 The U.S. policy of “strategic ambiguity” avoids overly specific commitments to defend Taiwan so as to restrain Taipei, while accommodating China’s long-range objective of peaceful reunification.71 In past crises, the United States issued more pointed threats at China, as it did in the 1995–96 confrontations. But after 2001, the value to the United States of deep cooperation with China increased, while Taipei repeatedly lurched toward declaring independence. At this juncture, the political action within the extended-deterrence triad departed from the standard script. In moments of heightened tension triggered by Taiwan’s initiatives, China hawks within the U.S. administration, and even the president, issued pointed public statements declaring U.S. opposition to Taiwan independence.72 In other words, the defender’s “immediate” political signals in such contexts leaned heavily in the direction of restraint and accommodation, and this appears to have muted China’s impulse to challenge Taiwan with overt military pressure. At the level of high policy, the stability of extended deterrence rests on a common understanding between Washington and Beijing that the major short-term threats to peace are the political ambitions of Taiwan and that one condition necessary for China’s military moderation is American due diligence in restraining its protégé.73 This does not mean that the United States has abandoned defensive partnership with Taiwan; instead, defensive military measures come with a political posture that constrains Taiwan’s ambitions. Without the concomitant efforts to restrain Taipei, military measures to deter China would have to be greater, and even they would be likely to fail if Taiwan declared independence. In the Middle East, the conflict between Turkey and the Kurdistan Workers Party (PKK) sheltering in Iraqi Kurdistan raises the ongoing prospect of war among three U.S. protégés and clients: Turkey, the central government in Iraq, and the Kurdish Regional Government (KRG). In late 2007, as PKK operations escalated, killing scores of Turkish soldiers, it became clear how such a conflict imperiled major cross-cutting U.S. interests. Washington’s NATO and bilateral security commitments to Turkey and its declared sup-

the endurance of extended deterrence

297

port for robust military action against terrorist threats bumped up against concerns about undermining Iraq’s central government, alienating the Iraqi Kurds who strongly supported U.S. occupation and destabilizing the Kurdish province that had long been Iraq’s most stable.74 Negotiations between Ankara and Baghdad had little effect because the KRG was the real authority in the area, and it was averse to moving against the PKK. Meanwhile, close to 100,000 regular Turkish troops massed near the border with Iraq, and Ankara threatened to invade to destroy the PKK. The awkward corner called for a complex, multidirectional deterrence policy: it precluded the leveling of direct, public, military threats in any direction, least of all toward Turkey. Political factors would be paramount, and, as the situation escalated, the United States had to pose delicately balanced sanctions and incentives quietly. How these threats and promises were manipulated bore little resemblance to conventional models of crisis bargaining, and there was no room for audience-cost gamesmanship. How could the president sell the U.S. public on the need to restrain its long-term NATO ally from using force against a State Department–designated terrorist organization, operating from a safe haven in a failed state? How could the president explain to the U.S. public that that safe haven was in the most secure part of Iraq, which had been under a U.S. security umbrella since 1991? Adding to the conundrums were explosive domestic politics on both sides: as the Turkish parliament moved to give its military a green light to invade Iraq, the U.S. House of Representatives began debating a resolution to condemn the Armenian genocide of 1915. The structural complexity of the situation and a public sphere inimical to diplomacy meant that the defender’s signals would have to be contradictory rather than consistent and communicated in secret or muffled in the public domain. U.S. policy worked in multiple directions. It strove to reassure Turkey and to slow it down. It publicly pressed the Iraqi government and the KRG to act against the PKK. In early November 2007, Secretary of State Rice stopped in Ankara to give public gestures of support. In private she counseled against a major escalation and at the same time promised U.S. intelligence to support Turkish strikes against PKK targets. Subsequent weeks brought a flurry of Turkish cross-border strikes, presumably approved by the United States. Then came the large incursions of Turkish ground troops, which were not welcomed by Washington. They elicited from Rice a declaration that “no one should do anything that threatens to destabilize the north [of Iraq].”75 After the largest one, in later February 2008, President Bush stated that the

298

t i m o t h y w. c r aw f o r d

“Turks need to move quickly, achieve their objective, and get out.” Defense Secretary Gates echoed that “they should wrap this thing up as soon as they can.”76 A few days later, Turkey announced that it had pulled back all of its troops—but not due to any U.S. pressure. Iraq’s foreign minister, a Kurd, called U.S. intervention “instrumental” in the Turkish withdrawal.77 If the United States can be criticized for lacking decisiveness in its efforts to maintain stability between Turkey and Iraq, two things are worth considering. First, given the complexity of the problem, it is likely that the strongest U.S. moves to restrain the parties have occurred outside of the public eye. Second, taking a posture of partial support for, and limited ability to stop, Turkey’s military action not only protects the broader United States–Turkey relationship, it is also a useful way to weaken Kurdish nationalists in the KRG, who might otherwise be tempted by U.S. patronage to expand their ambitions. These two sets of issues are emblematic of the sorts of complex conflicts that are likely to invite U.S. extended deterrence in the future. A U.S. nuclear guarantee for Israel or other allies in the region could encourage those allies to take more military risks against Iran. Growing United States–India security ties ensure that a meltdown in Afghanistan or another regime crisis in Pakistan, could trigger an Indo-Pakistan confrontation, which would demand U.S. deterrence diplomacy of labyrinthine complexity. Russia’s military moves against Georgia in 2008—with a crisis prologue that featured a flamboyantly reckless protégé and ineffectual quiet efforts to restrain it, coupled with morally hazardous public signals and a challenger motivated partly by insecurity—delivered a sharp reminder of the cross-cutting political currents that may come into play with any United States extended-deterrence role versus Russia. Moscow’s September 2008 defense pacts with South Ossetia and Abkhazia multiplied, in turn, the dimensions of political complexity in the extended-deterrence situation.78 Beyond that, with increasing U.S. military support to Azerbaijan, Washington could plausibly become engaged in immediate deterrence attempts against Russia or Iran or in a conflict between Azerbaijan and Armenia.79 In such circumstances, conventional force deployments, the “shadow” of nuclear crisis, and other military factors will be essential background concerns, but the political dynamics of coercive diplomacy are likely to move in surprising directions because power, interests, and motivation are likely to stack up in ways that defy simple plotlines of a status quo defender, a greedy challenger, and a passive-defensive protégé. Extended deterrence remains useful to students of international relations because its various logics often approximate foreign policy and the strategic

the endurance of extended deterrence

299

situations confronting policy makers.80 As long as leaders take into consideration the strength and credibility of their adversaries’ alliances when deciding resort to force, and as long as states use security commitments to generate protective and restraining effects, extended-deterrence theory will remain relevant for describing the central dynamics of international politics. The extended-deterrence research program endures because it helps categorize the complex forces and relationships that shape the efforts of policy makers to extend deterrence to mitigate conflict.

notes 1. Keith Payne, “Moving beyond the Cold War’s Nuclear ‘Balance of Terror,’” Comparative Strategy 23 (April / May/ June 2004): 121; Phillip Bobbitt, Terror and Consent: The Wars for the Twenty-First Century (New York: Knopf, 2008). 2. Richard Betts, Nuclear Blackmail and Nuclear Balance (Washington, DC: Brookings Institution, 1987), 11; George Quester, The Future of Nuclear Deterrence (Lexington, MA: Lexington Books, 1986), 83. 3. Paul Huth, “Deterrence and International Conflict: Empirical Findings and Theoretical Debates,” Annual Review of Political Science 2 (1999): 25–48; James Morrow, “Alliances: Why Write Them Down?” Annual Review of Political Science 3 (2000): 63–84. 4. Thom Shanker and Norimitsu Onishi, “Japan Assures Rice That It Has No Nuclear Intentions,” New York Times, October 19, 2006, 14. 5. William C. Mann, “U.S. Reaffirms Promise to Defend S. Korea,” Associated Press, October 21, 2006. 6. See Robert S. Ross, “Comparative Deterrence: The Taiwan Strait and the Korean Peninsula,” in New Directions in the Study of China’s Foreign Policy, edited by Alistair Iain Johnston and Robert S. Ross (Stanford: Stanford University Press, 2006), chapter 2. 7. Josef Joffe, “Ally with the Sunnis,” New Republic, November 27 and December 4, 2006, 21. 8. See James Fearon, “Iraq’s Civil War,” Foreign Affairs 86 (March /April 2007): 12. 9. Glenn Kessler, “Analysts Divided on Clinton’s Arab Defense Plan,” Washington Post, May 4, 2008, A10. 10. See Kathleen J. McInnis, “Extended Deterrence: the U.S. Credibility Gap in the Middle East,” Washington Quarterly 28 (Summer 2005): 169–86. The calculus has precedents from the early cold war, when it was believed that effective security guarantees through NATO could prevent West Germany from building its own bomb. 11. Joseph F. Pilat, “Reassessing Security Assurances in a Unipolar World,” Washington Quarterly 28 (Spring 2005): 159–70. For a skeptical view, see Ted Galen Carpenter, “Closing the Nuclear Umbrella,” Foreign Affairs 73 (March /April 1994): 8–13. 12. Morgan, this volume. 13. See Barbara Walter, Committing to Peace: The Successful Settlement of Civil Wars (Princeton: Princeton University Press, 2002). 14. See Timothy W. Crawford and Alan J. Kuperman, eds., Gambling on Humanitarian Intervention: Moral Hazard, Rebellion, and Civil War (New York: Routledge, 2006). 15. One avenue of progress that will not be discussed here concerns the problem of “selection effects.” The basic theoretical insight is that challengers motivated enough to breach general extended deterrence will be very hard to deter in the immediate deterrence phase. See Jack S. Levy, “Quantitative Studies of Deterrence Success and Failure,” in Perspectives on Deterrence, edited by

300

t i m o t h y w. c r aw f o r d

Paul C. Stern et al. (New York: Oxford University Press, 1989), 117–20; Christopher Achen and Duncan Snidal, “Rational Deterrence Theory and Comparative Case Studies,” World Politics 41 ( January 1989): 161–63; James Fearon, “Selection Effects and Deterrence,” International Interactions 20 ( January–March 2002): 5–29. 16. Michael Clarke, “West European Politics and Extended Deterrence,” in Nuclear Deterrence: New Risks, New Opportunities, edited by Catharine McArdle Kelleher, Frank J. Kerr, and George Quester (Elmsford, NY: Pergamon-Brassey’s, 1986), 201. 17. Alexander George and Richard Smoke, “Deterrence and Foreign Policy,” World Politics 41 ( January 1989): 172; Abraham Kaplan, The Conduct of Inquiry: Methodology for Behavioral Science (New York: Chandler 1964), 298. 18. See Patrick M. Morgan, Deterrence: A Conceptual Analysis (Beverley Hills: Sage, 1977), 38–42; Paul Huth, Extended Deterrence and the Prevention of War (New Haven: Yale University Press, 1988). 19. Lawrence Freedman, “The First Two Generations of Nuclear Strategists,” in Makers of Modern Strategy: From Machiavelli to the Nuclear Age, edited by Peter Paret (Princeton: Princeton University Press, 1986), 735. Also see Freedman, Evolution of Nuclear Strategy, 3d ed. (New York: Palgrave-Macmillan, 2003), 404–5. 20. See Wallace Theis, Friendly Rivals: Bargaining and Burden-Shifting in NATO (Armonk, NY: M.E. Sharpe, 2003). 21. Henry Kissinger, “Missiles and the Western Alliance,” Foreign Affairs 36 (April 1958): 389. 22. Alexander L. George and Richard Smoke, Deterrence in American Foreign Policy: Theory and Practice (New York: Columbia University Press, 1974), 536–40. 23. Richard Betts, Surprise Attack: Lessons for Defense Planning (Washington, DC: Brookings Institution, 1982), 22. 24. Michael O’Hanlon, “Stopping a North Korean Invasion: Why Defending South Korea Is Easier Than the Pentagon Thinks,” International Security 22 (Spring 1998): 135–70; Michael O’Hanlon, “Why China Cannot Conquer Taiwan,” International Security 25 (Autumn 2000): 51–86. 25. Glenn H. Snyder, Deterrence and Defense: Toward a Theory of National Security (Princeton: Princeton University Press, 1961), 15; George Quester, “Some Thoughts on Deterrence Failures,” in Stern et al., Perspectives on Deterrence, 63. 26. Vesna Danilovic, When the Stakes Are High: Deterrence and Conflict among Major Powers (Ann Arbor: University of Michigan Press, 2002), 13–17; Huth, “Deterrence and International Conflict,” 30–31; Robert Jervis, “Deterrence Theory Revisited,” World Politics 31 ( January 1979): 289–324. 27. Thomas C. Schelling, Arms and Influence (New Haven, CT: Yale University Press, 1966), 43–55. 28. See James D. Fearon, “Domestic Political Audiences and the Escalation of International Disputes,” American Political Science Review 88 (September 1994): 577–92; Kenneth Schultz, Democracy and Coercive Diplomacy (Cambridge: Cambridge University Press, 2001). 29. Kenneth Schultz, “Looking for Audience Costs,” Journal of Conflict Resolution 45 (February 2001): 32–60. 30. See Shuhei Kurizaki, “Efficient Secrecy: Public versus Private Threats in Crisis Diplomacy,” American Political Science Review 101(August 2007): 543–58. 31. Robert Jervis, The Logic of Images in International Relations (Princeton: Princeton University Press, 1970), 75. 32. For key statements see Schelling, Arms and Influence, 55; Schelling, The Strategy of Conflict (New York: Oxford University Press, 1960), 29. 33. Glenn H. Snyder and Paul Diesing, Conflict among Nations (Princeton: Princeton University Press, 1977), 185–89; Huth, “Deterrence and International Conflict,” 32–34. Key reputation

the endurance of extended deterrence

301

skeptics include Patrick M. Morgan, “Saving Face for the Sake of Deterrence,” in Psychology and Deterrence, edited by Robert Jervis, Richard Ned Lebow, and Janice Gross Stein (Baltimore: Johns Hopkins University Press, 1985), 125–52; Ted Hopf, Peripheral Visions: Deterrence Theory and American Foreign Policy in the Third World (Ann Arbor: University of Michigan Press, 1995); Jonathan Mercer, Reputation in International Politics (Ithaca: Cornell University Press, 1996); Daryl Press, Calculating Credibility: How Leaders Assess Military Threats (Ithaca: Cornell University Press, 2005). For a recent expression of reputation optimism see Anne Sartori, Deterrence by Diplomacy (Princeton: Princeton University Press, 2005). 34. Mercer, Reputation in International Politics, 213–14. 35. Press, Calculating Credibility. 36. Morgan, “Saving Face,” 134. 37. See Mercer, Reputation and International Politics, 5; Danilovic, When the Stakes Are High, 16–17, 172–73. 38. See Press, Calculating Credibility, 153–59; Christopher J. Fettweis, “Credibility and the War on Terror,” Political Science Quarterly 122 (Winter 2007 / 08): 607–63; Fettweis, “On the Consequences of Failure in Iraq,” Survival 49 (December 2007): 83–98. 39. Todd S. Sechser calls this the “hegemon’s curse,” a major theme of his forthcoming book, Winning without a Fight. 40. Danilovic, When the Stakes Are High, 17–19, 22–23; Jervis, “Deterrence Theory Revisited,” 314–16; Stephen Maxwell, “Rationality in Deterrence” (Adelphi Paper no. 50, London: International Institute of Strategic Studies, 1968). 41. Schelling, Arms and Influence, 74–75, and 84. 42. Robert Jervis, Perception and Misperception in International Politics (Princeton: Princeton University Press, 1976), 101. 43. David A. Baldwin, “The Power of Positive Sanctions,” World Politics 24 (October 1971): 19–38; George and Smoke, Deterrence in American Foreign Policy, 604–10; Snyder and Diesing, Conflict among Nations, 254–56; Janice Gross Stein, “Deterrence and Reassurance,” in Behavior, Society and Nuclear War, vol. 2, edited by Philip E. Tetlock et al. (New York: Oxford University Press, 1991), 8–72; Huth, Extended Deterrence, 51–53. Also see James W. Davis, Threats and Promises: The Pursuit of International Influence (Baltimore: Johns Hopkins University Press, 2000). 44. Stein, “Deterrence and Reassurance,” 31; Michael Howard, “Reassurance and Deterrence,” Foreign Affairs 61 (Winter 1982 / 83). 45. See Snyder and Diesing, Conflict among Nations, 243. 46. Stein alludes to this as “reassurance through restraint”: “Deterrence and Reassurance,” 35–38. 47. See Richard Ned Lebow, Nuclear Crisis Management: A Dangerous Illusion (Ithaca: Cornell University Press, 1987), 107–8. 48. See Snyder, Deterrence and Defense, 29–30, 246–49; Schelling, Arms and Influence, 75, 84–85; Jervis, Logic of Images, 123–30; David A. Baldwin, “Thinking about Threats,” Journal of Conflict Resolution 15 (March 1971): 75–76; George and Smoke, Deterrence in American Foreign Policy, 529–30. 49. Schelling, Arms and Influence, 75. 50. Snyder and Diesing, Conflict among Nations, 217. 51. Timothy W. Crawford, Pivotal Deterrence (Ithaca: Cornell University Press, 2003); Frank Zagare and D. Marc Kilgour, “The Deterrence-Versus-Restraint Dilemma in Extended Deterrence: Explaining British Policy in 1914,” International Studies Review 8 (December 2006): 623–42; Brett V. Benson and Emerson M.S. Niou, “A Theory of Dual Deterrence: Credibility, Conditional Threats, and Strategic Ambiguity” (paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, April 12–15, 2007). 52. Curtis S. Signorino and Ahmer Tarar, “A Unified Theory and Test of Extended Immediate

302

t i m o t h y w. c r aw f o r d

Deterrence,” American Journal of Political Science 50 ( July 2006): 586–605; Fearon, “Selection Effects and Deterrence”; Frank Zagare and D. Marc Kilgour, Perfect Deterrence (Cambridge: Cambridge University Press, 2000), chapter 6. 53. See Huth, Extended Deterrence, chapter 3. 54. Crawford, Pivotal Deterrence, 9–10, 16; and Stephen L. Quackenbush, “Not Only Whether but Whom: Three-Party Extended Deterrence,” Journal of Conflict Resolution 50 (August 2006): 563. 55. Raymond Aron, Peace and War: A Theory of International Relations (New Brunswick, NJ: Transaction, 2003), 83. 56. See Michael I. Handel, Weak States in the International System (London: Frank Cass, 1990), 46. 57. Ronald Krebs, “Perverse Institutionalism: NATO and the Greco-Turkish conflict,” International Organization 53 (Spring 1999): 343–77; Crawford, Pivotal Deterrence, 100–34, 173–77. 58. Schelling, Arms and Influence, chapter 2. 59. Jeremy Pressman, Warring Friends: Alliance Restraint in International Politics (Ithaca: Cornell University Press, 2008); Patricia Weitsman, Dangerous Alliances (Stanford: Stanford University Press, 2004); Glenn H. Snyder, Alliance Politics (Ithaca: Cornell University Press, 1997); Paul Schroeder, “Alliances, 1815–1945: Weapons of Power and Tools of Management,” in Systems, Stability, and Statecraft: Essays on the International History of Modern Europe (New York: Palgrave, 2006); George Liska, Nations in Alliance: the Limits of Interdependence (Baltimore: Johns Hopkins University Press, 1968), 139; Hans J. Morgenthau, “Alliances in Theory and Practice,” in Alliance Policy in the Cold War, edited by Arnold Wolfers (Baltimore: Johns Hopkins University Press, 1959), 190. 60. David Hume, Political Essays, edited by Knud Haakonssen (Cambridge: Cambridge University Press, 1994), 159. 61. George and Smoke, Deterrence in American Foreign Policy, 370. 62. Snyder and Diesing, Conflict among Nations, 436. 63. Glenn H. Snyder, “The Security Dilemma in Alliance Politics,” World Politics 36 ( July 1984): 461–95. 64. James Fearon, “Signaling Foreign Policy Interests,” Journal of Conflict Resolution 41 (February 1997): 84. 65. Gregory D. Miller, “Hypotheses on Reputation: Alliance Choices and the Shadow of the Past,” Security Studies 12 (Spring 2003): 66–67. 66. Danilovic, When the Stakes Are High, 366. 67. See Benson and Niou, “Theory of Dual Deterrence”; T. V. Paul, Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge: Cambridge University Press, 1994), 31–33. 68. Huth, Extended Deterrence, 52–53; George and Smoke, Deterrence in American Foreign Policy, 604–10. 69. Crawford, Pivotal Deterrence, 177–87. 70. Robert S. Ross, “Navigating the Taiwan Strait: Deterrence, Escalation Dominance, and U.S.-China Relations,” International Security 27 (Fall 2002): 48–85; Thomas J. Christensen, “Posing Problems without Catching Up: China’s Rise and Challenges to U.S. Security Policy,” International Security 25 (Spring 2001): 5–40. 71. Richard Bush, “The US Policy of Dual Deterrence,” in If China Attacks Taiwan: Military Strategy, Politics and Economics, edited by Steve Tsang (London: Routledge, 2006), chapter 3; Nancy Bernkopf Tucker, “Strategic Ambiguity or Strategic Clarity,” in Dangerous Strait: The USTaiwan-China Crisis, edited by Nancy Bernkopf Tucker (New York: Columbia University Press, 2005), 186–212; Steve Chan, China, the U.S., and the Power Transition Theory: A Critique (New York: Routledge, 2008), chap. 9. 72. Crawford, Pivotal Deterrence, 197–99; President Bush and Premier Wen Jiabao, “Remarks

the endurance of extended deterrence

303

to the Press, The Oval Office,” Washington, DC, December 9, 2003, http://www.whitehouse.gov / news / releases / 2003 / 12 / 20031209–2.html. 73. Jing-dong Yuan, “China, U.S. Delicately Juggle Taiwan,” Asia Times Online, September 25, 2007, http://www.atimes.com / atimes /China / II25Ad01.html; Thomas J. Christensen, U.S. Deputy Assistant Secretary for East Asian and Pacific Affairs, “A Strong and Moderate Taiwan,” Speech to U.S.-Taiwan Business Council, Annapolis, MD, September 11, 2007. 74. The KRG president, Massoud Barzani, has publicly asked the United States to deploy a continuing force in the province, as a “deterrent to intervention by the neighboring countries.” Cited in Christopher de Bellaigue, “The Uncontainable Kurds,” New York Review of Books, March 1, 2007, 36. 75. David Stout, “Bush Sharply Criticizes Iran and Praises Turkey,” New York Times, January 8, 2008. 76. Lolita C. Baldor, “Gates to Turkey: Halt Incursion in Iraq,” Washington Post, February 28, 2008. 77. Sabrina Tavernise and Richard A. Oppel, “Turkey Withdraws Troops from Northern Iraq,” New York Times, March 1, 2008. 78. See Oksana Antonenko, “A War with No Winners,” Survival 50, no. 5 (2008): 23–36; Helene Cooper, C. J. Chivers, and Clifford J. Levy, “U.S. Watched as a Squabble Turned into a Showdown,“ New York Times, August 18, 2008. 79. Jim Nichol, “Armenia, Azerbaijan, and Georgia: Security Issues and Implications for U.S. Interests,” CRS Report for Congress (Washington, DC: Congressional Research Service, February 2007), 38–41. 80. Frank C. Zagare, “Deterrence Is Dead: Long Live Deterrence,” Conflict Management and Peace Science 23 (2006): 115–20.

chapter thirteen R

The Revolution in Military Affairs Impact of Emerging Technologies on Deterrence michel fortmann and stéfanie von hlatky

deterrenc e had a profound effect on strategic thinking during the cold war. This concept was central to the conduct of American defense policy and military planning for over forty years. The revolution in military affairs (RMA), which today is more often referred to as “transformation,” is often presented as the new keystone concept of U.S. grand strategy. In the context of a more complex and uncertain world, where smaller threats are proliferating, the RMA makes an appealing promise. New hardware and doctrine will produce an increase in power-projection capabilities, while decreasing friendly and even enemy casualties. This new U.S. military instrument is supposed to offer a multipurpose and flexible way to address several threats simultaneously, in the spirit of cost efficiency. Transformation represents a major shift from cold war strategic thinking. With the demise of America’s Soviet archenemy, the United States left the old precepts of nuclear deterrence behind, which led to a restructuring of its military forces, smaller defense budgets, and different strategic interests. But might deterrence retain its relevance even in the face of these strategic shifts? Is deterrence still relevant in this new security environment? Is the technology associated with the RMA an effective instrument for deterring hostile actors today? To answer these questions, the chapter first explains what is meant by an RMA and the various scholarly debates sparked by the concept. It then describes the impact of the RMA on conventional and nuclear deterrence, especially if the RMA continues to blur the distinction between conventional and nuclear warfare or if RMA technologies are incorporated fully 304

the revolution in military affairs

305

into nuclear arsenals. It then concludes by offering some observations about the RMA and why the application of information-age technology to warfare can be a source of stability and instability in world politics.

the disputed rma concept The RMA presupposes drastic changes in military technology. But exactly what does the RMA mean? Today’s RMA is based on four technological developments.1 First, improvements in computers and electronics permitted major advances in all-weather precision-guided weapons and unmanned reconnaissance systems, which translate into more efficiency in terms of the size of the force needed to achieve a given objective. The RMA also holds out a great potential for robotic systems, but, so far, the deployment and use of autonomous weapons (e.g., cruise missiles and battlefield androids) have been relatively limited. Second, air and space sensors have worked to lift the so-called fog of war by making several types of conventional battlefields (ocean surfaces, deserts, airspace) transparent. Third, weapons platforms continue to become more fuel efficient, faster, and stealthier, making networked combat forces more rapidly deployable and lethal. In other words, the RMA has produced a leaner and meaner military machine. If properly exploited and integrated into military organizations and doctrine, these technical trends hold the promise of bringing an RMA that may constitute the greatest advance in warfare since the dawn of the nuclear age in 1945.2 The RMA concept and its attendant technologies also have proven to be politically attractive because they fit extremely well within the framework of what Lawrence Freedman and Martin Shaw have called the “new Western way of war.” This approach favors wars that are fought in distant places, involve a minimum of casualties and collateral damage, emphasize airpower or long-range precision weapons, and, most importantly, deploy a minimum of troops on the ground.3 Because this type of warfare holds out the prospect of generating minimal friendly casualties, the RMA actually makes the use of force a more attractive option to officials who find it increasingly difficult to justify the use of force to their public. Like any normative concept, the notion of a military revolution should be perceived as more or less useful rather than true or false.4 As such, the concept of an RMA was certainly useful for a military institution that had lost its main enemy and needed an intellectual blueprint to design a leaner, more adaptive armed force for the post–cold war world. In the ensuing decade, the notion of a military revolution certainly served that purpose well.

306

michel fortmann and stéfanie von hlatky

The other side of the coin, of course, was that the RMA carried with it the dream of a U.S. conventional military supremacy that could be maintained over the long term. Origins of the RMA Concept The changes associated with the current RMA stem from developments that emerged in the 1970s and 1980s. The notion of a “military-technical revolution” was introduced by the Soviets to describe the possible consequence of integrating emerging information and microprocessing technologies into existing military systems. U.S. strategists were interested in acquiring a technological edge to counter the Russians’ superior numbers by utilizing emerging technologies to destroy Soviet second and third echelon units in what was called the “Follow on Forces Attack.”5 The RMA thus came out of the perceived necessity to battle massed armored formations on the inter-German border and to “deep strike” Soviet reinforcements as they moved toward forward operating areas. In the minds of Soviet strategists, U.S. technological prowess would inevitably exploit these new opportunities, raising the possibility that Soviet conventional superiority in Europe might eventually disappear. Beyond that, several strategists in the United States foresaw that the same technologies held out the promise of “strategic use of non-strategic weapons,” meaning that the United States “could accomplish with conventional means and at tolerable costs the same objectives assigned since 1945 to (essentially unusable) U.S. nuclear forces.”6 In a nutshell, deterrence was about to mutate into war fighting.7 Ten years after North Atlantic Treaty Organization (NATO) strategists began to exploit emerging technologies, Operation Desert Storm demonstrated to those interested in military affairs that the world was confronted with a technological breakthrough in real military operations. Despite the success of this international effort to oust Iraq from Kuwait, some experts remained hesitant to recognize the existence the RMA, or the demise of “industrial” warfare. Arguments persisted over what actually constituted an RMA and whether or not the coalition victory over Iraq provided evidence of an RMA or just a more incremental evolution in warfare. Generically, a military revolution is described as a period of discontinuous change that renders obsolete existing means for conducting war. An RMA is produced by simultaneous and spontaneous changes in military relevant technologies, tactics, concepts of operation, and methods of organization.8 According to Mike Vickers and Robert Martinage, what is currently being termed the

the revolution in military affairs

307

information-age RMA is defined by key developments in military affairs: “the ability to strike with great accuracy independent of range; the ability, through the use of stealth, to penetrate defenses with impunity; the emergence of unmanned warfare; the tactical and operational exploitation of space; and the ability to move information rapidly and widely across a joint battle network and exploit the effects of increased joint force integration.”9 A list of technological improvements, however, does not prove the existence of an RMA. In fact, the current term used by the U.S. Department of Defense to refer to the RMA—transformation—suggests that changes with regard to military technology, organization, and doctrine do not just happen. Instead, the term denotes a lengthy and arduous process whereby officials and senior officers attempt to foster changes within the military in a step-by-step fashion. The RMA and Its Critics Scholars usually point to several reasons to be skeptical about the current RMA.10 They criticize the use of the “revolution” label, arguing that improvements in information technologies are a permanent staple of military history and that current technologies have simply highlighted the ability to conduct combined-arms operations as the basis of success on the battlefield. Additionally, military developments spanning over three decades can hardly be qualified as a revolution, which is a sudden and intense event.11 Others point to the limited scope of current changes in technology, tactics, and doctrine—hence the need for a deliberate policy and process of transformation. These critics are skeptical that the all-encompassing changes expected in the wake of an RMA are occurring.12 The RMA is a normative concept that inspired U.S. conventional military reforms during the 1990s. Trendy labels, such as star wars, battlespace dominance, the system of systems, network-centric warfare, and transformation, were all influenced by the idea of an RMA, which has taken on positive connotations among Americans who prize technological innovation and novelty. Many of the documents produced by the U.S. military establishment during the 1990s and early 2000s still bear the RMA imprint. even if they do not explicitly mention the concept.13 The George W. Bush administration, in a bit of political marketing, favored the term transformation instead of RMA. Nevertheless, it embraced the technological and organizational advances made possible by RMA technologies. What was key for the administration was the potential for cost

308

michel fortmann and stéfanie von hlatky

savings created by the shift from “legacy” weapons, technologies, and organizations to smaller, more capable military units and flatter command structures. The RMA held out the promise of lighter and more densely networked fighting units, which would foster closer cooperation among land, air, sea, and space forces, along with a more responsive “just in time” logistical support system. Resistance to this vision actually came from the military itself, because new technologies would have to be paid for by reducing the size of standing forces. Because RMA technologies will likely involve drastically lighter, more fuel-efficient, faster, and stealthier weapon platforms on land, on sea, and in the air, there also is a reduced role for warriors in some of these visions for future warfare. Although urban combat has changed little from World War II, on some battlefields, new technologies, not humans, are dominant. Global precision strike involves autonomous delivery systems flying intercontinental distances following computer-constructed maps to their targets. The RMA is an American phenomenon. Not only does the United States enjoy the leading edge in the latest information warfare technologies, but it is developing, acquiring, and employing these capabilities faster than any other state. By contrast, U.S. allies have been unwilling or unable to make the same kinds of investments in RMA-related technologies. Coalition and alliance operations are thus complicated by the fact that only a few militaries could actually contemplate undertaking a combined operation with their U.S. counterparts. Allied militaries are assigned more or less quiet sectors to undertake small-scale combat operations, peacekeeping activities, or military support to civil authorities.

the rma and conventional deterrence The general aim of deterrence is to achieve a reversal in menacing behavior.14 The target of deterrence is thus presented with a threat and an opportunity to comply. If compliance occurs, deterrence is successful. If compliance does not occur, deterrence fails. Since deterrence was the common currency of cold war crises, it became deeply intertwined with the general superpower confrontation and the prospect of mutually assured destruction (MAD) as the likely outcome of a Soviet-American nuclear war. Now that the cold war is over, the nuclear deterrence paradigm is not only faced with a variety of new confrontational relationships, but leaders now have a wider set of military technologies to choose from when making deterrent threats. The all-or-nothing game of nuclear deterrence has been dramatically altered,

the revolution in military affairs

309

and the logic of MAD no longer dominates most deterrence relationships. Furthermore, with no peer competitor to challenge the status quo, the United States is left to police the international system according to its own strategic imperatives. As a general rule, the RMA literature does not take into account nuclear weapons or nuclear deterrence. There are several reasons why contemporary theorists fail to consider the impact of the RMA on the theory and practice of deterrence. First, the RMA is about conventional deterrence, which is based on the capability to deny aggressors their battlefield objectives with conventional forces.15 Conventional deterrence is a coercive strategy that rests on an assessment of the capabilities possessed by both parties engaged in a crisis.16 According to Richard J. Harknett, unlike nuclear deterrence, conventional deterrence is “contestable” in the sense that the target of the deterrent threat can entertain the possibility of avoiding threat by defeating in battle the party attempting deterrence. By contrast, nuclear weapons are so destructive that there is no way to prevent unimaginable death and destruction following the execution of a nuclear deterrent threat.17 This notion of contestability in conventional deterrence also sheds light on asymmetric confrontations, during which the weaker state may call the defender’s bluff in hopes that there is insufficient will for retaliation. Under these circumstances, the diplomatic context of a confrontation, the political will behind opposing positions, and actors’ beliefs and perceptions become paramount in predicting whether or not efforts at deterrence will succeed. Because the RMA will never measure up to nuclear weapons in terms of threatening unacceptable costs, it must be used in a way to convey the incontestability of deterrent threats to achieve credible deterrence.18 The target to be deterred cannot believe it possesses the capability to circumvent or mitigate the defender’s conventional deterrent threats. Second, the RMA alters the deterrence equation by reducing the costs associated with intervention should the threat fail to produce compliant behavior. Most RMA theorists thus see these new technologies as primarily providing a war-fighting instrument, best used to preempt an opponent’s attack or as part of a preventive war strategy to stop the acquisition or use of weapons of mass destruction (WMD). However, theorists are missing an important relationship between the RMA and deterrence. A reputation for resolve enhances credibility, and resolve can be bolstered by the fact that certain RMA strategies hold the promise of reducing casualties, especially when objectives are kept limited. The RMA can thus be seen to bolster the credibility of deterrent threats because the reluctance associated with

310

michel fortmann and stéfanie von hlatky

carrying out deterrent threats evaporates when the potential costs to the party making good on the threat are minimal. Moreover, because retaliatory strategies that rely on RMA technologies are far less controversial than nuclear retaliation or the deployment of ground forces, RMA-based threats would probably appear more credible to policy makers. There are fewer political restraints on policy makers when they contemplate the use of forces that rely on the RMA. Third, the RMA is often disconnected from deterrence because there is no existential threat against the United States. The issue of deterrence itself has faded into the background compared with the cold war preoccupation with the possibility of a nuclear exchange between the United States and the Soviet Union. As a result, the American deterrence posture is geared toward extended deterrence or enforcing the status quo on a global scale.19 Strategists rightly tend to see the extended-deterrence issue as a matter of war fighting; the RMA helps to preserve the dramatic gap in military capabilities between the United States and its potential regional challengers. But given the enormous disparity in the military capability possessed by the United States and its likely adversaries, political factors become extremely important in the opponents’ estimates of whether or not deterrence threats will be executed. Given their relative lack of military capability, challengers will have to be especially motivated before risking a confrontation with the United States. In this sort of deterrence relationship, the RMA offers an important advantage to U.S. policy makers: by reducing the potential costs of war to the United States, the credibility of deterrent threats is likely increased in the minds of the beholder. In sum, the RMA may have its biggest impact on extended deterrence. It is more difficult to assess both the risks the protector will run and the costs it will bear in protecting its ally. The aggressor can play on this uncertainty. But if the aggressor believes that the threat will be enforced, then it might pursue a more cautious strategy.20

the risks created by the rma RMA technologies dramatically improve power projection by making flexible strikes against specific targets a reality. The RMA allows the United States to devise diversified tactics against rogue states and nonstate actors. The capability to launch surgical strikes to achieve limited objectives also increases the credibility of deterrent threats against rogue states and great-power competitors. Although Patrick Morgan remains skeptical about this “tailored ap-

the revolution in military affairs

311

proach” to war, he nonetheless concedes that the RMA makes power projection easier and thus enhances the credibility of extended deterrence.21 The capability to launch precision strikes in support of extended deterrence, however, can potentially lead to destabilizing consequences. The promise of lowering casualties and collateral damage could prompt states to be more aggressive. If the potential costs of war appear to be extraordinarily low, why not simply remove potential irritants before they become major problems? Because of the apparently low costs of military action, compellence might be increasingly considered as a viable option, which could lead to reckless behavior in crisis situations. Indeed, the ideas embodied in the RMA appealed equally to the military enthusiasts who dreamed of “full spectrum dominance” or “complete ascendancy in every form of warfare” and neoconservative activists clamoring for a new Pax Americana built on U.S. military supremacy.22 The George W. Bush administration’s willingness to intervene in Afghanistan and Iraq thus reflected, partly at least, the belief that the RMA was the key to U.S. military dominance in the post– September 11 world. The RMA may make the exercise of extended deterrence seem easier for a leading power like the United States, which might serve the cause of peace by making violent challenges to the status quo less likely. Alternatively, the RMA might elicit more aggressive responses on the part of targets because they may perceive deterrent or compellent threats as a prelude to preventive war. RMA technologies seem to breed overconfidence. The same way leading powers are said to lose preeminence through overextension, the technological edge implied by the RMA could lead to the adoption of overly ambitious objectives. RMA technologies are best suited for limited strategies.23 For example, there is a clear contrast between the way American-led operations in Iraq unfolded in 1991 and in 2003. The current lack of security in Iraq is not due to the RMA’s failure but to the fact that U.S. strategists asked more of their precision-guided and networked arsenal than it could deliver. The power vacuum that was created by the overthrow of Saddam Hussein led to a situation that is not readily addressed by RMA technologies, namely, guerrilla warfare and urban combat. The situation in Iraq thus suggests that in order to contribute to the full range of military operations likely to be encountered in the future, RMA forces and technologies must be combined with better counterinsurgency strategies and tactics and a more serious attitude toward stability, security, transition, and reconstruction operations. Leveraging high technology to make it more relevant to low-intensity conflicts might, however, be only part of the solution. The real issue is the

312

michel fortmann and stéfanie von hlatky

well-known American aversion to open-ended involvement in limited wars. In the words of Jeffrey Record, “The U.S. military aversion to counterinsurgency is a function of 60 years of preoccupation with high-technology conventional warfare against other states and accelerated substitution of machines for combat manpower.”24 Can information technologies and precision-guided weapons replace boots on the ground? The answer is probably not. But from a more strategic point of view, RMA-associated technologies could enhance the credibility of threats by shifting the focus of war directly to leaders, making them more responsive to threats. By targeting them directly, or the forces, infrastructure, or resources needed to keep their regimes in power, RMA-based threats might be more effective. To more fully exploit the potential for generating effective deterrent threats, however, RMA strategies also should be coupled with reasonable incentives for leaders to change their behavior. The RMA should not be viewed as a technical fix to the problems of deterrence but rather as an opportunity to diversify the types of strategies conceived under its banner. As a recent RAND report noted, “the essential nature of devising effective coercive strategies has not changed. One must understand not only an adversary’s capabilities and the threshold that must be reached to coerce him, but also effectively communicate to him that you have both the will and the capability to prevail.”25

the rma and nuclear deterrence How do conventional strategies inspired by the RMA impact nuclear deterrence? On the one hand, the RMA might enhance crisis stability. Improved reconnaissance capabilities can increase transparency and decrease fears of surprise attack, especially if they are embedded in an arms-control regime that fosters strategic communication among potential adversaries. Better early-warning capabilities and better strategic intelligence should promote crisis stability. Crises sometimes escalate because the emerging threats are detected at the eleventh hour, which leaves a narrow window for diplomacy or deterrence. In contrast, early detection of emerging threats can promote more measured responses; deterrent or compellent threats could be made before a fait accompli is presented to the international community or events spiral out of control. The RMA also might reduce the attractiveness of maintaining a large nuclear arsenal, especially among great powers. States might choose to forego nuclear weapons because the RMA could be perceived as providing an adequate military capability for most contingencies.26 The RMA also may

the revolution in military affairs

313

redirect regional powers’ defense efforts toward conventional rather than nuclear weapons. Paul Nitze, who shaped U.S. defense policy during the cold war, stated in 1999 that conventional weapons could replace nuclear weapons. He even argued for the elimination of the U.S. nuclear arsenal on the grounds that the RMA made it obsolete.27 Andrew Krepinevich and Steven Kosiak also noted that a shift to smart long-range conventional weapons will offer U.S. leaders greater flexibility in responding to military threats, thereby enhancing deterrence.28 Even the Pentagon’s Quadrennial Defense Review in 1997 suggested that “[a]dvanced military technologies that merge the capabilities of information systems with precision-guided weaponry and real time targeting and other new weapons systems may provide a supplement or even an alternative to the nuclear arsenal of the Cold War.”29 Relying on RMA technologies instead of nuclear weapons as the basis for deterrent threats represents a significant change in the means of deterrence. In terms of nuclear weapons, the threat relied on the overwhelming and indiscriminate destruction promised by the use of more than a few nuclear weapons. Because of the scope of this destruction, to say nothing about the prospect of retaliation in kind, the credibility of the threat to use nuclear weapons remained relatively low. In contrast, the destruction produced by RMA technologies is extraordinarily discrete and limited, while the credibility of RMA-based deterrent threats remains very high. These observations raise an important question: can transformation be made compatible with the indiscriminate destruction implied by nuclear weapons? The George W. Bush administration actually proposed to rethink the place of nuclear weapons in the U.S. arsenal. Its 2002 Nuclear Posture Review recognized the importance of nuclear forces as a weapon of last resort, but it also sought to integrate them within a broader strategy that included missile defense and conventional precision-guided munitions. The term “strategic deterrent” was substituted for “nuclear deterrent” in the administration’s political lexicon, as the strategists came to think of this new force as capable of undertaking preemptive strikes and preventive war, especially against states threatening to use WMD. Although the Nuclear Posture Review reduced the role played by nuclear weapons in overall U.S. defense policy, it did propose developing a hybrid arsenal that combined missile defenses, RMA-engineered precision strike capabilities, information-driven technologies, and low-yield “boutique” nuclear weapons to neutralize primitive nuclear capabilities. Nuclear weapons offer unique advantages in destroying hard-to-reach targets, such as deeply

314

michel fortmann and stéfanie von hlatky

buried underground facilities.30 The long-term danger of integrating nuclear capabilities with RMA technologies in a unified conventional force is that it imperils the tradition of nonuse.31 But the combination of small nuclear weapons and precision-guided technologies also can bolster the credibility of deterrent threats because policy makers might be more willing to authorize nuclear strikes if their effects can be limited to specific military targets and missions, for example, counterproliferation strikes against rogue states’ nascent nuclear arsenals. The paradox is that by reducing the indiscriminate destructiveness of a deterrent force, one can increase the overall credibility of deterrent threats and the battlefield performance of an arsenal. The RMA could create an instrumental nuclear force; it might be possible to use nuclear weapons to achieve limited aims. Observers ignored the improved deterrent effects created by the new “strategic triad” that was proposed by the 2002 Nuclear Posture Review and were quick to criticize the Bush administration for an apparent attempt to conventionalize nuclear capabilities. By reducing the prospect of indiscriminate nuclear destruction, they believed that the administration was eroding the nuclear taboo. Critics believed that a doctrine that combined nuclear and conventional capabilities would only blur the distinction between the two types of weapons, while increasing the likelihood of nuclear escalation. Critics also believed that progress in RMA-related technologies, in the realms of stealth, space reconnaissance, flexible targeting, and responsive command and control, generally serve to bolster both conventional and nuclear offensive force postures and doctrines. Observers worried that ballistic missile defense systems, for example, could easily be married to nuclear and conventional precision-strike systems, shifting entire defense postures from deterrence to a clear-cut offensive orientation. Although coupling the RMA to nuclear weapons could have enhanced the credibility of U.S. nuclear deterrence and war-fighting capabilities, the U.S. Congress took an extraordinarily dim view of the effort to develop an arsenal that directly linked these two technologies. Congress refused to fund even a study of the feasibility of this type of program and instead directed the Bush administration to increase its ongoing efforts to reduce the size of the U.S. nuclear-weapons stockpile by accelerating the destruction of surplus nuclear weapons. Congress also has exhibited little enthusiasm for development of a reliable replacement warhead, which is designed to maximize safety and shelf life and not yield-to-weight ratios or exotic nuclear-weapons effects. Nevertheless, it remains theoretically possible to derive some deterrent and war-fighting benefits from marrying nuclear and RMA technol-

the revolution in military affairs

315

ogies, although no state or nonstate actor has chosen to explore fully this technological possibility.

the systemic impact of the rma The RMA could hold several implications for great-power relations. First, the RMA could increase state competition by fostering an arms race or by creating threatening strategic imbalances as technologies spread across the world. Some argue that the United States should curtail the sale of RMArelated technologies in order to stall these types of developments.32 It is difficult to imagine an RMA arms-control regime, however, because many of these technologies are dual use in nature. In fact, the RMA is largely dependent on commercially developed technologies that are simply adapted for military use. Technology diffusion will continue, and it is likely that this technology will exacerbate existing enduring rivalries or aggravate simmering disputes. Second, conventional superiority vis-à-vis a weaker nuclear power can lower the nuclear threshold. The Russians, for example, have realized since the 1980s that they cannot compete with the West’s technological and industrial prowess, especially in the realm of RMA-related technologies. It is thus not surprising that they have announced that they reserve the right to use nuclear weapons to offset the increasing inferiority of their crumbling conventional capabilities. In 2000, Russia announced a new military doctrine, including among its provisions that “The Russian Federation reserves the right to use nuclear weapons in response to large-scale aggression involving conventional weapons in situations that are critical for the national security of the Russian Federation and its allies.”33 The continuing modernization of Russia’s nuclear arsenal also demonstrates that Moscow is attempting to preserve its nuclear options in the face of growing conventional inferiority.34 To be sure, the RMA is only one factor among many in explaining the evolution of Russia’s recent strategic choices. Nevertheless, it has played a part in the decision to reemphasize the role of nuclear weapons in Russian defense doctrine. Third, the RMA could ignite asymmetric competition, exacerbating regional rivalries or leading to a general deterioration in world politics. China’s recent test of an antisatellite missile, for instance, can be interpreted as an asymmetric response to U.S. RMA-related conventional superiority. Beijing is indeed keenly aware of the role that communications, surveillance, navigation, and data-relay satellites play in Washington’s ability to

316

michel fortmann and stéfanie von hlatky

extend and sustain conventional operations globally but most particularly in Asia. Since the 1990s, Chinese military planners have recognized that the United States relies on space systems for its military operations and its participation in the global economy. Attacks against U.S. satellites could thus neutralize one of the key components of America’s worldwide military and economic superiority. Such a capability could help deter, impede, or even prevent U.S. military moves in Asia, reducing the likelihood that the United States would rush to intervene in a military confrontation between China and Taiwan.35 Moreover, China’s antisatellite capability creates problems for Japan and Taiwan because it could negate the protection the United States has provided them for half a century, thereby undermining American extended deterrence in the Asian theater. It could also increase the pressure on Tokyo and Taipei to acquire missile defenses or an eventual nuclear deterrent.

conclusion The RMA is gaining ground in an era in which the United States is faced with several types of security challenges. This chapter has explored the theoretical and practical implications of the RMA for both nuclear and conventional deterrence. The major argument is that the RMA is now altering one of the inherent limits when it comes to putting deterrence theory into practice by minimizing the costly and unpopular facts of war: casualties, collateral damage, and the deployment of large numbers of troops in foreign lands. The RMA can make extended deterrence and global security management easier for the United States by supplying efficient and flexible power-projection capabilities. Although the challenger’s willingness to absorb costs and to engage in asymmetric strategies to deny the United States free use of its RMAequipped military has to be taken into account by strategists, the RMA can bolster deterrence especially when threats can be directed against targets of high value to a given regime. The chapter has made several other points as well. The RMA also is blurring the distinction between conventional and nuclear weapons, although the George W. Bush administration has failed to integrate RMA technologies fully into the U.S. nuclear force structure to create a new strategic deterrent. Second, U.S. conventional superiority can spur WMD proliferation as other states attempt to narrow the gap in military capability with the United States and its allies. Third, conventional superiority vis-à-vis a weaker nuclear power also can lower the nuclear threshold because leaders under attack by

the revolution in military affairs

317

precision-guided forces can panic and choose to use rather than lose their limited nuclear capabilities. Fourth, RMA-ignited asymmetric competition can provoke regional instability, especially in Asia. Fifth, the RMA makes the prospect of a decapitating first strike against a weak nuclear power more than just a theoretical possibility, although a state contemplating such a strike could be deterred by the remote possibility of nuclear retaliation. Will the RMA strengthen deterrence? For the RMA to bolster deterrent threats, potential targeted leaders have to believe that new weapons and technologies will work as advertised. RMA successes have been well publicized. During Kosovo and Operation Desert Storm, the RMA technologies delivered stunning victories, which seemed to usher in an era of low casualties and high battlefield efficiency. So far, the RMA has not suffered credibility-undermining failures, but it has not yet established a strong record of deterrence or compellence success.36 Saddam Hussein presented the world with a fait accompli in Kuwait and then ignored the threat of an overwhelming attack by a global coalition. Serbia seemed to ignore an impending precision air campaign. The current war in Iraq also demonstrates that the RMA needs to be linked to security, stability, transition, and reconstruction capabilities if militaries are to avoid winning wars and end up losing the peace.37 But unlike nuclear weapons, which have a long-established tradition of nonuse, RMA technologies are establishing a tradition not only of use but of usefulness. New conventional weapons and operations can constitute credible deterrents and deliver tailored punishment in the event of deterrence failure. The RMA, however, is a work in progress. Even though technological advances have been made in many areas, they have not been consolidated to create a truly revolutionary departure in warfare. The RMA has yet to demonstrate its effectiveness across the entire spectrum of conflict, which is crucial for the practice of deterrence. The war in Afghanistan and the war in Iraq have not produced decisive victories, which suggests that asymmetric responses to the RMA are in fact available, especially if opponents are willing to pay a high price in terms of death and destruction. Americans have demonstrated high resolve in the war on terrorism, but there is a perception that in the end the United States lacks the will to sustain bloody engagements. Nevertheless, the RMA might become the basis of a new age of deterrence, because precision-guided warfare can be less destructive than nuclear war. RMA deterrent threats are credible because if deterrence fails, policy makers, despite the experience in Afghanistan and Iraq, are likely to make good on their threats.

318

michel fortmann and stéfanie von hlatky

notes 1. Michael E. O’Hanlon, Technological Change and the Future of Warfare (Washington, DC: Brookings Institution, 2000), 2. See also G. Michael Vickers and Robert Martinage, The Revolution in War (Washington, DC: Center for Strategic and Budgetary Assessment, 2004), 7–68. 2. O’Hanlon, Technological Change, 3. 3. Lawrence Freedman, “The Transformation of Strategic Affairs” (Adelphi Paper 379, International Institute for Strategic Studies, London, March 2006), 50; see also Martin Shaw, The New Western Way of War: Risk-Transfer War and Its Crisis in Iraq (Cambridge: Polity Press, 2005). 4. Colin S. Gray, “Nuclear Weapons and the Revolution in Military Affairs,” in The Absolute Weapon Revisited: Nuclear Arms and the Emerging International Order, edited by T. V. Paul, Richard J. Harknett, and James J. Wirtz (Ann Arbor: University of Michigan Press 1998), 17. 5. Ariel E. Levite and Elizabeth Sherwood-Randall, “The Case for Discriminate Force,” Survival 44 (Winter 2002 / 03): 84. 6. Andrew J. Bacevich, The New American Militarism (Oxford: Oxford University Press, 2005), 162. 7. Ibid., 163. 8. Richard Huntley, Past Revolutions, Future Transformations: What Can the History of Revolutions in Military Affairs Tell Us about Transforming the U.S. Military? (Santa Monica: RAND, 1999), 8–11. 9. Vickers and Martinage, Revolution in War, 4. 10. Michael O’Hanlon, “Beware the RMA’nia!” (paper presented at the National Defense University, Washington, DC, September 9, 1998). 11. Colin S. Gray, “Nuclear Weapons,” 110–13. 12. See Elinor Sloane, The Revolution in Military Affairs (Montreal and Kingston: McGillQueen’s University Press, 2002), 32–56. 13. General John Shalikashvili, Joint Vision 2010 (Washington, DC: Department of Defense, 1996); H. H. Shelton, Joint Vision 2020 (Washington, DC: U.S. Government Printing Office, 2000). 14. For an introduction on deterrence theory, see Lawrence Freedman, Deterrence (Cambridge: Polity Press, 2004). 15. John Mearsheimer, Conventional Deterrence (Ithaca, NY: Cornell University Press, 1983), 15. 16. David E. Johnson, Karl P. Mueller, and William H. Taft V, Conventional Coercion across the Spectrum of Operations (Santa Monica: RAND, 2002), 6. 17. Richard K. Betts, “Conventional Deterrence: Predictive Uncertainty and Policy Confidence,” World Politics 37 ( January 1985): 98–100. 18. Richard J. Harknett, “Information Warfare and Deterrence,” Parameters 26 (Autumn 1996): 95. 19. Freedman, Deterrence, 36. 20. Ibid. 21. Patrick Morgan, Deterrence: A Conceptual Analysis (Beverley Hills, CA: Sage, 1983), 221. 22. Bacevich, New American Militarism, 171. 23. Mearsheimer, Conventional Deterrence, 201. 24. Jeffrey Record, “The American Way of War. Cultural Barriers to Successful Counterinsurgency,” Policy Studies, no. 157, September 2006, Cato Institute, Washington, 5. 25. Johnson, Mueller, and Taft, Conventional Coercion, xii. 26. T. V. Paul, Power vs. Prudence: Why Nations Forgo Nuclear Weapons (Montreal and Kingston: McGill-Queen’s University Press 2000), 15. 27. Paul Nitze, “A Threat Mostly to Ourselves,” New York Times, October 28, 1999, quoted in James L. Geick, “Nuclear Weapons and the RMA” (unpublished M.A. thesis, Naval Postgraduate School, 2000), 11–12.

the revolution in military affairs

319

28. Andrew Krepinevich and Steven Kosiak, “Smarter Bombs, Fewer Nukes,” Bulletin of the Atomic Scientists 54 (November–December 1998): 26–33. 29. Philip A. Oden, “Transforming Defense: National Security in the 21st Century,” Report of the National Defense Panel (Washington, DC: United States Government Printing Office, 1997), 8, quoted in Geick, “Nuclear Weapons and the RMA,” 15. 30. Hans M. Kristensen, “U.S. Nuclear Strategy Reform in the 1990s” (working paper, Nautilus Institute, Berkeley, 2000), 15. 31. T. V. Paul, The Tradition of Non-Use of Nuclear Weapons (Stanford: Stanford University Press, 2009), Introduction. 32. William W. Keller and Janne E. Nolan, “The Arms Trade: Business As Usual?” Foreign Policy 109 (Winter 1997–98): 113. 33. Nicolai Sokov, “Russia’s 2000 Military Doctrine,” Nuclear Threat Initiative ( July 2004), http://www.nti.org / db / nisprofs / over/ doctrine.htm. 34. R. S. Norris and H. M. Kristensen, “Russian Nuclear Forces,” Bulletin of the Atomic Scientists 61, no. 2 (2005): 70–72. 35. Frederick W. Kagan, “Star Wars in Real Life: Political Limitations on Space Warfare,” Parameters 28 (Autumn 1998): 117. 36. Patrick M. Morgan, “Taking the Long View of Deterrence,” Journal of Strategic Studies 28 (October 2005): 765. 37. Lawrence Freedman, “Deterrence: A Reply,” Journal of Strategic Studies 28 (October 2005): 798.

Conclusions R JAMES J

.

WIRTZ

few topics in the history of twentieth century social science have received as much attention as the impact of nuclear weapons on world politics and the theory and practice of deterrence. It would therefore not be surprising if our authors managed to make only a small contribution to these subjects. Nevertheless, our volume marks a significant departure in theorizing about nuclear deterrence because it identifies the network of relationships, trends, and issues that characterize the contemporary practice of deterrence in world politics. We clearly do not offer the final word on the role of nuclear weapons or deterrence in twenty-first-century international relations, but we do generate important insights into how technological, scientific, social, and international developments have raised important issues about the theory and practice of deterrence and compellence. Deterrence has been recently overshadowed by an American preoccupation with preemption and preventive war as the primary instruments in the battle against al-Qaeda and in the effort to halt the proliferation of weapons of mass destruction. Our contributors suggest, however, that the practices of deterrence and compellence are only changing, not fading from the international scene. Deterrence remains an important security strategy, despite the fact that it is sometimes undertaken now in situations barely imaginable a few decades ago. This chapter highlights this rich tapestry of issues that characterizes our thinking about deterrence in world politics. It begins by briefly discussing the structural considerations that order contemporary deterrence relationships. As our authors demonstrate, these structural relationships offer more than hypothetical constructs. Instead, they reflect new kinds of deterrence 321

322

james j. wirtz

and compellence situations that create distinctive challenges and opportunities for the parties involved. It then explores the social, behavioral, and technical issues that shape the contemporary practice of deterrence. It concludes with a few thoughts on the emergence of complex deterrence.

structural considerations As T. V. Paul notes in the introduction, the modern practice of deterrence centers on five types of actors and relationships: (1) among great powers; (2) among new nuclear states; (3) between nuclear great powers and regional powers armed with chemical, biological, and nuclear weapons; (4) between nuclear states and nonstate actors; and (5) by collective actors. In contrast to the bilateral and relatively symmetrical deterrent relationship that characterized the nuclear balance between the superpowers, a structural depiction of today’s deterrent relationships captures three fundamental changes in international politics since the end of the cold war. Power asymmetries now characterize many deterrent relationships. Deterrence and compellence relationships no longer primarily involve state actors but can involve international governmental organizations, nonstate actors including terrorist networks, and even domestic or transnational audiences. In the relations among great powers, however, deterrence is increasingly a background consideration as globalization forces great-power competition and cooperation increasingly into economic, political, and social spheres. Power Asymmetries and Deterrence The new deterrent and compellence relationships that garner most of the world’s attention reflect overwhelming power asymmetries. The United States, for instance, has enormous nuclear and conventional military advantages over regional powers armed with chemical, biological, or nuclear weapons. Traditional conceptions of deterrence would predict that on the grounds of military efficacy alone, U.S. threats should be highly credible. Freed from the constraints of mutual assured destruction or in some cases the possibility of even weak retaliation in kind, the United States and its allies should enjoy great success in deterring weaker states or compelling them to comply with their wishes. The same theory would predict that because of their ability to muster resources and legitimacy, collective actors would be effective when it came to issuing threats to gain compliance with international norms or the desires of the international community. Simul-

conclusions

323

taneously, nonstate actors—terrorist networks or transnational political or social movements—are both the targets and users of deterrent and compellent threats. One might predict that these nonstate actors will enjoy relatively little leverage over traditional nation-states, especially when it comes to deterrence and compellence. These asymmetric deterrence and compellence relationships are complicated, however, because they seem to be more animated by the resolve of the parties than by differentials in power. A good illustration of this phenomenon is Robert Jervis’s discussion of the George W. Bush administration’s apparent lapse of faith in deterrence. Jervis identifies a strange paradox in current U.S. foreign policy. Although the United States has overwhelming, full-spectrum military superiority against potential adversaries, members of the Bush administration have little faith in the efficacy of U.S. deterrent or compellent threats. They express alarm about the threats grossly inferior adversaries may mount, which actually increases the ability of governments or nonstate actors wielding militarily meaningless arsenals of chemical, biological, or nuclear weapons to make those threats effective. Ivan Arreguín-Toft suggests that this paradox is explained not by power asymmetries but by an asymmetry in interests at stake and the resolve of the parties. Militarily weak actors can deter or even compel far stronger competitors if they are willing to suffer horrendous casualties to achieve their objectives; they count on the belief that the strong will not bring the full force of their military might to bear or that they will at least abandon a dispute before it becomes a Pyrrhic victory. About a dozen countries now have the military capability to kill millions of people quickly, while scores of states and nonstate actors can kill hundreds of thousands over the course of a few months. Given this diffusion of the capability to inflict death and destruction at relatively low cost, the willingness, not the ability, to inflict death and destruction apparently is used by policy makers to estimate the credibility of deterrent and compellent threats. Under these circumstances, precision-guided conventional munitions become important instruments in the contemporary practice of deterrence and compellence because they can destroy an opponent’s ability to fight while inflicting minimal friendly and even enemy casualties. As Michel Fortmann and Stéfanie von Hlatky note in their chapter, technologies associated with the so-called revolution in military affairs (RMA) bolster the credibility of deterrent threats not only by making democratic leaders more likely to carry out threats but by denying the weak the opportunity to use their casualty acceptance as a weapon against the strong. The impact of the RMA on

324

james j. wirtz

deterrence reflects the general theme that emerges in the survey of asymmetric relationships. Threats are deemed credible not based on the ability to inflict significant death and destruction, a capability that is possessed by both state and nonstate actors. Instead, credibility is based on the perception that an actor is willing to suffer significant costs to achieve what are often judged to be less than existential objectives. The RMA thus creates a paradox of its own: it bolsters deterrence because it greatly reduces the costs involved to all concerned of actually making good on a threat. Nontraditional Actors Deterrence theory and practice generally involve state actors, although important exceptions have always existed to this rule of thumb. The institutions and consultative procedures that stand at the center of the North Atlantic Treaty Organization, for instance, bolster the credibility of the alliance’s deterrent and compellent actions by demonstrating that there are standing planning processes, command procedures, and capabilities to execute threats. Today, compellence and deterrence relationships increasingly seem to involve nonstate actors, such as international governmental organizations, international nongovernmental organizations, transnational and local terrorist and criminal networks, and even domestic and transnational audiences. Deterrence theory can account for the behavior of these various actors, but in practice serious issues emerge that raise doubt about the viability of deterrence as a universal strategy. In a technical sense, deterring nonstate actors is challenging. Terrorist networks might have few assets to hold at risk. It might even be impossible to identify which organization was actually responsible for some terrorist incident. The fundamental conditions needed for successful deterrence might not exist in confrontations with nonstate actors; preventive war or preemption might be the only viable options. Deterrence might be inappropriate because the conditions for success—as posited by standard theory—might not be present. It also might not make strategic sense to attempt deterrence or compellence against all types of nonstate actors. As Paul Kapur notes, nonstate actors that embrace positive goals might not be beyond the reach of deterrence or compellence if state actors can threaten to deny the objectives sought by the nonstate actor. It would be politically challenging to put this type of denial strategy into practice and announce it as official policy, but it could emerge in subtle ways or through strategic interaction. Years of bloodshed in Iraq, for example, may have compelled

conclusions

325

all the parties concerned to help eliminate extremists, abandon their most ambitious objectives, and seek a compromise position. In other words, state actors could create “threats that leave something to chance” when it comes to deterring nonstate actors with positive goals: the state actor might not actually create a regional disaster if the nonstate actor persists in its undesirable policy, but it might also fail to head off catastrophe once forces are engaged. In contrast, if terrorist organizations adopt negative goals, that is, the objective of creating mayhem is seen as a desirable goal in itself, then the execution of deterrent threats is likely to play into the hands of the terrorists. When nonstate actors embrace negative objectives, it makes no sense to help these actors achieve them, even though such a policy could fall under the rubric of deterrence. This observation was highlighted in the chapters written by Emanuel Adler and Janice Gross Stein. When actors embrace negative goals, preventive war, preemption, or mitigation efforts—not deterrence— seem to be the only options available to policy makers. Kapur’s vision of deterrence vis-à-vis nonstate actors raises questions about resolve and the willingness to inflict death and destruction, although it does seem inherently credible that policy makers would be willing to bear and inflict costs to deter mass casualty terrorist attacks. Because avoiding mass-casualty terrorism is a national priority, policy makers might not be “self-deterred,” so to speak, from executing deterrent threats. The “weak deterring the strong,” which characterizes the scenarios described by Arreguín-Toft, might not be so effective when core national values and security concerns are at stake. Collective actors face a slightly different set of challenges and opportunities when it comes to making deterrent and compellent threats. Deterrence and compellence are fundamentally at odds with the liberal and transformational ethos of collective actors. Collective actors are supposed to offer a cooperative and collaborative alternative for resolving disputes rather than the threats and violence that characterize the tradition of realpolitik that has dominated much of world politics. The fact that the pledge of collective defense can have a deterrent effect is generally welcomed by the governments that constitute collective actors. But generating a real military response to an impending or unfolding threat can create an ideological and practical crisis for the collective actor as principles clash with realities and members are asked to bear real costs to defend the peace. The need to take military action quickly highlights the variations in the degree of commitment of member states to execute threats. Weaknesses in planning, command and control,

326

james j. wirtz

doctrine, and resources become readily obvious when quick and decisive actions are most needed. As Patrick Morgan noted, the weaknesses inherent in any organization’s effort to take collective action can make it difficult for these types of actors to make credible deterrent threats. Even if they possess a significant capability to act, challengers to the status quo could be justified in believing that, barring some truly egregious insult to the international community, collective actors cannot muster the political will to act, especially when the costs of action are not trivial. Collective actors can enjoy a great international legitimacy, which can bolster the credibility of their threats. It is therefore not surprising that dominant states may attempt to “hijack” collective actors, hoping to use their legitimacy to cloak narrow national interests. But the paradox of the international legitimacy enjoyed by collective actors lies in the fact that acting to execute threats may prove corrosive to the consensus that backstops that international legitimacy in the first place. Although collective actors often stand idly by in the face of interstate or intrastate violence, sometimes they avoid gridlock and surprise aggressors with an effective military response. Indeed, one is left to wonder whether the behavior of collective actors is too unwieldy or unpredictable to serve as a reliable stabilizing force during crises. A third nontraditional actor that emerges somewhat tangentially in our volume is national and transnational audiences when it comes to the practice of deterrence. The desirability and legitimacy of deterrent and compellent threats are increasingly subject to scrutiny by domestic public opinion, judgments made by ethnic communities in other states, attitudes on some foreign “street,” or a global consensus articulated by a collective actor. Purists might argue that there is nothing new about this phenomenon; after all, all war is about politics. But in an age of globalization, when information technologies invite people everywhere to develop and express an opinion about international events, strategists and policy makers seem to be highly aware of the power of the political audience. Jervis’s discussion of the Bush administration, for example, highlights the possibility that U.S. officials may have believed that the American public would not tolerate another deterrence failure, making preventive war the only politically acceptable alternative open to the administration. The weak want to engage public opinion, if not actually get local folks to put their bodies directly on the firing line, in an effort to make or break deterrent and compellent threats. By contrast, the powerful use RMA-based weapons systems to increase deterrence credibility by detaching the public from war, turning combat into entertainment by vir-

conclusions

327

tually eliminating collateral damage and significant casualties. Once again, one is left to wonder whether support, opposition, and indifference, among relevant audiences, might be a growing factor in the success and failure of deterrent and compellent threats. Deterrence and the Great Powers Among the great powers, there is evidence that the marginalization of nuclear weapons is occurring.1 Admittedly, the motives for marginalization are not always pure. Chinese nuclear force modernization has probably been slowed by the nuclear test moratorium; the Russians could barely afford to keep their military afloat in the 1990s; and the United States has embraced the war-fighting potential of the RMA as a more credible deterrent than nuclear weapons. Nevertheless, the salience of nuclear weapons is fading in the strategic policies of some governments, although there is little official support for abandoning nuclear weapons or abolishing nuclear deterrence. The generally benign atmosphere that characterizes today’s great-power politics is largely responsible for the minimal salience of conventional forces and nuclear weapons in relations among the great powers or, for that matter, among all advanced industrial nations. There really are few other factors that can account for the apparent lack of concern among them about deterrence, including the nuclear and conventional balance of forces. For instance, the leaders of the world’s nuclear-weapons states have not become abolitionists. Although nuclear modernization programs have slowed and existing arsenals have lost some of their cold war punch, these arsenals still consist of weapons that posed the tangible threat of Armageddon just a few decades ago. Even more unnerving is that military balances among the great powers now shift continuously, with the United States generally ending up far ahead of all potential competitors. Different modernization rates, uncoordinated force drawdowns and retirements, the military application of new technologies, and efforts at offense-defense integration also can alter the military status quo, despite the absence of much concern about potential strategic consequences. The great powers also conjure up various insults to international decorum with few immediate consequences. Bold Russian pronouncements about their reliance on nuclear weapons when facing conventional threats, U.S. policies of preventive war, and a Chinese test of an antisatellite weapon all seem to bank on the forbearance of the other great powers. As today’s Cassandras are fond of noting, stability among the great powers has not prevented advances in military technology, diplomatic or

328

james j. wirtz

strategic insults, or changes in force structures. If these military imbalances ever became highly salient, during an unanticipated crisis, for instance, there might be hell to pay. As Morgan and Paul note, leaders tend to ignore this changing military balance because they believe that military force is not a particularly valuable tool in their relations. Military force, especially nuclear weapons, helps guarantee that other great powers will not be tempted to use violence to achieve their objectives because the costs and risks of aggression would easily outweigh any conceivable political gains. Because little in their political disputes is really worth fighting for, nuclear and conventional forces remain largely irrelevant in great-power relations today. Nuclear deterrence has become a secondary or tertiary issue in the minds of officials and officers because there is not much interest or motivation for war among the great powers.

the changing social, scientific, and strategic setting Overlaid across the structural considerations that shape the contemporary practice of deterrence is a set of social, scientific, and strategic developments that suggests that the logic of deterrence theory may no longer accurately describe the dynamics or forces unleashed by modern conflict. This sort of thing has happened in the past: a British professional military found it impossible to defeat politically motivated amateurs in the American revolution; the eighteenth-century theory and practice of warfare produced unmitigated disaster in the early nineteenth century; and the massive ground armies of World War II were made largely obsolete by the nuclear-armed air forces of the cold war. In other words, history is replete with situations in which what were thought to be the verities of strategy no longer reflect strategic realities. Society, science, and strategy have continued to change since the dawn of the nuclear age, and advances in deterrence theory have not kept pace with these changing circumstances. It is time to take a more nuanced view of the theory and practice of deterrence. The previous chapter highlighted the impact of RMA-based technologies on the practice of deterrence. The information revolution, however, has also spawned a host of new technologies that have generally empowered individuals, grassroots organizations, and transnational actors at the expense of traditional bureaucracies and state actors. As globalization continues to foster all sorts of commercial, professional, and social interaction around the

conclusions

329

world, will international borders become increasingly meaningless, at least for the “globalized” class? Will deterrent and compellent threats defy logic and reason when they are directed against one’s business interests, friends, or even family members who happen to live outside one set of national borders? Admittedly, there is nothing new about the failed expectation that “the bankers will never allow it” when it comes to describing war’s negative economic consequences. But more individuals than ever have a vested interest in now keeping open growing communications and commerce with people around the world, and they act on that interest every day. Under these circumstances, will deterrent threats inherently lack credibility? Another challenge to deterrence theory is raised by Janice Stein’s discussion of the recent findings of cognitive and neurocognitive scientists about the way individuals respond to threats. One widespread assumption behind deterrence theory is that individuals make reasonable assessments of the costs and benefits associated with an initiative, forgoing some action if the expected costs outweigh the expected gains. Cognitive psychologists, by contrast, are suggesting that affect—subjective likes and dislikes and visceral responses to stimuli—shapes assessments of developing situations and likely consequences. Because people dislike being threatened, especially when something they value is at stake, they are likely to react emotionally, not rationally, to a threat, making it difficult to undertake the cool cost-benefit calculation at the heart of deterrence theory. It might be premature to offer this judgment, but the making of deterrent and compellent threats might be akin to waving a red cape in front of a bull or yelling “fire!” in a crowded theater. Both acts will get the audience moving but not because of some rational calculation and not always in the desired direction. Emanuel Adler and Timothy Crawford also identify a strategic weakness inherent in deterrence theory. Actors with negative goals can harness others’ deterrent threats to achieve their objectives. According to Adler, governments can face a deterrence trap when they react with force to a provocation. Sometimes such a reaction is necessary not because of the values threatened in the present showdown but simply to preserve a reputation for honoring deterrent threats, a reputation that will come in handy when core values are actually threatened. Actors can thus face a dilemma: they can preserve their reputation and play into the hands of provocateurs or practice what amounts to a form of appeasement. The deterrence trap has always existed, but it is less pronounced when deterrence is practiced among equals who seem to have less to gain by attempting to harness the death and destruction of their opponent’s arsenals

330

james j. wirtz

to affect the perceptions of their own citizens or other interested audiences. Deterrence theory, however, has little to say about the possibility that actors may provoke the execution of deterrent or compellent threats to hijack another party’s military force to achieve their own objectives. The deterrence trap is partially a manifestation of the information age, which has empowered individuals in terms of communication, organization, and analytical capabilities but has not made individuals or small organizations the equal of states when it comes to undertaking sustained military operations. Terrorists and nonstate actors seek to maximize their power in world politics or at least to make significant threats that are out of proportion to the limited resources under their direct control. It is therefore not surprising that they will seek military capabilities where they find them, even their adversary’s military capabilities, to motivate their followers or intimidate their opponents.

in lieu of conclusions Complex deterrence is a reality. From a structural perspective, complexity is created by the fact that (1) deterrent relationships now exist among different types of international actors, (2) deterrent relationships are often characterized by significant power asymmetries, and (3) credibility now appears largely to be a matter of the opponent’s perceived willingness to bear the costs involved in executing a deterrent or compellent threat. The fact that the great powers have pushed issues of nuclear weapons into the background highlights the political nature of deterrence. When there is a potential for conflict, deterrence, the finer points of the nuclear balance, and declaratory postures become salient. Deterrence retains a role among the great powers, but given the relatively benign nature of their relationships, most officials take the deterrent power of nuclear arsenals and the general stability of great-power relations for granted. From a social, scientific, and strategic perspective, complex deterrence is shaped by changing technology, globalization, and the ability and willingness of weak actors to hijack the deterrence and compellence strategies of others to achieve their objectives. It also appears that the fields of psychology and neuroscience are on the verge of a significant reconceptualization of human cognition, which calls into question some of the assumptions common in both the theory and practice of deterrence. This layer of social, strategic, technological, and scientific changes interacts with the asymmetric conflicts that gain the lion’s share of international attention, helping to determine whether or not deterrence and compellence are successful.

conclusions

331

The structure and trends highlighted in our volume, however, have not fully emerged on the world stage, which makes it impossible to predict the impact of complex deterrence on international relations and the prospects for stability. But these trends and structural factors continue to emerge, despite the fact that scholars and officials alike generally treat deterrence and nuclear weapons as a vestige of the increasingly remote cold war. Perhaps they are right; as globalization continues, conflict and the use of force hopefully will be banished to an increasingly inconsequential and geographically limited portion of the planet. But if hostility and competition again dominate relations among the great powers, increasing the salience of military balances and nuclear postures, it will be a mistake to assume that future deterrence will closely resemble the deterrence of the past. Today and for the future, deterrence and compellence are characterized by a new complexity.

note 1. John Baylis, “Nuclear Weapons, Prudence, and Morality: The Search for a Third Way,” in Alternative Nuclear Futures, edited by John Baylis and Robert O’Neill (Oxford: Oxford University Press, 2000), 85.

Contributors

editors t. v. pau l is James McGill Professor of International Relations in the Department of Political Science at McGill University, where he has been teaching since 1991. Paul specializes in international relations, especially international security, international conflict and conflict resolution, regional security, and South Asia. He received his undergraduate education from Kerala University, India; an M.Phil. in international studies from Jawaharlal Nehru University, New Delhi; and an M.A. and Ph.D. in political science from the University of California, Los Angeles. Paul is the author or editor of eleven books and thirty-five scholarly articles and chapters. He is the author of The Tradition of Non-Use of Nuclear Weapons (Stanford University Press, 2009); India in the World Order: Searching for Major Power Status (with Baldev Raj Nayar, Cambridge University Press, 2002); Power versus Prudence: Why Nations Forgo Nuclear Weapons (McGillQueen’s University Press, 2000); and Asymmetric Conflicts: War Initiation by Weaker Powers (Cambridge University Press, 1994). Paul is the editor of The India-Pakistan Conflict: An Enduring Rivalry (Cambridge University Press, 2005); and coeditor and contributor to four volumes including Balance of Power: Theory and Practice in the 21st Century (with James Wirtz and Michel Fortmann, Stanford University Press, 2004). He was professor of national security affairs at the Naval Postgraduate School, Monterey, California (2002–3), visiting scholar at Harvard University’s Center for International Affairs and the Olin Institute for Strategic Studies (1997–98), and visiting affiliate at the Center for Nonproliferation Studies, Monterey (2002–3). He currently serves as the director of the University of Montreal–McGill Research Group in International Security. patrick m. morgan is Tierney Chair in Peace and Conflict Studies at the Department of Political Science, University of California, Irvine, where he has taught since 1991. His research interests include international politics, national and international

333

334

list of contributors

peace and security issues, U.S. foreign relations; U.S.-European relations; and international relations theory, with specific reference to deterrence theory, strategic surprise attack, arms control, and related subjects. He has also had a long-standing interest in theoretical approaches to the study of international politics. Currently, he is involved in projects on the theory and practice of deterrence in the post–cold war era, security strategies for global security management, and security in Northeast Asia. His publications include Deterrence: A Conceptual Analysis, 2nd ed. (Sage Publications, 1983); Deterrence Now (Cambridge University Press, 2003); and ReViewing the Cold War, coedited with Keith Nelson (Praeger, 1999). j ame s j . wi rt z is acting dean of the School of International Graduate Studies and a professor in the Department of National Security Affairs, Naval Postgraduate School, Monterey, California. He is the editor of the Palgrave Macmillan series Initiatives in Strategic Studies: Issues and Policies. Wirtz earned his degrees in political science from Columbia University (M.Phil., 1987; Ph.D., 1989) and the University of Delaware (M.A., 1983; B.A., 1980). He was a John M. Olin Fellow at the Center for International Affairs, Harvard University (1983–85). In 2004–5, he was a visiting professor at the Center for International Security and Cooperation, Stanford University. Wirtz is a former chair of the Intelligence Studies Section of the International Studies Association and a past president of the International Security and Arms Control Section of the American Political Science Association. Wirtz is the author of The Tet Offensive (Cornell University Press, 1991, 1994); coeditor with T. V. Paul and Richard Harknett of The Absolute Weapon Revisited: Nuclear Arms and the Emerging International Order (University of Michigan Press, 1998, 2000); coeditor with Peter Lavoy and Scott Sagan of Planning the Unthinkable: New Powers and Their Doctrines for Using Chemical, Biological and Nuclear Weapons (Cornell University Press, 2000); and coeditor with T. V. Paul and Michel Fortmann of Balance of Power: Theory and Practice in the 21st Century (Stanford University Press, 2004).

contributors e m a n u e l a d l e r is the Andrea and Charles Bronfman Chair in Israeli Studies in the Department of Political Science at the University of Toronto and coeditor of International Organization. His research interests include the international politics of identity and peace, Israel and the Mediterranean, European security institutions, and international relations theory. Adler earned his B.A. and M.A. degrees at Hebrew University, followed by a Ph.D. from Berkeley. His publications include The Power of Ideology (University of California Press, 1987); Progress in Postwar International Relations (coedited with Beverly Crawford, Columbia University Press, 1991); Security Communities (coedited with Michael Barnett, Cambridge University Press, 1998); and articles in International Organization, European Journal of International Relations, and Review of International Studies. His current projects include a study of rationality and the Israeli-Palestinian conflict, a constructivist reconsideration of strategic logic, includ-

list of contributors

335

ing deterrence, in post–cold war international security, a project on the role of practice in international relations, and a study on Europe as a civilizational state. i van ar r e gu í n - t o f t is research fellow in the International Security Program at Harvard University’s Kennedy School of Government and visiting assistant professor of political science at Wellesley College. His work focuses on the interaction of strategy and conflict outcomes, with a particular emphasis on asymmetries of material power. His first book, How the Weak Win Wars: A Theory of Asymmetric Conflict, was published by Cambridge University Press in 2005. His current book manuscript evaluates the consequences of the systematic harm of noncombatants as a strategy in war and is tentatively entitled The [F]utility of Barbarism. timothy w. crawford is associate professor at Boston College where he teaches courses in international security, causes of war, the United Nations, and intelligence. His research interests include deterrence and wedge strategies in alliance politics, intelligence cooperation, and executive power in U.S. foreign policy. He is the author of Pivotal Deterrence: Third Party Statecraft and the Pursuit of Peace (Cornell University Press, 2003), which was the winner of the 2003 Edgar S. Furniss Book Award. He is also coeditor with Alan J. Kuperman of Gambling on Humanitarian Intervention: Moral Hazard, Rebellion, and Civil War (Routledge, 2006). mi c h e l fort m a n n is cofounder of the Research Group in International Security (1996). His research and teaching interests include contemporary military history and the evolution of warfare. He is the author of Les cycles de Mars. Révolutions militaires et édification étatique, de la Renaissance à nos jours (Economica, Paris, 2009). He is also coeditor with T. V. Paul and James Wirtz of Balance of Power, Theory and Practice in the 21st Century (Stanford University Press, 2004). He has written several books and numerous articles / book chapters, including A Diplomacy of Hope: Canada and Disarmament, coauthored with Albert Legault (McGill-Queen’s University Press, 1992); and Le système politique américain, mécanismes et décisions (4th ed.), coauthored with Pierre Martin (Presses de l’Université de Montréal, 2008). frank p. h arv e y currently holds a University Research Professorship of International Relations and is a former director of the Centre for Foreign Policy Studies at Dalhousie University. His books include The Homeland Security Dilemma: Fear Failure and the Future of American Insecurity (Routledge, 2008); Smoke and Mirrors: Globalized Terrorism and the Illusion of Multilateral Security (University of Toronto Press, 2004); Millennium Reflections on International Studies (coedited with Michael Brecher, University of Michigan Press, 2002); Using Force to Prevent Ethnic Violence: An Evaluation of Theory and Evidence (with David Carment, Praeger, 2001); Conflict in World Politics: Advances in the Study of Crisis, War and Peace (coedited with Ben Mor, Macmillan Press 1998); and The Future’s Back: Nuclear Rivalry, Deterrence Theory and Crisis Stability after the Cold War (McGill-Queen’s, 1997). pat r i c k j ame s is director of the Center for International Studies and professor of international relations at the University of Southern California. His teaching and

336

list of contributors

research interests are in the fields of international relations (conflict, crisis, and war), comparative politics (Canadian politics), rational choice (collective action, expected utility, and game theory); and empirical methods (research design and statistics). His most recent book is International Relations and Scientific Progress: Structural Realism Reconsidered (Ohio State University Press, 2002), and its sequel, currently in progress, focuses on creating a version of neorealist theory called elaborated structural realism. He is the author or editor of twelve books and over one hundred articles and book chapters. He is a past president of the Midwest International Studies Association and the Iowa Conference of Political Scientists. From 1999 to 2003 he served as editor of International Studies Quarterly. Currently he serves as the vice president of the International Studies Association. rob e rt j e rv i s is Adlai E. Stevenson Professor of International Affairs at Columbia University. He specializes in international politics, security policy, and decision making. Among his books are The Psychology and Politics of Intelligence and Intelligence Failures (Cornell University Press, forthcoming); American Foreign Policy in a New Era (Routledge, 2005); System Effects: Complexity in Political and Social Life (Princeton University Press, 1997); The Meaning of the Nuclear Revolution (Cornell University Press, 1989); Perception and Misperception in International Politics (Princeton University Press, 1976); and The Logic of Images in International Relations (Columbia University Press, 1989). Jervis also is a coeditor of the Security Studies Series published by Cornell University Press and a past president of the American Political Science Association in 2000–1. In 2006, he received the National Academy of Sciences’ award for contributions of behavioral science to avoiding nuclear war. s . pau l k ap ur is associate professor in the Department of National Security Affairs at the United States Naval Postgraduate School, Monterey, and a faculty affiliate at the Centre for International Security and Cooperation, Stanford University. He is the author of Dangerous Deterrent: Nuclear Weapons Proliferation and Conflict in South Asia (Stanford University Press, 2007) and several articles in journals such as International Security, Security Studies, and Asian Survey. His research interests include South Asian politics and security, the strategic impact of nuclear weapons proliferation, ethnoreligious violence, and deterrence theory. j e f f r e y w. k n o p f is an associate professor at the Naval Postgraduate School, Monterey, where he has been on the faculty since fall 2000. Knopf received a Ph.D. in political science from Stanford University (1991) as well as an M.A. from Stanford and a B.A. from Harvard University. Knopf is the author of Domestic Society and International Cooperation: The Impact of Protest on U.S. Arms Control Policy (Cambridge University Press, 1998) and the coauthor of U.S. Arms Exports: Policies and Contractors (Ballinger, 1988). He has also published articles in International Organization, Journal of Conflict Resolution, Journal of Peace Research, Security Studies, Review of International Studies, The Nonproliferation Review, Contemporary Security Policy, and various edited books. Knopf ’s current research focuses on the role of deterrence after September 11.

list of contributors

337

di ns h aw mi s t ry is associate professor of political science and director of Asian Studies at the University of Cincinnati. He was previously a fellow at the Belfer Center for Science and International Affairs, Harvard University, and the Center for International Security and Cooperation, Stanford University. Mistry specializes in international relations, security studies, technology and politics, and Asian security. He is the author of Containing Missile Proliferation (University of Washington Press, 2003) and numerous articles in scholarly journals such as Asian Survey, Current History, Security Studies, and International Security. j ani c e gross st e i n is Belzberg Professor of Conflict Management and Negotiation in the Department of Political Science and the director of the Munk Centre for International Studies at the University of Toronto. She is the author of Rational Decision Making: Israel’s Security Choices, 1967 (Ohio State University Press, 1980, coauthored with Raymond Tanter), Choosing to Cooperate: How States Avoid Loss ( Johns Hopkins University Press, 1991); We All Lost the Cold War (coauthored with Richard Ned Lebow, Princeton University Press, 1994); Powder Keg in the Middle East: The Struggle for Gulf Security (coedited with Geoffrey Kemp, University Press of America, 1995), and Networks of Knowledge: Collaborative Innovations in International Learning (coauthored with Richard Stren, Joy Fitzgibbon, and Melissa MacLean, University of Toronto Press, 2001). Her most recent book is The Unexpected War: Canada in Kandahar (coauthored with Eugene Lang, Penguin Press, 2007). She is a member of the Order of Canada, the Order of Ontario, and an Honorary Foreign Member of the American Academy of Arts and Science. stéfanie von hlatky is a doctoral candidate at the Université de Montréal and holds a B.A. from McGill University. Her dissertation, entitled The Great Asymmetry: America’s Closest Allies in Times of War, focuses on how allies respond to American war initiatives. More specifically, it looks at how power asymmetries shape intra-alliance interaction patterns over time. She is also deputy director for the Research Group in International Security. Von Hlatky is the recipient of doctoral awards from the Fonds québécois de recherche sur la société et la culture and the Security and Defence Forum.

Index

active defense, 195, 289 Adler, Emanuel, 16, 75, 325, 329 Afghanistan, 26n40, 87, 120, 128n47, 167, 205, 210, 213–214, 218, 260, 268, 288, 298, 311, 317 African Union, 159 Ahmadinejad, Mahmoud, 101 Albright, Madeleine, 220, 231, 233, 238 Alexander, Jeffrey, 87 Algeria, 205, 213 Allison, Graham, 109. 116 Al-Qaeda, 5, 15, 31, 47, 73, 79, 96, 97, 99, 101–102, 110, 114, 119–123, 127n31, 128n47, 129n52, 215, 285, 321 Annan, Kofi, 237–241 anti-ballistic missile system (defenses), 44, 50, 143, 268, 270, 273, 274, 314 Anti-Ballistic Missile Treaty, 142 Argentina, 212 Armageddon, 121–122, 129n59, 327 Armenia, 25n22, 298; genocide and, 297 Arreguin-Toft, Ivan, 15, 95, 323, 325 Asahara, Shoko, 121 asymmetric: attacks, 18; capability, 14–15, 185, 188, 200; competition, 315, 317; conflict, 85, 87, 103, 204–206, 210, 220n13, 309, 330; deterrence, 13–15, 94, 96–97, 102, 323; strategies, 316; structural conditions, 91, 97, 99, 103; threats, 31, 34; warfare, 23,

75–77, 85, 87, 94–97, 99, 101, 103, 185, 199, 205 audience costs, 284, 297 Aum Shrinrikyo, 16, 110, 119, 121–122, 129n59, 129n60 Australia, 122, 209; government, 56n28 Azerbaijan, 25n22, 298 Aziz, Tariq, 231, 237–239 B-52 bomber, 235 Baldwin, David, 37 Baldwin, Stanley, 55n21 bandwagoning, 164, 167 Barzani, Massoud, 303n74 Berlin, Isaiah, 64 Betts, Richard, 282 biological systems, 25n20 biological weapons, 7, 9, 13–14, 18, 40, 44, 75, 109, 141, 194–195, 249, 262, 322; agents, 26n32; attack, 1, 13, 121, 249; proliferation, 44 bipolarity, 2, 17, 20, 44, 70, 135 Blair, Tony, 240, 242 Boer War, 154 Bolton, John, 80n1 Bonaparte, Napoleon, 266 Borden, William, 135, 137 Bosnia, 171, 214, 225, 228 Boulding, Kenneth, 18

339

340

index

Brazil, 177 Brennan, Donald, 46 Brodie, Bernard, 2, 44, 89, 135, 275n5 buck-passing, 16, 266, 281 Bush, George H. W., 14, 221n28, 225, 227, 229, 235 Bush, George W., 58, 61, 72–73, 156n21, 243–244, 297; administration of, 5, 20, 31, 50, 55n15, 128n45, 134, 141–155, 161, 179, 205, 218, 285, 307, 311, 313–314, 316, 323, 326; critics, 42, 146, 151; doctrine, 31 Butler, Richard, 238, 241 Butler Report, 241 Canberra Commission, 56n28 Caribbean Crisis, 140 Carter, Jimmy, 214 Castro, Fidel, 141 Chamberlain, Neville, 66 Chechens, 16 chemical weapons, 7, 9, 13–14, 16, 18, 26n32, 40, 44, 109, 141, 194–195, 199, 238, 241, 244, 262, 322–323; accidents, 121; attack, 1, 13, 121, 249; nonuse, 199; proliferation, 44 Cheney, Richard, 143 China, 1, 10–11, 23, 101, 133, 142–143, 177, 183, 205, 215, 236–237, 242–244, 246–247, 261, 266–267, 269, 272–274, 275n7, 288; conventional forces, 269; invasion of Korea, 215l; military doctrine, 276n22; nuclear policy, 264, 270, 273, 275n12, 275n18, 281; position on North Korea, 181n15; relations with Taiwan and the United States, 10, 159, 271, 282, 296, 316 Christopher, Warren, 231 Churchill, Winston, 52, 154 Clinton, Bill, 15, 142, 151, 185, 228, 231–232, 235–236, 240, 242; administration of, 164, 233–235, 239 Clinton, Hillary, 278 coalition of the willing, 164, 167, 181 coercive diplomacy,4, 21, 25n27, 26n39, 60, 73–74, 161, 174, 180, 190, 222–223, 235, 244–245, 284, 289, 298 cognitive psychology, 60, 62–64, 66, 89, 329 Cohen, Jonathan, 69 cold war deterrence framework, 25n27 cold war deterrence social structure, 87 collateral damage, 96, 100, 153, 207, 215, 305, 311, 316, 327

collective-actor coercion, 165, 173 collective-actor deterrence, 3, 9, 16–19, 26n40, 141, 158–180, 180n8, 278, 286–288, 295, 325–326 collective-actor nonmilitary coercion, 181n12 compellence, 1, 14, 20–21, 25n27, 25n28, 26n39, 37, 88, 96, 99, 108n65, 134–135, 137, 152, 160–162, 167–170, 172–176, 178, 180n8, 184, 187, 189–190, 197, 199–200, 222–250, 317, 321–325, 330–331 concatenated theory, 279–280 consequence management (CM), 40, 55n15 conventional deterrence, 2, 18, 25n22, 165, 186, 217, 271, 282, 308–309; failures, 47 cost-benefit calculation, 3, 5–6, 90, 92, 102, 104, 208, 220n11, 249, 329 Crawford, Timothy, 15, 172, 180n8, 329 crisis stability, 93, 186, 194, 196, 312 Cuban Missile Crisis, 96–97, 140, 197, 202n39, 222, 260 Czech Republic, 273 Czech smuggling case, 113–114, 127n31, 127n32 Dallaire, Romeo, 216 Damage limitation, 262 Damasio, Antonio, 68 Danilovic, Vesna, 293 Darfur, 26n40, 177, 216–217, 219, 287 Davis, Glyn, 234 Davis, Jay, 124 “deep strike,” 306 defensive cognitions, 64 Defense Planning Guidance, 143 deterrence: basic, 277, 282–283; by denial, 2, 24n6, 38, 40–41, 44, 49, 88–89, 93, 110, 115–117, 119, 122, 129n51, 205, 214, 324; by punishment, 2, 24n6, 38, 40–41, 49, 88–90, 110, 117–120, 122–123, 125, 128n44, 130n64, 134–135, 247, 289; demand side, 110, 114–115, 117–120, 122–123; homeland, 135; mutual, 11, 14, 21, 24n15, 55n21, 94, 136, 262, 267–268, 272; pivotal, 172, 180n8, 294–295; supply side, 110–115, 118–120, 122; tripwire, 278, 283, 287 Deterrence Operations Joint Operating Concept (DO-JOC), 39–40, 55n13 deterrence theory: core premises, 2–3; rational (RDT), 47–50, 56n38, 57n40, 58–80, 222–223; self-deterrence, 5, 7, 18–19,

index 153–154, 216, 325; traditional, 33, 36, 53, 226, 322 deterrence trap, 75, 85, 88, 95–96, 99–104, 329–330 Diehl, Paul, 170 Diesing, Paul, 290, 293 dirty bombs, 5, 16, 75, 198 egocentric bias, 66 Egypt, 67, 101, 185, 199, 200n3, 240 Ekeus, Rolf, 26n32 Emile, 220 enduring rivalry, 4, 18, 190, 259–260, 265–166, 275n16, 315 Epistemic community, 34, 93, 103, European Security Strategy, 101 European Union, 101, 168, 178 existential deterrence, 10, 15, 265–266 F-15, 228 F-16, 228 failed states, 7, 16, 297 Fair, Christine, 190 Falk, Richard, 218 Falkland Islands, 212–213 Fearon, James, 56n36, 293 Fick, Nathaniel, 100 fissile material, 110–116, 118–124, 126n22, 126n25, 127, n31, 127n32, 129n51, 195, 198 fog of war, 305 “Follow on Forces Attack,” 306 Fortmann, Michel, 11, 323 Foucault, Michel, 91 France, 1, 10, 13, 227, 231, 233, 235–236, 243–247, 261–264, 269, 274 Freedman, Lawrence, 34, 37, 39, 100, 205, 209, 211, 220n11, 280, 305 free-riding, 17, 171, 281 Frost, Robin, 109 Gaddis, John Lewis, 143, 208 Galluci, Robert, 109, 124 general deterrence, 10–11, 24n6, 39, 55n11, 72, 89, 135, 164, 172, 259, 265–266, 275n20, 279–284, 293–294; extended, 279, 282–283, 294, 299n15; nuclear, 71 George, Alexander, 47, 54, 79–80, 155, 160, 248, 279, 292 George Washington, 231

341

Georgia, 298 Germany, 70, 212–213; West, 299n10 Glimcher, Paul W., 82n30 Global Strike, 267, 308 globalization, 11, 18, 141, 159, 259, 272–273, 322, 326, 328, 330–331 Gray, Colin, 48–49, 57n41 Greece, 206, 295 Grenada, 216 Guam, 235, 271 hair-trigger alert, 24n15 Haiti, 164 Halliday, Denis, 240 Hamas, 16, 99, 199 Hamdoon, Nizar, 232, 240 hard-target kill capability, 267 Harknet, Richard J., 309 Harvey, Frank, 14, 50 Hezbollah, 88, 96–102, 105n4, 199 Hiroshima bomb, 111, 113–114 Hitler, Adolf, 66, 70, 266 Hobbes, Thomas, 209 Holocaust, 101 House of Commons, 55n21 Hukbalahap, 206 Hume, David, 292 Huntington, Samuel P., 212 Hussein, Saddam, 14, 26n32, 47, 50, 52, 61, 73, 148, 151, 215, 221n28, 226–227, 231–232. 235–236, 239–240, 243–245, 247, 249, 311, 317 Huth, Paul, 291 Hutu, 214, 216 immediate deterrence, 14, 18, 24n6, 55n11, 135, 266, 275n20, 281–283, 289, 294, 298, 299n15; extended, 280, 284, 291, 293–294 India, 4, 11–13, 23, 177, 183–200, 201n14, 201n19, 272, 275n18, 295, 298; Cabinet Committee on Security, 194; cold start strategy, 12, 194; National Command Authority, 194; National Security Advisory Board, 193; nuclear test, 12, 183, 193, 197; Political Council, 194; Strategic Force Command, 194 Indian Ocean, 231 Indochina, 205–206, 213 “industrial” warfare, 306

342

index

intercontinental ballistic missiles (ICBMs), 261, 264, 267, 270 intergovernmental organizations, 159, 171 International Atomic Energy Agency (IAEA), 145, 237–240, 247 Iran, 4, 11, 13, 15, 58, 97, 99, 101–103, 145–147, 149, 153–154, 199, 278, 288, 298, Iraq, 4, 13, 15, 21, 26n32, 52, 99, 101, 112, 154, 166, 222–249, 268, 278, 288, 296–298, 324; Ba’athist, 231, 240–241; first Iraq War, 163, 199, 212–213, 216, 306, 311; Republican Guard, 231, 233; Revolutionary Command Council, 240; second Iraq War, 50–51, 55n17, 61, 66, 73–74, 96, 100, 143, 150, 154, 164, 167, 205, 215, 218, 221n28, 222, 285, 311, 317 Israel, 4, 11, 13–14, 66–67, 77, 88, 96–102, 105n4, 108n65, 142, 145, 185, 195, 198– 199, 200n3, 275n18, 278, 287, 298 James, Patrick, 14 Japan, 12, 12, 121, 179, 215, 271, 277, 316 Jervis, Robert, 2, 15, 34, 47–48, 323, 326 Joint Task Force-Southwest Asia ( JTF-SWA), 227–229 Joyu, Fumihior, 121 “just in time,” 73, 308 Kahneman, Daniel, 69, 95 Kapur, S. Paul, 16, 187, 201n11, 324–325 Kargil, 4, 185, 189 Kargil War, 4, 12, 185, 188–189 Kashmir, 4, 184–185, 187–188, 190–191, 199, 201n22 Kennedy, John F., 143, 206 Khalizad, Zalmay, 143 Khan, A. Q., 16, 112–113, 142, 193, 195 Khan Research Laboratories, 112 Khrushchev, Nikita, 140 Kim, Jong Il, 47 Kim Il Sung, 140 Knopf, Jeffrey, 171 Korean War, 140, 260 Kosiak, Steven, 313 Kosovar Albanians, 216–217 Kosovo, 153, 171, 216–217, 225, 295, 317 Kozyrev, Andrei V., 233 Krauthammer, Charles, 51 Krepinevich, Andrew, 313

Kurdish Regional Government (KRG), 296–297, 303n74 Kurdistan Workers Party (PKK), 296–297 Kurds, 224–227, 234–236, 297, 298 Lahore Accords, 189 launch on warning, 10, 24n15, 263, 270 Lebow, Richard Ned, 47, 56n38 Lee, Steven, 18 “legacy” weapons, 308 Lewis, Jeffrey G., 113–114, 126n25, 127n31 Libet, Benjamin, 68 Libya, 112, 142, 150–151 Litwak, Robert, 150 Locke, John, 209 “Long Peace,” 208, 260 magnetic resonance imaging (MRI), 81n26 Magsaysay, Ramon, 206 Malaya, 206, 219 Martinage, Robert, 306 massive retaliation, 35, 49, 138, 152, 278 McNamara, Robert, 140, 211 Mercer, Jonathan, 285 MIG-25, 228 “military-technical revolution,” 306 Milosevic, Slobodan, 214, 225, 245 Mistry, Dinshaw, 12, 272 Mogodishu, 216; battle of, 214, moral hazard, 71, 291–295 Morgan, Patrick, 10, 27n50, 35, 39, 144, 278, 285, 295, 310, 326, 328 Mutual Assured Destruction (MAD), 2, 35, 46, 93–94, 143, 260, 308–309 Napoleonic Wars, 205 neuroscience, 60, 62, 67–70, 78, 95, 330 New York Times, 225, 251n20 Ngo, Dinh Diem, 206 Nixon, Richard, 136, 139 Nodong ballistic missile, 113 no-fly zone (NFZ), 224–225, 227-, 231, 233, 236, 245–247, 251 nongovernmental organizations, 73, 159, 162, 324 nonproliferation, 168, 178, 278, 288 North Atlantic Treaty Organization (NATO), 11, 25n28, 26n40, 153, 158–159, 164–166, 170–171, 177, 208, 216–217, 269–270, 273– 274, 280–282, 292, 295–297, 299n10, 306

index North Korea, 4, 12, 14, 112–113, 142, 146–147, 149, 151, 153, 159, 164, 179, 181n12, n15, 185, 198, 214, 269, 273; nuclear test, 277 nuclear material. See fissile material Nuclear Posture Review (2002), 313–314 nuclear revolution, 2, 260 nuclear weapons: first strike, 44, 196, 260, 262, 263, 268, 317; no-first-use policy, 193–195, 261; proliferation, 4, 20, 44, 50, 74, 165, 167, 171, 263–264; second-strike, 2, 193; taboo, 18, 314; tradition of nonuse, 2, 18–19, 26n42, 193, 314, 317 Nye, Joseph, 94 “Of the Balance of Power,” 292 Operation Desert Fox, 236–242 Operation Desert Storm, 306, 317 Operation Desert Strike, 234–236 Operation Enduring Freedom, 214 Operation Northern Watch, 224–227 Operation Provide Comfort, 224–227 Operation Southern Watch, 227–230 Operation Vigilant Warrior, 230–234 Organization of American States (OAS), 158–159 Osirak nuclear facility, 13 Pakistan, 4–5, 11–12, 112–113, 126n22, 142, 183–200, 201n14, 202n39, 272, 275n18, 295, 298; Development Control Committee, 192; Employment Control Committee, 192; Joint Chiefs of Staff Committee, 192; National Command Authority (NCA), 192–193l; nuclear test, 12, 183, 197; Strategic Employment Committee, 192; Strategic Plans Division, 192 Panama, 216 Parachini, John, 109 passive defense, 292, 298 Patriotic Union of Kurdistan, 235 Paul, T. V., 10, 154, 216, 223, 227, 250n3, 322, 328 Pax Americana, 311 Payne, Keith, 48–49 People’s Republic of China. See China permissive action links (PALS), 193 Perry, William J., 233, 236 Pipes, Richard, 49 Pluta, Anna, 109 plutonium, 111–112, 202n39

343

Poland, 273 Posen, Barry, 173 Powel Doctrine, 220n17, 220n22 Powell, Colin, 216, 219 precision-guided, 267, 311; forces, 317; munitions, 196, 313, 323; technologies, 314; warfare, 317; weapons, 305, 312 Press, Daryl, 285 preventive war, 4, 12, 19, 21, 32–33, 36, 45–46, 50–52, 54, 99, 101, 143, 152, 200, 267, 269, 278, 309, 311, 313, 321, 324, 325–327 precision air campaign, 317 precision strike, 313–314 preventive strikes. See preventive war Proliferation Security Initiative, 181n12 Qatar, 243–244 Quadrennial Defense Review (1997), 313 radiological weapons (materials), 5, 13, 16, 109, 114 Ramadan, Taha Yasin, 242 RAND, 312 Rashomon, 140, 149 Reagan, Ronald, 46; administration, 51, 57n46 recessed general deterrence, 9–10, 164, 259, 265 Record, Jeffrey, 312 reputation, 18–19, 66, 70–71, 76, 86, 105n9, 136, 139, 171, 285–288, 291, 293, 300n33, 309, 329 revolution in military affairs (RMA), 11, 23, 195–196, 267, 304–317, 323–324, 326–328 Rice, Condoleezza, 277, 297 Ritter, Scott, 238–239 Robertson, George, 240 rogues, 15, 141–155; actors, 1, 266; nonstate actors, 310; regimes, 15, 245; states, 1, 31, 36, 41–42, 44, 47, 133–134, 137, 141, 266, 268, 310, 314 Rousseau, Jean-Jacques, 220 Rumsfeld, Donald, 277 Russia, 1, 10–11, 13, 16, 101, 104, 122, 133, 143, 215, 231, 233, 235, 242–247, 270, 272–274, 298, 315; military doctrine, 276n22, 315, 327; nuclear arsenal, 263–265, 315; uranium suppliers, 113 Rwanda, 170, 214, 216–217, 219, 278

344

index

Saafaraniay industrial complex, 229 Sadat, Anwar, 67 sarin gas, 121 Saudi Arabia, 101, 227, 233, 240; royal family, 142 Schelling, Thomas, 37, 55n13, 59, 76, 108n70, 136, 147, 211, 283, 289–290 Schlieffen Plan, 213 Scud missile, 199 second-image, 43, 142, 145, 150 security dilemma, 11, 168, 293 Selassie, Haile, 213 September 11, 2001, terror attacks, 5, 31, 40, 42, 44, 45–46, 53, 72, 73, 99–100, 120, 128n42, n47, 139, 141–142, 155, 179, 191, 195, 215, 285, 311 Serbia, 153, 171, 216–217, 317; nationalists, 214 Shah, Prakash, 241 Sharif, Nawaz, 185 Shaw, Martin, 305 Shia, 99, 147, 154, 228, 247 Shultz, George, 51–52 Singapore, 288 Sino-Indian Crisis, 272 Six Day War, 222 six-party talks (negotiations), 4, 159, 164, 179 Smoke, Richard, 47, 54, 79, 160, 248, 279, 293 Snyder, Glenn, 38, 88, 138, 290, 293 Somalia, 216, 228, 285; Habr Gidr clansmen, 216; Task Force Ranger, 216 South Korea, 14, 145–147, 198, 277, 282; “Sunshine Policy,” 198 Soviet Union, 3, 25n20, 46, 51–52, 71, 93, 96, 124, 133, 135–136, 138–139, 141, 143, 146–147, 198, 201n19, 208, 214–216, 262, 275n7, 275n18, 310; former (FSU), 112; Intervention in Afghanistan, 87, 260; Nazi occupation, 215; nuclear test, 202n39 “standing start” attack, 282 Sri Lanka, 16 stability-instability paradox, 4, 138, 147, 183, 185, 187–189, 199, 201n11 Stalin, Joseph, 136, 140 Stein, Janice, 16, 47, 95, 325, 329 Strategic Defense Initiative (SDI), 46 Strategic Offensive Reductions Treaty, 263 Strauss, Leo, 142 submarine-launched ballistic missiles (SLBMs), 261, 264, 267, 270

Sudan, 26n40, 181n12, 287 suicide bombing, 75–77, 98, 199 Sunni, 99, 101, 154 surface-to-air missiles (SAMs), 228–230 Surprise Attack, 282 surprise attack, 99, 175, 281–282, 312 Syria, 13, 101–102, 112, 142, 199, 240, 278 Taiwan, 269, 282, 296, 316 Taiwan Strait, 281, 296 Taliban, 26n40, 120, 214–215 Tamil Tigers, 16, 76 Tellis, Ashley, 190 terrorist, 31, 35, 39, 50, 86, 95–96, 109–117, 123–125, 126n25, 127,n31, 127n33, 128n44, 126n45, 1238n55, 130,n64, 141, 143, 147, 153, 190, 195, 286, 325, 330; attacks, 41, 44, 53, 100–101, 109, 124, 141, 143, 186, 244, 325; cataclysmic, 5; conventional, 125; groups, 1, 15–16, 90, 102, 111, 115, 119, 123; negative goals, 115, 117–119, 122, 325, 329; networks, 85, 95–98, 100–101, 103, 322–324; nuclear, 110–113, 117–120, 122–124, 128n46, 198; objectives, 114; organizations, 1, 41, 45, 89–90, 102, 105n4, 115, 117, 297, 325; positive goals, 115, 117–119, 122, 324–325; provocateurs, 103; state, 18; strategy, 96; targets, 35, 124; theft of nuclear device, 111; threat, 41, 297; WMD use, 55n15 Tetlock, Philip, 63–65 The Day After, 51 Time magazine, 135 Tomahawk cruise missile, 229 transformation, 267, 304, 307, 313 Treaty on Conventional Armed Forces in Europe, 273 Trident missile submarines, 264 Turkey, 224, 231, 243, 295–298 2002 State of the Union Address, 144 unconventional deterrence, 210–219 unipolar, 7, 8, 20, 44, 263 286 United Arab Emirates, 231, 233 United Kingdom, 1, 10, 65, 190, 235 United Nations, 17, 26n32, 214, 222, 225–226; forces in Korea, 176; General Assembly, 231–232; inspections, 4; Monitoring, Verification and Inspection Com-

index mission (UNMOVIC), 242–244; Security Council, 9, 158, 232, 243, 259; Security Council Resolution, 688 (UNSCR 688), 224, 227; Security Council Resolution 949 (UNSCR 949), 234; Security Council Resolution 1154 (UNSCR 1154), 237; Security Council Resolution 1205 (UNSCR 1205), 242–244, 246; Special Commission (UNSCOM), 237–242, 244, 247 United States, 3–5, 8–11, 13–15, 19–20, 22, 32, 41, 44–46, 50, 65, 71–74, 88, 93, 96—87, 99–102, 104, 119–120, 124–125, 128n42, 133, 135–138, 140–150, 152–155, 157n27, 159, 165–167, 176–177, 186, 190–191, 197–198, 201n19, 204, 206–207, 213–218, 221n35, 222—224, 226–230. 232–233, 235–250, 252n53, 253n77, 260–261, 263–265, 267–271, 273–274, 275n13, 275n17, 277–278, 280–283, 285–288, 295–298, 303n73, 304, 306, 308, 310–311, 315–317, 322–323, 327; Code, 125n7 U.S. Air Force, 228, 267 U.S. Central Command, 227 U.S. Congress, 267, 281, 314; House of Representatives, 243–244, 297; Senate, 243 U.S. Department of Defense, 307 U.S. Department of Homeland Security, 128n42 U.S. Marines, 233, 238 U.S. State Department, 297 U.S. Strategic Command, 39 U.S. Threat Reduction Agency, 124 uranium, 113, 122, 202n39; highly enriched uranium (HEU), 111–114, 126n25, 127n31, 127n32, 128n42; smugglers, 113; yellowcake, 61

345

Vegetius, 34 Viet Cong, 206–207 Vietnam, 285, 288; North, 206–207, 211, 260; South, 206–207 Vietnam War, 87, 207, 211, 218, 260 Vickers, Michael G., 306 von Clausewitz, Carl, 43 von Hlatky, Stefanie, 11, 323 VX, 237 Wall Street Journal, 149 Waltz, Kenneth, 145 Walzer, Michael, 211 war on terror, 5, 72, 128n42, 287, 317 Washington Post, 51 weapons of mass destruction (WMD), 9, 12–15, 21, 25n22, 36, 39–41, 47, 55n15, 141–143, 151, 153, 160–161, 175, 223, 239, 242, 246, 249, 309, 313, 316, 321 Weber, Max, 6 Weinberger Doctrine, 220n17, 220n22 Wendt, Alexander, 90 Wohlstetter, Albert, 137, 142 Wolfowitz, Paul, 143 World Trade Organization, 164, 181 World War I, 215 World War II, 46–47, 70, 204–205, 207, 211, 213, 215, 217, 308, 328 World War III, 133, 136, 141 Yeltsin, Boris, 242 Yom Kippur War, 222 Yugoslavia (former), 278 Zakho, 116–227 Zimbabwe, 142 Zimmerman, Peter D., 109, 113–114, 126n25, 127n31